Chrysokamino I: The Metallurgy Workshop and its Territory (Hesperia Supplement 36)

HESPERIA Supplement

THE

CHRYSOKAMINO

METALLURGY AND

36

ITS

WORKSHOP TERRITORY

* Philip P. Betancourt American Classical

2006

School of Studies

at Athens

THE CHRYSOKAMINO METALLURGY WORKSHOP AND ITS TERRITORY

Hesperia

Supplements

6* J.H. Oliver, The SacredGerusia (1941) 7* G. R. Davidson andD. B. Thompson, Small Objectsfrom thePnyx: I (1943) 8* Commemorative Studies inHonor ofTheodoreLeslie Shear (1949) 9*

Fine, Horoi:

J. V. A.

in Mortgage,

Studies

Real

and Land

Security\

inAncient

Tenure

Athens (1951) 10*

L. Talcott,

B. Philippaki,

G.

R. Edwards,

Small

and V. R. Grace,

Objects from

the

P?^: II (1956) 11* J.R. McCredie, FortifiedMilitary Camps inAttica (1966) 12* D. J.Geagan, TheAthenian Constitution after Sulla (1967) 13

J. H.

Marcus

Oliver,

Aurelius:

of

Aspects

Civic

and Cultural

in the East

Policy

(1970)

14 J. S.Traill, The Political Organization ofAttica (1975) 15 S.V. Tracy, The Lettering of anAthenian Mason (1975) 16

M.

K.

A

Langdon,

Sanctuary

of

onMount

Zeus

(1976)

Hymettos

Shear Jr.,Rallias ofSphettos and theRevolt ofAthens in 268 b.c. (1978)

17 T.L

18* L. V. Watrous,

Lasithi:A

History

of

on a Highland

Settlement

Plain

in Crete

(1982)

19 Studies inAttic Epigraphy\History, and TopographyPresented toEugene Vanderpool

(1982) 20

Studies

in Athenian

Architecture,

and

Sculpture,

to Homer

Presented

Topography

A. Thompson (1982) 21 J. E. Coleman, Excavations at Pylos inElis (1986) 22

E. J.Walters,

Attic

Grave

That

Reliefs

Represent

Women

in the Dress

of

Isis

(1988)

23 C. Grandjouan, Hellenistic ReliefMolds from theAthenian Agora (1989) 24

J. S. Soles,

The

atMochlos

Cemeteries

Prepalatial

and Gournia

and

theHouse

Tombs

ofBronze Age Crete (1992) 25

S. I. Rotroff

and J. H.

Oakley,

Debris

from

a Public

Place

Dining

in the Athenian

Agora (1992) 26

I. S.Mark,

The Sanctuary

ofAthena

Nike

inAthens:

Architectural

Stages

and Chronol

?.gy (1993) 27

N. A. Winter, Terracottas

on Greek Architectural Conference December 12-15,1991 (1994)

International ed., Proceedings of the and Hellenistic Classical the Periods, of

28 D. A. Amyx and P. Lawrence, Studies inArchaic Corinthian VasePainting (1996) 29 R. S. Stroud, TheAthenian Grain-Tax Law of 374/3 b.c. (1998) 30 J.W. Shaw, A. Van deMoortel, P.M. Day, and V. Kilikoglou, A LMIA Ceramic Kiln 31

J. K.

in South-Central Papadopoulos,

and

Crete: Function Ceramicus

Redivivas:

Pottery The Early

Production Iron

Age

(2001) Potters'Field

in theArea

of theClassicalAthenian Agora (2003) 32

and K.

J.Wiseman

Zachos,

eds.,

Landscape

Archaeology

in Southern

Epirus,

Greece I (2003) 33 34

A.

: inHonor Sara A. Immerwahr (2004) Essays of L. A Historical and Economic and Davis, J. Bennet, J. Geography in the 18th Greece: The Southwestern Morea (2005) Century

P. Chapin, F. Zarinebaf, Ottoman

ed., XAP

35 Gloria S.Merker, The Greek Tile Works at Corinth: The Site and theFinds (2006)

of

* Out ofprint

THE CHRYSOKAMINO METALLURGY WORKSHOP AND ITS TERRITORY

Philip

WITH

P. Betancourt

BY

CONTRIBUTIONS

Eleni A. Armpis

Yannis

Bassiakos

Curt

Beck

Ruth

F. Beeston

Mihalis

Doniert R. Catapotis Brigit Crowell Evely William Farrand Susan C. Ferrence R. Noel Gale Cheryl Floyd B. Hafford William Donald C. Haggis Jones Glynis Byron Koukaras

Eleni Nodarou James D. Muhly George H. Myer Lada Onyshkevych Palatinus Natalia Joe Poulou-Papadimitriou Robert S. Powell S. Reese David Ann Schofield B. Elizabeth Shank Christine

Evi

The American 2006

Sikla

Zofia

M. Thompson

Stos

Edith

C. Stout

Tanya Yangaki

School of Classical

Studies at Athens

Charles

P. Swann

Copyright ? 2006 The American

All

School

of

at Athens

Studies

Classical

rights reserved.

To

contact:

order,

Out-of-print Hesperia supplements may be purchased from:

(in North America) The David Brown Book Company

Swets & Zeitlinger Service

www.davidbrownbookco.com

Royal Swets

Tel.

P.O. Box

800-791-9354

(outside North Oxbow Books

Backsets 810

2160 SZ Lisse America)

The Netherlands E-mail:

[email protected]

www.oxbowbooks.com

+44 (0) 1865-241-249

Tel.

Cover

illustration: The metallurgy

Library Betancourt,

of Congress

site during

Cataloging-in-Publication

cm.

Includes

Data

P., 1936

Philip

and its territory / Philip P. Betancourt.

The Chrysokamino Metallurgy Workshop p.

excavation

?

(Hesperia

bibliographical

supplement references

; 36) and

index.

ISBN-13: 978-0-87661-536-2 (alk. paper) ISBN-10: 0-87661-536-1 (alk. paper) 1. Chrysokamino

Site

3.Minoans?Greece?Chrysokamino 5. Kavousi Kavousi Region. (Princeton, DF221.C8B55 939r.i8oi?dc22

N.J.).

Supplement

2006 2006047906

(Greece)

Site. Region ;36.

Site.

2. Metallurgy?Greece?Chrysokamino 4. Excavations

(Greece)?Antiquities.

(Archaeology)?Greece? I.Title.

II. Series: Hesperia

CONTENTS

ix

List of Illustrations List of Tables Preface

xvii xix

and Acknowledgments

PART I:THE CHRYSOKAMINO TERRITORY Chapter 1 Introduction by Philip

P. Betancourt

3

Chapter 2 The

Natural by Philip

Environment P. Betancourt

andWilliam

R. Farrand

19

PART II:THE METALLURGY WORKSHOP Chapter 3 The

of

Excavation

the

Metallurgy

Workshop

by Philip P. Betancourt, James D. Muhly, Eleni A. Armpis, Robert S. Powell, Elizabeth B. Shank, Evi Sikla, and 47

Tanya Yangaki 4 Chapter The Apsidal by Philip

Structure P. Betancourt

55

P. Betancourt

67

Chapter 5 The

Pottery by Philip

Chapter 6 The

Stone

Tools

byDoniert Evely

99

Chapter 7 The

Furnace

Chimney

Fragments

by Philip P. Betancourt 109

VI

CONTENTS

Chapter 8 The

Pot

Bellows

by Philip

P. Betancourt

and James D. Muhly

125

Chapter 9 Miscellaneous

Ceramic

by Susan C. Ferrence

Artifacts

and Byron Koukaras

133

Chapter 10 Other

Metallurgical

by Philip

Materials

P. Betancourt

137

Chapter 11 Faunal

Remains

by David

S. Reese

149

Chapter 12 at

the

for

Evidence the

Early

of Threshing

Use

Minoan

Remains

Metallurgical

by Glynis Jones andAnn Schofield

Workshop

153

Chapter 13 in the

Chrysokamino

of

History

Early

Metallurgy

by JamesD. Muhly 155 Chapter 14 of

Discussion

the

and

Workshop

Reconstruction

of

by Philip

P. Betancourt

the

Practices

Smelting

179

PART III: THE SURFACE SURVEY Chapter 15 to

Introduction by Philip Chapter 16 Topography

the

Surface

of

Survey

193

P. Betancourt

the

Region

Chrysokamino

by Lada Onyshkevych

andWilliam

B. Hafford

197

Chapter 17 A

Summary

CHRYS

of

the

Habitation

O KAMIN O-ChOM

by Cheryl R. Floyd

Site

at

ATAS

205

Chapter 18 Edith

Hall's

Theriospelio by Philip

Excavations

in the

Cave P. Betancourt

and Cheryl

R. Floyd

Chapter 19 Chrysokamino

in Context:

Archaeological

Survey

byDonald C. Haggis

221

A

Regional

215

VII

CONTENTS

Chapter 20 The

Geographic

of

Boundaries

Chrysokamino

Farmstead P. Betancourt

by Philip

the

Territory

233

Chapter 21 Land

on

Use

the

Chrysokamino

P. Betancourt

by Philip

Farmstead

241

Chapter 22 Survey

Conclusions P. Betancourt

by Philip

257

APPENDIXES Appendix A and

Petrography of

Analyses

X-Ray

Slags

H. Myer

by George

Appendix B SEM/EDAX

Diffraction

and

Furnace

and Philip

Chimneys

P. Betancourt

281

Analysis

by Yannis

Bassiakos

293

Appendix C Lead

and

Isotope from

Slags

Chemical

of

Analyses

Chrysokamino

by Zofia

Stos and Noel Gale

299

Appendix D Arsenic

Content

of

A Study

Applying

PIXE

by Susan C. Ferrence

Copper

and Charles

Prills: P. Swann

321

Appendix E Slag

by Wavelength

Analysis

Dispersive

Spectrometry by Christine

M. Thompson

325

Appendix F of

Reconstruction Process

at

Analysis by Yannis

the

Copper

Chrysokamino

of

Ore

and

Bassiakos

Smelting

Slag

and Mihalis

the

on

Based Samples Catapotis

329

Appendix G Register by Philip

of Anthropogenic P. Betancourt,

Features

Lada Onyshkevych,

and

William B. Hafford 355 Appendix H The

Minoan by Philip

Pottery P. Betancourt

from 377

the

Survey

VIII

CONTENTS

Appendix I for

Evidence

Beekeeping

by Susan C. Ferrence

and Elizabeth

B. Shank

Ottoman

Pottery

391

Appendix] The

Byzantine

the

Survey

to

by Natalia

from

393

Poulou-Papadimitriou

Appendix K of

Excavation

The

AF

Terrace

Cave

AF

9 and

22b and Philip

by Brigit Crowell

P. Betancourt

399

Appendix L and

Soils at

from

Sediments

Natural

Deposits

Chrysokamino

by Eleni Nodarou 403 AppendixM Residue

Organic from

Analysis

of

Joe Palatinus,

Curt Beck,

Pottery

Sherds

Chrysokamino by Ruth F. Beeston,

and

413

Edith C. Stout Appendix N Petrographic

Analysis

Neolithic Metallurgy

Index

of Two the

Final

Chrysokamino

Location

by Eleni Nodarou References

from

Sherds

429

433

457

ILLUSTRATIONS

of eastern Crete

1.1.

Map

1.2.

Three-dimensional

topographic plan of the Chrysokamino and habitation sites and the surrounding territory

metallurgy 1.3.

Aerial

1.4.

Map

1.5.

Toponyms

1.6.

Olive

of the Chrysokamino

7

territory

of the area near Chrysokamino

trees growing site the habitation

in Chordakia,

1.8.

Chylopittes,

a travertine cave

Theriospelio

9

looking northeast

activity at Chrysokamino

formation

11

at the coast near the

11

as seen from Kavousi, with

The Kambos

from

10

site of the metallurgical

The

5

the village of

photograph showing Chrysokamino, Kavousi, and the surrounding region 6

1.7.

1.9.

4

the coastal hills

behind it 12 1.10.

Lakkos Ambeliou, site 13 habitation

1.11.

The

1.12.

A fragment

in the Kambos,

burial cave called Theriospelio

Detailed

bedrock map

2.2.

Bedrock

geology

2.3.

Schematic

2.4.

Plattenkalk

2.6.

Metacarbonate

by Angelo Mosso

of the Chrysokamino

of eastern Crete

section

south from the

13

of industrial pottery published

2.1.

2.5.

looking

through

limestone with formation

15

area 23

24

the nappe pile of eastern Crete calcite veins

25

near the metallurgy

site

column through the Phyllite nappe Stratigraphie area 28 east of the Chrysokamino-Kavousi

26 somewhat

25

x

illustrations

2.7.

Cliffs

north of the metallurgy 30

toward the

site, looking

island of Pseira 2.8.

Caves

and cliffs south of Chrysokamino

2.9.

clay and the entrance 31 Agriomandra

31

to the ravine leading

Red

to

2.10.

a small harbor at the foot of the ravine that Agriomandra, connects the coast with the Kambos 31

2.11.

Soft, powdery phyllite 32 workshop

2.12.

Kink fold in phyllite 33

2.13.

Paleoravine

2.14.

Detail

uphill

in a trench at the metallurgy

exposed

from the metallurgy

site

35

of the paleoravine showing the contact between the reddish soil in the ravine and the underlying pale-colored

phyllite 36 2.15.

Colluvium

2.16.

Terrace

consisting

The

3.2.

The metallurgy

surface of the metallurgy site during

at the metallurgy

3.3.

Sieving

3.4.

James Muhly

building

and other rocks

22, located near the habitation

system AF

3.1.

of reworked phyllite

cleaning

site before excavation

site

excavation

36 40

48

48 49

workshop

the upper part of the apsidal

50

3.5.

Plan of the grid-squares

4.1.

Aerial

4.2.

Plan of the lower floor

4.3.

The

4.4.

Level

at the metallurgy

excavated

view of the apsidal building

56

56 57

apsidal building

4 in grid-square N lower floor at the eastern

lower floor, showing

Apsidal building, the doorway 58

4.6.

Plan of the apsidal buildings

the location

4.7.

Plan of the apsidal buildings

5.1.

Location

lower floor

dark-burnished

the slag pile 69 cup 75

72

5.2.

Sherds of conical

5.3.

Pottery

from the slag pile (1-23)

5.4.

Pottery

from the slag pile (24-43)

57 of

floor and cross section

middle

59

of Neolithic

the

19, looking north, showing side of the apsidal building

4.5.

of the building

site

74 80

60

sherds within

51

XI

illustrations

82

from the slag pile (44-59)

5.5.

Pottery

5.6.

Pottery from the slag pile (60-72) structure (73-82) 85

5.7.

Pottery

from the apsidal structure

5.8.

Pottery

from the apsidal

5.9.

Pottery

from the apsidal structure

6.1.

Stone

6.2.

Stone

6.3.

Cutting

in bedrock

6.4.

Stone

tools (122-124

and the apsidal

90

(83-90)

structure

92

(91-101)

95

(102-121)

102

tools (122-127)

obsidian tools (128-133), 104 in bedrock cutting

tools (142-144),

at the metallurgy

workshop

and 127-129)

180

7.1.

Front

7.2.

no. X 213, Chimney preserving fragment of an olive leaf 110

7.3.

Alternative

fragment

reconstructions

105

from the metallurgy 106

site and back of chimney

and

110 the impression

of the industrial

ceramics

111

fragments

114

7.4.

Profile

drawings

of chimney

fragments

(145-162)

7.5.

Profile

drawings

of chimney

fragments

(163-178)

118

7.6.

Profile

drawings

of chimney

fragments

(179-188)

122

8.1.

of workmen Drawing Rekh-mi-re 126

8.2.

Reconstruction

8.3.

Bellows

8.4.

Profile

drawings

of pot bellows

(189-194)

129

8.5.

Profile

drawings

of pot bellows

(195-203)

131

9.1.

of the pot bellows

fragment

Miscellaneous

using bellows,

189

artifacts

127

135

(204-209)

of slag from unit M

Lumps

10.2.

Slag resting

14.1.

East Cretan White-on-Dark

on bedrock,

sites

in use

of

128

10.1.

and nearby

from the Tomb

22-3

138

looking

south

Ware

139

from Chrysokamino

181 of a bowl

14.2.

Reconstruction

16.1.

Map

16.2.

Map of anthropogenic farmstead 199

16.3.

Topographic

furnace from Chrysokamino

of the greater region around Chrysokamino

map

features

showing

in the Chrysokamino

terraced areas

200

198

184

XII

illustrations

for the Chrysokamino

Map

16.5.

farmstead, with Map of the Chrysokamino 202 of all points measured electronically

16.6.

map Topographic farmstead 203

17.1. 17.2.

of grid points

Western

portion

201

farmstead

16.4.

indications

of the territory of the Chrysokamino

of the plan of the habitation

206

the LM

site showing

State plan of the habitation 207 architecture

site

III

17.3. The LM I bronze dagger fromRoom A (X 830) 17.4.

of EM

Examples

1321, EM IIB Vasiliki Ware

17.5.

ceramics: X 541 and X

of MM

Examples

I pyxis lid closed

ceramics: X

fragment; X 1179, EM 208 vessel base

208

1448, cups; X 1497, base of a sherd of a large, closed vessel

from carinated

fragments tumbler; X 932, body 17.6.

of LM IB ceramics: X 1480, base of a bell cup; Examples 208 X 808, cup rim; X 836, neck of a jug

17.7.

Stone bowl

17.8.

LM

IIIA hearth

of LM

Example

I stratum

(X 818) from LM above a LM

209

(Wall C), but below

Iwall

210

aLMIIIAwall(Wall9) 17.9.

IIIA wall

at the habitation

construction

210

site entrance

into the LM

17.10.

East

17.11.

Examples

of LM

17.12.

Sealstone

(X 208)

17.13.

Examples

of ground

17.14.

The

18.1. 19.1.

208

Map

from Room

drawings

210

212

IIIA ceramics

2 212

stone tools from the habitation in Room

ritual deposit

Profile

IIIA complex

213

11 213 cave

of sherds from the Theriospelio

of the Kavousi

site

219

survey area in the late Prepalatial

period 222 19.2.

Map

of the Kavousi

survey area in the Protopalatial

period 222 19.3.

Map

19.4.

The

of the Kavousi

area in the late Prepalatial

phyllite terraces of Chordakia seen from the west (the Kambos),

19.5.

An

olive crushing

19.6.

A millstone

19.7.

Theriospelio

basin

(trapetus)

and the Kavousi

223

plain

224 at Kephalolimnos

(orbis) at Kephalolimnos cave from the west

period

224 225

224

xiii

illustrations

survey area in the Neopalatial

19.8.

Map

of the Kavousi

19.9.

Map

of the north Kambos

Neopalatial

and Tholos

period

Ravine

20.2.

Boundaries

21.1.

Land

of Type

1, used for horticulture

21.2.

Land

of Type

2, used for pasture

21.3.

Land

of Type

3, used for dry farming

21.4.

Steep

22.1.

22.2.

and Agriomandra

areas unsuitable

Map of Clusters 275 features Pressures

harbor

for the farmstead

rocky

1-3, with

influencing

land for agriculture

Slag sampleCHR 97 282

A.2.

Slag sampleCHR 87 283

237

in LM

III 239 243

249 251

for farming

closely

99 showing

pyroxene

sample CHR

97 291

A.4.

Slag sample CHR 283 cuprite

A.5.

Furnace

C.l.

The Anorthite-Wollastonite-FeO

C.2.

The

C.3.

Lead

C.4.

Lead

C.5.

Lead

chimney

253

related anthropogenic

Slag sample CHR 99 showing pyroxene and cuprite 283 magnetite,

Anorthite-Si02-FeO

with

in the

Bay

greater or lesser use of marginal 277

A.l.

A.3.

230

231

20.1.

or

period

crystals,

crystals and

ternary phase diagram

309

ternary phase diagram 309

of ores from Crete isotope compositions those for slags from Chrysokamino 312

compared

of slags from Chrysokamino isotope compositions and ores from Lavrion and the Cyclades 315

some isotope analyses of Prepalatial alloy artifacts Protopalatial copper-based

D.I.

PIXE

D.2.

PIXE measurements

and 317

analyses of copper prills arranged by increasing arsenic content 322 showing

the variation within

individualprills 323 E.l.

Backscattered

E.2. White

electron

image of CHR

55 326

light photomicrograph of CHR 55 326

E.3.

Backscattered

electron

E.4.

Photomicrograph of CHR 50 327

E.5.

Photomicrograph of CHR 8 328

image showing

core of CHR

48

326

XIV

illustrations

E.6.

Backscattered

F.l.

Copper residual

F.2.

Iron ore from Chrysokamino of calcite 333

electron

74 328

image of CHR

ore (ORE 1) from Chrysokamino sulfides 333

showing

rhomboid

crystals

of ores and slag from Chrysokamino 335 ternary system CaO-FeO-Si02

in the

F.3.

The

F.4.

The

showing

position

position of slags from Chrysokamino smelting sites from the southern Aegean

and other EBA in the ternary

335

system CaO-FeO-Si02

(+A1203)

F.5.

Sample A04: magnetite slag "layers" 340

"bands" forming

F.6.

Sample B18: Copper prills showing massive of a second arsenic-rich phase precipitation

F.7.

The

position

of matte

Schematic smelting

of the proposed representation 352 process at Chrysokamino

G.I.

The metallurgical

G.2.

Location

Segment

copper

and its nearby

workshop

features AF

of anthropogenic

G.3. Well(AF2) G.4.

in slags from

found

345

Chrysokamino F.9.

inclusions

microstructures

Typical

356

territory

1,AF

2, and AF

3

358

358 of awall,

a

possibly

boundary marker

G.5.

Destroyed threshing floor (AF 4) 359

G.6.

Collapsed field house (AF 5) 360

G.7.

One

G.8.

Collapsed

field house

(AF 7), looking west

G.9.

Collapsed

field house

(AF 7), looking

Downhill

from the habitation

G.10.

342

(Cu3As)

present in slags from in the ternary system Cu-Fe-S 342

Chrysokamino F.8.

three

of the terrace walls

(AF 3)

359

(AF 6) 361

south

362 362

site (AF 29) to scatter

of sherds (AF 8) and oval enclosure (AF 32) 362 G.ll.

Standing field house (AF 15)

G.12.

Sections

G.13.

View

of wall

of AF

365

(AF 21), probably

boundary wall

367

21 from above

G.14.

Section ofwall (AF 21) 368

G.15.

Section

G.16.

General

of boundary wall

a

(AF 27)

view of the habitation

370

site (AF 29)

371

367

XV

ILLUSTRATIONS

the east side

site (AF 29) showing

G. 17.

The

habitation

G.18.

The main

G.19.

Opening of awell (AF 39)

G.20.

The modern

entrance

to the habitation

road (AF 40)

site (AF 29)

375

Minoan

pottery

from the surface, AF

17 to AF

H.2.

Minoan

pottery

from the surface, AF

32 383

H.3.

Minoan

pottery

from the surface, AF

32 384

H.4.

Minoan

pottery

from the surface, AF

32, AF

Location

1.2.

The

J.l.

Byzantine

of the beehive

beehive

fragment to Ottoman

371

374

H. 1.

LI.

371

fragment

32

379

33 386

392

392 pottery

from the surface of the

territory (J-l to J-5) 395 J.2.

Ottoman

pottery

from the surface of the territory

396

(J-6toJ-15) 400

Kl.

CaveAF9

K.2.

The

K.3.

Terrace

K.4.

Terrace AF

L.l.

Analytical

L.2.

The

L.3.

Analytical

data for the road exposure

409

L.4.

Locations

of the two sets of analyzed

soil samples

cave s entrance group AF

400 22 401

22b 401 data for agricultural

sediment

deposit

terrace AF

at Lakkos Ambeliou

22b 407 408

409

TABLES

variations

at Hagios

2.1.

Seasonal

5.1.

Final Neolithic

5.2.

Early Minoan

I-IIA

5.3.

Early Minoan

IIB sherd from the slag pile

5.4.

Early Minoan

II?III

5.5.

Early Minoan

Ill-Middle

20

Crete

Nikolaos,

sherds from the slag pile

76

sherd from the slag pile

76

77

sherd from the slag pile 77 IA sherds from the

Minoan

slagpile 87 sherd from the apsidal structure

5.6.

Final Neolithic

5.7.

Early Minoan

5.8.

Ill-Middle Early Minoan structure 97 apsidal

5.9.

Comparison

Plant

SEM/EDAX increasing

B.2.

SEM/EDAX

C.l.

Chemical

C.2.

Results

C.3.

97

from the

fragments

analyses of copper prills arranged by arsenic content 293 analyses of metallurgical analyses of Chrysokamino

of applying the Bachmann analyzed slags from Chrysokamino

Lead

88

IA sherds from the

Minoan

in chimney 154 workshop

impressions

metallurgical B.l.

II sherd from the apsidal structure

of Early Minoan Ill-Middle Minoan IA structure from the and the slag pile shapes apsidal

pottery 12.1.

87

remains

294-295 306-307

slag computations 308

isotope analyses of copper ores and galenas

Crete

to the

from 311

C.4.

Lead

C.5.

ores and copper Cycladic slags isotopically with slags from Chrysokamino 314

isotope

ratios of pieces of Chrysokamino

slag 313

consistent

XVIII

C.6.

tables

artifacts isotopically Early Minoan copper-based consistent with the slags from Chrysokamino

E.l.

CHR

55, Points

1-5

E.2.

CHR

55, Points

6, 7

E.3.

CHR

48, Points

1-4

326

E.4.

CHR

50, Points

1,2

327

E.5. CHR 50, Point 3

325 326

327

E.6.

CHR

8, Points

1-3

327

E.7.

CHR

8, Points

4,5

328

E.8.

CHR

74, Points

328

1,2

E.9. CHR 74, Point 3

328 of copper ore and

F.l.

EDS

F.2.

EDS

F.3.

EDS

analysis of main components glass phases in slag samples 339

of silicate and

F.4.

Results

and calculated

F.5.

EDS

analysis of copper prills in slag samples

F.6.

EDS

analysis of matte

L.l.

Analytical

analysis (average composition) iron ore samples 334 analysis of slag pellets

336-337

ofM?ssbauer spectroscopy Fe+2/Fe+3 ratios of slag samples 340

M.l.

Descriptions

M.2.

GC-MS

inclusions

of sherds analyzed parameters

341

in slag samples

data from natural deposits

instrument

318-319

at Chrysokamino

for organic 417

343

residues 415

404

PREFACE AND ACKNOWLEDGMENTS

some at Chrysokamino metallurgical interesting workshop posed was located in a rural part of eastern Crete, away from site The challenges. sites. It had been first visited by archaeologists the large archaeological

The

nature was appar early in the 20th century, and although itsmetallurgical ent to almost everyone from the its the nature of the date and beginning, activities undertaken there were controversial during most of the 20th cen as ancient, and other times asMedieval or tury. Sometimes regarded Early Modern, We

lay unexcavated until 1995. Chrysokamino as a the approached workshop multidisciplinary

research project, were fields. large complement Specialists invited to contribute information of several types to help explain what we knew would be a complex series of problems. In addition to excavation and a we decided to use an intensive surface large program of scientific analysis, to the small the context of its immediate survey place workshop within a context was needed in order to address territory. We felt that questions that could not be answered by examining the metallurgical site in isolation. it possible to support a metallurgical Was installation at Chrysokamino on based local ores, personnel, fuel needs, and other resources? Would with

a

of scientists

in different

or would such a workshop have operated independently, it have been an a nearby farmhouse, like herding or agriculture? How of large appendage was the territory associated with this small rural site? How would this ter s and what was Chrysokamino ritory have been organized and managed, was a relationship to similar sites and to the larger region of which it part? How would such a rural establishment have changed through time, and how did its situation compare with the Classical, Roman, and later systems used for agriculture and craft production? We knew from the beginning that these would not be easy questions to address. The results of this some progress in make study resolving them, but they do not provide final answers. Yet even if the evidence does not at present, it is to be allow complete explanations hoped that by posing the questions and trying to address them, we will take an important series of steps toward understanding many of the fundamental issues about this life. Small farms and workshops situated away from aspect of Minoan events theMinoan and the palaces dynamic taking place there must have

XX

PREFACE

AND

ACKNOWLEDGMENTS

some of the basic culture.We cannot building blocks ofMinoan at the Cretan Bronze Age without small workshops looking and farms as well as the larger architectural complexes.

constituted

understand

uses study approaches that have not been applied in the study of Crete before, and these approaches have led to many new conclu sions. Such results have been possible only because of the hard work of The

Minoan

a very fine team of students

and colleagues who have assembled consid erable data on the Bronze Age workshop and its surrounding territory. Because of this team's efforts, we know much more about the beginnings ofMinoan than we did before the project began. In addition, metallurgy we can

to define

begin

can

system of agricultural land use, and we and fields using the archaeological data

theMinoan

between

distinguish collected from

gardens survey. This information provides valuable insights about the small rural establishments that were an important part of Minoan society.

Acknowledgments This

project

entailed

three field

in 1995 with

Work

seasons

followed

by

a

a

period of study. and an electronic

began investigation geological were conducted instrument survey and mapping project. Excavations in 1996 and 1997, under the direction of Philip P. Betancourt, with the

assistance

of co-directors

(assistant director, Floyd conducted at the INSTAP

James D. Muhly 1996; co-director

(1996-1997)

and Cheryl R. seasons were

1997). Study for East Crete in 1998 and 1999, and when the project was completed except for preparation of manuscripts was The with laboratory analysis. sponsored by Temple University project the collaboration of the University of Pennsylvania Museum of Archaeology and Anthropology.

It was

School

Studies

Ministry Aegean Museum

Study Center

the auspices of the American a permit issued by the Greek was of Culture. Financial support provided by the Institute for the University of Pennsylvania Prehistory, Temple University, Dennis and Janice Versiegen, of Archaeology and Anthropology,

of Classical

conducted

at Athens,

under with

and other donors. are

to the citizens

of the village of Kavousi, who assisted with the project in many ways, as well as to the large staff of in the project. colleagues and students who participated The

directors

grateful

Staff members

included Mary A. Betancourt, apotheki supervisor and sur Lada supervisor for the electronic Onyshkevych, registrar (1995-1999); team (1995-1998); William B. Hafford, computer spe vey and mapping Carola R. Farrand, geologist cialist (1995-1998); William (1995-1996); excavation consultant (1996); (1995); Polymnia Muhly, Steams, geologist archaeobotanist Mark Hudson, archaeobotanist (1996); Ann Schofield, Susan C. Ferrence, artist (1997); Lyla Pinch Brock, artist (1996-1997); E. artist Ann Gleit, artist (1996); Laura A. (1996); Stephanie (1998); Foster, Ian Versiegen, artist (1998); Sari K. Uricheck, Labriola, artist (1996-1997); conservator Baxter Shank, assistant cataloger (1996), (1996); Elizabeth assistant pottery

specialist

(1998); Gerardo

I.Medrano,

assistant

cataloger

AND

PREFACE

XXI

ACKNOWLEDGMENTS

assistant cataloger (1998); Gayla Weng, assistant (1997); Jane D. Hickman, to the director (1998); Doniert Evely, stone tools specialist (1997); David S. soils specialist (1998-1999); Reese, faunal analyst (1997); Eleni Nodarou, and Natalia Poulou-Papadimitriou, pottery specialist (1996). Byzantine Robert Huber assisted with layout (1998). team members assisting Lada survey and mapping Onysh B. Hafford included Jaime J. Alvarez (1998), Leigh kevych andWilliam Ann Bingham P. Brennan (1998), Katherine May (1995), (1995),Terrence Robert S. Powell (1996), Stephanie Takaragawa and Jonathan (1996,1998), Electronic

(1995). The team used aTopcon GTS 303 Electronic Total Station 2000 (486) laptop computer for data collec interfaced with a Gateway tion. In 1995, the software used was PC AMP (Archaeological Mapping

Wallis

by the Center

Field Training, Pima In 1996, Easy Survey Plus was County Community College, Tucson, AZ. 13 and 14 were employed for additional pro used. Surfer 6 and AutoCAD run To the II Solar Panel (Keep it Simple computer, aMercury cessing. 32 S. Suite 330, Helena, Montana, 59601) provided electric Systems, Ewing, Program),

provided

for Archaeological

power generated by solar energy. Trench supervisors included Eleni A. Armpis (1997), Barbara J.Hayden (1996), Katherine May

(1996), Brigit Crowell (1996), Eleni Nodarou

(1996-1997), Robert S.Powell (1996),Elizabeth Baxter Shank (1997),Evi Sikla (1997), Suzanne

Stichman

(1997), Stephanie Takaragawa

(1997), and

Tanya Yangaki (1996). Staff members during

the 1999 study season, in addition to James D. and the writer, included Susan C. Ferrence, Muhly, Mary A. Betancourt, D. BethAnn Judas, Elizabeth B. Shank, Gim?nez, Hickman, Jane Alejandra and Gayla Weng. Support and technical

services were provided by the following person nel at the INSTAP Study Center for East Crete: Thomas M. Brogan, direc tor (1997-1999); chief conservator (1997-1999); Stephania Chlouveraki, Ann N. Brysbaert, conservator (1997); Katherine May, chief photographer (1997-1999); Westley assistant J. Huffman,

Bernard, assistant photographer to the director (1997-1999).

(1997); and Eleanor

was undertaken Photography by Katherine May, Philip P. Betancourt, William R. took Farrand, and Lillie Floyd. Donald Haggis Gayla Weng, was the photographs for Chapter 19. Aerial photography for the site done on a kite (1997). The Soviet a camera mounted Driessen, Jan using by for the region. Elizabeth B. Sputnik program provided aerial photography Shank completed the plan of the metallurgy site. are expressed to the Directors Thanks of the American School of Classical Studies atAthens, the lateWilliam D. E. Coulson (1987-1997) and James D. Muhly and to the Director of the 24th Epho (1997-2002), rate for East Crete,

the late Nikos Papadakis.The representatives from the were excavation Charoulis (1996) and Gior Ephorate during Giorgios was Katsalis in with Crete furnished by (1996-1997). gios Help logistics andMaria Tsagarakis. The team is grateful to the proprietors of the at Kavousi, Aristidis Chalkiadakis, Maria Chalkiadaki, and their son Ioannis Chalkiadakis.

Manolis Tholos

Beach Hotel

PREFACE

XXII

AND

ACKNOWLEDGMENTS

cave are grateful to authors of the chapter on the Theriospelio the following persons for assistance with various aspects of their research: The

and Emily Pamela Russell, Cornelius Douglas Haller, David Romano, Watson. and Vermeule, Wendy are extended to our many friends in the village of Ka Special thanks vousi, and especially to the former mayor, Dimitris Kophinakis, who offered Kaso invaluable assistance at the beginning of the project, and toManolis as in and foreman who excavation acted many ways. takis, provided help Betancourt Philip R

2004

PART I:THE CHRYSOKAMINO TERRITORY

i

CHAPTER

Introduction byPhilip P. Betancourt

a The Chrysokamino territory is located along the south side of hill named on the eastern side of the Gulf ofMirabello in northeast Crete Chomatas, was first settled (Figs. 1.1,1.2). The land adjoining the Gulf ofMirabello a substantial in the Neolithic and it had during the period,1 population were in Bronze Age. Its people culture that developed part of theMinoan and gradually became an influential force within the Aegean. Those lived near Chrysokamino, however, did not participate much in the and cultural society. Their small dynamic political expansion of Minoan to contributions territory remained essentially rural, making only modest

Crete who

the advance of Cretan

society. an site in the Chrysokamino territory is largest archaeological isolated building constructed on a foundation of megalithic dolomite blocks. can be as a farmstead as defined The establishment regarded by Pettegrew, The

name to any "isolated, rural unit of habitation and center applies the more of agricultural operations."2 The Chrysokamino territory contains an area than is especially interesting be isolated farmstead, however. The cause of the presence of a located a little over half metallurgical workshop a kilometer from the farmstead, on a cliff the sea (Fig. 1.2). overlooking in this volume studied the metallurgical The project published location

who

and the excavations at the workshop and its regional context, undertaking an intensive surface survey of the with farmstead, nearby along territory sites. around these two archaeological informal visits were made

area of by the author to the Chryso kamino in the 1980s and the early 1990s. The information acquired from these visits, in conjunction with a study of the relevant literature, suggested that the location offered an opportunity for the discovery of much new evidence about the area's history. Archaeologists knew of an ancient met Several

1. Betancourt 2. 3.

1901, Boyd 1908, p. 33. 5. Faure

1999. p. 189. p. 156; Hawes

2001,

Pettegrew

4.Mosso

even before Harriet at Chrysokamino allurgy workshop Boyd visited the never in site but the had been excavated. Surface examination 1900,3 region of the location had yielded ambiguous conclusions, and the workshop had to modern.5 Only new work been assigned dates ranging from Minoan4 et al.

1908; 1910, pp. 289-292. 1966,

pp. 47-48.

excavation could properly date the use of the metallurgy site and the details of its history. The metallurgy workshop and the farmstead were not the only archaeo

involving elucidate

logical features

in the immediate

area. An

ancient burial cave, stone terrace

CHAPTER

I

lerapetra

0 10 km

other land walls, eroded threshing floors, old field walls, abandoned wells, and was abun for Evidence were observed also farming nearby. scape features ter seen immediate in the dant, and many agricultural terraces could be to an at Chrysokamino would present opportunity ritory. An excavation its local context. A recent intensive survey of the study the workshop within as a part of the Kavousi Archaeo been whole region had already completed be a great advantage to a new regional survey would logical Project.6The itwould allow conclusions on the interactions between the project, because and the larger, regional picture. tiny territory of Chrysokamino

Figure

1.1. Map

of eastern

Crete

LOCATION OF THE CHRYSOKAMINO TERRITORY The

has a rugged and mountainous history.7 The Isthmus of lerapetra,

island of Crete

plex geological chain running

a com landscape with a break in the mountain

across the island from east to west, affords the only easy coast to the south. It was probably traversed by passage from the north in all periods of Crete's human history. The Gulf of an important road of the Isthmus, is a large bay with several harbors at north the Mirabello,

6.Mook andHaggis 1990;Haggis 1992,1993b, 1995,1996a, 1996b, 2000, 2005; Haggis andMook 1993. 7. Creutzburg

et al. 1977.

For

a use

ful bibliography of the island's geology, see Fassoulas

20?1,

pp. 93-98.

INTRODUCTION

5

Habitation Location

>

S a** Metallurgy Location

Figure

1.2. Three-dimensional

topo

graphie plan of the Chrysokamino and

metallurgy the surrounding from

the

sites

habitation territory

and

as seen

sea

across

shelter for seagoing craft. Traffic and cargo transported the Isthmus could have connected with several different seaports.

that offer

An

east-west

coastal

road

meeting

the

north-south

at

passage

the

of the Isthmus, somewhere near modern Pacheia Ammos, would seem to be a logical result of the topography, but the actual situation is

north more

complex.8

Before

at Pacheia Ammos

the

construction

harbor

of modern

and it was

was

works,

the

port

safe only in good of any settlement

extremely exposed, weather.9 The location was not suitable for the growth interests. In addition, the coastal road could not with year-round maritime at Pacheia proceed east for very far beyond the beach and small harbor Ammos. The town at the north of the Isthmus connected easily with places farther west, but the land rises so quickly and becomes somuch rougher to the east, that beyond Kavousi itwas only a steep path before the modern on road system was built. Even in the 19th century, travelers commented the difficulty of the road leading east from this part of Crete.10 These geo to the fact that the main Minoan graphic limitations probably contributed town developed farther west, at Gournia, of Gournia had smaller settlements. The

and the region immediately farmstead at Chrysokamino

one of the smallest of these sites. tiny habitation from is Pacheia Ammos visible Chrysokamino

8. Betancourt 9. Seager 10. Spratt

1865,

more

p. 6. pp.

was

it sits on a

a hill called headland east of the small harbor, on the western slope of it is not far from Pacheia Ammos Chomatas Bay, (Fig. 1.2). Although to the itwould not have been easy to travel by land from Chrysokamino

2004a.

1916,

because

east

157-158.

accessible western

ties reaching

sites.Wheeled

the location before

vehicles would

the construction

have had difficul

of the modern

road, and

6

CHAPTER

I

??tm'f

Figure 1.3. Aerial photograph show ing Chrysokamino, the village of and

the

at 6:52

a.m.

Kavousi, taken

surrounding region, on June 5,1988,

with a 3.00 x 3.38 km field of view. ?

1999 Aerial

Sovlnform

Images,

Sputnik

Inc., and 1999

INTRODUCTION

Figure 1.4.Map kamino

7

of the Chryso

territory

11. Seager 1916;Harrison 1993, pp. 190-191; Spanakis 1993, p. 618. 12.Haggis 1996a. 13. Betancourt

2004a.

ancient traffic probably moved more sea than by by land. The closest modern runs the of located highway Kavousi, through village directly inland from the territory of Chrysokamino (Fig. 1.3). The part of the Cretan coast where the Chrysokamino territory is located is not fertile not it and is farmland, particularly adjacent to any harbor. The is in illustrated the schematic topography drawing (Fig. 1.2) and in the local map (Fig. 1.4). The farmland consists of a series of low, rounded hills. The spaces between them are sometimes extremely eroded, and small ravines and gullies are common. The soil is partly terra rossa and partly eroded and decomposed phyllite. Steep hillsides and several cliffs demarcate the boundary between land and water. The architecture of the main habitation site, situated on a rocky outcrop of pale gray dolomite, a commands but it is set back more good view of the Gulf of Mirabello, than half a kilometers walk from the sea. The lies metallurgy workshop on a headland the overlooking Aegean. The hilly nature of the terrain and the absence of any adjacent boat make the location of less desirable for seafaring landing Chrysokamino than the two good beaches located at Pacheia Ammos to the southwest11 to the northeast.12 Like other andTholos nearby parts of Crete, the rough not a does allow coastal road,13 and the east-west highway is topography as itwas in located well inland today, antiquity.

8

I

CHAPTER

both Pacheia Ammos and Tholos had towns by the Late Although Bronze Age, no largeMinoan sites have ever been found along the rough, 8 km long coastal strip that lies between them (see Fig. 1.1). Regional survey has shown that while this region was already inhabited by the Final Neolithic

in period, before 3000 b.c.,14 it supported only tiny settlements the Bronze Age.15 In this publication we discuss the three main sites that occupy the Chrysokamino territory: a farmstead last inhabited in LM IIIB, site ca. 600 m northwest of it, and a burial cave named The metallurgy found northeast of the metallurgy location (Fig. 1.4). Among riospelio these three sites, only the habitation location has evidence for long-term

a

occupation.

MODERN TOPONYMS Several locations near Chrysokamino have modern names (Fig. 1.5). Some of the labels are used for specific topographical features, while others are areas. In several cases, a name is used both for an to applied geographic area and for the feature that area its name. Interviews with many gives the local residents in the village of Kavousi provided substantial information on nomenclature.

local people are well aware of the variations exist in regard to the names of several of the locations. They explain the alternatives stem from different oral traditions. The

that that

Agriomandra at the foot of a ravine that leads downhill from Lakkos is called Agriomandra (Aypi?jLiavxpa). The name is derived or from agrios (?Vypioc), wild savage, and mandra (uxivTpa or uav?poc), a pen for animals. The word mandra is sometimes used in Crete for unfenced

A

small harbor

Ambeliou

locations where animals are gathered (such a location is found on Pseira

to be loaded on board together ships The beach at Agriomandra Island). has rock outcrops on both sides, and animals were herded together here was used until after the middle of the 20th century for easily. The harbor animals and other cargo. An alternative, Ayiomandra (from the shipping church of Hagios Ioannis in the ravine uphill from the location), is also used occasionally.

Chalepa (XocA?tcoc) is the name of a rugged hill south of the ravine that Chalepa The name is derived from chalepos leads to the beach at Agriomandra. hill is so rocky and eroded that it can difficult. The (xaXen?q), meaning not be used for any agriculture, sheep and goats. 14.Haggis 1992, p. 269; 1993a; 1996c; Betancourt 1999;Hayden 2003a, 2003b.

and its only modern

15.Mook 1992,1993b,

use is as pasture

for

andHaggis 1990; Haggis 1995,1996b, 2005.

INTRODUCTION

9

Pyrgos Chrysokaminou

Gulf of Mirabello

Chomatas

Chrysokamino JQ Chordakia

Katsoprinos

Kambos

Kephalolimnos Agriomandra

Lakkos

Figure near

1.5.

Toponyms

Ambeliou

200 m

of the area

Chrysokamino

Chomatas called a (Xcojucctocc) is the name of a coastal hill (sometimes mountain) with two peaks, both of which are northeast of the Chrysoka mino territory. The southern part of the hill extends to the east of slightly the location marked Chordakia on area 1.5. The of the excavations, Figure where the settlement site and the are located, is on the metallurgy workshop western slope of the southern part of this hill. The name is applied locally to the hills crest, a bare area with little vegetation, because the lower slopes have other names. Chomatas is derived from ch?ma (%cbuxx), soil, meaning soil that ground, or dust, and it refers to the dusty, pale-colored, phyllitic covers a substantial part of the hill. Chomatas

IO

CHAPTER

I

Figure 1.6. Olive trees growing in Chordakia, looking northeast from the habitation site (1999) Chordakia (Xcop?cxKia) is used for an area of olive groves on name is used of Mt. Chomatas only slopes (Fig. 1.6). The for the cultivated part of the hill. It is a contraction from Chorioudakia

The

name Chordakia

the inland

(Xu)pioi)8(XKia), from chorio (xcopi?), meaning village. The term is often used for a place where an old settlement occurs. The reference here is to a Hellenistic, and Byzantine site with remains Roman, Kephalolimnos, from at least as early asMiddle Minoan times. Substantial ruins are visible the ground.16 Choriodakia is a variation recorded (Xcopio?aKia) is another local variant. Both Khordakia and by Boyd,17 and Choridakia are used name is Khoridakia Kanta18 and The by by Haggis.19 applied to above

to the modern the region from the edge of the Kambos fence that sets off the public land at the seacoast, not just to the archaeological sites. the cultivated plots (Chordakia) The sharp boundary between and the barren landscape at the coast is clearly visible in the satellite photograph shown

in Figure

1.3.

Chrysokamino Chrysokamino a Semitic gold, and kaminos

(Xpuaomurvo) means gold furnace, from chrysos (xpuao?), loan word that appears in Greek as early as Linear B,20 (Kajiwoc) or kamini (kgcu?vi), a furnace or kiln. It is used

locally both for the actual location of the ancient metallurgical activity site (Fig. 1.7) and for the part of the hillside on which the metallurgical name is located. A local variation is in the plural, Chrysokamina.21 The is used by the current project for the territory because it Chrysokamino is a unique name, while Katsoprinos, the local name for the location of more one the habitation is used for than site, place in this part of Crete, as and it is already in the published the name for a different literature archaeological

site (see Fig.

1.5).22

16. Boyd 1901, p. 156; Sanders 1982, p. 141;Haggis 1992, pp. 160 161, locus 2; Fotou 1993, p. 102. 17. Boyd 1901, p. 156. 18.Kantal980,p.l45.

19.Haggis 1996b, p. 381. 20. Chadwick 1973, pp. 136 and 343.

21. See Lamb 1929, p. 5;Haggis 1992, locus 88; 1993a, p. 28; 1996b, p. 381.

22. Haggis 2005, p. 35, locus 28.

II

INTRODUCTION

Figure

1.7. The

site of the metallur

^

gical activity at Chrysokamino

Figure formation

1.8.

Theriospelio

Chylopittes, at the coast

u?e*S?&

a travertine near

the

cave

Chylopittes sea clearly visible from the (Fig. 1.8) is named is It north of cave, on a cliff (X\)^07utt??). Chylopittes Theriospelio a sea. name is cooked mass of above the The derived from chylos (xuXo?), or t?tcoc), a pastry or to the pie. It refers ground cereal, and pitta (nrina

A

travertine

formation

pale brown stalactitic pastalike food.

formations

that look like chylopitta (xu^Otuttoc),

a

Kambos coast atTholos Bay is called the large valley that extends inland from the kambos Kambos, from the word (KOurco?),which means plain or flat field term is borrowed from the Italian word campo, derived from (Fig. 1.9). The

A

12

I

CHAPTER

Figure

1.9. The

Kambos

as seen

from Kavousi, with the coastal hills behind it campus. The valley is inland from Chordakia, as defined in this volume. An territory Chrysokamino the Latin

and it is outside

the

intermittent

stream,

the Platys River (called the Kavousianos River by Faure),23 flows in the spring and empties into Tholos Bay. the Kambos

through

Katsoprinos (Kaxa?rcpivoc), a local variant of katsoprini (mTCOnpivi), the Katsoprinos holly oak, is also called katsiprinos (KctToinpivo?) in Crete.24 A local variant is katsoprinios. This evergreen shrub, Quercus cocci/era, is a member of the in Crete on the lower slopes of hills.25 Fagaceae family, and it grows widely It has dark green spiky leaves and seldom grows taller than a few meters. shrub gives its name to the dolomite outcrop where the habitation site is located because several of the shrubs grew there before excavation name is not used exclusively for this farmstead, as it is also began. The

The

applied remains

Lakkos

to a spot lower on the hills can

be

seen

above

east where slope to the

other Minoan

ground.26

Ambeliou

It is a low (Aolkko? p\\mzk\oxi) is south of Katsoprinos. a has of red sediment where substantial of the Kambos) valley (a part deposit accumulated from the erosion of the nearby hills (Fig. 1.10). The deposit,

Lakkos Ambeliou

is a pedon, is most likely a result of alluvial action that deposited the sediment in a lake.27 The sinkhole has been used as a clay source by a local company engaged in making brick, and residents of the village of Kavousi regularly take the red sediment to use it as soil for their gardens

which

in a substantial

and vineyards. The removal of the sediment has resulted called Lakkos Ambeliou. excavated area. The location is appropriately Lakkos (Axxkko?) means a hole in the ground, and ambeliou is the genitive is of ambelos (aurce?o?), meaning vineyard. The variant Ambeli Lakkos used occasionally.

23. Faure 1984, p. 47; 1989, p. 320. 24. Alibertis

1994,

25. Chavakis

n.d.,

1987,

pp. 77-78. p. 187; Sfikas

pp. 32-33.

26. Haggis

1996b, p. 400; 2005,

p. 35, locus 28. 27. Morris 2002,

chap.

3.

INTRODUCTION

13

Figure 1.10. Lakkos Ambeliou, in the Kambos, looking south from the habitation site (1999)

Figure 1.11. The burial cave called Theriospelio Therio (0epio) is a rugged, uncultivated, barren hillside northeast of the location. The name, which means wild beast, is a reference to metallurgy the rugged nature of the landscape. The word is recent, as the ancient word

Therio

is spelled with

an eta.

Theriospelio (0epi?o7tT|A,io) is the preferred term for a cave northeast of the site (Fig. 1.11). It has several names. Theriospelio (alsoTherio metallurgy a reference to the wild nature of the is spelaio) surrounding landscape, from therio (6epi?), wild beast, and spelio (any\\i?)y cave. The variation Therio uses A spelos (0epi?G7cr|?,o?) is given by Zois, and Haggis Thergiospilio.28 Theriospelio

28. Zois 1993, p. 340; Haggis

1992,

p. 171.

29. Faure 1964, p. 227; 1966, p. 48.

synonym, Theriotrypa

(Oepioxpurca),

is recorded by Faure.29 Agriospelio

CHAPTER

14

I

from (AypiocKX]kio) was used by a few residents of Kavousi in the mid-1980s, a or for wild therio. (Kotao savage, synonym Kolonospilios agrios (?ypioc), voojcnAaoc) is given by Faure,30 and the neuter gender variation Kolonos names are from kolona pilio (Kotauv?G7rr|Xio) is added by Haggis.31 These (ko?-covoc), one of many a reference

column,

to

Italian loan words the

cave

s

stalagmite

in use in Crete

(from color?na, a

formations).

Discussion for the land features near Kavousi

belong to several classes. Most of names near are Greek. for the territory the modern Chrysokamino Among the exceptions are the word kolona in Kolonospilios, derived from Italian, and the word kambos from the same language; they are presumably from

Terms

was borrowed period of Crete. The chrysos in Chrysokamino from Semitic as early as the Linear B documents, and it has been common in Greek since the end of the Bronze Age. the Venetian

in Crete, the names used for locations on and toponyms a or about Chomatas signify something about place's appearance events or features associated with it.None of the names necessarily derives Like most

near Mt.

or from any earlier time. The period only designations and Chordakia. Chrysokamino that allude to antiquities are Chrysokamino is not the name of an actual town, refers to the wrong metal, and Chordakia in the vicinity. All of the names but an indication of an early settlement some at to least the and of them are probably much back 19th go century, from the Roman

older, but the dates when they were first used are not known. They have been handed down in an oral tradition, so it is not surprising that more than a single variation or name exists for some places.

EARLIER WORK IN THE CHRYSOKAMINO TERRITORY the second half of the 20th century, the Chrysokamino metallurgy During location and its nearby territory has been one of the most controversial visited the region between modern ancient sites inGreece. Archaeologists and the sea shortly before 1900, and the region iswell known from the archaeological literature. The metallurgical workshop's period of use and it has been assigned dates ranging has been particularly problematic, to modern. The location has even been considered as a site from Minoan

Kavousi

as an alternative to a metallurgical installation. The for lime production as industrial ceramics on the surface have been regarded pieces of crucibles, furnaces, covers for furnaces, and the floor of a large kiln. before 1900. Arthur knew about Chrysokamino Archaeologists

Evans

the region in the 1890s. Harriet Boyd excavated Roman remains at over Chordakia (east of the Chrysokamino territory) in 1900, and shewalked much of the region. In an article about her 1900 excavations at Kavousi, visited

she reported the Roman discoveries at Chordakia and also noted that "early to Iron Age remains) was architecture" (a term she employed forMinoan on top of the coastal hill.32 The architecture is probably the hilltop site

30. Faure 1964, pp. 32,227; 1966, p. 48.

31. Haggis 1992, p. 171. 32. Boyd 1901, p. 156.

INTRODUCTION

IS

recorded here as Pyrgos Chrysokaminou (Fig. 1.5), but it could also refer to the farmstead excavated by this project. was first discussed in the The metallurgy location at Chrysokamino of the excavations at Gournia.33 Harriet Boyd Hawes final publication a site called "Golden reported that in 1900 the local residents showed her a few of where she collected Furnace," copper. She said that "rock " pieces obtained from the adjacent cliff contained traces of copper. Her report noted that the ground was strewn with fragments of "an ancient furnace." She did no work the

earlier

there personally,

and a later report simply

summarized

one.34

Angelo Mosso Joseph Hazzidakis

discussed

the site in more

detail.35 He

(who was

reported that visited the site

of Crete) then Ephor-General twice, once in 1906 and a second time either before or later. Hazzidakis loca collected a piece of metal and specimens of slag from the metallurgy cave. not who have visited the and he also Mosso, tion, may Theriospelio was not it and he visited the did realize cave, natural, completely actually

as the mine for the ore that was smelted nearby. He reported regarded it Minoan pottery from the cave, but no sherds aside from what he Early as sections of "crucibles" were found at the metallurgy location regarded itself. Pieces

of "scoria"were

also discovered

in the cave. An

illustration

of

one of the "crucible" (reproduced here in fragments from Chrysokamino were that artifacts the ceramic shows the pieces with pierced Figure 1.12) holes and a glassy deposit adhering to one side that still cover all unexca vated portions of the site.Mosso analyzed the piece of metal and reported Figure 1.12. A fragment of industrial pottery published by Angelo Mosso. After Mosso

1910, fig. 164. Scale

1:4

that it contained

45% copper. He also analyzed the "scoria" collected from Chrysokamino. a trace of copper, and as this result it that contained complemented reported for extracting his other researches, he concluded the site was aworkshop Mosso

copper from its ore. He reported finding pieces of similar "scoria" on the beach at Pacheia Ammos, suggesting another site where copper may have been smelted. was a Bronze believed Chrysokamino Mosso Age installation. He did not publish any pottery from the metallurgy workshop, but he did describe visited the ceramics from the nearby cave. He reported that Hazzidakis aswell as from cavern and collected sherds from "the primitive Early period" on a dark paint Minoan period. A ground, material that contained copper was also collected from the cave, but its exact s brief is difficult to comprehend from Mosso identification description. Edith Hall and Richard Seager investigated the location more than

Minoan

II. One vase from the cave had decoration

so itwas either from EM

33. Hawes

et al. 1908,

34. Hawes

and

Boyd

p. 33. Hawes

1909,

p. 38.

35.Mosso 1908; 1910, pp. 289-292 and fig. 164. 36. Unpublished letter of April 24, 1910 by Edith Hall in the archives of the University of Pennsylvania Museum of Archaeology and Anthro pology;

see also Becker

and Betancourt

1997, p. 109. 37. 473.

Schachermeyr

1938,

pp. 472

inwhite

III or from theMiddle

letters.36 once, and a visit they made in 1910 was recorded in unpublished sea Hall called the the after Pseira. headland by visiting They approached times." site "the place where they smelted their bronze [sic] inMinoan cave on the same trip, and Hall returned for two days of They visited the excavation there the following week. The limited information concerning her excavation is reported in this volume. Fritz Schachermeyr visited the territory in the 1930s and made a ground location as a part of his survey of the region.37 He visited the metallurgy was there, he a of Vasiliki Ware he sherd While (the picked up exploration.

i6

I

CHAPTER

location, confirming pottery for EM IIB) at the metallurgical the Early Minoan date of the site. A few discussions of the metallurgy workshop appeared in various in the years after Hall worked in the area. Until the scholarly writings

definitive

was as a 1960s, Chrysokamino occasionally mentioned mining, smelting, or were location.38 These all based on the research reports copper-working conducted before 1910 and on Schachermeyr's 1938 report of his visit to the small promontory. New work began

in the 1960s. LM

IIIB:1

tombs at Chordakia

were

in 1962.39 Chordakia

is the region just inland from were the LM III and tombs from Chrysokamino, probably related either or to the to the settlement at Katsoprinos site. habitation Chrysokamino Paul Faure surveyed the area is the early 1960s.40 He explored the excavated

by Davaras

site. For the lat cave, and he also examined the metallurgy Theriospelio ter, he suggested that the ceramic fragments on the site were not from crucibles but from a single large furnace used for smelting chalcopyrite ore, no evidence for this on the state of although he provided assumption. Based of the glassy waste products, the absence of soil accumulation, and the fact that he believed the ore reduced was a sulfide, he regarded the

vitrification

site as "relativement a smelting correctly

operation rejected

suggested it represented the remains of for copper using either charcoal or coal as fuel. He

r?cente." He

the

nearby

cave

as a mine.

Keith Branigan location. He also visited the metallurgical the deposit of "cinders, ashes, slag, and crucible fragments" as tion of copper remelting (not smelting), and he gave the site Electron Microprobe date.41 Based on an analysis by Geoscan showed that a temperature of 1150?F or more had been which

regarded an indica a different Analysis, reached in

the slag and that calcium was present, he suggested the site was later than Roman. He used literary sources to propose the 12th century A.D. as the most likely date. Branigan correctly realized that the cave was not amine but a natural cavern in limestone, with in the

no visible

trace of nonferrous metals

rock.

60 years of surface collecting by archaeologists, diagnostic sherds were no longer present on the surface of the site when Faure and metallurgy After

on Branigan visited this part of Crete in the 1960s. They based their dating the characteristics of the slag, and they cannot be criticized for their conclu was based on research in sions because the methodology they used, which as valid at the time. This was widely rested regarded methodology Cyprus, on the mistaken belief that before the Roman period, early metallurgists could never achieve high temperatures with their primitive technological any slag with evidence for substantial vitrification knowledge. Therefore, had to be more recent.42 The metallurgy site, like the rest of the territory around it, did on Minoan not figure prominently in the scholarship Crete after the 38. Fimmen 1921, pp. 17 and 120; Glotz 1923, p. 39; Lamb 1929, p. 5; Davies 1932, p. 987; 1935, pp. 7,113, 264,270; Money-Coutts

Pendlebury,

Pendlebury,

1935-1936,

p. 104;

and

Forbes 1950, p. 364; Hutchinson pp. 40,247. 39. Kanta nos. fig. 56,

1980, 6 and

pp.

144-145,

8.

40. Faure 1964, p. 32; 1966,

1962,

pp. 47-48. 41. Branigan 42. Koucky summarize and

table

1968, pp. 50-51. and Steinberg nicely the system (1982a, p. 156

1).

INTRODUCTION

V

after this time of early Aegean metallurgy written of the modern project often omitted the site the beginning was Medi completely.43 Several writers accepted the theory that the site eval.44 A general history of metallurgy written byTylecote suggested the 1960s. Discussions

and before

site was

used for the remelting of copper.45 An exception to the general a considered the site to be from the rule, study of daggers by Nakou, Bronze Early Age.46 Karen Foster published a catalogue of the collection ofMinoan objects in 1978, including pieces from in theMount Holyoke College Art Museum cave excavated Edith the Theriospelio Other pieces from the Hall.47 by same excavation were included in a the of Pennsyl of catalogue University vania Museum of Archaeology and Anthropology.48 In 1983, Noel Laboratory, Oxford

Gale

and Zofia

of the Nuclear

Stos-Gale

Physics as University, samples from Chrysokamino a part of a in archaeometric of the survey larger metallurgy Aegean.49 They collected slag and pieces of the industrial ceramics for scientific analyses collected

types. The results of their analyses showed that the site was ex in the Aegean Bronze tremely important for the early history of metallurgy to establish the 3rd millennium b.c. Age. They used thermoluminescence of various

date of the industrial

ceramics. Their lead isotope studies indicated that the copper might have originated in Lavrion or Kythnos. The Chrysoka mino territory was included in the regional survey of Donald C. Haggis. in conjunction with This was a doctoral dissertation project undertaken excavations at Kavousi.50 The archaeological laid the foundation for the more detailed investigation

ofMinnesota

the University

survey by Haggis reported in this volume, inasmuch settlement inwhich Chrysokamino

the regional pattern of

functioned.

studies of the soils of the area around Chrysokamino with the Kavousi Archaeological Expedition.

detailed

Two were made

as it established

in connection

Michael

Timpson investigated the vicinity of Pacheia Ammos series of soil pedons, including

Late Quaternary alluvial sediments from and Kavousi.51 Michael Morris studied a the one at Lakkos Ambeliou

just south of two the Chrysokamino habitation site.52The investigations complementary on considerable information the of soils the gathered region. The most recent investigation of the area before the current project was carried out by Antonios Zois. He explored the cave, where he began collected Final Neolithic and Early Minoan pottery, and he also surveyed area. He raised the the metallurgy workshop possibility that the latter site was

a modern

limekiln.53

In 1995, logical 43. A.

C. Renfrew

1972;

47. Foster

Branigan

1974; Healy 1978;McGeehan-Liritzis 1996. 44.

Stos-Gale

p. 59; Haggis p. 341. 45.

Tylecote 46. Nakou

and Gale

1984,

1992, p. 170; Evely 2000,

the present project began with survey and topographical mapping. 1978.

48. Betancourt 49. Gale,

and Gilmore

1985; Gale and Stos-Gale 1989; Stos Gale 1989; 1993, p. 124; 1998, pp. 720 721.

1976, 1995.

p. 19, table

19.

50.Mook Haggis

preliminary

season of geo

pp. 373-379, fig. 7; 1996a; 1996b, pp. 380-381 and 401-403; 2000; 2002;

1983.

Stos-Gale,

a

andHaggis 1990; 1992, Loci 50 and 88; 1995,

see also Haggis pp. 287-288.

and Mook

1993,

51.Timpsonl992. 52. Morris 1994,2002.

53. Zois 1993, pp. 340-341.

i8

I

CHAPTER

GOALS OF THE CHRYSOKAMINO PROJECT towns in substantial information was known about Minoan Although this part of Crete before this project began thanks to the excavations at Gournia, Pseira, and Kavousi, much less was known about the tiny rural farmsteads that consisted of isolated architectural complexes. No Minoan had ever been

farmstead

studied with

the intent of ascertaining territory. A Minoan

the strate

gies employed for the use of its agricultural site had not been excavated previously.

smelting

a better of human goals of the project included understanding on in the Chrysokamino with the metal territory, emphasis to its and the lurgical workshop relationship surrounding territory. The project planned to address several separate items: the date and nature of The

habitation

the metallurgy

the types of site, the technology used by the metallurgists, at the the relation of the metal metallurgical products produced workshop, to the nearby habitation of the site, the size and characteristics lurgy site local territory, the prevailing subsistence strategy, the diachronic history of the territory, and its relation to the larger region and Crete in general. The archaeological plan that was developed to address these goals was as an interdisciplinary study involving researchers in several It was based on the nature of the landscape to be investigated, the questions that needed addressing, and the available resources of staff, time, and finances. Because itwas anticipated that much of the material that would formulated

fields.

require scientific analysis for its proper understanding, were invited to metallurgical participate both on-site during the specialists excavation that uncovered the evidence and as laboratory researchers after be discovered would

the material was

excavated. The

plan included

1. Excavation

at the metallurgy

site

2. Excavation

at the habitation

site

3. Small-scale

excavations

at other

selected

the following

aspects:

locations

4. An

electronic instrument survey and mapping project combined with an intensive surface survey to record visible anthropogenic features

5. A program ceramics,

6. A

of laboratory and

other

analysis for metallurgical

remains,

materials

series of corollary studies including geology, geomorphology, topography, soil studies, examination of land use, toponyms, other investigations

and

7. Study and publication research plan was devised in 1994. Excavations were planned for the farmstead and the metallurgy location, along with additional programs of and geomor small excavations at locations to be decided later. Geological

The

phological studies, topographic mapping, soils science, and other specialized studies were planned. A large supporting program of laboratory analysis was conducted at the was assembled. Study of the excavated materials Kavousi

of the INSTAP Study Center until the completion in Pacheia Ammos. Several preliminary reports have been

schoolhouse

for East Crete, published.54

54. Betancourt

1997;

Betancourt,

Floyd, andMuhly 1997; Betancourt, Muhly, and Floyd 1998; Betancourt et al. 1999;Muhly 1999, p. 17; 2002, pp. 79-80; 2000-2001.

Betancourt

and Floyd

reports Preliminary in this volume analyses published et al. 1985; in Gale also included

on

the

are Stos

Gale 1989,1998; Gale and Stos-Gale 1989;

Betancourt

and Myer

1999.

CHAPTER

2

The

Environment

Natural

byPhilip P. Betancourt and William R. Farrand1

CLIMATE AND ITS RELATION TO CHRYSOKAMINO climate of eastern Crete

The

cal factors. Like

is one of the regions most important ecologi it affects the choice of plant com

soils and topography, and the agricultural methods

munities

necessary

for successful

farming. climate, and many authors have and winds.2 The general picture local conditions must be considered as

Eastern Crete has a typical Mediterranean discussed its precipitation, temperature, is important, well because

but some specifically local factors seem to have played

a decisive

role in the situa

tion at Chrysokamino. As awhole, mean temperatures and precipitation aremoderate, but the crucial factor is not the mean but the seasonal variation. Table 2.1 illustrates east Cretan agriculture a specialized not will thrive without substantial care. many crops eastern part of Crete receives less moisture than the western and

the seasonal differences endeavor The

that make

inwhich

central parts of the island. Precipitation, mostly in the form of rain, is largely confined to the period between September and May, with the heaviest amounts in December and January. Modern records show that the mean rainfall for January is 96.3 mm, while for July it is only 0.3 mm (Table 2.1). Rain is extremely rare during the hot summer months. It iswetter during the winter, when dry continental air masses move southeast from Europe and cross over the warmer Ionian Sea, with evaporation and atmospheric over the island. Eastern Crete instability leading to precipitation nearly always

receives

less

rain

than

western

Crete

because

the

air masses

move

from west

to east, and the mountains that run along the length of the island as a shield and cause a act for the Isthmus of except Ierapetra majority of to fall in the west. the precipitation

1. The sections

climate of this

P. P. Betancourt. was written

were

chapter The geology

byW.

resources

and natural

written

by

section

R. Farrand.

2. See esp.Allbaugh 1953;Mariolo poulos 1961; Zohary andOrshan 1965; Gat

and

1980.

Magaritz 3. Zohary and Orshan

1965,

suppl.

are moderate.

As the chart in Table 2.1 demonstrates, the temperature in January is substantially colder than it is in July, but it is seldom cold enough to kill hardy plant life. The mean temperature for eastern Crete (as recorded at on the south coast) is 13.2?C for Ierapetra Temperatures

January and 27.2?C for August.3 the summer, fall, and winter, the are from During prevailing winds the northwest. They can be very strong, especially the winter, and during

20

CHAPTER

TABLE 2.1. SEASONAL NIKOLAOS, CRETE

VARIATIONS

January Mean Mean

Temperature Number of

AT HAGIOS

April

12.4?C

16.9?C

6.8/month

4.2/month

2

July 25.8?C

October 20.9?C

0.6/month

4.5/month

Cloudy Days Mean

96.3 mm

Rainfall

27.0 mm

0.3 mm

53.4 mm

1961.

After Mariolopoulos

salt spray inland, damaging land and crops, are not un at low coastal sites.4 In late summer, the north wind (meltemi)

storms that wash common

blows almost every day, and days with wind speeds of up to 7 Beaufort (i.e., 32-38 mph) are not unusual. The force of the breezes varies with the time a are strongest in the early afternoon because of well-known warms situation inwhich the land up more quickly during meteorological the day than the sea does, causing air over the land to rise, which increases the force of the sea breeze. The result is that the wind is strongest during of day.Winds

part of the day. For the human residents, this situation is an It makes the coast more comfortable than inland areas because

the hottest

advantage. "the sea breeze

an the temperature important role in moderating plays "a welcome relief of land the seacoasts," offering strips along near the summer heat for residents who live the shore."5 It is one

of narrow from

of the reasons

that residents

(like eastern Crete)

the world

founding establish

new

In the Chrysokamino territory, which above sea level, these conditions are more

a

in warm parts of number of coastal towns. large north is faces and elevated well settlements

extreme

than would

be the case

so windy and dry that in protected landscape. The land is times it has remained unfarmed, and it is used only for sheep and modern site, which consists of a troughlike goats. The topography at the metallurgy depression between higher exposures of bedrock, is always the windiest part

with

a more

the wind blows from the north. The trough funnels across the site, and few plants can survive at this dry and barren Similar topography was chosen for Bronze Age smelting opera

of the hillside when the breezes

location. tions on Kythnos6 and Seriphos,7 suggesting that the early metalworkers a deliberately chose particularly windy location for their furnaces. are not constant throughout the year. In the spring, they are Winds than in the summer or fall, leading to different condi years. The south wind (sirocco) is especially warm, and it

much more variable tions in different can

to the south Aegean, including Crete.8 The result bring Saharan dust a in the amount of difference is seasonal variations considerable of these moisture

in the soil. The

same wind

that contributes

to the comfort

of the

to some types of crops, because the soil is damaging population cannot retain itsmoisture during the long, rainless period when the weather and harvesting must iswarm and the winds are fairly constant. Planting

human

be arranged according to this seasonal schedule. The local climate encourages the cultivation of crops that can withstand months with many dry, hot days and are not inhibited by periods of dry soils. The triad of olives, grapes, and grains is often regarded as the main group of such crops.9 A convincing case has been made for the growing of more care a a legumes as fourth staple,10 but this is crop that requires much

4.

Seager

1916,

p. 6.

5.Miller andAnthes 1980, p. 94. 6. Hadjianastasiou and MacGilliv ray 1988. 7. Gale and Stos-Gale 1989, pp. 24 25.

8. Pye 1992. 9. Vickery 1936. 10. Sarpaki 1992.

THE

NATURAL

ENVIRONMENT

21

than grains, olives, or grapes, so itwould agricultural strategy, with extra care.

have required a different

type of

exists for a somewhat

Evidence

in comparison with for pollen suitable

the situation

different climate in the Bronze Age in the 20th century. East Cretan searches have not been successful,11 but several

for analysis studies from western Crete suggest

a wetter to period in the Neolithic to current Bronze with the climate Early gradual change leading Age, by the 1st millennium b.c.12 These conditions may have prevailed in eastern as well.

Identification of wood charcoal from cooking fires on the island of Pseira, mostly from LM I contexts, shows that pine was present in addition to the expected trees.13 Pine does not grow on Pseira today because the climate is too dry, although it is present in some of the ravines and other shaded places in this part of Crete. It is probable, Crete

offshore

climate was very much like the present, but therefore, that the FN-LMI with more rain in the FN-LM I periods, and with a gradual drying by the end of the Bronze Age. The Bronze Age countryside has been aptly described patches and

byMoody, of woodland

other

As

who

envisions

and maquis,

a with

setting of rolling hills composed of cultivated fields of grain, olives,

crops.14

the climate became

drier, the agricultural potential of the region increasingly strained. The zonation of the ancient plant would have been affected by both the natural moisture avail

must have become communities

able from rain and human the construction

activities

that altered the environment, of organic matter

of terraces and the addition

including to the soil

to help it retain moisture. These factors must have contributed

to long-term region. Elsewhere

ecological changes in this volume, the taking place in the Chrysokamino evidence for substantial care of the Chrysokamino gardens, in the form of chemical traces in the soil and the deposition of small, isolated sherds associated with manuring practices, is presented. The climatic conditions suggest that plants requiring extra attention would have been harder to grow on the coast of eastern Crete than in the western part of the island or inland areas. Local conditions would have required increasing human effort to make

the landscape productive. factor to be considered is the gradual rise in relative sea level. In the part of Crete near Chrysokamino, the sea level may be as much as 3.5 to 5 m higher than itwas in the Bronze Age.15 This condition will not as much as sites located on the seacoast, but have affected Chrysokamino it must have been a factor in the viability of the harbor at Agriomandra, Another

11. Coring was done byG Rapp; see

Hayden, Moody, p. 307.

and Rackham

1992,

12. Bottema 1980; 1994, pp. 57-58; Moody ham,

1987a, and Rapp

13. Betancourt

chap. 2; Moody, 1996. and

Hope

Rack

Simpson

1992. 14.Moody 1987a, p. 120. 15. Fairbridge 1972; Flemming 1972; Dermitzakis, Karakitsios, 1995; see also comments Lagios Farrand below.

and by

16.Desborough 1964,1972; Drews 1993. 17.Willetts 1955, pp. 46-51; 1990; Jameson

1992.

used by the local residents. The abandonment of coastal settlements

at the end of the Bronze Age has usually been attributed to warlike raiding and other unsettled political conditions.16 This explanation maybe partly correct, but additional factors must be invoked to a part of the coast that had explain why supported was not resettled in the more stable two residents for millennia permanent new Dorian periods. The farming methods using serf labor on large estates may have been one of these factors.17 It is also likely that the climate had become too dry to support a local population on the coastal As a result of climatic change, the strip at Chrysokamino.

Classical

and Roman

land, which had always been less than desirable, may settlement. poor that it did not merit permanent

now have been

so

22

CHAPTER

2

GEOLOGY OF THE CHRYSOKAMINO AREA area from report results from two visits to the Chrysokamino July 21 ex to 28,1995 and July 23 to August 2,1996. Detailed field observations tended somewhat beyond the limits of the map in Figure 2.1 and included

This

of outcrops, recording and sampling road cuts and excavation mapping trenches, and discussions with the topographic survey team. Reconnais sance observations extended still farther into the hills above Kavousi, the beach area, and into the Ierapetra graben. Detailed studies of the soils of the immediate area and of the surrounding region, reported in two dissertations from the University of Tennessee,18 proved very helpful. our This section ismodified previous report of the geol slightly from

Tholos

area are very ogy of Pseira19 because Pseira and the Chrysokamino-Kavousi similar in their relations with respect to their general geological setting in eastern Crete. They are separated by only 2.5 km of water, and the distance to the metallurgical site is only 6 km. eastern in is of Crete detail but relatively simple geology complex in its broad outlines. For our purposes, the broad outlines will suffice for

from the center of Pseira The

an

understanding the Chrysokamino fieldwork, mostly and of the Aegean

of the regional relations of the major bedrock units of area. The summary is based on a spate of following in the 1970s, dealing with the tectonic setting of Crete

as awhole.20 region On the broadest scale, Crete constitutes the central part of the Hellenic Arc, a major tectonic feature that curves from the P?loponn?se through of southwestern Turkey. The Hellenic Arc Crete to the Taurus Mountains and European marks the collisional junction of the African lithospheric now plates where portions of the Earths crust, formerly situated inwhat is the central Aegean, have been thrust southward to override a thick section platform limestones constituting the bulk of the "basement" rocks of Crete. The overriding rocks, in the form of relatively thin sheets called nappes, arrived in successive waves (Figs. 2.2,2.3).21 of marine

of

Bedrock

Eastern

Crete

most specifically to the part of Crete east following comments apply lowland shown in Figure 2.2. The first ofMirabello of the Ierapetra-Gulf overthrust sheet is the "Phyllite" nappe that directly overlies the autoch thonous, relatively thinly bedded, cherty limestones known as "Plattenkalk"

The

(Fig. 2.4). The metamorphism, ward through

rocks of the Phyllite-Quartzite nappe exhibit low-grade ismost intense in its lower part and fades out up the nappe. which

and the overlying phyllites between the Plattenkalk Stratigraphicalfy is a rock unit that appeared to us at first to be of a character transitional it appears between those formations. However, upon further examination, a that this "transitional" unit is really the upper part of the Plattenkalk, the regional interpretation. Wachendorf as gray marly schists and colleagues described the top of the Plattenkalk some examination of one of section interbeds.22 Thin limestone with conclusion

that is consistent with

these rocks showed both

the carbonate mineralogy

and the metamorphic

1992;Morris

18.Timpson 2002. 19. Farrand

and Steams

20.

E.g., Wachendorf, Gwosdz 1975; Baumann Baumann, Seideletal.

Best,

1994,

2004. and Best, et al. 1976;

and Wachendorf

1977;

1981.

21. For a diagramatic of explanation see the emplacement of these nappes, also Baumann, Best, and Wachendorf

1977, p. 517, fig. 8. 22. Wachendorf

et al. 1974.

500 m

Gulf of Mirabello

Figure 2.1. Detailed

bedrock map of the Chrysokamino

area

i

h?i?i?i?i?i?i?i?i?i?i?i?i?i?i?i?r

Neogene

and Quaternary

Tripolitza

Series

i

i

i

i

r

i?i?i?i?i?i?i?i?i?i

Aegean Sea

Phyllite Series Plattenkalk ^^ I? t ] Diorite-Granite \"::\ Magmatite of the Phyllite Series

Gulf ?yX

Pseira

Mirabello Chrysokamin

Figure

2.2.

Bedrock

geology

of eastern

Crete.

After Wachendorf

et al. 1974

THE

NATURAL

ENVIRONMENT

Subpelagonian Series Pindos

Series

Tripolitza Series 2.3.

Figure through Crete.

The

tenkalk) the Only this mino

Schematic

the nappe

of eastern pile limestone (or Plat

cherty is the basement

nappes overlying the three lowest occur figure area. The

the Phyllite trated,

in the chaotic

as is the breccia

Baumann

2.4. Figure with calcite

across

were

which

thrust. in

formations

Phyllitic Series ^^C^

Chrysoka m?lange

of

series is clearly illus

the Tripolitza After

section

at the base

(dolomite) series. et al. 1976

Plattenkalk veins

limestone

of

Cherty Limestone

25

26

CHAPTER

2

2.5. Metacarbonate Figure tion near the metallurgy

foliation

as dominant

features,

and so we have chosen

to call these rocks

metacarbonates Next

(Fig. 2.5). above the Phyllite

nappe is another carbonate series, theTripolitza nappe, comprising massively bedded and reefal carbonates and dolomites. Rocks of the Tripolitza in the far east of Crete, nappe are widespread especially south and east of Siteia. Still higher in the pile are the Pindos and Subpelagonian in eastern nappes, which are less widely distributed are Crete. Rocks of the Subpelagonian nappe important, however, at the near Kalo Chorio, where one finds a chaotic head of the Gulf ofMirabello mixture of igneous and metamorphic rocks (lavas, diorite, marble, etc.). Both

the igneous

rocks and the activity witnessed by the Subpelagonian in the well those rocks before metamorphism Phyllite nappe originated were thrust into their present not which did take position, place until mid times. These

rocks have been detached

from their deep-seated are to be found well to the which sources, magmatic (high-temperature) north of Crete. Some of them have been uplifted and are now exposed in islands of the central Aegean. Tertiary

eastern Crete is characterized a foundation of Plattenkalk by overlain by a succession of nappes that represent displaced ter ranes of central the southward transport of these origin. During Aegean were the rocks within them deformed and sheared to various nappes, as a function of the degrees physical characteristics of the included rocks. Thus, limestone

forma site

THE

ENVIRONMENT

NATURAL

27

shearing is particularly inherent weakness of these

evident in the Phyllite nappe, owing to the foliated rocks. A conspicuous thin-bedded, character of the phyllite series is the occurrence of phacoids: isolated, len ticular blocks of more competent rock types (e.g., limestone) resulting from the fracturing and shearing of original sedimentary beds or lenses.23 These a few centimeters to hundreds of meters, but phacoids range in size from can be seen as blocks a few meters commonly long. A prominent phacoid Such

occurs about midway lower

up the slope between

the metallurgical

site and the

road.

Since the Phyllite nappe is an important element in the Chrysokamino to it isworthwhile area, aswell as on Pseira and in the hills around Mochlos, some a describe its rocks in detail. Figure 2.6 shows typical section through traverse some 10 km east of this nappe, specifically based on a north-south Mochlos.24 Actually, only the lowermost six units described in Figure 2.6 were seen in the area, from the upper part of the Plattenkalk Chrysokamino into the lowermost units of the Phyllite series. These rocks will be described in more detail below. In the middle of the section, relatively thick units of volcanic rocks (especially unit 11, Fig. 2.6) are shown as "Magmatites" in Figure 2.2 and described as intrusive diabase on the Geologic Map of volcanic rocks are slightly metamorphosed (to chlorite and lenses of mica schists, and some grade), they incorporate phyllite, of in marble. Elsewhere the large phacoids Phyllite nappe, but not shown are masses in Figure 2.6, of gypsum, notably in the hills southwest of the

Greece.25 These

are they being mined extensively. of the in eastern Crete are not known with rocks ages because fossils are rare. This situation, of great precision well-preserved course, leads to some differences of opinion in published ages. Some sources

village ofMochlos The geologic

where

is probably of Jurassic to Paleo suggest that the autochthonous Plattenkalk The series is of Permian toTriassic age, gene (Early Tertiary) age. Phyllite and the Tripolitza limestones and the Pindos series are both of Jurassic (?) to Eocene age. The igneous complex of the Subpelagonian nappe has to 88 to 90 million years (Late Cretaceous).26 been dated radiometrically Others have suggested somewhat younger ages, ca. 75 million years (still for the diorites

and granites of the Kalo Chorio area.27 the tectonics that thrust the Following mid-Tertiary compressional succession of nappes onto the Plattenkalk basement of eastern Crete, a

Cretaceous),

tectonics. This situation led period of relaxation gave rise to extensional to new structures bounded by normal faults, oriented roughly high-angle the island of Crete into a number of com northeast-southwest, breaking area is Of for the partments. particular importance Chrysokamino-Pseira

23. Durkin

and Lister

provide

a use

ful description of the origin of phacoids (1983, p. 95, figs. 9 and 10). 24. This is described byWachendorf et al. 1974. 25. Geologic Map of Greece. et al. 1976; 26. Baumann Geologic Map of Greece. et al. 1981. 27. Seidel

the Ierapetra graben, a complex downdropped block that includes the Gulf and its extension through the lowland across mainland ofMirabello cutting Crete to Ierapetra on the south shore. Here, multiple northeast-southwest in a number trending faults have lowered the floor of the Gulf ofMirabello as one of steps or splinters, clearly seen in the present topography proceeds westward from theMochlos Pseira and the bedrock ridge leading vicinity. southwest towards Gournia includes Chrysokamino, (Fig. 2.2), which are horsts, or elevated blocks between downfaulted areas. to According the local boatmen, this faulted topography is expressed on the sea floor on the west

side of Pseira, where

steep cliffs, presumably

part of a fault

CHAPTER

28

2

in 15. Phacoid Series, marble phacoids carbonate a marly matrix; compact layers at the base, which break up into phacoids upward.

600m

^Hl?a

is

14. Clastics, brownish gray, of variable and milky grain sizes (with volcanic quartz

components).

13. Phyllite, olive colored to gray; inter m calated with chlorite schist ( eta at the base and with marble volcanic) at the top. 12. Chlorite and mica schist cm- to dm- thick marble base marble

h 400m

phacoids

with

individual

lenses; at the in the dm-size

range. volcanite, marbled with toward the feldspar, decomposed; east interfoliated with chlorite- and mica schist, as well as large marble

11. Basic

11

phacoids. 10. Marble,

laminated.

schist grayish (metavolcanite), concretions green, with amygdaloid and phyllitic interlayers.

9. Chlorite

8. Mica schist and phyllite, garnets up to the rock is strongly 5 mm diameter;

10

decomposed.

h 200m

7. Basic

volcanic,

dark olive

green,

decomposed. 6. Marble

thin-bedded phacoids, in phyllites.

calcphyllite

5. Limestone, thin-bedded

dark gray, at top.

laminated;

in part with 4. Phyllite, red and green, dark gray mica schist; at the top, red phyllite with dm-thick marble beds. 3. Marble,

structureless,

resembling

lens-shaped,

phacoids.

2. Marly schist, gray with flaser structures beds. as well as individual Plattenkalk

L-Om

1. Plattenkalk

(slabby limestone), with intercalations

crystalline, red and green

coarsely of marly

limestones.

some 300 m to a submerged bench about 300 m line, plunge downward in turn is terminated by another abrupt dropoff, presumably wide, which from the the lowland extending fault. East of Chrysokamino, another to Tholos Beach is another, smaller graben separating village of Kavousi to the east. A small bench horst from the mountains the Chrysokamino it east of the Tholos Beach Hotel; in this graben is located immediately now cemented covered with is it is apparently bedrock (although gravel), and it is probably

another

tectonic

sliver.

Figure

2.6.

Stratigraphie

column

some through the Phyllite nappe what Kavousi

1974

east

of the Chrysokamino area. After Wachendorf

et al.

NATURAL

THE

ENVIRONMENT

29

Neotectonics vertical movements

Differential

above have continued

related to the tectonic

activity described times up until the present. Former and Quaternary date have been mapped at various studies show that eastern Crete stood relatively

into Quaternary

shorelines of lateTertiary on Crete.28 These places sea level at the end of theMiocene high above Epoch, after the thrusting of the various nappes. The area was then submerged at least 450 m early Epoch, only to emerge again progressively throughout the sea Period, during which time oscillating glacial-interglacial Quaternary level changes were superimposed on tectonic uplift. Relative sea-level changes affecting young archaeological sites docu in the Pliocene

adjustments even within the last 2,000 years. A synthesis of these sea-level changes shows that Crete is not reacting as a single block.29 Crete is rising rather rapidly, at rates of about 4 m per 1,000 years, Western ment

tectonic

central Crete has undergone relatively little change. In eastern Crete, corner is being the northeastern submerged, while the southeast is rising. The Pseira-Chrysokamino horst is situated within the Ierapetra graben,

while

a rate late Tertiary subsiding at considerable throughout time.30 The amount of post-Bronze Age subsidence of the area cannot yet be determined. However, et al. give a value of Flemming 0.75 m subsidence in the last 1,900 years for Psyra [sic]; the basis for this not given in their article.31 The same authors list values of 0.25 m figure is and 1.75 m subsidence since 1,900 and 3,600 years ago, respectively, for ar

which

has been

and Quaternary

chaeological features atMochlos. The evidence for these values presumably comes from the interpretation of the presently submerged isthmus between on the Cretan shore and Mochlos Mochlos Island, a submerged village street on Mochlos Island, and drowned fish tanks (piscinae) of presumed Roman age atMochlos village.32 It is possible that the area around Chrysokamino a has undergone amount to of subsidence than the figures Mochlos, greater according cited above.33 Field evidence of uplifted Quaternary shorelines along the shore, just east of Ierapetra, shows a very strong increase in from east to west across the same eastern boundary fault that

south Cretan subsidence

one area.34 Therefore, passes through the Mochlos suggest that might the sea floor off Chrysokamino has subsided perhaps two or three times as much asMochlos times. The has, or some 3.5 to 5 m, since Minoan on the promontory of the metallurgical site is we were not because at do know which faults active problematic, however, a some which times. In any case, difference of 5 m would not significantly effect of these movements

change

the altitude

of the metallurgical

site, which

et al. 1982; Peters, and van Harten 1985.

28. Pirazzoli Troelstra,

29. Flemming, 1973.

Czartoryska,

Hunter

sits 38 m above

33. Hunter

30. Fortuin (1978) gives details of structural

fig.

history.

31. Flemming,

Czartoryska,

and

1973,

table

1.

32. Soles 1978.

and

Hunter

this

now

sea.

the

Flemming, 1973.

34. Ang?lier 5.

Czartoryska, et al. 1976,

and

p. 436,

CHAPTER

30

2

Figure 2.7. Cliffs north of the metal lurgy site, looking toward the island of Pseira

Local

Bedrock

site is on an isolated point of land metallurgical Chrysokamino m above the sea with no immediately about 38 adjacent boat perched a but site is backed available. The by sharp, passable ridge rising landing

The

another

80 m without

areas. The

point of land where the cove on the north side (Fig. 2.7) cliffs and receives the brunt of heavy

significant level two broad coves. The

site sits separates is flanked by very steep, limestone cove waves generated by prevailing winds from the northwest. The south more is less and bordered steep slopes by protected (Fig. 2.8) is somewhat on the softer nonviolent In times of and related rocks. weather, phyllite small ships could anchor in the south cove, but offloading would be dif ficult because of the narrow beach and the steep slope up from the water. about 800 m southwest of the site, around a small but prominent However, a cuts through the local dolomite rocky peninsula, major ravine (Fig. 2.9)

bedrock down to a small, but sheltered cove (Agriomandra, Fig. 2.10), ideal for small ships (see Figs. 1.5 and 2.1). The cobble beach at the mouth of this ravine contains many exotic rock types, i.e., rocks not found in the local area (for traces of copper alteration example, basalt, diorite, and rocks with that the exotic rocks were dumped here from products). It is conceivable access to flat fields above ravine ballast. This ship provides relatively easy situated south of the farmstead at Katsoprinos (near Lakkos Ambeliou), remnants of awell-built road (aTurkish kalderimi) (Fig. 1.5). At present, are still present ravine. the through The Plattenkalk limestone forms the structural support of the Chryso as cliffs kamino headland, well as forming the impressive, near-vertical some 200 m out of the sea to the northeast

site of the metallurgical with interlaced thin-bedded but solid, gray, (Fig. 2.7). to distinguish conspicuous white calcite veins (Fig. 2.4), which make it easy between the Plattenkalk limestone and the local dolomite. Some of the rising

It is dark bluish

away in the immediate vicinity of the water s The Plattenkalk contains a number of small fissures. leaving open

veins have been dissolved edge,

THE

Figure 2.8. Caves and cliffs south of Chrysokamino

Figure to the

Figure harbor connects

2.9.

Red

ravine

clay

and

leading

2.10.

to

Agriomandra

Agriomandra, at the foot of the the coast with

the entrance

a small ravine

that

the Kambos

NATURAL

ENVIRONMENT

31

2

CHAPTER

32

Figure 2.11. Soft, powdery phyllite exposed workshop

of the overriding by deformation during the emplacement site, the Plattenkalk phyllite nappe. To the southwest of the metallurgical on the north side sea at to into down the of 25 but 35?, angles dips steadily of the excavated area, the limestone cliff drops precipitously almost 40 m to the water (Fig. 2.7). folds produced

Metacarbonate rocks, which sit stratigraphically at the top of the Plat tenkalk, are particularly obvious in two areas. Northeast of the metallurgical site, at the base of the rather steep slope that descends from the area of the road gate, the metacarbonate rocks crop out around the mouth of the dry rocks here streambed where itmeets the sea (Fig. 2.5). The metacarbonate are so that their bedding planes, although discernible, are foliated strongly less obvious than the planes of foliation. The metacarbonate rocks aswell as are clearly folded and offset by small faults in this the adjacent Plattenkalk area. Greenish rocks can also be seen along the and reddish metacarbonate foot of the slope due south of the metallurgical site, just below the phyllite contact

were not and above the upper limit of wave action. However, they as a distinct unit in this area because of their gradational relations

mapped with the Plattenkalk

proper. site sits on phyllite bedrock, just a few meters away The metallurgical from the contact between the phyllite and the Plattenkalk limestone. The

(Fig. 2.11) ismostly the white, so residue of weathering, very powdery (silty), calcium-carbonate-rich typi cal of most phyllite outcrops in this area (see below). In addition, in several phyllite

exposed

in the excavation

trenches

carbonate ledge, probably a phacoid, was exposed, across the trenches (strike N40-55E, In angling diagonally dip 40-50S). contrast to the orientation of this phacoid, the Plattenkalk just west of in orientation the trenches strikes N70W and dips 35S. This discordance trenches

a resistant

supports the interpretation of the phyllite the autochthonous Plattenkalk.

as an overthrust

nappe

above

In the cove between Chrysokamino and Agriomandra (Figs. 1.5 and 2.1), the complex nature of the phyllite series is abundantly illustrated by the chaotic nature of the included rock types: well-foliated phyllite (Fig. 2.12),

in a trench

at the metallurgy

THE

NATURAL

ENVIRONMENT

Niatofe"""^* *s

Figure 2.12. Kink fold in phyllite

and gypsum, lenses, strongly sheared limestone masses, all of which may minor and with sandstone pebble conglomerate, along this is a m?lange.35 have heavy travertine coatings. Geologically, hard carbonate

above, the phyllite outcrops are commonly very powdery as a result of in situ weathering rock. At of the calcite-rich, fine-grained first glance, the outcrops look very much like loess (windblown silt), but are cleaned structure of the the foliated when sufficiently, phyllite they As mentioned

ismuch more conspicuous here than on silty weathering inNewark conducted X-ray Pseira. Victoria Herbert of Rutgers University of of rock and the the degraded weathering prod analyses samples phyllite ucts. She reported that the rock specimen (phyllite) was composed mainly of chlorite and quartz, with minor K-feldspar and plagioclase feldspar. No can be seen. This

or muscovite/illite). was detected The (e.g., biotite phase it and but also contained chlorite, quartz, feldspars, phyllite a 10-? micalike and calcite.The powdery, phase (probably illite)

10-A micalike weathered contained

very pale soil was chemically similar to the weathered phyllite, confirming its identification as a late stage of the degraded rock. It contained chlorite, illite, quartz, calcite, and minor feldspars of the pale colored soil A clay separate (approximately <2micrometers) are the only clay minerals. It is unlikely confirmed that chlorite and illite that the sample contains any kaolinite. In the separate, the illite peak is it is more abundant. Probably larger than the chlorite peaks, suggesting no is present. swelling clay The 35. Wachendorf

et al. 1974.

Tripolitza

southern nappe

of the dolomite part of the map area is dominantly contact A the dolomite clear 2.1). very separates (Fig.

CHAPTER

34

2

from the adjacent phyllite series. The farmstead building is on the dolomite contact (in the parking area). The immediately adjacent to the phyllite to light gray, to is predominantly dark dolomite gray, but itweathers gray or white. It is massively bedded, with the bedding planes grayish brown, so the local dip and strike. that it is rarely possible to determine obscured it is commonly rather steep to the the dip can be determined, Where as cove caves across in south of the metallurgical the the south, e.g., large the other hand, the dolomite along the contact with the phyllite into angular rock fragments. They have is highly fractured or brecciated been recemented together, indicating brittle fracture during overthrusting. site. On

seen along the contact west-northwest readily be at dolomite also occurs as a thin from the farmhouse site Katsoprinos.The cover or as isolated outcrops above the phyllite in the area between the

This

characteristic

can

and the metallurgical site, especially in the form of a long "fin of the in the north-central part map (Fig. 2.1). Also, a small, isolated ger" (too small to show on the map) occurs outcrop (or outlier) of dolomite on the east just inside the road gate, and another larger outlier is situated farmhouse

hills just above the extensive olive groves in slope of the Chrysokamino Beach valley. Large boulders of dolomite have fallen the Kavousi-Tholos down to the shoreline as part of a landslide in the middle of the cove south of the metallurgical The dolomite

site. is particularly

as to solution weathering, susceptible are of lapies, which sharp, small-scale

shown by widespread development on solution features are and grooves exposed surfaces. Larger-scale ridges also present. First, several relatively deep fissures occur in the dolomite a short distance west of the farmhouse site. Four of these were investigated. are 1 to 2 m wide at the top and up to 4 m deep at least. The bot They toms are obscured, being partially filled with colluvium and fallen blocks the observed fissures are not strictly parallel, they all of dolomite. Although trend. Undoubtedly, have a generally northeast-southwest they follow joint are the two caverns features in solution the dolomite. Still patterns larger at sea level on the lower slopes of the dolomite, facing north toward the site. These caves are not very extensive horizontally, but they metallurgical have high arching roofs, which make them very conspicuous when viewed caves occur on the walls

of the sharp cove of mentioned Agriomandra, into a small chapel, but the above. One of these caves has been modified other is in its natural condition. All of these solution features appear to from the metallurgical ravine leading down

site. Smaller

to the sheltered

be part of amajor karstic system that pervades the dolomite; however, the nature of this bedrock probably precludes the development highly fractured of

long

cavern

systems.

in other bedrock formations as well. In the case developed of the phyllite, some small caves appear to have been dug by humans or phyllite is very soft and easy perhaps enlarged by animals (the weathered area can seen to the be along phyllite slopes in the dig). Several examples Caves

are

of the large agricultural terraces just northeast of the farmhouse is large enough largest of these, discussed further inAppendix K, and goats. sheep

site. The to shelter

THE

NATURAL

ENVIRONMENT

35

occurs in the Plattenkalk (Fig. 1.11), just largest cave,Theriospelio over 200 m north 50? east of the from site, metallurgical directly upslope travertine headland called Chylopittes the white (see Fig. 1.5), at about 50 masl. It has a small opening, only 1.5 m high at the entrance, but at The

room with many large to be actively dripping. stalactites This room continues into a dark cavern that appears to be extensive, but cave is discussed in we were not Chapter equipped to explore farther. The 18 of this volume. about 5-6 m

the cave opens into a very and stalagmites, none of which seemed inside,

Colluvium,

Landslides,

and

Soils

area is surface in the Chrysokamino generally barren of soil a few cover or unconsolidated with exceptions. Along the shoreline deposits, in the middle of the south cove is a small landslide deposit (LS on Fig. 2.1) some larger blocks of do composed of rubble from the phyllite series and m some 80 lomite fallen from the outcrop upslope. Directly upslope from the metallurgical site, on the uphill side of the lower road at about 100 masl, is a fossilized ravine (Fig. 2.13). It is obvious by its reddish brown color,

The

bedrock

contrasting with the underlying, pale-colored phyllite (Fig. 2.14), and by a to the fact that it appears be filling small ravine cut into the phyllite. Its contents include rock fragments from the Plattenkalk limestone that crops out a bit higher on the slope. Elsewhere on the phyllite rocks is a thin colluvial cover, usually 20 to cm 30 thick, of reworked phyllite and associated rock types that have been moved downslope by sheet flow and soil creep (Fig. 2.15). This colluvium is easy to detect because it has a very clear contact with the underlying can be seen at the top of road phyllite and is browner than the phyllite. It cuts between the metallurgical and farmhouse sites, as well as along the road on the east side of the hills leading down into the Kavousi valley. site The field of red clay colluvium and soil south of the farmhouse (at the location

2.13. Figure the metallurgy

Paleoravine site

uphill

from ?*? *** ****

~ " ys?r

v*

called Lakkos Ambeliou)

is described

below.

CHAPTER

36

2

Figure 2.14. Detail ravine

showing

of the paleo

the contact

between

the reddish soil in the ravine and the 8***

underlying pale-colored phyllite

Figure 2.15. Colluvium consisting of reworked phyllite and other rocks Economic

Geology

a search of the obvious metallurgical interest of Chrysokamino, for possible sources of ore, but none was found. In fact, none of the local bedrock types seem obvious environments for mineralization. The

Because

was made veins

in the Plattenkalk

appear barren of any minerals other than calcite. as 10 cm thick) and are much coatings (as fillings of travertine associated with outcrops of dolomite, especially in and near the two large

Extensive

caves mentioned

was seen. Also, a above, but no mineralization large out on most travertine 1.8 is of andTR of which 2.1), crop (Fig. brilliantly Fig. occurs on the white, but with zones that are red, brown, or honey-colored, north side of the cape about 300 m north-northeast of the metallurgical

THE

ENVIRONMENT

NATURAL

37

site (Fig. 2.1). The deposit is known locally as Chylopittes (Fig. 1.5). It is associated with eroded stalactites and capped by Plattenkalk limestone, all of which plunges below sea level at an angle of 45-50?. This travertine ap pears to be a filling of an ancient karstic cavity, mostly destroyed by coastal was obvious. erosion, but no trace of metallic mineralization raw materials that are locally available include both Other geologic and clay. Two modest outcrops of gypsum were noted (GX on one coast the about 250-300 m southeast of the metal along Fig. 2.1), on the m east-southeast lurgical site and the other high ridge about 600 gypsum

of the site (just outside the road gate, below the stone hut recorded asAF the phyllite series. 15 inApp. G), both within red clayey sediment fills an upland basin named Lakkos Abundant Ambeliou, which appears to be a large sinkhole, located at the head of the major ravine discussed above, down the slope and almost due south of the terra rossa soil eroded from the site. This clay is redeposited farmhouse surrounding dolomite slopes; it seems to be a reasonable source of clay for ceramics. The deposit is at least 4 m thick in the center and covers some 25,000 to 30,000 m2. It comprises two layers of red clay: a younger one, as as 1.10 m,36 with occasional Middle Minoan (MM) sherds on the deep artifacts. The lower deposit, surface, and an older clay deposit without at 2.1 m below the present surface, is beginning brighter red and harder than the upper one and is characterized by light-colored calcium carbonate and by manganese staining. The upper deposit is at least than older and perhaps much older. Much of 4,000 years old, i.e., MM, the terra rossa in this area is considered to be relict from wetter ("pluvial")

veins and nodules

conditions Morris

that prevailed during times of northern glaciations. studied the red sediment in detail as a part of his dissertation

sites in eastern Crete.37 He presents a study of soils related to Minoan of laboratory analyses including grain size determination, constellation as a mineralogical analysis, and chemical analysis. He classifies the soil that it is a highly colored Vertisol developed Chromoxerert, meaning under seasonally dry climate. A Vertisol is a clay-rich soil with a strong causes to open in the dry season shrink-swell that cracks capacity large and turbation

or

and drying. mixing of the soil upon sequential wetting causes a phenomenon called "slickensides" to appear on the faces of the peds (structural units) within the soil, that is, shiny surfaces with lines or grooves produced as the peds rub against each other during swelling of the soil. This particular soil is unusual in that its clay minerals include only illite and kaolinite; no expandable clays (such as smectite) are The

turbation

present.38 Morris believes that the strong turbation in this soil is a product of the very strong seasonal contrast between the wet and dry seasons on Crete. From an industrial point of view, the lack of expandable clays would be an advantage in ceramic production because of the reduced likelihood of cracking upon firing. Morris concludes that "movement

36. Morris

2002,

p. 57.

37. Morris 2002, chap. 3. 38. However, all clays will shrink and swell to some extent upon wetting and drying. 39. Morris 2002, p. 59.

of the artifacts through the profile seems to be on the dependent limiting diameter of the artifacts and the of the vertical cracks."39 He believes that the artifacts below morphology there through cracks. A depth of 1.1 m appears to be the MM sherds have supposedly moved depth to which

the surface moved the maximum from

the

present

surface,

where

they

are

also

found.

CHAPTER

38

2

Summary The

bedrock of the Chrysokamino tions. The Plattenkalk limestone

area consists

essentially of three forma out to the north and crops immediately site and continues in the high cliffs to the north

west

of the metallurgical dolomite begins at the farmhouse site and continues east; the Tripolitza in the high hills to the south and southwest; and the Phyllite series lies between these two carbonate formations. Whereas limestone and dolomite

rocks form uplands of resistant rocks, the phyllite is highly weathered and easily eroded, forming unstable, less steep slopes. The phyllite has to an arable soil and has been exploited for terrace weathered agriculture it crops

wherever

out.

no metallic mineral

were found in the area, resources Although deposits in the form of travertine, gypsum, and clay are present. Carbonate solution features are found in the limestone and the dolomite. A deep cave with and later use occurs in the Plattenkalk limestone, and lapies, fis are caves in and smaller found the dolomite. The topography of the sures, was area suggests that the access sea means of the from the only safe by to the southwest of the sites, which connects small cove of Agriomandra site. with a trafficable ravine leading up toward the main habitation Neolithic

is a tectonically active area, but it appears in the immediate area of Chrysokamino deformation Crete

that any geological has not changed times. Sea level may

since Bronze Age the local landscape significantly have changed by a few meters, but the landscape would affected because of the steep slopes along the coast.

hardly have been

NATURAL RESOURCES resources include rocks, minerals, soils, and other geological features can make the differ regions local plants and animals. They ence between a successful economy and a hopeless failure. Rich natural

Natural as well

as a

resources may be factors in the original decision to settle an area, and they may significantly affect the lives of those who have them. To achieve the successful maintenance of a settlement, balance must be achieved between must take steps to ensure enough food population and resources.40 Someone and other supplies to support the population from one harvest to the next, can if resources (local and increases in population only be accommodated are increased or imported) proportionately. near has relatively few natural resources. Its region Chrysokamino its soils are poor, and the topography does rocks have little mineral wealth, not support a good road system. Access to the sea is limited, and the nearest The

stream, the Platys River, does not flow most of the year.41 Chrysokamino's in the Bronze Age and in marginal economic and political character, both its later history, may be largely attributable to its restricted base of natural did not grow and become a town because Chrysokamino never have supported a large population. local landscape could

resources.

one notable

the

exception, the natural geological materials available in the territory of Chrysokamino pose few problems in their interpretation. With

40. Binford 1968b, p. 327; Smith 1972, p. 7;Wilkinson 1981. 41.Faurel984,p.47.

THE

NATURAL

ENVIRONMENT

39

region has soils of two types and several kinds of rock formations, all of which have been carefully mapped. The other natural resources are also easily discussed. The problem concerns the possible presence of copper a ore.With operating for centuries, the possibility of smelting workshop

The

a local ore source is an important

issue.

Soils terraced areas occur

in the vicinity of Chrysokamino (Fig. 2.16), was once farmed. The basis of this that land the showing farming, the soil, may be divided into two classes. A red soil, called terra rossa, occurs over the local limestone and dolomite bedrock, and a pale brown phyllitic soil Several

occurs over the phyllite bedrock. Both have different characteristics. They

soils have been thoroughly

studied.42

terra rossa contains

The

substantial carbonate, kaolinite, and quartz. over of soil forms bedrock. Most carbonate researchers feel it type consists primarily of the insoluble residue remaining after the dissolution

This

of the parent rock, with the red color resulting from late oxidation of iron minerals.43 The distribution of terra rossa in the Chrysokamino region to the of carbonate bedrock (Fig. 2.1). An especially presence corresponds site, at the low area large deposit is found downhill from the habitation named Lakkos Ambeliou, and the soil also occurs on the carbonate hillsides this location. The red soil is considered suitable for raising overlooking hole," a reference to grapes, and the name of the locality means Vineyard the excavation of the terra rossa soil for use in local vineyards and gardens. The sediment from Lakkos Ambeliou is particularly important for the because it seems to have been the material metallurgy workshop used for the furnace chimneys (see App. A). Eleni Nodarou studied this soil (App. L), and Michael Morris also terra rossa it.44 show The that the consists of sediment analyzed analyses a mixture of several different minerals, containing including enough clay (kaolinite) to make it plastic enough for pottery.45 The sediment is red

Minoan

of the presence of with Munsell colors of 2.5YR 4/6 dry Fe203, 3/6 moist. A substantial amount of silica in the analyses comes from quartz as well as kaolinite and other silicates, and the sediment also

because

and 2.5YR includes

other

oxides. These impurities including calcite and manganese a studies confirm that the soil is useful material for making industrial pottery or coarse storage jars or cooking vessels. It is not really suitable, however, for fine ceramics that require a smooth surface. The

soil that forms from the decomposition pale-colored cal phyllite has very different characteristics. Both Nodarou also analyzed this soil.46 It is more calcareous richer in some other elements, such as sodium. 42. Timpson see also Nodarou,

1992;Morris 2002; App.

43. Nevros and Zvorykin 1936; Danin,

Gerson,

and

Garty

Model

Plasma

Atomic

Emission

Spec

trometer [ICAP-AES] at theUniversity

L.

1983.

44.Morris used aThermo JarrelAsh

of Tennessee, pp.

Knoxville;

113,119-120, 45.

In the 20th

seeMorris

Kavousi century,

3

2002,

pedon. the soil at

Lakkos

of the lo and Morris

than the red soil, and it is In addition, it retains more

Ambeliou

was

used

as a source

of clay for brick making. 46. Nodarou, App. L in thiswork; Morris 2002, pp. 112 and 119, Kavousi 2

pedon.

CHAPTER

40

2

Figure 2.16. Terrace system AF 22, located near the habitation site. The small cave discussed inAppendix K (AF 9) is at the upper left, and ter raceAF 22b is infront of the cave than the terra rossa, so that it is more suitable for dry agricul ture.47 It also consolidates more firmly and does not erode as easily. Local conditions have stripped many of the carbonate hills of much of their red

moisture

soil cover, while The pale-colored

the phyllite slopes are still covered with deep deposits. soil is useful for agriculture, but not for the manufacture

of pottery.

and

Rock

Mineral

Resources

area is described above. Most of the land near the geology of the site at Katsoprinos has a covering of soil, and the outcrops of habitation bedrock are fairly small. Rocks include phyllite, dolomite, limestone, and a few small deposits of other rocks. The limestone and the dolomite are good are available, they are the main stones used they building materials. Where

The

habitation site is built over the for nearby terrace walls. The Chrysokamino on it uses dolomite blocks for its and of dolomite the hill, largest exposure source are a of lime for plaster. This also good walls. The carbonate rocks period, although might have been needed locally in theMinoan no evidence for its ancient use at Chrysokamino has been found. The local sources were definitely used for lime in later periods, and a limekiln from

material

the first half of the 20th century stands near Lakkos Ambeliou. The stone is also a good material for ground stone tools, especially for oval hand tools used in pounding and grinding. Small amounts of gray to black chert occur as inclusions in the dolomite and limestone. As in some other parts of the southern Aegean, however, the as the chert found inmainland Greece.48 As a result, not as good quality is were made of obsidian imported from East Cretan many chipped tools In post-Minoan periods, the residents of northeastern Crete used imported cherts and flints for threshing sledge blades and other purposes. eastern also used many other lithic materials. Although The Minoans

Melos.

Crete

has few deposits

of metals,

it has awide

variety of rock types. The

47. Morris 48. Jameson,

2002,

p. 76.

Runnels,

Andel 1994, pp. 302-303.

and van

THE

NATURAL

Minoans

ENVIRONMENT

collected

41

of them for various

purposes, including tools, other and items.49 The Chryso religious objects, in size and wealth, and many periods are poorly but the residents of the area certainly used east Cretan resources many

ornaments, stone vessels, kamino sites are modest preserved, as materials

for

querns,

stone

other

tools,

stone

vases,

and

other

objects.

Small deposits of several specific rocks and minerals occur in the Chry sokamino Some of them have territory and nearby parts of Crete. Minoan and later uses, while others do not. A deposit of coal lies north of Chrysokamino's did not find any use for territory,50 but the Minoans it.Although central Cretan gypsum was used inMinoan buildings,51 the local gypsum is too porous for architectural blocks (App. G, AF 15). A few other local deposits consist of stones used elsewhere in Crete, but whether were ever used is not known. In the specific deposits near Chrysokamino this latter category are the many small lenses of banded white and brown travertine including the one at Chylopittes (App. G, AF 34). They are of stone vases,52 but no examples of this mate the type used for Early Minoan rial come from the Chrysokamino excavations. Iron oxides and carbonates occur on the hill near the site, and they could have been used metallurgy as fluxes in but little of the smelting, original deposits survives today, and the extent of the original deposits is not known. On balance, the territory has little mineral wealth. The local phyllites and carbonates could be used as building materials, but the other deposits of useful stones are very small. They were suitable only for domestic tools and a few other items.

Copper

Ore

source of copper ore used at is particularly Chrysokamino impor tant in view of the date of FN-EM the III early metallurgy workshop. copper deposits exist in Crete,53 the modern geological survey Although The

no trace of any copper ore. territory has recorded that the rocks of the region probably never had any because have little nonferrous mineralization. copper Furthermore, they none of the modern residents of Kavousi who were questioned know of of the Chrysokamino It has been suggested

the presence of any copper in the vicinity. On the other hand, archaeologists who visited 1900 published

the region shortly after

statements:

very positive

On a headland about three miles east of this town are fragments of an ancient furnace, called figuratively by the peasants the "Golden some an and Furnace"; specimens of rock from adjacent cliff have shown on analysis a low percentage of copper, not sufficient to induce modern enterprise, toric inhabitants.54

49. The

most

the lithic materials Crete taken 1996.

inMinoan for Pseira;

study of complete used in eastern

50. Faure

times was

1912. 52. Compare Seager sources 53. For lists of copper

under

see Betancourt

1994

p. 48.

1966,

on

this

not

neglected

by the prehis

Branigan 1968, pp. 51-53; 1974,

51. Chlouveraki 1998.

and bibliography

but evidently

subject,

pp. 59-66; McGeehan-Liritzis pp. 386-387. 54. Hawes see

p. 38.

and Boyd

1996,

Hawes

1909,

42

CHAPTER

2

have repeated the suggestion that copper ore was found near Others, however, have rejected these claims. Branigan, Chrysokamino.55 who spent considerable time tracking down references to copper deposits, did not list the site among Cretan copper locations.56 Several writers

In view of the erosion

and the other geographical changes since the sea in Bronze Age, including level, the question of change a local copper source remains open, but the available evidence indicates that a nearby deposit is highly unlikely. Ore deposits that were exploited in the Bronze Age leave more than traces of ore. Ancient miners routinely a considerable

ore bits, the ore-bearing rock to remove the high-grade leaving behind waste rock in large amounts, aswell as broken and discarded stone crushed

tools.57 None

of this evidence

has been recorded from the Chrysokamino of an ore deposit, it is possible that the early archaeolo some stray ore discarded gists found pieces of by the smelting workshop. In the unlikely event that an ore deposit did once exist nearby, it must hillside.

Instead

have been very small. It did not fill the local needs, and foreign ore was as shown certainly brought in by ship, by the presence of copper ore on the small beach at Agriomandra and by the fact that the lead isotope pattern in the local slag matches Lavrion and Kythnos of inclusions (App. C). some distance from the ore is, in fact, common worldwide.58 at Smelting Furnace operations must adhere to schedules based on factors including the availability of labor and weather conditions that do not affect mining. no reason There for their schedules to coincide is, moreover, compelling with mining activities because mining output is based on other variables, such as intermittent discoveries of rich ores.

and

Forests

Other

Plants

the wild trees of ancient Greece in general in particular.59 They agree that timber played an impor tant role in the ancient world, though they differ in details of interpretation, especially in the extent of ancient Greek forest.60 The role of trees is nicely expressed by Rackham when he says that trees "are not part of the scenery

Several authors have discussed and those of Crete

are actors in the history; they play."61 In regard to the eastern part of the Gulf ofMirabello, the ancient vege can tation Several factors are relevant to only be partly reconstructed. Bronze Age conditions. The climate, discussed in the first section of this one of the most chapter, is important of these factors. The Mediterranean of the theater of human

sets the limits for the type of vegetation and the species that can grow and flourish under local conditions. The modern vegetation of the strip of coastal land near Chrysokamino also provides helpful information. It is mainly phrygana, which is a low climate

vegetation

often found

in Crete where

55. Fimmen 1921, pp. 17 and 120; Glotz 1923, p. 39; Lamb 1929, p. 5; Hutchinson 1962, p. 40; Faure 1966, p. 48.

1991; Gale

and Otta

and

1991.

way

58. Rothenberg 1972, pp. 208-210; Lechtman mann,

56. Branigan 1970a, p. 80, fig. 17; 1974, p. 60, fig. 1; 1988, p. 80, fig. 17. 57. D. Gale

the land has been overgrazed

Bach 1991; Hauptmann, and Maddin 1994, p. 4; O'Brien

1996. 59. See esp.Meiggs 1982; Perlin 1989; Rackham andMoody 1996;

Rackham

2001.

60. See

in

particular

the discussion

by Rackham (2001), who suggests the amount

of ancient

forest

overestimated. 61. Rackham

2001,

p. 5.

has been

NATURAL

THE

ENVIRONMENT

43

has been destroyed by fire. Very few trees grow on the taller vegetation remains are located. The trees that coastal strip where the archaeological are present are low shrubs of the type called maquis, species that would taller under favorable

grow

conditions

but remain

stunted

if browsed

by and other undershrubs also grow here. animals.62 Thyme was once agricultural terraces, however, show that the region in selected terraces found only post-Minoan farmland. Excavations remains, or other

goats Abandoned

suggesting

that many

of the terraces were

not built until after the Bronze

was used even though the survey indicates that this land (App. K), already during the Minoan period (App. G). How much forest was present? Pollen analysis and other forms of evidence suggest that, like some other parts of Greece, portions of Crete wet period following the retreat of probably had substantial forests in the

Age

the last glacier.63 The date at which these hypothetical forests disappeared from eastern Crete is not clear, however. Certainly, Crete was never com most of the endemic Cretan plants (species that pletely forested, because do not occur anywhere else) are not forest plants but species that grow on more open land.64At there was surely no forest in the Late Chrysokamino, Bronze Age because few wild animal bones are present in the archaeological record, and large numbers of sheep/goat bones indicate extensive grazing. No forest could grow on extremely eroded hills like Chalepa. The erosion that removed soil from this hill and deposited large amounts of terra rossa at Lakkos Ambeliou, which was earlier thanMiddle Minoan based on the of soil (App. G), presence of sherds on the surface of the accumulation indicates that the date of the erosion was earlier than the 2nd millen nium

b.c. This

erosion must

have

taken place

after the natural ground

cover had disappeared. It is likely that human

activities during the Bronze Age curtailed or cover local forest still in place before the Final Neolithic. any to be cleared for agriculture, and grazing would also have needed

eliminated Fields

inhibited natural plant growth. Grain impressions in the furnace chimneys show that agriculture was already prevalent at the end of the Early Bronze or Age (Chap. 12). In addition, activities like making quicklime for plaster65 the copper smelting operation itself could have required considerable fuel, and the cutting of trees for charcoal may have been a factor in destroying or reducing any remaining forests.66 By theMiddle Bronze Age, the forest was surely restricted to areas not used for farming. the presence of extensive local forests, however, wild an resource. Timber from the hills been would have important rising plants inland from the coast could have been exploited. Both maquis and larger trees would have grown in the gullies and ravines in all eras, including pe Even without

an agriculture. These plants would have been important source of charcoal and wood for fuel. Herbs, edible greens, and other useful wild plants would always have been available. riods of extensive

62. Rackham 63.

Zohary

2001, p. 8. and Orsan 1965; Turner

andGrieg 1975;Turner 1978. 64. Rackham 65. Betancourt 66. For as a result 1983; ham

the concept of

for an 2001,

pp. 7-8. 2004b.

2001,

metallurgy,

opposing pp. 33-34.

of deforestation seeWertime view,

see Rack

Forests

and other uncultivated

areas would

have provided other useful resources, in addition to the plants themselves. They would have sheltered wild animals, including hares and birds that could be hunted. The bone of a hare found in the faunal remains at the site (Chap. 11) can metallurgy come from hunting. Flowers were always available, and they only have would have supported beekeeping (discussed further inApp. I).

CHAPTER

44 Fish

and

Other

Marine

2

Resources

of the Aegean Sea must also be regarded as a resource. Al is not a coastal site in the traditional sense, itwas though Chrysokamino close enough to the sea to make some use of its potential. The harbor is within easy walking distance, and trails lead down to the shore near both the The

proximity

settlement

and the metallurgy

on

fishing at nearby Pseira of the sea as a food source consisted

location. Research

suggests that the Pseiran exploitation resources available near the coast?both fish only of coastal fishing.67 The have been and other marine life such as sea urchins and mollusks?would to the residents

at Chrysokamino

aswell,

and the archaeological a the exploitation of the sea. The marine environment, sea can coast and the shells abundant with shallow water, life, support rocky and small fish from here would have formed a useful adjunct to protein from other sources. Salt, an additional resource from the sea, would also available

record confirms

have been available.

67. Rose

1996.

PART II:THE METALLURGY WORKSHOP

CHAPTER

3

The

of

Excavation

Metallurgy

the

Workshop

byPhilip P. Betancourty James D. Muhly, EleniA. Armpis, Robert S. Powell, Elizabeth B. Shank, Evi Sikla, and Tanya Yangaki

After

in 1985, a formal several informal visits to Chrysokamino beginning to the site in 1994 accompanied from by representatives

visit was made the village Kophinakis,

Plans were made

to begin an excavation. Dimitris the mayor at that time, extended the full support of the local This included free access to the site (which is public land), the

of Kavousi.

government. offer of the Kavousi

as a storage and for use, without charge, new roadwork to access. Other study facility, and the promise of improve were a research was made for the and logistics arranged, plan project to summer. the following begin The goals for the project included a better understanding of the nature schoolhouse

location and of its relationship to the local terri of the workshop was seen as a part of a regional study that would elucidate its broader context. The nearby habi tation site, ca. 600 m away on a dolomite outcrop on the southwest slope of the hill of Chomatas, was excavated as an integral part of the Chryso and date of the metallurgy tory. From the beginning,

the excavation

kamino

Project.1 Even before excavation began, the extent of the metallurgy location was was covered with easily visible because its surface chimney fragments and a no soil was visible in such concentration that almost slag heavy (Fig. 3.1). Only a small amount of grass and a few other plants grew on this bare and

1. The third archaeological location in the territory, the burial Theriospelio, cave on the slope of the hill north of the metallurgy in this project.

site, was The

not

three

investigated locations were

all first inhabited in the Final Neolithic period,

and

they

were

all in use during

the Early Bronze Age and the opening phase of theMiddle Bronze Age. The habitation the Late

site continued Bronze

Age.

to be used

in

in a single location, with infertile spot. The slag heap was concentrated a at its sharp line edge where the slag stopped and the surrounding soil was on a promontory The location the Aegean, with a began. overlooking sea view of the good (Fig. 3.2). The position of the metallurgy site within the local landscape and its to the closest habitation site can be seen in the relationship in meters of shown 1.2. The 600 drawing Figure landscape two locations is easy to walk, with no major natural barriers on a cliff at the bottom of two places. The slag heap, found a coastal headland with between

an elevation

between

the

a hill, sits on

It lies on a small saddle

on the west

and phyllite on the east, with to the north and the south.

of limestone

outcrops the land sloping down both The surface bedrock, discussed

formation

of ca. 38 m.

topographic between the

of the headland.

in Chapter 2, plays a crucial role in the The western end of the promontory, directly

CHAPTER

48

3

Figure lurgy

3.1. The site before

surface

of the metal

excavation

Figure 3.2. The metallurgy .-**&?

..

during

site

excavation

above the sea, consists of Plattenkalk limestone, which is hard and durable. tectonic plate.2 The adjacent bedrock is a much It is part of the African tectonic plate that has been thrust softer phyllite, part of the European over the African

one. Because the limestone bedrock at the western tip of more resistant than the is the headland adjacent substantially harder and two is formations and because interface between the the especially phyllite, the limestone remains as an outcrop that rises higher subject toweathering, than the adjacent rock. The result is a small natural saddle that slopes more toward the north than the south (Fig. 3.2). Here, on this natural trough established that terminates in cliffs at its two ends, the early metallurgists their workshop. road stops well inland and uphill from the metallurgy site (Fig. 1.4). The promontory must be approached by descending a steep can also be see Figure 2.1). The headland slope (for the contour Unes, The

modern

2.

Papastamatiou

et al. 1959.

THE

EXCAVATION

OF

THE

METALLURGY

. ?s-

Figure 3.3. Sieving at the metallurgy workshop

WORKSHOP

49

* '^w^t^m

on the western side when the sea is calm; a steep path approached by boat near the a from small leads up ending landing place to the promontory, location where the dry sieve is shown in Figure 3.3. The steep hill between the metallurgy site and the modern road above has no agricultural terraces, and the absence of sherds suggests it has had little recent use aside from occasional

grazing. surface survey began in 1995 with mapping and geological studies, at the settlement site and it continued for two additional years. Excavations The

and the metallurgy workshop were conducted in 1996 and 1997. This vol ume describes the results of the survey and the work at the site.3 metallurgy The 1995 season lasted only two weeks. It included geological study, instrument survey, and mapping. This season was intended to be a prepa ration for the excavation scheduled for the next two seasons. No artifacts were installed near the location, Survey points metallurgical of the area. and a topographic grid was established to aid in the mapping Under the supervision of Lada Onyshkevych, this work was accomplished with aTopcon GTS 303 Electronic Total Station attached to a Gateway 2000 Handbook (486) laptop computer. were inserted at 2 m intervals on the the rectangular Using grid, stakes were collected.

border of the area to be excavated.

In the two seasons when

excavation was

conducted (1996 and 1997), string was attached to the stakes to create a se ries of 2 x 2 m grid-squares. A letter-plus-number system was used to facili was excavated tate recording. The site by grid-square (Fig. 3.4), and the slag a was a pile sampled in series of squares that created long trench north and a more in with the southwest south, quadrant (Fig. 3.5). complete coverage A large part of the slag pile was left unexcavated because something should always be preserved for future archaeological investigation. Because the boundaries of the site could be easily traced before the excavation began (since the ground was completely covered with slag and small pieces of industrial ceramics), measuring the slag heap s outer border 3. The site will

excavations be

published

at the habitation separately.

was

a simple matter. The deposit extended tance of 34 m from location E5015 N4989

from north to E5015

to south for a dis N5023,

and east to

CHAPTER

5o

3

Figure 3.4. James Muhly cleaning the upper part of the apsidal build ing, visible at the bottom of the level with

west

for a distance

averaging

between

grass

roots

10 and 15 m from location E5007

N4998 toE5019 N4998.The deposit had clearly eroded off the cliffs into overall area of slag and clay of about 375 to 380 m2. No architecture was visible

the sea at both the north and south sides. The

fragments consisted before excavation, and (aside from the fragments of industrial ceramics) only a few sherds were present. None of the sherds was easily identifiable as to date. Many of the ceramic pieces had awhite calcareous coating adhering to them, especially on their inner surfaces. Almost were under 5 cm in size. fragments Excavation

was

conducted

all of the slag and clay

in 1996 and 1997 under

the supervision supervisors included

of James D. Muhly and Philip P. Betancourt. Trench Eleni A. Armpis, Robert S. Powell, Elizabeth B. Shank, Evi Sikla, and teams of three workmen. Tanya Yangaki. The supervisors worked with Manolis

was foreman. Objects collected from the modern surface and the letter S. Excavated units with their grid-square

Hairetakis

were

numbered

were

numbered

sequentially beginning with the first pass.4 Unfortunately, this site did not yield a single artifact with any evidence that itwas in its use. All pottery consisted of fragments, all of which had original place of been scattered in antiquity. no stratigraphy was present, excavation was conducted in 10 cm Where was followed where it existed. All soil units. Natural stratigraphy deep except for water-sieving and material remaining

was dry-sieved, samples and other soil samples on the screens was washed with fresh water after

remove dust and facilitate visibility. No soil was dumped without sieving to wet. In 1996, all pottery, and sieving subsequent examination, both dry and faunal rounded stones, unusual stones, and slag remains, chimney fragments, were cm saved; similar collecting strategies fragments above the size of 2 was not saved. water Samples for employed in 1997 except that slag were as the collected directed other and purposes supervisors, by sieving collected from each level.Water-sieving with samples for water-sieving was conducted onsite (Fig. 3.3), using fresh water brought in by truck to a reservoir on the modern road uphill from the excavation. The water was were

4. The volume that from

of artifacts catalogues the find contexts, include

the reader

can ascertain

the same grid-square

what and

in this so came stratum.

THE

EXCAVATION

OF

THE

METALLURGY

WORKSHOP

51

Cliff

N A

Small Hut

5007E 5002N

5007E 4992N Figure excavated

3.5.

Plan at the

of the

grid-squares site.

metallurgy

E. Shank

to the site by channelled details irrigation pipe using gravity. Additional of recovery and recording are documented a on in handbook methodology file in the INSTAP Study Center for East Crete library.5 With the exception of the grid squares inwhich the apsidal building was discovered (discussed in Chapter 4), the same excavation method was repeated in every grid square, with similar results overall. After removal and bagging of any stone tools or sherds found on the surface, the first pass was excavated. It invariably yielded small slag fragments, pieces of furnace chimneys with black vitreous deposits on their interior surfaces, of phyllite, and a mixture of soil, ground-up slag, and tiny The first roots also contained from pass fragments. grasses and

angular pieces 5. Betancourt

1995.

chimney

CHAPTER

52

3

other plants. Infrequently, the workmen would find amarine shell, a piece of a clay vessel, or some other artifact. Except for the roots, the lower passes were similar. Some passes had no pottery except for the chimney pieces. was visible anywhere in the stratigraphy slag pile, except for the squares inwhich the apsidal structure was located. was concern The methodology rigorous enough to allow conclusions ing what materials were not present. As previously noted, all excavated

No

was

material

dry-sieved.

The

coarse

material

remaining

on

was

screen

the

in order to remove dust and reveal any green, blue, or ore was ever found this way. In no seasons red copper ore, but spite of two awater of extensive water-sieving using separation machine operated by with water

washed

not a single seed or fragment of charcoal was recov skilled paleobotanists, was ered. Nor any piece of partly consumed fuel found in either season of excavation.

acces the 1996 season of excavations, the study, conservation, During were in and of finds undertaken the Kavousi school sioning, cataloguing house. In 1997, work was transferred to the newly constructed INSTAP Study Center also

for East Crete

in Pacheia Ammos

where

the artifacts were

stored.

work was done at these study facilities during the course of the seasons. Trench excavation supervisors studied their own pottery under a brief report the the supervision of the directors, writing day following was after the washed and dried. They also selected excavation, pottery were at that time. Accessioned initial objects for accessioning objects Much

drawn, were

photographed,

accessioned.

and catalogued.

Accessioning

and

Subsequently,

cataloguing

strategy

additional was

objects

more

com

than at some sites. All waterworn

stones and all pottery sherds were plete a catalogued except for few samples retained for possible future destructive were acces of the analysis. Samples fragments of the furnace chimneys sioned. All other artifacts were accessioned. The registrar, Mary Betancourt, supervised the cataloguing operation. Further analytical work on finds from the metallurgy location was done at the following laboratories and institutions: 1. Isotrace Laboratory for Archaeology (Research Laboratory the History of Art), Oxford University, England 2. Mineralogical Laboratory, Geology Department, Temple University, Philadelphia, Pennsylvania 3. Archaeological Laboratory, Department

of Art History,

and

Temple

University, Philadelphia, Pennsylvania 4. Bartol Research Institute, Department of Physics, University Delaware, Newark, Delaware 5.Museum

Science Center for Archaeology (MASCA), Applied of Archaeology and of University Pennsylvania Museum

Pennsylvania Philadelphia, Anthropology, 6.Materials of Materials Characterization Facility, Department Science and Engineering, of Phila University Pennsylvania, delphia, Pennsylvania

of

THE

EXCAVATION

OF

THE

METALLURGY

WORKSHOP

53

7. INSTAP

Study Center for East Crete, Pacheia Ammos, Crete 8. Exhibit Museum, of Michigan, Ann Arbor, Michigan University 9. Department of Archaeology and Prehistory, University of Sheffield, England inArchaeology 10. Center forMaterials Research and Ethnology and the Center

for Geochemical

Analysis, Massachusetts Institute of Technology, Boston, Massachusetts 11. Laboratory for Archaeometry, NCSR "Democritos," Hagia Greece Paraskevi, Attica, 12.Martin Chemical Laboratory, Department of Chemistry, College, Davidson, North Carolina Research Laboratory, Department of Chemistry, New York College, Poughkeepsie,

Davidson 13. Amber

Vassar

The

Structure

Apsidal

byPhilip P. Betancourt

a series of 2 x 2 meter square trenches arranged project excavated from north to south on the long axis of the ancient slag pile in order to determine whether any features were preserved within or below the deposit of smelting debris (Fig. 3.5). Only a little vegetation was present on the

The

surface of the slag, but a few bushes grew in grid squares L 18, 18.Most of the excavated grid squares showed no change in most places, stratigraphy anywhere between the surface and bedrock. In

modern M

18, and N

the deposit consisted of small pieces of black, glassy slag mixed ments of industrial pottery containing holes pierced before firing, same phyllite found in nearby outcrops, angular pieces of the other artifacts. The only area that differed lay in the southwest

with

frag occasional and a few

quarter of the deposit where the bushes were growing. It later became apparent that the slag. plant roots had taken advantage of ancient lenses of soil within was an area A small building first recognized when of pale-colored soil

was

in Trenches

L 18, L 19, N 18, and N 19 (Figs. 3.4-4.4). contrasted with the darker color of the slag, began to the surface, at the level of the roots of the few plants in these grid squares. It was first visible at 37.90/37.96 masl.

discovered

pale soil, which be visible just below

The

growing When the area was

fully revealed by excavation, it became apparent that of three superimposed lenses, each one with an apsidal its long axis oriented north to south and the rounded end at

the soil consisted

shape with the north (Figs. 4.1, 4.2). The lenses of soil were were

floors within

a small structure whose

walls

of wood and other perishable materials. The building had three strata, visible as successive apsidal floors. The highest floor was covered with a small amount of slag, and additional slagwas found beneath constructed

the lowest floor. All

of the floors were

distinct, and the stratigraphically six grid squares. A few circular features were present at the margins of the lenses in the lower two strata, and they were most clearly visible at the south and at the east of the lowest floor. structure

could be traced within

of cylindrical cavities filled with dark, pulverized slag, had migrated down into the cavities from the sides and from above. The slag-filled cavities, apparently postholes, defined the outer margin of the apsidal area of soil. Eight postholes could be securely identified

They which

consisted

CHAPTER

56

4

Figure 4.1. Aerial view of the apsidal building; squaresM 18 and L 18 show in the

wC-:;:?

>3*v-#?g

floor.

N

the middle

floor;

photograph Kite photograph

show

other the

squares lower

by Jan Driessen

N20

M 20

L20

0

1m

Figure 4.2. Plan of the lower floor

THE

APSIDAL

STRUCTURE

57

Figure 4.3. The apsidal building

Figure

4.4.

Level

4 in

grid-square

N 19, looking north, showing the lower

floor

at the eastern

side of the

apsidal building structure was the more were clearly present originally. This (Fig. 4.5), but at site. After the lowest of discovered the built feature metallurgy only the three strata was exposed, the building was reburied to preserve it for future

generations.

was renewed three superimposed layers of soil show that the floor twice after the small structure was built (Fig. 4.6). At the north, the outer at the same place, indicating that edges of the superimposed lenses occurred at after the wall the north edge of the soil was built for the first phase of The

the structure, its position was not moved. At the south, however, the posts were moved for the middle floor. This floor was larger than the original one. The uppermost floor, found surface, was very just below the modern at not be measured the edges. The soil in all three floors disturbed. It could was

the same: a pale colored disintegrated phyllite typical of this part of Hill. The pale soil contrasted significantly with the darker color of the deposit of slag below and beside it.

Chomatas

is known of the upper floor because itwas just below the modern surface, and its soil was very disturbed. It appears to have been about the same size as the lower two floors. Its upper surface was irregular and poorly Little

CHAPTER

58

N

Sharp Edge

A

4

N20

of

Floor

M 20

L20

Probable

Soil Spread

Doorway

Outside

Doorway

P = Post

Hole

Figure 4.5. Apsidal building, floor,

1m was visible at ca. 37.90 to 37.96 masl (because of the soil preserved, but it color), presumably close to its original surface. The middle floor (Fig. 4.6), formed of the same pale-colored phyllitic soil as the other two floors, lay at an elevation of ca. 37.91/38.05 masl. At the east, the edge of the middle floor (visible in Fig. 4.1) extended slightly was found at the beyond the edge of the lower floor. A lens of burned soil an x area cm ca. 65 in size. oval 50 Two large stones north, covering irregular on the lowest floor one as well. The soil used resting projected above this to form the floor had many sherds mixed within it, and a few fragments found on the surface of the floor (marked on the drawing, formed complete vessels, and in some cases fragments of same were the vessel its soil, indicating both on top of the floor and within that the pottery was not used in this building. Possibly itwas used instead near the place where the soil originated. The southern part of the build was ing enlarged relative to the previous period of use, and the posts were in seated slightly different locations.

of pottery were Fig. 4.6). None

at ca. 37.75 to 37.94 masl, sat directly The lowest floor (Figs. 4.4,4.5,4.7), on top of the of and other smelting debris. Between pulverized slag deposit 37 and 45 cm of slag lay below this lowest floor, and bedrock was found

doorway

showing

the

location

lower

of the

THE

APSIDAL

STRUCTURE

59

N

A Bellows 189 . Sherd 83

Burned Area

Disturbed

mv*>

Sherd 102

*.

'

^

_,^

77

++^-Sherd

Bellows

2Q2

Cup 78

Stone Tool 139

38.04

2m

GRID SQUARE M 18

3815

GRID SQUARE M 19 -

_

^_?_ UPPER FLOOR

37.88

|-

38.10 above

sea

level

37.43

Figure 4.6. Plan of the apsidal build ings middle floor and cross section of the building

6o

CHAPTER

4

Pot Bellows

203 Probable Post Hole

Channel from Plant Root

Probable Post

Hole

Figure 4.7. Plan of the apsidal buildings lower floor the slag. The floor was packed and firm, with only two sherds lying upon it (fragments of pot bellows 190 and 203). A burned area was found at the north. It consisted of about 50 x 65 cm of red soil. Its northern side was 15 cm from the north edge of the lens of soil. Two large approximately stones (neither with any signs of working or use on the upper surface) and several smaller ones rested on this floor (Fig. 4.4, at left of photograph). below

shape of the lowest lens of soil can be best seen in the plan shown 4.7. The boundaries of the floor could be easily traced by the of the and the edges of the soil floor. Postholes were postholes configuration The

in Figure

were less distinct at the clearly visible at the south, east, and north. They west, either because they were never present or because of disturbance from roots. The soil here (at the west) was extremely disturbed from plant the roots of bushes still growing at the time of excavation, and the exact outer (western) margin of the lens of soil could not be traced. The south

modern

of a row of three postholes, with awider distance between a ones (forming doorway). The eastern and northern a sides of the building had scalloped appearance at the edge of the floor, and only a few postholes could be seen. The wall must have been made side consisted

the two western

of some perishable and perhaps flexible material that was posts. Postholes were about 15-20 cm in diameter.

attached

to the

THE

APSIDAL

STRUCTURE

6l

OBJECTS ASSOCIATED WITH THE BUILDING few fragments of vessels and bellows were found both on and within the two upper floors. The lowest floor was almost completely clean on its

A

surface, and itwas not removed by excavation. The middle floor, which was ca. 5-10 cm above the lowest one, had the most sherds on it and within it were on com not Most found the it. within of No soil, top (Fig. 4.6). objects vases were on any of the floors; plete everything consisted of casual sherds. The objects are discussed in the other chapters of this volume. They indicate a date in the EM III-MMIA period of eastern Crete, when East was

in use (see Chap. 5). The three phases of the structure are successive, but they are separated by too little time for any occur. It must be that no evidence change in pottery style to emphasized was found for the actual use of the bellows, shallow bowls, and other jars, Cretan White-on-Dark

Ware

ceramic objects in this building itself. The artifacts must be regarded as part of a secondary deposit. Most sherds came from within the soil, not on top of the floors, indicating that the objects must have been used elsewhere, near the location where the soil was excavated to form these presumably successive

floors.

THE ARCHITECTURE

eastern Crete

are uncommon

in Crete, a few structures from the island.1 They are more common in structures with curved than elsewhere. None of theMinoan

apsidal buildings Although with curved ends are known walls

because provide good parallels for the example from Chrysokamino all previously discovered apsidal buildings from Minoan Crete are built of stone or mudbrick. The is, at present, apsidal structure at Chrysokamino on the island. are more common much in Anatolia unique Apsidal plans and northern Greece.2 Although most northern are built Aegean examples on stone foundations, a few are also known.3 of wood buildings The use of perishable materials for buildings is also uncommon (or seldom recognized) inMinoan Crete, but the evidence from Chrysokamino is very clear. The presence of posts iswell attested, but no traces of clay,

plaster,

or

other

weatherproofing

materials

were

present.

Postholes

were

especially clear at the south and east of the lower floor where the remains were well preserved. no trace of any wood remained, the outline Although of the posts could be recognized because the soil from the floor had spread around the wood when the site was in use. The phyllitic soil was not very so it did not mobile, spread into the cavity as the loose slag did when the wooden

the site posts disappeared (perhaps from being pulled up when line of at A soil the of the floor abandoned). sharp pale-colored edge where it had been laid up against a wall or some other barrier could be

was 1. Giannouli

1995.

2. Hood 1986, pp. 41-42; Werner 1993. 3. See Hood 1986, p. 41; Renfrew 1986a,

p. 191.

traced easily, especially at the north. This some sharp edge showed that at divided interior from exterior the of the wall the floor; i.e., thing edge was built before the soil was out was in the interior. The soil spread only spread outside the structure at the south, where it had been scuffed outside

62

CHAPTER

4

at the doorway. In addition, the position of the sherds and other artifacts interior and exterior everywhere except indicated a sharp break between at the south, to confirm that this was not just a set of posts at the helping a lens of soil but an actual barrier of preventing anything inside from edge the interior was strewn casually being kicked out of the structure. Although with casual debris (Figs. 4.6,4.7), not a single sherd was immediately north of the edge of the soil. No

evidence

survives for a roof. Since the walls were made

of perishable it is likely that a roof would have been perishable aswell. How materials, ever, one could also reconstruct the walls as a simple barrier with no roof at all, or with a partial covering of branches providing some shade but no real shelter. The hearth area at the north suggests that probably nothing the place where the fire was built. architectural tradition that uses posts driven into the ground is called pile-based construction. Several conclusions regarding this practice at may be inferred from the surviving evidence. Most Chrysokamino likely the builders began by driving posts into the deposit of slag to form a U-shaped outline with the entrance at the south or a J-shaped outline with both covered

The

the south and west

was

then placed either against the posts, creating a solid barrier forming the wall. Finally, soil was to make the floor, and itwas spread out evenly placed inside the space to create a relatively flat surface extending as far as the posts and the wall attached to them. open.

Some material

or beside

the aspects of the small hut suggest northern connections: stone the of the absence foundations, megaron plan, pile-based construction, and the use of wood without mudbricks. Pile-based architecture, in fact, has Its technology is still been present in northern Greece since the Neolithic.4 Several

known

as a folk tradition

in the 21st century, and architectural studies of building practices inThessaly and Thrace demonstrate

modern

pile-based structures.5 Several aspects of the techniques used for this class of wooden the modern pile-based buildings help to explain the archaeological evidence

at

Chrysokamino.

use vertical posts with their bases set firmly into buildings the ground to form the core of the walls. The sturdy vertical timbers form the skeleton of the building. Flexible limbs of smaller diameter are often to the posts, with the limbs placed alternately inside attached horizontally Pile-based

and outside

the vertical

timbers

to create the walls. The

alternation

in the

a

scalloped wall that is placement of the flexible horizontal branches makes at the floor. This construction may be the explanation for the visible easily at the outer edge of the soil floor at Chrysokamino scalloped appearance of rigid vertical piles set into the ground (Figs. 4.5-4.7). A combination and flexible materials as

for the walls

is a combination

that has been regarded It iswell suited

resistant to the lateral forces of fierce winds.6

particularly to the windy conditions on the hill with the workshop. The framework provides the underlying structure, but a covering is necessary to make the building complete. The walls of modern Thessalian or mud to make a daub are regularly smeared with clay pile-based buildings at traces that of daub survive structure.7 No suggesting Chrysokamino, if the walls were covered, the builders used leather, fabric, or some other more windproof. perishable material to make the shelter

4. See

text, plans,

and bibliography

(p. 88) inHiller andNikolov 1988. 5. Efstratiou 1997; Eres 2003. 6. Rapoport 7. Efstratiou

1969, 1997.

pp.

112-113.

THE

APSIDAL

STRUCTURE

63

THE PURPOSE OF THE STRUCTURE location of the apsidal feature on top of the slag pile may seem unusual, but the explanation for its placement is easy to understand because if such a structure were required for the workshop,- itwould have been more difficult

The

around the elsewhere in the vicinity. The area immediately on of of is hard limestone bedrock very very compact, consisting slag pile the side nearest the sea (the west) and phyllite bedrock with very little soil cover on the inland side. Bedrock is near the surface, with little soil cover. to construct

as easier than Perhaps driving posts into the slagwould have been regarded cover was sufficient to drive the a where soil the piles deeply finding place to seat into them the enough ground properly. All the objects in the structure were fragmentary. They included pieces of clay cooking dishes, bowls, cups, and jars in small to large sizes as well as pot bellows fragments. Many of the sherds suggested the possibility of food preparation and consumption. The pot bellows fragments were the only evidence

a few activity aside from slag and furnace like those found outside the structure. No crucible

for metallurgical

chimney

fragments or other fragments, mold fragments, copper prills, copper waste, objects were the present. suggesting casting process Activities inside the small structure cannot have involved more than three or four persons because of the lack of space. The presence of a burned area on all three floors shows that a role in these activities and heating played that the fire continued to be used in the same place through all three stages of the structures history. A number of possible interpretations are suggested.

Option

1. The

Hut

Was

Used

for

Smelting

has published a suggestion that the lens of soil was used as a place for smelting and that the pot bellows were placed inside what he calls the "so-called postholes."8 Unfortunately, this idea is not supported by the

Watrous

postholes have dimensions of 15 to 20 cm, while have dimensions of 30 to 50 cm. The bellows were hot enough evidence. The

the bellows to have soil

fused to them, but neither fused soil nor any other evidence for intense heat was found at the edge of the floor. All the evidence from scientific analysis indicates that the smelting at Chrysokamino generated products that could only have been created inside a bowl furnace excavated into the ground (Apps. C and F), and no such bowl furnace is present in the lens of soil. And finally, since most of the pieces of bellows were found within the floor packing, they were brought to this location along with the soil, so they were used and broken elsewhere.

Option Metallurgical

2. The

Location

Was

Used

for

Some

Other

Activity

hearth area is only a slightly burned area, visible by its reddish color. It is not the base of a bowl furnace, but one might still consider whether a it was a location on which purpose was something with metallurgical heated. Smelting was only one step in ancient The metallurgy. smelting of

The

8.Watrous

2001,

p. 216.

copper under Bronze Age

conditions

resulted

in very impure metal. After

CHAPTER

64

4

the smelting, they had to purify the copper completed by remelting it in crucibles before they could pour the molten metal into molds to make useful objects. the metalworkers

to this theory, the According apsidal areawould not have been the floor a a of building, but lens of soil used to work on, and the posts driven into the slag would have supported something other than the walls of a build one cannot ing (perhaps a heat barrier?). Although completely rule out the was practiced inside a closed structure, several that possibility metallurgy points argue against the hypothesis of activity inside an actual building at inside a hut, space would have been with any metallurgical major problem activity, and the high heat would a a to hut made of perishable materials. have been If metallurgy danger a roof because were the activity, then the posts would not have supported this location.

If the lens of soil were

a

of the smoke and sparks that needed venting. In addition, a strong wind a role in all operations where the heat must be intense, so plays metallurgical a very choosing windy site and then placing the location for the activity out of the wind is somewhat illogical, and itwould prevent the best results. If the structure were used for metallurgy, one might envision the posts or framework construction, some type of either to pro supporting bracing or use of the bellows. tect the metallurgists perhaps in connection with the this hypothesis, the scallops in the edges of the soil might be regarded as the traces left by the edges of the pots for the pot bellows, which would be used in large numbers to force air into a portable, freestanding furnace or hearth or brazier that would be over the burned area. A very placed

With

serious problem with this reconstruction, however, is that the pots used as seem set to into bellows have been the ground surface and seated with mud, which fused to them during use, and no traces of this fused mud survive on the floors where

this theory. In addition, on the soil surface, pieces of them placed were not would have been high up in the slag pile, and they present. In addition, no evidence survives at Chrysokamino for a casting or a single crucible fragment comes from the site, remelting operation. Not and the slag is all smelting slag, rather than the residue from remelting for they would if the furnaces or crucibles were

have been with

of implements (one obvious difference between smelting is and from the higher iron content of the former). slag remelting slag Little physical evidence supports this possibility. the manufacture

3. The

Option

Hut

Was

a Dwelling

had to live somewhere, but the character of this structure The metallurgists tasks could not have been practiced inside it. indicates that all household The hearth takes up so much room that very few people could be inside at once, even sitting down. Insufficient space is present for storage or for structure was not very weatherproof, and the It is that likely sleeping. a were in winter. If the hut itwould have been cold and uncomfortable dwelling, place

one would

have

to assume

that many

domestic

activities

took

outdoors.

as awhole is not the range of artifacts from the workshop to represent a habitation site. No ground stone querns enough

In addition, complete

APSIDAL

THE

STRUCTURE

65

nor stone nor loomweights spindle whorls no cook vases. Pottery is specialized (mainly dishes, cups, and jars, with no almost the and few luxury items, pouring vessels). Among ing pots, for the from the Ware White-on-Dark sherds, corpus example, workshop includes no fragments of the decorated jugs and teapots that are usually are present. The

site has neither

found at settlements

from this period.9 Even food remains (animal bones, marine other and remains) are not numerous shells, enough to suggest either permanent or seasonal settlement at the location.

Option

4. The

Hut

Was

a Kitchen

for cooking is another possible the small hut at the metallurgy workshop.

A

location

explanation for the presence of need not have The windbreak

to because it would have been designed completely weatherproof a on too from small fire this prevent cooking burning rapidly windswept of the finds (domestic pottery hill. It need not have been roofed. Many animal bones, marine shells, stone [including cooking dishes],obsidian,

been

tools of small size, and cups and other serving vessels) suggest that at least some food was in the immediate vicinity. If the prepared and consumed lens of soil were

the floor of a small building, itmay have been a kitchen even if the metalworkers returned to their

a noon meal was cooked, homes at night.

where

Unanswered questions, however, also exist with the hypothesis that the structure was used as a kitchen. The nature of the hearth area (a flat space dis colored by fire) indicates that fires were built over it, not on it; if a fire were made here, the remains were either carefully cleaned away after use or the fire inside a portable hearth. The only cooking vessel sherds from the site are from a shape called a cooking dish, a large, shallow vessel whose thin walls indicate itwas used in place over a fire and not moved. Such a vessel would

was

have been supported by stones or by the edges of a portable hearth, prevent ing it from tipping over. Fires would have been built around it.However, this configuration was not preserved here, although itmay present. The main Minoan cooking vessel of EM III-MM a vessel that is common at the Chrysokamino cooking pot, half a kilometer away, but does not occur here. The absence of

once have been IA is the tripod habitation site

broken cooking no was of that indicates vessel broken course, here, offering pots, only tripod no

evidence

as to whether

or not

the

shape

was

used

nearby.

A roof over a cooking area is not desirable because of the smoke that would rise from the hearth. With the building used as a kitchen, one would that the expect posts supported only a screen that was partly roofed to a little shade. The structure would have been a temporary shelter provide rather than a full building.

Option

5. The

of the contents results

1984.

Was

an Apothecary

of clay vases from the workshop reveal some were of the vessels Several interesting (App. M). sufficiently well to yield evidence of their contents. The residues in some of the preserved vases yield evidence for spices and herbs, including saffron, camphor, and

Analyses

9. Betancourt

Hut

66

CHAPTER

4

rue.

some of these herbs may have been used for Although flavoring food, uses. The absence of normal would also have had medicinal they cooking residues such as plant oils or animal fats in the vessels containing these raises the possibility that some of the mixtures had a specialized use. Some of the herbs have been as traditionally employed analgesics for skin problems such as burns. One should also note that arsenic poisoning causes skin lesions that would have needed treatment. The configuration of herbs

area, suggesting use of a portable hearth rather than a cooking not does contradict this hypothesis. However, because the sherds were fire, found mostly within the floors and not on them, we cannot conclude that the hearth

anything

found

in the structure was necessarily

on

Conclusions

the

Hut's

used there.

Use

all the evidence, a metallurgical process other than smelt or the are the most of medicines ing, simple cooking, preparation likely on at the hut. Perhaps the most plausible interpretation activities carried In considering

for the apsidal structure is that itwas a small kitchen associated with the in this isolated location may have workshop. The metallurgists working in the evening, even wanted a cooked meal at noon or perhaps occasionally if they returned to their homes at night, because they could not interrupt to prevent the cooking the smelting process once it had begun. A windbreak fire from burning too quickly would have been absolutely necessary due to the windy conditions on this bare hillside, and a small structure (probably roofed) would have provided for the needs of the cook. The that herbal medicines were also prepared in such a structure possibility cannot be dismissed, particularly given the specialized nature of the pottery from the site and the absence of sherds from tripod cooking pots, which are only partly

common

in the EM

III-MM

IA corpus at the nearby habitation

site.

CHAPTER

5

Pottery

The

byPhilip P. Betancourt

few sherds come from the metallurgy workshop. Their dates Relatively (FN) to EM III-MM IA, with the largest range from Final Neolithic number of pieces coming from the final period. Relatively few shapes are a represented, and the pottery is clearly specialized assemblage. Even the pottery from the final phase lacks several crucial shapes, including the tri are present. The sherds come from pod cooking pot. No complete vessels the slag, from the soil used for the floor of the apsidal structure, and from the ancient ground surface that existed when the apsidal structure was standing. scattered

locations within

THE POTTERY PERIODS In eastern Crete, EM III-MM IA cannot be divided stylistically into two of Most the sherds from before this phase come from the slag pile. periods. one sherd, the FN are all from the some of Except for pieces slag pile, and them come from deep within the deposit (though not at its base). All of the I?II fragments are also from the slag deposit. This pattern is extremely important for the dating of the workshop, because it clearly shows that the was not derived from an earlier settlement on the early pottery existing was was before the Nor it in founded. with promontory workshop brought the soil for the floor of the apsidal building, because it occurs within the

EM

35-45

cm

deep slag deposit found below the level of the small hut. The FN and EM I?II sherds are definitely a part of the slag heap that existed IA phase of the workshop, before the EM III-MM and the foundation of the site must

predate the EM III apsidal structure. little pottery is present from any of the periods before the metal Very lurgists decided to build the apsidal structure. Only 14 sherds, representing 10-14 vessels, predate EM III-MM IA.This assemblage is a limited corpus, but the fact that several different shapes are represented, including a pyxis (12) and at least one probable storage vessel (14), indicates that the small number of sherds is not necessarily ameasure of the level of activity. Several contrasts can be noted between the pre-EM III pottery and the ceramics

from the final phase

of the site. First of all, the number

of

68

CHAPTER

5

sherds ismuch

larger in the later period. Secondly, the assemblage includes several new classes of pottery, including cooking vessels, more open bowls, and more storage vessels. The EM III-MM IA pottery represents a sub in the activities

stantial change

carried on at the site. The workers

now

spent enough time to build a small apsidal structure that was used long enough for the floor to be renewed twice after its initial construction. The new ceramic classes (especially the cooking vessels and the storage shapes) demonstrate that the preparation of food, cooking, storage, and probably increased food consumption

were

presence

in the vicinity. taking place somewhere of the likely represents expansion activity. The metallurgical as of pot bellows sherds in the corpus, in Chapter discussed 8,

supports

the

This most

an

same

conclusion.

pottery from the site can best be studied by dividing it into two the pottery from contexts, the slag pile and the apsidal structure. When one can see some clear each of these contexts is examined individually, differences both in date and in character between the two assemblages. The

These

differences

are very important

for the reconstruction

of the history

of the small workshop.

THE POTTERY FROM THE SLAG PILE the deposit of slag was reached in all of the the slag pile except for those that were directly under the apsidal building. In all trenches, the deposit of slag and furnace on sterile soil or bedrock. No traces of chimney fragments rested directly were earlier settlement encountered any anywhere within the trenches that

The

bedrock

trenches

were

surface below

excavated within

excavated. The

sterile stratum consisted mostly of very underlying highly calcareous phyllite. The soil, which ranged in

pale, disintegrated, character from hard and compact to soft and unconsolidated, in character to the local soil found adjacent to the slag pile. The character of the slag pile did not vary significantly

was

similar

between

the

surface and its lower limit, except within the trenches that revealed the area structure. the it consisted of small apsidal investigated, Throughout amounts small of bits of furnace of soil, and a chimneys, slag, clay pieces few stones. Only a few sherds of vessels were discovered, and some of the excavated levels (ca. 10 cm deep) yielded no fragments of vessels at all. The present publication presents all of the sherds noted by the excava tors except for a few scraps kept separate for potential in future generations. The context of each catalogued the other descriptive is evident. Final Neolithic

with

destructive analysis sherd is given along and the lack of clear stratification

information, sherds come from the surface as well

as from

sherds come from the sterile deep within the slag pile. No Final Neolithic a concentration of several early sherds in surface below the slag, although the middle levels inTrenches N 20 and O 20, below the level of the earliest floor in the apsidal structure, indicates a generally early level in this central part of the slag pile (Fig. 5.1). The scattered and mixed nature of the pot the deposit, shows its dis tery, along with the lack of stratigraphy within turbed nature. Perhaps

the metallurgists

disturbed

the deposit

by digging

THE

POTTERY

69

CLIFF

Level 3: 1 sherd (no. 1 )

i

Surface

1 sherd

cleaning:

Level 2: 2 sherds

(no. 73)

(nos. 8, 10)

Level 2: 1 sherd (no. 9)

Level 3 (4 sherds

Level 2: 1 sherd Level 1 (Surface):

Surface:

SLAG

Figure 5.1. Location of Neolithic dark-burnished sherds within the

(nos. 3, 5, 6, 7)

(no. 4)

1 sherd (no. 2)

1 sherd (no. 11)

slag pile

into it or moving the construction

the material

(presumably when of the small building and when

the area was its posts were

leveled for seated

in

the ground). The pottery deposits from the slag pile and the small apsidal structure are organized in separate tables (Tables 5.1-5.5 and 5.6-5.8; see Catalogue below) in order to show their character, following a system first used at Kommos.1 The tables provide a clear picture of the deposits by recording each of the sherds by ceramic class, vase shape, and position on the vessel, with an additional column at the right of each table giving the minimum number of vases represented by the surviving fragments. If a group of sherds is restorable as a vessel on paper, it is recorded as a single piece in the column called "whole" (no complete vases were found). The list with each table provides references to the catalogue in this chapter where the sherds are described more fully.

The

Final

Neolithic

Pottery

is all made from a single fabric (see pottery at the workshop a to is It with brown dark brown color and a heavily bur coarse, App. N). are nished surface. Inclusions, which abundant, include quartz in several both white and dark mica, and sandstone. These forms, phyllite containing The

1. Kommos

II; Kommos

III.

FN

inclusions

are not distinctive

enough

to suggest

a

place of manufacture,

CHAPTER

7?

5

in eastern Crete. The but they do not preclude a local source somewhere fabric is substantially different from the main fabric used for the EM III MM IA ceramics at the site (which is tempered with in the inclusions to diorite series), and it can be easily distinguished. granodiorite vessels have thick walls with a uniform The Final Neolithic

thickness.

are sturdy and well made using a few They hand-forming techniques. Only are present: bowls, a a and vessel with jars, pouring shapes globular body and are present, a horn on the one suggesting that preserved handle. No cups local drinking vessels in this period may have been made of gourds or wood. Coarse, dark-surfaced, burnished pottery is the earliest ceramics from this part of Crete. It occurs atmost of the sites in the region, and published finds come from Vasiliki,2 Mochlos,3 Sphoungaras,4 Pseira,5 Kalo Chorio,6 some of the unexcavated the Vrokastro Hagios Antonios,7 region,8 and from comes from survey.9The pottery also in eastern Crete.10 The chronology of this pottery still needs elsewhere more work. It definitely in the Final Neolithic (if not earlier), and begins in the Kavousi-Thriphti

sites recorded

at some sites, such as Pseira, the same fabrics persist into EM I, although EM I fabrics occur at and pattern-burnished the harder, well-burnished Mochlos

and elsewhere.11

consist of shapes in the region that includes Chrysokamino are Handles often wide and vessels without spouts. pouring jars, bowls, is extremely rare. Horned vessels straps or small, pierced lugs. Decoration like the jarwith the horned handle from Chrysokamino (6) are not known in this part of Crete from any of the published EM I sites, suggesting that Neolithic

the earliest pottery from the metallurgy ning of the Bronze Age.

The

EM

I-IIA

is earlier than the begin

Pottery

in the eastern Gulf

I fabrics

EM

location

of Mirabello

thinner and harder than in the Final Neolithic

region

were

sometimes

were period.12 New shapes vessels with spouts, pouring

manufactured during the period, including chalices on conical bases, and a series of pyxides of globular to ovoid as pattern burnishing, made their shape. New decorative techniques, such as well.13

appearance

over I period. long distances in the Early Minoan Pottery was traded vases IWare Imports into eastern Crete included Hagios Onouphrios from

south-central

2. Seager 3. Seager Davaras

1912, pp.

1912,

5. Betancourt pp. 29-30, 134. 6.

Haggis 7. Haggis

tomb

p. 212. p. 93; Soles and 179-180.

pp. 46-48. and Davaras 2; Pseira

Ware

from

from north-central

1904-1905,

1996,

4. Hall

Lebena

Crete,14

region,15 Pyrgos Ware

VII,

1990, pp.

133

in

somewhere

Crete,16

the

same

and several Cycladic

8.Hayden 2003a, 2003b. 9.Haggis 1992,1996b. 10. SeeManteli (1992) for Kastelli Phournis; Schlager (2002) forHagia Triada and Petrokorio (p. 181) and for the Katharades Acropolis (p. 196). 11. On

the chronology in Crete situation

1996c.

on

1993a.

this time, seeHood

the

of EM in general

1990a, 1990b;

I, and at

Wilson 1996c. Mirabello 12.

andDay 2000, p. 53;Haggis For

additional region,

Compare

sos inHood

discussion

of the

see Betancourt the

situation

1999. at Knos

1990a.

13. Betancourt

1985,

pp. 23-34. 1997.

15. Betancourt

1985,

16. Betancourt

1985,

pp. 31-32. pp. 26-29.

14.Whitelawetal.

THE

POTTERY

71

possible imports from outside of the immedi shapes.17 At Chrysokamino, on the ate region include a pyxis (12) and a vase with basket impression base (14); the pyxis is from EM I-IIA, while the vessel with the basket more impression cannot be dated closely than EM I?III. some other Although pieces might be from this period, the only certain EM

sherd from the metallurgy workshop is a tiny fragment of a pyxis in the slag pile (12). It has a combed decoration incised into the

I-IIA

found

are as EM ILA, but the usually regarded chronology is not absolutely certain. Good parallels come from central Crete,18 and a similar pyxis comes from a tomb at Gournia.19

exterior. Such vessels

The

EM IIB Pottery

One

sherd from a Vasiliki Ware

The

EM

goblet is present from the slag pile (13). Vasiliki Ware, the mottled pottery of EM IIB, is one of the most distinc tive and easily recognized wares ofMinoan Crete.20 The gray fabric of the sherd from this location is typical of the ware in this part of Crete.21 The surface is slipped and burnished in the usual fashion.22

III-MM

White-on-Dark

IA Pottery

is the definitive pottery for EM III-MM IA in It begins before the end of EM IIB,24 and it persists into that is contemporary with the start ofMM IA in central Crete.25 Ware

eastern Crete.23 a

period The ware

it is easily recognizable used because the MM

a by linear white decorations on dark background; both by its style and by the type of white paint IB white slip in this part of Crete ismuch whiter

than the EM

IA variety*26 Both

is characterized

III-MM

shapes and decorations

are very

distinctive. Most of the pottery from the slag pile is from EM III-MM IA. Both fine fabrics and coarse fabrics are present in a number of typical shapes. The fine fabrics are the pale, slightly of the Mirabello pastes "gritty" clay Fabric

that is typical of this phase in this part of Crete. The typical coarse fabric from this phase is red to yellowish brown. It varies in coarseness from vessel to vessel. Included fragments of stone, in the granodiorite/diorite

the definition for Mirabello Fabric.27 A series, place both fabrics within useful p?trographie is of this fabric in connection description published with examples found at Pseira.28 Mirabello Fabric is also the main local fabric for this region

17. Davaras

1971.

18.Xanthoudides 1924, pp. 34-36; Wilson 1985, p. 305. 19. Soles 1992, p. 32, fig. 13, no. GIII-4,

pi.

13:G.

21. Myer 1981.

court

1972;

1979. 1979; Myer

22. Schachermeyr (1938) collected another surface

and Betan

piece of the

of Vasiliki Ware metallurgy

from

the

workshop.

23. Zois 1968; Betancourt 1984; 1985,

20. See period 2 inWarren Betancourt

in later phases

pp. 55-61.

24. See period 2 inWarren 1972. 25.Momigliano 1991, p. 227; 2000, p. 102; Betancourt 2003b. 26. Noll, Holm, and Born 1971,

of theMiddle

Minoan

period.29

p. 617; Noll 1982, pp. 190-193; Fer rence,

Swann,

27. Myer

and Betancourt 1979; Myer

?002.

and Betan

court 1981;Day 1991, pp. 91-101; 1997, p. 223. 28. court

Myer, 1995.

Mclntosh,

29. Haggis andMook 2000.

and Betan

1993; Haggis

CHAPTER

72

5

THE POTTERY FROM THE APSIDAL STRUCTURE pottery from the apsidal structure comes from a level just below the modern surface. It comes only from inside the building and from the small scatter of soil in front of the are doorway. Only two periods represented.

The

One

sherd

from EM

is from III-MM

the Final Neolithic, and the rest of the pottery is date is certain because it is based on several

IA. The

well-dated

Ware pieces, including sherds of East Cretan White-on-Dark ware at this end of EM its the latest date 5.2). IIB,30 (Fig. begins Although is contemporary with MM IA in central Crete.31 The EM III andMM IA

parts of the phase cannot yet be subdivided on stylistic criteria. The pottery from the apsidal structure (Tables 5.6-5.8) is presented in tables similar to those used for the slag pile. These tables allow the two contexts

Final

to be compared

and contrasted

easily.

Neolithic

only pottery fragment from the apsidal structure predating EM III is a a body sherd from bowl of FN date (73). Its isolation within a substantially later deposit suggests it is an heirloom.

The

Early

Minoan

III-Middle

Minoan

IA

of 32 vessels. pottery from the apsidal structure represents aminimum There aremany ambiguous pieces whose original shape is uncertain because the sherds are too small for identification. These ceramics are presented in The

their entirety in the catalogue and charts in this chapter. Where the evi is sufficient for judgment, small sherds have been combined to form sherds; in the cases where joining has not incomplete vessels of nonjoining been possible, sherds are catalogued individually. The fact that many of the sherds come from within the soil of the dence

in the building proves that the soil, with most or all of its included sherds, was brought into the workshop from a settlement, so that the sherds catalogued here need not have been used originally within the apsidal struc ture itself. On the other hand, the date, style, and shapes of the EM III MM IA pottery from inside the apsidal structure and from outside it in the slag pile are so similar that the activities must be reconstructed from both assemblages. This similarity can be seen easily by a comparison be tween the lists of shapes from the two contexts (Table 5.9; see Catalogue IA contexts below). These similarities indicate that the two EM III-MM are similar parts of a IA people single range of pottery. The EM III-MM were area not the entire of the it is and using workshop, surprising that their pottery is spread over the entire area. The fragments of cups, pouring vessels, shallow bowls, storage jars with wide mouths, and cooking dishes observed in the sample attest to floors

food preparation activities and the consumption of food and drink. Not a a as a sherd from such that single luxury vessel, pyxis might have contained

30. See period 2 inWarren 1972. 31. Betancourt 1984; 2003b, p. 119; MacGillivray

et al. 1992.

THE

POTTERY

73

or oil, or from any vase that could be called a ritual vessel, is pres perfume ent. A few painted cups and pouring vessels are represented, but the truly as teapots, are not found. elegant painted shapes from the period, such Also missing are transport vessels with narrow mouths (such as amphoras), and all other vessels from aside cooking cooking dishes. tripod cooking pots, be that this is a limited

a small, assemblage from specialized site without many of the vessels that would be commonplace within the pottery assemblage of a larger and more diverse settlement. IA pottery, like the earlier ceramics, ismostly lo The EM III-MM The

must

conclusion

cal. The main

fabric is theMirabello

Fabric of the Gournia

to Priniatikos

Pyrgos region, and the style of the painted vases is typical of this part of the use of an apsidal structure and the possible importa Crete. Although a ore tion of from region north of Crete may suggest foreign connections, was the local pottery demonstrates that the community at Chrysokamino was to ceramic it the local This tied that system. firmly production implies as with of the other local cultural well. aspects system integrated

CATALOGUE In the following catalogue, the vessels are arranged by context. Finds from the slag pile are listed first, followed by those from the apsidal structure. in this and subsequent catalogues are in centimeters; heights Dimensions include the handles; the maximum axis across the sherd. Nomenclature Minoan

Pottery?2 Clay from

Pottery Final 1

is the length of the longest dimension follows the conventions of The History of colors are described using theMunsell system.33

the

Slag

Pile

Neolithic

(X 410, from R 20-3) Bowl,

rim

Diam.

of rim

Fig.

5.3

sherd. ca. 24.

A coarse fabric (black,with an unevenly colored surface,mostly dark reddish brown,

5YR

Open,

2

3/3);

burnished

shallow

bowl.

on

(X 1860, from K 17-Surface) Bowl,

rim

Diam.

of rim

interior

Fig.

and

exterior.

5.3

sherd. ca. 10-11.

A coarse fabric (unevenly fired, black, with the surfacemostly reddish brown, 5YR 5/4); burnished on interior and exterior. 3

(X 694, from N 20-3)

Fig.

5.3

Bowl, body sherd. 32. Betancourt 33. Kollmorgen ration 1992.

Max.

1985. Instruments

Corpo

dim.

4.8.

A coarse fabric (unevenly fired, with 3/2); burnished on interior and exterior.

the surface mostly dark brown, 7.5YR

CHAPTER

74

5

> ) J

7

/

/

Xr 6,7

\

I

/

12

11

10

)

7

?i

\\ 18

19

n

20

16

15

14

13

21

?

n

^f 22

17

23

the slag pile Figure 5.3. Pottery from 1:3 Scale (1-23).

THE

4

POTTERY

75

(X 671, from L 17-2)

5.3

Fig.

Bowl, body sherd. dim.

Max.

3.6.

A coarse fabric (unevenly fired, black, with the surfacemostly reddish brown, 5YR 5/4); burnished on interior and exterior. 5

(X 718, from N 20-3) Carinated

vessel,

open of

Diam.

ca.

body

5.3

Fig.

15-16

sherd. body at the carination.

A coarse fabric (unevenly colored, mostly dark reddish brown, 2.5YR 3/4, with a reddish brown surface); burnished on interior and exterior. 6

(X 421, from P 20-3)

Fig. 5.3 handle with sherd horned rim. Jar,

Restored

ca. 25.

H.

A coarse fabric (black,with an unevenly colored surface,mostly reddish brown, 5YR 5/4); burnished on exterior. Parallels: For FN closed vessels with horns on top of the handles, see 1992, fig. 8 Vagnetti 1975, p. 85, fig. 75, especially no. 8 (Phaistos); Manteli (Kastelli

Horned

Phournis).

and Bankoff

example, Stefanovich Cuka). 7

handles

(X 717, from N 20-3) of

interior

of Balkan

1998, p. 299, fig. 26:a-d,

ceramics

for

(see,

from Kamenska

5.3

Fig. neck sherd with handle. Jar,

Diam.

are a feature

ca. 8.

of neck

A coarse fabric (unevenly colored, mostly dark gray to black, with an unevenly colored surface, mostly light reddish brown, 5YR 5/4); burnished. Wide strap handle with small hole pierced in it,made before firing. 8

(X 652, from O 20-2) Jar, handle dim.

Max.

5.3

Fig.

sherd. 5.4.

A coarse fabric (unevenly colored, with a reddish brown surface, mostly reddish brown, 5YR 5/4); burnished. Wide strap handle from ajar. 9

(X 5, from O 22-2) Jar, handle dim.

Max.

5.3

Fig.

sherd. 5.4.

A coarse fabric (unevenly colored with a reddish brown surface, mostly reddish brown, 5YR 6/3); burnished. strap handle

Wide

10

from

(X 621, from O 20-2)

A

of

coarse

body fabric

light

ajar.

Fig. Closed vessel, body sherd. Diam.

light

5.3

ca. 22. (unevenly

colored,

with

dark reddish brown, 5YR 3/2); burnished. Comments: Probably a jar.

an

unevenly

colored

surface,

mostly

CHAPTER

76 11

(X 941, from K 21-Surface) Closed

vessel, dim.

Max.

5

Fig. 5.3

sherd.

body

3.1.

A coarse fabric (unevenly colored, mostly black, with the surface mostly dark gray, 5YR 4/1).

TABLE 5.1. FINAL NEOLITHIC SLAG PILE Handle

Rim 1.

Base

Spout

SHERDS FROM THE

Leg

Body

Whole

3.

1

4. 5.

3

Min. No.

4

2 Ill

2

2.

Total

1

11 3 2

2

2 2

Coarse fabric sherds: (1) bowl (1-4, X 410, X 1860,X 694,X 671); (2) carinatedbowl (5,X 718); (3) jarwith horned rim (6,X 421); (4) jar (7-9, X 717, X 652,X 5); (5) closed vessel (10,11, X 621, X 941).

EM 12

I-IIA (X 460, from N 17-2)

Fig.

5.3

Pyxis, body sherd. Max.

dim.

2.4.

A fine fabric (brown, 7.5YR 5/4). Globular

pyxis

with

decoration

combed

incised

into

the

exterior.

Comments: Small globular pyxides with stamped and incised decoration begin in EM I and continue into EM IIA. Parallels: Xanthoudides 1985, p. 305 1924, pp. 34-36 (Koumasa);Wilson (Knossos); Soles 1992, fig. GIII-4,

pi. 13:G (Gournia).

TABLE 5.2. EARLY MINOAN SLAG PILE Rim

Base

Handle

Spout

I-IIA SHERD FROM THE

Leg

Body

Whole

Total

Min. No.

Ill

1. Fine fabric sherd: (1) pyxis (12,X 460).

EM IIB 13

(X 750, from N 20-4)

Fig.

5.3

Goblet, rim sherd. Restored

Diam.

ca. 8-10.

A fine fabric (pinkish white, 5YR 8/2); slipped and burnished. Goblet with slightly inturned rim; Vasiliki Ware. Comments: The small goblet is one of the most common shapes in Vasiliki Ware (Betancourt 1979). It is distributed throughout Crete, but the majority of the

examples

are from

the

eastern

Parallels: See Betancourt

end

of the Gulf

1979, pp. 42-45,

of Mirabello

region.

for a list of other examples.

THE

POTTERY

77

TABLE 5.3. EARLY MINOAN SLAG PILE Handle

Rim

Base

Spout

IIB SHERD FROM THE

Leg

Whole

Body

Total

1.1

Min. No.

11

Fine fabric sherd: (1) goblet, VasilikiWare (13,X 750).

EM 14

II-III (X 497, from P 20-3)

Fig.

5.3

Jar (?), base sherd. Max.

10.6.

dim.

A coarse fabric (yellowish red, 5YR 5/8, to very dark gray, 5YR 3/1). Base

of a closed

basket

vessel;

impression

TABLE 5.4. EARLY MINOAN SLAG PILE Rim

Handle

Base

Spout

on bottom

of base.

II-III SHERD FROM THE

Leg

Whole

Body

Total

1.1

11 coarse

Very

EM 15

Min. No.

fabric

sherd:

III-MM

(1) closed

IA

(X 459, from N 17-2) Rounded Max.

cup,

dim.

(14, X 497).

vessel

rim

5.3

Fig.

sherd.

2.6.

A fine fabric (light red, 10R 6/6). Comments: White-on-Dark

Most

of

Ware,

but

the

published examples several from examples

of

this

shape

Chrysokamino

only a band on the rim. For the shape, see the list in Betancourt from

16

eastern

are

plain,

as with

1984, p. 43, all

Crete.

(X 668, from L 17-2) Rounded

cup, of rim

Diam.

are decorated

rim

Fig.

5.3

sherd.

ca. 9-11.

A fine fabric (light reddish brown, 5YR 6/4). Dark slip: band on rim. 17

(X 698, from L 17-2) Rounded Max.

dim.

Fig. 5.3

cup, body sherd. 2.6.

A fine fabric (light red, 10R 6/6). 18

(X 796, fromM 22-4) Rounded Max.

dim.

cup,

base

sherd.

2.0.

A fine fabric (pink, 7.5YR 7/4).

Fig. 5.3

CHAPTER

78 19

(X 711, from N 20-2)

5

5.3

Fig.

Cup (?), body sherd. Max.

dim.

Convex

2.

profile.

A fine fabric (reddish yellow, 5YR 6/6). Dark not

slip on exterior. Added white

preserved.

(X 719, from N 20-3)

20

5.3

Fig.

Cup (?), body sherd. Max.

dim.

1.7.

Convex profile. A fine fabric (reddish yellow, 5YR 7/6). Dark slip on exterior. Added white not preserved. (X 1223, fromM 20-2)

21

rim

Cup, Diam.

5.3

Fig.

sherd.

of rim

8.

rim.

Straight

A fine fabric (between pink, 5YR 7/4, and reddish yellow, 7.5YR 7/6). Dark slip on inside of rim and on exterior. Added white: bands with zigzag band at rim.

22

(X 3, from M 22-2) Shallow Diam.

rim

bowl,

Fig.

5.3

sherd.

of rim 56.

coarse fabric (reddish yellow, 7.5YR 6/6). rim.

A medium Thickened Comments:

a list of

For

such

open

bowls,

all from

eastern

see Betan

Crete,

court 1984, pp. 38-39. Complete examples are illustrated by Hall 1912, fig. 20 and Betancourt 1983, nos. 114,115 (all from Sphoungaras). Both plain and decorated are known.

bowls

In

several

cases,

the

decoration

is restricted

to

the

rim. The

distinction between the shape of the bowls and the cooking dishes is somewhat arbitrary

when

the

only

rim

survives,

but

the bowls

are often

decorated,

while

the

cooking dishes are left unpainted. 23

(X 1865, from K 20-Surface) Bowl,

rim

Diam.

of rim

Fig.

5.3

sherd. ca. 38.

Mirabello

Fabric (reddish yellow, 7.5YR 6/6). Thickened rim. Dark band (red) on the rim.

24

(X 219, from R 20-2)

Fig.

5.4

sherd. jar, spout Bridge-spouted Diam. of rim not measurable.

Mirabello Thickened paint

not

Fabric (reddish yellow, 5YR 6/6). rim. Dark paint on exterior and on inside of rim. Added white

preserved.

Comments: East Cretan White-on-Dark Ware, based on the typical rim pro file with paint on the interior of the rim. For a list of East Cretan examples from EM III-MM IA, see Betancourt 1984, pp. 45-46. 25

(X 607, from O 20-2) Vessel,

body

sherd.

Fig.

5.4

THE

POTTERY

Max.

79

dim.

11.6.

A coarse fabric (reddish yellow, 5YR 6/6). (X 6, from O 22-2)

26

5.4

Fig.

Cooking dish, rim and body sherds. Diam.

of rim 44.

Mirabello

Fabric (red, 2.5YR 5/8).

(X 448, from P 20-1)

27

dish, body sherd.

Cooking Max.

2.3.

dim.

Fabric (red, 2.5YR 5/6, and reddish brown, 5YR 4/3).

Mirabello

(X 658, from K 17-1)

28

dim.

2.6.

(X 672, from L 17-2)

29

5.4

Fig.

dish, rim sherd.

Cooking Diam.

5.4

Fabric, fine grained (red, 5YR 4/6).

Mirabello

of rim

Mirabello

ca. 30.

Fabric, fine grained (red, 2.5YR 4/6). rim.

Straight

(X1851, fromM

30

19-Cleaning) Fig. 5.4 dish, rim sherd with wide, knoblike handle attached at the bottom

Cooking of

Fig.

dish, rim of body sherd.

Cooking Max.

5.4

Fig.

the

rim.

Diam.

of rim not measurable.

Fabric (red, 10R 5/6). and parallels: For discussion of the cooking dish, a shallow open vessel that appears in Crete as early as EM I and often preserves marks of fire from cooking, see Betancourt 1980. This fabric is typical of EM III andMM cook Mirabello

Comments

ing dishes in this part of Crete. Compare Banou 1995b, p. 113, no. ADC p. 116, no. ADC 74 (Pseira). 31

(X 577, from P 20-3) Shallow Diam.

bowl, of rim

Mirabello 32

rim

Fig.

Vessel,

rim

5.4

sherd.

ca. 30.

Fabric (from red, 2.5YR 5/6, to light reddish brown, 5YR 3/4).

(X 1853, from J 17-Cleaning) Diam.

42, and

Fig.

5.4

sherd.

of rim not measurable.

coarse fabric (weak red, 10R 4/4). Vertical, slightly offset rimwith holes pierced in the rim before firing. Comments: It is possible that this vessel is aNeolithic "cheese pot" from the Cyclades. See Katsarou and Schilardi 2004, pp. 38, 39, nos. 1-8, fig. 10. A medium

33

(X 8, fromM 22-Surface)

Fig. 5.4

Bucket jar (or jar or basin?), base and body sherd. Diam.

of base

Mirabello

greater

than

56.

Fabric (reddish yellow, 5YR 6/6).

CHAPTER

8o

5

V

C7|

24

/ 26

27

/ / 25 01

28

^ 29

7

-

31

30

I

f

O 38

I I

35

34

33

32

T 39 40

37

36

^ 41

42

43

/

i

THE

Figure 5.4 (opposite). Pottery from the slag pile (24-43). Scale 1:3

34 Bucket

POTTERY

(X 1850, from H 19-Cleaning)Fig. 5.4 jar>base sherd Max.

dim.

Mirabello 35

dim.

Mirabello

Max.

jar, dim.

Mirabello

5.4

Fig.

6.4.

Fabric (reddish yellow, 5YR 6/6).

(X 1864, from G 18-Cleaning) Bucket

37

13.5.

Fabric (reddish yellow, 5YR 6/6).

(X 1852, from J 18-Cleaning) Bucket jar, body sherd. Max.

36

8l

rim

Fig.

5.4

sherd.

8.1.

Fabric (reddish yellow, 5YR 6/6).

(X 69, from Q^20-2) Closed

vessel,

Diam.

of rim

rim ca.

Fig.

5.4

sherd. 16.

A coarse fabric (yellowish red, 5YR 4/6). Probably a jug. 38

(X 99, from S 20-2)

Fig.

5.4

Jar (or jug?), rim sherd. of rim not measurable;

Diam.

Mirabello

(X 659, from K 17-Surface) Closed Diam.

dim.

3.8.

Fabric, fine grained (red, 2.5YR 4/6). rim.

Outturned

39

max.

vessel,

rim

Fig.

5.4

sherd.

of rim not measurable;

max.

dim.

3.4.

Mirabello

Fabric, fine grained (yellowish red, 5YR 4/6). rim. outturned Slightly

40

(X 422, from P 20-3) Jar, handle Max.

dim.

Mirabello 41

Jar, handle dim.

Mirabello

5.4

sherd. 6.7.

Fabric (red, 2.5YR 4/6).

(X 661, from K 17-Surface ) Max.

42

Fig.

Fig.

5.4

sherd. 8.4.

Fabric (light reddish brown, 5YR 6/4).

(X 660, from K 17-1)

Fig.

5.4

Jug, handle sherd. Max.

dim.

Mirabello 43

4.8.

Fabric (red, 2.5YR 4/6).

(X 405, from R 20-1)

Fig.

Jarwith thickened, flat-topped

5.4 rim, rim sherd.

CHAPTER

82

5

1 11 T 45

44

II 47

46

I ! '

i

i i

I

i i

/ i i

48

51

49 50

11

r i

i

i

i

i 55

54

53

52

i i

i i i i

i i

I I I

/ 56

57

58

59

Figure 5.5. Pottery from the slag pile (44-59). Scale 1:3

THE

POTTERY

Diam.

Fabric (red, 2.5YR 4/6).

(X 1859, from R 20-Cleaning) Jarwith thickened, flat-topped Max.

dim.

Fabric (red, 10R 5/6).

(X 67, from P 21-1)

Thickened 46

Vessel,

48

Max.

Mirabello 49

5.5

3.8.

Fabric (yellowish red, 4YR 4/6). Fig.

5.5

sherd.

body dim.

Fig.

sherd.

body

(X 68, from S 20-2) Vessel,

sherd.

Fabric (light red, 2.5YR 6/6).

dim.

Mirabello

5.5

Fig.

jar, rim ca. 22.

(X 7, from O 22-2) Max.

sherd.

Fabric (light reddish brown, 5YR 6/4). rim.

(X 664, from K 17-1)

Mirabello

5.5

ca. 28-32.

Bridge-spouted Diam. of rim

47

Fig.

jar, rim

Bridge-spouted Diam. of rim

Mirabello

Fig. 5.5 rim, rim sherd.

8.8.

Mirabello 45

ca. 32-36.

of rim

Mirabello 44

83

3.7.

Fabric (yellowish red, 5YR 4/6).

(X 218, from R 20-2)

Fig.

5.5

Vessel, body sherd. Max.

dim.

4.6.

A coarse fabric containing phyllite and other stones (red, 2.5YR 5/8). 50

(X 403, from R 20-Surface)

Fig.

5.5

Jar or bellows, body sherd. Max.

dim.

Mirabello 51

3.6.

Fabric (unevenly colored, from red, 2.5YR 5/8, to gray).

(X 413, from R 20-Surface)

Fig. 5.5

Jar or bellows, body sherd. Max.

dim.

Mirabello 52

4.3.

Fabric (red, 2.5YR 4/8).

(X 457, from N 17-2)

Fig. Jar or bellows, body sherd. Max.

dim.

5.5

3.6.

A coarse fabric containing phyllite and other stones (red, 2.5YR 3/6). 53

(X 458, fromM Jar or bellows,

17-2)

Fig. body sherd.

5.5

CHAPTER

84

Max.dim. 5.Figure 5.6 (opposite). Pottery from Fabric (unevenly colored, from reddish yellow, 5YR 6/8, to yelthe slag pile (60-72) and the apsidal lowish red, 5YR5/6). structure (73-82). Scale 1:3

Mirabello

54

5

(X462,fromOl8-Surface)

Fig.

5.5

Jar or bellows, body sherd. Max.

dim.

6.4. Fabric

Mirabello

colored,

(unevenly

from

red, 2.5YR

4/6,

on

the

exterior,

to

yellowish brown, 10YR 5/4, on the interior). (X 467, fromM

55

17-Surface)

Fig.

5.5

Jar or bellows, body sherd. Max.

dim.

Mirabello

5.2.

Fabric (light red, 2.5YR 6/8, with a gray core).

(X 470, from M

17-Surface)

Jar

or bellows,

sherd.

Max.

dim.

56

coarse

A

body

5.5

7.2. fabric

containing

red, 10R 5/6, to gray). Comments: Discolored 57

Fig.

(X 633, from R 20-2)

phyllite

and

other

stones

(unevenly

colored,

from

by partial reduction. Fig.

5.5

Jar or bellows, body sherd. Max.

dim.

Mirabello 58

4.2.

Fabric (red, 2.5YR 5/8).

(X 663, from K 17-1)

Fig.

5.5

Shallow bowl or cooking dish, body sherd. Max.

dim.

Mirabello 59

(X 666, from K 17-1) Jar

or bellows,

Max.

dim.

Mirabello 60

body

Mirabello

5.5

sherd.

Fabric (yellowish red, 5YR 5/6, and dark reddish gray, 5YR 4/2).

vessel, dim.

Fig.

5.8.

(X 674, from L 17-2) Open Max.

61

5.5.

Fabric (red, 2.5YR 5/6).

body 3.7.

Fig.

5.6

sherd.

Fabric (yellowish red, 5YR 4/6).

(X 675, from L 17-2)

Fig.

5.6

Jar or bellows, body sherd. Max.

dim.

Mirabello 62

6.

Fabric (red, 10R 4/6).

(X 676, from L 17-2)

Fig.

5.6

Jar or bellows, body sherd. Max.

dim.

Mirabello

4.3.

Fabric (dark gray, 10YR 4/1).

THE

POTTERY 85

1 1 1

1

I

(

I

60

I

62

61

63

64

1 1

1 1

1 1

I

i

\ 65

1

1/

67

66

/ 1

68

69

/

70

1 1

71

72

mi

73

74

?,'

1

75

I

r

-v

77

76

l:0>

78

tI

/1 !? \

NL 79

1 80

81

82

86

CHAPTER

(X 685, from N 21-Surface)

63

Fig.

5

5.6

Jar (or bellows?), base sherd. Diam.

ca. 38.

of base

A coarse fabric (unevenly colored, from red, 2.5YR 4/8, to light red, 2.5YR 6/6). (X 689, from L 17-2)

64

Fig.

5.6

Jar or bellows, body sherd. Max.

dim.

Mirabello

4.4.

Fabric (red, 2.5YR 4/8).

(X 720, from N 20-3)

65

Vessel, Max.

5.6

body dim. 2.4.

Mirabello

Fabric (dark brown, 7.5YR 4/2).

(X 1722, from N 17-2)

66

Fig.

sherd.

Fig.

5.6

Jar (?), body sherd. Max.

dim.

Mirabello

4.7.

Fabric (red, 2.5YR 5/6).

(X 1855, from O 19-Cleaning)

67

Vessel, Max.

Vessel,

Fig.

5.6

Fabric (red, 10R 5/6).

(X 1857, from O 19-Cleaning) Max.

5.6

body dim. 5.4.

Mirabello 68

Fig.

sherd.

sherd. body dim. 4.3.

Mirabello

Fabric

(exterior surface weak

red, 10R 5/4; dark interior and

core).

69

(X 1858, from O 19-Cleaning) Vessel, Max.

Vessel,

Fig.

5.6

sherd. body dim. 7.9.

Mirabello 71

Fabric (red, 10R 5/6).

(X 1862, from 119-Cleaning) Max.

5.6

body dim. 9.0.

Mirabello 70

Fig.

sherd.

Fabric (red, 10R 5/6).

(X 1863, from Q_19-Cleaning)

Fig.

5.6

Fig.

5.6

Vessel, body sherd. Max.

dim.

Mirabello 72

4.2.

Fabric (red, 10R 5/6).

(X 1867, from N 20-Cleaning) Vessel, Max.

body dim.

Mirabello

sherd. 3.9.

Fabric (red, 10R 5/6).

THE

POTTERY

87

TABLE 5.5. EARLY MINOAN III-MIDDLE SHERDS FROM THE SLAG PILE Rim

Handle

Base

Spout

Body

Leg

21

\.

Total

Whole

1 2. 1 2

1 3. 4. 1 5.or 2 1 6. 7.

1

8.

1

9.

3

4

3

3

11 12 2 or623

3

1 11 11

22 3

5

1 2

3 2

2

11.

1

11

12.

Min. No.

4

10. 2 2

13.

IA

MINOAN

2

2

1

3

2 14.

2

2725 10

Fine fabric sherds: (1) rounded cup (15-18, X 459, X 668,X 698,X 796); (2) cup, painted (19-21, X 711, X 719,X 1223).Medium coarse fabric sherds: (3) shallow bowl (22,X 3); (4) vessel (25, 32,X 607,X 1853).Mostly Mirabello Fabric sherds: (5) cooking dish (26-30, X 6,X 448, X 658,X 672,X 1851); (6) bowl (23,X 1865); (7) shallowbowl (31,X 577); (8) bucket jar (33-36, X 8,X 1850,X 1852,X 1864); (9) jaror jug (37-39, X 69,X 99,X 659); (10) jar (40,41, X 422, X 661); (11) jug (42,X 660); (12) jarwith thickened, flat-topped rim (43,44, X 405, X 1859); (13) bridge-spouted jar (24,45,46, X 219, X 67,X 664); (14) vessel (47-72, X 7,X 68,X 218, X 403, X 413, X 457, X 458, X 462, X 467, X 470, X 633,X 663,X 666,X 674, X 675,X 676, X 685, X 689, X 720,X 1722,X 1855,X 1857,X 1858,X 1862,X 1863,X 1867).

from

Pottery Final 73

the

Apsidal

Structure

Neolithic (X 1870, from O 19-Cleaning) Bowl, body sherd.

Max.

dim. coarse

A

5.6

Fig.

4.3. fabric

colored,

(unevenly

with

an

unevenly

colored

surface,

brown

to reddish brown, 2.5YR 4/4).

TABLE 5.6. FINAL NEOLITHIC APSIDAL STRUCTURE Rim

Handle

Base

Spout

Leg

SHERD FROM THE

Body

1.

Whole

Total

Min. No.

Ill

Coarse,

dark, burnished

fabric

sherd:

(1) bowl

(73, X

1870).

EM II 74

(X 66, from S 20-2) Fig. 5.6 Jar, body sherd. Max.

dim.

A medium 6/4).

9.7.

coarse, gritty, chaff-tempered

fabric (light reddish brown, 5YR

CHAPTER

88

5

Large closed vessel, probably a jar; trickle decoration in dark paint. Comments: This atypical vessel cannot be closely dated. Good parallels for the decoration exist from EM II atMyrtos (Warren 1972). The fabric is not typi cal for Chrysokamino.

TABLE 5.7. EARLY MINOAN APSIDAL STRUCTURE Rim

Handle

Base

II SHERD FROM THE

Spout

Leg

1.

Total

Min. No.

Ill coarse

Medium

EM 75

Whole

Body

fabric

(1) jar (74, X

sherd:

III-MM

Fine

IA:

(X 167, fromM

Fabrics

18-1, M

Conical

cup,

rim

Restored

H.

ca. 6.5-7.0;

66).

18-2, andM

Figs. 5.2, 5.6

18-3)

sherds.

and base

Diam.

of rim

ca. 10; Diam.

of base

ca. 5.5-6.0.

A fine fabric (pink, 7.5YR 7/4); burnished. Dark slip on inside of rim and on Added

white

Comments:

East

exterior.

with

and without

triangles

with

paint:

White-on-Dark are common

handles,

corners

hatched

is one

Ware.

of the

76

2, nos.

p. 24, motif

1984,

(X 148, fromM

19-3)

Fig.

corners.

hatched Decorated

in the ware.

Parallels: For the shape, see Betancourt see Betancourt

with

triangles

pendant

Cretan

standard

The

conical

decoration

cups,

both

of pendant

types.

1984, p. 39, shape 2A; for the motif

3-5.

5.6

Rounded cup, handle and body sherds. Max. dim. (largest sherd) 4.4. A fine fabric (light brown, 7.5YR 6/4). One

handle

See

Comments:

77

circular

with

section.

15.

(X 182, from N 19-2, N 19-3, M Rounded

cup, of rim

Diam.

rim

and body

19-3, M

19-5)

Fig.

5.6

sherds.

ca. 9-10.

A fine fabric (pink, 7.5YR 7/4). Dark slip; band on rim. 78

(X 210, fromM Rounded Diam.

cup, of rim

19-2B, M

19-3, andM

19-4)

Fig.

5.6

fragmentary. ca. 8.

A fine fabric (pink, 5YR 7/4). Band on rim, inside and out; line on outside of handle. One

handle

79

with

sherds

Several

circular

(X 415, from N 19-3) Rounded Max.

section.

burned.

Fig.

5.6

cup, handle sherd. 3.8.

dim.

A fine fabric (reddish yellow, 7.5YR 6/6). Dark slip; line on outside of handle. Handle

80

with

circular

(X 534, fromM Rounded

section.

18-3)

Fig.

5.6

cup, body, base, and handle sherds.

POTTERY

THE

89

ca. 9-10.

of rim

Diam.

slip; line on outside of handle.

A fine fabric (pink, 7.5YR 7/4). Dark 81

(X 636, from N 19-3)

cup (?), handle sherd.

Rounded Max.

5.6

Fig.

dim.

4.5.

A fine fabric (light reddish brown, 5YR 6/4). Dark slip: line on outside of handle. with

Handle

82

section.

circular

(X 149, fromM

5.6

19-3)

Fig. Closed vessel, body sherd. Max.

dim.

5.2.

slip on exterior; no added white paint

A fine fabric (pink, 5YR 7/4). Dark preserved.

Comments: From the thickness of the sherd and the quality of the dark slip, this vessel ismost likely a bridge-spouted jar of East Cretan White-on-Dark Ware with the white paint missing (Betancourt 1984, pp. 45-46). 83

(X 181, from N 18-2 and N 19-2) Closed

vessel,

Max.

dim.

of

body

sherds.

largest

sherd

8.5.

A fine fabric (unevenly colored, mostly added

white

vessel,

most

exterior;

covering Large

closed

East

Comments:

Cretan

are often

in this ware

decorated

spirals,

with

spirals

and

jar. jars Large bridge-spouted are more other motifs. They

IA

1984, pp. 45-46.

See

(X 466, fromM Closed

vessel, dim.

likely a bridge-spouted

jar.

83.

17-Surface)

body

Fig.

5.7

sherd.

3.2.

A fine fabric (reddish yellow, 5YR 7/6). Dark

slip: covering exterior; added

traces.

paint:

closed Large Comments:

vessel, See

most

likely

Fig. Closed vessel, body sherd. dim.

a

bridge-spouted

jar.

83.

(X 1675, from N 18-2)

Max.

slip: covering exterior; added

lines.

diagonal

paint:

Max.

86

slip

lines.

sherd helps provide a secure EM III-MM

Fig. 5.7 Closed vessel, body sherd. Diam. of body of vessel above 32. A fine fabric (light brown, 7.5YR 6/4). Dark

Comments:

white

diagonal

(X 271, from N 18-2)

Large closed vessel, most

85

light brown, 7.5YR 6/4). Dark

a likely bridge-spouted Ware. White-on-Dark

Parallels: See Betancourt

white

bands,

paint:

common than jugs or teapots. This date for the deposit.

84

5.7

Fig.

5.7

5.1.

A fine fabric (reddish yellow, 5YR 7/6). Dark slip covering exterior; traces of added white paint. Large closed vessel, most likely a bridge-spouted jar. Comments:

See

83.

chapter

9?

5

/ // 84

83

! 86

85

7 87

89

7 90

Figure 5.7. Pottery from the apsidal Structure

(83-90).

Scale 1:3

THE

POTTERY

EM

III-MM

91

IA: Medium

(X 145, from M

87

Shallow

bowl,

Diam.

and

Fabrics

Coarse

5.7

Fig.

sherds.

ca. 30-35.

of rim

coarse fabric (mostly pink, 5YR 7/4). Dark slip: band on the rim,

A medium inside

19-3)

rim

to

Coarse

out.

Straight, thickened rim. Comments:

is a common

This

(X 146, fromM

88

see 22.

shape;

Fig. 5.7 Shallow bowl, base sherd with handle. Diam.

19.

of base

coarse fabric (reddish yellow, 7.5YR 6/6).

A medium

See

Comments:

89

19-3)

22.

(X 168, from M

18-2 andM

18-3)

Fig.

5.7

Shallow bowl, rim sherd. Diam.

of

the

coarse fabric (reddish yellow, 7.5YR 6/6, with a darker surface on

of

groups

paint:

a on surface paler short lines on rim.

and

exterior

ca. 39-41.

rim

A medium

the

interior).

Dark

slip

on

rim. Added

white

rim.

Straight

See

Comments:

22.

Decoration

is often

only

on

the

rim,

as on

this

ex

ample.

90

(X 274, from N 18-2) Fig. Shallow bowl, rim sherd. Diam.

rim

of

A medium Comments:

Shallow

Straight Comments:

19-4)

rim

bowl, of rim

A medium

Fig.

5.8

sherd.

ca. 30-32.

coarseMirabello

Fabric (light reddish brown, 5YR 6/6).

rim. See

22.

(X 150, from M 19-3) Fig. 5.8 Shallow bowl or cooking dish, rim sherd. Diam.

of rim

Mirabello

ca. 25.

Fabric, fine-grained

(X 212, fromM Shallow Diam.

(red, 2.5YR 4/6).

rim.

Straight

93

22.

See

(X 1662, fromM Diam.

92

ca. 30.

coarse fabric (reddish yellow, 7.5YR 6/6). Band on rim.

rim.

Straight

91

5.7

bowl, of rim

18-3 andM rim

19-4)

94

5.8

sherds.

ca. 20.

Mirabello with

Fig.

Fabric, fine-grained (reddish brown, 2.5YR 5/4). Shallow bowl rim. Band in dark slip on rim, inside and out. straight

(X 270, from N 18-2) Fig. 5.8 Shallow bowl, rim and body sherd.

CHAPTER

5

91

93 i

7

92

y

94

95

> 97

96

98

y' 99

/

100

101

THE

POTTERY

Figure 5.8 (opposite). Pottery from the apsidal

structure

(91-101).

Scale

1:3

Diam.

of

93

ca. 28; max.

rim

3.2.

dim.

Mirabello Fabric (unevenly colored, from red, 2.5YR 5/6, to light reddish brown, 5YR 6/4). 95

(X 275 (from N 18-2, M

Fig.

5.8

dish, rim sherd.

Cooking

ca. 55-60.

of rim

Diam.

18-2)

Mirabello

(yellowish red, 5YR 4-5/6, with

Fabric, fine-grained

a browner

surface).

Straight, thickened rim. Comments:

96

See

25.

(X 411, from N 18-2) Cooking

of rim

Diam.

5.8

Fig. dish, rim sherd. ca. 55-60.

Mirabello

Fabric, fine-grained Straight, thickened rim.

(brown, 10YR 4/3).

(X 532, fromM

5.8

97

18-3) Fig. rim sherd. dish,

Cooking Diam.

of rim

ca. 45-55.

Mirabello Fabric, fine-grained Straight, thickened rim. 98

(red, 2.5YR 4/6, with a darker exterior surface).

(X 1661, from N 19-3) Cooking Max.

dim.

3.3.

Fabric (yellowish red, 5YR 4/6). Burnished

Mirabello 99

(X 641, from N 19-4) or

Bowl

Mirabello 100

in the exterior.

5.8

Fig. rim

dish, cooking of rim not measurable.

Diam.

5.8

Fig. dish, body sherd.

sherd.

Fabric (reddish yellow, 5YR 7/8).

(X 2, from L 17-2, L 18-2, M

18-1, M 20-2, N 18-2)

Fig. 5.8

Bucket jar, rim and body sherds. Max.

dim.

Mirabello

of

largest

Fabric

sherd

11.6.

(light brown, 7.5YR 6/4, to light reddish brown, 5YR

6/4). 101

(X 169, from M Bucket H.

20.4;

Jar,

p. 114) from probably

as larnax bucket

fragments or basins.

jars

are

5.8

of base

35-37.

19-3) Fig.

inside

5.9

sherd. of rim

ca. 32.

A coarse fabric containing quartz and other stones (reddish brown, 2.5YR

34. The sherds identified by Haggis (2005,

rim

Diam.

Fig.

Fabric (light brown, 7.5YR 6/4). Band on rim, extending

(X 144, fromM Diam.

18-2)

jar, complete profile. Diam. of rim 35-36;

Mirabello

102

18-1, M

5/4). Thickened,

flattened rim.

chapter

94 (X 211, fromM

103

Jar, rim

19-4, N 18-3)

Fig.

5

5.9

sherd. ca. 40-42.

of rim

Diam.

Fabric (from red, 2.5YR 5/8, to black). Thickened, flattened rim.

Mirabello

(X 291, from N 18-2)

104

Jar, rim

5.9

Fig.

sherd. of rim

Diam.

ca. 40-42.

Fabric (red, 2.5YR 5/6). Thickened, flattened rim.

Mirabello

(X 1659, from N 19-2)

105

rim

Jar,

5.9

sherd. of rim not measurable.

Diam.

Mirabello Thickened,

Fabric (yellowish red, 5YR 5/6). flattened rim.

(X 1663, fromM

106

Fig.

Jar, rim

sherd. of

Diam.

5.9

Fig.

18-1)

rim

ca. 30-32.

Fabric (between yellowish red, 5YR 5/8, and dark reddish brown,

Mirabello 5YR3/4).

(X 530, fromM

107

Bridge-spouted of rim Diam.

Mirabello

rim jar, ca. 30-32.

sherd.

Fabric (yellowish red, 5YR 4/6).

(X 414, from N 19-3)

108

5.9

Fig.

18-3)

Fig.

5.9

Closed vessel, probably a jar, handle sherd. dim.

Max.

10.4.

Mirabello Fabric (light red, 2.5YR 6/6), with a darker core (between pale red, 2.5YR 6/2, and weak red, 2.5YR 5/2). (X 1664, from N 19-2)

109

Jar, handle dim.

Max.

Mirabello 110

111

Fabric (red, 2.5YR 5/6).

Closed

vessel,

Diam.

of base

probably ca. 17-20.

Jar

or basin, dim.

Mirabello

5.9

Fig. ajar,

sherd.

base

Fabric (pink, 7.5YR 7/4).

(X 525, fromM Max.

112

sherd. 6.

(X 1658, from N 19-2)

Mirabello

5.9

Fig.

body of

8-3, M

19-5)

Fig.

5.9

sherds.

largest

5.7.

sherd

Fabric (between light red, 2.5YR 6/8, and red, 2.5YR 5/8).

(X 568, from L 18-2)

Fig.

Jar or bellows, body sherd.

5.9

THE

\

102

POTTERY

95

\

104

i

i

103

105

T

106

107

110 109 108

I 0/

I 111

113

112

/

114

115

/ /

/

/ 116

117

118

I

' /

119 Figure 5.9. Pottery from the apsidal structure (102-121). Scale 1:3

120

'

/

121

CHAPTER

96 Max.

dim.

Mirabello 113

5.7.

Fabric (reddish brown, 2.5YR 5/4).

(X 601, fromM

19-3)

Fig. Jar or bellows, body sherd. Max.

5

dim.

5.9

4.8.

Fabric (unevenly colored, from red, 2.5YR 4/6, to dark reddish

Mirabello gray, 5YR 4/2). 114

(X 602, from M

19-3) Fig. or bellows, body sherd. Jar Max.

dim.

Mirabello 115

5.9

4.6.

Fabric (reddish brown, 2.5YR 4/4, with a gray surface).

(X 610, from M

19-5)

Fig.

5.9

Jar or bellows, body sherd. Max.

dim.

Mirabello 116

(X 643, from N 19-4) Open Max.

vessel, dim.

Mirabello 117

Vessel,

Fabric (yellowish red, 5YR 5/6). 19-3)

Vessel,

Fig.

5.9

Fabric (brown, 7.5YR 5/4) and (reddish yellow, 5YR 6/8).

(X 1660, from N 19-3) Max.

5.9

sherds. body dim. 3.2 and 3.8.

Mirabello

Fig.

5.9

sherd. body dim. ca. 2.8.

Mirabello 119

body

Fig.

sherd.

4.3.

(X 816, fromM Max.

118

5.6.

Fabric (reddish brown, 2.5YR 4/4, with a gray surface).

Fabric (yellowish red, 5YR 4/6).

(X 1721, from N 19-2)

Fig.

5.9

Jar (?), body sherd. Max.

dim.

Mirabello 120

7.3.

Fabric (yellowish red, 5YR 5/6).

(X 1737, fromM Vessel, Max.

Mirabello

Fig.

5.9

sherd.

body dim.

19-Surface)

7.8.

Fabric (dark reddish brown, 5YR 3/4, with yellowish red, 5YR 5/6,

and dark gray core). 121

(X 1781, from N 19-3) Closed Max.

vessel, dim.

Mirabello

base

Fig.

5.9

sherd.

5.9.

Fabric (yellowish red, 5YR 5/6).

THE

POTTERY

97

III-MIDDLE MINOAN TABLE 5.8. EARLY MINOAN SHERDS FROM THE APSIDAL STRUCTURE Rim

Handle

Base

Spout

Leg

Body

Whole

1. 4

2.

1

3. 1

4. 5. 6.

Total 1

1 5

7

1

7

4

12

5 1

7.

9

9.

11 1 2 1

11. 12.

Totals

29

7

10

Min. No.

14

1

10.

IA

2 4 1 2 1

16

16

5

33

82

32

Fine fabric sherds: (1) conical cup, painted (75,X 167); (2) rounded cup (76-81, X 148, X 182,X 210, X 415, X 534, X 636); (3) closed vessel, paintedwhite on dark (82-86, X 149,X 181,X 271, X 466, X 1675).Medium coarse fabric sherds: (4) shallow bowl (87-91, X 145,X 146,X 168,X 274, X 1662).Mirabello Fabric sherds: (5) cooking dish (92-98, X 150,X 212, X 270, X 275, X 411, X 532, X 1661); (6) bowl (99,X 641); (7) bucket jar (100,101, X 2,X 169); (8) jar,flat thickened rim (102-106, X 144,X 211, X 291, X 1659,X 1663); (9) bridge-spouted jar (107,X 530); (10) jar (108,109, X 414, X 1664); (11) closed vessel (110,X 1658); (12) vessel (111-121, X 525, X 568,X 601, X 602, X 610,X 643,X 816,X 1660,X 1721,X 1737,X 1781).

III TABLE 5.9. COMPARISON OF EARLY MINOAN IA POTTERY SHAPES FROM THE MIDDLE MINOAN SLAG PILE AND THE APSIDAL STRUCTURE Slag Pile

Shape. Conical

yes

cup

Rounded

Apsidal Structure

yes

yes

Shallow bowl

yes

yes

Bowl

yes

yes

yes

yes

Bucket

cup

jar

jar Bridge-spouted thickened, Jar with flat-topped Miscellaneous closed vessel

rim

yes

yes

yes

yes

yes

Jug

yes

yes no

Cooking dish

yes

yes

The

Tools

Stone

byDoniert Evely

in spite of a collecting strategy (Figs. 6.1-6.4). From the ca. 40-50 m3 of deposits excavated (ca. 20%-25% of the slag pile), com furnace chim prised mostly of small pieces of slag and parts of demolished neys, only 23 stone tools were recovered. The corpus includes three pieces of obsidian, one potential surface for working, and a number of limestone hand tools, along with some cuttings in bedrock (Fig. 6.3). This is a small

Only

a few stone tools were

found in the workshop, that saved 100% of the excavated material

number of tools for a site where

the breaking up of furnace slags to retrieve of the working copper prills major part cycle. The three pieces of obsidian from the workshop and 144) (142,143, on as none of the pieces has site, may be debris from working the material either the size or other attributes that would merit identification as a tool. was

a

ores (see App. C), must have a useful material, and its presence is the from It imported Cyclades. here is not surprising. So little has been recovered that nothing can be de

The

obsidian,

like at least some of the metal

been

duced from it except the likelihood that itwas worked and used at the site. The largest class of stone tools consists of water-worn limestone and dolomite

cobbles.

In terms of size and shape, most examples fall within a are in one hand, either hafted essentially tools wielded

small range. They or unhafted. Lengths vary from 10-15 cm, and the weight is normally up to half a kilogram. No uniformity of shape can be recognized among the cuboid, bar-shaped, elliptical, ovoid, and nearly spherical forms. All the owe stones their shapes to the effects of water action, having been tumbled in watercourses and/or on the seashore so that their angles have become rounded and their surfaces worn to a smooth and even texture. The sole in the matter of size is 125, a larger implement whose exception shape, a softened the of effects resembles that of water-erosion, by though large boulder, split lengthwise. The piece is heavy, and itwas probably intended some to provide an immobile surface on which activity (other working was than pounding) carried out. The tools seem to have been collected as natural cobbles, and none in order to modify of them appears to have been deliberately worked the natural

shape. Suitable

gathering

places for the waterworn

cobbles would

CHAPTER

IOO

6

have been found at the sea s edge, with the nearest readily accessible source a at the mouth of a small ravine, being pebbly beach at Agriomandra, about a kilometer away (west of the Chrysokamino farmhouse). All of the were tools pieces at the site, whether they to the not occur do transported workshop by humans; they among the angular, slab limestones and layered phyllites of the

rounded

limestone

and dolomite

or not, were naturally promontory.

of tool types is difficult. Many of the pieces cannot nor can as identified their be reliably function be established, because tools, no wear marks are visible on the surface. Eight stones have been excluded are included their from serious consideration, basic descriptions though in the catalogue. At the most generous level of interpretation, five tools An

identification

as

and 133), two were pounders (124,126,129,132, used for grinding actions (122,123), and one (125) may be regarded as a or support. The surface uncertainty concerning use stems mainly working from the uniformly poor state of the surfaces. This condition has been called could be considered

not clear if this state is the "degraded" in the catalogue descriptions, and it is of the corrosive effects of the debris produced result of natural weathering, in the metalworking of several factors. In processes, or of a combination any case, the surfaces of the stones are somewhat powdery, and details are in blurred. In addition, patinas and carbonate deposits (called concretion the catalogue) are also present. are

the three represented within the corpus. Excluding obsidian, the tools are all native limestone pieces of (presumably Melian) and dolomite. The most common kind by far is a dark gray stone similar to Few materials

rocks used as materials

carbonate

for tools throughout the Bronze Age in of off-white to yellow limestone (125), and

Crete. A single example ismade one tool is a carbonate with (136). Analysis larger crystalline composition a limestone breccia. The dark gray 131 is shows that byMyer (unpublished) carbonates would have made a reasonably suitable material for the percus sive work stones

likely to have been carried out at the site, although

would

have

been

are rare in the immediate to Gournia). and itmight other

better.

However,

sources

of

finer-grained

fine-grained

limestones

are obtainable closer they of an unusual limestone,

neighborhood (though pale yellow tool (125) ismade have been used for a different sort of activity from that of the

The

tools.

small number of tools can be explained in several ways. Perhaps tools were broken or damaged, they were slung to one side, an action down the that would almost certainly ensure their eventual disappearance The

when

sea. If any of the tools were hafted, slopes of the saddle into the have been taken away when work ceased. Patterns tools. They

they might

of wear from usage can be seen on only a few of the limestone can be placed in the following categories:

loss of small surface chips as a result of direct percussive action of the stone tool against an item of considerable hardness (such as slag) areas of 2. The accidental removal of large flakes from the working the tool 1.The

THE

IOI

TOOLS

STONE

creation, as a result of grinding action (possibly combined with a degree of crushing), of a surface texture varying from smooth to polished

3. The

cut into the stone tool by the agitation of in the course of work (such as stray grains of harder materials sand or glassy silica-based substances in the slags)

4. Occasional

In addition

scratches

to the tools that have been recovered, one would expect to find items. Apart from more pounders and grinders, the most are items with a firm surface against which lacking set of tools

some additional

obviously the slag could have been broken up. The single instance of a stone-work was not used in this way. It is ing surface (125) possible that the limestone on-the with natural slablike hillside, patterns of cleavage, would outcrops a have supplied surfaces, although no damaged ready supply of working areas were

encountered

in the excavation.

Two

interconnected

circular

(35 cm across and up to 7 cm deep) were found in situ in the depressions at the site (Figs. 6.2,6.3), bedrock and they could have been used in virgin this way, but one would expect that many more would have been created. For the rest of the manufacturing processes undertaken at the site, the and surfaces would (mostly missing) working pounders, grinders, have sufficed for breaking up the charcoal and other fuels and demolishing same

the furnaces after each smelt. Sharper edges for cutting and slicing actions (obsidian and metal blades) or for chopping actions (axes) have relatively minor roles to play in the craft activities envisaged for Chrysokamino. and food preparation activities, however, if occurring in and near Cooking the small hut, might have required obsidian and other cutting tools aswell pounding and grinding tools. It is assumed that the clay and straw mix

as

consumed prepared use

or

in such quantities in forming the furnace chimneys was probably away from the smelting site and brought in ready for immediate storage.

long-term

CATALOGUE Stone

Ground 122

Tools

(X 451, from N 17-2)

Rod-shape

with

somewhat

triangular

section;

6.4

6.1,

Figs. Hand tool (?), fragment; ends lost, sides broken. Max. L. 10.7;W. 4.9; Th. 3.9;Wt. 333 g. Limestone (gray); layered texture. flat face

smooth

to touch,

angled

(rounded) face slightly smoother. No

scratches

Comments:

missing 123

visible. The

ends might

smooth

imply pounding

(X 681, from O 10-4) Hand

zones

suggest

possible

grinding

or

rubbing

actions;

(with loss due to fracture). Figs.

tool (?), complete; surface degraded. L. 10.5;W. 6.3; Th. 4.9;Wt. 580 g. Dolomite (very pale gray to off-white).

6.1,

6.4

CHAPTER

6

123

122

124

126

-

fc ^

,!

>

*

^ ??,

125

127

THE

Figure 6.1 (opposite). Stone tools (122-127).

Scale

STONE

TOOLS

cobble.

Elongated 1:3

Sides plane,

ends

and

are not

but

IO3

Comments:

no

show

of use. The

signs

no

smooth;

especially The flatter

zones

Hand

tool,

L. 13.4;W. Limestone Cuboid

depression

a flat

have

or

grinding

action.

rubbing

6.4

6.1,

degraded.

with

a natural cobble. angles; on narrower end. Main

rounded

of

damage pounding centers (the more approximate

ghosts at

each

7.5; Th. 5.9;Wt. 1.220 kg. (pale gray to gray).

appearance,

Possible

surfaces

Figs. surface

complete;

for

argue

might

(X 712, from N 20-2)

124

two main

scratches.

area

distinct

faces

have

cm

is 2.5

slight

across

and

0.3 cm deep). Comments:

125

Pounder-hammer,

hafted.

perhaps

Fig. 6.1 stone naturally fractured and damaged by the sea. Support (?), complete; L. 23.1;W. 16.1; Th. 7.1;Wt 3.555 kg. Limestone (off-white and yellow). Somewhat amorphous in form; natural boulder (split?). (X 453, from O 18-1-Surface)

One

is

surface

and

flat

fairly

to the

smooth

so it was

touch,

not

to

subjected

percussion, but only to being ground or rubbed. Comments:

The

have

may

object

a

been

surface

passive

on which

work

was

performed.

126

(X 942, from J 18-Surface)

Fig. 6.1 Hand tool (?), complete; surface degraded. L. 7.5;W. 5.4; Th. 3.7;Wt. 208 g. Limestone (gray). Somewhat wedge-shaped Possible

smaller

small

of

and slightly irregular. at either

activities

pounding/bruising

end, more

so on

end.

Comments:

127

zone

The

(X 667, from M

piece

was

a

perhaps

pounder.

6.1, 6.4 Figs. Hand tool (?), complete; concretion and surface degraded (by heat?). L. 12.1;W. 9.3; Th. 5.6;Wt. 954 g. Limestone

(pale

17-2)

some

gray);

breccia

to one

adheres

side.

Ovoid, flattish. so

Surface Comments:

128

that destroyed function The

no wear is not

marks obvious;

can be the

piece

(X 217, from P 20-2) Hand

mately Max.

a

Figs. tool (?), fragment; end and middle

quarter L.

seen. would

suit

as a

pounder.

6.4

6.2,

section of larger cobble (approxi

survives).

8.3; max. W.

4.6; Th.

8.0.

Dolomite (pale gray to gray). From a cuboid (?) cobble. remnants

Possible Comments:

129

The

of a flatter function

(X 452, from N 17-2) Hand

tool (?), complete.

on side, plane is not obvious.

Fig.

either

natural

6.2,

or from

use.

6.4

CHAPTER

i04

6

?

128

129

130

I

V

I

'I

>

133

131

132

i 142 (1:1)

143 (1:1)

I 144 #.-?; Cutting

in Bedrock

THE

in bedrock. cutting otherwise indicated

Scale 1:3 unless

IO5

8.5; Th. 5.3;Wt. 657 g. (pale gray); soft.

L. 5.5;W. Dolomite

Figure 6.2 (opposite). Stone tools (128-133), obsidian tools (142-144), and

TOOLS

STONE

cobble;

Humped

rounded

from water

angles

one more

with

action,

angular.

over small portion of one end. Perhaps a hint of pounding damage The

Comments:

is not

function

(X 232, from L 21-1)

130

obvious.

6.2

Fig.

tool (?), fragment; part of end of a cobble; surface damaged and dis

Hand

to heat. by exposure Dims. 7.1x4.6x3.0.

colored

Limestone traces

No

(gray). of work The

Comments:

is not

(X 768, from N 30-3)

131

now.

discernible function

obvious.

6.2

Fig.

Hand

tool (?), fragment; part of side of larger cobble.

Dims.

8.1

x 6.0

traces

No

x 3.2.

breccia (black and rose elements).

Limestone

of work The

Comments:

discernible. is not

function

obvious.

The

material

by George Myer, Department

lyzed by optical microscopy

of

the

tool was

ana

of Geology, Temple

University.

132

(X 1753, from T 23-Cleaning) Fig. 6.2 a Hand tool, fragment; part of larger cobble. P. L.

12.0.

Dolomite

(pale gray).

Rectangular Peck marks

shape. on extant

end.

Comments:

The

was

piece

(X 1866, from K 20-Cleaning)

133

Fig.

from

use

as a

pounder.

6.2

tool, fragment of larger cobble; small piece.

Hand

6.3.

L.

Max.

broken

possibly

Limestone

(gray). Comments: No signs of use, but the fragment may have broken off the cobble in a percussive

action.

(X 637, from O 20-3)

134

Hand L.

tool,

complete;

surface

degraded.

11.3.

(pale gray to gray); partly calcined. Ovoid, humped cobble. Comments: No signs of wear, though surface below the calcined (?) crust Limestone

indicate

might

No

clear

signs

of use.

cracked

and

(X 455, from Q21-Surface)

135

Hand L.

Figure 6.3. Cutting in bedrock at the metallurgy workshop

pounding.

tool,

complete;

surface

11.2.

Limestone (gray). Ovoid, flattened cobble. Comments:

No

signs

of use.

fissured

heavily.

CHAPTER

6

122

123

124

127

128

129 Figure 6.4. Stone tools (122-124 and 127-129) from the metallurgy site

THE

STONE

136

(X 495, from O 20-2) Hand

TOOLS

tool,

IO7

nearly

complete;

or

large flakes

segments

lost.

L.11.9.

Limestone/calcite (?) (gray and off-white). Ovoid, flattened cobble; stone fissure throughout. No

Comments:

of use.

signs

(X 496, from O 20-2)

137

Hand

tool, about half missing

(split lengthways); surface condition poor in

places. L.

8.7.

Limestone

(pale gray). cuboid

Subrectangular, Comments:

138

No

cobble. of use.

signs

(X 456, from N 17-2) Hand L.

tool,

Limestone Humped

condition

poor.

(dark gray). cobble.

Comments:

139

surface

complete;

7.6.

No

of use.

signs

(X 442, from L 19-2) Hand L.

tool,

6.8; W.

on

concretion

complete; 2.8.

surface.

Limestone (gray to dark gray). Ovoid pebble. Comments: No signs of use. Found on the middle floor of the apsidal build ing (see Fig. 4.6). 140

(X 454, from O 18-1) Hand L.

tool,

concretion

complete;

(calcined

crust)

on

part

of

surface.

4.8.

Limestone (gray to pale gray). Ovoid pebble. Comments:

141

No

of use.

signs

(X 231, from L 21-1) Hand

tool,

of

fragment

large

cobble;

ca. one

surface. Max.

L.

3.8.

Limestone

(gray).

somewhat

Ovoid, Comments:

No

spherical. of use.

signs

Obsidian 142

(X 945, from J 18-1)

Fig.

6.2

Small chunk/flake, complete. Dims.

1.0 x 0.9; Th.

Obsidian

0.5; Wt.

ca. 1-2.2

(black); fairly lustrous.

g.

quarter

survives;

concretion

on

CHAPTER

io8

in core

Removed

preparation;

implies

143

obsidian

on

preparation

(X 944, from J 18-1) Chunk,

Fig.

at the implies

6.2

Bladelet,

4 g.

18-2)

Fig.

fine,

but

its

below 1 g.

tapering.

No

a core. The or the

obvious,

complete.

Comments: working the use

or of accidental

6.2

L. 1.6;W. 0.5; Th. under 0.5 mm; Wt. Obsidian (black); semilustrous. Very

side.

site.

is either a result of core preparation fragment core. The is not of this piece function of the tip on the site. the use of obsidian

(X 1208, from M

144

the

The

Comments:

presence

on one

complete.

1.5 x 1.0; Th. 0.65;Wt. Obsidian (black); lustrous.

Dims.

severance

of cortex

function is not obvious (too small), but the presence of the

Comments: The flake

elements

slight

6

working

signs of edge damage. This function of

of the flake

this material

on

is unclear, the

site.

piece is probably debris from but

its presence

shows

either

CHAPTER

7

The

Furnace

Chimney

Fragments byPhilip P. Betancourt

of the metallurgy location taken before it was excavated Photographs the small illustrate many fragments of industrial ceramics present at aptly Hundreds of thousands of pieces survive in this artifact Chrysokamino. to over 10 cm in the in sizes from that class, range microscopic longest dimension. Most of the pieces are tiny, and they come from every level and every square meter of the deposit of slag on the site. No ment can be discerned within the class.

stylistic develop

of ceramic vessels in the deposit of slag provide the best Fragments clues for dating, but problems exist with this evidence. Pieces of Final vases come from the modern Neolithic surface aswell as from deep within In addition, most parts the deposit, indicating that the debris is mixed. of the slag pile contain no vessel sherds at all. Because of these problems with the context, none of the individual industrial ceramics fragments can be assigned a secure date based on the stratigraphie level. are Typical examples of the industrial ceramics presented in the cata at The the end of this includes logue pieces from bases, chapter. catalogue bodies, and rims of cylindrical clay objects with open tops. The fragments a provide large enough selection to illustrate the range. The total number of fragments present from the site is enormous. Be land surface, many pieces have eroded sides the fragments on the modern so into the sea, of the deposit are not known. that the original dimensions The extant size of the deposit is ca. 200 m2, with a depth that varies from ca. 60 cm. nothing to The fragments have a slight curvature, and computation of the degree of arc indicates they are from cylinders with diameters of ca. 16-44 cm. Rim sherds prove that one or both ends of the cylinders were open, with were straight to slightly everted rims that usually slightly rounded at their upper edge. No base sherds from the center of any base are present, but a few angular pieces from the point at which a vertical wall met a base sug gest that the cylinders sometimes had small ledges at the bottom to allow them to sit with more stability. Wall thicknesses are usually in the range cm. of 1-2 of the bases have diameters that are substantially larger Most than most

of the rims, suggesting

that the shapes were

usually

tapered.

no

CHAPTER

7

I

I

I

The

measure 20-30 generally than 40 cm in diameter.

rim diameters

cm, while

Figure 7.1. Front and back of chim ney fragment 180, showing dark, glassy deposit on the interior (right)

the bases measure

more slightly No decoration

or occurs on any fragment, but circular holes made slip before the clay was fired are present on all fragments of more than a few centi meters inwidth (Fig. 7.1). Diameters of the holes average about 2 cm. In all cases, the configuration of the holes shows that they were made by through the wall from the exterior of the cylinder, thrusting something either at a right angle to the wall or at a slight oblique angle. Experiments with moist clay cylinders of the correct thickness indicate that these holes were almost certainly made by the potters by thrusting their fingers through the clay. Holes are irregularly spaced, typically between 5 and 15 cm apart, and randomly distributed from base to rim. The fired fabric is red on the exterior of the cylinder and dark gray to black on the interior. It is coarse and porous with many voids from burned out organic matter. A detailed of the fabric is description p?trographie in and the A. The study of the presented Appendix p?trographie analysis in the fabric show that the organic material is mostly chaff impressions (short lengths of grain stems remaining aswaste material after threshing). the molds from burned out chaff are voids in the shape of barley unidentifiable other plant remains, and one olive leaf (Fig. 7.2).* grains, in the fabric are not like the igneous rocks found in Stone fragments

Among

Fabric, which was used for most pottery at the site. They consist, instead, of carbonate, crystalline and cryptocrystalline quartz, and are in the local soils at minerals that inclusions phyllite fragments, typical and that also occur in the clay of Lakkos Ambeliou.2 The Chrysokamino the Mirabello

is close enough to similarity with the clay deposited at Lakkos Ambeliou was as source that this feature used the for the clay to suggest probably make the cylinders. of the fragments have a coating of dark, glassy, highly vitrified Many material on the interior side of their curved surfaces (see Chap. 10). This on rim sherds. It was once deposit is not usually present highly viscous, with drips, flow lines, and circular voids attesting to its former liquid state. On some sherds, the coating was once thicker than presently preserved, away, while on other sherds, the surface of the vitrified Small lenses of copper alteration products under 0.5 coating cm occur occasionally within the glassy coating. and it has broken

is smooth.

1. For discussion remains

preserved

of the botanical in the clay of these

chimneys, 2. Morris

see

Chap. 2002,

12. pp. 52-59.

Figure 7.2. Chimney fragment no. X 213, preserving the impression of an olive leaf, compared with a leaf from

a modern

olive

tree

THE

Figure

7.3. Alternative

reconstruc

tions of the industrial ceramics fragments: (a) a cylindrical furnace; (b) a cover for a bowl furnace; (c) a chimney for a bowl furnace

FURNACE

CHIMNEY

III

FRAGMENTS

THE FUNCTION OF THE INDUSTRIAL CERAMICS industrial ceramics fragments differ in several ways from the sherds Their fabric is distinctive, of clay vessels found at Chrysokamino. their are not those ofMinoan vessels, and their coating of glassy material shapes use. The presence of copper alteration is not consistent with household

The

products

in the vitreous

deposit

proves

the pieces played

a role in copper

metallurgy. The fragments were obviously part of some aspect of the smelting include the possibilities that the small frag process. Previous suggestions ments were parts of crucibles3 or that a they joined together to form pierced floor in a single large furnace.4 The restoration of the fragments as cylin ders invalidates these suggestions. The objects were freestanding, tapering a narrows the range of for cylinders, shape that considerably possibilities their use. Alternatives

are illustrated

in Figure

7.3.

Freestanding, cylindrical smelting furnaces (Fig. 7.3:a) have a long his in tory metallurgy. They were used inNigeria until theMedieval period5 and in Shaba Province, Congo called (formerly Katanga Province, Zaire), until more recently.6 The furnaces inNigeria and Congo were used for smelting, and they were only used once before being broken up into pieces.

3.Mosso p. 50. 4. Faure

1910,

p. 291;

Branigan

1968,

p. 47-48. 1983; Bernus

1966,

5. Bernus

and Echard

1985. 6. Forbes

1972,

7. Betancourt

pp. 26-27. et al. 1999,

p. 362.

8. For the involved, principles see the discussion in Olsen 1973, pp. 36-39.

were also The possibility that the Chrysokamino cylinders freestanding furnaces has been considered,7 but the evidence indicates that this theory is not tenable. If the cylinders were the furnaces inwhich the ore was smelted, smelting slagwould be found adhering to their inner surfaces, at least in their lower parts. The interiors do not have any such smelting slag attached. Covers (Fig. 7.3:b) are another alternative. Small bowl furnaces lose heat upward, and a cover to help hold the heat and increase efficiency would be an advantage inmaintaining since high temperatures. However, a number of rim can be in the of corpus fragments recognized fragments, the cylinders must have had open tops. are the third alternative are an Chimneys (Fig. 7.3:c). They important or of kiln furnace of also aspect design (and fireplaces), because proper venting is necessary in order to raise the temperature by increasing the flow of oxygen.8 In systems using only a natural draft, the exit venting system must be as large as the inlet system because if the outlet flue is restricted, combustion will be slowed, and efficiency will be lowered. If the system adds

CHAPTER

112

7

oxygen by employing artificial means such as bellows to increase the draft, then the venting system is even more critical. Tall chimneys increase the velocity of the upward draft, driving the temperature higher as long as

more

proper fuel is present and adequate air is forced into the firing chamber. At where we know that bellows were used, chimneys would Chrysokamino, have been a logical part of the furnace system. as is of the Chrysokamino cylinders chimneys on interior the of of the the further supported by presence glassy coating of hot vapors the fragments. This coating was formed by the condensation The

identification

a as they moved upward through the enough to leave deposit occur in any situation where the vitreous Such heating deposits cylinders. materials, and they are common in the up process volatilizes mineral-laden per parts of both lime and pottery kilns. The presence of cuprous inclusions that cooled

within

the coatings at Chrysokamino from the intense heating

were made

shows that these particular deposits that included copper.

of minerals

are not unique to Chrysokamino. In cylinders used for smelting the closest parallels are with ceramic cylinders from smelting the Aegean, are similar to the ones from sites on Kythnos.9The fragments from Kythnos Clay

they seem to come from tapering open rims (and possibly open bases). They aremade of coarse cm with holes pierced through the walls. The holes, often about 2-3 clay, the pierced in diameter, are spaced irregularly from rim to base. Evidently are part of a technology used for the production regional Aegean cylinders of copper.

Chrysokamino. cylinders with

Like

Reconstruction From with

the Cretan

of

the

pieces,

Chimneys

the curvature of the walls, the clay fragments must be from cylinders are present in the cm. Many holes ranging from 16 to 44 to a just below occurring in random pattern from just above the base

diameters

walls, the rim. The

holes would have allowed drafts of air to enter and exit. The is open. The bases were probably not closed, because although upper part a few of their angled edges can be identified, no fragments coming pieces from the bottom away from the edge are present. If the cylinders were as suggested by the glassy deposits on their interiors, then they chimneys,

had no bases. They must have been used over bowl furnaces built into the not been discovered.10 ground in locations that have are not known. While are it uncertain. Heights Some of the details is likely that the chimneys tapered slightly toward the top, with the rims the surviving fragments are all too having smaller diameters than the bases, small to establish their shapes with certainty. If pot bellows were used with the furnaces, the most efficient system would have been for the nozzles to be inserted into apertures of the bowl furnaces, not into their chimneys, because the oxygen should be directed at the fire, not above it.One tuyere was recovered from the excavations, indicating the workshop used the usual for conveying the draft from the bellows (or an earlier blowpipe) method to the fire (see Chap. 9). The interior diameters of the bowl furnaces may have been about the same as the maximum

diameters

of the chimneys.

The

largest diameter,

9. Y. Bassiakos 10. A is discussed

fragment in Chap.

(pers. comm.). of the furnace 9.

lining

THE

FURNACE

CHIMNEY

FRAGMENTS

113

ca. 44 cm, is near the maximum diameter of experimental kilns used in smelt was found to be effective could be size maximum The that ing copper.11 on the fact that airwould ca. 20 cm into a computed based only penetrate furnace loaded with a charge of copper ore and fuel, using an ore-to-fuel ratio of 2:1. This suggests that the furnaces were fairly small, and they would

not have produced and

Technological Chimney

an abundance

Economic

of the smelted metal.

Implications

of

the

Fabric

fabric of the chimneys is interesting for several reasons. It is unlike the fabric used for vessels at the site, indicating that its production was

The

characteristics are an incorporation of distinguishing matter of identifiable as chaff (Chap. 12), a use of organic large quantities low-fired, porous clay, and a brown to red color on the exterior combined specialized.

with

Its main

a darker color on the interior. Minoan

pottery employs many fabrics, but the addition of large quan is not a common practice for Bronze Age vessel tities of organic material an exception, see 74). Fabrics with large production in Crete after EM I (for amounts of organic matter are used mainly for objects intended towithstand straw is heat, such as crucibles and molds used in metallurgy.12 Chopped a porous particularly useful in fabrics of this class because it creates body that can withstand thermal shock, and similar recipes have been used for molds made in some parts of Asia even in the 20th century.13 The evidence from Chrysokamino shows that the general principles of manufacturing use of for objects metallurgical operations, including the specialized clay were a established is a recipes, already by very early period. The workshop a were of tradition skills that inherited part using specialized technological in later times. It was not an anomaly outside the general development of traditions. metallurgical fabrics also provide information on the relation of the metal to the local economy. The stone inclusions suggest the manufacture lurgy of the furnace chimneys was local. The metallurgists had access both to local clay and chaff derived from the harvesting of grain. As noted earlier, a local clay the petrography of the clay shows that it matches deposit at The

The cylinders must have been made shortly after the harvest, in the fall of the year, when chaff was available. Perhaps the work was seasonal, performed in late summer to early fall when the harvest was the seasonal north wind (the meltemi) was blowing fiercely, and completed, Lakkos Ambeliou.14

the autumn

rains had not yet begun.

CATALOGUE 11.

Bamberger

1985.

12. Cf. Blitzer 1995, pp. 504-508. 13. Forbes

1972,

p. 26.

14.App. A; for the local clay see Morris

2002,

chap.

3.

Rim 145

Fragments (X 103, from O 22-4) Furnace Diam.

rim chimney, fragment. of rim ca. 24.

Fig. 7.4

CHAPTER

114

i i

7

146

i a

147

145

'

i

'

I 151

150

148

155

153

152

156

154

D"l

1

1 158

157

159

r 161

1

i

i

160

7 162 Figure 7.4. Profile drawings of chim ney fragments (145-162). Scale 1:3

THE

FURNACE

A

coarse

FRAGMENTS

CHIMNEY

fabric with

from

voids

II5

burned-out

organic

matter

(unevenly

colored,

mostly black). Furnace and

exterior;

chimney vitreous

with

rim; straight on lower

deposit

(X 111, from O 22-4)

146

Furnace Diam. A

coarse

rim

chimney, of rim ca. 30. fabric with

three of

part

holes

below

the

interior.

the rim;

rough

interior

7.4

Fig. fragment.

from

voids

burned-out

organic

matter

(unevenly

colored,

mostly dark reddish brown, 5YR 3/2). Furnace ous

with

chimney

outturned

rim;

rough

interior

and

exterior;

no vitre

deposit.

(X 112, from O 22-4)

147

Furnace Diam. A

coarse

Fig.

7.4

rim chimney, fragment. of rim ca. 36. fabric with

from

voids

burned-out

organic

matter

(unevenly

colored,

mostly yellowish red, 5YR 4/6). Furnace chimney with straight rim; hole below the rim; rough interior and vitreous

exterior;

148

on

deposit

lower

(X 113, from Q20-2) Furnace Diam. A

coarse

Fig.

part

of

the

interior.

7.4

rim chimney, fragment. of rim ca. 36. fabric with

voids

from

burned-out

organic

matter

(unevenly

colored,

mostly yellowish red, 5YR 4/6). Furnace chimney with outturned rim; hole below the rim; rough interior and no vitreous

exterior;

149

deposit.

(X 198, from Q20-3) Furnace Diam. A

coarse

rim

chimney, of rim ca. 32. fabric with

Fig.

7.4

fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from reddish yellow, 5YR 6/6, to black). Furnace chimney with straight rim; hole below the rim; rough interior and no vitreous

exterior;

deposit.

(X 222, from R 20-2)

150

Furnace Diam. A

coarse

Fig.

7.4

rim chimney, fragment. of rim ca. 18. fabric with

voids

from

burned-out

organic

matter

(unevenly

colored,

mostly very dark brown 10YR 2/2). Furnace chimney with straight rim; two holes below the rim; rough interior and

exterior;

151

no vitreous

deposit.

(X 299, from M 20-6) Furnace Diam. A

coarse

rim

chimney, of rim ca. 54. fabric with

Fig.

7.4

fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

mostly yellowish red, 5YR 5/6). Furnace chimney with straight rim; hole below the rim; rough interior and exterior;

no vitreous

deposit.

CHAPTER

ii6

152

(X 319, from R 20-3) Furnace Diam. A

rim chimney, of rim ca. 34.

coarse

fabric with

7.4

Fig. fragment.

from

voids

burned-out

organic

matter

(unevenly

colored,

mostly black). Furnace

with

chimney

straight

rim;

Fig.

7.4

interior

rough

and

no vitreous

exterior;

deposit.

(X 320, from R 20-3)

153

Furnace Diam. A

rim fragment. chimney, max. of rim not measurable;

coarse

fabric with

voids

from

dim.

3.5.

burned-out

organic

matter

colored,

(unevenly

mostly dark reddish brown, 5YR 3/2). Furnace

with

chimney

rim;

straight

interior

rough

and

no vitreous

exterior;

deposit.

(X 321, from R 20-3)

154

Furnace Diam. A

rim

chimney, of rim ca. 24.

coarse

fabric with

7.4

Fig. fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from reddish yellow, 5YR 4/6, to very dark gray). Furnace and

chimney no vitreous

exterior;

155

with

Furnace

A

coarse

of

three

holes

below

the

rim;

rough

interior

deposit.

(X 426, from N 18-3) Diam.

rim;

straight

7.4

Fig.

rim chimney, fragment. rim ca. 24. voids

fabric with

from

burned-out

organic

matter

(unevenly

colored,

mostly dark gray). Furnace

chimney

with

rim;

straight

rough

interior

and

exterior;

no vitreous

deposit.

(X 427, from N 18-3)

156

Furnace Diam. A

coarse

rim

chimney, of rim ca. 26.

fragment.

voids

fabric with

7.4

Fig.

from

burned-out

organic

matter

(unevenly

colored,

from red, 2.5YR 5/6, to dark gray). Furnace

chimney

with

straight

rim;

rough

interior

and

exterior;

no vitreous

deposit.

(X 429, from N 18-3)

157

Furnace Diam. A

coarse

rim chimney, of rim ca. 26. fabric with

Fig.

7.4

fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from reddish yellow, 5YR 6/6, to black). Furnace

chimney

with

straight

rim;

rough

interior

and

exterior;

no vitreous

deposit.

(X 491, from P 20-2)

158

Furnace Diam. A

coarse

rim

chimney, of rim ca. 50 fabric with

Fig.

7.4

fragment.

(or from voids

from

a flat

area

burned-out

from yellowish red, 5YR 5/8, to black).

on

a

warped

organic

or matter

elliptical

artifact).

(unevenly

colored,

7

THE

FURNACE

FRAGMENTS

CHIMNEY

straight rim; hole below the rim; rough interior and

Furnace chimney with no vitreous

exterior;

deposit.

(X 500, from L 19-3)

159

Furnace Diam. A

rim chimney, of rim ca. 28.

coarse

II7

fabric with

7.4

Fig. fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from yellowish red, 5YR 4/6, to very dark gray). Furnace

chimney

with

rim;

straight

interior

rough

and

no vitreous

exterior;

deposit.

(X 640, from N 19-4)

160

Furnace Diam. A

7.4

Fig.

rim chimney, fragment. of rim ca. 32.

coarse

fabric with

voids

from

burned-out

organic

matter

(unevenly

colored,

from red, 2.5YR 4/6, to gray). Furnace

chimney

with

rim;

straight

interior

rough

and

no vitreous

exterior;

deposit.

(X 752, from N 20-4)

161

Furnace Diam. A

rim

chimney, of rim ca. 28.

coarse

fabric with

7.4

Fig. fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from yellowish red, 5YR 5/6, to very dark gray). Furnace ous

with

chimney

outturned

rim;

rough

and

interior

exterior;

no vitre

deposit.

162

(X 753, from N 20-4) Furnace Diam. A

rim chimney, of rim ca. 22.

coarse

fabric with

7.4

Fig. fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from yellowish red, 5YR 4/6, to dark reddish brown, 5YR 3/2). Furnace ous

with

chimney

outturned

rim;

rough

interior

and

exterior;

no vitre

deposit.

Base

Fragments

163

(X 100, from S 20-2) Furnace Diam. A

mostly

base chimney, of base ca. 42.

coarse

fabric with

fragment.

voids

Furnace deposit

with flat base chimney on the interior.

(X 110, from O 22-4) Furnace Diam. A

mostly

from

burned-out

organic

matter

(unevenly

colored,

red, 2.5YR 5/8, with a black core).

vitreous

164

7.5

Fig.

coarse

voids

from

from

of wall;

rough

interior

near

interior

and

exterior;

exterior;

base.

burned-out

organic

matter

(unevenly

red, 2.5YR 5/6).

Rough

and

7.5

Fig.

chimney, fragment of base ca. 35-40. fabric with

at bottom

vitreous

deposit

on

the

interior.

colored,

CHAPTER

ii8

7

ii t

J

?

165

164

163

i

i

/ i

J

169

168

167

166

i

ii

J

>l 172

171

170

II I i

I I

I II

I

I

175

174

173

i

?

i

I

O n

/

i

?

l

176

178

177 Figure ney

7.5. fragments

Profile

drawings

(163-178).

of chim Scale 1:3

THE

FRAGMENTS

CHIMNEY

FURNACE

(X 114, from Q^20-3)

165

Furnace Diam. A

base

fabric with

voids

Furnace vitreous

deposit

with flat base chimney on the interior.

(X 129, from S 20-2) Furnace Diam. A

from

burned-out

organic

matter

(unevenly

colored,

red, 2.5YR 4/6, with a black interior).

mostly

166

7.5

Fig. fragment.

chimney, of base ca. 44.

coarse

119

at bottom

of wall;

rough

interior

and

exterior;

7.5

Fig.

base fragment. chimney, of base not measurable.

coarse

voids

fabric with

from

burned-out

organic

matter

(unevenly

colored,

mostly red, 2.5YR 4/8, with black interior). Furnace chimney with flat base at bottom of wall; two holes; rough interior and

exterior;

vitreous

on

deposit

(X 205, from O 22-4)

167

Furnace

Fig.

base

chimney, ca. 22.

interior.

7.5

fragment.

of base

Diam. A

the

coarse

fabric with

from

burned-out

with flat base chimney on the interior.

at bottom

voids

organic

matter

(unevenly

colored,

mostly very dark gray). Furnace vitreous

168

deposit

(X 214, from R 20-2) Furnace Diam.

base chimney, of base ca. 44.

Fig.

of wall;

rough

interior

and

exterior;

7.5

fragment.

A coarse fabric with voids from burned-out organic matter (black). Furnace chimney with flat base at bottom of wall; rough interior and exterior; vitreous

169

on

deposit

the

interior.

(X 129, from N 20-2) Furnace Diam. A

base chimney, fragment. of base not measurable.

coarse

fabric with

voids

Furnace vitreous

deposit

with flat base chimney on the interior and

(X 739, from N 20-3) Furnace Diam. A

from

burned-out

organic

matter

(unevenly

colored,

red, 2.5YR 4/8, with a black interior).

mostly

170

7.5

Fig.

coarse

Fig.

at bottom

voids

from

rough

interior

and

exterior;

7.5

base fragment. chimney, max. of base not measurable; fabric with

of wall;

exterior.

dim.

burned-out

of

sherd organic

4.3. matter

(unevenly

colored,

from yellowish red, 5YR 5/6, to black). Furnace

with flat base chimney on the interior. deposit

vitreous

171

(X 188, from O 22-4) Furnace Diam. A

coarse

Fig.

base chimney, of base ca. 28. fabric with

at bottom

of wall;

rough

interior

and

exterior;

7.5

fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

mostly dark reddish brown, 5YR 3/2). Furnace one

hole;

chimney

vitreous

with

deposit

flat base on

the

at bottom

interior.

of wall;

rough

interior

and

exterior;

I20

CHAPTER

Body

Fragments (X 1757, from P 20-2)

172

Furnace dim.

Max.

chimney, 4.1.

coarse

A

body

fabric with

Furnace the

7.5

fragment.

voids

chimney

with

from

straight

burned-out

matter

organic

colored,

(unevenly

wall;

one

hole;

thick

unusually

vitreous

deposit

interior.

(X 105, from O 22-4)

173

Fig.

red, 2.5YR 4/8).

mostly on

7

Furnace dim.

Max.

chimney, 7.6.

coarse

A

body

fabric with

voids

mostly red, 2.5YR 5/8). Furnace chimney with

Fig.

7.5

fragment.

from

burned-out

matter

organic

colored,

(unevenly

on the in straight wall; one hole; vitreous deposit

terior.

(X 106, from O 22-4)

174

Furnace Max.

coarse

A

body

chimney, dim.

Fig.

7.5

fragment.

7.1. voids

fabric with

from

burned-out

matter

organic

colored,

(unevenly

red, 5YR 4/6).

mostly

Furnace

with

chimney

straight

wall;

one

vitreous

hole;

deposit

on

the

in

terior.

(X 107, from O 22-4)

175

Furnace

A

chimney, dim.

Max.

coarse

body

Fig.

7.5

fragment.

7.3. fabric with

voids

from

burned-out

matter

organic

colored,

(unevenly

red, 2.5YR 4/6, to gray).

mostly

Furnace

with

chimney

straight

wall;

two

vitreous

holes;

deposit

on

the

in

terior.

(X 108, from O 22-4)

176

Furnace

coarse

A

body

chimney, dim.

Max.

Fig.

7.5

fragment.

7.3. voids

fabric with

from

burned-out

organic

matter

(unevenly

colored,

mostly yellowish red, 5YR 4/6). Furnace

with

chimney

straight

three

wall;

holes;

vitreous

deposit

on

the

interior.

(X 109, from O 22-4)

177

Furnace

chimney, dim.

Max.

coarse

A

body

Fig.

7.5

fragment.

8.4. fabric with

voids

from

burned-out

organic

matter

(unevenly

colored,

mostly very dark gray). Furnace

chimney

with

straight

wall;

interior.

178

(X 120, fromM Furnace Max.

chimney, dim.

5.9.

19-2) body

Fig. fragment.

7.5

thickened

area; vitreous

deposit

on

the

THE

CHIMNEY

FURNACE

coarse

A

fabric with

voids

FRAGMENTS

from

121

burned-out

matter

organic

colored,

(unevenly

mostly yellowish red, 5YR 5/6). Furnace chimney with straightwall; two holes; unusually thick vitreous deposit on

interior.

the

179

(X 130, from S 20-2) Furnace Max.

chimney, dim.

coarse

A

body

Fig.

7.6

fragment.

4.2. fabric with

voids

from

burned-out

matter

organic

colored,

(unevenly

mostly reddish yellow, 5YR 6/8), with a dark gray interior. Furnace chimney with straight wall; one hole; vitreous deposit on the in terior.

180

(X 140, from S 20-2) Furnace

chimney, ca. 29.

Diam. coarse

A

body

fabric with

voids

7.6

7.1,

Figs. fragment.

from

burned-out

matter

organic

colored,

(unevenly

from reddish brown, 2.5YR 5.4, to black). Furnace

with

chimney

wall;

straight

six holes;

vitreous

on

deposit

the

in

terior.

181

(X 215, from S 20-2) Furnace Max.

chimney, dim.

coarse

A

body

Fig.

7.6

fragment.

3.9. fabric with

voids

from

burned-out

matter

organic

colored,

(unevenly

from dark grayish brown, 10YR 4/2 to black; light red, 2.5YR 6/6, on exterior). Furnace

with

chimney

wall;

straight

one

vitreous

hole;

on

deposit

the

in

terior.

182

(X 310, from M 20-3) Furnace Diam. coarse

A

7.6

Fig.

chimney, body fragment. not measurable; max. dim. fabric with

voids

from

5.9.

burned-out

organic

matter

(unevenly

colored,

from reddish yellow, 5YR 6/8, to dark gray). Furnace 2.3);

rough

183

with somewhat chimney straight wall; hole no vitreous interior and exterior; deposit.

(X 323, from R 20-3) Furnace Diam. A

coarse

chimney, ca. 27-30. fabric with

body

voids

Fig.

larger

than

usual

(Diam.

7.6

fragment.

from

burned-out

organic

matter

(unevenly

colored,

from light red, 2.5YR 6/6, to black). Furnace

chimney

with

straight

wall;

three

holes;

vitreous

deposit

on

the

interior.

184

(X 324, from R 20-3) Furnace Diam.

chimney, ca. 30.

body

Fig.

7.6

fragment.

A coarse fabric with voids from burned-out organic matter (unevenly colored, from light red, 2.5YR 6/6, to black). Furnace

chimney

with

straight

wall;

six holes;

vitreous

deposit

on

the

inte

CHAPTER

7

^ I

K J

I I/

1

181

179

i

ii

v

J

i i

i

I

i

184

183

182

n

f

\

i

i

Ol 186

185

187

188 Figure ney

7.6.

fragments

Profile

drawings

(179-188).

of chim Scale 1:3

THE

FRAGMENTS

CHIMNEY

FURNACE

(X 1806, fromM 22-2)

185

Furnace dim.

Max. A

chimney, 6.8.

coarse

body

fabric with

Fig.

123

7.6

fragment.

voids

from

burned-out

organic

matter

(unevenly

colored,

from red, 2.5YR 5/6-8, to dark gray). Furnace chimney with straight wall; three holes; vitreous deposit on the interior.

186

(X 101, from S 20-2) Furnace

chimney,

Diam.

of artifact

coarse

A

Fig.

body fragment. not measurable;

fabric with

voids

mostly red, 2.5YR 5/6). Furnace chimney with

from

7.6

max.

dim.

burned-out

5.8. organic

matter

(unevenly

colored,

straight wall; two holes; vitreous deposit on the in

terior.

187

(X 104, from O 22-4) Furnace

A

chimney, dim.

Max.

coarse

body

Fig.

7.6

fragment.

9.6. fabric with

voids

from

burned-out

mostly yellowish red, 5YR 5/8). Furnace chimney with straight wall;

organic

matter

(unevenly

colored,

three holes; vitreous deposit on the

interior.

(X 1756, from P 20-2)

188

Furnace Max. A

dim. coarse

chimney, 9.4. fabric with

body

voids

Fig.

7.6

fragment.

from

burned

out

organic

matter

(unevenly

mostly light red, 2.5YR 6/8). Furnace chimney with straight wall; four holes; unusually deposit

on

interior.

colored,

thick vitreous

CHAPTER

8

Bellows

Pot

The

byPhilip P. Betancourt and James D. Muhly

is a device used to provide a forced draft for forges, metal in which high temperatures are lurgical furnaces, and other installations in the eastern Mediterranean desired. It was used extensively during the

The

pot bellows

a by several authors.1 It consists of a flexible leather cover cylindrical ceramic vessel fitted with having one-way cover up and down in air through the the draws alternately flaps. Moving

Bronze Age,

and it has been discussed

out through a nozzle. The device iswell known both flaps and forces it from illustrations in Egyptian paintings and from actual examples surviv record. Modern ing in the archaeological examples with similar designs are known from Asia and Africa where a role in they play metalworking and other crafts.2 Before

the discovery of the bellows from Chrysokamino, only known from later periods.3

were examples earliest use of the pot bellows is not known. Although the bellows themselves do not survive, the device has been suggested for smelting op erations as early as the Chalcolithic.4 Other than the finds from Chrysoka

Cretan

The

mino, the earliest known examples are from the beginning of theMiddle Bronze Age.5 a from the Tomb of Figure 8.1 shows painting of the 18th Dynasty at Thebes.6 Metalworkers Rekh-mi-re here use the bellows to melt cop per in small crucibles. The bellows are used in pairs so that alternation in their pumping action can produce amore steady draft. The devices in this painting consist of cylindrical vessels with nozzles to allow the attachment

1.Davey 1979;Tylecote 1981; Forenhaber pp.

106-107, 2. Forbes 3. Blitzer

Kommos;

1994; M?ller-Karpe 76. fig. 1950, p. 116.

1994,

can be lifted with from

pp. 508-509,

1995,

1997,

Dimopoulou

p. 435,

from Poros; Evely 2000, p. 363, from Zakros

and Palaikastro.

4. Hegde and Ericson 1985, p. 66; Rothenberg

1985,

p. 124.

1994,

pp.

106-107,

fig. 76. 6. For pi.

17.

the context,

see Davies

the air to the fire. A

leather cover over each vessels

a cord to inflate the bellows

on the leather forces the air stepping through shown for each small fire.

rim

and suck in the air, while the nozzle. Four bellows are

of a minimum of nine pot bellows come from Chryso Fragments kamino. All examples are presented here (see the catalogue in this chapter). are somewhat different in They design from the bellows known from other sites, although the principle of operation is the same. Perhaps they can be as regarded early, experimental examples of the device. are The bellows from Chrysokamino clay cylinders with drum-shaped

5. Davey 1979;Tylecote 1981; M?ller-Karpe

of a reed to conduct

1943,

closed

tops, with per surfaces. The

one or more design

large holes rather crudely cut in the flat up of the lower edge (i.e., the open part facing down

CHAPTER

126

Figure

8.1.

of workmen

Drawing

using bellows, from the Tomb of Rekh-mi-re.

After

Davies

1943, pi. 23

the object was in use) is not certain, but sherds with straight rims, are between 28 and 53 cm. such as 191, maybe from this shape. Diameters A nozzle iswell preserved on bellows 189 (see Figs. 8.3 and 8.4 below).

when

The

nozzle

is attached

on the upper part of the near the hole cut cylinder,

in the upper surface. The bellows all show signs of darkening from a hot, reducing atmo a sphere in their interiors. This characteristic is confirmation of their iden tification, showing that the devices were used before they were broken. The act of pumping air into a furnace was evidently not so efficient that it drew in only cold air to pump into the fire. Some of the hot air and some sparks seem to have been drawn back out of the furnace and into the bellows, a hot, inside the bellows and resulting in reducing atmosphere producing a dark color in the clay, as its red ferric oxide was altered to black (Fe203) oxide (FeO). The process that creates black iron oxides in clay fabrics by the chemical change called reduction iswell known.7 Contact ferrous

the hot, oxygen-depleted atmosphere accounts for the condition of the interiors; this heat is one of the reasons why these devices were made of clay instead of a flammable material.

with

of fused mud are on the exteriors of several bellows, always Deposits away from the end with the holes cut in it. Sherd 191, which may be a rim from a bellows, is one of the pieces with fused mud on its exterior. It is were in use, and were set into the ground when they likely that the bellows the join with the ground was sealed with mud to prevent air from escaping so that the mud baked and fused to the vessel during the pumping action, use. by the heat drawn into the bellows from the furnace during are from EM III? All the bellows fragments from Chrysokamino MM IA. The date is based on several considerations, with the most im portant one being that pieces are found within the soil used for the floor of the small hut, in association with White-on-Dark is found anywhere on the site.

Ware.

No

later pottery

7. Betancourt 1988,

p. 80.

1985,

p. 6; Noble

THE

8.2. Reconstruction

Figure bellows

in use

based

made

by Harriet Lyla Pinch Brock

on

Blitzer.

POT

BELLOWS

127

of the pot suggestions Drawn by

The bellows cannot be earlier in date because they are all made from a type of Mirabello Fabric consisting of a local clay paste with abun dant large rock fragments, a fabric not used before EM III-MMIA. Mira bello Fabric uses numerous fragments of angular white stone in the grano diorite/diorite series.8 It has several subclasses based on the coarseness of and their percentage in the fabric, ranging from a fine fabric a few inclusions used for Vasiliki Ware9 to a very coarse class just used for storage containers. Deposits of this igneous rock occur at several locations at and near Gournia,10 and the pottery using it has been analyzed the inclusions

with

several times.11 The as defined

igneous rock analyzed in the pottery from Pseira is and Guilford12 because it Donaldson, by MacKenzie, consists of plagioclase, amphibole, and biotite, with very little quartz, while the material analyzed from Mochlos is called granodiorite by Day, Joyner,

diorite

and Relaki.13 of the bellows from Chrysokamino suggests a reconstruction a leather bag with slits and flaps attached at the top, and the down and buried in the earth, sealed with mud to prevent air from escaping during use and to prevent the bellows from dislodging Blitzer

as a pot with "rim" upside

the leather bag up and during the pumping operation (Fig. 8.2).14Working down would pull in air through the one-way flaps and force it through the nozzle

and

into

the

furnace.

The widely distributed bamboo called kalami, which grows in many parts of Crete, has stems with a hollow interior that would fit the size of the aperture of the tuyere found at the site (see Chap. 9). The presence of the plant onMinoan pottery and wall paintings shows that it grew in Crete during the Bronze Age.15 It would have made a suitable conduit to allow air from the bellows to flow to the furnace, where the clay tuyere would the hollow

protect

8.

and Mook

Haggis

fabric court

1995, 9. Myer

p. 268. 10.

1993,

p. 273, and Betan

2; Myer,

Mclntosh, pp. 144-145. 1979; Whitelaw

et al. 1997,

11.Myer 1979; Day 1991, pp. 91 101; Myer,

Mclntosh

1995,

144-145;

Relaki

pp.

2003,

pp.

12. MacKenzie, Papastamatiou

et al. 1959.

stem from the heat.

and Betancourt Day, 17-18.

Joyner,

Donaldson,

Guilford 1982, p. 103.

and

pp.

13. Day, 17-18. 14. H.

Joyner,

and Relaki

Blitzer

(pers. comm.). see pottery, Popham and c; for wall painting,

15. For and

2003,

pi. 76:a Platon 2002,

pi. 48:a.

1967, see

CHAPTER

128

8

CATALOGUE lists all bellows and probable bellows fragments found at the into 15 catalogue entries combining pieces similar in thick site, grouped ness, fabric, shape, and general appearance. In many cases one cannot tell which sherds belong together, but aminimum of nine bellows is estimated The

catalogue

from this assemblage. 189

(X 143, from L 17-3, L 18-2, L 19-2, M 18-2, M 18-3, M 19-3, M 19-5, N 19-2, N 19-3, P 19-1, R 20-1, locus E5050N5016)

Figs. 8.3, 8.4

Pot bellows, large part of side and closed top and many nonjoining Diam.

of

top

Fabric (unevenly colored, red to black, with the interior darkened

Mirabello a

from

sherds.

ca. 37.

reducing

atmosphere).

Cylindrical vessel with almost straightwalls and closed top (lower part ofwalls

190

at side,

nozzle

missing);

near

closed

top;

rectilinear

in closed

hole

top.

(X 609, fromM 19-5) Fig. 8.4 Pot bellows, sherd from closed top. Max.

dim.

of

18.9.

sherd

Fabric (red, 2.5YR 4/8 on interior; dark brown, 7.5YR 4/2, on

Mirabello exterior).

Sherd from the closed top, from near the center; the edges of two long holes survive. This

Comments: than

a

191

single

sherd cut

opening

demonstrates

that

(X 420, from N 19-3, N 20-3) Jar

or

pot

Diam.

of

bellows, rim

the upper

part

sometimes

had more

in it.

rim

8.4

Fig.

sherd.

ca. 38; Diam.

of body

ca. 47-50.

Mirabello

Fabric (reddish yellow, 5YR 6/6). rim. Straight Comments: Burned soil fused to the body on the exterior. The

the

exterior

of

the vessel

than

a

192

(X 317, fromM andM 19-5)

this

suggests

artifact

may

have

been

a

pot

fused soil on bellows

rather

jar.

simple

18-1, M

18-2, M

18-3, M

19-2, Fig.

8.4

Pot bellows, body sherds. Diam.

of body

ca. 47-48.

Mirabello Fabric (unevenly colored, red, 2.5YR 4/8, to reddish brown, 2.5YR 5/4, to black; dark in the interior from a reducing atmosphere). Cylindrical vessel with almost straight walls. Comments:

193

Burned

(X 531, fromM

soil

fused

18-3) Fig. Pot bellows (?), body sherd. Diam.

of

Mirabello

body

to the body,

on

the exterior.

8.4

ca. 40.

Fabric (gray core; red surface, 2.5YR 5/6).

THE

POT

BELLOWS

129

190

189 (1:5)

191

1 1

192

193

Figure 8.4. Profile drawings of pot bellows (189-194). Scale 1:3unless otherwise

indicated

194

CHAPTER

130

194

(X 418, from M Pot

bellows, of

Diam.

reducing

sherds

many

19-3, N 18-1, N 18-2, and

from walls

closed

Fig.

8.4 N 19-2, N 19-3)

top.

ca. 28.

body

Fabric (black to red, 2.5YR 4/8; extremely dark in the interior from

Mirabello a

18-3, M

8

atmosphere).

Cylindrical vessel with almost straight walls; rectilinear hole in closed top. (X 294, from N 18-2, N 19-2)

195

Fig.

8.5

Fig.

8.5

Pot bellows (?), body sherd. Diam.

of

ca. 48.

body

Mirabello

Fabric (red, 2.5YR 5/8).

(X 293, fromM

196

18-2, M

18-3)

Pot bellows (?), body sherd. Diam.

ca. 45.

of body

Mirabello reducing

atmosphere,

in the

especially

(X 483, from P 20-2)

197

Pot

sherd

bellows,

Diam.

to almost completely black from a

Fabric (very dark gray, N3,

of closed

interior).

8.5

Fig.

the upper

from

part.

ca. 36.

end

Fabric (brown, 7.5YR 5/4; dark in the interior from a reducing

Mirabello atmosphere).

Cylindrical vessel with almost straight walls; closed top with hole in it. 198 M

(X 416, from K 17-1, L 19-2, M 19-2b, M 19-5, N 18-2, N 10-2, N 19-3)

Two Diam.

or more rim

of

of made pot bellows ca. 20; restored Diam.

similar of

19-3,

8.5

Fig.

rim and clay, body ca. 52-53.

sherds.

body

Mirabello Fabric (unevenly colored, from dark red, 2.5YR 3/6, to reddish brown, 2.5YR 5/4, to dark brown, 7.5YR 4/2, to light red, 2.5YR 6/6). Burned

Comments:

on exterior;

soil fused

one

sherd with

with a diameter very different from the body sherds. The would have helped seat the bellows in the ground. (X 1712, fromM

199

19-3)

rim, heavily

reduced,

ridge below the mouth

8.5

Fig.

Pot bellows, body sherds. Max.

dim.

Mirabello

M

largest

Burned

Comments:

200

of

sherd

6.2.

Fabric (red, 2.5YR 4/6). soil

fused

on

exterior.

(X 142, from L 17-3, L 19-2, M 18-3, M 19-2b, 19-3, M 19-4, M 19-5, N 18-2, N 19-2, N 19-3)

Pot

bellows, of rim

Diam.

rim, base,

and body

ca. 18; Diam.

8.5

Fig.

sherds.

of base

ca. 20;

restored

Diam.

of body

ca. 36.

Mirabello Fabric (unevenly colored, from red, 2.5YR 5/6, to reddish brown, 2.5YR 5/4, to dark reddish gray, 5YR 4/2). Comments: Hole cut in top; darkened from reduction in the interior; burned soil

201

fused

on

exterior.

(X 1713, from N 19-3) Pot

bellows,

rim

sherd.

Fig.

8.5

Figure 8.5 (opposite). Profile drawings of pot bellows (195-203). Scale 1:3

THE

195

196

POT

BELLOWS

131

197

199 1

1

198

1

200

201

J

202

1

203

1

1

CHAPTER

132

Diam.

Comments:

202

ca. 28.

of rim

Mirabello

Fabric (dark brown, 10YR 3/4). Burned

soil

fused

(X 183, from N 19-2) Pot Diam.

8

rim

bellows,

on

exterior.

8.5

Fig.

sherds.

ca. 48.

of rim

A coarse fabric containing phyllite and other stones (red, 2.5YR 5/8). Horizontal molding near the thickened rim. Comments:

Burned

soil

on

fused

the

exterior.

This

object

an atypical shape in comparison with atypical fabric, and it is regarded was shape

as bellows used

fragments. as a bellows. The

sherd

The

molding

that more suggests on the exterior would

is made

than have

one

ceramic

helped

the bellows in the ground during use. 203

(X 1666, from N 19-2)

Fig.

8.5

Pot bellows, body sherd. Max.

dim.

Mirabello Comments:

13.9.

Fabric (yellowish red, 5YR 5/6, dark on interior and core). Burned

soil

fused

on

the

exterior.

an

from

the other sherds seat

Miscellaneous

Ceramic

Artifacts by

Susan

C. Ferrence

and

Byron

Koukaras

Several ceramic objects from Chrysokamino do not fit easily into the other were found within the hut, in items discussed this volume. Two categories the others were discovered

in the slag area. None of the objects were were in their original place of use; they all fragments found in casual debris. to various (see categories They belong Fig. 9.1 below).

while

of the catalogued items, a tuyere and a piece of furnace lining, for evidence the reconstruction of the metallurgical activi provide important to our knowledge ties at the site. They make a significant contribution of the workshop. Some of the other objects, however, have unusual characteris tics that do not correspond with any of the pottery groups or industrial Two

at the site, and they are difficult to explain. The objects are all from the latest stratum (an EM III-MM IA context), but the poorly a an unusual vessel, is very handmade fashioned, object 204, possibly piece of classes found

crude for the latest phase of the Early Bronze Age. It may be an heirloom was context. in its archaeological produced long before it deposited

OBJECTS FROM THE APSIDAL HUT The

fragmentary pieces of 204 (X 272) consist of rim and body sherds to a vessel with a maximum rim diameter of 22 cm. The most belonging vase is its chaff-tempered, distinctive characteristic of the fine-textured fabric. It is a soft, pinkish white fabric that erodes easily into a very fine, an chalky powder. The vessel also has extremely irregular, lumpy surface. It as demonstrated in the coil method, by its obvious breakage too is coil lines. The porous to hold liquid. along object probably This fabric is unique for the site, and it is difficult to determine what function the vessel served at this industrial location. The shape cannot be is handmade

reconstructed.

Both

the fabric and the technology of manufacture differ from the other artifacts at the workshop and also from other

completely pottery known from this part of Crete. This as an

piece should surely be regarded

import.

Another object with

unusual a

consisting of two pieces of an body approximately 47-48 cm in diameter. The

item is 205

cylindrical

(X 1667),

CHAPTER

134

9

in its sides (like the furnace chimneys), but it ismade in a hard, dense clay fabric, not the porous, soft fabric filled with chaff used for the usual chimneys. It iswell fired to a uniform red color, but it shows no no evidence that itwas ever used as a signs of fire chimney. It displays or burning. item has holes

OBJECTS FROM THE SLAG PILE away from the hut may have had some function in the activities. Two of the artifacts, a fragment of furnace lining and workshop a a were tuyere, certainly used in the smelting. The other objects piece of are a sherd possibly fired accidentally and a unique triangular artifact of Objects

unknown

found

function.

Fabric, (X 614) is a severely burned piece of Mirabello on one side. The smooth, bubbly surface of the curved but flat mostly sherd indicates itmay have been near or trapped inside a burning furnace. It could be a sherd that was accidentally fired by one of the furnaces. Number

206

a object 207 (X 940) is triangular block of well-fired Fabric. It comes from the surface of the slag pile. The artifact Mirabello is 7.7 cm long on its hypotenuse and 1.3 cm thick. The triangular object, a was cut of clay before being fired to produce a from slab which carefully hard, dense object, has no marks from use. Its purpose is unknown. The

anomalous

208 (X 1755) is a tuyere fragment that was found in the slag end of the hut. It would have been used as the incombustible outside pile a to a hollow reed or the furnace. pot bellows piece of bamboo connecting in is only partly preserved, was originally cylindrical The tuyere, which an aperture with a a center. in It had hole with the longitudinal shape diameter of ca. 2.5 cm. Number

discusses tuyeres in detail, and he lists examples from many Tylecote The use of a tuyere is essential ancient sites in the Eastern Mediterranean.1 in operations involving pumping air to a furnace using combustible conduits because something must protect the end of the hollow reed from the in tense heat of the furnace. The clay tuyere would be fitted to the end of the reed, and only itwould come in contact with the furnace. In comparison the example from Chrysokamino the tuyeres illustrated by Tylecote, has a relatively small exterior diameter and a relatively thin wall, but its hole is the proper size for passing the draft from a bellows or a blowpipe to the furnace. Its 2.5 cm diameter is exactly in the center of the 2-3 cm for a tuyere's hole in order size range suggested as the optimal dimension

with

to pump sufficient air into the fire. Number 209 (X 1869) is a fragment of a coarse clay object that was one of the bowl furnaces. It has no finished probably part of the lining for sides, and its thickness (over 4 cm) indicates that it cannot have vessel. The piece has been fired extensively, and been part of any Minoan it has a black exterior color from complete reduction, a feature that would exterior

be expected if it was buried in the ground, lining not in contact with oxygen while itwas heated.

was

on all sides.

a bowl

furnace, and is broken roughly It 1. Tylecote

1981.

CERAMIC

MISCELLANEOUS

ARTIFACTS

135

207 208 209 Figure

9.1. Miscellaneous

(204-209).

Scale 1:3

artifacts

C?TALO

GUE

Objects

from

the

Hut

Apsidal

(X 272, from N 18-2, N 18-3, N 19-2, N 19-3)

204

9.1

Fig.

Vessel (?), rim and body sherds. Diam.

of rim 22.

A chaff-tempered, fine fabric (pinkish white, 5YR 8/2). a shallow

Perhaps Comments:

Very

bowl. crudely

made

by hand

and broken

at the places

where

coils

joined. 205

(X 1667, from N 18-1, N 19-3)

Fig. 9.1 vessel or object, two body sherds.

Unidentified Diam.

of

body

ca. 47-48.

A coarse, hard fabric (red, 2.5YR 4/8). Cylindrical form with holes thrust through the body before firing. Comments:

The

artifact

is similar

to the

furnace

chimney

fragments

except

that the fabric is very different and the item was never subjected to fire after it was

made.

CHAPTER

i36

206

the

from

Objects

(X 614, from L 17-1) dim.

9.1

Fig.

curved lump of clay with a flat side.

Elongated Max.

Pile

Slag

5.1.

Fabric (dark gray, 2.5YR 4/0). Comments: Severely burned and vitrified.

Mirabello

207

Fig. 9.1

(X 940, from J 17-Surface) Triangular

block,

Max.

5.0

dim.

complete. x 4.5 x 7.7;

1.4.

thickness

Fabric (red, 2.5YR 5/6). Triangular object with a slightly convex profile.

Mirabello

208

(X 1755, from P 20-2) Tuyere Diam.

fragment. of exterior

ca. 4; Diam.

Fig.

9.1

ca. 2.5;

of hole

original

length

unknown

at both ends). A coarse fabric (red, 2.5YR 5/6). Cylindrical profile. 209

(X 1869, from P 20-Cleaning) Lining Max.

from p. dim.

a bowl

furnace,

Fig.

fragment,

9.1 broken

on

all sides.

7.8.

A coarse fabric, highly fired in a reducing atmosphere (black).

(broken

9

CHAPTERIO

Metallurgical

Other Materials

byPhilip P. Betancourt

different materials discovered at Chrysokamino Many provide evidence for the activities carried out at the site. The excavation strategy was de awide range of manmade artifacts as well as geological signed to collect on the samples and other items that could contribute information history of the workshop uncatalogued

and the technology used there. This chapter presents the ore materials, metallurgical including slag, fragments, and

copper prills.

THE SLAG slag has been applied to awide variety of vitreous materials found in the archaeological record.1 It has been used for glassy waste products from several different processes including the production of lime, the fir the field ing of ceramics, the making of faience, and metallurgy. Within

The word

of metallurgy, the term has been applied to the residue from smelting, on various surfaces. in crucibles, and condensation remelting Vitreous materials resulting from smelting and other metallurgical practices have sometimes been divided between furnace slag (formed in the furnace), tap slag (formed when the furnace is tapped to remove mol ten metal), and crucible slag (formed in a crucible). all of these Although share the characteristic of being vitreous, they can differ from materials one another in their texture. and microscopic composition by Description formation process based on visual examination and on the archaeological context is useful for and describing the technology, but it understanding can be uncertain or impossible when only tiny fragments survive or when the context of the slag is unknown. An understanding of the original com can best be done to the and that led the final processes position products by analyses of various

types that demonstrate

both

the chemistry

and the

microstructure.

The

1. Bachmann dockl995,p.20.

1982,

pp.

1-6; Crad

slag resulting from smelting operations to obtain copper is regu in color and fairly heavy from inclusions of iron oxides, such as dark larly it usually occurs in larger quantities than slag magnetite. Archaeologically, resulting

from remelting

of copper

in crucibles. Estimates

have suggested

CHAPTER

i38

IO

Figure 10.1. Lumps of slag from unit M 22-3

that smelting results in quantities greater than the amount of metal

of slag that are between that is produced.2

10 andl5

times

fall into two broad classes. Some pieces The slags at Chrysokamino are loose, thick chunks that may or may not preserve an original smooth are on surface. Other examples surface coatings the interiors of furnace are dark and vitreous, (Fig. 7.1, right). Both classes have prills of copper in them. The surface coatings on the chimneys are thin (usually under 0.5 cm in thickness), and they have a smooth surface. The loose chunks, on the other hand, are thicker, solid pieces of glassy material (Fig. 10.1). It is likely that the solid pieces of slag chimney fragments and they sometimes

resulted from the melted

charge inside the furnace, and the coatings inside the chimneys formed partly from the condensation of vapors and partly from small liquid droplets and solid particles carried upward and deposited inside the chimneys

The

during

Chrysokamino

Before

Slag

process.

Deposit

a began, deposit consisting mainly of the saddle that forms the metallurgy

the excavation

the surface of most

the smelting

of slag was on site at Chryso

kamino

the slag formed the largest component, the (Fig. 3.2). Although also included clay chimney fragments, pieces of natural phyllite, deposit a few bits of ore, and the soil resulting from the phyllite s disintegration, various artifacts (pottery, stone tools, animal bones, shells, and other items in this volume). The small fragments of slagwere densely packed, and tiny bits of pulverized slag were also present. Presumably, most of the was broken slag by human activities. Freeze/thaw cycles during winter weather might also have broken some of the vitreous material, but breakage from the freeze/thaw cycles would have been most intense at the surface, so the fact that deeply buried parts of the deposit consisted of the same small as the surface suggests that natural was minimal. fragments breakage The deposit was low and wide, extending from sterile bedrock at the east and west of the a ca. 60 cm near the deposit to maximum height of discussed

center of the saddle a north-south (Fig. 10.2). It covered depression in the on west between of limestone the and phyllite exposures landscape higher on the east. The upper surface of the was level before excavation deposit (i.e., it did not form amound, but consisted instead of a flat area between the two higher exposures of bedrock). The original size of the slag pile could not be determined

because it had partly eroded into the sea at the and the south where the saddle terminates in cliffs, but it covered an area of ca. 200 m2. An estimated 60 to 70 m3 of was present. slag

north

2.Evely2000,p.353.

OTHER

Figure looking

10.2.

Slag

resting

METALLURGICAL

MATERIALS

139

on bedrock,

south

The phyllite on the east was extremely altered, and at the surface much of it had disintegrated into a white calcareous mass that crumbled easily. White material from this phyllite, which had washed downhill powdery over much

of the area, looked very much like lime, leading to the under was once the site standable but incorrect suggestion that Chrysokamino of a limekiln.3 The nearly white soil created a covering of dust that tended to adhere to everything. It made the surface of the slag pile appear paler was in actual composition; than it after the slag was washed, most of itwas almost black.

Vitreous

The

on

Coatings

the

Chimneys

of the fragments of clay furnace chimneys had dark, glassy material seems adhering to their interior surfaces (Fig. 7.1, right). The material to have covered most of the inside of the cylindrical chimneys, although it did not reach up to the open upper rims. The thickness of the coating

Most

was

usually

dence

less

than

0.5

that itwas once

cm. The

was

coatings

surface

was

smooth,

in surface flow

and

evi

structure and

preserved liquid drips. The orientation of the drips proved that the cylinders were used in a vertical fashion, with their open rims at the top, as is to be expected in artifacts interpreted as cylindrical chimneys. The vitreous material inside the chimneys seems to have formed mostly as a result of condensation of hot vapors, although liquid droplets and small solid particles carried up ward inside the chimneys by the force of the draft probably added to the condensation collected while the smelting operation was in were cooler than the furnace, and the because the progress chimneys coating solidified as soon as its temperature dropped sufficiently. indicates that at the microscopic level the material is com Analysis a numerous of matrix iron inclusions of the oxide posed glassy containing as occurs A and dendritic B).The magnetite magnetite (Apps. crystals that formed when the glassy material cooled quickly. A few very tiny grains of quartz are present as relics of the original charge. A copper chloride rec one can be an altera ognized in sample (App. B, CHR 93) interpreted as tion product of a tiny amount of copper deposited along with the glassy material. The

3. See Zois

1993.

material.

CHAPTER

140

The

Chunks

of

IO

Slag

in small pieces under 2 cm in size. solid chunks of slag occur mostly is pulverized. Because the pieces form a continuous Much of the material The

to cm or larger in series from microscopic particles fragments about 2-3 their greatest dimension, the total number of fragments cannot be calculated except by an estimate, which would be in the millions. The surfaces have different appearances depending on the degree of weathering. Weathered are dark, vitreous masses with dull, black exteriors. pieces By contrast, can be seen on fresh breaks. When the bright, highly reflective surfaces it is smooth and often original exterior surface of the slag is preserved, rounded (Fig. 10.1). A few rare stalactitic pieces demonstrate that some of the slag reached itsmelting point during the smelting process (Fig. 10.1:a). most of the slag is dense and heavy, a few pieces exhibit con Although siderable variation in physical characteristics. A little lightweight, "frothy" material

is present.

slags have been analyzed by a number of different laboratory tech some cases it has been niques (see the individual appendices). In possible to more than one so that the results analyze specific samples by methodology, one another. The can consists of a glassy complement typical microstructure matrix containing crystalline phases. Among these microscopic inclusions, as is common in ancient copper are numerous, magnetite crystals especially The

are often dendritic structure (the branching smelting slags.4 The crystals in This of is CHR inclusion is quite apparent 87). type App. A, typical of as opposed to those from slags derived from smelting oxidic copper ores, or more which of have lower iron con pure metal, remelting scrap ingots forms from the iron oxides that act as a flux during tents.5 The magnetite the smelting operation, and it contributes substantially to the black color of the final vitreous material. The composition shows and heavy weight that some of the slag cooled slowly enough for pyroxene group minerals to crystallize, as they are recorded as inclusions within the microstructure are fayalite crystals (App. A, (App. A, CHR 96, CHR 99). Also present CHR 99), another feature of slags from oxidic copper ores.6 Circular to are present from elliptical voids entrapped gas bubbles. In addition to the are other minerals, copper prills present in the slag as inclusions up to a some across. centimeter of of the fragments record considerable Analyses at the microscopic level (App. A, CHR 99). The analyses prove at that the this site results from smelting copper ores. slag conclusively of the slags are the result of several factors. Ele The final compositions ments of the final composition from the ore, the fuel, include contributions copper

the flux (whether itwas an addition or part of the ore), the air introduced even the walls of the furnaces. Some of into the furnaces, and potentially the crystalline phases are relics of the original charge placed while others formed during the smelting and subsequent as material. Additional weathering products phases formed use of the ary depositions long period between the during

in the furnace, cooling of the and as second 4.

the workshop

and the time of excavation. A few relics from the original charge can be recognized. These unaltered include phases that were insufficiently affected by the tempera minerals tures to alter their composition. Good examples include the quartz and

fig.

Craig 11.8:d.

and Vaughan

5. Bachmann Gale

et al. 1985, 6. Bachmann

3:b; Gale

1982,

1994,

p. 345,

p. 21, fig. 3:c;

p. 88.

1982, p. 12, figs. et al. 1985, p. 88.

3:a,

OTHER

METALLURGICAL

MATERIALS

141

chert grains visible in thin section analysis (App. A, CHR 97, CHR 93). Quartz has also been noted as a relic in comparable slags from the Near East.7 Tiny traces of the primary minerals that were not oxidized in the ore bodies the also survived the smelt workshop secondary exploited by were detected ing process. They Dispersive Spectrometry byWavelength and under the CHR examination 8) E, by polarizing microscope (App. where they were identified as pyrite, and SEM-EDS (App. F, ORE-1), covellite, chalcopyrite, and chalcocite. include several mineral phases. Late alteration products are of within voids visible (App. A, carbonate occasionally depositions of the glassy has also caused devitrification CHR 93). The weathering some in of CHR the formation of 93), resulting parts fragments (App. A, Postburial

new minerals

such as hedenbergite (App. B, CHR 93). The devitrifica is responsible for the dull surfaces on the portions of slag left tion, which to weathering in the upper layers of the slag pile, is visible with exposed

eye. One cannot be sure, but some of the cuprite may also be a product of recent weathering (App. A, CHR 99). Most of the mineral phases in the slags crystallized during the cooling of the material. These minerals are helpful in establishing some of the details of the process of smelting. Fayalite, the iron-rich end member of the olivine the unaided

group, has been regarded as a characteristic of slags formed under highly In slags produced under more oxidizing conditions, reducing conditions. the formation of the pyroxene hedenbergite would be typical.8 Olivine also tends to form with

high iron content in the melt (see App. F). Fayalite is some in of the present slags produced at Chrysokamino (App. A, CHR 87, CHR 99; App. B, nos. 7, 19, 20, 36), and magnetite and wustite, which also form under reducing conditions with high iron content in the melt, are present

as well

nos. 6, (App. A, CHR 97, CHR 87, CHR 93; App. B, see were also conditions 13, 23, 28, 33, 38; App. F). Reducing probably usual, but the presence of slags with substantial amounts of pyroxene and means cuprite (App. A, CHR 99) conditions with lower iron content

that more

(and/or oxidizing conditions in the melt) were also occasionally pres ent. The amount of cuprite is very small in comparison with the amount amount in ancient of cuprite of cuprite in the ingots has ingots.9 The large been interpreted as a result of the remelting of copper under more oxidiz than those existing in the smelting furnaces. The crystal ing conditions lization

of the magnetite and the other phases varies considerably from more euhedral to some with fragment pieces having larger and fragment, in crystals, indicating variations in the cooling rate as well as differences composition.

demonstrate that the smelting practices at Chryso not completely consistent from one operation to the next. ore differed in from time to time, and the de Probably the composition over to of the ratio of control fuel the amount of draft, and other ore, gree elements of the operation were not strictly controlled. The workshop was These

variations

kamino were

7. Bachman 8.

Hauptmann,

Maddin 9.

1980,

pp.

108-110.

Bachmann,

and

1994, p. 6.

Hauptmann, 2002, pp. 7-12.

Maddin,

and Prange

successful in removing copper from oxidic ores, but repeated operations did not duplicate prior conditions exactly, and the amount of metal remaining in the slag indicates that the process was not always a completely efficient smelting

operation.

CHAPTER

142

IO

THE ORE of copper ore are present at several locations in Crete,10 and are known from amounts larger Kythnos and elsewhere in the Cyclades.11 None of the known Cretan sources are close to Chrysokamino. The sources Small amounts

for copper on Kythnos and Seriphos have lead isotope patterns that are consistent with many of the artifacts found in EM Crete,12 and these were deposits and those at Lavrion already being mined and smelted in seem period.13 Sources north of Crete good candidates for the copper used for most of the EBA objects found on the large Minoan island. Analyses of copper prills in the slag from Chrysokamino by lead isotope analysis show that the patterns here also match those of Lavrion the Neolithic

and Kythnos,14 suggesting that some (or all?) of the ore smelted at Chryso in the north. kamino originated Small pieces of malachite and azurite come from the metallurgy loca are tion. These minerals and they are easier secondary copper minerals, a copper silicate that is also easy to smelt than sulfide ores. Chrysocolla, to smelt under ancient conditions,15 is also present. The pieces are small in length), and very few fragments are present in the (about a centimeter This situation is not unusual for smelting sites in record. archaeological the Eastern Mediterranean. to the ore from the smelting site, one piece of malachite This small, waterworn also comes from the small harbor at Agriomandra. ca. 2 cm in size with small amounts of a bright green mineral in it, pebble, on was Noel the beach. Gale confirmed the presence of copper picked up In addition

at the Isotrace Laboratory and the (Research Laboratory for Archaeology of No formations that Oxford could Art), History geological University. are present near the site, and it must have have produced the malachite been brought in by ship at some time in the past. The beach is a little over a kilometer

from the smelting workshop.

Affecting

Considerations Smelting

the

of

Location

Sites

to the logic that itwould be easier to transport smelted metal Contrary instead of the heavier ore, the location of smelting sites away from the mines is a typical practice in the ancient Mediterranean. Similar patterns recognized all the way from early mines at Feinan in Jordan16 in Palestine17 to Bronze Age mines in Britain and Ireland.18 Ores have been found at Shiqmim and Bir Safadi, both of which are located have been

and Timna

away from any ore deposits.19 A In a common metal production model

well

10.

Branigan

1989;

Stos

1998;

17. Rothenberg 1972, pp. 208

see also 210.

C. 15. See Tylecote's

comments

in

Koucky and Steinberg 1982b, p. 140.

Gale 1998. 1989,1993;

Gale andMacdonald 13. Stos-Gale

exists elsewhere.20

places, only crushing

14. Stos-Gale App.

and Stos-Gale

12. Stos-Gale

used inmany

1968; McGeehan

Liritzis 1996, p. 387. 11. Gale

similar situation

16. Hauptmann,

Stos

1991.

and Gale

Maddin 2003,

p. 87.

Bachmann,

and

1994, p. 6;Hauptmann 2000, pp. 167-168; Weisgerber 2003, p. 76.

18. O'Brien

ments byMuhly 19. Weisgerber 20. Lechtman

1996;

see also

1997, p. 772. 2003, 1991.

p. 76.

com

OTHER

METALLURGICAL

MATERIALS

143

and sorting occur near the mines, while the ore is transported to another location for smelting. Good reasons exist for this pattern. The crucial resources that determine the location of the smelting operation involve several factors in addition to the availability of ore. Some of these factors are related to availability of other materials, while others concern markets and social concerns. and charcoal, the most common are fuels employed, very bulky, and bundles of limbs take time to load and transport. After trees are cut down, several years pass before timber is again available from the same region. If the residue from olive oil production Fuel

is the first consideration. Wood

used as fuel for smelting, as is tentatively suggested in this volume, this product would arrangements with settlements capable of providing be required.

was

in choosing a smelting of flux can also be a consideration Availability site. At Chrysokamino, evidence for a deliberate addition of iron minerals as flux is ambiguous, but enough goethite is available locally for use as flux in copper smelting if this material was desired. At Feinan, no evidence survives for any deliberately added flux until the Late Bronze Age,21 but at

is apparently different. Iron oxide was obvi Chrysokamino ously present in the furnaces in some fashion, as is proved by the high iron content of the slag. It could have been part of the ore, or it could have been the picture

as a separate material. If the Chrysokamino smelters to use as a flux, the proximity of suitable sources could have been one of the influences on the choice of location. added deliberately needed iron oxide

The presence of a suitable topographic situation for smelting may also have been a factor. At Chrysokamino, the windswept cliff chosen for the seems to be a deliberate because the wind would aid workshop preference in providing the draft for the furnaces. A funneling effect associated with the north-facing is located helps to intensify trough where the workshop all northern breezes. Windswept hills were also chosen for smelting on the islands of Seriphos22 and Kythnos.23 An important additional factor

is the steady employment of a work are force trained in smelting. Ore concentrations in nature seldom uniformly distributed even within a single deposit, and miners must sometimes spend considerable time excavating worthless rock to get to the next part of the ore body. Smelting can proceed quickly, and it can often handle more ore can to stop the than be produced locally. It is not economical smelting while the workshop waits for additional ore, so smelting sites often use ore from more than one mine. Even a rich tin-mining region like Cornwall has sometimes relied on imported tin ore from South America.24

21. Hauptmann 2001, 22. Gale and Stos-Gale pp. 24-25. 23. Hadjianastasiou

p. 426. 1989,

and MacGil

livray 1988; Gale and Stos-Gale 1989, pp. 25-27.

24. See Gowland 1930, p. 522.

Proximity to markets and to the agricultural and social infrastructure to support a necessary may also be important metallurgical workshop considerations. Mines maybe found in remote locations, away from settle ments that would be sources of food, water, shelter, and steady supplies of to make furnaces. If like pottery, tools, or the raw materials as may have been the case the smelting work was conducted seasonally, at Chrysokamino, the ore may have been an alternative to transporting to the sources and all of its necessities transporting the entire workshop commodities

of copper ore.

CHAPTER

144

IO

ore is an additional factor. Before it can be preparation of the smelted, ore is typically prepared by beneficiation. This process involves the removal of bits of ore from their matrix to create a richer charge and improve The

the smelting process by giving a higher yield of copper with less slag. In less fuel is required during addition, because the ore ismore concentrated, is simple but time-consuming. the smelting. The process of beneficiation The rock containing the ore can be crushed with stone hammers, and the ore can be sorted is bright by color. Because azurite is bright blue, malachite green, and cuprite is red, the sorting operation can be done even by children. If the ore is heated first, it cracks and makes the crushing easier. In some cases where beneficiation has been studied, it has been observed that the ideal size of the ore pieces to be smelted is about 1 cm in length.25 This is Because this time-consuming task the same size found at Chrysokamino. use to can be done not it is metalworkers for unskilled labor, profitable by not to for this it, and itmay be preferable stop the smelting process type of activity. a result of all these factors, ancient smelters often seem to have ore. Min to transport only the beneficiated convenient

As

it more

found

at the mine itself, ing, sorting, and beneficiating would have been done to the smelting site. Transportation, either by land before transportation to accomplish with bags of tiny pieces or by sea, would have been easier waste ore rock. than with of beneficiated large tonnage that included ore sea would, in have been of fact, very simple, be by Transportation cause bags of ore could have been used as the ballast ships stable.

required

to keep early

The archaeological evidence from Chrysokamino proves that wherever at Chrysokamino itself. The it was not beneficiated the ore originated, absence of large quantities of waste rock and discarded low-grade ore at the site indicates that this process did not occur on the headland. The same shows that the mine was not adjacent to the workshop. Mining so the mining large quantities of useless rock along with the ore, is the the like beneficiation site, site, dumps of waste easily recognizable by has neither the waste rock nor material. Because the site at Chrysokamino the discarded tools that are common at such sites,26 one must conclude evidence removes

that the workshop was not the location of either mining or beneficiation, a conclusion confirmed by the geological survey of the region.27 not locate any ore deposit at did the geological survey Although or any formations that might have once had cuprous Chrysokamino metal

that a small deposit once existed has been the possibility cave immediately north of the metallurgy The site, previously. was later as a mine Mosso and others,28 interpreted properly by

deposits,

proposed regarded as a natural

cavern.29

ore on the cliff near Chrysokamino.30 Boyd reported seeing a was saw amount of ore brought there from some small what she

Harriet Whether

else or a natural geological occurrence, however, is not known. No was present, and she did not describe what she saw in any detail. geologist Local chlorite deposits are green, and someone not trained in geology or some ores. Even if a locality mineralogy might mistake them for copper where in the vicinity did once produce a small amount of ore, the deposit

where

was are

not adjacent present

there.

to the workshop

site, because

no

signs of beneficiation

25. Hauptmann,

and

Bachmann,

1994, p. 5.

Maddin

26. D. Gale

1991; Gale

and Otta

1991.

way

27.

See Chap.

28. Mosso Hutchinson 29. Faure

2 in this volume.

1910, 1962, 1966,

pp. 289-292; p. 40. pp. 47-48;

pp. 50-51. and 30. Hawes Boyd 38. p.

Branigan

1968,

Hawes

1909,

OTHER

METALLURGICAL

MATERIALS

145

of the ore was just one step in a complex process. procurement to be coordinated with needed the establishment of a acquisition

The Ore

the training of personnel, the procurement of fuel and possibly workshop, flux for the furnaces, and the construction of the furnaces with their clay to the general logistics involved with the maintaining personnel engaged in the work, the workshop a distribution system for the smelted metal. chimneys.

Ore

In addition

feeding and also needed

Sources most

for the identification of sources of satisfactory methodology ore combines lead with the prehistoric isotope analysis study of elemen tal compositions.31 Lead isotope analyses provide a pattern based on the

The

can date of the deposit, while elemental geological composition give some indication of the ore. of the For the composition original prills from lead isotope studies have revealed patterns consistent with Chrysokamino, sources in Kythnos or Lavrion.32 The high percentage of arsenic in some of the prills and the occasional presence of other metals (see App. B) sug gests that at least part of the ore may have been derived from the oxidized upper parts of a deposit containing some arsenic and possibly other metals in addition to copper.33 Such a deposit has not been identified in either Crete or the Cyclades. The percentages of arsenic from the copper sources on

are lower than at some arsenic is Kythnos Chrysokamino, although ores are there. the At Lavrion, present primary iron-manganese deposits and zinc-iron-lead but other minerals, many deposits, including copper, occur

in smaller amounts.34 Malachite

iswidespread within the region of arsenic Lavrion, arseniosiderite, containing (arsenopyrite, and many others) occur aswell. We have no way of knowing the composi tion of the upper levels of the deposits at Lavrion because they have been use as a mined away. However, of the Lavrion any copper source region must have exploited were for Chrysokamino well away from deposits that the two main classes of primary ores, because manganese and lead are scarce in the at no is The truth that known prills Chrysokamino. deposit and minerals

in the Aegean and the exactly matches both the elemental composition lead isotope fingerprint of the prills at Chrysokamino. If Chrysokamino a one is Lavrion polymetallic deposit, exploited geographic region that has both the potential for an elemental general comparison and an isotopic comes close to the ores from Chrysokamino. At fingerprint that matching our present level of as a cannot source it be excluded for the knowledge, ore. cannot be excluded the of lower levels either, presence Kythnos given

31. Hauptmann, Schmitt-Strecker 32. Stos-Gale

Begemann, 1999. 1998;

of arsenic from the region, the match in lead isotope fingerprint, amount of evidence for interactions with Crete large Cycladic areas during the Early Bronze Age.35

and

see also App.

C.

33. See Craig and Vaughan (1994, for similar occurrences in pp. 247-250) central and several other parts Europe of the world. 34. Marinos

and Petrascheck

Hanke 1994. 35. Karantzali 2000.

1996;

Broodbank

1956;

and the in other

in the slags and copper however, prills from Chrysokamino, elsewhere in this volume (see esp. Apps. A, B, and C), is a strong cautionary signal suggesting the possibility of a number of different ore sources If this was the case, and during the long life of the workshop. more a source than used of ore, then the situation Chrysokamino single is extremely complicated, and we cannot be sure that any one elemental or lead composition isotope fingerprint will provide a secure identification of the source of ores for the site as awhole. Diversity as discussed

CHAPTER

146

IO

THE COPPER PRILLS Small prills of copper are present in some of the slag fragments. Their size is always small, ranging from microscopic (under 50 micrometers across) a few to about a centimeter in size. Although of size small prills especially are in the on the interior of the most of them coatings chimney fragments, come from the chunks of was slag. Their small size suggests that the metal were extremely valuable, and that larger pieces carefully removed and saved. If occasional copper lenses or ingots were collected, no remains of them were found in the excavated material. The were only prills found at the site small masses

within the vitreous waste material. They have a been examined by number of analytical techniques, providing different of information. types Most analyses of prehistoric metals have been directed primarily toward disseminated

the determination

of provenance, but results have been very problematic. are variable even within a ore or compositions single piece of a sources. few conclusions single artifact, permitting regarding original More encouraging results have come from analyses aimed at understanding ancient technological processes. Elemental

thin section analysis indicates that other copper minerals are Optical to the prills of native copper found in the slag. The in addition present occasional presence of tiny crystals of the copper oxide cuprite may be a result of the original crystallization (App. A, CHR 99), but it is also possible that it represents oxidation of small amounts of native copper after burial. The copper and The cuprite is always very tiny (under 100 micrometers). iron oxide delafossite, revealed by X-ray is likely to have formed under oxidizing

diffraction conditions

(App. B, CHR 96), during the smelting

operation.

of the recovered prills are extremely altered, and inmany cases no pure copper survives. Information on the copper corrosion is provided by was used to examine the surfaces PIXE analyses (App. D).This technique Most

of altered prills, with analyses of several parts of the surfaces. It showed that the chlorine content was enriched on the surface from the formation content on the surfaces was highly variable, of copper chlorides. Arsenic and itwas depleted by the onset of corrosion products. analysis by SEM/EDAX (App. B) shows that the copper contains amounts the metals, iron and arsenic are elements. small of many Among present up to several percent, with iron occurring up to 4.9%. The arsenic The

is extremely variable at the microscopic that is too small to detect to 26%.

content

The

level. It ranges from a trace

is an important aspect of the copper from this site. of arsenic is a feature of copper artifacts from the 4th

arsenic content

A high percentage and 3rd millennia

area in the eastern Mediter B.c. in a broad geographic as as in well central and southeastern Europe37 and the northern ranean,36 of Early Cycladic It is a characteristic artifacts39 and Early Caucasus.38 Minoan objects.40 Arsenic can be present in copper tools and weapons as a result of several alternative processes. Smelting arsenic-rich copper ores, adding arsenic ores to the smelting charge, or introducing arsenic in subsequent melting opera

36. Charles 1967. 37. Budd and Ottaway 1991. 38. Ravich and Ryndina 1995. 39. Gale tables

40. Gale nou

and Stos-Gale

1989,

1990; Mangou

and

1-3.

1998.

Ioan

MATERIALS

METALLURGICAL

OTHER

H7

tions are all capable of producing arsenical copper. Most coppers contain amounts ranging from a trace to asmuch

ancient

arsenical

as 7%-10% of the

raises questions concerning the characteristics of the resulting over unalloyed copper. and the possible benefits of arsenical alloys alloys It also raises questions as to whether the ancient metallurgists were aware the arsenic was added of these advantages (if they existed) and whether

metal. This

deliberately. Several

authors have discussed the possible advantages of arsenical over metal. Favorable characteristics have been suggested both pure copper in regard to preventing oxidation during casting and improving annealing.41 Arsenical copper would also have had a paler color than the pure metal, it recognizable by sight and, perhaps, aesthetically preferable. An advantage in producing metal for sheets has been proposed for the alloy aswell.42 Cautions have been voiced, however, particularly with respect to

making

in the heterogeneous metals achieved by ancient processes.43 Experiments not it is have arsenical shown that copper casting necessarily advantageous because too many other factors are involved in creating a superior alloy, casting temperature, pour rate, and the presence of other im Most researchers have considered that at least 2%-3% of arsenic purities.44 is required to suggest a deliberate alloy.45 Percentages of arsenic between including

the best results.46 Amounts

2%-6% yield When

41. Caley 1949;Mar?chal 1958; Charles 1967; 1985, p. 25; Gale and Stos-Gale

1989,

42. Lechtman

p. 30. 1996.

43. Budd 1991. 44. Budd and Ottaway 1991, p. 135. 45. Hauptmann, Schmitt-Stecker

Begemann, p. 9.

and

1999,

47. Pollard et al. 1991, pp. 130-131; Thomas,

and Williams

age of arsenic above 5%, and much of the copper surely had less than 3%. since the arsenic is present within the prills in the slag, itwas Obviously, included in the charge in some fashion, either from within the copper ore or from the addition of a different ore. The amount of arsenic is so variable it that it seems unlikely that the metal was added deliberately. Whether deliberate or accidental, it probably resulted from the use of arsenic minerals along with the copper ores. Experiments have shown that if arse

was

nic minerals below

138-139.

Pollard,

little effect.

are included

in the charge, concentrations of the metal below even in the with temperatures easily smelting process it ismost likely that small amounts of 1000?C.47 At Chrysokamino,

5% can be achieved

46. Budd and Ottaway 1991, pp.

less than 2% have

percentages exceed 7%, the metal becomes brittle with cold working. The copper from Chrysokamino does not seem to have had a percent

1991.

arsenic minerals

were

crushed

ore.

copper

present

in the charge along with

the small pieces of

CHAPTER

II

Faunal

Remains S. Reese

byDavid

faunal remains are divided by the two contexts identified at the met two contexts, the small allurgy site. These apsidal structure and the rest of the slag pile, have different dates. The slag pile is amixed deposit with to Early Minoan Ill-Middle pottery that ranges from Final Neolithic

The

IA. The

Minoan

structure

apsidal

is a closed

deposit

that comes

from

EM III-MM IA. Within The

these contexts, the finds are organized by grid-square and level. are a presented with whole and restorable individuals before and fragments whose complete size is uncertain after a semicolon.

numbers

semicolon Dimensions The

of the whole

faunal remains

individuals

include

are

presented

inmillimeters.1

the following:

Alvania

montagui (gastropod, small) Cerithium vulgatum (cerith, horn shell) Columbella rustica (dove shell) Fasciolaria lignaria (tulip shell)

Monodonta Paracentrotus Patella Pisania

turbinata lividus

(top shell) (sea urchin)

coerulea (limpet) maculosa (dove shell)

Lepus europaeus (hare) Only a few faunal remains come from the pile of slag and industrial ceram ics fragments. They include 1 Lepus vertebra, 61 Patella, and 6 topshells (Monodonta). The shells are all edible species that live in shallow water on rocky

shores.

the time of excavation, the samples were isolated by specific loca could be recognized because faunal tions, but no pattern of distribution remains were distributed fairly uniformly over the excavated areas, so for At

1. Abbreviations: Number includes sieving.

MNI

of Individuals; material

ws

recovered

=Minimum = the deposit from water

this presentation, the remains are given by grid square. more marine remains come from the apsidal structure. Relatively 1 fish 34Monodonta, include 3 Paracentrotus, 1 crab 35 Patella, bone, They 8 and other shells of 5 the Like shells from the slag pile, the claw, species. are on species here rocky shores. mostly edible species from shallow water All of the marine invertebrates could have been collected casually from

CHAPTER

i5o

the seashore and fish bones

in the vicinity of the workshop site. The rarity of mammal a meat not that did suggests significant role in the diet play

here.

consumed

THE SLAG PILE Mammal

Bone

Q20-2 (ws) 1 Lepus-sized

Marine

vertebra

Invertebrates

K17-2 3 Patella,

25.25;

2 broken,

small/medium,

3MNI

K18-1 1 Patella

M17-2

fragment

(ws)

9/18 Patella,

M20-2

II

5 broken, 2 large, 9MNI

16.75,18,19,20;

(ws)

1 Patella

fragment 4Monodonta fragments,

1MNI

M20-3 1 Patella,

exterior partly burned black, broken,

1 Patella,

broken, medium

2 Patella,

26.25,1

large

M20-4 (?)

M20-5 broken, burned,

large, 2MNI

M20-7 1 Patella,

30+, broken,

large

N17-2 31/4 Patella, 26.25,29.5, 4Monodonta

(2), 22,23,23.75,25.5,25.75, small; most broken, 31 MNI fragments, 2 lips, 2MNI

18.5,19.5,19.75 34.5,1

N20-4 (ws) 1/1 Patella, broken, large, 1MNI 2Monodonta fragments, 1MNI O20-4 1/1 Patella,

20 (burned); broken

fragment,

2MNI P20-3 (ws) 1/1 Patella, ca. 15,1 MNI 1Monodonta fragment, medium/large

possibly

burned,

REMAINS

FAUNAL

151

Q20-2 (ws) 1 Patella

fragment 1Monodonta fragment

Q20-3 4 Patella,

18.75,23.25;

small/medium;

medium/large,

4MNI

R20-3 (ws) 2/4 Patella,

all broken, 2MNI

THE APSIDAL STRUCTURE Fish

Bone

M18-2 (ws) 1 fish bone Marine

Invertebrates

L19-2 internal center, 2MNI 2/1 Patella, 16,18.75; no near 25 Monodonta apices (1 fragments, 7 lips, apex), 7MNI L19-3 1 Patella,

burned gray, 30.5 6Monodonta fragments, 5 fresh, 3 lips (3MNI),

1waterworn,

large

M18-2

(ws)

8 Patella

fragments, 3MNI 38Monodonta fragments, 4MNI 1Alvania montagui, small 1 Cerithium 1 Columbella

vulgatum, fresh, small, broken lip rustica, lip fragment columella fragment, water-worn

1 gastropod Paracentrotus, 6 internal pieces, ments, 1MNI

6 spine fragments,

11 test frag

M18-3 1 Patella, burned 3Monodonta fragments; 1 Fasciolaria

M19-2

large, 1MNI lignaria, bumpy, open mouth,

broken

lip, length 35

(ws)

1/1 Patella, broken, 1MNI 2Monodonta fragments, large, 1MNI

M19-3

(ws)

6/1 Patella,

27.5,29.5, 6MNI

32.75,

34.25; 2 broken,

1 small, 1

medium, 6 complete, 6 lips, 11MNI 3/51 Monodonta, 1Fasciolaria lignaria, fresh, broken lip, length 31.75

CHAPTER

152

M19-4

(ws)

1MNI 1/2 Patella, medium/large, 1/1 Monodonta, complete example is small; distal piece with ancient break in half, 2 MNI 1 Fasciolaria Paracentrotus,

lignaria, body fragment 3 internal pieces, 2 test fragments,

is large,

1MNI

M19-5 1Monodonta

fragment,

body, large, encrusted

N18-2 17.25, ca. 20,24.5; 6MNI medium,

6 Patella,

1Monodonta,

1 center fragment,

2 broken,

complete, medium

N18-3 3 Patella,

3MNI 17.25; 1 small, 1 medium, 1Monodonta fragment, burned, body, medium/large 1 Pisani,

fresh, complete,

length

18.75

N19-2 (ws) 5/9 Patella, 18.5,19,25.25; 2 MNI 1/12 Monodonta, 1 Columbella, Paracentrotus, fragments,

2 broken,

complete, fresh 12 internal pieces,

2 medium,

7MNI

38 spine fragments,

36 test

1MNI

N19-3 (ws) 1/1 Patella, 23.75; broken, 1 crab pincer fragment

2MNI

N19-4 3 Patella,

3MNI

2 burned

center, broken;

1 unburned,

25, broken,

II

CHAPTER

12

for

Evidence Threshing Early

the

Remains Minoan

of

Use at

the

Metallurgical

Workshop by Glynis Jones andAnn Scho?eid the 1997 excavations

of the metallurgical workshop, furnace chim were with of evidence botanical collected from ney fragments impressions the surface of the site. Latex casts were made of these impressions, and they were identified (at x 10 to x40 magnification) subsequently by reference to botanical collections at the Department of Archaeology and Prehistory, The are of Sheffield. These identifications listed inTable 12.1. University During

The majority of casts were identified as barley (Hordeum L.).The most inclusions were rachis fragments, and all of them could be iden tified (either definitely or probably) as two-row barley (Hordeum disti on chon L.) the basis of the sterile lateral florets characteristic of this species

common

(in six-row barley, all the florets are fertile, i.e., grain-bearing). The shape of the internodes (parallel-sided) was also characteristic of the two-row species (rachis internodes of six-row barley are trapezoidal). The sterile lateral florets of two-row barley remain attached to the rachis when the ears are threshed to release the grain from the central fer tile florets, and so these fragments, lacking the central grain, apparently repre sent threshing remains. One of the rachis fragments (cast X 991) is from the lower part of the ear and is still attached to the culm (straw). Other grass culm fragments, not attached to straw resulting from threshing. A few definite and possible grains of identified, one of which appeared to be two-row barley (six-row barley has both lateral grains). The grains were all rather grains that might

remain attached

rachis, could also represent barley barley (the hulled variety) were also as would be straight, expected for central straight grains and twisted

sort of underdeveloped to the rachis during threshing. The most small?the

likely interpretation of this material is that it represents the chaff and straw resulting from the threshing of two-row hulled barley, which was used as of the chimneys. temper in the construction One fragment has been tentatively identified as a glume (chaff frag ment) of wheat (Triticum L.), which may have been aminor contaminant of the barley crop. A single olive (Olea L.) leaf (cast X 213) and a large indeterminate object (cast X 990) are presumably chance inclusions. Few Minoan

they are, these remains provide rare evidence of Early at this time, and it agriculture. Barley was clearly cultivated though

CHAPTER

154

TABLE 12.1. PLANT IMPRESSIONS WORKSHOP METALLURGICAL

IN CHIMNEY

FRAGMENTS

FROM THE

Cast Number

Spe 213 Hordeum distkhon L. H.

cf. distkhon L.

rachis frags.

cf. distkhon L.

hulled grains

978

981

982

985

988

989

990

991

993

994

997

998

1000

1001

1002

hulled grains

sp.

cf.Hordeum

977

rachis frags.

H

Hordeum

12

sp.

grains

cf. Triticum sp.

glume

indeter. cereal

grain

indeter. Gramineae

culm, frags.

Olea sp.

leaf

+ chaff

indeterminate

to identify with certainty the two-row hulled type. possible were also found in mud at the Early plaster Impressions of hulled barley but in this case identified, at least in part, settlement of Myrtos,1 Minoan as the six-row species (Hordeum impressions also vulg?re L.).The Myrtos has been

it of chaff and straw and, as at Chrysokamino, predominantly was floor. from the that threshing they represented sweepings suggested coarse fabric of a few At Early Minoan Debla,2 impressions from the emmer as storage bins were identified barley grains (along with grains of settlement of Pseira, a con wheat and oat), and at the Middle Minoan consisted

of charred hulled barley grain was tentatively identified to the and Pseira contrast with species.3 The finds from Chrysokamino those from the rest of Bronze Age Greece, where six-row barley ismore ismuch less reliable be usual.4 The evidence for wheat at Chrysokamino centration two-row

cause it is tentatively identified and may be no more than a contaminant of the barley. The olive leaf, too, could be derived from awild or cultivated tree (Fig. 7.2).

1972. l.J. Renfrew 2. and Warren Greig 3. Jones and Smith, 4. Hansen

1988.

1974. forthcoming.

1003

CHAPTER

13

in

Chrysokamino

of

Early

the

History

Metallurgy

byJames D. Muhly

research on Crete since about 1980, especially in the north Archaeological eastern part of the island, the Bay of Mirabello, has established that the of Cretan coincided with the establishment copper metallurgy beginnings of new settlements period,

in the eastern part of the island in the Final Neolithic ca. 4500-3500 b.c.1 This association of new settlements with

extended to the far east coast (Palaikastro) incipient copper metallurgy and the site of Phaistos.2 The evi and possibly as far west as the Mesara at Phaistos dence for FN metallurgy is, however, very inconclusive,3 and under discussion the associations here are best attested in the eastern part of the island. Inasmuch as metallurgical technology must have been to it from the Crete outside should be stressed that the island, imported Final Neolithic

has long been

recognized

as the first international

age in

Cretan

prehistory.4 on Crete Prior to 1996, the evidence for incipient copper metallurgy consisted of isolated finds that would be considered "trinket metallurgy" in a comparable Mesopotamian context.5 This situation changed with the 1996-1997

Here, for the first time, was an a site, copper smelting site that seemed to have itsmain b.c. (EM III), in the later part of the 3rd millennium

excavations

actual metallurgical

at Chrysokamino.

period of activity but was associated with Neolithic.

the date, see Johnson 1996, a for date, see p. 271; slightly different 1999. Johnson 2. Vagnetti 1975, esp. pp. 94-95. 1. For

3. Vagnetti and Belli 1978, p. 131. 4. Betancourt

p. 13; also 38-39. 1996, pp. Vagnetti 5. Moorey 1988, p. 29. 6. Evans vol. 2, pp. 1921-1935, 15, fig. 3:f. 1985,

7. Evans

1971,

8. Phelps,

Varoufakis,

1979.

p. 115. and Jones

14

sherds from all earlier periods, going back to Final It is true that no direct association can be made between metal

lurgical activities and the FN sherds, but it is hard to imagine any other reason for human presence on this isolated, windswept point of land. How are we to evaluate this new evidence for the early development on the island of Crete? How does it relate to our of metal technology current understanding of the beginnings of metalworking in the Eastern Mediterranean and the Near East? First of all, itmust be recognized that the Final Neolithic does not represent the very beginning of metal usage on Crete. That

honor goes to a Late Neolithic flat axe from Knossos, found a by Sir Arthur Evans.6 Evans published only drawing of this axe; to my no exists, and the axe can no longer be located in knowledge, photograph the Herakleion Museum. does it come from a secure Nor, unfortunately, context.7 It does, however, seem to be of a common Late Neolithic type, known

from sites on the Greek mainland

sites in the Balkans.8

and especially

from numerous

CHAPTER

i56

13

as an axe is probably best understood import in Late Neo lithic Crete, most likely from the Balkans, as this type of axe was never pro duced on Crete itself. A Balkan origin would not be surprising. Throughout The Knossos

for a Balkan origin of Greek Neolithic metallurgy seems exist not only with copper Balkan connections quite overwhelming. even lead. but also with the of gold, silver, and possibly metallurgy, working I have argued the case for Balkan influence on several occasions,9 and Balkan the Aegean,

the evidence

connections

have also been pointed

out by Branigan

and Nakou.10

EARLY AEGEAN METALLURGY AND THE BALKANS increases

The

quantity of metal finds from the Greek Neolithic of D?moule and Perl?s every year. In the terminology

view of Neolithic with

Neolithic), ca. 5500-4500

almost

in their recent re

Greece, most of these finds come from Phase 5 (Final a small handful of finds from Phase 4 (Late Neolithic,

b.c.).11 Only one tiny amorphous lump of copper from ca. 6000-5500 is said to be Middle Neolithic, b.c.12 level 14 at DikiliTash Recent finds from excavations at Ftelia, on the north coast of Mykonos, have produced a small collection of copper objects, especially awls and as well as a circular disk of a central pins, gold with perforation.13 A series of 10 calibrated radiocarbon dates from Ftelia support a date in the first I in the terminology b.c.,14 or Late Neolithic in of the excavator.15 Yet the Balkans, gold appears only inwhat are now it appears, b.c.16When called Late Chalcolithic contexts, ca. 4250-4000

half of the 5th millennium

however, it is attested in large quantities; some 6 kg of gold have survived from Eneolithic southeastern Europe.17 The metal finds from the site of Dimitra, just north of Sitagroi in some controversy. are eastern Macedonia, the of subject According currently to Grammenos, the excavator, five beads of copper come from aMiddle context in trench II, and four beads of copper, along with two Neolithic context in trench I.Metal finds beads of gold, come from a Late Neolithic from the site also include a possible fishhook of copper and another of gold to Andreou, Fotiadis, inwhat seem to be unknown contexts.18 According and Bronze Age northern Greece, to Sitagroi I to III, with the site of Dimitra has three phases, corresponding metal finds, including gold and copper beads, from all phases.19 Accepting a date for b.c.,20 this would mean that Dimitra Sitagroi I at 5500-5200 has some of the earliest copper finds from anywhere in Europe, as well as the earliest objects of gold. and Kotsakis,

1985a, 1996.

9.Muhly 10. Branigan 1995,

in their review of Neolithic

1974,

see I. Sikka,

p. 98; Nakou

11.D?moule and Perl?s 1993; for of the evidence,

p. 15. 14. Facorellis

%pi)aa(pi December

gttj 24,

2002,

pp. 4,21.

a summary

M?kovo,"

"Neo^iBiK?

Kathimerini,

see

Muhly

1996, pp. 78-80; 2002, p. 78. 12.Muhly 1996, p. 76. 13.Maxwell 2002; for the gold disk,

p. 311. 15. Sampson 16. Pernicka 17. Pernicka

and Maniatis

2002,

18. Grammenos the analysis et al. 1997.

19. Andreou, p. 312. 20. Elster

1997,

of these

finds,

Fotiadis,

p. 270; for see Mirtsou

and Kotsakis

2001, 2002a. et al. 1997, et al. 1997,

p. 51. p. 41.

p. xxvii.

and Renfrew

2003,

CHRYSOKAMINO

EARLY

AND

METALLURGY

*57

the island of Andros, and jewelry, copper implements, weapons, with architecture and marble have been discovered vases, along impressive at the fortified settlement of Strophilas.These finds are dated to the middle of the 5th millennium of the Final Neolithic b.c., or the very beginning are also a few Final Neolithic There copper objects from the Thar period.21 rounia cave in Euboia, principally pins or needles.22 For Sampson, the Final On

Neolithic

IIA) was numbers, make

(his Late Neolithic

always in insignificant Mainland Greece and the Aegean."23 Two aspects of the development

the period "when copper objects, their appearance at every site of

of Final Neolithic Aegean metal to be emphasized. The first is its polymetallic character. The lurgy need saw the real Final Neolithic of in and silver development gold metallurgy the Aegean world. The number of gold artifacts known to date from this has greatly increased pendants from the Theopetra period

in recent years, with the discovery of gold cave in near and from Anavissos Thessaly

as well as a inMagnesia, and Platomagoulia gold strip from the Zas cave on the island of Naxos.24 Silver pendants have been found in the cave in theMani were discovered Alepotrypa peninsula. They along with Yannitsa

a hoard of silver jewelry that included two pairs of bracelets and 168 very in unpublished small perforated silver beads, now duplicated finds from an EM I cemetery at Gournes in central Crete. Other silver pendants have in the Amnisos

been found

cave on Crete

and in the Cave of Euripides

on

the island of Salamis.25 Most

of these pendants, both the gold the characteristic "ring-idol" design, with are awell-known feature of Balkan pendants the Karanovo VI or Gumelnitsa culture, also with the Aegean period, roughly contemporary

and silver examples, are of a central Such perforation. from the time of metallurgy referred to as the Eneolithic Final Neolithic.26

The most

famous examples come from the cemetery at Varna, on the Bulgarian shore of the Black Sea. The Varna finds have been featured in many books and articles by the lateMarija Gimbutas.27 21.

"Neo^iGiKO? Kathimerini,

Av?po,"

From

axr|v

oiKiGuo?

p. 15.

2001, 22.

1993,

Andreopoulou-Mangou on for more

these

pp. 435-437;

made

apparently see below. 23.

Sampson

of arsenical

objects,

copper,

the largest known pendant of this type.28 An example in copper is known from Period IX at Emporio on Chios.29 A stone pendant comes from the on Andros. Strophilas settlement Itwould be hard to imagine more convincing evidence for the dominant

p. 164.

2002a,

24.Muhly 2002, p. 78. 25.Muhly 2002, p. 78. 26. For

see von G?rs

chronology,

dorf and Bojadziev 1996. 27. E.g.,

Gimbutas

Balkan

1977.

28. For all this, seeMuhly 2002, p. 78; also rary exhibit

Athens

that

chemical

of a tempo of Museum

(Demakopoulou 1998).

29. This from

see the catalogue in the National

is the earliest

metal

artifact

site; for the remarkable composition

Muhly 2002, p. 78.

of this piece,

the most

spectacular collection of such ring from the antiquities market. In two comes, unfortunately, one at (October 1, 1997) and a second in sting operations, Vouliagmeni Patras (May 17,2000), Greek police confiscated a large cache from would be sellers of looted antiquities, including 67 ring-idol pendants made of thin sheet gold. The most impressive piece, 15.5 cm in height, is, in fact, world,

idol pendants

7,

September

the Aegean

see

character of Aegean Final Neolithic metallurgy. Yet this interpreta tion is still anathema to many no part Aegean prehistorians. They want of it, for they feel that it reeks of old-fashioned diffusionist archaeology. This brings us to a discussion of the second main feature of Aegean Final Neolithic metallurgy. Whereas earlier periods produced only isolated ar us with evidence for the actual the Final Neolithic tifacts, period provides of such artifacts. Now, for the first time, we can speak with production assurance of the actual of metal artifacts on Greek soil. The production

CHAPTER

i58 most

13

comes from the site of assemblage of such evidence Sitagroi inMacedonia. The excavators of Sitagroi are also the most outspoken op ponents of what can be called the "Balkan Connection." extensive

site of Sitagroi, also known as Photolivos, was excavated in three under the joint direction of Renfrew and Gimbutas. (1968-1970)

The seasons

in 1986, dealt report, published on the and with studies archaeology chronology, along specialized area. environment and economy of the The second volume, dealing with

The

of the final excavation

first volume

with

in 2003. At the very beginning crafts and technology, was published of the excavation, Renfrew wrote an extremely influential article on "The of the South-East Autonomy European Copper Age."30 In this article, he outlined

his arguments for the independent invention of metallurgical in various parts of the ancient world. He has now returned

technology to this theme Artifacts

in the chapter on "Metal Sitagroi volume, a series of studies based on or Meanwhile, of the finds from Sitagroi appeared between 1969

in the second

and Metallurgy."31

including discussion and 2003. Among them, the most important from the perspective of early is a study by Veronica McGeehan-Liritzis and Noel Aegean metallurgy and Early Bronze Age metal Gale dealing with the analysis of Neolithic artifacts from Greece.32 This

is not

the occasion

is important

What

here

for another

is the evidence

evaluation

for metalworking

of all this work.33 at

Sitagroi. The seem to have remarks Renfrew diffusion34 recently published by regarding in the late 1980s and have no bearing on current discussions. been written I talk about diffusion I have inmind such works as the recent book When by Oppenheimer, deriving all ancient civilization from Southeast Asia.35 inAegean Neolithic metallurgy To argue for a strong Balkan component does not, inmy opinion, constitute diffusion.36 of copper artifacts is present already in Sitagroi II, a now as 5200-4800 b.c.37Most of period regarded early Late Neolithic, a come from of uncertain the metallurgical materials Sitagroi III, period exact date (Sitagroi IV-Va is ca. 3200?2500 b.c., with III covering some or all of the intervening centuries between 4800 and 3200 B.c.).38 The metal A

small number

a as objects from Period III include pins, beads, and bent strip of copper, well as a small gold bead.39 Upon final publication, however, the evidence for actual metalworking has proved quite problematic. most likely frag Sitagroi III has 36 sherds with copper incrustation, are ments of crucibles. As the incrustations found not only on the interiors of the sherds, but also on the exteriors and the rims,40 the logical conclusion is that the crucibles were used to pour molten metal in some sort of casting a operation. The lack of molds is something of problem in this context, but contexts prior to the Early molds are virtually absent from all metallurgical Bronze Age. Sitagroi III, then, provides evidence only for the crucible melt even native copper. This is true as ing of metallic copper, perhaps probably well for the isolated finds at other contemporary sites, such as Giali (near inwestern Macedonia.41 the Alepotrypa cave, and M?ndalo Nisyros), at Sitagroi or at any other cannot be documented smelting Copper can site in the Aegean. Nor the source(s) of the copper ore being Neolithic

30. Renfrew

1969.

31. Renfrew

and Slater

2003.

32. McGeehan-Liritzis

and Gale

1988.

33. SeeMuhly 34. Esp. pp. 317-318.

1991.

Renfrew

35. Oppenheimer 2000. 36. See

Diamond

the

and Slater

2003,

1999; Terrell

important

remarks

in

1995.

37. Andreou,

Fotiadis,

and Kotsakis

Fotiadis,

and Kotsakis

2001, p. 308. 38. Andreou,

2001, p. 308. 39. Renfrew fig. 8.1:b-g. 40. Slater 41. Renfrew for problems

and Slater

2003,

p. 305,

p. 303. and Slater

2003,

p. 309;

2003,

of interpretation

Sitagroi III, seeMuhly 2004a.

regarding

AND

CHRYSOKAMINO

EARLY

METALLURGY

159

in the Aegean Neolithic be identified. The analytical evidence to is date very unsatisfactory. The artifacts themselves are small; published the samples that the analysts were allowed to take were obviously much exploited

smaller, and they often contained only corrosion products, not actual metal. Some of these samples, taken from a single artifact, were analyzed by three different analytical techniques.42 The results vary by over 260%.43 Obvi can be said about provenience on the basis of such evidence. ously, nothing lead isotope evidence is also inconclusive,44 but it does suggest that sources to the north, were as perhaps in Bulgaria, being exploited. Indeed, Renfrew concludes, "The likelihood of a northern origin for the earliest on other was practices at Sitagroi grounds, metallurgical proposed above

The

and these findings would harmonize with it."45And that, in reality, is all I mean when I talk about the role of the Balkans in the development of Aegean metallurgy. The reasons for turning to the north, of metallurgical understand the beginnings

to the Balkans

to

in order

in the Aegean

technology technological priority than with simple typological or with flat axes and was comparisons, ring-idols. What being done in the was also in in the itwas be done but the Balkans Balkans, Aegean being on amuch centuries done earlier and scale. The earliest copper ing larger have more

to do with

in the Balkans, an awl from the site of Balomir in Romania, is a massive object, 14.3 cm in length, and it comes from an Early Neolithic b.c.46 By the second half of the 5th millennium context, 7000-6000 B.c., the Balkan copper industry was producing hundreds of massive shaft-hole artifact

axes and axe-adzes.47 At

the same time, Balkan metallurgists were mining ores in copper underground mines with shafts and galleries,48 and they were ores from these of capable smelting the complex secondary sulfide of gold ring-idol pendants known today from the is quite impressive, but it does not begin to compare with Aegean the quantities from the Balkans. Just four of the richest graves from the Varna cemetery produced a total of 2,200 gold objects.50 mines.49 The

number

world

We

now have

some

perspective for comparing Cretan metallurgical with what had gone on to the north, on the Greek mainland developments and in the Balkans. Metallurgy later in Crete and on a much developed smaller scale, with striking differences. The ubiquitous ring-idol pendant is virtually unknown on Crete. The silver example from a burial near the Amnisos

cave seems to be an isolated find. Flat axes never became

of the Cretan

metal

only one ever found were

quite

common.51

developing 42. McGeehan-Liritzis

1988. 43. SeeMuhly 44.

Stos-Gale

45. Renfrew

and Gale

1991, p. 361.

and Slater

2003,

46.Muhly

1996, p. 78; Bogn?r

as

always,

example similar axes made

seems

to have

repertoire of types (e.g., the triangular of Eneolithic

1978;

Pernicka

et al.

1993. 48. Chernykh 1978; Jovanovic 1978, 1995. 49.

p. 313.

Crete,

Kutzi?n 1976, p. 70. 47. Todorova

2003.

a

assemblage, on the island, although

See Ryndina, Kolossova 1999, with

Indenbaum, detailed

and analysis

a part

remains

and the Knossos

its own

gone

the

of stone way,

and the long daggers crucibles,

pp.

1063

1066. 50. Renfrew Varna

finds,

Europe. 51. Strasser

1986b, p. 148. For the see First Civilization in

2002.

CHAPTER

i6o

13

of the EM period) very different from that attested in other regions.52 Yet even so, there are three silver daggers from Early surprising similarities.The a Minoan remarkable resemblance to a silver dagger from Koumasa53 bear the Hungarian

Middle Bronze Age.54 In his 1969 article, Renfrew had very little to say about the Aegean. His goal was to establish the validity of radiocarbon dating and to use the dates it provided to demonstrate the autonomy and cultural priority of cultures. The "Copper Age" of the Balkans could not as scholars, from Anatolia, possibly following especially Troy, was had long believed. This conclusion Childe, necessary because the were far earlier than comparable finds from the Balkans anything known from Troy or any other Anatolian site. Even when early metal-using sites European

Neolithic

be derived

such as ?atal H?y?k and Cay?n? Tepesi came to light in Anatolia, the tiny bits and pieces of copper from these sites could not possibly compare, itwas argued, with the massive copper implements of the Balkan Eneo and Eastern Europe were dismayed, for careers upon the primacy of Troy. Never they had based entire academic theless, Renfrew's arguments carried the day, especially after his dramatic of the case in a popular book entitled Before Civilization: presentation lithic. Prehistorians

The Radiocarbon

from Central

Revolution

and Prehistoric Europe.55 Renfrew was so sure of the cultural primacy of the Balkans, and of was to draw a "fault line" Europe in general, that he prepared chronological on from the and the Near East. separating Europe Aegean, Anatolia, Only new seem to the European side of this fault line did the radiocarbon dates force drastic, upward revisions of the traditional archaeological chronology. The map showing the fault line appears in the book mentioned above,56 two years earlier in an extremely influen but it had also been published tial Scientific American article entitled "Carbon 14 and the Prehistory of was This article and the fault line times, map reprinted many Europe."57 became justly famous. It was not long, however, before it became obvious that the basic premise illustrated by this map was rather misleading. The caption to the map indicated that the fault line only concerned dates after ca. 3000 b.c. At the time, however, the dates after 3000 b.c. from the east of the fault line were based on historical chronologies, while those from the based on comparisons with Troy, according to the chronology then accepted. Early radiocarbon dates from the Eastern Mediterranean

west were

and

the Near

East

were

the historical

ported and Mesopotamia. Neolithic

from

contexts

after

ca.

3000

b.c.,

and

they

sup

based on written

chronologies Early radiocarbon

tell sites, and they indicated

records from Egypt dates from the Balkans came from a

chronology

much

earlier than the

accepted "Trojan-based" chronology. the "fault line," but in reality it did not exist. With the wide Hence use of calibrated radiocarbon dates, discussed in 1973,58 by Renfrew spread revision of traditional chronologies became necessary almost everywhere, a and Anatolia. The past decade has witnessed especially in the Aegean our in of the Neolithic of the chronology transformation understanding a on our had in this and has Greece and Anatolia, profound impact period areas. in of metallurgical both evaluation of the development technology of cultural priority and technological autonomy, Conceptions secure be just 20 years ago, now have to be reformulated.

thought

to

52. Branigan 53. Branigan 54. Mozsolics

1967. 1968, 1967,

pl. 15. 55. Renfrew

1973.

56. Renfrew

1973,

57. Renfrew

1971.

58. Renfrew

1973,

p. 63. pl. 45.1;

Kov?cs

p. 105, ?g.

21.

1994,

pp. 69-83.

CHRYSOKAMINO

EARLY

AND

METALLURGY

EARLY METALLURGY

IN ANATOLIA

has change been more of metalworking

Nowhere

the development to be something

of a backwater

l6l

than in our understanding of in Anatolia. Once thought technology

dramatic

the Neolithic

period, Anatolia has of the ancient world, especially emerged dynamic parts once called the Aceramic Neolithic, but more correctly during the period referred to as Prepottery Neolithic A and B, or PPNA and B, ca. 11,000 a PPNA 6200 B.c.59The earliest known settlement atHallan ?emiTepesi, during

as one of the most

site dating to ca. 10,200 b.c., already had a series of semi-subterranean round houses, stone bowls with incised decoration (one showing a proces sion of dogs), and the bones of what seem to be domesticated pigs, along with

use of wild plants and animals.60 to is the located of modern Diyarbakir. To the northeast ?emi in the vicinity of modern Urfa, lie a series of sites in the area

evidence Hallan

for extensive

southwest, known asUpper Mesopotamia. The earliest of these, G?bekliTepe Ziyaret, structures with interior has a series of circular or oval-shaped T-shaped some 2 m in pillars. The pillars, height, have remarkable relief carvings

wild boars, lions, snakes, foxes, and perhaps even wild Asiatic These finds are slightly earlier than 9000 B.c.61 donkeys. to the the PPNB north, in the foothills of the Taurus Mountains, Just site of Nevali ?ori has five levels of occupation covering the second half depicting

B.c. Nevali ?ori has monumental stone buildings, stone monumental with likely shrines, freestanding sculptures in the round depicting human figures and a bird in flight. Smaller sculptures

of the 9th millennium most

a veritable bestiary: panthers, lions, wild boars, wild horses, bears, depict and vultures.62 A single copper bead from Nevali ?ori is one of the earliest metal objects from Anatolia.63 From the PPNB site of A?ikli H?y?k, 25 km a series of beads made of rolled in central Anatolia, thin sheets of copper provides even more extensive evidence for the early use of copper. A?ikli represents a real town surrounded by a city wall, the earliest inAnatolia, and it had an extensive obsidian industry. The site is dated to the first half of the 8th millennium B.c.64 of Aksaray

southeast

The best evidence for early copper metallurgy inAnatolia still comes from the site of Cay?n? Tepesi, one of the first Aceramic Neolithic settle ments to be discovered, and one of the earliest of this Excavations period. now uncovered some 7,000 m2 in the course of underway since 1964 have 17 field seasons. Most

of the metal artifacts, made of native copper ham into the shape of objects such as awls and fishhooks, come from the b.c. Two calibrated radiocarbon dates suggest the site mid-9th millennium was b.c. (also known from occupied much earlier, 10,000-9400 Cay?n? Nevali ?ori) has elaborate buildings with terrazzo floors and the earliest mered

for woven

evidence 59. Bar-Yosef

2001;

2002a,

pp. 379

table, p. 30; Aurenche

and

Kozlowski (1999, p. 68) put G?bekli

381. 60. berg

logical

textiles, which were made

Rosenberg et al. 1995.

1994,1998;

Rosen

ca. 8200

61. Beile-Bohn et al. 1998; Bischoff 2002. 62. Bischoff

at ca. 8500

2002,

with

chrono

b.c.

and Nevali

?ori

at

b.c.

63.Yal?in2000,p.

65.Muhly 1989; ?zdogan and ?zdogan 1999;Maddin, Muhly, and Stech from

19.

64. Esin 1999;Yal?in and Pernicka 1999; Yal?in 2000, pp. 18-20.

of linen.65

1999. The the

earliest

site of Mehrgarh

known

cotton,

in southern

Pakistan, dates to the first half of the 6th millennium

2002).

b.c.

(Moulherat

et al.

CHAPTER

i6z

the onset ofMellaart's

At

excavations

13

at the Konya

basin site of ?atal in 1961, the scholarly world was totally unprepared for the wealth H?y?k an area then as of finds from aNeolithic site inAnatolia, provin regarded cial and of little significance compared to Syria and Mesopotamia. Mel laart's excavations revealed elaborately decorated shrine buildings, wall paintings of hunt scenes and bizarre funerary rituals with headless corpses and vultures, and clay figurines depicting women in the act of giving birth. These finds came from contexts now dated to the first half of the 7th millen nium B.c.66There were

also small finds of copper, including beads made of like those from A?ikli H?y?k. These artifacts, strangely rolled-up were never but they are in the Konya Mu enough, published byMellaart, seum.67The small beads from level VIA, identified as lead in the publication, sheets

are

actually made

of the lead sulfide galena.68 can be seen as the culmination

?atal H?y?k in Neolithic Anatolia

of developments underway for over 3,000 years. This was the period known as a time when the drying and cooling of the environ

the Younger Dryas, ment stimulated a shift from a way of life to agriculture hunter-gatherer and settled communities.69 The Anatolian evidence clearly shows that the introduction of metal and the development of metal technology took place

of profound social and economic change, periods that saw the of specialized artistic work in crafts, highly sophisticated development was but one stone and clay, and monumental architecture. Metalworking in periods

innovations aspect of awhole series of technological of intense creativity. This remarkable fluorescence

taking place in periods of culture in Anatolia

came to an end toward the end of the 7th millennium

b.c. Severe climatic

b.c., as seen in the ice cores from Greenland changes dated and the pollen cores from Anatolian lakes, brought about the drought that an across to PPNB cultures Anatolia and the Levant.70 One of end put to ca. 6200

the last products of this remarkable age is the massive from level 2B at Can Hasan, southeast of ?atal H?y?k. to ca.

6000

copper macehead This find is dated

B.c.71

is, not surprisingly, serious opposition among some Aegean to the whole Younger Dryas climatic Eastern prehistorians and the argument that a sudden reversal of climate at change hypothesis the end of the last Ice Age led to the development of sedentary agricultural There

and Middle

In a 1997 monograph entitled Naissance des divinit?s, naissance de the Cauvin shifted argument from ecology and environment l'agriculture, to one based on human mentality of new religious and the emergence to Cauvin, Childe's ideas and symbols.72 According so-called "Neolithic in the world view of the Revolution" was a fundamental transformation

66. Levels 67. Yener

of the day, as attested by the elaborate architecture and religious societies, especially those of Anatolia.73 iconography of PPNB This issue is of some significance here because most of the scholars

who 6200 2200

believe

cultures to an end ca. changes brought PPNB that another sudden climatic change occurred ca. to this the Akkadian empire of Sargon I.74According

that climatic

B.c. also believe B.c., ending

climatic changes across the Eastern Mediterranean theory, contemporary of copper smelting activity coincided with the apparent intensification in the EM III period.75 Chrysokamino

at

Mellaart

1967.

pp. 23-24.

68. Sperl 1990.

societies.

societies

IX-IV; 2000,

69. Bar-Yosef

2001,

pp.

70. Bar-Yosef

2001,

pp. 26-28;

Bar-Yosef

2002b,

17-26.

p. 122.

71. Yal?in 1998. 72. Cauvin

1994.

73. Cauvin

s monograph,

noticed

in the Anglo-American

little world

of prehistoric has now archaeology, in an English translation appeared by

TrevorWatkins

(Cauvin 2000) and has

been

of a review

the subject

the

Cambridge Archaeological et al. 2001). (Cauvin 74. Weiss 1997.

75.Weiss 1998.

feature

in

Journal

and Courty 1993; Courty

CHRYSOKAMINO

AND

EARLY

METALLURGY

163

THE USE OF ARSENICAL COPPER In recent decades, two basic themes have dominated studies of the devel of Mediterranean and Near Eastern Eastern tech opment metalworking ca. b.c. 8000-2000 nology The first theme is the development of furnace technology, or the stages metalworkers first from moved whereby heating native copper to melting native copper in crucibles, then to smelting simple oxide and carbonate copper ores in crucibles, and later to smelting simple (and complex?) cop per ores in specially built furnaces. Initially, these furnaces may have been more than or without crucibles as linings pits in the ground, with nothing or on and collars top of the pits to increase the draft. This stage chimneys was followed by the use of the on the fully developed shaft furnace which, basis of present evidence, seems to have been a product of the Late Bronze the history of these developments attempt at writing Age. A pioneering has now been made by Craddock.76 other theme concerns the development of alloying technology, a to of related the of furnace complex developments integrally development no is evidence for the manufacture of any artificial or technology. There The

alloys with native copper as their base. Itwas only with the smelt ores that copper of ing complex copper alloy artifacts appeared. The first of these contained ca. 2%-6% arsenic. Later alloys contained 4%-10% tin.

manmade

The 4000

of arsenical copper alloys during the period from predominance to 2000 b.c. is one of the most significant discoveries of modern

research. Contemporary archaeometallurgical metallurgists, working chiefly in industry, long ago forsook any interest in arsenical copper. Their objective was, contrarily, to remove all trace of arsenic from metallic cop per. In the 20th century A.D., copper was of interest chiefly for its ability to conduct

traces of arsenic electricity, and the presence of even minute greatly reduced this capability.77 It was only when the Cambridge metal lurgist James Charles examined some of the Early Cycladic daggers in the at Oxford Ashmolean Museum that the importance of arsenical copper alloys in Early Bronze Age metallurgy was rediscovered.78 a Beginning with pair of articles published by Charles and Renfrew in the 1967, significance of arsenical copper from the Indus River to the Brit ish Isles has been documented in great detail through a series of analytical Much of the technical metallurgical research on the production programs. and the metallurgical properties of arsenical alloys has been carried out on artifacts from the New World, especially from the arsenical copper industries of ancient Peru.79 Some similar work has also been done on early arsenical alloys from the Old World.80 It has often been stated that arsenical copper and bronze (the alloy with tin) had more or less equal mechanical was properties and that it only the medical hazards involved in working

76. Craddock 2001. 77. Lechtman

1996,

p. 479.

78. Charles 1967; for his scholarly career,

see Charles

79. Lechtman

2000. 1996;

Lechtman

and

Klein 1999; Shimada andMerkel 1991. 80. Northover 1989; Budd and 1991. Ottaway 81. Northover

1989,

pp.

112-114.

arsenic that prompted a shift to tin. This is simply not correct. Bronze a is harder, tougher alloy. The main advantage of the arsenical alloy is its and the medical hazards have been greatly exaggerated, as all ductility,81 was done out in the open air, not in an enclosed environment. It smelting has also been established in laboratory work, making use of modern alloys

with

with different

concentrations

of arsenic, that the presence

of less that 2.0%

CHAPTER

164

13

arsenic has virtually no effect on the properties of the copper artifact.82 The shift from arsenical copper to bronze was a major technological develop ment of great cultural significance. use of arsenical

copper in the pre-Bronze Age Aegean must, at as be present, regarded problematic. The presence of arsenical copper is claimed in copper artifacts from Final Neolithic Kephala and the Zas cave, but these analyses are unpublished.83 The use of arsenical copper in the The

Thessalian

is said to have been demonstrated

Final Neolithic

of two metal

artifacts

by the analysis but this cannot be correct. Of the

from Dimini,84

two objects in question,85 only one, a flat axe, was to be 99.7% copper with no trace of arsenic.86 The axes from Sesklo87 were 99.3% copper. The metal

also made

analyzed. It was found two contemporary flat

of very pure copper, having

99.9% and

cave in Euboia

remain enigmatic. From contexts that Sampson regards as LNIIA (late 5th to early 4th mil lennia B.c.)88 come four copper artifacts, a dagger and three needles, made of arsenical copper. They have an average of 3.12% arsenic.89 It is curious and Ioannou publish the Tharrounia that Mangou analyses in their study finds from theTharrounia

of prehistoric copper-based artifacts from mainland Greece, but they say use of arsenical copper in a Final Neolithic this anomalous about nothing context. They are only interested in the presence of some 2% nickel of the arsenical copper needles from the Tharrounia cave.90

in one

It is theoretically possible that complex sulfide and arsenide ores were being smelted in the Aegean already during the Final Neolithic period. As mentioned above, such technological sophistication has already been claimed on the in the Eneolithic analysis of crucible period,91 based to ca. 4300 B.c.92 site of dated from the Durankulak, securely fragments Similar claims, involving the use of both copper and arsenic-rich ores, have now been made for contemporary contexts in Anatolia.93 The analytical for the Balkans

however, still supports the reconstruc one that an produced pre-Bronze Age metal industry, on the basis of surviving estimated 4,700 kg of copper, calculated solely was based on a very pure copper that can only artifacts. This industry represent the use of melted native copper.94 for such a copper smelting technology The Aegean documentation is even more ambiguous than the data from southeastern Europe. The evidence

from southeastern

Europe,

tion of an extensive

Sitagroi evidence, cited in the past in support of copper smelting activity at that site during Phase III, now seems more compatible with the simple cru as discussed above. Another a melting of (native?) copper in crucible, a B.c. at western context in M?ndalo cible from good late 5th-millennium Macedonia

has extensive vitrification

82. Northover 1989, p. 113; Budd and Ottaway 1991, pp. 138-139. 83. Sherratt 2000, p. 68. 84. Andreou,

Fotiadis, n. 46.

p. 267, 85. McGeehan-Liritzis

2000,

nos.

and Gale

34, 35.

p. 203, 86. McGeehan-Liritzis

1988,

and Kotsakis

and Gale

in its interior,95 but the way p. 217. 87. McGeehan-Liritzis

inwhich

1988,

1988,

p. 219,

nos.

91.Ryndinaetal. 92. Pernicka

and Gale

1996, p. 73. Sampson 89. Andreopoulou-Mangou 1993, and Ioannou p. 436, table 1;Mangou 1999, p. 84, table 1. and

Ioannou

p. 127.

2000, p. 35. et al. 1997, 94. Pernicka

pp. 41,

93. Yener

68, 69.

88.

90. Mangou

1999. et al. 1997,

1999,

p. 86.

121. 95. Papanthimou

and Papasteriou

1993, pp. 1209,1213, fig. 2.

AND

CHRYSOKAMINO

EARLY

METALLURGY

165

this crucible was

used has yet to be determined. Fragments of crucibles site of Kephala on the island of Kea;96 also come from the Final Neolithic were found with they pieces of slag originally thought to represent a fluid was too advanced for the Final Neolithic slag that period.97 Now, in light of the Chrysokamino evidence discussed in this volume, the Kephala slags need to be studied in greater detail. Two small amorphous lumps of metallic copper, found with four daggers and other copper artifacts in good Final Neolithic

at the cave,98 suggest some sort of local Alepotrypa a more but definite activity, interpretation must await future

contexts

metalworking

investigation. on Crete is still the Early metal production and usage subject of much debate and considerable confusion. Certain basic assumptions, eminently analytical

in the early years of the 20th century, still influence our present of the archaeological evidence. In publishing the metal finds from the Trapeza cave, excavated in 1936, Money-Coutts wrote, "Itmust reasonable

evaluation

be remembered

inhabitants

II period metal of any kind was practi so that the use and the skill rapid spread of its shew the receptiveness and ingenuity of the early

that until the EM

in Crete, cally unknown was it with which worked

of the island."99

or Yet nothing about the deposits within the cave itself necessitated even a conclusion.100 such nature discussion of the supported Pendlebury's of these deposits shows that the metal finds, both of precious and of base metal, came from a context that included a jumble of pottery, ranging toMM from Final Neolithic II. Because of the assumption made explicit in the foregoing quotation, none of the metal any earlier than EM II.101

96. Keos 97. The

I, p. 4, pis. 22, 66. was studied by S. R. B. slag

in the 1970s

Cooke

(Keos I, p. 114).

98. Papathanassopoulos 228. p. 99. Money-Coutts

1996,

this.

101. Pendlebury, Pendelbury, 1935-1936.

and

Money-Coutts

102.Xanthoudides 1918. 103.Muhly 2004b. 104. Davaras Betancourt

1971; Davaras

106. For 1984,

and

2004.

105. Day,Wilson, the finds,

to be

The degree to which old assumptions still influence current thinking is even more apparent in the case of the metal finds from the Pyrgos cave, in 1918.102 Recent discussions excavated by Xanthoudides of the finds from this cave show a consensus that the pottery is EM I, but there is a reluctance to call the associated metalwork EM I, because the metalwork includes a series of long daggers claimed to be no earlier than EM II.103 the pottery is EM I but the Thus, we are left with a situation in which isEM II. In the case of Hagia Photia, a metalwork large cemetery thought to cover the EM I?II period,104 there is also a general reluctance to admit that any of the metal finds from the more than 263 tombs could be EM I,

and Kiriatzi 1998. see

though the bulk (if not all) of the pottery clearly dates to late EM I.105 for theMesara tholoi, which were used for burials during the EM and MM periods, none of the bronzes from these tombs, including the 54 from As

1935-1936,

100. She would probably have been to admit

thought

even

p. 102. the first

artifacts was

Branigan

pp. 30-31.

107. Blitzer 1995, pp. 502-505.

Hagia Triada, the 56 from Koumasa, and the 95 from Pl?tanos, has ever been thought to be as early as EM I.106 The metal finds from Hagia Photia show all too clearly why scholars have been so reluctant to accept the obvious conclusion, one supported by the archaeological evidence. If the metalwork from these tombs is indeed EM I, then one has to accept the use of long daggers, small saws, knives, at B.c.Moreover, and fishhooks the of the 3rd millennium chisels, beginning a with the these tombs metalwork, along produced pair of footed crucibles, of a type known from slightly displayed in the Hagios Nikolaos museum, on later contexts at Kommos,107 as well as from Chalandriani Syros and

i66

CHAPTER

Thermi molten

13

on Lesbos.108 The Hagia Photia crucibles were clearly used to cast and would thus seem to testify to the actual working copper, they

on Crete come during the EM I period.109 Over 30 metal artifacts a a from this site, including silver zoomorphic lead bead, and a pendant, socketed spearhead of probable LM III date. In the 1980s, the Isotrace

of metal

at Oxford, then under the direction of Noel Gale and Zofia Laboratory 16 metal artifacts from Hagia Photia. The lead isotope Stos-Gale, analyzed seven ratios for of these objects were published in 1993,110 but the chemical composition data has only been published in the form of a histogram.111 The use at histogram makes it clear that the dominant alloy in Hagia Photia was

copper with 1.0%-6.0% arsenic. Is this the earliest arsenical copper in the Aegean world, apart from cave? Perhaps it is, on the basis of the the objects found in the Tharrounia to date. However, when published, evidence published the site of Poros, as the harbor town of Knossos, will take pride of place the most important arsenical

Imetallurgical site on Crete. Arsenical copper was certainly Early Minoan the alloy being produced and cast at Poros.112 Arsenical copper was also for the dominant alloy at Early Cycladic sites, especially daggers, but the or the Cycladic objects have traditionally been dated either to the EC II II?III period.113 Noel Gale has further suggested that the triangular daggers from Hagia Triada, also of arsenical copper, could be an EM I?II date,114 probably slightly later than the copper-based assigned transitional

EC

un analyses of nine site of Petromagoula, show that seven of these objects were made

Photia. The

artifacts from Hagia

electron microprobe artifacts from theThessalian

published copper-based located southwest of Sesklo,

copper, averaging 2.9% arsenic.115 The authors regard Petro as an EB II site, but this does not seem to be correct. Andreou, magoula at the very end of Fotiadis, and Kotsakis put the material from Petromagoula or I.116Johnson assigns the beginning of Early Helladic the Final Neolithic of arsenical

scheme.117 the site a similar date, ca. 3700-3300 b.c., in his chronological on the basis of It is hard to see any pattern in these developments our present knowledge. The arsenical copper artifacts from Petromagoula seem to be somewhat earlier that those from the so-called Palace hoard in copper objects found in level VIA at the site of Arslantepe eastern Turkey.118 Of the 22 artifacts making up this hoard, 8 have been or INAA.119 All analyzed using instrumental neutron activation analysis

of arsenical

proved

to be made metal

contemporary cesultan in western The

copper, averaging 4.16% arsenic. The at Late Chalcolithic hoard from level XXXIV Bey also attests to the use of arsenical copper. Anatolia

analyzed

eight

of arsenical

averaged

objects

108. Day,Wilson, and Kiriatzi 1998, p. 136; for the historical development see Mohen

of the crucible,

109. For

contemporary

110.

1993,

evidence

p. 122,

table

11.1. 111. Gale

1990,

p. 303,

Gale (1998, p. 725) says that only 10

discrepancy. 112. P. Day

see further Azzurro, in the text below.

Stos-Gale

Early

and Walter

Poliochni

discussion

arsenic.120 Among

were objects analyzed; only full publica tion of the data can clarify this seeming

1994. from

1.33%

fig.

1. Stos

(lecture,

Athens,

March

22,2004). 113.

Sherratt

p. 217,

table 4.

and Kot

117.Johnsonl999,p.320. et al. 2002. 118. Hauptmann et al. 2002, 119. Hauptmann 49, table 5. and Pernicka

114. Gale

1988,

Fotiadis,

p. 271.

pp. 47,

120. SAM 11774-11781; Muhly

2000.

1990, p. 301. 115. McGeehan-Liritzis

116. Andreou, sakis 2001,

and Gale

1992,

p. 312.

AND

CHRYSOKAMINO

METALLURGY

EARLY

167

5 contexts of the early 3rd millennium b.c., the site of Hassek in southeastern Ana of the Taurus Mountains located south just H?y?k, a rich collection of metal artifacts. Of the 76 objects that tolia, produced have been analyzed, not one had any tin, but 52 were made of arsenical copper, averaging 1.90% arsenic.121 It should be pointed out that this Late Ninevite

in Anatolia also involved the use of silver. The metallurgy of some of the short swords from Arslantepe VIA were decorated silver inlay, and a silver ring was found in the Beycesultan hoard. This

Chalcolithic handles with

silver is probably just slightly earlier than the silver pendant in northeastern Crete.122

from Hagia

Photia

was the dominant preceding evidence shows that arsenical copper use ca. B.c. This is true for Iran,123 4000-2500 during the period alloy in Syria,125 the Levant,126 and, of course, Anatolia.127 It has Mesopotamia,124 The

now been well

established that the Early Bronze Age was, for most of its an duration, age of arsenical copper. The actual production of this arseni cal alloy, however, still remains a matter of great controversy. It certainly was not made arsenic to molten is no copper. There by adding metallic that arsenic was

as a separate metal in ancient times, recognized a language had word for "arsenic." There are, however, various arsenic sulfide minerals that could have been added to the copper or could have been present already in certain types of copper ores that, upon smelting, would have produced an arseni cal copper alloy. Recently, it has been proposed that arsenical copper could have been produced by a simple cosmelting operation using copper oxide ores or iron together with either copper sulpharsenide sulpharsenide.128 evidence

and no ancient

What

about the production of arsenical copper in the EBA Aegean? the 1980s, it has been argued that the Cycladic island of Kythnos a and site. This site, known very impressive copper smelting possessed large as Skouries, has massive blocks of copper on today smelting slag still lying its surface. Skouries is dated by associated pottery to the Early Cycladic II period, with further support from two radiocarbon dates on pieces of Since

in the slag.129When calibrated, the radiocarbon the site fall in the first half of the 3rd millennium B.c.130

charcoal found

Of even greater potential of

the

smelting

site,

a

of

significance

copper

mine

said

just to the south

is the discovery, to contain

dates for

arsenical

copper

ore.

copper prills in the slag from the Skouries smelting site also proved to have a high arsenic content, with microprobe analyses of six different prills from one Kythnos slag sample averaging 6.3% arsenic.131Moreover, the lead

The

isotope analyses of the trace element lead in the ores and slags of Kythnos gave a lead isotope "fingerprint" that seemed to match the "fingerprints" obtained from analyzing the lead in many of the arsenical copper objects from

121.

Schmitt-Strecker,

Begemann,

and Pernicka 1992; for themetallurgy of the Ninevite 5 period, seeMuhly and Stech 2003. 122. Davaras p. 181. For

early 123. Thornton

and Betancourt silver, see Maran et al. 2002.

2004, 2000.

the Early Cycladic

period.132 It also matched

124. Stech 1999, p. 63. 125. De

Ryck,

128. Lechtman and Adams

Adriaens,

2003. 126. Golden,

Levy,

and Hauptmann

2001. 127. Palmieri,

Chernykh 1993.

Sertok,

and

the "fingerprints"

129.

Hedges 130. Stos-Gale 131. Gale

and Klein

p. 226. p. 720.

1998,

and Stos-Gale

p. 34, table 6. 132. Sherratt

1999.

et al. 1990,

2000,

1989,

pp. 72-92.

of

i68

CHAPTER

13

artifacts from theMinoan

Mesara tholoi (including Pl?tanos, Triada, Koumasa, Kalathiana, Marathokephalo, Hagia Hagios Onouphrios, and Porti)133 and even the copper-based objects from Hagia Photia.134 Arsenical copper tools also occur in an Early Cycladic hoard of al now in the British Museum. leged Kythnian origin that is They average copper-based

3.2% arsenic.135 In comparison, the 20 arsenical copper EC artifacts in the Ashmolean Museum average 4.31% analyzed by Gale and Stos-Gale can sure one never arsenic.136 Although be of the provenience of such ma terial, there seems to be no reason to accept the recent suggestion that the Kythnos hoard actually The original information to be

came from the Zas cave on the island of Naxos.137 that it came from Kythnos

itself ismore

likely

correct.138

In the 1990s,

it looked as if the combination

of chemical analysis and one of the solved had major problems of prehistoric analysis Aegean metallurgy. The arsenical copper artifacts from the Early Bronze seemed to be made of copper from arsenical copper ores Age Aegean the various mined and smelted on the island of Kythnos. Unfortunately, lead isotope

to threads of this argument soon started to unravel. Attempts replicate in subsequent the earlier results of Gale and Stos-Gale archaeological work at various Kythnian copper smelting sites and in further analyses of same results.139This issue was slags and ores have not yielded the at the Third Round Table on Aegean Archaeology, devoted to the for topic of "Metallurgy in the Early Bronze Age Aegean: New Evidence and at a colloquium devoted to "Early and Consumption,"140 Production new in the Aegean Islands."141 Much analytical work Copper Metallurgy

Kythnian discussed

on ores, slags, and copper prills from copper smelting sites by Bassiakos on to detect the presence of arsenic. This situation is has failed Kythnos raises it serious but questions regarding the theory that Kythnos puzzling, was

the source of arsenical

copper for the Aegean Early Bronze Age. At the same time, Gale and Stos-Gale are also having serious problems with their definition of a "Kythnian" lead isotope field. It now seems certain that not all the copper ores being smelted on Kythnos actually came from that island. The true picture appears much more complicated, with ores for from a number of different islands having been brought to Kythnos seems now to consist both of a high smelting.142 The Kythnos field itself to conclude, "We need still more leads Stos-Gale and a low field. This to be finally sure what the term analyses of slags and ores from Kythnos seem that we have yet to solve all 'Kythnos field' really means."143 Itwould the problems presented by arsenical copper in the Aegean Early Bronze the smelting of arsenical copper ores at Chrysokamino Age. Obviously, seen as must be but one chapter of a complex story, one that we are still in the process of understanding. 133. Gale 1990, p. 315, fig. 7. 134. Stos-Gale 1998, pp. 720, 725. 135. Craddock 1976, pp. 98,106. 136.

Sherratt

2000,

pp. 72-92.

137. Fitton 1989; Sherratt 2000, p. 92.

138.Muhly

1999, p. 17.

139. Work Service

by the Greek Archaeo was under the direction

logical of Olga Philaniotou-Hadjianastasiou. was undertaken The analytical work at the Laboratory at Demokritos, Center

of Archaeometry the Greek National

for Scientific

Research,

led by

Yannis

Bassiakos.

140. This

conference

Sheffield University 141. This

was

held

at

in January 1998.

colloquium

was

Demokritos

on April 28,2001.

142.

Stos-Gale

1998,

143.

Stos-Gale

1998,

held

at

pp. 723-724. n. 36.

p. 726,

CHRYSOKAMINO

AND

EARLY

METALLURGY

169

THE USE OF BRONZE copper, of course, is only half of the technological history of Early more Bronze Age copper-based alloys. Even complex is the development of the copper-tin alloy known as bronze (and the term "bronze" should be

Arsenical

is and tin-bronze used only for this alloy; arsenical bronze is a misnomer, s of the Pare Since 1985 redundant). survey my study problem,144 magisterial the situation in the Aegean still presents has appeared.145 Nevertheless, serious problems, both archaeological and analytical. At the outset, a dis tinction needs to be made between the limited use of bronze, the existence of bronze as but one of several competing copper alloys, and the full use to the exclusion of other alloys. Pares discussion and his map use in the of of the bronze the full bronze put showing metallurgy spread ca. 1600-1400 same now at the date for b.c.,146 Syria and Aegean suggested such a date does not mark the introduction of Obviously, Mesopotamia.147 or even the first serious use of bronze. bronze metallurgy In recent decades, it has been proposed that bronze may have been B.c. in Chalcolithic used during the second half of the 5th millennium of bronze

now been put to rest. In speculations have Europe.148 These we take a all the Pernicka and evidence, reviewing colleagues conclude, "If closer look at the proposed early finds of tin bronze in southeastern Europe southeastern

not

a

one

single

What Aegean Minoan

needs

stands

up

to close

scrutiny."149

in detail here are developments in the b.c. and the use of bronze in millennium early

to be examined

in the mid-3rd

Crete. It is in the first half of the 3rd millennium, especially in the later part of this period, that we find a substantial quantity of bronze at the time of the Royal Cemetery of Ur. InMeso in use inMesopotamia,

144.Muhly 145. Pare

2000.

146. Pare

2000,

147. De

Ryck,

Ilia or 2600-2500 b.c.150 The tin potamian terms this is Early Dynastic seems come to have from central Asia and Afghanistan,151 for this bronze

1985b. p. 26, fig. Adriaens,

1.14. and Adams

2003, p. 579. 148. Schickler 1981. 149. Pernicka

and Mari.153

et al. 1997,

p. 125.

150. Stech 1999, p. 63. 151. Alimov

et al. 1998;

153. Pinnock,

See

Boroffka

1985b.

the articles

and Scandone

154.Muhly

inMatthiae, Matthiae

1995.

1985b; for the complex

of these terminology metallurgical and Bachmann texts, seeWaetzoldt 1984. 155. Palmieri

and Hauptmann

2000, p. 1260. 156. Palmieri

and Hauptmann

2000, pp. 1261-1264. 157. Stech and Pigott 1986, pp. 55 58.

158.Muhly 1999, pp. 18-20. 159. SeeMuhly 1993.

The

texts from Ebla, dating to Early Dynastic IIIB and the following contain so-called for period, recipes making bronze, adding

Old Akkadian

et al. 2002.

152.Muhly

along with gold and lapis lazuli.152 So-called Royal Treasures and Royal Tombs containing significant amounts of lapis lazuli and gold are known from contemporary or slightly later contexts in Syria, especially from Ebla

specified quantities of tin to specified amounts of copper in order to produce a quantity of bronze for a specific even purpose.154 Some of these texts of "classic" bronze with 10% tin.155The bronze describe the manufacture now been objects from Ebla that have analyzed contain up to 13% tin.156 In 1986, Stech and Pigott presented a very convincing argument for B.c. bronze and tin as status metals in the mid-3rd millennium They argued, route one or went the main that tin trade for this however, by sea, from more was to across It then distributed from Anatolia Syrian ports Troy. Troy as well. In and, presumably, to the north Aegean exchange, Troy supplied with Anatolian silver.157 Such a trade pattern is, Syria and Mesopotamia of course, possible, but as I have already argued, it ismost unlikely.158 The situation inAnatolia is ambiguous at best, even if one iswilling to discount the possibility of Taurus sources of tin.159The lapis lazuli trade route cer went as farwest as Ebla, where blocks of raw tainly lapis from Afghanistan

CHAPTER

170

have been found

in Palace G

in late-3rd-millennium

13

b.c. contexts.160 The

of EBA Anatolia, however, seem to have had no interest in this semiprecious stone. The only exception is the enigmatic battle-axe of lapis lazuli from Treasure L atTroy.161 One isolated example, in amost unusual inhabitants

context

that also included an axe made

to support aTroy-Byblos What makes things

of jadeite, is hardly enough

evidence

trade route.

even more our present understand complex is at at the situation For least the past 100 years, Troy II ing of Troy itself. and its north Aegean Poliochni Giallo (V) have been seen counterpart as part of an EB II of In Gold." the mid-3rd millennium b.c., this "Age across the Old World on Crete to from Mochlos phenomenon spread Troy II, Alaca H?y?k, Eskiyapar, Ebla, Tell Brak, Mari, Ur, and perhaps even as far east as the sites along the Indus river. The EB II Harappan of Gold" hypothesis has, however, been called into question recently "Age of romantic sensationalism who refers to it as "a combination by Nakou, and an overwhelming to historiase tendency

desire for order in the past" and as part of a "general the Early Bronze Age."162 Nakou is certainly correct to the fact that the evidence comes from a series of

in calling attention hoards that do not necessarily represent finds from any one period. They are better seen as was current for consid "palimpsests of material which erable time, and would probably have escaped deposition were it not for extraordinary circumstances in the late millennium."163 In other words, what in precious and base has preserved for us is the metalwork, happenstance was in that circulation the mid-3rd millennium B.c., exactly metals, during the period when bronze became an important prestige alloy throughout the Eastern Mediterranean and the ancient Near East. does all this mean for Early Minoan Crete? To answer that ques a is of comparative archaeological tion, proper understanding chronology B.c. does not represent the essential. First of all, the mid-3rd millennium first use of bronze on Crete. I have long argued for an EM I date for the What

two small daggers from the Krasi tholos, located just outside the northern over 70 years border of the Lasithi plain and excavated byMarinatos ago.164 was a careful excavator, and his detailed report makes it clear Marinatos that the daggers in question came from the earliest burials in the tomb, associated with

the distinctive

EM

pottery known as that "the small primitive blades from

Pyrgos Ware. Branigan has argued Krasi might well belong to the very Crete."1651

would

160. Pinnock

I pattern-burnished

start of the Early

Bronze

Age

Treister

in

agree.

and Pyrgos170 cave

1.85%

averaged

arsenic.171

B.c. terms, however, the mid-4th millennium a new new of the emergence marked complex, involving metallurgical as and flat such of axes, types double-spiral pendants, daggers, objects In wider

comparative

pp.

162. Nakou

1997,

163. Nakou

1997,

p. 637. p. 637.

1929. p. 215.

1967,

166. SeeMuhly fig.

1973, pp. 339-342.

167. Marinatos

1929,

13, nos.

SAM

30-31;

and

Tolstikov, 148-152.

165. Branigan

pp. II/3,

119,131, nos.

9447,9449.

is scarce. The

from Salami169 roughly contemporary metalwork is all of arsenical copper. The five objects from the Pyrgos

from Crete

1996,

164. Marinatos

two daggers, however, are made of bronze. Analyzed by the SAM Project in the 1960s,166 they proved to contain 10% and 6% tin.167 to current understanding of Early Minoan these chronology, According These

b.c.168 Is the use daggers belong to the second half of the 4th millennium of bronze at such an early date a realistic possibility? Comparative material

1988.

161. Antonova,

168. Warren 169. This

1980, comprised

p. 489. two

daggers:

SAM II/3, nos. 9444 and 9445. 170. This

comprised

three

daggers,

needle, and a chisel: SAM II/3, nos.

9366-9370.

171. SeeMuhly 2004b.

a

CHRYSOKAMINO

new

AND

EARLY

METALLURGY

171

over a broad complex extended expanse, including southeastern Europe (Usatovo), western geographic eastern Anatolia Anatolia VII and VIA), and (Ihpinar IV), (Arslantepe The metalwork Palestine from (Nahal Mishmar).172 Sitagroi IV is part of along with

sources of metal. This

this complex. The site has artifacts made of bronze, two metal fragments with 3.2% and 5.9% tin respectively.173 The scholars involved in the pub lication of these analyses have gone out of their way to urge caution in the interpretation of the analytical data,174 but these results are not necessarily as anomalous as they once seemed. The most use of bronze interesting early comes not from Krasi but from the site of Kastri on in the EBA Aegean the island of Syros, excavated over 100 years ago by Tsountas. There is no a point in belaboring here the confused analytical results from single set of Kastri metal

samples originally taken for the SAM Project.175 For analyti it is best to deal instead with results of the SAM Project cal consistency, in 1967.176What Bossert has always been remarkable about published by seem to document a analyses is that they metallurgy involving of arsenical copper and bronze. I would combination still regard this as a transitional were period, when metalworkers representing adding tin the Kastri

sense can the material from Kastri be copper.177 But in what as transitional? regarded The pottery from Kastri is now the subject of amajor international re search project because it seems to document the arrival of Anatolian colo to arsenical

in the Cyclades, paralleling the EM I arrival of Cycladic colonists in eastern Crete, in the of Photia.178 Central to this particularly region Hagia discussion is the date of the ceramic assemblage from Kastri. In Sotirako nists

s recent study of the "Kastri Group," said to include Kastri, Lefkandi, poulou the pottery is classified as transitional EH II/III,179 Rafina, and Manika, but this study simply ignores the metalwork from all of these sites. a new In a 1986 essay, Mellink of proposed radically understanding 172. Pernicka

et al. 1997,

173. Renfrew

and Slater

table

p. 306,

8.2, nos.

EBA Troy and its role in the Eastern Mediterranean during the 3rd mil lennium B.c. This reconstruction of traditional archaeological sequences has profound of the development of implications for our understanding

p. 57. 2003,

and Gale

McGeehan-Liritzis

1988,

p. 216, table 4, nos. 24 and 25. 174. McGeehan-Liritzis and Gale 1988,

pp. 220-221;

2003,

pp. 313-314;

Renfrew

and Slater

see also Begemann, and Pernicka 1992,

Schmitt-Strecker, p. 223.

175. SeeMuhly p. 18. 176. Bossert

1991, pp. 362-364;

1999,

1967,

p. 76,

table

178. Betancourt Sotirakopoulou

1;

2003c. 1993.

180.Mellink 1986, pp. 146-148. 181.Mellink 1986, p. 151. 182. For Troy II and Poliochni Giallo as EB IIIA sites, see Efe and Ilasil997,p.600,fig.2.

seen as

at the end of Troy correlating with destructions Ij, Poliochni Verde was followed the Transitional (III), and Thermi V What period associated with pottery assemblages from Rafina, Manika, Lefkandi I, and Poliochni Rosso was

Muhly 1985a, p. 127. 177.Muhly 1999, p. 18. 179.

our metallurgical technology and appreciation of the de s InMellink velopments documented by the excavation of Chrysokamino. at of II the the the close EH destructions of the that pottery, reading period put an end to Lerna III and the "House of the Tiles" at that site are now 3rd-millennium

5624,5625;

(IV). The a

handmade,

distinctive one-handled,

pottery

type fossil of this transitional

red-polished

period

tankard.180

But this tankard is not to be found atTroy because, with level Ij having been destroyed byWest Anatolian marauders from Tarsus, the site of Troy lay in ruins during the Transitional period. In the following EB IIIA period was

as a

"triumphant station built by a successful warrior new breed ofWest Anatolian king of the navigators and traders."181This was with at Kastri and Poliochni similar settlements Troy II, contemporary Troy

Giallo

resettled

the exception of Lefkandi, all of these sites from the (V).182With and from EB IIIA have associations. Itmust period important metallurgical

CHAPTER

172 be remembered, however, that very little of EBA Lefkandi site) has been excavated to date.

13

(the Xeropolis

at Rafina uncovered two Rescue excavations conducted by Theocharis areas of copper activities with apparent remains of a smelting smelting no proper with and furnace, along tuyeres.183 Unfortunately, slags, molds, was made at the time of excavation, and any finds study of the material can no to inAthens the National Museum brought longer be located. The metal

finds

from Manika

have been

in detail by Stos artifacts from the site,

studied

the more

and Mangou.184 Of than 50 metal 20 copper-based objects and 1 of silver were analyzed by atomic absorption spectroscopy (AAS). The copper-based objects contain both arsenic and tin, but in contrast to the situation at Kastri, a clear separation was maintained

Gale

one the two alloys.185With exception not to be considered here, the objects contained either arsenic or tin, but not both. The analytical results establish very clearly that both arsenical copper and bronze were available to the smiths who made the metal artifacts found atManika. The

between

12 objects of arsenical copper (i.e., those with over 1.0% arsenic) averaged 3.5% arsenic. The five objects having over 2.0% tin averaged 5.9% tin. The possible reasons behind the choice of alloy are not at all clear. Pins and knives were made

of both alloys, while

only of arsenical copper. The situation at Poliochni late 1980s, the Heidelberg-Mainz artifacts. This study used both

daggers

and chisels were made

the most complex of all. In the team analyzed a total of 97 copper-based instrumental neutron activation analysis

is perhaps

(INAA) and atomic absorption (AAS), the latter only for spectroscopy lead and bismuth.186 The samples came from all four metal-using periods at Poliochni, and the number of samples taken (shown in parentheses) gives a (6), Verde good indication of the growth of metal usage at that site:Azzurro terms In of and Giallo Rosso (28), (52). (11), comparative archaeological these four periods cover almost the entire Early Bronze Age, chronology, of Troy I, if not earlier, and with Azzurro dating to at least the beginning Giallo

certainly extending Arsenical copper was

to the end of Troy II. already in use in the Azzurro

phase, with four of the six analyzed objects averaging 1.96% arsenic. There was no tin, but bronze made a slight appearance in the Verde phase. Of 11 analyzed objects, 6 were of arsenical copper, averaging 2.2% arsenic, and 2 were of bronze, use of bronze increased during the Rosso phase, averaging 9.25% tin. The toMellink's corresponding period following the destruction of Troy Ij.Of were 28 analyzed objects, 17 of arsenical copper, averaging 2.8% arsenic, and 7 were of bronze, averaging 8.16% tin. A single, very unusual artifact had 19.8% tin and 11.71% arsenic; it is hard to imagine what such an aberrant alloy might It is remarkable

represent. that during

the Poliochni Giallo phase we find the 17 copper (with objects averaging 1.94% arsenic), bronze (with 14 objects averaging 9.17% tin), as well as the same mixture of the two alloys found at contemporary Kastri. There are 12 objects with more than 1.0% arsenic and 2.0% tin. Taken together, they average 1.34% arsenic

use of arsenical

and 6.68%

tin.

183. Theocharis 1955. 184. Mangou 185.

Stos-Gale,

Sampson,

and

1996.

Stos-Gale, Sampson, 1996, p. 51, table 1. Mangou et al. 1990. 186. Pernicka

and

AND

CHRYSOKAMINO

EARLY

METALLURGY

173

figures do not take into account the artifacts. Already in the Azzurro phase, all had over 1.9% lead, averaging 3.1% lead. This assume that there was a high trace element only the arsenical copper ores being utilized at Poliochni. These

Poliochni

incidence

of lead in the

the six analyzed artifacts ismost unusual; one can concentration of lead in

activity Metalworking of some sort was underway at Poliochni already during the Azzurro phase, team analyzed the slagged I;187 the German contemporary with Thermi interior of six crucible fragments from this period at Poliochni.188 Very little lead was

in the objects from the Verde and Rosso phases, or in those of the Giallo, but there are several notable exceptions in the lat are made of leaded arsenical ter from Two Poliochni Giallo objects phase. found

lead, and two are of leaded At least three objects have lead. bronze, averaging small amounts of arsenic, tin, and lead. Clearly the copper-based alloys in use were more in all Poliochni Giallo than those found during complex copper,

averaging

previous The

2.01%

arsenic

9.55%

tin and 9.33%

and 9.05%

phases at the site. site of Thermi on Lesbos

has played a significant role in all discus sions of early Aegean metallurgy for over 70 years. Lamb, in her excellent final report on excavations at the site,189 paid careful attention to all the

evidence uncovered, and she even had some of the metal ar metallurgical tifacts analyzed by Desch, the scholar who best deserves the title of "Father of Archaeometalfurgy." in there has been a long discussion Consequently, the scholarly literature regarding the crucible fragments from the lowest levels at the site (Thermi I) and, from the same context, a bronze pin, with 13.1% tin, often cited as the earliest object of bronze from the Aegean world.190 From his detailed analysis of all the relevant pottery, Podzuweit concluded that Thermi I and II were contemporary with Troy la and that III was contemporary with Troy lb.191A bronze artifact of such an as we have seen, not without early date is unexpected but, precedent. The context and analysis of this pin have now been called into question,192 but

Thermi

it is probably best to accept the original report. Unfortunately, not included in the new programs of Thermi analyses because

the pin was it could not

in the museum.193

be located

recent programs of analyses194 included 19 objects from Thermi I. were all made of arsenical copper, averaging 1.91% arsenic. Of the They seven artifacts studied from Thermi II, only one was bronze, with 2.4% tin. Five were arsenical copper, averaging 2.36% arsenic.195 The Oxford The

187. Schmitt-Strecker, Begemann, and Pernicka 1992, p. 221. et al. 1990, p. 269, 188. Pernicka

189. Lamb 1936. 190. Lamb 1936, pp. 214-215, no. 31.64.

The

a division

metallurgical

issued

present special complications. metal from the site were artifacts

1979,

pp. 38-40.

team with

the

authorities.

The

strong British

192. Gale 1996, p. 118.

members

193.

because

Begemann, and Pernicka 1992,

Thermi

Schmitt-Strecker, p. 223.

protest, School

support of the Greek team Oxford lodged with

should Thermi

the

support

of the

at Athens, have was

its claiming been consulted

originally

exca

vated by the British. This resulted in

of the sides samples. Both to each tried and publications,

discredit

sampled by theHeidelberg-Mainz

table 4.

191. Podzuweit

194. The studies

the work

of the other. Having to reopen what has been a most I pres unfortunate controversy, ent here what I feel to be a reasonable no desire

a

account of the development of EBA at Thermi. metallurgy 195. Begemann, and Pernicka, Schmitt-Strecker 1995, p. 125, table

1.

CHAPTER

m

13

analysis of these samples gives somewhat different results. Of 34 analyzed I and II, 1 was bronze (a punch or drill with 4.12% objects from Thermi tin) and 24 were

arsenical copper, averaging 2.62% arsenic.196 Of the 16 ob III, again only 1was bronze (a pin with 4.2% jects analyzed The 12 of arsenical tin). copper averaged 2.22% arsenic. The nine objects no objects of bronze, and the six analyses from Thermi IV produced objects of arsenical copper averaged 2.43% arsenic. The eight analyses of objects from Thermi

two objects of bronze, V produced averaging 10.7% tin, and six of arsenical copper, averaging 3.41% arsenic. In all five periods at a clear Thermi, separation existed between arsenical copper and bronze; no amixture of both metals. No leaded bronzes were had object analyzed, but two objects were leaded arsenical copper. One artifact from Town III from Thermi

had 2.83%

arsenic and 2.64%

lead; one from Town

IV had 3.8% arsenic

and 3.2% lead. Given the relatively small number of objects involved, little can be said about the comparative use of the two alloys. Single bronze pins came from Thermi I, II, III, and V, but 11 pins were made of arsenical copper: 4 from Thermi III, 4 from Thermi IV and 3 from Thermi V Arsenical copper was also used

to make

chisels: 2 from Thermi

III and 3 from Thermi

to these figures. Thermi

V No

was

great significance should be attributed clearly a center of Bronze Thermi major Early Age Aegean metallurgy. Although seems to have had one of the earliest examples of bronze from the Aegean, as well

as the use of arsenical copper only tin object from the area,197 the seems to have at the site. This would support the idea that predominated a time contemporary with at the occupation of Thermi ended probably

Poliochni

Verde,198 before

the real expansion of bronze metallurgy. as itwas be study has shown that arsenical copper,

This

comparative ing produced in the Aegean Early Bronze Age, most often had an arsenic content of just under 2.0%. Yet the modern metallurgical studies cited above all conclude that an arsenic content below 2.0% has little effect on the then, did physical properties of the metal being produced. Why, so an Bronze metalworkers Early Age consistently produce alloy that gave them a metal with the same basic properties as unalloyed copper? I have no

answer

to

this

question

at

the

present

time.

In many respects, all of this research began over 100 years ago in the Troad. From the beginning of his work at Troy in 1870, Schliemann an interest in the development demonstrated of metal technology and a to have metal artifacts examined and willingness analyzed by competent the basis of this analytical work, he concluded that there was no use of tin in what is now called an Troy I, opinion supported by later time the of his Great Treasure, which Biegen by scholarship.199 However, was use of bronze. A total in for the there evidence Troy Ilg, placed ample scholars. On

of nine flat axes from this treasure, analyzed for Schliemann by metallur gists in England, France, and Germany, averaged 5.20% tin.200 The work.201 itswork

use of bronze

inTroy

recent analytical team the results of published

II has been confirmed

In 1984, the Heidelberg-Mainz in northwestern Anatolia.202 One

by

table gives the analytical results from the study of 50 artifacts contemporaneous with Troy II from north

196.

bracelet

1992, p. 175, app. 3. a remarkable tin

Stos-Gale

197. This

was

1936, pp. 171,215, pl. 25. 198. Begemann, and Pernicka 1992,

199.Muhly Begemann,

IVa; see Lamb

from Thermi no.

30.24,

fig. 50,

Schmitt-Strecker, p. 224.

1985b, pp. 283-284;

Schmitt-Strecker,

and

Pernicka 1992, p. 223. 200. Muhly and Pernicka 1992, p. 310. 201.Easton2002,p.327. et al. 1984. 202. Pernicka

AND

CHRYSOKAMINO

EARLY

METALLURGY

75

these 50 objects, 30 had over 4.0% tin, averaging were at Troy itself, 28 objects analyzed. Results were published only in terms of whether the object was made of copper, arsenical copper (having over 2.0% arsenic), or bronze (having over 1.0% tin).204 On this basis (and it should be noted that the Heidelberg

western 7.95%

Anatolia.

Of

tin.203 From Treasure A

definitions of arsenical copper and bronze are the reverse of those used here), only 1 object was made of arsenical copper; 16 pieces were of bronze. This table also includes the results of analytical work done on a hoard of bronze objects from the Troad, purchased by Istanbul University

Mainz

and published by Bittel.205 The objects consisted mainly of attachments for a type of metal "teapot," an example of which had been excavated by Schliemann.206 Of the 22 objects analyzed, 15 were bronze and 6 were ar senical copper. Bittel had already published these analyses in 1959, giving actual percentages.207 Twelve of these bronzes had the "classic" amount of ca. 10% tin. The six objects of arsenical copper averaged 3.27% arsenic. team published In 1985, the Oxford its own analytical work, making use of the same use in the SAM taken for samples originally Project. They published neutron activation analyses (NAA) of 15 objects from Troy II.208 Nine proved to be bronze, averaging 8.23% tin, and four were arsenical copper, averaging 2.45% arsenic. But this hoard of metalwork, designated as the Great Treasure, Treasure A, or Priam sTreasure, is no longer dated to the end of Troy II (Troy Ilg). Recent work directed at Troy by Korfmann would now place this metalwork in the middle of the Troy II phase, with an absolute date of ca. 2500 B.c.209 Korfmann does not accept Mellink's 1986 revision of Trojan stratigraphy. The gold and bronze metalwork from now as seen in other is with that from the words, Troy, contemporary Royal Cemetery of Ur. This same combination of gold jewelry and bronze metal can be seen in central Anatolia, lurgy especially at the site of Alaca H?y?k, a where group of so-called royal tombs roughly contemporary with the "treasures" of Troy II were excavated.210 The metalwork from Alaca shows the same combination of arsenical copper and bronze. Of 36 objects from analyzed by the SAM Project, 12 had over 5% tin, and 13 had over 1.0% arsenic.211 The very strong Caucasian connections of the finds from are of considerable Alaca H?y?k interest.212 Alaca

In all the studies dealing with copper-based artifacts from Anatolia and the Aegean undertaken by the Heidelberg-Mainz and Oxford teams, chemical elemental analyses have been combined with the study of lead years lead isotope ratios have isotope ratios. Over the past twenty-some come to be seen as the best way to the study provenience of copper, silver, has generated an enormous bibliography and a great over deal of controversy, especially Lavrion and Cyprus as sources of copper for the Aegean Bronze Age. Both the and the Oxford Heidelberg-Mainz and lead. This work

teams agree that the real introduction 203. table

204. 579,

Pernicka,

et al. 1984,

p. 575,

3. Pernicka

et al. 1984,

table 4.

205. Bittel 1959.

pp. 578

206. Bittel 1959, p. 14, fig. 28. 207. Bittel 1959, p. 34. 208. Gale,

Stos-Gale,

p. 148, table 209. Korfmann

1985,

and Gilmore

occurring

p. 380.

in the middle

210. Mellink 1956. 211. Stech and Pigott 1986, p. 54.

212. Mansfeld 2001.

1. 2001,

of bronze,

CHAPTER

176

13

b.c. along with the use of tin rather than arsenic as is associated with the appearance of a new, so-called the alloying metal, "exotic" lead isotope signature. Bronze artifacts, especially those from Troy and Poliochni, tend to plot right off the lead isotope diagram. That is, they

of the 3rd millennium

a lead that is older than or from anything known from the Aegean ore This lead could only have come from Precambrian deposits 700-900 million years old, and such deposits do not exist inGreece, Turkey, or theMiddle East.213 Cyprus, contain

Anatolia.

is being studied here, of course, is the geological date of the lead as a trace element in artifacts. As this "exotic" lead copper-based

What contained

does not appear in the arsenical copper artifacts, the logical conclusion would be that it came from the tin. This lead might in turn provide evidence alluvial cas provenience of Bronze Age tin. However, of tin and presumably the source of this "exotic" lead, to contain virtually no lead.214 It is now being proposed

for the long-sought siterite, the dioxide has been

shown

that the lead came with

the copper. In the mid-3rd millennium B.c., sources of copper were being exploited by metalworkers, and the northeastern Aegean.215 those in northwestern Anatolia new

therefore,

especially Such copper is even found in objects and II.216This copper must have come favored as the source of Bronze Age was the copper mixed with the tin in

of bronze, as inThermi I from central Asia, the very area now tin (see above). But how and where

not made

order to produce bronze? If this was in it is possible that Early Bronze Age Asia done central then itself, being in metalworkers were, fact, smelting stannite, an ore containing both copper and tin. Such ore is found in central Asia, especially at the site ofMushiston

inTajikistan. This without iron. The

area has deposits of stannite high in copper and tin, but ores "would result in copper with smelting of these high

stannite, the usual iron content was replaced cannot it and be accidental that some of the objects from Thermi by zinc, to be so in zinc that they were considered possible examples proved high of the early production of brass.218 tin contents."217 InMushiston

the distinctive

Given Eastern

Mediterranean

typology

of Early Bronze Age

it seems

world,

artifacts from the

to assume

reasonable

that

this metal

in the form of bronze ingots and was turned into into the Aegean at all of The evidence for metalworking artifacts by local metalworkers. at Poliochni and the relevant sites supports this conclusion. Metalworking

came

has already been discussed; the evidence for Troy II has recently bronze ingots are something summarized by Easton.219 Admittedly, seems no an feasible at the present time. but other of anomaly, explanation

Thermi been

lead isotope analyses indicate that this metal, with "exotic," very early contexts inOman and the United to be found in contemporary is also lead, Arab Emirates.220 The implications of this, obviously, are enormous. Recent

213. ments,

For see

a summary of these and Pernicka

Muhly

216.

Stos-Gale

217.

Boroffka

218. There

pp. 312-315.

214. Muhly 1978, p. 45, table 1. 215. Pernicka 1987, pp. 702-703; Muhly 313.

and Pernicka

1992,

1992, p. 170. et al. 2002, p. 141. are four averag objects

argu 1992,

pp. 312

ing 7.6% zinc; p. 175, app. 3. 219. Easton

220.Weeks

see Stos-Gale

2002,

2003.

p. 328.

1992,

CHRYSOKAMINO

AND

EARLY

METALLURGY

CHRYSOKAMINO AND THE DEVELOPMENT METALLURGICAL TECHNOLOGY

177

OF

seems to In light of the preceding discussion, the site of Chrysokamino a of into threshold the Middle Bronze On the one occupy something Age. terms in its of what be called hand, smelting technology, might proto-shaft furnaces and the use of pot bellows, seems to be a harbinger of what is to come; on the other hand, its use of arsenical copper reflects the technol ogy of previous centuries. Arsenical copper has already been discussed. It is appropriate, therefore, that this chapter close with a brief discussion of over a century, scholars have smelting technology. For struggled to under stand the curious fragments of perforated clay that littered the surface of In 1910, Mosso took them to be fragments of crucibles.221 Chrysokamino. Our current thinking on this subject is well described by Betancourt in 14 of this volume. Similar fragments are known from the Chapter Early II smelting site of Skouries on Kythnos, mentioned above. Al are seem come to these from though fragments largely unpublished, they more substantial, structures than the Chrysokamino larger, and thicker were used, and the idea of their some sort However examples. they being of a cover for a bowl furnace ismost attractive, they must have been de signed to increase the flow of air into what was awind-powered smelting operation (with additional draft supplied by the pot bellows in the EM III Cycladic

period). Wind-powered smelting technology was probably widely used all over the world. It has now been well documented for iron smelting furnaces inMedieval

Sri Lanka.222 Roughly contemporary with are Chrysokamino the Early Bronze Age examples from the site of Feinan in Jordan.223 It is interesting that 20th-century copper smelters were unable to achieve a separation of slag and metal in replication experiments done such a separation is now being proposed for con was achieved with the additional temporary Chrysokamino. Perhaps this draft provided by the pot bellows, but there is also no evidence for any sort at Feinan,224

221. Mosso

1910,

pp. 290-291,

fig. 164. 222. Juleff 1998; Craddock 2001, p. 162, n. 47. 223. Bunk 224.

Bunk

et al. 2002. et al. 2002,

225. Golden, 2001.

Levy,

p. 336. and Hauptmann

226. Golden,

Levy,

and

2001, pp. 956, 958, fig. 8.

Hauptmann

although

on of perforated structure at Feinan. Yet a primitive structure positioned site of top of the bowl furnace was already being employed at the Negev to a dated the mid-4th millennium B.c.225 At bowl Shiqmim, Shiqmim, furnace, consisting of nothing more than a pit dug in the ground, possibly lined with a crucible, was surrounded on the surface by a sort of clay collar about 5 cm in thickness, 9 cm in height, and 30 cm in diameter. The original collar was about 80% of a circle, leaving an open space for the draft of air.226 It is probably not too fanciful to see this Shiqmim collar as the precursor on of the structures used at Skouries on Kythnos and at Chrysokamino a Crete millennium later.

CHAPTER

14

of

Discussion and

the

Workshop of

Reconstruction

Smelting

the

Practices

byPhilip P. Betancourt

The

main

period of the smelting workshop in EM III-MM IA, but the quantity

at

can be Chrysokamino of slag beneath the stratum

placed from this period shows that the workshop had a substantial earlier history on the spot before the EM III The number of Final Neolithic period. sherds found stratified within the deposit of slag below the EM III level indicates

that the installation was

has been

demonstrated

already in operation by this period. As in the previous chapter, the Final Neolithic of in western Asia) inherited a Crete (often equated with the Chalcolithic long earlier tradition of copper technology. From the lead isotope anal copper objects, we know that the ore deposits ysis of Neolithic Aegean on were and Kythnos, Seriphos, Siphnos already being exploited dur ing this period.1 Since these ores do not consist of native copper, this exploitation must have involved smelting; the situation at Chrysokamino cannot be unique. The

in the Aegean presence of specialized metallurgical technology an before the beginning of the Bronze Age is not isolated event, and it fits well with a rapidly growing body of evidence for specialized technology in other crafts as well.2 Pottery production was more becoming sophisti are and the earliest marble from the Neolithic cated, Cycladic sculptures as well.3

On the other hand, the exact date of the heavily burnished pottery in this part of Crete is far from secure. It definitely begins during the Neo lithic,4 but its latest date is at least contemporary with EM I, and it may even later. The persist length of time over which the workshop was used cannot be determined at our present level of knowledge. was not the at Chrysokamino workshop only cop eastern were in Crete. similar Other, facility workshops even in operation during the Early Bronze no Age, though

The metallurgy per

1. Stos-Gale

and Gale

2003.

2. Perl?s and Vitelli 1999. 3. Papathanassopoulos 4. Betancourt 1999. 5. Mosso

1910,

1981.

pp. 289-292.

smelting

undoubtedly additional examples have been excavated. Two other locations for probable are recorded. In at Chrysokamino, the metallurgy workshops describing Mosso mentioned another place with similar perforated ceramic fragments, somewhere

on or near the beach at Pacheia Ammos.5

not recorded. Another

location with

The

exact location was

a similar sherd was found by Haggis

CHAPTER

i8o

14

came from piece during his regional survey in the vicinity of Kavousi.6The no an excavation within the village of Kavousi, and other evidence for the was present. It is site at Chrysokamino the furnace that unlikely workshop would

have been

extraction without workshops

of metal the hints once

the only location where a process as important as the from its ores would have been practiced, so that even of these other locations, one would suppose that other

existed.

TOPOGRAPHY is an important component topographic situation of the smelting site a limestone outcrop at the west and a phyllite of its location. It consists of

The

outcrop at the east, with a lower space between them, situated at the top cliff slightly over 38 m above the sea. This topography of a north-facing it blows creates a fiinnel-like the wind when trough that concentrates situation has parallels from elsewhere in the Aegean. copper smelting locations on Kythnos and Seriphos Slagheaps are found on similarly situated north-facing cliffs high above the sea.7 In all of these cases, the topographic placement suggests a deliberate choice to maximize the exploitation of the north wind.

from the north. The

at ancient

The workshops location was obviously based on factors other than to the ore. The geological survey shows that no copper ore was proximity and it had to be imported. That present in the vicinity of Chrysokamino, ore some distance away from the mines in order to smelt it transporting at in antiquity is demonstrated by the similar situation ores several other places. At Feinan in Jordan, for example, Chalcolithic were over 100 km for smelting.8 Similar examples have been docu shipped

was not unknown

in Europe.9 Ore was especially easy to ship by sea made have because good ballast. Smelting sites might have been situated near other necessary resources like fuel and labor, in preference to

mented

from elsewhere itwould

locations

near

the

ore,

because

it was

easier

to

ship

the

ore

than

it was

to

the fuel.

transport One presumes that even if the smelters arrived as strangers in a new some needed commodities. Re territory, the local region could provide water for of and the included furnaces, making chimneys clay quirements fuel for smelting, additional laborers, and and food for the metalworkers, the for the final product. For Chrysokamino, however, whatever no evidence that the smelt situation was when the workshop began, there is

markets

ers were

IA activity. By then, foreigners by the period of the EM III-MM were within the local culture, as is the workshop integrated fully personnel sources access to and chaff (the two ingredients local clay shown by their of the chimneys), the presence of local styles of pot in the manufacture

site (with no foreign products), and the use of local tery at the workshop for all the ground stone tools. The abundance of East Cretan materials situation that whatever pottery and stone tools, in particular, demonstrates earlier, by the period of the greatest integral part of eastern Cretan society.

existed

activity, this workshop

was

an

6. 2005, p. 109, a chimney Haggis two holes found in soil with fragment of Kavousi. from a site in the village 7. Gale

et al. 1985,

Hadjianastasiou

pp. 90-91; and MacGillivray

1988; Stos-Gale 1998. 8. Hauptmann p. 425. 9. O'Brien

2000,

2001,

1996.

pp.

167-168;

OF

RECONSTRUCTION

SMELTING

l8l

PRACTICES

mj^J

?7 14.1. East Figure from Dark Ware

Cretan

7

V7

White-on

Chrysokamino

and

nearby sites: (a) Chrysokamino 75; (b)Myrtos Fournou Korifi, EM IIB; (c)Vasiliki, EM III-MM IA; (d-f )Gournia, EM III-MM IA; (g) Pseira, EM III-MM IA.

1972, no. P25; C: after A, B: afterWarren after Betancourt Pseira VI, 111. 13B; D-F: and Silverman 1991, nos. 313-315; G: after Pj?'nz VI, 111. 13A. Scale 1:3

ORGANIZATION

OF THE WORKSHOP

on Kythnos and Seriphos, the deposit of slag slag heaps at Chrysokamino in Crete is not near any large settlement. The nearest site is 600 meters away, and the nearest substantial town known habitation or at was located either at Kavousi, at Tholos Beach, at Pacheia Ammos, none of which were in times later all of which (but Gournia, important Fabric habitation have excavated Final Neolithic levels). The Mirabello IA pottery in the most important phase of the used for the EM III-MM Like

the Cycladic

the ware was site suggests a possible relationship with Gournia, where occurs at a in class this The in this the main fine pottery pottery period.10 whole series of sites near Chrysokamino (Fig. 14.1). intense period of IA date for the final and most The EM III-MM is Ware is secure. East Cretan White-on-Dark activity at the workshop at ware the at site. The the this fine from the only begins phase pottery end of EM IIB with simple motifs like pendent chevrons (Fig. 14. lib), and IA.11 Conical it develops fully during EM III-MM cups decorated with white hatched triangles (like the one illustrated in Fig. 14.1:a) are among the most common cups from the period. The end of the ceramic phase is IA in central Crete.12 contemporary with MM could not have existed in isolation. The The Chrysokamino workshop is very clear on this point: the site has no houses, no complete range of domestic pottery, and no household debris. The metalworkers slept and at The itself. the workshop lived somewhere nearby, not workshop must have been an appendage of some other site (presumably the habitation location situated half a kilometer away) because the workers needed food, evidence

water,

10. Hall

1904-1905;

1984. 11. Betancourt

1984.

12. Betancourt

2003b.

Betancourt

housing,

and

other

necessities

as well

as access

to

local

commodi

ties such as pottery, fuel for the furnaces, and clay and chaff to make the to have access to markets, and they depended on chimneys. They also had someone who could coordinate the acquisition of the ore. Since the lead and Lavrion in the slag matches Kythnos isotope pattern of the copper ore must have been shipped in by sea, requiring access to the (App. C), the aswell as personnel to unload, transport, and store it. port atAgriomandra,

CHAPTER

l82

14

At the regional level, the nature of the workshop would have neces sitated some type of directing authority. Importing ore would have required the coordination of a number of different activities, not all of which were same individuals: to find sources of metallic performed by the prospecting ore of the the ores, operation mines, preparing by removing waste rock ore an to the embarkation point, and ship transporting (beneficiating), sea over a distance of many kilometers. After the ore was received ping by and smelted at Chrysokamino, the copper was then transported to places where the copper was remelted in crucibles and cast inmolds to form use seems to have been ful articles. The metalworking industry in east Crete into specialized stages, as proposed for other regions by Ottaway.13 was most a small part of at Chrysokamino The small workshop likely only a overseas connections. large operation with both Cretan and divided

in the area, as shown by the chimney presence of other workshops from Pacheia and Ammos Kavousi, suggests a regional produc fragments tion system designed to provide the many villages and towns of the eastern The

Gulf

ofMirabello

from

with

local resources.

an important commodity not available exclusively In order to supply these workshops with ore, fuel, some agency must be envisioned. The coordinating

and other necessities, some or economic authority already existed in implication is that political the region by EM III-MM IA, and that itwas directing the Chrysokamino as well as other aspects of the industry. workshop it was palatial or authority, and whether a private, is complicated problem that cannot be resolved with the pres In western Asia, private companies could man available evidence. ently ventures like long-distance international trade, and their age comparable vast quantity of personnel often had kinship ties.14 On the other hand, the The

nature

of this higher

IA pottery from Gournia15 suggests that whoever was in a was of this which would have site, charge palace by LM I,16 already moot its economic issue base. The be because the modern may expanding in between private and palatial may have had little meaning dichotomy

EM

III-MM

society, where kinship may be the cornerstone of both in Early Minoan tensification of craftwork and the emerging political hierarchy.17 The lines between public and private may have been very blurred. At the local level, the metalworkers must have slept, cooked their meals, in the Chrysokamino and lived their daily lives somewhere territory. The EM III-MM record for all before IA indicates that sparse pottery periods the site was definitely not used continuously the during long period rep resented by its surviving artifacts; indeed, it is possible to argue that itwas in between. used only for very brief episodes, with periods of abandonment in years when could not have operated itwas active, the workshop were year-round. The evidence from the furnace chimneys suggests they

Even

an important ingredient probably all made in the fall because the chaff, in the clay fabric, would only have been available after the threshing that followed the grain harvest. This period also coincides with the windiest season, when

there is no rain but the north wind the conditions

every day, optimizing the draft for the furnaces. The didate

to help provide

at Chrysokamino

place of residence,

even

is the best can

though

it is 600 m

see Garelli

1963; Veenhof 1972; Kraus 1982; Larsen 1967,1976; Dercksen 1996; for short

see Foster

summaries,

15.Hall 1904-1905.

location of the later farmhouse for the metalworkers'

called the meltemi blows

for its exploitation

13. Ottaway 2001. 14. For Karum Kanesh,

16. Soles 17.Warren

1991. 1987,

p. 52.

1977,1987.

RECONSTRUCTION

OF

SMELTING

PRACTICES

183

away. No architecture has been found for the Early Bronze Age, but the IA pottery small settlement has Final Neolithic through EM III-MM from later contexts, and the pottery's style exactly matches that found at the workshop.18 In addition, the early settlement pottery has amore complete range of shapes than exists at the workshop, including cooking pots and other and serving vessels, storage shapes, and even a cooking dishes, cups few luxury pieces. Its pottery convincingly supports the argument that the was settlement the long period represented by the occupied throughout pottery at the workshop. On the other hand, the 600 meters of distance be tween the two locations raises questions. In particular, the soil for the three successive floors of the small apsidal structure was carried to the workshop from somewhere

away from the pile of slag. This soil, with many sherds and bellows pot fragments included in it,must surely have come from wherever the metallurgists lived. Unless the soil had a symbolic meaning (which may very well have been the case), itwould have been simpler to bring it from somewhere closer than a spot over a half kilometer away.Wherever the not the location has been excavated. Our information lived, metallurgists about the metallurgists, therefore, must come almost exclusively in the workshop evidence unearthed itself.

RECONSTRUCTION PROCESS

from the

OF THE SMELTING

In order to form the chemical

reactions that separate copper from other in secondary ores like malachite and azurite, the ore must be heated in a reducing atmosphere in the presence of fluxing agents.19 The flux, usually quartz and iron oxides, will eventually form much of the slag. a fuel like wood charcoal in a small, can create Burning mostly closed space elements

the reducing agent, carbon monoxide. The chemical reactions take place in and azurite, heating to ca. 500?C decomposes the stages. With malachite carbonate ores to create copper oxide. On higher heating, the carbon mon oxide unites with

the oxygen in the copper oxide and forms carbon dioxide, the copper behind. The reaction can be achieved at temperatures leaving under the melting point of copper (1083?C), and lower temperatures can be used if heat ismaintained for longer periods of time. comes from several items for the process at Chrysokamino in the archaeological record. The expedition uncovered a fragment of furnace lining, pieces of slag, chimney fragments, bits of ore, a fragment of a tuyere, stone tools, and pieces of pot bellows. This physical evidence substantial information for a reconstruction of the smelting provides Evidence

found

of the slag fragments and their included prills provide processes. Analyses information on the operation as well. in the smelting of copper have demonstrated Modern experiments several of the principles and techniques that must have been present in 18. Betancourt

and

Floyd

2000

2001. 19. For useful process,

Maddin

summaries

see Bachmann

1996, pp. 9-10.

20. Bamberger

1985.

1982,

of the p. 121;

the operation at Chrysokamino. Bamberger produced good results with a a furnace having diameter of up to 40 cm, preheated for 1.5 hours.20 He found that air would not penetrate more than approximately 20 cm into the charge, so he suggested ca. 40 cm (which is also the maximum diameter as the maximum of the Chrysokamino diameter for a simple chimneys)

CHAPTER

i84

14

Furnace Chimney

Bellows

14.2. Figure bowl furnace

Tapped Slag Tuyere

based

ratio of ore-to-fuel

from

fragments

Chrysokamino in the found

excavation

Charge furnace. The

on

of a

Reconstruction

ore was used along with results. good

was 2:1. Artificial

as fuel, with three air intakes yielding a few remain, most of the steps in smelting can be questions reconstructed. The evidence suggests that the technical process in Crete

charcoal

While

was not very different from other Chalcolithic in the eastern Mediterranean.21 ing operations

to Early Bronze Age smelt Before the smelting could

a coarse chimney from clay, build begin, the workers needed to construct a bowl furnace, and raw materials and fuel. the gather together rossa at a were made from local Chimneys clay. A deposit of terra about a kilometer away from the nearby sinkhole called Lakkos Ambeliou, is the closest clay deposit whose mineralogy matches the fabric workshop, site than to the of the chimneys. The deposit is closer to the habitation itself. Its sediment includes both abundant kaolinite and small workshop stone fragments, so it (or a similar deposit) would have worked well as a raw material for ceramics without the addition of extra materials, aside from the organic material that would burn away and leave the fabric more this organic material was chaff and straw from porous. At Chrysokamino, the threshing of barley stems of plant material as crucibles

used

clay fabric made

cut pieces of the (Chap. 12). The addition of short, was not unusual for molds and other items such

in Cretan metallurgy, because the creation of a porous ceramics the less likely to break under the stress of high

were built by hand to form cylinders that temperatures.22 The chimneys were open at both top and bottom, with slightly smaller diameters at the were in the end. Holes upper cylinders by thrusting fingers through pierced the walls while the clay was still wet and soft. The chimneys were probably not fired before use. a con smelting process would have taken place in bowl furnace in a small hollow in the ground (Fig. 14.2). The furnace would have been small, 20-40 cm in diameter, and lined with clay. Ore, already The

structed

into small pieces, would have been placed in the furnace along with the fuel and flux to form the charge. The absence of waste rock or of stone tools for preparing the ore indicates itwas brought to the site ready for

broken

use.

over the furnaces would have aided in the draft, and, by Chimneys EM III-MM IA if not earlier, the natural draft would have been augmented 21.

use of pot bellows. by the Three

copper

minerals,

azurite,

malachite,

and

chrysocolla,

occur

in

are

the archaeological material excavated from the site (Chap. 10). They all present in very tiny fragments, and no direct evidence survives for their source. Several writers have considered the possibility of Cretan copper

See,

among

others,

Bachmann

1980; Rothenberg 1972; 1985; Haupt mann

1989; Hauptmann,

and Maddin

Bachmann,

1994; Hauptmann

2000.

22. Blitzer 1995, pp. 504-508; Evely 2000,

pp. 353-356.

OF

RECONSTRUCTION

SMELTING

PRACTICES

185

island does have small copper deposits,23 and some writers needs.24 Others that they could have supplied Early Minoan

ores. The suggested discounted

have have

are some of these sources,25 but the use of small deposits that a or that were as yet undiscovered remains depleted long ago possibility. all the ores may have been imported from sources outside Alternatively, Isotope analyses of the lead in copper prills trapped within the a source outside of the island. They indicate slag support the argument for the isotope that the minute quantities of lead in the prills best matches

of Crete.26

pattern of ores found at both Lavrion and Kythnos.27 sea. In calm weather, The ores probably arrived at Chrysokamino by the ships could have landed at the foot of the cliff at the metallurgy site, but inwindy weather, landing at the small beach of Agriomandra would have stone containing malachite safer. A piece of waterworn (discussed in was found on the beach at the small harbor of This Agriomandra. App. C) a waterworn is beach with sheltered covered located cove, stones, small, by at the foot of a ravine that begins just downhill from the Chrysokamino been

habitation.28 It is farmhouse, the most likely location for the metallurgists' is one of the Malachite the closest harbor to the Chrysokamino workshop. ores one at cannot the site. found determine copper metallurgy Although when

the malachite

condition

indicates

was on the beach, its waterworn specimen deposited it is not a recent arrival. It also indicates that copper

ore was

to at some time in the past. brought Agriomandra are all ores found at the ores secondary copper workshop easily is is their colors and azurite (malachite blue, green, bright recognized by The

is pale blue-green). Even novice metalworkers would have no ores or the difference between higher-grade recognizing problem sorting and lower-grade shipments. These minerals are easier to smelt than cop chrysocolla

per sulfides, and they would have yielded suitable amounts of copper, but not account for all of the present in the slags from they do mineralogy iron contents of the slags In the and silica Chrysokamino. particular, high are not

present

in these

ores.

One

must

assume

that

an

iron-rich

flux

was

present in the charge along with the copper ores and fuel. The variable arsenic content also requires an explanation (Apps. B a ores not is and D). It constituent of any of the three copper found at the site. Arsenic

is of particular interest because of its presence in arsenical cop artifacts from Crete and elsewhere in the Aegean,29 aswell as in objects per from western Asia.30 The presence of arsenic inAegean copper artifacts has been discussed many times.31 The element can occur in nature in brightly colored green and blue-green ores such as clinoclase and pharmacosider ite. Such ores may have been recognized as yielding copper with different 23. Faure 1966; Branigan 1968, pp. 50-52; 1974, pp. 57-63; McGee han-Liritzis

1996,

p. 117.

25.Wheeler, Maddin, andMuhly 1975; Becker 1976; Gale 1990,

probability 18. chap. 28. For

1982;

of a match,

see Baxter

Agriomandra,

see Faure

2000, p. 459. 29. Gale

and Stos-Gale

and

See App. C; for comments to achieve statistical methods

27. the

p. 387.

24. Faure 1966; Branigan 1974,

pp. 304-305. 26. Gale

Gale 1990. on the 2003,

Ioannou

p. 4.

31. Charles 1967,1980; Coghlan 1972; Gale 1990; Zwicker 1991;Tyle cote 1991; Katsa 1997;Mangou and Ioannou

and Stos-Gale

1989,

tables 1-3; Gale 1990;Mangou

1998.

30. Ravich and Ryndina 1995,

Chap.

1998; 13.

see also discussion

in

i86

CHAPTER

14

their addition, or they may from copper smelted without characteristics for the more common copper ores. have been accidentally mistaken The high arsenic content in some of the prills (App. B) can only mean that arsenic minerals were present in some of the charges placed in the microscopic level, the arsenic is irregularly present, in one tiny mass that is as high as 26% arsenic (App. B, no. 48), in most prills, the arsenic content is much lower. The element

the furnaces.32 At resulting

although is not uniformly

and many prills contain only a trace.When distributed, traces in ancient metalwork arsenic has been regarded as the occur, only were not fully aware an accidental inclusion.33 Even if the metallurgists to the conclusion must be that in addition of their process, however, the sometimes contained the malachite, and azurite, charge chrysocolla, ores with small amounts small quantities of more complex polymetallic even of several metals, gold including arsenic, nickel, lead, and (rarely) no. 1). (App. B, Fluxing agents were also present. The analysis of the slags from Chryso shows that the original charge used for the smelting operation was amounts of highly siliceous, with the additional presence of substantial and the iron The silicates A and iron oxides and/or hydroxides B). (Apps.

kamino

either present in the ore or in a separate fluxing agent that needed to be added to the charge. To a certain extent, the point ismoot, because both fluxing agents were clearly present in sufficient quantities to

minerals

were

produce The

the desired product.

ore and fuel and smelting operation could begin after the charge of flux was placed in the small bowl furnace. Fuel will probably have included some charcoal, but additions of high-efficiency fuel like the residue from not out. cannot do know what fuel was used extraction be ruled We olive oil some at the site. In contrast with of the other early smelting operations, no trace of unconsumed excavation. fuel comes from the Chrysokamino Charcoal, however, is not always found along with Bronze Age slag if the is usu waste pile is located away from the furnace itself.34 This material ally assumed to be the fuel of choice for early smelting operations, but the at Chrysokamino, along with the absence of high temperatures achieved on the excavated part of the site, suggest anywhere that seeds and skins from olives may have been used as an alternative or an addition to the charge. In the centuries before the use of large presses, olive oil extraction would have involved the use of mortars and hand pestles, either ash or charcoal

a residue that was a crushed mass of seeds and skins.35 This oily producing and itwould have burned mixture would have been highly combustible, an intense heat, accounting for the high temperatures, the absence of fuel, and the lack of unsmelted ore in the slags. It was used as a Its efficiency as a fuel is attested for fuel pottery kilns at nearby Mochlos.36

with

unburned

at modern

limekilns

and pottery

kilns

in Crete, where

it is regularly used

the fire was

started, the cylindrical chimney would have been to aid in the draft, which was created by the natural placed above the furnace were wind as well as by the use of bellows. The bellows at Chrysokamino to air for conduct the the reed with nozzles drum-shaped, cylindrical pots a to cut for leather in surfaces allow their upper toward the fire and holes After

32.

Thompson

33.

as a fuel.37

Hauptmann, Schmitt-Strecker

1958,

p. 4.

Begemann, 1999, p. 9.

and

34.Herdits2003,p.71. 35. Soles 2003, p. 24.

36. Sarpaki and Bending 2004, p. 126. 37. Betancourt

2004a.

RECONSTRUCTION

OF

SMELTING

PRACTICES

187

inflatable pumping mechanism. They were set into the ground and fixed there with mud so they would not move around during use. Several will have been used for each furnace, with a reed that terminated in a tuyere were not very efficient conducting the air from bellows to fire. The bellows at this site, and in addition to the fresh air, they sucked in enough heated air from the furnaces to darken their interiors and to bake the mud that fixed them to the ground sufficiently to turn the mud into pottery. The pumping mechanisms (presumably made of leather and wood) could not have lasted very long. The technology that produces either a copper ingot at the bottom of a furnace or a series of prills inside the slag iswell known, and modern have experiments replicated the process.38 In small furnaces, ancient met a temperature of ca. 1200?C and maintain it for allurgists had to achieve so that as an droplets would flow downward and collect in the bottom of the furnace.39 At lower temperatures, however, even ingot as much as 200?C below the heated melting point of copper, malachite the slag. These under reducing conditions will form small prills within not flow to the bottom of the furnace, so prills stay within the slag and do that their removal requires that the slag be broken into small pieces. about four hours

a the smelting operation at Chrysokamino produce high enough cause a mass to to at of collect the of the fur bottom copper temperature nace? Phase diagrams that show the chemical compositions of slags40 can Did

demonstrate

the liquidus temperature (a reasonable estimate of the lower achieved in the furnace). An alternative method derives tem

temperature based on the fractional perature estimates from the chemical composition cation composition.41 By using both of these methods, Stos-Gale and Gale a near estimate and 1200?C for the slags at Chysokamino, temperature of they note

fall into two groups, with some 100? higher than others approximately

that the slags they analyzed

examples reaching temperatures a diameter of 1 cm are sufficient for prills with (App. C). These temperatures to flow downward and collect at the bottom of the furnace, with tiny prills the slag. The presence of slag fragments with surfaces remaining within that the material flowed out of the furnace and cooled quickly showing a (Fig. 10.1) provides additional proof that the process resulted in pool of copper at the base of the furnace. This would have necessitated tapping the furnace to remove the copper, and the tapping process would also have caused the slag to flow out and form rivulets of molten siliceous material on (note the flow formations Fig. 10.1:a and c). No rivulets would have formed if the smelting resulted only in amass of slag inside the furnace. The

small size of the chunks of slag from the workshop is typical of was where the broken up to remove early copper smelting operations slag the copper.42 As at Chrysokamino, these roughly contemporary operations some or all of the copper as isolated the slag, and prills within was necessary to retrieve the metal. Even if (as seems to have been crushing some of the metal collected at the bottom of the case at Chrysokamino)

produced 38. Tylecote andMerkel 1985. 39.Maddin 1996, p. 14. 40. Bachmann

1980,

pp.

120-131.

1982, pp. 21-22; Rothenberg 1985,

the furnace, the remainder of the copper was obviously desirable enough to be retrieved even if it consisted the only of tiny prills trapped within The accumulation of the small of broken created slag. pieces resulting slag

p. 124.

the slag heap at the site (Fig. 10.2).

41. Nathan and Van Kirk 1978. 42. Bachmann 1980, pp. 108-110;

i88

CHAPTER

14

indicate that at least some of the prills slags at Chrysokamino an to flowed down form ingot. Many pieces of slag from Chrysokamino are a result of than the other known consistently higher temperatures The

ore or intermediate stages early smelting sites. No partly smelted pieces of between ore and slag are present. The site does not have any of the incom pletely smelted material (sometimes called furnace conglomerate) present in Chalcolithic smelting operations at Feinan, Jordan43 and elsewhere.44 Also, some of the pieces of demonstrate the presence slag from Chrysokamino of a liquid stage during the smelting. Small drips, stalactite-like forma are the tions, and other traces of amelted among stage present fragments is also occasionally of slag (Fig. 10.1:a). Flow-banding recognizable at the level (App. A, CHR 99). microscopic The high temperatures, however, were not always achieved success in the slags at the microscopic level and the varying fully. The differences amounts of arsenic in the copper prills indicate that surviving slag fragments were

formed under different

conditions.

Some of the slag reflects differ ences in the amount of reduction as is shown during smelting (App. A), amounts of of formed the relative characteristic under by fayalite (a slags and which indicate pyroxene group minerals, highly reducing conditions) amore the arsenic, experiments have oxidizing atmosphere.45 Concerning of this element within shown that incorporation copper prills in the slag can result if arsenic minerals are present and if the temperature is not raised a high enough to result in molten metal that would be removed by tapping the furnace.46 The evidence indicates that conditions varied substantially between than

other

different

firings,

and that temperatures

were

sometimes

higher

times.

The high temperatures that would have been necessary to produce these are important because of Chrysokamino's early date. The nature of slags the slags from this site contradicts an opinion that was once generally held, and that that slags may be dated on the basis of their physical description, are a high temperature) "glassy" slags (i.e., highly vitrified pieces indicating on or In for Roman later.47 this basis based from fact, always slags dating, as a was once general pattern, widespread, Cyprus and regarded incorrectly itself was once assigned a date in theMedieval and Chrysokamino period or later on this basis.48 After the smelting operation was completed, the slag was broken into it.The impor small pieces to retrieve even the small bits of copper within tance of the availability of water for this process has been emphasized by because it adjoins Herdits.49 The sea is easily accessible at Chrysokamino the site, and wetting the hot slag with sea water would have shattered it at the relative scarcity of stone hammers and easily quickly, explaining was of the waste pile, which this site (Chap. 6). The composition mostly to attests of of with the final corners, tiny pieces slag angular composed step in the process of smelting (Fig. 10.1:a). In conclusion, a combination of evidence with

from scientific analysis along from the excavation has resulted in a useful re

the physical evidence construction of the main steps in the smelting process. This reconstruction were successful in at Chrysokamino indicates that the methods achieving

43. Hauptmann,

Maddin

44. Bachmann

1980,

45. Hauptmann,

Maddin

and

Bachmann,

1994, p. 5. p. 108.

Bachmann,

and

1994, p. 6.

46. Pollard 47. The

et al. 1991,

theory

is well

p. 132. summarized

by Koucky and Steinberg 1982a, p. 156, 1; see also 1982b. 48. Faure 1966; Branigan

table

49. Herdits

2003,

pp. 69-70.

1968.

RECONSTRUCTION

OF

SMELTING

PRACTICES

189

could have been disseminated good results. Copper from Chrysokamino to several sites in Crete, to be made into chisels, daggers, awls, and other a was the That metal is products. resulting highly valued commodity extreme at the was not the workshop. Ore economy practiced proved by (the largest piece found at the excavation was a centimeter across). small prills were desirable, and the slag was shattered into small pieces to recover the tiny bits of metal. Copper was clearly a highly prized

wasted Even

material.

PART III: THE SURFACE SURVEY

CHAPTER

15

to

Introduction

the

Surface

Survey byPhilip P. Betancourt

a small Minoan A metallurgy workshop, habitation site, and a burial cave an area of less than a were present in the region within Chrysokamino to In kilometer. order better these understand sites, an intensive square surface survey was planned as one component of the investigation. The was a to main goal better diachronic understanding of the local develop use. to its with and land In particu settlement, respect economy, territory was to context it essential elucidate the of the lar, absolutely metallurgy in order to interpret its significance. The local context was by no workshop means understood before the project began. Was copper ore present in the as a consequence of local vicinity? Did the metallurgy develop mining and the exploitation of a small ore deposit, as had been claimed by some of the early investigators,1 or did it exist in isolation from the natural resources for its development? How did the metallurgy site fit into the local was the relation between the What site and the small economy? metallurgy cave claimed at one time as a copper mine?2 Did the live at metallurgists

necessary

the habitation

site or somewhere

else? Did the metallurgy workshop and definable territory, and could one learn anything specific the history of that territory during the Bronze Age and later? a

within

exist about

to questions about the metallurgy operation, many general to needed be The date of the earliest settlement of the addressed. questions local area had not yet been established. When did the human occupation of the region begin, and how did it develop? Were Final Neolithic sherds found elsewhere as well as at the metallurgy workshop? was an The nature of the region s development through time important or did issue. Was the local development and continuous, interruptions was the relation between the habitation recessions exist? What site, the as a and the location? Did the site habitation function cave, metallurgy an or center for estate at any time agricultural farm during its history, and In addition

if it did, could the boundaries of the local territory be determined? Did the nature and extent of this territory change through time?What were the characteristics of small individual (location, description, and distribution) sites and features 1. Hawes

and Boyd

Hawes

1909,

p. 38.

2.Mosso pp. 289-293.

1908, pp. 518-521; 1910,

were

in the territory of the main interrelated?

domestic

complex,

and how

all these locations The

exploitation one of the important

of the land and its natural resources must

have been

factors in local history. To answer the many

questions

CHAPTER

194

15

site, the project needed to establish the potential of posed by the metallurgy the region, including its geological materials (soils, mineral wealth, stones as as for building and other purposes) its favorable environmental well such as access

to natural harbors, landing places for ships, and access to water, the pres routes. to It also needed establish transportation ence of different types of with varying potentials for exploitation, landscape

factors

and resources

such as forests and wild

the local environment area needed were

game. These and other attributes of to the settlement of the that may have contributed before we

to be defined

could determine

how

such resources

used.

on Bronze was a Age land management special focus of the was to define of the landscape characteristics important goal key as a basis for it into units different different classes of (i.e., dividing spatial as land), using asmany factors possible: agricultural potential as determined Information

survey. An

by soil type and presence of mineral nutrients, water retention capability of the soil, use of the land in antiquity as determined by the presence of or to sherds and other artifacts, proximity distance from settlements, and setting (e.g., steepness of the land and other factors). The topographic was to use these characteristics to establish different classes of land plan as part of the evidence for how the and landscape may have been managed used. Agriculture must have been important. Did the exploitation of the the land use of the Bronze Age the farmland change through time?Was foundation for the later agricultural exploitation of the territory, or could in later periods? be recognized even some of these the con questions would complement Answering clusions to be drawn from the data collected by excavation. Placing the

differences

operation within of the economic

metallurgy

something would hopefully contribute aswell as the general history

a historical

context

and trying to understand it operated and social systems within which nature into the of the local history in insights of the small metallurgical

installation

and its

role in the local region.

PROCEDURES OF THE INVESTIGATION A research plan was developed answers.

ponents, the same

It was

organized

to address the many

as an

some of which were

intensive

extensions

surface

questions survey

with

that required several

com

of the excavation

project and used such as the instrument

required for the excavation, personnel survey team. Other components required the expertise worked only on the intensive surface survey.

of specialists who

were an of the Chrysokamino territory important in determining the extent of the survey. They were defined component of several factors including topography, natu based on a combination The

boundaries

ral barriers elsewhere, were being

in the landscape, information from

analogies with the excavation

territorial

boundaries

that showed what

from

resources

of nearby sites. Because and the proximity this exploited, not be in detail before the assembled could project methodology employed its data, preliminary estimates of the territory were made at the inception of the work,

and the plan was modified

as work

progressed.

TO

INTRODUCTION

THE

SURFACE

SURVEY

*95

An especially close examination of the landscape was required because of the many questions raised by the metallurgy operation. One of the most important of these questions involved the possible use of local copper ores, an issue raised several times in the early 20th century.3 The presence or absence of former ore beneficiation sites, mining operations, or locations for stockpiling of ore could only be determined by examining the entire for example, involves the enrichment of bulk ore landscape. Beneficiation, of the removal rock that does not contain any ore, and the practice results by in large quantities of waste rock.4 Even if a small ore deposit is completely exhausted, the piles of waste rock will survive as dumps. Since the excava tion of the metallurgy site demonstrated that no beneficiation occurred at a waste the workshop search for in rock the of itself, piles Chrysokamino the presence or absence of former mining territory could help demonstrate or on the absence of ore stockpiling sites. Any potential conclusions based deposits (i.e., using negative evidence) could only be valid after the entire was examined landscape closely enough to be certain that such waste rock was

not

present.

surface survey of the Chrysokamino specialized territory was It differed from the intensive traditional planned. substantially walking survey.5 A traditional survey had already been conducted in this territory,6 and most of the pottery fragments on the surface had already been collected. In 1983, a team of metallurgists headed by Noel Gale and Zofia Stos-Gale had also examined the territory as a part of their metallurgical survey of A

sites in the Aegean smelting and mining (App. F). The soils of the region had also already been studied.7 The new investigation was planned as an extension of the previous work. new intensive

The

carried out between

surface survey included 1994 and 2003:

the following

components,

1. Geological

the geology of the Survey. In order to understand local territory, William Farrand and Carola Steams conducted a survey of the and the bedrock geology. They geomorphology

across the entire territory and recorded the relevant in 1994 and 1995 (see information Chap. 2). of the Landscape. A team of three to five persons used 2. Mapping a Topcon Total Station interfaced with a laptop computer to map the Chrysokamino territory, walking over the landscape at and recording points suitable intervals. The team recon walked

firmed

and accurately mapped sites recorded by the earlier collected additional sherds, and recorded all features. survey, The team set up the survey instrument and recorded points it at intervals of 5-10 m, noting whatever anthropogenic the area being measured. The sur present within vey station was then moved to record new territory. This aspect of the project took four field seasons (1995-1998). The team around

features were

3. Hawes

et al. 1908,

p. 33; Hawes

and Boyd Hawes 1909, p. 38;Mosso 1910, pp. 289-292. 4. D. Gale 1991;Gale and Ottaway 1991.

5. Among

many

others,

see the

Davis,

methodology of Keller and Rupp 1983; Cherry, Gamble, and Shennan 1978; Cherry

et al. 1988; Wells,

Zanger

1990,

pp. 214-216;

Runnels, Cherry,

and

and Mantzourani

1991,

espe

Mantzourani,

and

Whitelaw 1991;Wells 1996. 6.Haggis 1992,1995,1996b,

2005.

cially Cherry,

7. Morris

Davis,

2002.

CHAPTER

iq6

15

its time between working on the survey and providing and mapping for the excavations. The electronic

divided

measurements

maps and charts mapping project produced computer-generated as well as detailed information on the nature of the landscape (see Chap. 16 and App. G). 3. Soils Survey. A survey of the soils in the territory was made by (see Apps. L and N). 4. Analysis Program. Several analytical studies were undertaken to contribute information on the territory (see Apps. A-F, M, and N). They examined artifacts and materials by a series of different techniques, contributing substantial information for Eleni Nodarou

the project. 5. Register of Anthropogenic Features. All architectural features terrace recorded walls) were mapped, (including groups of with the survey instrument on a three-dimensional grid in the and described (App. G). The instrument computer, measured, team the features as a part of their mapping recorded survey architecture were also mapped, project. Sherd scatters without and sherds were

(see Apps. H and J). Given the nature as the occurrence of a previous survey, the new collection was regarded as a sample possibly not rep the new information was resentative of all periods.8 Therefore, collected

as well

of the evidence

coordinated

the data obtained

with

to Anthropogenic

6. Revisits

ters were

revisited

staff for additional 7. Selective

Excavation.

metallurgy genic

of the several times by different members and sherd collection. study, photography, In addition

a small

to obtain

excavated

8. Study, Analysis,

to the excavations

cave

and

additional

an

terrace,

agricultural

information Revisits. After

and Additional

at the

site, two anthropo

and the habitation

workshop

features,

from the previous survey.9 features and sherd scat

Features. All

were

(App. K). study and

analysis, specific locations, regions, and features were again visited and additional data (including more electronic survey were obtained as needed. This extra step allowed points) maps to be checked against the actual computer-generated and preliminary written conclusions the physical situation on the ground, information could be collected as needed.

to be com

topography pared with

so that addi

tional

Information

from the excavation workshop from

the

was

site and the metallurgy interpret the data

of the habitation

and used to help

coordinated

survey.

In order to determine

land use, the land was divided into types of soils and their agricul

classes based on topography, tural potential, archaeological uity (including location within in new

resulted

ment

evidence for land use in antiq terrace sherds, walls, and other characteristics), the territory, and other factors. These studies conclusions

on

the

organization

and manage

of farmland

9. Conclusions published

during antiquity. and Publication. Conclusions

both

in preliminary

8. Read

were

completed reports and in this volume.

and

9.Haggis 2005.

1986.

1992,1995,1996b,

2002,

of

Topography

the Region

Chrysokamino

byLada Onyshkevych and William B. Hajford

is located in a region of terraced site at Chrysokamino The metallurgy the Gulf hillsides and steep cliffs west of modern Kavousi, overlooking of Mirabello. this area, the ground surface is covered with Throughout a dense, scrubby vegetation (maquis), occasional large thorny bushes, and few lone olive trees. The bare landscape near the coast is divided from the to the territory from the village of inland fields by amodern fence. Access is provided by a modern road, with a gate (to prevent sheep and crosses the road. The at from the goats point where the fence trespassing) road terminates at theMinoan habitation location.

Kavousi

area

surveyed for the maps in this volume includes the Chryso kamino metallurgical workshop and a habitation location with the stone remains of a Late Minoan farmhouse (Fig. 16.1). Anthropogenic features are discussed in Appendix G and shown on Figure 16.2. They include The

terrace walls, field houses, threshing floors, boundary walls, small wells, features. None of these are in use anymore, and their and other manmade distribution contributes substantial evidence for the past history of land use in the region. The survey area to the north extends to the coastline beyond theTherio spelio cave, while to the south, it continues to the field called Lakkos Ambe liou and to the lower slopes of Chalepa Hill. The hills north of the territory, where the small unexcavated hilltop tower called Pyrgos Chrysokaminou is located, are beyond steep cliffs that form a natural boundary north of The coastline, with additional steep cliffs, forms the western Theriospelio. On the east, the surveyed area includes terraced olive groves east boundary. site (Fig. 16.3). In terms of the grid created by the survey team in 1995, the area surveyed extends from approximately 4000N to 5250N, and from 4500E to 5600E, with some points measured as well see also (Fig. 16.2; Fig. 16.4 for the locations of beyond these lines of the Minoan

habitation

the datum and other grid points set in cement). as as of the Maps of the excavated areas, well larger region linked to them, were sea instrument electronic survey produced by during the 1995-1998 sons. were cover measured electronically, and 7,000 points Approximately age averaged every 2-3 m surrounding In

areas

farther

away

from

the

excavation

the excavated locations,

locations coverage

(Fig. 16.5). was

sparser,

CHAPTER

198

N

16

Pyrgos Chrysokaminou

Metallurgy Location

Theriospelio

Habitation Location Katsoprinos of the greater 16.1. Map around with Chrysokamino, indications of the habitation and

Figure

region

locations, Theriospelio beach and the cave, Agriomandra to it, the road within ravine leading the ravine, the road and the fence of

metallurgy Kephalolimnos Agriomandra

the ravine road '

Chrysokamino sites

neighboring

CHALEPA

Kephalolimnos,

LAKKOS AMBELIOU

Chrysokaminou

at an average of every 5-6 m, because of the considerable extent of land areas (such as needing to be surveyed. Very distant Chalepa, the hill south of Agriomandra received beach) coverage averaging every 15-20 m. In areas as the ravine of and treacherous addition, very steep footing (such on the northeast

or the lower ravine area of Chomatas Hill side ofTherio, the two excavated locations) could in some cases receive only between coverage: a line of points taken along the highest ridgeline representative and the lowest line of the ravine. The software to generate topographic areas 16.6). (Fig. This coverage was sufficient

were resulting data

satisfactory

topographic

sufficient for the lines for those

for detailed mapping. The work also surface survey of the territory, with the recording of all anthropogenic features noted by the team (see App. G). The intent was a series of accurate maps to showing the topography and to locate produce included

an intensive

the abandoned

architectural

remains and other manmade

features within

the territory, so that both natural landforms and anthropogenic could be studied in more detail as a part of the overall project.

features

region, of and

and

Katsoprinos, Pyrgos

the

TOPOGRAPHY

OF

THE

REGION

CHRYSOKAMINO

199

5200 H

5100 H

5000 H

4900

4800 H

4700

-\

4600

-J

4500 H

4800

4900

Figure 16.2.Map features

of anthropogenic

in the Chrysokamino

stead (seeApp. G)

5000

farm

5100

5200

5300

5400

5500

5600

5700

the survey personnel gathered almost all of the data for the Although in this volume maps electronically, many of the coastal cliffs at Chrysoka mino could not be surveyed directly because of their dangerous slopes and treacherous footing. Therefore, the maps were produced with the aid of a coastline digitized from an old Greek military map and merged with our own

files. The military electronically surveyed data in Surfer and AutoCAD to topographic features and map, however, could not be precisely matched landmarks currently visible, due to a number of factors. First, the sea level, and consequently over the decades

the apparent line of the coast, had undergone changes since the military map was made. Second, the original purpose of the military map did not require the same level of precision as that yielded by modern surveyed data, and itmade use of older technology. these two bodies of data, itwas necessary to make some Thus, inmerging in the data recorded on the military adjustments rotation of the coastline axis west of north.

map,

including

a

slight

CHAPTER

l6

N A

Metallurgy Location

Gulf of Mirabello

Figure 16.3. Topographic showing

A few areas of the coastline and coastal cliffs, however, were accessible for surveying, and they helped determine the correct alignment of the mili area can be tary map's data and the electronically surveyed data. One such found on the coastline immediately southwest of the metallurgical location, location. curving south and southeast into the bay below the habitation A number of coastal cliff areas that descend almost directly into the sea could also be surveyed electronically, and they yielded data consistent with these areas were Chylopittes, the rotated coastline. Among location's headland, northeast of the metallurgical

Therio

the headland

of

the cliffs south

beach, location, and the cliffs flanking Agriomandra the small beach or harbor including parts of Chalepa Hill. Agriomandra, at the bottom of the ravine southwest of the metallurgical and habitation

west

of the habitation

also yielded data consistent with the adjusted alignment of the our own data. Thus, the coastline used in military map's coastline and extent not the maps in this volume, while directly surveyed in most of its and not as precise as other surveyed data in this study, can be regarded as locations,

reasonably

accurate.

is situated downhill from a location of Chrysokamino small saddle south of the dome of Chomatas Hill, at an average elevation the north and of approximately 120 masl, in an area called Katsoprinos.To once terrace modern A walls. it is bordered east, sheepfold (mandra) by The

habitation

terraced

areas

map (shaded)

OF

TOPOGRAPHY

THE

C H RYS O KAM

IN O

REGION

Figure 16.4. Map of grid points for the Chrysokamino farmstead (Main Datum,

East

Point,

South

Point,

Villa North, Villa West) bordered

it to the south and partly extended over the ancient remains prior architectural complex itself slopes to the southwest at

to excavation. The

an average a distance of 1 m features a gradient of 0.14 (i.e., change in to continues elevation of 0.14 m).This down the southwest, flattening slope out into a fairly level field (about 106 rh in elevation), which contains terra rossa sediment and a construc large ellipse of rubble wall. This elliptical see an old now tumbled down (Fig. 16.2:32; likely sheepfold, a to rises AF The terrain small for and then 32, date). description App. G, hill of about 110 m in elevation, descends to a lower hill, and continues tion ismost

a down again until it reaches Lakkos Ambeliou, large and fairly flat basin (about 75 m in elevation) filled with more of the terra rossa sediment a steep ravine (revma), with sides of Lakkos Ambeliou, (Fig. 16.6). West at maximum down of 3.0, stretches to dropping gradients approximately the sea, ending at Agriomandra beach. Chalepa Hill rises south of the at an average gradient of 0.4, while another ravine and Lakkos Ambeliou hill rises north of the ravine in two peaks of about 100 m in elevation (with a

of Agriomandra gradient of about 0.3 from the ravine edge). Northwest a from beach is that is completely inaccessible small, craggy peninsula land because of a deep crevasse separating it from the cliffs of the north side of the ravine; topographic information for this small peninsula was estimated from the military map and from visual data, and so cannot be regarded West

as

highly accurate. of the habitation

location is a rocky outcrop, which descends a steep ravine, which in turn and along knobby ridge rocky cliffs into to terrain down the coastline. of this and the South ravine, drops ridge a an to descends along average gradient of 0.5) very gentler incline (with m a at into the of North which 44-50 3.5. sea, cliffs, steep drop gradient a

CHAPTER

Indicates points measured

electronically

Indicates points estimated

for topographical continuity

Coastline

in inaccessible

l6

areas

based on Greek military map

location, the terraced hillside steps down along a shallow ravine. Then, at an elevation of approximately 80 m, it drops into a a rather steep and very gradient of 0.53) slippery cliff and ravine (with

and west

to

the

sea

of the habitation

coast.

north, the modern dirt road (partly created in the spring of location along the terraced hillside from the habitation extends 1996) onto the southwestern side of the same hill (Chomatas). One branch Farther

side where the gate to a area is located, and lower branch descends along the the Chrysokamino side western side of Chomatas Hill and stops abruptly. The southwestern of the hill descends along a relatively bare and unterraced stepped ridge, down to a very rocky headland above the sea where the Chrysokamino curves around

the dome

of the hill

to the north

Figure 16.5. Map kamino of all

of the Chryso

farmstead,

with

measured

points

indications electronically,

all points taken from the Greek map, military to generate contours

and

all

satisfactory

(see

legend)

points

inserted

topographic

TOPOGRAPHY

OF

THE

REGION

CHRYSOKAMINO

203

CHYLOPITTES N Theriospelio A

Metallurgy Workshop

Gulf of Mirabello

16.6. map Figure Topographie the of the Chrysokamino territory farmstead

of

LAKKOSAMBELIOU

is located. The metallurgical location is on a low metallurgy workshop saddle on this headland,1 at an upper elevation of approximately 38 m, drop a ping sharply to the west at gradient of 6.5, to the north at a gradient of to at a gentler gradient of 0.65. The seacoast and the southwest 1.1, sloping can be ease approached with relative along the southwestern descent, while the west,

northwest,

and

northeast

sides

are

too

steep

to

allow

access

to

the sea. Located

area about 222 m below the lower road, the metallurgical itself requires a walk for ascent or descent (gradient of 0.28). The direct distance between the metallurgical location and the habitation location is ca. 585 m. Around

the curve of the cliff to the north and northeast

of the metal

lurgical location (average gradient 5.5), another headland (Chylopittes, part to a rocky of the hillside named Therio) protrudes into the bay, descending coastline at an average gradient of 0.56 before dropping in stepped cliffs to the water on the west and southwest. The northern corner of this headland, however, steps down directly to sea level, although the wide gaps between its ridges make

actual access to the sea difficult or

1. Haggis 1996b,

1992,

pp. 380-381,

p. 170, locus esp. n. 22.

88;

here. A cave, called The is located 44 masl on this headland

riospelio, Agriospelio, Kolonospelio, this cave has a small entrance, but a large interior (Figs. 16.1,16.2:34,16.6); a cave is located at a with (not mapped) sloping and slippery floor. The ca. m direct distance of 230 from the metallurgical location and ca. 610 m

CHAPTER

204

l6

from the habitation

location. Farther north, the coastline contains another small inlet, from which a ravine ascends to the area of the modern gate (with a gradient of 0.3 along the base, and 0.6 along the sides); a few ter

race walls

remain

along

this

steep

ravine.

larger Chrysokamino region also contains the remains of a number of features that probably marked the natural boundaries or extent of the ter ritory utilized by the inhabitants in the Bronze Age. The terraced hillsides immediately to the northwest of the habitation location (Fig. 16.3) find paral The

lels on the hills farther north, up to the modern gate to the Chrysokamino and to the east, where an olive grove cur (Fig. 16.2:19, 21, 25-27), on the inland (eastern) side of the modern an area rently thrives fencing, in terrace called Chordakia the walls, built (Figs. 16.2:17, 16.6). Although to recent in of schist or dolomite be rubble, appear date, these relatively area

areas may well in the Bronze Age poses modern. atively South of theMinoan terraced

have been used for grazing or agricultural pur of rubble buildings are also rel

aswell. A number farmhouse

are two fields with

terra rossa-derived

sediment

for the (see Fig. 16.2:32 for the upper field and Figs. 16.1,16.6 lower field, named Lakkos Ambeliou). Area residents currently use the as a "clay mine" for brick larger field in the low basin (Lakkos Ambeliou) as for numerous apiaries. soil, as well making and for gardening Access to the sea for this region was provided from antiquity onwards means of a substantial ravine, which curves west of Lakkos Ambeliou by and north of Chalepa Hill, between two extremely steep cliffs (gradient 3.0), to the small beach of Agriomandra (Figs. 16.1,16.6). Currently, this ravine is lined with walls of relatively recent date (Venetian or Ottoman?), which from one cliff side to the other for the length support a built path winding of the ravine, often on top of terrace walling constructed especially to sup port the path. A small church lies in a small cave about two-thirds of the way down eastern

this ravine, while

entrance

of

the

the remains of a limekiln

can be found at the

revma.

that the natural features noted here delineate the outer Assuming of the region (see Chap. 20), a total area of approximately boundaries 687,000 m2 can be ascribed to the inhabitants of the territory of Chryso or trade purposes. This kamino for agricultural, pastoral, manufacturing, area within the modern fenced land, territory mostly coincides with the at the north, to the north from northern coastline of Therio the extending to the south, and from the fence on the east edge of Lakkos Ambeliou to the sea and cliffs on the north and west.

ern

CHAPTER

17

A

Summary at

Site

of

the

Habitation

Chrysokamino

Chomatas by CherylR. Floyd

site at Chrysokamino is located on habitation near the coast of northeastern Crete, northwest of 600 m from the EM Kavousi.1 It is approximately smelting operations apparently ended in the EM

The

the hill of Chomatas the modern

village of site. The

metallurgy III-MM IA period, remains at the habitation site date to the

while

the surviving architectural LM IB and LM III periods (Figs. 17.1,17.2), with potsherds ranging in to Venetian or Ottoman date from the Final Neolithic periods. Although a considerable number of MM come from the habitation loca potsherds tion, no associated architecture has been identified. The preserved archi tectural remains are mostly from LM IIIA. The latest ceramic material to Ottoman recovered consists of a few Byzantine and Venetian sherds from

strata.

surface

in more detail in a future volume, but a be published of is the results of the summary necessary for a proper understanding conclusions from the Chrysokamino survey. The

site will

THE EARLY PERIODS site are Final Neolithic The earliest identifiable remains at the habitation to Early Minoan I sherds from strata below the level of the LM I and LM

IIIA floors and from fill in Room A, Room 8 at the north. In addition, occasional

toWall

in various diately west

over the site (including of the northern part of Room

trenches

Room

2, and in levels adjacent early sherds were found 6 and the area imme

2), but none

occurred

in any

concentration.

Larger quantities of EM II-MM pottery were unearthed from floor over most of the site, a considerable like packing and fill strata indicating structures were located in the immediate area. lihood that EM II-MM 1. For tion

information

complex,

pp. 401-403,

see fig.

Haggis

on

the habita 1996b,

13; Betancourt,

No

permanent

structures

from before LM

IB have been

found

at this

site. Floyd,

andMuhly 1997; Betancourt, Muhly, and Floyd 1998.

Only the final two passes inTrench 38 (nearWall passes in Room A were unmixed, early strata (FN-MM).

8), and the lowest The early stratum

2o6

CHAPTER

E5380

?j

E5390

N4580

Figure plan

17.1. Western of the habitation

of the portion site showing

the LM Iwalls in black. Rooms/ are indicated spaces by L. Labriola, P. Betancourt

letters

A-G.

THE

Figure

17.2.

State

plan

of the habita

tion site showing the LM III archi are indicated Rooms/spaces L. Labriola, numbers 1-16. by P. Betancourt tecture.

HABITATION

SITE

AT

C H RY S O KAM

IN O

-

C H O M ATA

S

20J

CHAPTER

208

17

Figure 17.3. The LM I bronze dag ger from Room A (X 830)

V
I

7

X541

X1448

X1321

\

17.4.

Examples

of EM

Figure 17.5. Examples ofMM ceramics: X 541 and X 1448, frag

ceram

ics:X 1321, EM I pyxis lid fragment; X1179, EM IIB Vasiliki Ware closed vessel

base.

Scale

X 932

X1497

X1179 Figure

1

O

ments

from

carinated

cups; X

1497,

base of a tumbler; X 932, body

1:3

sherd Scale

of a large,

closed

vessel.

1:3

??

mm

X808

X1480

X836

Figure 17.6. Examples of LM IB ceramics: X 1480, base of a bell cup; X 808, cup rim;X 836, neck of a jug. Scale 1:3

1

THE

HABITATION

inTrench

SITE

AT

CH RYSO

38 was found at a level below

and it could not be associated with

KAM

INO"C

the foundation

any architecture.

HOM

ATAS

ofWall In Room

20Q

8, however, A, below

I floor with its dagger (Fig. 17.3), a stratum contain sherds but no later material was also revealed. FN potsherds

the level of the LM ing FN-MM were

in both number and shape; the forms represented included and jars. The vessels were formed from either gritty-textured gray brown fabrics or reddish brown gray fabrics. The walls of the vessels were were dark in color and often burnished. generally thick, and their surfaces limited

bowls

Examples of fragmentary EM vases were also recovered, including a sherd in a fine gray fabric from an EM I cylindrical pyxis lid with pierced lugs, and sherds from EM IIB Vasiliki Ware vessels (Fig. 17.4). The range of forms was

considerably greater; numerous fragmentary cups, bowls, were recovered, as and closed vessels in light-on-dark kalathoi, technique were MM of and dishes pieces cooking pots (Fig. 17.5). This full comple ment of vessel forms supports the a that in the MM hypothesis period

MM

settlement

site existed at this location.

THE LM IB PERIOD LM IB architectural remains are confined to the western portion of the site (Fig. 17.1). Parts of several rooms from one or more structures have been excavated, revealing a technique of wall construction that differs from the later, LM IIIA technique. Architects in the LM IB period generally

The

built directly atop the dolomite bedrock and used more regularly shaped stones than did their LM IIIA counterparts. The paucity of LM IB architecture makes it difficult to saymuch about the nature of the site during this period. LM I pottery recovered from the lower strata in the western part of the site (Fig. 17.6) included vessels asso ciated with general domestic activities, such as eating, drinking, storage, and food preparation. This evidence suggests that the associated struc tures (or structure) were utilized, at least in part, as domiciles. Other finds from LM IB strata included ground stone tools, a broken obsidian blade, and clay loomweights. LM I luxury goods, as represented by a complete bronze dagger (X 830, Fig. 17.3), a complete stone bowl (X 818, Fig. 17.7), and the rim of another stone vessel, indicate a degree of wealth for the LM

Figure 17.7. Stone bowl (X 818) from LM I stratum. Scale 1:3

IB inhabitants.

2IO

CHAPTER

17

Figure 17.8. LM IIIA hearth above a LM Iwall Wall ( C), but below a

LM IIIAwall (Wall9)

Figure 17.9. Example of LM IIIA wall construction at the habitation site. Note

the

seating

of

larger

stones

on soil and smaller, chinking/leveling stones

Figure 17.10. East entrance into the LM IIIA complex, delineated with upright slabs

THE

HABITATION

THE

POST-LM

SITE

AT

C H RYS O KAM

IN O

-

C H O M ATA

S

211

IB PERIODS

site is post-LM IB in of the preserved architecture at the habitation date. Several phases of construction took place after LM IB, the latest of which occurred in LM IIIA. Sometime after the destruction of the LM IB

Most

a hearth was constructed above the level of Room Bl (Fig. 17.8). building(s), This hearth was below the LM IIIA wall running east-west as a dividing wall in Room 3 (Fig. 17.2), but above the level of the LM Iwall below it can be associated with this hearth. (Fig. 17.1). No architecture Following the phase represented by the hearth, the first recognizable LM IIIA phase can be isolated, represented by the east-west wall within Room 3, which contained LM IIIA sherds in its construction material. This wall, however, is earlier in construction

sequence

than most

of the LM

IIIA architecture

at the site. In fact, it is actually incorporated into the massive west wall of the latest phase complex (also LM IIIA in date). Many of the trenches in their excavated on the eastern part of the site had LM IIIA potsherds lowest strata, usually above bedrock. latest structure is a series of adjacent rooms and open spaces con thick enclosing walls. The construction techniques employed contrast with those of the LM IB period on the site. The builders of this The

tained within

to seat their walls on bedrock. More complex frequently did not bother often, they merely built on soil, using small chinking stones to help bed initial irregular courses of larger stones (Fig. 17.9). The type of masonry used was rubble construction with small to large irregularly shaped stones set in amortar

of soil. Two doorways from this phase were articulated using slabs on either side of their thresholds (Fig. 17.10, the east

large, upright entrance into the architectural builders were

complex). Due to the slope of the site, the to construct terraces stepping down from north to

required terraces were

formed using buttressing walls in conjunction shelves (see Fig. 17.2, the walls at the north of Room 1 and at the south of Rooms 3 and 6). Along the south of Room 6, this system was especially massive and complex, having been built buttressing in up stages. The largest enclosed space, Space 10, appears to have been an interior several built features were located. Some courtyard in which rooms or spaces were paved with slabs where the bedrock did not project (for example, see the exterior space north of Room 1 and south ofWall 8, and Room 6). The numerous built features found at the site included two south. These

with

natural bedrock

hearths, possible Not

depressions with cists or bins.

channels

pecked

into the bedrock,

and several

all of the building survives. The northeastern portion of the site is very poorly preserved, and it remains enigmatic. The walls defining Rooms 13,14, and 15 at the south were truncated at some point after their construction, probably during the construction of amodern mandra imme south of the complex. In addition, Walls 8 and 24 are incompletely to the complex and and/or hence their excavated, preserved relationship is not entirely understood. They may have functioned as further enclosing walls, or they may date to an earlier period. diately

of the pottery from this latest phase was LM IIIA in date (Fig. 17.11). Vessel forms represented in fine fabrics included cups, handled Most

212

[WW?

X1533

CHAPTER

17

X791

X1399

X1506

Figure 17.11. Examples of LM IIIA ceramics: X 791, cup rim;X1387, rim of a bowl; X1533, kylix stem;

X194 r X18

i \ X89

X1399, champagne cup foot; X1506, stirrup jar neck; X18, leg of a tripod cooking vessel; X 89, jar rim;X194, Scale

rim of a cooking dish.

1:3

bowls, kylikes, stirrup jars, jars, jugs, and other closed vessels. Many pot sherds from cooking and storage class vessels made in tempered fabrics were

also

present.

classes of artifacts were also recovered from the LM IIIA com a chisel (X1534), a hook (X1287), a plex. Bronze finds included fragmen tary knife (X 1025), a buckle (X 1026), and an elaborate, fragmentary jug (X 1149). In addition, several strips of folded lead were found in Room 5, and a sealstone (X 208, Fig. 17.12) came from Room 2. Other

stone tools at the site included saddle-shaped querns (Fig. 17.13) as as tools that were probably used pounders, rubbers, and grinders. were at the site, only a stone tools commonplace couple ground

Ground aswell While

of pieces of obsidian were found, and these derived from LM IB strata. and debitage Quite possibly this scarcity of chipped stone implements and materials from the pre indicates a change or shift in tool production ceding LM IB period. site yielded some animal bones, which for the diet of the inhabitants, as many of them showed were most common, signs of butchering. Although bones of sheep and goats a a shrew, bird, and an agrimi or examples of cattle, hares, weasels, pigs, Most

trenches at the habitation

serve as evidence

Figure 17.12. Sealstone Room

2. Scale

1:3

(X208) from

THE

HABITATION

SITE

AT

C H R Y SO KA M

IN O

-

C H O M ATA

S

213

Figure 17.13. Examples of ground stone

tools

from

the habitation

site

deer were

also found. Strata in Room A

the

(LM IB), Room 5/7 (LM IIIA), the greatest concentrations of bones on

11 (LM IIIA) yielded

and Room site.

Of the numerous

species of marine shells recovered from the site, limpets (Patella) and top shells (Monodonta) occurred most frequently. Several whole or nearly complete triton shells (Charonia) were also found in LM IIIA contexts. A number

examples came from the site. was discovered in Room 11. This One interesting deposits two pairs of long goat horns still attached to parts of deposit consisted of the skulls of the animals, a complete triton shell, a handle from a ladle of a a small a broken storage Mycenaean shape, jar, conical cup, and ground of fragmentary

of the most

stone hand tool (Fig. 17.14). The LM IIIA complex

at presents a series of con Chrysokamino rooms and spaces with built features. The was complex apparently constructed with a concern for defensible architecture. The nature of the

nected

architectural

features

as well

as the finds of ceramics, metal implements, faunal and botanical remains, etc., suggest that in part, as a residence for individuals engaged in

stone tools, loomweights, the complex functioned, agricultural pursuits, animal husbandry, and small-scale, related industries. or had access to some These inhabitants, however, also either manufactured Figure 17.14. The

ritual deposit in

in situ,

Room

11,

horns,

triton

handle,

and

showing stone shell,

other

ceramic

the goat tool,

ladle

fragments

finer objects such as the bronze jug and the sealstone, and they engaged in certain cult activity, as indicated by the ritual deposit in Room 11. The remains of the final architectural sufficient, outlying for its Bronze Age

community inhabitants.

phase at the site attest to a relatively self in a variety of capacities that functioned

CHAPTER

l8

Hall's

Edith

in

Excavations

THE ThERIOSPELIO

CAVE

byPhilip P. Betancourt and Cheryl R. Floyd

a short excavation in the In the spring of 1910, Edith Hall conducted cave. The results of the excavation were never published. Theriospelio This report is based on a study of pottery from the excavation and records of Pennsylvania Museum of preserved in the archives of the University and Anthropology. Archaeology cave is a natural cavern in limestone. The

It contains many stalactites called Therio, on a steep and almost

and stalagmites. It is in the region bare hillside. In direct distance, the cave is 230 m from the metallurgy site site. It has a small entrance (about 1.15 m and 610 m from the habitation 10 m into the earth onto a large room high), but it opens approximately and several smaller ones. From the entrance to the back, the cave is about 60 m in length. Traces of the excavations of Edith Hall or others, includ burnished pottery, and ing disturbed soil, pieces of coarse, dark-surfaced, are on human bones, visible the surface inside the cave. are known

Few details

about Edith Halls

were

Richard

work

in the cave. She and

near Gournia

at

excavating Seager Sphoungaras and they visited the cave on a Saturday when their workmen home for the weekend. They looked at the cavern on April returned on April pieces

of

25 to excavate with are

ceramics

from

preserved

six men the

in 1910,1 had returned

23, and Hall for two days. Only a few

excavation.

is contained in two let ters written by Hall to her parents. The relevant passages are reproduced that the party here. The first letter, written on Sunday, April 24, mentions of visitors had stopped the day before on the way home from a trip by sea to the island of Pseira. They landed on the coast, where they examined Most

of what

is known

about the excavations

times and the place where they smelted their bronze inMinoan where there is a big cave like those I dug inWestern Crete. I am to go there again tomorrow, with Nikolaos from the foreman [the excavations at Sphoungaras] and six men to clear this cave. It will take two days and after that comes Easter

probably the men

won

holiday when

t work.

1.Hall 1912. Letters of Edith Hall are

preserved

University

in the archives of Pennsylvania

of the

Museum.

The tion

second letter was written about

the

excavation:

on

May

1st. It provided

additional

informa

CHAPTER

2l6

18

I had begun to dig the cave when I was I there and Tuesday you. only two days, Monday a was to to the first of this week. According plan I sleep there in tent guarded by my six workmen but later we decided itwas poor I cant remember whether last wrote

to pitch a tent and that steep, rocky, mountain-side ground?a on I had better ride back and forth horseback. That meant three hours a day in the saddle and over [as] steep ground as I had ever taken a horse before but no mishaps. These Cretan ponies are sure-footed. The cave had a very small opening so marvelously that we had to enter on all fours, but once inside itwas fairly lofty, took in hanging supported by splendid stalactite columns. We on ropes strung them with reflectors and strong suspended lamps men to from stalactite had candles too. It was a stalactite. The weird

sight corners with

to see the men

tending to their work in dim dark their only light.We got out very candle smoking I hope we may have time for more caves later on.

a

good pottery.

Sherds from the cave are in two collections

in the United

States. Six sherds

are in the Mount

in South Hadley, Mas Holyoke College Art Museum, are two in the of Pennsylvania and additional sachusetts, University pieces in Philadelphia. Edith Hall Museum of Archaeology and Anthropology, (later Edith Hall Dohan) taught at both institutions during her career, and she arranged for an official gift of sherds from the Candia Museum to be exported inHerakleion) Museum (predecessor to the Archaeological for teaching purposes. This pottery is divided between the two institutions. Sherds inMount Holyoke were catalogued by Foster.2 The ceramic pieces in Philadelphia The Mosso

have also been previously described.3 in print several times.4 site and its excavation have been mentioned discussed the cavern in 1910 and noted that both he and Hazzidakis the site before Hall

collected pot excavated there. Hazzidakis or EM I), FN sherds what he called tery, including "primitive" (presumably as "MM III."5 EM II pieces, and one fragment described byMosso Haggis is surely correct in suggesting that the "MM III" sherd with white paint on had visited

a dark ground ismore likely to be EM in a few other publications.7 mentioned collected

III.6 The Zois

site has also been briefly the cave in 1990 and

visited

several sherds.8

THE POTTERY The

pottery

Coarse

in the American

Fabric

with

collections

Burnished,

belongs

Dark

to three classes.

Surface

2. Foster

1978.

3. Betancourt

in a coarse, dark fabric are in theMt. Holyoke a bowl with a H collection One (H 1, 2). piece is from vertically pierced on the exterior, and the other is from a bowl or chalice. The fabric is lug a coarse and gritty. It is heavily burnished and fired in reducing atmosphere

Two

sherds of vessels made

to create a dark brown surface. The

fabric

a dark brown, unevenly colored I fabrics from Mochlos and like the FN-EM

to black fabric with

is not

Haggis

1983,

most

4. The

complete

p. 14. discussion

is

1992, pp. 170-173.

5. Mosso

1910,

p. 290.

6. Haggis

1992, p. 171.

7. Becker

and Betancourt

1997,

p. 114;Haggis 1996b, pp. 380-381, fig. 4. 8. Zois

1993,

p. 340.

IN

EXCAVATIONS

THE

CAVE

THERIOSPELIO

Pseira, which have phyllite from Sphoungaras.10

2TJ

but it is visually

inclusions,9

similar to a fabric

un chronology for this class of coarse, burnished pottery is still certain. Similar vessel shapes (not always made in identical fabrics) have been recognized from Kavousi,11 Alykomouri/Hagios Antonios,12 Sphoun and several other places.17 No clear garas,13 Pseira,14 Vasiliki,15 Mochlos,16 The

on

based

has been found,

sequence

stratigraphie are

between

comparisons

and chronological

the

surface

shapes,

synchronisms and

treatments,

and firing technology with Final Neolithic pottery from manufacturing elsewhere in Crete.18 The class may persist into EM I, but it does not ap pear in the late EM I assemblage from Hagia Photia.19

with

Fabric

Gray

Pattern

Burnishing

Four sherds of pattern-burnished gray pottery (Pyrgos Ware) from the cave are in the American collections. Two pieces are in South Hadley, and two are from chalices, and one piece are in pieces Philadelphia. Three pieces (H 3) is either from a bowl or a chalice.20 This class of pottery, called Pyrgos fine, hard fabric with few visible inclusions. The Ware,21 has a medium surfaces of the vessels are rubbed with implements when in a leather-hard surface. This stage to align the clay particles, causing a shiny, burnished class sometimes has burnished patterns achieved by rubbing over specific areas. The vessels are fired in a reducing atmosphere, which accounts for their usual gray color. This ware

is particularly

sist a little

longer. Pattern would be needed

characteristic

of EM

I, although is widespread

burnished

itmay per in Crete.22

pottery to determine the relation of this pottery (if Analyses as defined to the later "Fine Gray Ware" of EM I-EMIIA any) byWilson and Day.23 Several authors have discussed

the chalice. The general development set out by Hood many years ago,24 and more detailed analyses have been made by others.25 Based on the fabric and the shapes, the examples

was

here can be assigned to EM I (H 3-H 5) andEM I or EM IIA (H 6). Vasiliki

Ware

Two pieces of Vasiliki Ware are in the assemblage (H 7, H 8). They aremade of the fabric typically used for the finest versions of this ware, consisting of amedium fine fabric in the gray to pink color range with a slipped surface 9. Banou 10. Foster

1995b, 1978,

p. 109; no. 2.

1998,

p. 15.

16. Seager 1909, p. 279; 1912, opposite

nos. p. 82,

29-42,

fig. 48; for

11.Haggis 1992, locus 17/92; 1996b, pp. 389-393. 12.Haggis 1992,, locus 58; 1996b,

the FN date, seeVagnetti and Belli

pp. 389-393. 13. Hall

1999;Hayden 2003a; 2003b; 2004,

court

1983,

Betan pp. 46-48; p. 46, nos. 110-113. 1912,

14. Banou 1995b, p. 109; 1998, p. 15. 15.

Seager

1904-1905,p.

212.

p. 137. 17. Betancourt

19. Davaras 20. For

the

1999;

Branigan

p. 42.

18.Vagnetti 1973; Vagnetti and Belli 1978; Vagnetti, Christopoulou, andTzedakis 1989;Hood 1990a; Manteli 1992.

see

1997.

Haggis

21. Branigan 1970a, p. 21; 1970b, p. 18; Betancourt

1978,

1971. shape,

1985,

pp. 26-29.

22. See the list in Betancourt 1985, pp. 26-27. 23. Wilson

and Day

1994,

pp. 4-22.

24. Hood 1971, p. 38, fig. 14. 25.Wilson 1985;Momigliano 1990;Hood 1990a;Haggis 1997.

2l8

CHAPTERl8

that was

intentionally mottled during the firing process. The sherds are in the collection of theMount Holyoke College Art Museum. is the definitive ware for EM IIB.26 It is first used in Vasiliki Ware this period, and it does not occur in deposits from East Cretan EM III? MM IA. Recent research suggests that this ware was produced by a limited of East Cretan workshops.27

number

CATALOGUE H 1

(Mount Holyoke College BAI.3d) Fig. 18.1 Bowl with vertically pierced lug. Max. dim. 4.8; Diam. of vessel ca. 14; thick (0.8) wall. A gritty fabric; core, dark gray (10YR 4/1); exterior surface, reddish brown (2.5YR 4/4); interior surface, very dark gray (10YR 3/1). Slightly convex profile;

small

at Gournia

found

on

exterior

and

interior.

fabric is similar to an early fabric from theMirabello

Comments: The area

Burnished

vertically.

lug pierced

and

Sphoungaras.

Marked

"Chryso

Kamino"

Bay

in Greek,

in

pencil. FN (to EM I?). Bibliography: Foster 1978, no. 5 (incorrectly published as from Knossos). H 2

(Mount Holyoke College BAI.3g) Chalice (or bowl?), rim sherd. Max.

dim.

of rim

7.9; Diam.

ca. 15;

Fig. thin

18.1

(ca. 0.5) wall.

A gritty fabric, reddish brown (5YR 4/5); gray areas on exterior surface.Convex rounded

straight,

profile;

rim. Burnished

on

exterior

and

interior.

Comments: The fabric is like H 1.Marked "Chryso Kamino" in Greek, in pencil. FN (to EM I?).This vessel ismore likely to be a bowl than a chalice. Com pare Xanthoudides 1918, p. 153, no. 84, fig. 10. If it is a chalice, it ismore globular than

as is the

usual,

case with

an

example

from Mochlos

in the Museum

of Fine

in Boston (09.655), a piece from Kanli Kastelli (Alexiou 1951, no. 5, pi. 14, fig. 1), and one from Pyrgos (Xanthoudides 1918, p. 153, no. 76, fig. 10). Bibliography: Foster 1978, no. 12.

Arts

H 3

(University of Pennsylvania Museum Bowl

or chalice,

Diam.

of rim

rim

42-34-2-A)

Fig.

18.1

sherd.

ca. 15.5.

A fine gray fabric (gray, 10YR 6/1); heavily burnished. Straight rim. Burnished vertical

lines

forming

panels.

Comments: Pyrgos Ware. The date is late in EM I (based on the presence of the shape atHagia Photia Siteias, forwhich seeDavaras [1981], fig. 2; 1977, pi. 603; and 1989b, p. 266, pi. 149). Parallels: For vertical burnished lines, seeHaggis 1997, pi. 92, fig. c. For the 1918, p. 150, nos. 42-45, especially 42 and shape of the rim, see Xanthoudides 45, fig. 8; p. 151, nos. 56-59, especially 58, fig. 9; p. 153, nos. 75 and 77, fig. 10, republished by Haggis 1997, pi. 91, fig. 5:a, no. 3, fig. c, lower row nos. 2 and 4, all from the Pyrgos cave;Davaras 1977, pi. 603, from Hagia Photia. Bibliography: Betancourt 1983, no. 13. H 4

(University of Pennsylvania Museum Chalice, Diam.

rim of rim

sherd. ca.

15.5.

42-34-2-B)

Fig.

18.1

26. Warren court

1972, 1979;

pp. 29-30; Betan

pp. 94,108-109; 1985,

27. Whitelaw Wilson,

1970a,

Branigan

pp. 43,45-48. et al. 1997; Day,

and Kiriatzi

1997.

IN

EXCAVATIONS

THE

CAVE

THERIOSPELIO

219

1 *

r

HI

H2

7 H3

H4

7

K

H5

H6

AJ HI

H7

Figure 18.1. Profile drawings of sherds Scale

from 1:3

the

Theriospelio

A fine gray fabric (gray, 10YR 6/1); heavily burnished. Straight rim.Burnishing

cave. over

entire

exterior

Comments:

interior. Ware.

Bibliography:

H 5

and

Pyrgos Betancourt

EM

IB. See H no.

1983,

(Mount Holyoke College BAI.4a) rim

Chalice, Diam.

3.

14.

Fig.

18.1

sherd. ca. 19; max.

of rim

dim.

8.8.

A fine gray fabric (gray, 10YR 5/1) with browner areas on the surface; heavily burnished,

vertically,

Comments:

to create

patterns.

Ware.

Marked

Pyrgos

See H 3.

EMI.

Bibliography: Foster 1978, no. 8.

Straight "Chryso

rim. Kamino"

in Greek,

in

pencil.

220

CHAPTERl8

H 6

(Mount Holyoke College BAI.4b)

Fig.

18.1

Chalice, base of the conical bowl and upper part of the pedestal. Max.

dim.

ca. 9.6.

A fine gray fabric (gray, 10YR 5/1) with browner areas on exterior and interior on

burnished

surfaces;

exterior

and

interior

of bowl.

Incised

lines

on

exterior

of

pedestal.

Comments: Marked "Chryso Kamino" in Greek, in pencil. EM I-IIA. Parallels: Xanthoudides 1918, p. 151, no. 56, fig. 9, from Pyrgos; Haggis 1997, nos. KT53 and KT56, pi. 94, fig. a, from Kalo Chorio; for EM IIA, seeWilson 1985, p. 302,

11, no.

fig.

from

31,

Knossos;

Warren

1972,

p.

157,

no. P44,

from

Myr

tos.

Bibliography: Foster 1978, no. 20. H 7

(Mount Holyoke Conical

bowl,

Diam.

of base

College BAI.5a)

base

18.1

Fig.

sherd.

ca. 4.7;

thick

(0.6-0.9)

wall.

A fine fabric (reddish yellow, 5YR 7/6). Slipped on interior and exterior; red

mottled, on

exterior

and

just

areas.

black above

Burnished.

articulated

Flat,

base with

deep

groove

base.

Comments: VasilikiWare. EM IIB. For a list of conical bowls see Betancourt 1979, pp. 33-36. Bibliography: Foster 1978, no. 17. H 8

(Mount Holyoke College BAI.5b)

Fig.

18.1

Jug, body sherd. Max. dim. of sherd 6.4; thick (0.7) wall. A fine fabric (core dark gray, 10YR 4/1; exterior surface reddish brown, 2.5YR 4/4; interior surface very dark gray, 10YR 3/1). Slipped on exterior; mottled, red and black

areas.

Burnished.

Comments: VasilikiWare. Marked "Chryso Kamino" inGreek, in pencil. The curvature of the sherd, which is definitely from a closed shape, makes it unlikely that

it could

come

from

any vessel

form

other

than

a

EM IIB. For a list of jugs in this ware, see Betancourt Bibliography:

Foster

1978,

no. 21.

large

jug

with

a raised

1979, pp. 46-48.

spout.

CHAPTER

19

in Context:

Chrysokamino

A Regional

Archaeological

Survey byDonald C.Haggis

This

is an overview

chapter

of Bronze Age settlement patterns in the Ka is to provide a broad regional and chronological

vousi region. The purpose area of draw Chrysokamino, perspective for the analysis of the sites in the on the results of an intensive from conducted survey ing archaeological is not to reduplicate previous or forthcoming reports on the Kavousi survey, but to comment in detail on the specific area and periods relevant to the Chrysokamino project. Bronze Age settlement are on the two in the broader Kavousi patterns region brought to bear 1989 to 1991.1 The

intent

areas excavated: the FN-EM III metallurgical site of Chrysoka specific mino (site 32 in the Kavousi survey; Figs. 19.1,19.2) and the FN-LM III site on the spur of the hill of Chomatas 34 in habitation the Kavousi (site survey; Figs. Prepalatial the context

19.1,19.2). The and Protopalatial, of the southwest

principal focus of discussion will be the late and Neopalatial and Postpalatial periods in Kambos and southern end of Chomatas.

area of the sites excavated by the Chrysokamino project is composed of a number of locales with distinct toponyms (see Chap. 1). these localities, the most important is Chordakia or Choriodakia Among The

broader

("little villages"),2 situated on the southernmost spur of Chomatas and at the southwestern edge of the Kambos (Figs. 19.1,19.3). A small valley, Lakkos and from the hill of Chalepa Ambeliou, separates the hill of Chomatas drains westward down a narrow gorge into the small inlet and roadstead at (Figs. 19.1,19.3). A kalderimi (Venetian- or Ottoman-period Agriomandra cobbled path) links the Kambos with this port. Evidence would suggest that from the Roman period until World War II the small harbor served as an important gateway to the Bay ofMirabello. No doubt the same was true for the Bronze Age. Immediately off the shore in the depths of Agrio are deposits of amphorae of the 1st century A.D., testifying to the s use. In more recent times, boats would anchor them site Roman-period selves in this rocky cove to receive cargoes of carob, olive oil, soap, cheese, as oil made from cedar, sage, and laurel?bound and other products?such

mandra

1.Haggis 1992,1993b, 2005. 2. Boyd 1901, p. 156. 3.

Spanakis

1991,

1996b, 2002,

pp. 322-323.

for Herakleion,

and other Aegean and Mediterranean Athens, ports.3 as of in in this locale the Bronze modern times, lies importance Age, not only in its proximity to the port at in the but also facility Agriomandra, on the southeastern of the soils agricultural potential slopes of Chomatas The

CHAPTER

222

19

Theriospelio

Kambos

Agriomandra

Figure 19.1. Map survey

area

in the

of the Kavousi late Prepalatial

period

1000m

Figure 19.2. Map survey

period

area

of the Kavousi

in the Protopalatial

A

REGIONAL

ARCHAEOLOGICAL

SURVEY

223

,x7/Mt. Schiniasy

Metallurgy Workshop

Figure

19.3.

Map

\^7

of the Kavousi

in the late Prepalatial period

area

between the areas of Kephalolimnos and Katsoprinos rossa to terra contrast In marked the soils farther afield in (Fig. 19.1). rocky the central plain of Kambos, the phyllite alluvium of this small valley com

and in the Kambos

the optimal arable land for rain-fed agriculture (Fig. 19.4). The more retain soils moisture than the stony limestone phyllite plant-available and dolomite alluvium, and local farmers have traditionally reserved them prises

for small gardens, wheat (rather than the usual barley), and intensive olive planting. The oldest olives in the Kavousi area are found in phyllitic soils. In an olive Chordakia, the Roman settlement of Kephalolimnos (Fig. 19.1) has a basin and millstone (orbis), reminders of the impor crushing (trapetus) tance of this area and these soils in ancient olive oil production (Figs. 19.5, area were also found to be several 19.6). Sherd densities in the Chordakia times greater than in the main area of Kambos,4 a good indication of the intensity and continuity of land use since the Bronze Age. The

actual occupational and its surrounding history of Chordakia Roman the Settlement patterns and land use long predates period. in section the this discrete of Kavousi lowlands and coastal hills practices area

can be discussed the broader

4. Haggis 1996b.

as a unit. While

this area is not physically isolated from and the north Isthmus of Ierapetra, its late

region of Kavousi and sites illustrate characteristics of a pattern Prepalatial Protopalatial a self-contained households. may represent group of interdependent

that

224

CHAPTER

19

Figure 19.4. The phyllite terraces of Chordakia and the Kavousi plain (the Kambos),

seen

from

the west

Figure 19.5. An olive crushing basin (trapetus) at Kephalolimnos

Figure 19.6. A millstone Kephalolimnos

(orbis) at

A

Figure 19.7. Theriospelio

REGIONAL

ARCHAEOLOGICAL

SURVEY

225

cave from

the west

THE PREPALATIAL AND PROTOPALATIAL PERIODS In Chordakia

and its vicinity, evidence for EM I?II is not substantial. While

II pottery has been recovered from the cave at Theriospelio (see to cave not is in is AF 34 this material the likely belong App. G), Chap. 18; to a permanent habitation site (Figs. 19.1,19.7). Neither the position nor the topography of the cave, which lies on the precipitous western slope of FN-EM

Chomatas, suggest a proper location for a long-term settlement. Therio an is isolated extremely narrow, low entrance and spelio topographically, with little room or sign of activity on the terrace outside. A special function, such as burial site, cult activity, or periodic camp, may be reasonably hypothesized. individual sherds datable to EM I?II have been found through While out the Chrysokamino-Chordakia area, no single site initially suggested a of from surface survey. The systematic excava locus habitation significant tion at the Chrysokamino habitation site has changed the picture somewhat, in its earliest pre-architectural producing both FN and EM I?II material levels. The nearest substantial EM I?II settlements found in the survey, is some 1.5 km however, are located at Kavousi village (Fig. 19.3), which to the southeast of Chordakia,5 and atAlykomouri in the Hagios Antonios at the northeastern to 3 km the end of Chomatas about north, valley, lying are founded in the Final Neolithic and settlements (Fig. 19.3).6 These some in sites the Kavousi of the earliest represent permanent occupation are settlements. Each is located on region. They typical of Early Minoan near a a hill or rise above the plain, perennial spring, and in close proxim to alluvium. Kavousi ity village appears to have been abandoned phyllite in in EM IIB, while Alykomouri continues into EM III. The farmhouse area, like the Chrysokamino workshop continues into EM III. Given the evidence

the Chrysokamino

Theriospelio, habitation recovered 5. Haggis 1996b, p. 391. 6. Haggis 1993b.

in excavations

and the cave of for FN-EM

II

in the Chrysokamino-Chordakia area, site in the locality it is very likely that the original Prepalatial habitation was the site (Chap. 17). habitation Chrysokamino

CHAPTER

226

In the Chordakia-Chrysokamino in EM III-MM IA. This is a period and settlement in the Kavousi area.

19

area, several new sites are founded of major reorganization of land use the sites of function special Including

and the Theriospelio cave, there (the metallurgy workshop) Chrysokamino were no less than nine sites in use in the late Prepalatial period (Figs. 19.1,19.3). Seven of these sites?probably individual farmhouses?were established above the Kambos

on the south and southeastern

and slope of Chomatas sites clustered neatly around the rich

at the northern

end of Chalepa. The alluvium of Chordakia and the mixed

phyllite liou are all of similar size. Their

terra rossa of Lakkos Ambe

equidistant spacing (between 100-200 m) the areas between the houses to be used for sizable gar dens and household livestock. The primary agricultural focus was, however, the phyllite soils in the Kambos and on the east-facing slope of Chomatas.

would

This

have allowed

is the area of Chordakia

Kephalolimnos. trated precisely

between

It is noteworthy at the watershed

the localities

that the settlement

of Katsoprinos and concen activity is

that separates Lakkos Ambeliou from con would have This position (Figs. 19.1, 19.3). permitted trol of an important route and access to the port at Agriomandra. Clusters of sites also appear inEM III-MM IA in the Hagios Antonios the Kambos

area of Xerambela (Vronda) and Chon valley and above Kavousi village in the drovolakes (Fig. 19.3). Each cluster occupies an area with phyllite soil and a permanent water the overall pattern of settlement supply nearby. While is dispersed,

the clusters of houses represent distinct groups, each occupy concentrations of arable land. Social conditions were perhaps discrete ing similar to those in the nucleated villages of EM I?II; the inhabitants of these clusters may have had a sense of local identity, autonomy, and direct to the land in the immediate area of each group. clan or family connections cluster pattern implies that individual sites reflect family units. The as units could have been interdependent, sharing land and water supplies well as lineage ties. The

and 45 may have been the settlements whose Sites 28,34,35,36,41,44, the site at Chrysokamino occupants operated metallurgical (Fig. 19.1). aspect, they are protected from the north winds By contrast, the Chrysokamino smelting site, by the peak of Chomatas. is favorably exposed to the winds situated on the west-facing headland,

With

their southeastern

required by the smelting process.7 The settlements, probable control of the is no doubt significant because the smelting port facility at Agriomandra raw materials, operation would have required the importation of fuel and such as the copper ore itself, aswell as export of the copper. Located at the the sites would and Kephalolimnos, have communicated easily with the Kavousi plain, the Isthmus of Ierape was self-sufficient tra, and Mirabello Bay. The late Prepalatial community

watershed

between

Lakkos Ambeliou

a view to coastal and trade Aegean agriculturally, but clearly situated with via the Bay ofMirabello. It also had easy access to the Isthmus routes and inland passages through the Kavousi mountains. inMM In the subsequent Protopalatial IB, the period, beginning to be used, and several of the old Prepalatial sites continued new sites were established (Fig. 19.2). This growth represents continuity locations

7. See discussion

in

Chap.

2.

A

REGIONAL

SURVEY

ARCHAEOLOGICAL

227

at each of the Prepalatial sites. In the Chor region, the total number of sites doubled, with 16 dakia-Chrysokamino site IB-IIB. Only the metallurgical farmhouse-sized sites in use inMM and expansion

of habitation

were abandoned, and possibly the cave at Theriospelio IB?II their special functions. The settlement pattern inMM

at Chrysokamino emphasizing is essentially

a continuation

IA: single ofthat established in EM III-MM m ca. are 150 located above the Kambos farmhouses, spaced apart, plain at the Kavousi region, this the fringes of pockets of arable land. Throughout twofold growth of settlement is reduplicated in each of the site clusters in IB?II and Hagios Antonios. The MM Avgo, Xerambela, Chondrovolakes, of of sites represents the maximum and distribution settlement dispersal area. in maximum the Bronze the Kavousi Age population possibly

The

of

Significance

in the

Changes

Prepalatial

Landscape The

a late Prepalatial period witnessed profound restructuring of settle in EM I?II to the shift from isolated nucleated villages

ment?the

late Prepalatial system was to provide the basis and stability for the growth and expansion of settle ment throughout the subsequent Protopalatial the period. Furthermore, clusters of EM

farmhouse

III-MM

IA. This

IA landscape represent a decisive changes apparent in the EM III-MM of FN-EM break with the long-established nucleated communities II. In sum, the changes apparent summary discussion here. 1. New

sites were

sometimes

in EM

established

III-MM

in EM

IA are remarkable,

III, while

EM

deserving

IIB sites were

abandoned.

2. A new structure of settlement

appears in the archaeological record, the pattern changing from nucleated to dispersed, with farmhouses clustering in arable zones sharing agricultural land and water. The Chrysokamino site is perhaps the habitation location of the original nucleated EM I?II settlement from

the late Prepalatial pattern develops in the Chordakia area. Similar Chrysokamino trajectories of settlement develop ment and cluster formation can be traced inHagios Antonios, The community Kavousi, and Xerambela-Chondrovolakes. the cluster retained a local would have pattern represented by was reinforced as as communal ties that well identity by lineage

which

in the locality. Such activities might have included cooperative exploitation of fields between the houses and

activities

the immediate

gardens, communal stock terrace and wall building, staple rearing, irrigation, terracing and surplus storage, and transportation. The Chordakia

beyond

household

cluster is likely to have been responsible for of the metallurgical site at Chrysokamino. a 3.Within site the clusters, single building appears to have been particularly prominent because of its central location and large stone or Cyclopean masonry. Such architectural elaboration Chrysokamino the operation

CHAPTER

228

indicates

IC

social stratification within

the site clusters, with the a or to prominent house belonging important socially politically or individual. The economy of the site cluster family including would have probably required a leader with an Chrysokamino and organizational role. important managerial 4. Land use patterns changed new sites were established Crete, including Bebonas (700-800

IA, significantly. In EM III-MM at several places in this part of the upper elevations of Avgo (600-700 masl),

masl), andThriphti Argira (Hagia Anna) (600-700 masl). Xerambela (400 masl) was also an important site in this period. Such settlements, located high in the Kavousi are evidence of Mountains, expanding agricultural and pastoral of new systems of interests, perhaps signaling the development land use and land tenure. These changes conceivably involved the control and intensification of agricultural and pastoral a elite.8 Rapid expansion of settle production by land-owning ment into the Kavousi Mountains in EM III may also have tied directly to the need for combustible materials, forest resources, to operate the furnaces at Chrysokamino. zones may in the mountain The settlements conceivably been

i.e., have

from the coastal areas increasing need for such is possible that the deforestation It of Avgo, Papoura, supplies. to and Thriphti led the expanding agriculture in these areas.

benefited

or are clearly fortified sites, which common on in north Isthmus elsewhere the and the very island in the period EM II-MM IA,9 strongly suggests that were or site clusters the Kavousi-area either interdependent

5. The

absence of defensible

mutually beneficial. That is to say, the social and political structure across the landscape appears to have been well integrated,10 with strong connections between sites. 6. Imports of pottery with diorite/granodiorite temper, derived from the Gournia and Kalo Khorio regions west of the in the archaeological record Isthmus, increased dramatically of the late Prepalatial period. These fabrics made up a large coarse ware assemblage until MM III, part of the Kavousi-area wares when local phyllite-quartzite the replaced tempered Kalo Chorio 7. The

expansion

the growth

fabrics.11 site at Chrysokamino of the metallurgical and sites in the Chordakia of neighboring habitation

cluster is evidence for significant economic Chrysokamino in the region, involving links to potentially distant changes sources of raw materials and extraregional trade. The inter raw materials crucial in the production of regional exchange of both agricultural tools and luxury goods signals the emergence of awealth distribution economy fueled by a new and extensive staple finance system. was In the Protopalatial period, the settlement pattern dispersed and the site clusters, including the one at Chrysokamino, remained intact. In fact, the total number of sites more than doubled in most areas. This continu

8.

Manning

9.Manning

1994,

p. 234.

1994; Alexiou 1979,

1980.

10. Plog 1974. 11.Haggis andMook 2000.

1993; Haggis

A

REGIONAL

SURVEY

ARCHAEOLOGICAL

229

ity of settlement, pattern from EM Prepalatial

along with the constancy and longevity of the cluster III until MM II, underscores the importance of the late settlement system in establishing the trajectory of settlement

use. The area expansion and land occupational history of the Kavousi really anew in EM III with new in social and ways begins changes complexity of exploiting the landscape. These changes were no doubt felt throughout the subsequent

period.

Protopalatial

NEOPALATIAL AND POSTPALATIAL PERIODS occurred inMM IA when III-LM the Chrysokamino discontinuity a marked decrease in area dense cluster of The experienced population. and farmhouses into fell disuse Prepalatial Protopalatial finally (Fig. 19.8).

A

It was

situated Neopalatial houses: the by two prominently site 34 in habitation the Kavousi (site survey; Fig. 19.1) Chrysokamino with a southeastern aspect, overlooking Lakkos Ambeliou, and Katsoprinos replaced

(site 28 in the Kavousi the plains of Chordakia

survey; Fig. 19.1), facing east, immediately above and Kambos. The Chrysokamino habitation site, known both Evans and is the farmhouse excavated by probably long Boyd,12 as part of the same pattern of abandonment, Chrysokamino project. The at and Avgo, coincided with a shift Chondrovolakes, Xerambela, apparent to the lowland plain. Of par in agricultural emphasis from the mountains ticular interest to the Neopalatial inhabitants of Kavousi were the north Kambos plain area and Tholos Bay (Fig. 19.9). With the changes in land use came a dramatic nucleation in the plain, especially at of population (site 5) Hagios Antonios settlements. Other villages 11, 15, 17, 18 [Fig. 19.9], which runs the full length

and Tholos and hamlets

(site 1), which became grew up in the Kambos

town-sized

plain (sites and 23, 27 [Fig. 19.8]) along the Platys river, of the plain and drains into the sea at Tholos

Bay. It is likely that these sites intensively exploited the terra rossa soils of the lowland for the first time, perhaps utilizing the river for direct irriga tion of these otherwise arid fields. The overall pattern of nucleation and into the Kambos, Hagios Antonios, and Tholos population movement areas attests to economic changes in the transition from the Protopalatial to the Neopalatial period. towns and In LM IA Kavousi, dominated by the nucleated villages of the north Kambos, what was the role of single, isolated farmhouses such as the one at Chrysokamino? Each of the houses was situated in areas of the the clusters. The Chrysokamino habi precisely Protopalatial on tation site (like Katsoprinos the site at Avgo, and the site Chomatas, at

Skali) clearly exploited the phyllite soils in Chondrovolakes-Panagia their immediate vicinity, the best arable land in the Kavousi region. Near II Cyclopean walls of the house at Chondrovolakes the EM III-MM is

tree in the a trunk diameter of about region, with which farmsteads, megalithic frequently dominated discrete estates. controlled small Like the LM IB house involved valleys, probably near the eastern end of the Mochlos at Chalinomouri, in flax production the oldest

living olive

3.0 m. These

12. Boyd 1901, p. 156. 13. Soles 1997, pp. 429-430; Mochlos

IA, pp.

103-132.

plain,13 they must

have been

involved

in intensive

farming

of cash crops

CHAPTER

230

IQ,

agricultural activities.14 Intensive farming and crop suggest an economic or political hierarchy in the region. specialization Such a hierarchy is perhaps echoed in the ranked site sizes apparent in the and other

Neopalatial These

specialized

Figure survey

19.8. area

of the Kavousi Map in the Neopalatial period

settlement

pattern. in the settlement

structure suggest concomitant changes changes in land use. Larger contiguous areas of land appear to have been brought under cultivation. The farms, abandoned small, dispersed Protopalatial new to in the shift of population from the clusters villages and towns in I, were

into larger landholdings, turned into probably incorporated or in the north The of sediment left fallow. pattern pasturage, aggradation of these changes in land use in Kambos may enhance our understanding the Kavousi region in the Late Bronze Age.15 LM

In an exposed scarp in the Platys river channel, soil scientists of the an alluvial event marked by silty overbank de Project discovered a on top of a stratified lag channel. On top of this alluvial episode, posits

Kavousi

stable living surface had formed. The terminus ante quern for this episode is provided by the overlying soil formation, which has a radiocarbon date I date.16 Morris has suggested of 3040 ? 90 B.p. and sherds of MM-LM in the upper elevations, perhaps triggered by that landscape destabilization deforestation, maybe consistent with the aggradation episode.17 The pattern sites in the mountains south and east of Kavousi illustrates of Protopalatial the character of human activity, which was certainly dense, dispersed, and II and land use practices in EM III-MM of long duration. Habitation no doubt transformed the natural environment of Kavousi. The dramatic shift in settlement

patterns underscores

in LM

I, with

of population effects on the local landscape the nucleation

in the Kambos, the potential and the probable changes in land use practices. It is likely that the alluvial was an erosive event, of the ultimately caused by the abandonment episode

14.Hayden 1997; Hood 1971. Cadogan 15. Morris 2002,

1983;

16. Morris

2002,

p. 75. pp. 68-69.

17. Morris

2002,

p. 75.

A

REGIONAL

ARCHAEOLOGICAL

SURVEY

231

V\\ W // i ?7///h////-^\C0 U\ )?//l,,fi(l^=>^\\4i); Schinias *\ | I / / IM \ / / / Mt. /^BV

V ^^~^?r^^.

f

19.9. Map

of the north

Kam

bos and Tholos Bay in the Neopala tial period

MM

^

^ II household

i i I I

;

^

\\(
\

River Bedi Platys / / /

inyyyy

Figure

/ / / I I I

\ I HagiosAntonios ^^^^k^r I

'S//" < ^? \\ \ \J \ Mt.Chomatas "} %^B

z^?

/ /

?y^/it y//n=^yj !

oJ>Y\|i V^j^L)

;/

n

till: ^Jlllttl

farms and site clusters. New

were introduced herding practices of areas of the Kavousi mountains.

inMM

III-LM

intensive

farming and I, leading to the disuse

In the subsequent Postpalatial period, the Kavousi area experienced a remarkable population decrease, perhaps partly because of these drastic environmental changes. The overall number of sites dropped from 40 habi tation sites in LM I to a combined total of 10 habitation and burial sites in was LM IIIA-B.The LM I farmhouse near Chrysokamino reoccupied in LM IIIA as was the megalithic farmstead at Avgo.18 The LM IIIA reoc (site 5 in the survey; Fig. 19.9) cupation of the town of Hagios Antonios was

tions of pottery and

spotty,

characteristically

discrete

areas.

suggesting The

with

sparse

and

surface

remains

actually

concentra

discontinuous

reuse and squatting

in individual reflect

the

buildings

character

of

LM

III occupation evident in excavated contexts at Gournia, Pseira, and Mochlos. While the overall reduction in the number of inhabited sites in identifying LM II-IIIA may be slightly exaggerated by the difficulties on east sites of in date Crete, the decrease in the survey pottery Neopalatial total number of sites does appear to mirror the evidence from excavation, 18. Hayden

1997.

suggesting the magnitude LM IB destruction.

of abandonment

and resettlement

following

the

CHAPTER

232

19

CONCLUSIONS a number of broad area of Chordakia-Chrysokamino, comprising sites including those excavated by the Chrysokamino project, is essentially a microcosm of the Kavousi area as awhole. The various settlement sys

The

tems and the changing land use and settlement patterns apparent results of intensive survey and excavation are duplicated on various throughout the region. Although Chordakia-Chrysokamino linked to a broader region, the local settlement patterns are nevertheless

of extreme

of sociopolitical change and eastern Crete. What

in the scales

is perennially in this single area the implications

importance in understanding in the north Isthmus, the Mirabello Bay region, were the social, economic, and political dynam

involved in the changes from nucleated village ics and cultural processes cluster and from dispersed site clusters to nucleated town? These changes are asmuch dependent on local social systems and environ ment as on broad interregional currents of exchange and external political to settlement

in Bronze Age Kavousi influences. The normative mode of settlement was not the or the town, but the of interdependent village cluster?groups of arable land. The cluster pattern farmhouses occupying concentrations ostensibly represents the optimal settlement arrangement in this particular environment.

The expansion of settlements and exploitation of agricultural land from as early as EM III suggests a highly stable and dynamic social II. and economic system capable of sustaining a constant growth until MM the internal cohesion, The longevity of the cluster pattern emphasizes between clusters, and cultural identification with strong interconnections the region and local environment. This highly integrated human-landscape interaction fostered economic prosperity and population growth. Even so, the draw of a centralized economy, the the clusters could not withstand in regional production and exchange, and the direct control of or indirect centers higher-order palatial influence. Nor could the limited arable land of Kavousi ultimately support the population explosion of changes

a site hierarchy and town economy I?II. By the Neopalatial period, had replaced the archaic site clusters. The farmhouse near Chrysokamino had become part of awider world.

MM

CHAPTER

20

The

Geographic

the

Chrysokamino

Boundaries

of

Farmstead

Territory byPhilip P. Betancourt

small farming communities must have a close relationship with the land that provides their economic livelihood, any understanding of the cultural history of an agricultural society must consider the use of the land

Because

scape as an essential part of the scientific data. In applying this assumption toMinoan Crete, one of the first steps must be the definition of boundar ies. Establishing

the extent of the territory providing the hinterland for a foundation of discussions and conclusions.

site is an essential

specific In this study of the region around Chrysokamino, the word "territory" is defined following Wilkinson, who suggests that a territory is "the area a group in the course of its seasonal subsistence regularly exploited by degrees of exploitation may be recognized. cycle."1 Different for the study of the extent of the territory controlled by Methodologies agrarian estates have been more developed for the Roman Empire than for earlier periods of antiquity. The sizes of Roman estates seem to have been so that estimates of territorial size based on regional and local conditions, vary widely.2 Estimates for different regions have varied from 50-2,000 ha, and even if both extremes are considered atypical, the situation emphasizes the variability of land use under different conditions. In order to examine or a one an examina must the territory for estate, specific farm begin with it existed. tion of the conditions under which

THE CHRYSOKAMINO REGION The mos

on the Cretan coast between Tholos region Bay and Pacheia Arri was first inhabited in the Final Neolithic Bay period, and its Bronze lasted until LM IIIB. The evidence, which is discussed occupation

Age in Chapter

the occupation was 19, is not sufficient to establish whether or intermittent, but most pottery are periods represented. Until EM IIB, only a few dispersed sites existed, including the Chrysokamino site on the hill of Chomatas, habitation the nearby metallurgy workshop, continuous

1.Wilkinson 2. See Vallar

1981, p. 251. 1991,

pp.

10-11.

and the cave called Theriospelio. The pattern of small, dispersed sites eastern this of the of the side Gulf of Mirabello, part prevailed throughout

CHAPTER

234 It was

followed

20

inMM

IB-II by small clusters of settlements. The local sites passed out of existence afterMM IIB, leaving habitation location to command this part of the I to LM III.

clusters of neighboring only the Chrysokamino terrain from LM

most of the evidence for the Chrysokamino is farmhouse Although from the Late Bronze Age (Chap. 17), the residents must have farmed the land from the time they first settled in the region until they left the coast at the end of theMinoan period. Because few objects other than sherds survive from the early phases at the site, little can be said about the nature of the Prepalatial and Protopalatial occupation or even about the exact location or of the building buildings that may have been present. The presence of chaff from barley in the furnace chimney fragments, however, shows that the was region already producing grains in the Early Bronze Age (Chap. 12), and the location of the settlement near the best local land supports the same conclusion. The earliest surface pottery from the farmland in the vicinity of the dwelling is from the Final Neolithic (App. H, AF 32.1). Evidence for LM I is more fully preserved than for earlier periods, and the occupation from LM III is even better documented. A substan complex from LM III is present, and the location of on virgin soil indicates it must have increased in of the architecture part at size this time. The LM III building almost certainly housed a larger number of people than lived there in LM I. The residents from both of tial architectural

these periods exploited the landscape for agriculture aswell as for animal survives for flocks and herds (sheep/goat, husbandry. Evidence pig, and cattle bones), for agriculture (botanical remains of olive and cereals, aswell as querns and other and for some trade (pottery grinding implements), from elsewhere undertaken. probably domestic

This

in Crete). evidence

The fairly

Small-scale suggests

self-sufficient,

craftwork, including weaving, was that in LM I?III the residents were

depending

on the land for most

of their

economy.

situation

is not an unusual

one. The

of hilly terrain by in the southern Mediterra

settlement

small groups of agriculturalists iswell known nean; it is regarded as a standard type of land use until the Roman period.3 The evidence indicates that low hills were especially favored for exploita tion. The crops that were grown and the strategies that were employed for suited to this type of landscape. of botanical remains found in furnace fragments from Impressions the metallurgy location furnish a little evidence for the crops grown at impressions, Chrysokamino during the Prepalatial period (Chap. 12).These an olive leaf, suggest that olives the of and wheat, preserving shapes barley, growing

them were well

and grains were some of the staple crops. The situation is not surprising, because similar crops have been regarded as normal both for other sites during the Early Minoan period4 and for later periods in this part of the are a We Mediterranean.5 fairly stable agricultural probably observing in which successive groups of people adopted similar practices, to nature the of their landscape and its resources. Moreover, because owing similar crops required similar farming techniques, the land was probably a in the residents of the isolated exploited generally unchanging way by strategy

3. MacKinnon

1998.

4. J. Renfrew 1972; Blitzer 1993. 5. Vickery 1936; Forbes 1995; Halstead 1987,2000.

OF

BOUNDARIES

THE

FARMSTEAD

TERRITORY

235

IIB, and IIB, the cluster of tiny sites of EM III-MM I-III farmhouse. A fundamental question in regard to a farm or an estate such as one envisioned here is the extent of its associated land.What part of an was to local region available such establishment? Such estimates farm in FN-EM

the

LM

the the are

should difficult, but several clues help to suggest possible conclusions. We bear inmind that the territorial boundaries would probably have varied in different periods, depending on the proximity of other inhabited locations, at the main domestic the fluctuating population sites, and several other factors.

Today, the situation is very different from that of the Bronze Age. A metal fence divides a coastal strip, where the hills are mostly bare, from an used for raising olives. extensively cultivated region called Chordakia The

area is on the inland side of Chomatas

cultivated

called the Kambos, public land.

and it is divided

At the microregional is varied. The sea bounds

Hill, above the plain into private plots. The barren coast is

level, the topography on the coast near Kavousi the territory of Chrysokamino and the southwest on both the north and the west. Chomatas has

slope of Chomatas Hill two peaks, but neither one is visible from the archaeologi Chrysokamino a cal sites. The hills southwestern slope is series of lower ridges, with two and uphill from the habitation location (one at E5336 N4934 the other at E5478 N4758). A saddle lies between them. The slopes uphill from theMinoan architecture are mostly phyllite and phyllitic soil, with a few isolated outcrops of dolomite, limestone, and other rocks (see Chap. 2, low hillocks

on the geology of the area). One contact between carbonate is at the north side of the Chrysokamino The

the phyllite farmhouse.

and the

site is on the largest relatively level dolomite outcrop slope. To the east, the terraced land slopes to the Kambos (Figs. 1.6,19.4). To the north and northeast

habitation

on Chomatas

s southwestern

gently down is the hill of Chomatas, of the hills slope down

and farther north are additional bare hills. Many toward the sea until they terminate at ravines and the Aegean, and easy passage by steep trails down these

cliffs overlooking cliffs to the sea is possible at only a few locations. South and southwest of the habitation site, the land has eroded extensively (Fig. 1.10). Its bedrock is dolomite, and the hill slopes down to a low area of red soil named Lakkos Ambeliou. From here, a steep gorge descends to the west all the way to a A cobble road (kalderimi) goes down small harbor named Agriomandra. awalk of much less than an hour, a resident of the habi tation site could reach a small, protected beach, slopes with rich phyllitic soil, bare limestone hills, and rugged ravines. The Chrysokamino for a diversified region provided the potential

the ravine.With

life style. The harbor would have allowed access to the sea with all of its potential for trade, seafaring, and fishing. Soils of two types, terra rossa and decomposed phyllite, could have provided for farming, while barer hills could have been used as grazing lands. Outcrops of dolomite, limestone, and phyllite were suitable for ground stone tools.

sources of building materials

and raw materials

CHAPTER

236

20

TERRITORIAL BOUNDARIES FOR THE CHRYSOKAMINO FARMSTEAD factors may suggest territorial limits for a specific for the correlation of topography, settlement methodology Several

land use has been developed for the study of Roman can be and similar observations applied to the situation Five factors are the most important considerations: 1.The

site. A

careful

patterns, and farms and estates,6 at Chrysokamino.

topography, bedrock types, and soil types presence of natural barriers

2. The

3. The nature and proximity of other sites 4. Information on the type of landscape that was being exploited, as sources of information provided by excavation and other estimate 5. Population Estimates

of boundaries

the above factors. The in relation

considered

at the Chrysokamino estate have been based on and demographic features have been topographic to one another

as well

as to the excavated

evidence

from the three archaeological sites, especially the one with domestic archi tecture. The intensive survey of the region conducted by Haggis provides on the location of nearby sites that would have needed terri of their own.7 tory the factors considered for boundaries, the size of the popula Among tion at Chrysokamino is the most tenuous. Certainly, itwas never large. information

an a Only single farmhouse is likely to have been present in LM I. Even if extended family lived there, one cannot imagine more than 10-15 persons inhabiting the single building, and the number could have been smaller. In III, when the architecture was larger, the number of occupants may have expanded as well. For earlier periods, fewer than ten persons is an educated guess. If slaves were present, as they probably were, one might add to the number.

LM

The

suggested limits for territorial boundaries have several degrees of that represent physiographic Those barriers, such as the sea at certainty. west and gorges and uninhabited bare peaks at the northeast, the north and not be easily crossed. must represent borders that remained They at the east static the long history of the farmstead. Boundaries fairly during and tradition and south are likely to have been established by negotiation is far less certain, even with neighboring sites, so that their exact placement can be proposed. None of the Bronze Age though their general locations would

borders were fortified or even demarcated the archaeological neighbors. General

record, suggesting

in anyway that is recognizable in relations with all contiguous peaceful

have existed as early as the site's foundation a few Final Neolithic sites have been period. Only near from Kavousi. The three main Minoan archaeological boundaries

must

in the Final Neolithic identified

sherds, but the nearest other site near Hagios Antonios, well to that has been recognized is at Alykomouri a a It lies natural the northeast beyond (Fig. 19.3).8 boundary, defined by ravine near the modern gate that provides access to the southern part of locations

at Chrysokamino

all have FN

6. See DalTAglio andMarchetti 1991, with bibliography. 7.Mook andHaggis 1990; Haggis 1992,1995,1996b, 2000,2005. 8. See Haggis

1993a.

BOUNDARIES

Vt

OF

THE

FARMSTEAD

TERRITORY

237

Cave / Church Field House

N

Lime Kiln

A

20.1. Figure dra harbor

Ravine

2*3,

and

Built Road

Agrioman

ravine and the bare cliffs and peaks to the north of it form and northeast of these physiographic barriers, topographic the closest human outpost on the cliffs is a small walled settlement, referred

Chomatas.The

barriers. North

use is not to here as Pyrgos Chrysokaminou (Fig. 1.5). Its earliest date of known, but itmay have served as a lookout post or a refuge site. To the west and north, the sea forms the border of the territory. Except location and the burial cave, no Minoan for the metallurgy archaeological sites have been found

in this direction

between

the habitation

location

and

the sea.At

the southwest, the ravine leading to the harbor of Agriomandra is also a natural border (Fig. 20.1). the sites are located nearby to the south, suggesting No FN-EM and boundary could have extended into the poorer land of the Kambos or to Lakkos Ambeliou of bare hill The southwest of beyond. Chalepa, cannot have supported agriculture and is a natural bar Lakkos Ambeliou, rier itself. It is possible (even likely), of course, that FN and EM sherds near eventually be recognized at additional locations Chrysokamino, our its of territorial scope. refining picture In Chapter 19, Haggis suggests that the settlement pattern at the in EM III-MM I. southwestern spur of Chomatas Hill changed markedly

will

Several new

sites were

founded

near the habitation site, Chrysokamino sites. He makes a strong case that these

a cluster of resulting in nearby sites were probably related, and that they could have shared resources and as the site and the harbor at nearby locations such managed metallurgy at Their strategic location the head of the ravine leading Agriomandra. to the tiny harbor suggests they monitored the tiny port traffic between

CHAPTER

238

20

and places farther inland. Nevertheless, site at the presence of aMinoan the harbor itself suggests the situation could also be more complex. If the cluster functioned as a unit, asHaggis suggests, the northern convincingly and western territorial limits were probably similar to those of the FN is surely indicated, how EM IIB settlement. An expanded population ever, and the southern somewhat

and southeastern

boundaries

could have extended

farther.

To the east,Middle Minoan sites existed at Katsoprinos and Kephalo limnos. Because these MM neighbors must have needed farmland of their the Chryso own, a boundary of some type must have existed between kamino territory and the land controlled by these other sites. Although are not known, the precise boundaries proximity of the other settlements an area at Chrysokamino sites and its vicinity for the cluster of suggests across the upper terraced part of the southwest and southeast slope of Chomatas Hill, including part of Chordakia. The territory prob near the major ravine north of the modern gate. The outpost ably ended that extended

was either an or outlying extension (Pyrgos Chrysokaminou) (more likely) the next neighbor to the north. InMM III-LM I, Haggis suggests that the cluster of sites went out the land of existence, and a situation arose inwhich one site dominated to the north

scape. That site was the Chrysokamino first period with excavated architectural partly preserved. Other communities

habitation remains

is the

location. This

at the site. One

house

is

near here in nearby. The most important site was was later times, Kephalolimnos, already inhabited. A Bronze Age site east of the Chrysokamino habitation still present at Katsoprinos, location, to the east. These LM I settlements must farther and Afykomouri lay have required their own land, so that the territory of the Chrysokamino were

at this period. cannot have included all of Chordakia For LM III, the boundaries were probably as shown in Figure 20.2, with additional land used for pastures. The northern and western boundaries, are not likely to have changed much. established by the geomorphology,

farmstead

Both Kanta9 and Haggis10 have reported LM III pottery from Kephalolim still had the same site as a neighbor nos, so the residents at Chrysokamino to the southeast, suggesting of a boundary somewhere the continuation on Hill. of Chomatas the slopes within the region of Chordakia, remains a question mark. southern border, near Lakkos Ambeliou, some of the site disappeared, sites the MM south of habitation tiny estate may have expanded slightly to include more area the Chrysokamino The

When

near Lakkos Ambeliou, but itwould probably farmland than the cluster had held jointly. The

not have controlled

more

expansion in territory, ab be a natural outgrowth of the had would what cluster controlled, sorbing the increase in the size of the architectural complex at the main remaining

did not also absorb population. site, but only if Kephalolimnos In the diagram in Figure 20.2, the border is drawn at the bare hillside south of AF 32 (an oval enclosure), a region with no sherds suggesting LM I?III a border).11 Farther south, at Lakkos Ambeliou, activity (i.e., it may be

habitation

LM

III and later sherds suggest that activity continued

into the 1st millen

9. Kanta

1980,

pp.

144-145.

10.Haggis 2005, pp. 116-117. 11. For see

Apps.

the G, H.

situation

and

the pottery,

BOUNDARIES

OF

THE

FARMSTEAD

TERRITORY

CHOMATAS HILL

N

Burial Cave

Metallurgy Site

Katsoprinos Habitation

Site

Modern Road

Kephalolimnos Agriomandra

Modern Ravine

Modern

Road

20.2.

Boundaries

for

the

farmstead in LM III

Fences

LAKKOS AMBELIOU

CHALEPA HILL Figure

Road

200m

s habitation site was not inhabited. This Chrysokamino not is in LM included the region, therefore, territory. in By LM I, the estate probably had an eastern border somewhere to Chordakia. The primary territory extended northward the ravine north nium

b.c., when

gate, and it probably included the region west to the sea to somewhere near Lakkos Ambeliou, the boundary where to have varied somewhat from This broad may period period. region had a a one at at the stable the north, west, south and east, but shifting border and southwest. It always included the locations of the metallurgy workshop of the modern and south

and the cave, but both were abandoned before the Late Bronze Age. Evidence from the excavations contributes additional information. Besides

the growing of crops, the residents in LM I and LM III, the evidence Especially stantial amounts of meat, which surely means Access to the bare hillsides south and north of 12. For rather modern

the practice of herding on these hills in than farming see times, 1992, p. 75. Haggis

have

been

essential,

whether

the

pasturelands

pursued

animal husbandry. they consumed sub

suggests flocks and herds were the main were

farming

shared

kept. area would

or owned.12

were also a part of the record, suggest goods archaeological access to a harbor. were almost ing Many imports certainly brought in by Trade

CHAPTER

240

20

sea through rather than overland. Because of its regional Agriomandra, it is that any single community ever had exclusive use importance, unlikely of this type of harbor; as with grazing land, one must consider the possi a shared resource. bility of In summary, the evidence

a for the Minoan period suggests territory in Figure 20.2. Only one long-lived habitation location area. on most is this It found the stable bed recognized within

like that shown has been

rock outcrop in the region (where most of the land is soft phyllite and its semiconsolidated site is surrounded alteration products). The habitation a set at off natural boundaries the northeast, by the by primary territory by sea at the north

and west, and by the proximity of other Minoan the east and south. This primary territory was probably the main region controlled by the estate.

sites at farming

the primary territory are secondary territories that might Beyond have been shared. They include a ravine that terminates at a harbor with a small beach, aswell as bare hills more suitable for grazing than for farming located both north and south of the ravine. Bare hills are also found north of the territory, where they could have been shared with Kephalolimnos, s access to which would also have surely required pastures. Chrysokamino a a for and for primary territory agriculture secondary territory grazing, a small, along with access to the sea, helps to explain how semi-independent unit was able to subsist for more than amillennium and a half in the face of fundamental

changes occurring both locally and throughout the island. stable relationship between land use and population must have been established early in the regions history, based on constant factors such as the nature of the soils, the topography, crop management strategies that did

A

not

contributed deplete the soil, and simple animal husbandry. Metallurgy to the domestic economy until EM III-MMIA, but in later periods, stable factors such as gardening, farming, and herding were more significant in the long run.

CHAPTER

21

Land

Use

on

Chrysokamino

the Farmstead

byPhilip P. Betancourt

communities were the main economic force in the Classical Agricultural era,1 and they would have been at least as important in earlier periods. In to trade and craftwork, Minoan spite of commitments society must have at small sites been firmly based upon an agricultural economy. Especially located away from the large urban centers, subsistence activities would have been focused mainly on growing crops and tending animals. For the territory considered here, strong evidence indicates an agrar ian base. In the first place, accessibility to farmland seems to have been an site important factor in the choice of settlement location. The habitation was neither situated on the coast at the harbor of which Agriomandra, nor farther inland where land traffic seafaring, easier. No major natural resources aside from the land

would

have accommodated

would

have been

itself are present in the vicinity of the site. It was surely the arable land that attracted the residents. Other evidence for agricultural activities comes from the present excavation. This evidence includes the presence of querns for grinding grain, substantial numbers of animal bones from domestic animals (and very few wild species), pottery evidence for long-term storage of commodities, and the identification of chaff, barley, wheat, and olive in the metallurgical furnace chimney fragments. No doubt exists that the residents here engaged in farming and animal husbandry. The interaction between human occupation in this part of Crete

and

the slowly changing conditions of climate, natural resources, and other are crucial factors in the of the environment aspects physical history of at the Bronze Age community The natural Chrysokamino. setting of the is formed of territory Chrysokamino by specific types physical features that can be arranged into separate zones. Because these zones differ activity one from would have afforded different potential another, they markedly to the community as awhole. The contributions dynamic ways their potentials were used or neglected had an important impact on the history of the region. economic inwhich

1. Barker 1985;Halstead 1987; Wells 1992.

The Chrysokamino location provides the setting for a specific way of life based on farming and the keeping of domestic livestock. The pattern of cultural activity must have changed through time, but the need for farmland

CHAPTER

242

21

have been constant.

would manage climate,

the land with and

In all periods, itwould have been necessary to a strategy that was appropriate for the local soils,

crops.

In the methodology for this project, the landscape was developed first divided into different classes. The information used for this analysis came from distinguished relationships

the intensive

by considering between them:

1. Available

survey, and the categories of land were the following three types of data and the

surface

soils and their characteristics,

especially

to support agriculture2 presence or absence of a substantial

2. The

on

the

their potential

scatter ofMinoan

sherds

surface3

or flat; or covered with nature of the topography rocky (steep a substantial soil cover) sediment; with or without

3. The

types of soils are present on the land that makes up the Chrysokamino a territory. A red soil called terra rossa occurs over carbonate bedrock, and

Two

over soils are very clearly phyllitic soil is present phyllite.The by color. Because they also have very different characteristics, distinguished cannot be farmed in the same way. Any successful agricultural system they would have required an understanding of the differences between the two pale-colored

soils and the development of different strategies for their management. was farmed for The fact that the Chrysokamino territory approximately is that the residents their land's 2,000 years ample proof clearly understood were able to manage their soil resources and capabilities successfully. The presence or absence of a substantial scatter of Minoan and later sherds on the surface can be easily documented by intensive surface survey some parts of and In the (see Apps. G, H, I, J). territory, Chrysokamino the landscape are almost empty of sherds, with only an occasional piece m or more apart. Other portions of the land have spaced every 25-50 many ceramic fragments, with more than 10 pieces in a 10 m2 area. This circumstance obviously results from a difference in ancient activity, and it is, therefore, a significant factor in defining classes of land use. is extremely varied. It includes The topography around Chrysokamino areas of steep cliffs, gently rolling hills, a deep ravine, and bare hillsides with almost no soil cover. These different classes of land do not offer the same

must have been exploited in potential for agricultural use, and they different ways. Four classes of land have been recognized based on the three criteria of soil type, presence or absence of substantial numbers of ceramic sherds, and type of terrain. The first three classes are on flat or gently sloping land suitable for farming. Type 1 consists of land with terra rossa soil and a sub sherds on the surface. Type 2 consists of terra stantial scatter of Minoan rossa soil without

the substantial

scatter ofMinoan

sherds. Type 3 consists scatter ofMinoan sherds

of phyllitic soil, which does not have a substantial on the surface at The fourth type consists of land that is Chrysokamino. too steep for either agriculture or the keeping of livestock. These four clas and of the geological sifications, defined initially by simple observations

2. For

discussions

of the

importance

of this factor, see Bintliff 1977, pp. 99 1992, pp. 48-49. of this factor importance for the presence of subsurface features to ac and for its relation human past see et al. 1988, p. 170; tivity, Cherry 100; Carothers 3. On

for

its use

agricultural

the

in the reconstruction practices,

of

seeWilkinson

1982; Bintliff and Snodgrass 1988, pp. 507-508;

Betancourt

and Hope

Simpson 1992 (with substantial addi tional bibliography).

LAND

USE

ON

THE

FARMSTEAD

243

CHOMATAS HILL

Metallurgy Workshop

Katsoprinos

Kephalolimnos

Modern Road

CHALEPA HILL

Figure 21.1. Land of Type 1, used for horticulture substantial

(terra scatter

LAKKOSAMBELIOU

rossa with

of Minoan

200m

sherds)

that distinguish data, have many additional characteristics archaeological them from one another as well. The additional characteristics that are unique to each class help confirm the validity of the organizational system as awhole. A recognition of the differences between the categories leads on the to some significant conclusions and management of organization land at the Chrysokamino

FARMLAND,

TYPE

In the classification soil with

farmstead.

1: GARDENS

used in this study, land of Type 1 consists of terra rossa scatter ofMinoan sherds on its surface. The land of

a substantial

Type 1 is fairly restricted in extent (Fig. 21.1). A sloping hillside extends from the habitation site, which has an elevation of 120 masl, to the oval structure

32 (App. G), which has an elevation of 106 masl. rises slightly to a ridgelike hill with an elevation the land 32, of 110 masl and then descends in a long, relatively evenly sloping hillside all the way to Lakkos Ambeliou (elevation 75 masl). The land of Type 1 is an area south and southeast of the habitation site, extending downhill called AF

South of AF

from the architecture

to just over the ridge south of AF

32. The

roughly

CHAPTER

244

21

circular deposit of sherds extends across an area with awidth of ca. 100 m and a length of ca. 110 m. Within this area, the sherds are distributed as small, worn, pottery on the surface fairly evenly, occurring fragments of the ground. The area is too large for the sherds to have been dumped here and later scattered. The soil (red terra rossa) is different from the covers the settlement, so the material cannot have pale brown soil that hill eroded down the from the habitation and the site, simply ridge with a same conclusion because erosion of 110 masl demonstrates the height across the land shows that the flow uphill. The even distribution sherds did not spread out from any single location. The long hillside south of AF 32 has the same terra rossa soil, but it has no sherds on it until it It is not a part of this class of reaches the vicinity of Lakkos Ambeliou. cannot

landscape.

Another

thick scatter of sherds occurs at Lakkos Ambeliou.

This

fea

ca. 280 m ture is a pedon (a deposit of sediment) in a sinkhole measuring m east-west. Morris north-south studied this deposit in detail.4 by 410 He concluded that it developed as an alluvial deposit from the terra rossa on the surrounding glacial period. The

in a lake during the last hills, possibly accumulating was of surface the pedon dry by the Middle Minoan sherds were deposited on it.

period, when Many Minoan

sherds occur both within

the soil and on the surface of

site and across the ridge south the land of Type 1 between the habitation are III period, but of the farmhouse. Most of them from the Late Minoan and Late Minoan I pieces are present aswell.5 The earliest Middle Minoan sherd recognized

from the area is from the Final Neolithic.

Several

sherds

in the present study as typical examples of the material catalogued found here (see AF 32 in App. G). They are all similar in their physical in that they are small (mostly under 5 cm in size), worn, and preservation, are

on them, and they do not join with poorly preserved. Little paint survives or of their other pieces to make larger fragments parts of vessels. Most are

edges

rounded.

character of the pottery differs from the better-preserved sherds found at the settlement, but it is identical to sherds found away from sites. In comparison with the sherds from the Chrysokamino settlement The

site, the sherds from Site AF 32 and its vicinity are smaller and come from the same Minoan periods. In terms although they the number of sherds per square meter was much lower at Site 32 than at the settlement before excavation. Isolated sherds of this type

habitation

more worn, of quantity, AF

been called "offsite scatter."6 They are commonly found in surveys in eastern Crete and elsewhere. the soil. Morris found Mi small, worn sherds also occur within

have sometimes regional The

to a depth of 110 cm at nearby Lakkos Ambeliou. He that the buried sherds were not on an ancient, buried ground surface, and he suggested that they might have moved downward through vertical cracks in the ground.7 This conclusion may be correct for some of noan

sherds buried

demonstrated

the sherds in the sinkhole he studied, but it is not likely to be correct for all of them because he also identified a buried tillage zone extending from the surface to ca. 30 cm below the surface.

4. Morris 5. For

2002, pp. 45-59. their previous recognition,

seeHaggis 2005, p. 119, locus 53. 6. Bintliff and Snodgrass 1988; and Mantzourani Davis, chap. 3. 7.Morris 2002, pp. 57-59.

Cherry,

1991,

LAND

USE

ON

THE

FARMSTEAD

245

that such vertical movement usually takes place inmont Considering as not in kaolinite morillonite those of this sinkhole, some such clays, clays of these sherds may have been deposited through the practice of manur not In is the sinkhole the region of AF 32 where the within any case, ing. sherds must be explained in terms presence of a larger number ofMinoan of Bronze Age activity. manure to land used for was aworldwide Adding agriculture practice for agriculturalists before the introduction of commercial fertilizers.8 The value of manuring has been studied inmany regions, from Great Britain,9 to continental Europe,10 to North America,11 to Asia.12 The practice im proves the land in several ways. The working of manure and other refuse into the soil renews nutrients, adds organic matter to help retain moisture, aerates the soil, distributes nutrients more uniformly, and breaks up clumps soil to improve root penetration. The practice has often been as a normal part of gardening and farming. regarded The manuring of fields has also been previously documented archaeo

of denser

in the ancient Mediterranean logically.13 It must have been widespread because Greek and Roman writers described it as a routine practice.14 In on farming, for example, Cato regarded it as an essential part of agriculture (deAgri Cultura. 5.8,29.1). Given its broad distribution it would be surprising if elsewhere, was not also practiced in Bronze we now have manuring Age Crete, and on Pseira. The excavation firm proof of its use during theMinoan period of two agricultural fields on Pseira Island discovered sherds both on the his manual

surface of the fields and deeply buried within the soil.15The Pseiran field at location G 2 is particularly relevant to the situation at Chrysokamino. The two regions are in the same part of Crete, the fields have contemporary to Late Minoan are both Middle Minoan deposits, and the sites farming land composed of terra rossa soil. using poor quality, marginal in the Pseiran agricultural field, which was located well away from any settlement, were found within the soil all the way from the surface to bedrock. They were small and worn like their counterparts from AF 32 locations

Sherds

8.

Lyon

and Buckman

1929,

pp. 381-405; Wilkinson 1982; Courty, Goldberg, andMacphail 1989, pp.

133-134.

9. Pitty 1978, pp. 132,135. 10. Barker

p. 163.

1985,

11.Wines

1985, chap. 1. et al. 1964, pp. 264 12. Tamhane

269;Walls 1982, p. 26. 13.Wilkinson 1982,1989; Bintliff and Snodgrass 1988, pp. 507-508; Bintliff, Howard, and Snodgrass 1999; Bintliff et al. 2002, pp. 260-261. 14.Tilly 1973, p. 73; Garnsey 1992, p. 151. 15. Betancourt

and Hope

Simpson

1992. 16. Bull,

1999.

Evershed,

and Betancourt

at Chrysokamino.

Small amounts of other refuse occurring with them, sea and shells, scraps of obsidian, broken loomweights, including bones, other items, helped indicate that the material on this and other Pseiran terraces was settlement debris. Scientific a team of scientists at analysis by the University of Bristol has now confirmed that chemical traces of manure are present

in the soil from context G 2 on Pseira. There

can be no doubt,

of the deposit.16 the terra rossa at the location

about the formation

therefore, ' on Pseira, near the As Chrysokamino site is rather poor terra rossa soil. It has few nutrients, and it habitation has an even greater problem in its inability to retain much water. If this or over a soil was used for horticulture agriculture long period of time, a fertilizer would have helped restore nutrients, and it would have added retention. organic matter for better moisture Analysis of the soils contributes additional evidence that the terra rossa in this region was used for agrarian purposes. Human influence on soils can be determined by identifying the presence of specific residual chemical

CHAPTER

246

21

the natural surface horizon of elements, especially phosphorus.17 Although soils in this region averages only 15 cm in depth,18 Morris demonstrated that the red soils from Lakkos Ambeliou have a 20-30 cm deep surface zone

and other chemical relics of agricultural ac containing phosphorus as a tillage zone, and the presence tivity.19 This horizon may be identified of substantial numbers ofMinoan sherds shows that this part of the land, 32, was used in the Bronze Age. in sherd density is the most variation

not far from AF The

of the difference between interpretation 1 and Type Types 2-3. At Chrysokamino,

important

factor

in the

the Chrysokamino farmland of the extensive sherd scatter of

pottery exists only on a restricted part of the terra rossa soil. An occasional sherd occurs on the other two types of land in the Chrysokamino

Minoan

territory, but the contrast in the amount of pottery is sufficient to indicate a in land use. The explanation cannot be found in difference pronounced the type of soil because some of the terra rossa (i.e.,Type 2 farmland) does not have the extensive

sherd scatter.

of the land of Type 1 further help to separate it, albeit less dramatically, from the other two types of farmland at Chryso kamino. All of the Type 1 land is close to the settlement. In fact, it adjoins Other

characteristics

the habitation site, with no physical barrier in between, while parts of the other two farmland types are up to a kilometer away. It is restricted in size, in contrast with the larger territory covered by the other two types of land. It is close to a deep natural cleft in the bedrock, which would have been a no terraces because it is not very steep. source of water in rainy periods. It has use was clearly indicates that local land different for various parts of the landscape. This conclusion fits well with a use in ancient growing body of archaeological evidence for zonation in land and more recent periods inGreece.20 The land available for agriculture was even we no though usually have apparently often divided between crops, The

evidence

at Chrysokamino

in agrarian clear picture of which crops were on which plots. Variations on and on have been based situations21 practices recognized topographic the suitability of land for grazing and other uses.22 That selected crops are of intense care (hoeing, weeding, watering, and grown under conditions can also be manuring) recognized from the presence of certain weeds.23 is best understood The use of different classes of land at Chrysokamino in relation to research on ancient

strategies for raising various types of crops. An excellent summary of the subject is presented by Leach, who notes that a large body of evidence (linguistic, topographic, archaeological, literary, and ethnological, among others) indicates a very different ancient attitude toward gardens and fields.24 Several Roman writers, especially Cato and Varro, discuss these different attitudes.25 The different strategies arose as a natural consequence of the types of crops grown; some plants (such as as and many vegetables) require constant care, while others (such pulses not. and olive do trees) grains for the two classes of crops were very different in Classi Strategies times. Crops in the fields were handled as large units in terms of tilling, planting, and harvesting, while plants in the gardens were tended individually. Gardens were routinely placed near the settlements where they could receive intensive care, while fields could be located farther cal and Roman

17. Sjoberg 1976; Proudfoot 1976; Eidt 1977. 18.Timpsonl992,p. 19. Morris 2002,

137. pp. 52,59.

20.Wagstaff andAugustson 1982; Hayden 1995, p. 100;Acheson 1997, p. 175;Halstead 2000, p. Ill; Ather don

2000,

p. 63.

21.Wagstaff

andAugustson

1982,

p. 107.

22. Gamble 1982, pp. 163-164; Forbes 1995;Halstead 2000. 23. Jones 1987,1992; Jones et al. 1999;

Bogaard

et al. 2000;

2000. 24. Leach 1997. 25. Tilly 1973, p. 64.

Jones

et al.

LAND

USE

ON

THE

FARMSTEAD

247

in Homer and in Linear B tablets suggest that gardens away. References fruit trees, and grape vines were situated near the settle for vegetables, ments.26 The basic contrast in attitudes is even reflected in the linguistic origins of the words horticulture, associated with the home, and agriculture, derived from a root meaning wilder areas at some distance from the house. The situation typically resulted in a division of labor along gender lines. Women typically cared for the gardens, which required small amounts of labor on a daily basis. Men were more likely to work in the fields, away from the houses. The fields required longer periods of intensive work at certain stages of the growing cycle, as in the times of pruning, planting, and harvesting. It is this difference helps explain

in the amount of care needed

the differences

in the number

for some plants that of sherds on the land around

It is likely that the Bronze Age residents of this farmstead some crops that required intensive care, with the addition of

Chrysokamino. were growing substantial amounts

of manure, while to grow with much use of manure and rubbish

could be allowed Greater

they also raised grains and olives that less hoeing, weeding, and manuring. as fertilizer in the near the gardens

settlement would have resulted in the higher density of sherds recorded in the archaeological record. A combination of poor soil and plants requiring care than other crops is extra nutrients, more moisture, and more frequent the best explanation for the use of farmland of Type 1. There are additional reasons for the placement of a garden near the home. Many vegetables are routinely harvested in tiny amounts every day or two to ensure freshness, making frequent trips to the garden necessary. are stored, in contrast to crops such small amounts of vegetables Only as grains and olives that are harvested in large quantities all at once and stored until the next harvest. Planting garden plots close to home means that short trips to the garden to water sprouts, remove weeds, pick fresh or add a small amount of refuse as fertilizer can be produce, interspersed with other household chores. such as olives, barley, and wheat, all of which are attested at Early as as elsewhere in EM Crete,27 do Chrysokamino (Chap. 12), well

Plants Minoan

can be not require constant care. a settle They planted in fields away from In addition, many of the tasks associated with their management, including pruning trees and harvesting grain, require full days of very hard labor. Such tasks are not well suited to household members who are also or taking care of very small children. cooking in land use is clearly reflected at Chrysokamino. This distinction Land

ment.

of Type

1 contrasts with

to the distinction

land of Types 2 and 3, conforming between gardens and fields. Land of Type

very closely 1 adjoins the no natural barriers such as ra

site and is easily accessible with It has many small sherds on its surface, indicating that this land received more attention than the land of Types 2 or 3, a situation very compatible with garden plots. A number of crops were probably grown in the gardens. A strong case has been made for pulses as a staple in the Minoan and Cycladic diet.28 habitation

vines to be crossed.

26.

See

the discussion

in Palmer

1999, pp. 478-480. 27. J. Renfrew 1972; Greig and Warren

1974.

28. Sarpaki 1992.

If this hypothesis is correct, it confirms the need for horticulture. Pulses have to have been raised in gardens near the settlements because the

would

CHAPTER

248

21

cultivation of these plants is extremely labor intensive.29 Grapes may also care in have been grown. Viticulture requires extra pruning, staking, fertil on and other and thrive well-drained tasks,30 grapes izing, slopes. The land here as garden plots would have been well suited to this crop. this pattern, with gardens for vegetables aswell as dry farm Although for olives and grains, has not been previously noted in the Minoan ing record, itwas probably not unusual. Diversity has long been archaeological interpreted

as an

important aspect of Bronze Age farming,31 and different between gardens and fields different crops require strategies. Distinctions were in made later of antiquity in both Greece and Italy. routinely periods no seems reason to suggest that Crete was There substantially different regarded

from other places where the same crops were raised. The ancient Mediter ranean people had limited strategies for solving the problems of agricultural on

marginal land. A division between gardens and fields would not have been an unusual situation, but rather an ordinary aspect probably of Minoan land use. production

FARMLAND, TYPE 2: GRAZING LANDS of terra rossa soil without

sherds any substantial scatter of Minoan constitute farmland of Type 2. This category of land is farther away from site than the land of Type 1. It includes low hills where the the habitation

Areas

a as well as the Kambos where the terra underlying bedrock is carbonate rossa has been extensively deposited. This class of farmland is shown in Figure 21.2. Chalepa, eroded. Many

the hill south of the ravine leading to the harbor, is extremely parts of the hillside are composed of boulders and bedrock

rather than soil. The hill does not have any terrace walls could not be used very successfully for farming in its present, of the soil has washed into the valley of the Kambos Most an

Ambeliou,

event

that

seems

to have

taken

place

in two

on it, and it eroded state. near Lakkos stages,

which

are not suggests that much of the first precisely dated. Research byMorris erosion occurred of during the last glacial period, when probably period was a lake.32 The presence of Middle the sinkhole at Lakkos Ambeliou sherds on the surface here proves the final stage was earlier than and the resulting erosion have often been re this period. Deforestation as processes in the Aegean,33 though there are strong garded anthropogenic the universal opinions against application of this hypothesis.34 It is tempting

Minoan

to associate

the last period of erosion with clearing of the land for agriculture for copper smelting, because such operations

Minoan

to deforestation smelting often leads no direct evidence for the association the

two

events

to Early the Final Neolithic or with the use of the region

30. Hanson

in this case except for the date when

In spite of their bare condition, the hill of Chalepa and other carbon ate hills in the vicinity are used today (1990s to early 2000s) as grazing them land for sheep and goats. They support sparse vegetation, making even in late summer. No

cattle are kept

in this vicinity.

1985,

p. 73.

1992.

31. A. C. Renfrew 1972, p. 306; Palmer

require substantial fuel, and and subsequent erosion, but there is

occurred.

suitable for grazing

29. Barker

1999,

32. Morris

pp. 478-480. 2002, p. 58.

33. Bottema 1980;Thirgood pp. 68,153-155;

van Andel,

1981,

Runnels,

and Pope 1986; van Andel, Zangger, and Demitrack nels,

Run 1990; Jameson, and van Andel 1994, p. 325.

34. Rackham

p. 127;Moody 2000b.

and Moody

1996,

1997,2000; Forbes

LAND

USE

ON

THE

FARMSTEAD

249

N

Katsoprinos

Agriomandra

Kephalolimnos

Modern Road Ravine Road

LAKKOS AMBELIOU

Figure 21.2. Land of Type 2, used rossa with

for pasture

(terra

face

or with

sherds

a

light

no scatter

sur

CHALEPA HILL

of

sherds) Additional

hills where

Lakkos Ambeliou, north of the main bare of vegetation, It is suggested for the ancient

the bedrock

is a carbonate

lie between AF

32 and

and also north of the ravine that leads toAgriomandra, are mostly Chrysokamino region. All of these places but they support enough for sheep and goats. here that these hills would

farmsteads

west

of modern

have afforded grazing lands Kavousi, including Chryso is not clear, and the absence of

Early Bronze Age situation site could suggest either that little meat animal bones from the metallurgy or that this was in the diet Bronze did not the Early workshop during Age consume much meat while the were In other words, the working. personnel site may have been too specialized to give a true picture of diet in general; kamino. The

may have returned to their homes at night. By the Late period, however, there is substantial evidence for sheep and/or goats in the LM I?III strata at the habitation site, indicating that meat formed an in important part of the diet of the people who lived at the farmhouse the metalworkers

Minoan

that period. Land animals

35. Gamble Forbes

1982,

pp.

163-164;

1995.

36. Halstead

1981,1987,2000.

that

they

of Type must

have

2 would

have provided

nearby grazing

for the

managed.

near Specialized grazing sites agricultural communities have also been recognized from later periods of Greece.35 Animal husbandry, therefore, in which may often have been more complex than simple transhumance, flocks and herds are moved from upland pastures in the summer to lower grazing lands in the winter.36 It has been suggested that keeping animals locally

was very important

to small farmsteads, with

agricultural

sites being

CHAPTER

250

21

of their proximity to grazing lands as well as farmland.37 us an additional reason to suppose that at least some animals gives were managed locally at Chrysokamino. chosen because

This

If the situation

in the Kambos

before

the middle

of the 20th

cen

tury A.D.

is any indication of ancient times, the low-lying parts of the as plain may have suffered from droughts that made them undesirable even could have flocks and herds, farming areas, though they supported summer. This situation would leave the terraced in the particularly phyllite hills

(farmland of Type 3) for raising field crops. A similar pattern of land on the upper with animals maintained use, slopes and cereals and other on lower hillsides, has also been proposed for other parts of crops raised eastern

Crete.38

FARMLAND,

TYPE

3: FIELDS

is steadily accumulating in Early Minoan for olive cultivation refutes earlier doubts about its presence,39 Crete. Blitzer convincingly must olive that the have been a significant crop. It was probably arguing also present in the Final Neolithic, although evidence has not yet been discovered for this early phase. Evidence

must also have been important. Impressions of chaff, barley awheat grain, and an olive leaf in the furnace grains, fragments from the was site that the show region Chrysokamino already engaged metallurgy in diversified agriculture at the end of the Early Bronze Age (Chap. 12). Grains

the impressions in pottery are primarily chaff, it is likely that grain amounts. Grain and olives may even have been planted in substantial on the same farmland, with the gathering of chaff resulting in the planted casual inclusion of olive leaves.

Because was

and grains are crops that can withstand long periods without more land their cultivation is often called rain; dry farming. They require area of than the small farmland of Type 1, and the land of Type 2 is too Olives

poor to support them. The large area of Type 3, consisting of the most fertile land in the Chrysokamino territory, must have been the mainstay of dry farming, if for no other reason than that none of the other available as a semi land was suitable. The farm could not have been maintained was exploited. independent unit unless the phyllitic soil Land of Type 3 is shown in Figure 21.3. It consists of areas inwhich the bedrock is phyllite, with the overlying soil being formed by its disinte terra rossa of the region, gration. This soil is substantially richer than the and it retains moisture

better. The

manure

of the region still characteristics These purposes.40 scatter of sherds associated with

modern

to the red soil for agricultural prefer it may explain why it lacks the extensive

residents

and other refuse. In this region, an average of one sherd or less is an area of 1,000 m2. A tiny settlement with limited refuse

present within

to use its manure on the nearby gardens where itwas most might prefer needed, rather than on the fields where crops grew well. site adjoins land of The area of Type 3 is substantial. The habitation some at of 3 much of the land lies distance from the but Type Type 1,

37. Forbes 1995. 38. Hayden 1995, p. 100. 39. Blitzer 1993, p. 165; cf. Runnels and Hansen

40. Haggis p. 378.

1986,

pp. 305-306.

1996a, p. 188; 1996b,

LAND

USE

ON

THE

FARMSTEAD

251

Metallurgy Site

Katsoprinos

Kephalolimnos

Figure 21.3. Land of Type 3, used for dry farming (phyllitic soil)

LAKKOSAMBELIOU

Several areas in this part of the landscape have well-preserved terrace walls. They show that much more of this general region was farmed in the past than today.41 Terraced areas are shown in Figure 16.3. settlement.

Terracing tory.42

The

has been a regular practice

practice,

however,

was

not

in many

universal,

and

periods it may

his

of Greek have

been

based

on local factors, such as the desired crop, the type of soil, and the degree of slope. In northeastern Crete, agricultural terraces have been excavated to one another, revealing very different at two sites in close proximity were built terraces during the Bronze Age,43 practices. On Pseira, many at Chrysokamino after the only excavated terrace was constructed the Roman period (App. K). Rackham andMoody suggest three types of Cretan terraces.44 Parallel terraces. at the ends create braided terraces. Switchbacks walls form step

while

Small, often semicircular terrace walls create pocket terraces. All three types a fourth type, in which occur at the walls Chrysokamino (App. G), and are built inside a ravine to prevent it from eroding and becoming larger, is as well (App. G, AF 30 and AF 31). present It is not possible to prove that any of these terrace walls areMinoan from the available data. The only terrace excavated (App. G, AF 22b) was sherds recorded on the land Occasional Minoan Venetian or Ottoman. to local residents, 41. the According here not last farming was abandoned War II. long after World 42. Moody

and Grove

1990;

Foxhall

1996; French andWhitelaw 1999; Rackham andMoody 1992; 1996, pp. 140-145; Krahtopoulou 1997; Frederick

and Krahtopoulou

2000.

43. Betancourt

and Hope

IX, p. 287. 44. Rackham and Moody

Simpson

1992; Pseira

1992.

CHAPTER

252

21

in the Bronze Age, although a sherd from the surface does not prove that a terrace was first built in the Bronze Age. The terrace itwas constructed. Even may simply have incorporated earlier soil when the of included though phyllite would have released gradual disintegration indicate

itwas used

gradually into the soil, occasional additions of refuse from animal and settlement refuse piles must have occurred. pens In most places, the terraces are irregular and somewhat informal, and a series of even, they do not create regularly spaced steps. This type of most has been considered terracing appropriate for olive trees,45 and it is nutrients

as well. At the likely that grains would have been grown under the trees no survives direct evidence for theMinoan farmstead, crops Chrysokamino grown on the land that is now terraced. One

of the most

to have been water

in terrace systems is likely important consideration rather than the creation of level land. Ter conservation

races of the "check dam" class, inwhich successive terraces are built on the are a standard method a occur slope of hill, throughout the territory. They of preserving moisture. small buildings found in this terraced zone all seem to have been constructed after the Bronze Age (App. G). Their late date suggests that theMinoan farmers walked out to their fields from the settlement, taking their tools with them. A similar situation exists at nearby Pseira, where all or later.46The the field houses and threshing floors are Byzantine gradual The

drying of the climate in this part of the Aegean47 may have brought about the need for more terrace walls to retain moisture. This problem may not have been so acute until after theMinoan

period. Many of the slopes of the hills are not very steep, and crops could have been grown on them without substantial terracing.

crops did the residents of the estate grow? The triad of grain, olives, and grapes has been regarded as the foundation of Aegean agriculture for many years.48 Pulses were probably staples as well.49 Several pieces of evidence suggest that the main grain grown on poor land in this part of Crete was barley rather than wheat. The molds of cereals preserved in the What

furnace chimney fragments aremostly two-row barley (Chap. 12). Similarly, atMyrtos, the molds in clay have also been identified as barley,50 and this even at Knossos.51 grain appears to have been stored in quantity Barley on it and would have been a logical land than wheat better does, grows poor choice for the main cereal to be grown on the hillsides at Chrysokamino. is lacking for other crops, but they surely existed. A strong Evidence case has been made for diversification in the small outlying farms of Mi seems to be the best noan Crete,52 and this explanation for meager theory The use of land of more than evidence that exists from Chrysokamino. one type, the presence of more than one species of domestic animal, and the relatively small scale of the operation all argue for a diversified agri cultural base. 45. Lohmann 46. Betancourt

49. Sarpaki 1992.

1992. and Hope

1992; Pseira IX, pp. 296-300. 47.

See Chap.

2, pp.

48. Vickery 1936.

19-21.

Simpson

50. J. Renfrew 1972. 51. Evans 1921-1935, 52. Halstead

1992.

vol.

4, p. 622.

LAND

USE

ON

THE

FARMSTEAD

253

CHOMATAS HILL

Metallurgy Site

Figure unsuitable

21.4.

or areas Steep rocky HILL for farming

X

CHALEPA

RAVINES,

LAKKOS

CLIFFS,

AMBELIOU

AND

OTHER

STEEP AREAS

If the land in the vicinity of Chrysokamino is divided into zones, the loca tions that are too steep for farming make up a very small percentage of the total (Fig. 21.4). They consist only of the sea cliffs, some rocky outcrops, of the ravines that di steep areas on the hillside, and a few ravines. Most are not vide hillsides with phyllitic soil very steep, and some of them were terraced

in

post-Minoan

times

to

conserve

water

and

provide

spaces

for

are farming. Numbers AF 30 and AF 31 (App. G) examples of ravines with walls running up and down the slope to prevent the gorges from becoming wider through erosion, with subsequent loss of farmland. Cliffs and very steep slopes characterize the land at the edge of the sea at this on the coast. of the cliffs are too dangerous even point Many for goats, and the modern shepherd tries to keep his animals away from them (one animal was lost over the cliff in 1996). The steepest gorge is the one that forms the natural barrier at the southwest of the territory. Some at the north of this gorge, bordering on the Chrysokamino are sheer cliffs. In this are not territory, study, the hectares of steep land area in estimates of the total land counted available for agriculture and

of the bluffs

animal husbandry. On the seacoast, continuous. Numerous

the steep cliffs at the edges of the hills are not steep paths lead down to the sea, and residents

CHAPTER

254

21

use them to get down to the sea to fish as a routinely ease leisure time activity. The with which the paths are used suggests that the appearance of isolation from the sea is deceptive. The Minoans prob

of modern

Kavousi

ably walked down to the coast fairly often marine invertebrates as well as to fish.

to

gather

sea urchins

and other

gorges were probably once forested, as many ravines are in this ismost likely a result of recent part of Crete. Their present bare condition of their is stunted from foraging by goats). (most overgrazing vegetation The

In the Bronze Age, they could have furnished brush and wood for fuel and other purposes, small amounts of vegetation for animals, and ha vens for the wild birds, hares, and other creatures that were occasionally resources were hunted in all periods. These extremely important.53 Hare bones are present in the faunal remains from both the farmhouse and the site, suggesting metallurgy the local meat diet.

occasional

hunting

or

trapping

to augment

LAND OUTSIDE THE MAIN TERRITORY: AGRIOMANDRA ravine that leads to the harbor of Agriomandra is one of the most in areas in the access route from the Kambos teresting region (Fig. 20.1).The a area consists of from the of Lakkos Ambeliou steep ravine descending

The

narrow cobbled road (kalderimi), now with some parts to the importance of the route. The road may be testifies away, or the Venetian from the Byzantine period, because it provides originally a a access to small church built into cave part way down the route, but some of the better-preserved sections may be from the Ottoman period. The to the harbor. A

eroded

to Hagios church, on the north side of the gorge, is dedicated a entrance is vault, and the interior is covered with concrete The (opus incertum). Beyond the church, the road continues it reaches a tiny beach called Agriomandra.

Ioannis.54 and stone

to descend down the gorge until The small harbor was used as

a seaport by some of the residents of Kavousi until the middle of the 20th century.55 Because its small beach faces west, it offers some protection from the Bronze Age, the beach must have and eastern winds. During sea level in this part of Crete.56 the because of lower relative larger a series of farms and villages in The harbor would have accommodated and Pacheia Ammos. the territory between the two larger ports atTholos northern

been

It was

too surely important to have been used exclusively by any single small settlement. One assumes that those who lived nearby took care of and the harbor's maintenance. reports Minoan pottery from MM Haggis

LM worn

from this beach

53. Burford 1993, pp. 111-112. 54. Faure 1964, p. 58; 1979, p. 58; 1989, p. 244. 55. Haggis 1992, pp. 169-170; 2005, p. 118. 56. Flemming 1972; Flemming,

cannot know

Czartoryska,

I from a small site at the south side of the harbor,57 and he also records sherds from the beach.58 A piece of water-tumbled malachite was

in 1996 (App. C). if the inhabitants of sites like Chrysokamino had direct or indirect contact with the seafaring that must have entered Crete collected One

at the small harbor, but the presence of imported goods at allMinoan sites indicates that access to the sea was a simple and constant part ofMinoan

and Hunter

1973;

Piraz

zoli et al. 1982; Pirazzoli 1988. 57. Haggis 1992, pp. 169-170; 2005,

p. 118,

locus

45.

58. Haggis 2005, p. 118, locus 110.

LAND

USE

FARMSTEAD

THE

ON

255

the best route to a harbor was life. For Chrysokamino, the ravine to Agriomandra.

the easy walk down

at the for a major seafaring component scarce. Neither bones from is, however, deep percentages of foreign pottery testify to marine

evidence

Archaeological

farmhouse

Chrysokamino water fish nor significant seas contribution pursuits. The Late Bronze Age when

may have been modest, especially in the the evidence for all activities at the site is most

complete.

For earlier periods, when the metallurgy workshop was active, the evi are present, and lead iso ismore extensive. A few Cycladic imports ore some at of used at the smelting site that least the studies tope suggest came from Lavrion and Kythnos.59 Certainly the region benefited from

dence

its proximity to the sea and its access to a small local harbor where and small ships could land.

boats

LAND OUTSIDE THE MAIN TERRITORY: THE KAMBOS large valley called the Kambos extends inland from the beach atTholos, to the south dividing the coastal strip of hills from the higher mountains (Fig. 19.3). The valley, drained by the Platys River, extends all the way from that runs the beach to the northern terminus of the break in the mountains

A

to south across the Isthmus of Ierapetra. It adjoins the hills of is one of its extensions. and Chalepa, and Lakkos Ambeliou terra rossa eroded from nearby The soils of the Kambos are mostly hills. They have been studied at several locations near Kavousi.60 These from north

Chomatas

soils are subject to droughts the phyllitic soils.

because

they tend to retain less moisture

than

The Kambos was planted with olive trees in the years following World IL As a part of his survey of the region, Haggis interviewed several residents of the areawho could recall when the plain was not planted. Itwas

War

as undesirable land, and people preferred to do their farming on regarded the phyllite and dolomite slopes flanking it to the north and south.61 The digging of deep wells made water available for irrigation and transformed agriculture in this part of Crete,62 making the olive tree the most profitable crop and leading to the planting of the entire plain in olive groves, as illus trated in Figure 19.4. This

situation

is typical of low-lying plains and river valleys in the before the middle of the 20th century. In many

southern Mediterranean 59. Stos-Gale see also App.

1998,

pp. 720-721;

C.

60.Timpson 1992;Morris 1994, 2002. 61. Haggis 1996a, p. 188; 1996b, p. 378.

62. Allbaugh 1953, pp. 258-263. 63. See general 1985, p. 57.

comments

by Barker

64. Tilly 1973, pp. 45-46,156.

cases, only modern technology has allowed this type of land to be exploited successfully for agriculture.63 Even in places where the soil is better than at Kavousi, the seasonal variation in large plains ismore extreme than in in rainy periods and too hilly terrain, and they tend to be water-logged dry at other times of the year. The Roman writer Varro noted the unsuit use of low hills for ability of lowland plains to farming.64 He advised the uses. It is the with reserved for different farming instead, plains being use of the Kambos more probable, therefore, that in the Bronze Age the closely

resembled

its use in the period before World War

II than that of

CHAPTER

256 the present. The modern Kavousi an economically

wells

and irrigation that have made viable region65 have led to a major

21

the plain at changes in

agricultural practice. Instead of farming, low plains were traditionally used for winter pas tures. They collect abundant water from the runoff of the hills, and herds and flocks are mobile enough to exploit the resulting vegetation easily and to avoid the marshes.66 The

to be used in places where upland plains tend can also be the summer, with the flocks and pastures exploited during a in pattern of seasonal transhumance. If the situation herds being moved at Kavousi and other to the 1940s is any indication places in Crete67 prior of the role of the plain in antiquity, itwould have been used primarily for winter pasture in connection with transhumance rather than for growing crops. The animals would have been moved to higher summer, when droughts would have severely affected soils of the plain.

ground during the the dry terra rossa

it occurred in this region during the Bronze Age, or at the Bronze have either intensified the end of may Early Age begun when new settlements opened up much fresh territory in the mountains south of Kavousi (Chap. 19). An expansion of sites near Lakkos Ambeliou took place at the same time, and the new pattern may signal an increased sources. population needing additional food mountains south of Kavousi would the Kambos and the Exploiting If transhumance

local have required new cooperation and possibly outside direction. These not be regarded as isolated phenomena. Crete experienced should changes at the onset of the Proto profound economic and political transformations a involving the exploitation of larger palatial period. A local development economic is consistent the visible elsewhere with patterns region changing IA,68 and it surely played a in Crete. Gournia expanded in EM III-MM role in overseeing or safeguarding any greater regional system that provided a good. It cannot be coincidence that the pottery tempered with rocks in the granodiorite/diorite series, the hallmark of the Gour amounts at Chrysokamino at this time. in nia production, appeared large for the common

A regional transformation, associated with the expanding power of Gournia, seems to fit the evidence from Chrysokamino best.69 of extensive grazing lands began, it was Whenever the exploitation Bronze in the Late firmly place during Age. Substantial evidence for ani mal husbandry the habitation

was site. A

recovered

from LM

levels in the excavation of economic of subsistence, one that I?III

complex pattern included some arrangement for large flocks and herds, was definitely of the Late Bronze Age agricultural system at Chrysokamino.

a part

65. Allbaugh 1953, pp. 258-263. 66. Barker

1985,

67. Blitzer

2004,

p. 57. pp. 122-124.

68. Hall 1904-1905. 69. For changes without

an alternative

managed the need

see Forbes

2000a.

under

model, local

for outside

with control

direction,

CHAPTER

22

Survey

Conclusions

byPhilip P. Betancourt

of the Chrysokamino territory adapted to several changes area in northeastern Crete. The this small their of long occupation during first known inhabitants settled here at the end of the Stone Age. They found an area suitable to their needs and established permanent residence. Some

The

inhabitants

of the territory's features, such as the proximity of the soils, remained stable over an extended

of the sea and the nature

period. Other aspects of the environment, however, such as climate and vegetation, did not remain constant. The territory's relations with the other settlements on Crete also

changes during the regions long history. experienced many deep-seated The residents of Chrysokamino responded to changing conditions, but the strictures of limited local resources. their choices were made within Some aspects of the local economy, such as the copper smelting workshop, a successful stage, and then slowly, reached developed passed completely out of existence; others, such as agricultural practices, lasted much longer. of Chrysokamino shows that cultural change, as others have on is based suggested, usually complex forces of formation and dissolution rather than on a strictly linear, gradual development.1

The

history

FINAL

NEOLITHIC

TO EM

III-MM

IA

a In the Final Neolithic period, probably well before 3000 b.c., few settlers In western Asia , the arrived on the coastal strip known as Chrysokamino. and copper tools were already period would be called the Chalcolithic, new settlers founded at being made and disbursed in small quantities. The least one settlement, and at some time within the period, they either estab lished or allowed others to establish a small metallurgical installation for ore. were in Their dead the Theriospelio smelting copper probably buried cave. The

1. See

Sutton

1994.

residents of the area were farmers, and the evidence they left in the archaeological record indicates their choice for a place to live was dictated more by the availability of low hills where crops could be grown and animals could be grazed than by easy access to the sea or to inland tracts of virgin forest.

CHAPTER

258

sites have also been recognized at several other places and/or cemeteries have been Settlements

Final Neolithic

of Mirabello.2

the Gulf

around

22

noted at Kavousi,3 Mochlos,4 Pseira,5 Sphoungaras,6 Pacheia Ammos,7 Vasi liki,8 Vrokastro,9 and elsewhere.10 The pattern suggests that the settlement was part of a that brought numerous of Chrysokamino general migration residents to this part of Crete. was neither the most desirable nor the worst location Chrysokamino It was not easily visible from the sea for a small band of agriculturalists. one could walk an if forces hostile approached), but advantage (perhaps fairly easily. An outcrop of dolomite provided both to build. construction material for houses and a stable bedrock on which land was available for both gardens and fields, and a cleft in Agricultural site was almost certainly a source of south of the habitation the dolomite water in the rainy season. Rain was more abundant than in later times, and the success of the settlement proves the location was favorable. The to the water

down

established a stable pattern of land management residents of Chrysokamino a small group of next their that allowed people for the territory to support two millennia. to settle in a location

that would support diversified agri to placing the habitation site centrally on successfully, opposed a view one that Neolithic traditional class of soil, raises questions about only a on was based grains.11 No real evidence for such restrictive only agriculture one cannot prove Cretan Neolithic farming practice exists, and although the point, the first residents at Chrysokamino probably required land for in addition to cereals. The almonds, fruit trees, legumes, and vegetables site to good garden soil as well as to land for proximity of the habitation field crops was surely not accidental (for Final Neolithic pottery on the see App. H, AF 32.1). garden area, on small, diversification, semi-independent especially Agricultural one an was asset. in Failure farmsteads, crop from disease, drought, or other factors could be offset by success in another. A varied diet is both healthier and more pleasant than a restricted one. Small labor forces can because different crops are planted and be maximized with diversification harvested at different times of the year, allowing the labor force to move The

choice

as

culture

one

from

crop

to

another.

In

regard

to

some

storage,

crops

complement

or vinegar, made by process (for example, can be used as preservatives for more easily spoiled ing olives and grapes, are suited for different crops. Diversification allows soils Different harvests). the use of otherwise unusable land (for example, planting fruit trees in gul lies or placing grains in the underused spaces that are necessary between one another

2. Betancourt

3.Haggis 17/92.

olive oil and wine

1999.

1992, pp. 173-174, locus

1912, p. 93. Seager and Davaras 5. Betancourt

4.

1990;

Pseira VI and VII. 6.Hall 1912, pp. 46-48; Foster 1978,

p. 4, nos.

2-3;

Betancourt

1983,

nos. 110-113. p. 46, 7. M. Tsipopoulou 8. Seager 1904-1905,

(pers.

comm.).

p. 212.

9.Hayden 2003a, 2003b. 10.Haggis 1992, pp. 53-56; Betan court

1999.

11. A. 304-307.

C. Renfrew

1972,

pp. 280,

CONCLUSIONS

SURVEY

259

trees). All of these factors suggest the likelihood that several crops raised at Chrysokamino during the entire history of the territory. differs from the one in central The situation near the Gulf ofMirabello

olive were

in some important ways. Knossos had a substantial population much levels were extensive.12 Neolithic earlier than eastern Crete, and itsNeolithic Knossos covered an area of at least 4.5 hectares,13 and in comparison with town or even other Neolithic sites, itmust be regarded as a large, wealthy a small were levels at Phaistos also substantial,14 aswere city. The Neolithic in southern Crete.15 other early settlements Crete

a by the Late Neolithic period central Crete had had large near and for the the southern millennia, living population population eastern shores of the Gulf of Mirabello before the Final Neolithic period Although

have been much smaller, as it has left few traces. In fact, no large town from any of the Neolithic periods has been found in eastern Crete, area seems to have consisted the occupation and in the Gulf ofMirabello

must

only of tiny sites spread through the region until the late Early Bronze of places like Gournia.16 The small size and the scattered Age development distribution

of the eastern

in comparison development of the island. part

have played a role in their slower the more urban situation in the central

sites must with

Large towns developed in eastern Crete by the end of the Early Bronze to substantial growth at two places Age. Evidence from excavations attests that would later be important: Mochlos17 and Gournia.18 But Chrysoka mino, like most of the local settlements, remained tiny. The contrast with central Crete in terms of population distribution would have been an use for a district with important factor in local history. The system of land to hamlets would have have been different from that of tiny substantially one dominated centers. urban by large Settlement patterns appear to have remained stable during the Early three archaeological Bronze Age. At Chrysokamino, locations have sub stantial evidence for human social/economic/religious activity (Fig. 1.4): 1.A habitation 2. A metallurgy

(App. G, AF 29) workshop (App. G, AF 23) location

3.A cave (App.G, AF 34) Final Neolithic The

discussion

and Early Minoan pottery in Chapter 20 demonstrates

three FN-EM made

12. Evans

1921-1935,

vol.

1, pp. 32

55; Furness 1953; Evans 1964,1971, 1994;Hood and Smyth 1981, p. 6. 13. Evans 1964, p. 132; Broodbank 1992,

pp. 40-41.

from all three places. for this

that the boundaries

as a

were single topographic district, probably formed barriers to the north and west, and the proximity sites to the east and south, along with several other factors. The

territory, functioning by natural topographic of other

comes

locations are very different

separate contributions

types of sites, and they must have to the inhabitants of the territory.

14.Vagnetti 1975. 15.Vagnetti 1973. 16. For the large size of EM III MM IA Gournia, seeHall 1904-1905. 17. Seager 1912; Branigan 1991,

pp. 101-104; Soles 1992; Soles and Davaras 1994, pp. 394-396; 1996, pp. 178-180; Mochlos IA;Mochlos IB. 18.Hall 1904-1905; Soles 1992; Watrous

and Blitzer

1999.

26o

CHAPTER

Main

The

Habitation

22

Site

habitation site's setting is directly related to the geomorphol on a is It hard, stable, dolomite outcrop, at the highest point on the ogy. is visible at southwest slope of the hill of Chomatas where the dolomite the surface. The dolomite provides ample amounts of building material for stone architecture, as well as a stable and relatively flat foundation on

The main

to build.

In fact, this bedrock outcrop is the only place within the are two to conditions such a present Chrysokamino territory where these great degree. The location is the only logical choice as a site for substantial stone and mudbrick architecture because the rest of the territory is covered

which

terra rossa or unstable decomposed phyllite, with a irregular outcrops of few other stones.

with

only small and highly

site also has other good characteristics. It is near the entrance to to the best harbor in the area; it overlooks the

The

the ravine that leads down

at Lakkos Ambeliou, through soft soil down

where water

could be reached easily by and it has ample land table, digging It is for and animals. also high enough to raising crops nearby grazing receive good breezes, alleviating the heat of summer, and it is relatively far that would have troubled people living closer away from the mosquitoes depression

to the water

areas in the Kambos.

to marshy

show Pottery sherds found under ttie floors of the later farmhouse was and Early that the spot already settled during the Final Neolithic Bronze Age. The excavation of the farmhouse has produced significant pottery from these early periods, and the site was it is likely that the factor in the territory, although probably controlling other tiny sites lie buried in the vicinity. The residents of the area used for burial, and they must have had a relation to the cave at Theriospelio amounts

of domestic

the metallurgy workshop. The three locations are so close together that It is significant that Final Neolithic they must have been interrelated. pottery is the earliest pottery at all three places because it implies that the territory

was

already established

as a unit during

this early phase

of its

settlement.

in the Final Neolithic, its establishment the pattern of land stable enough that no major change seems to have taken place until EM III, when several small sites were founded near the previous set tlement. This long period of stability must have resulted from a successful After

use was

balance between support The

the size of the population

and the capacity of the land to

it.

in the furnace and olive impressions presence of barley, wheat, was shows that the from workshop agriculture chimneys metallurgy in the region during the Early Bronze Age. An already being practiced seems to have on based mixed economy farming and animal husbandry In this region, the prox been normal for early Aegean communities.19 sea allowed access to seafaring as well, and an outlet to the imity to the sea would have been a crucial factor in the other important element of on im installation depending the early economy here, a metallurgical ores. ported

19. A.

C. Renfrew

1972,

pp. 304

307; Halstead 1981, pp. 318-320, 326.

SURVEY

CONCLUSIONS

Metallurgy

The

261

Workshop

the main

site habitation site, the choice of location for the metallurgy is on a saddle also be based on topographic factors. The workshop between higher outcrops of bedrock with a cliff at the north. The saddle

Like

must

acts as a funnel that channels

the northern winds

and intensifies

them, so

needed a natural draft to aid in the smelting process, that if the metallurgy this spot is the windiest location on the southwest slope of Chomatas. A second factor may be the proximity of the sea. A path leads down to the and in calm weather

water,

If both

access

ships could anchor and unload cargo and pas to the sea and windy conditions were the factors

sengers. the saddle is the most suitable location necessary for the pyrotechnology, in the vicinity. A third factor could be the availability of fuel. The territory must have had an original forest cover when the Final Neolithic settlers arrived, and metallurgy iron

ores

were

used

needs

substantial

fuel. It is also possible

that local

as fluxes.

The metallurgy site is too tiny to have been a complete settlement. The presence of one flimsy hut less than 4 m long does not imply a permanent or even a self-sufficient are scarce, and community. Artifacts they do not include the full range of objects (such as cooking pots and storage vessels) one expects from a permanent settlement. The workshop must have been an

some other site nearby. appendage of at this isolated promon The presence of a highly specialized workshop access to specialized tech that early tory calls into question suggestions was

nology

in locations

concentrated

that were

centers of socioeconomic

a number of at Chrysokamino power.20 The workshop probably supplied sites with copper, but there is no reason to assume that this technology was established

because

the choice

is more

of any elite presence within the region itself. Instead, on wind conditions, to have been made based likely

access to the sea, availability of fuel, and proximity of markets. topography, an If elite segment of the population controlled the workshop, which may was well have been the control exercised from a distance. true, very

The

Burial

Cave

at

Theriospelio

site in the territory is a natural cave with stalactites and such a cavern is a significant and highly specialized stalagmites. Obviously, cave is found in Plattenkalk The feature. limestone. Although topographic

The

third major

its entrance one

can

stand

is fairly small, the space inside is over 40 meters up

inside

the main

in length, and

room.

for a settlement is adjacent to the cavern. Its entrance is situated on a steep hill, so that construction of houses would have been extensive terracing. No traces of houses or terraces are impossible without No

evidence

present at the location, for habitation.

E.g.,

Carter

1998.

isolated from any spot used

on the surface inside the cave clearly visible (in the disturbed soil either excavated by Edith Hall in 1910 or by others), the cave is interpreted here as a place for burial. The earliest pottery from Because

20.

and the cave appears

human bones

are

CHAPTER

262

22

the location is from the Final Neolithic burials in period. Final Neolithic this part of Crete were associated with a wide variety of tomb designs, caves, and small built chambers.21 Some including natural caves, manmade more one times than tomb type exists in the same cemetery.22 Choices for tomb design were clearly based on social factors about which we have little information. The Theriospelio cave, the only known natural cavern was in the Chrysokamino territory, apparently the social choice for the local community's burials. burial caves in eastern Crete were regularly used until much later cave was also used into III.23 It is likely that the Theriospelio excavation we cannot know the the Middle Bronze Age, but without Other

than EM details.

Comments the Early Bronze Age, the residents of the Chrysokamino terri During use are made of three of If the different site. three locations tory types as a an considered lived at the habita agrarian community single unit, tion site (and possibly elsewhere), buried its dead nearby, and engaged in periodic craftwork including metallurgy. The evidence suggests that the one locations at Chrysokamino archaeological complemented unit probably based another, interacting to form a single socioeconomic at the excavated habitation location. This unit seems to have been one of three main

a series of small settlements

established

in the Kavousi

region at the end

of the Neolithic.24 Very little evidence survives for the Early Minoan agricultural products that must have been the foundation of this stable system. Impressions of site grain and chaff in the furnace chimney fragments from the metallurgy was raised, and the one wheat impression of (Chap. 12) show that barley seed hints at wheat production. An olive leaf gathered up with the chaff that tempered the chimneys suggests that grain and olive trees were grow one another. ing in close proximity to Organic residues in sherds from the site show that wine, flavored with many herbs and spices, was metallurgy so grapes were were already present (App. M), being raised. Grains already in the Cretan Neolithic.25 the Minoan grown By period, the farm Early ers of Crete raised wheat, barley, and oats along with olives26 and several other crops.27 As at these other sites, the agriculture probably diversified. The Neolithic

was

at Chrysokamino

economy was supplemented by the smelting of copper. For Final and Early Minoan I, copper is known from only amodest amount

in this part of Crete,28 but the evidence is accumulating for other places (see Chap. 13). The EM I-IIA cemetery atHagia Photia shows on Cretan culture an that the metal was already making impact early in of evidence

21. Betancourt 22. For Mochlos, Soles 1992.

2003a, see

pp.

123-128. 1912;

Seager see Soles For Gournia, see Pseira VI and For Pseira, 1992.

VII.

Money-Coutts 1970a,

Pendlebury,

and

Branigan

Davaras

p. 387; Schlager andDollhofer 24. For

discussion

of some

1989a,

1998. of the

see 1992, pp. 53-56. Haggis 25. Evans 1968, p. 269.

others, 23. Pendlebury,

pp.

1935-1936; 153-154;

26. J. Renfrew 1972; Blitzer 1993. 27. Greig and Warren 1974. see 28. For Mochlos, 1912, Seager vol. 1, pp. 58, 1921-1935, p. 93; Evans 68.

SURVEY

CONCLUSIONS

263

isotope patterns observed in the copper and Kythnos,30 suggesting the ore was patterns was pre brought in by ship. The workshop produced only copper, which to neighboring sites to be remelted in crucibles and cast sumably taken into various objects. the 3rd millennium

b.c.29 Lead

slag match

from Lavrion

local smelting operation was only one step in the making of metal of copper from ore involved the activities of a objects. The production series of skilled specialists in addition to the smelters working at Chryso The

the prospectors who found the ore, the miners who recovered it from the ground, the workers who beneficiated the ore and packed it for crews who provided transportation, shipment, the ship captains and their

kamino:

the final products. To set up the operation, the resources to obtain good quality copper ore, arrange for its shipping, mine the rock to be used as flux, establish a run secure by craftsmen familiar with the smelting process, and workshop to for It markets the would also have been obtain copper. necessary potential and the metalworkers someone

needed

who made

to advance

furnaces fuel, food for the workers, and raw materials for manufacturing and chimneys. The implications of this complicated production system are very im our for of Cretan craft Such portant understanding early organization. over an extended disparate activities, carried out period of time inmultiple of a series of different persons rather locations, imply the involvement than a small workforce performing every task. One cannot expect that ship a captains would also have been prospectors, miners, and smelters. Such and supervision. The complex endeavor would have required coordination ore did not get transported to the right port or loaded on the right ship the shipment from mine automatically. Someone must have accompanied to destination to in order guarantee its arrival. Behind such an operation was

someone with

to

confidence

in its successful

conclusion

and the resources

a series of tasks whose

profits would only be realized at the same manager might also have found it neces

accomplish end of the process. That sary to oversee the intermediate

level supervisors, and perhaps to take as for armed guards or soldiers) to make sure (such arranging that the potential profits were not stolen somewhere along the way. Such an operation would be fully at home in a bureaucracy like the one re flected on the Linear B tablets. The implication is that some of the super visory roles that played such an important part in later Cretan administra tion had a long history inMinoan Crete, emerging well before the Late Bronze Age. some action

was probably not unique. Both Chrysokamino locally and elsewhere in Crete, successful communities would have needed to develop agricultural systems that were viable within the constraints of locally available land, to secure access to the sea or other modes of transport, and to exploit local 29. Davaras Betancourt sion

2004;

and 1971; Davaras see also the discus

in 13. Chap. 30. Stos-Gale 1998,

see also F. App. 31. E.g., Blackman 1977,

pp. 69-71.

pp. 720-721; and Branigan

and regional

resources.

of the agricultural base with crafts Augmentation as in the at of a metallurgical manufacturing, development workshop as a to have been undertaken natural response the Chrysokamino, might leisure time afforded by seasonal farming activities. Patterns of settlement or

and land use recognized at other locations in Crete31 are not incompatible the more detailed conclusions presented here.

with

CHAPTER

264

22

in this long period occurred at the The culmination of developments end of the Early Bronze Age. EM III-MMIA is defined in eastern Crete the of East Cretan White-on-Dark The ceramic style Ware.32 presence by begins at the end of EM IIB with over Vasiliki Ware surfaces.33 After

the application the destruction

of simple white level coinciding

lines with

Ware became more popular, and its IIB,White-on-Dark to of motifs include range expanded spirals, circle motifs, quirks, and other on or reddish brown designs. It used white linear designs uniformly dark of without of the characteristic the surfaces, any mottling previous period. The phase ended with destructions throughout the region, identifiable in the end of EM

the archaeological record atMochlos,34 Pseira,35 and Gournia.36 The final MM IA at Knossos.37 of the with stage phase overlapped This period is an important one in Crete. Several new developments can be and elsewhere on the island. The assigned to it at Chrysokamino at Knossos substantial deposits that survive suggest this was a prosperous time in north

In southern Crete, several new sites can be as expansion in population there well.39 Important

central Crete.38

recognized, suggesting from this period

an

is also found atMalia.40 Gournia, a site that would into an important center by LM I, expanded significantly. The large IA pottery at Gournia41 contrasts dramatically of EM III-MM

pottery

develop amount

the tiny amount of Vasiliki Ware from the previous phase.42 The period was a stable one in central and eastern Crete, and itwas not

with

by major destructions until its end. The beginning of polychrome on pottery, which should mark the boundary between EM III and painting MM that was only used on a small IA,43 is actually a rare characteristic number of vases in this period, even at Knossos.44 The red pigment was marked

not used during this phase the Gulf of Mirabello.

at the eastern

in the pottery workshops

side of

a new settlement pattern inwhich At Chrysokamino, Haggis has noted a cluster of sites site replaced the single location of the .main habitation are new sites from the of found downhill Most these original (Chap. 19). in the settlement, at the edge of Lakkos Ambeliou. Other new foundations south of Kavousi may signal advances into the higher country of herds and flocks. seasonal movements inland, perhaps to accommodate and an exploitation The new pattern suggests an expansion in population of new resources for intensified animal husbandry. This expansion coin mountains

the main period of the metallurgy workshop, which would have timber for fuel, also available in the higher hills. for other This pattern of scattered settlements has been proposed a as must of have been the result well.45 It parts of Crete larger population cides with

needed

needing

more

food. By spreading

32. Betancourt

the population

36. Hall 1904-1905,1908.

1984.

33. Boyd 1904-1905, pl. 25, no. 2; Warren 1983,

1972,

phase p. 69, no. 236.

34. Seager

1909,

p. 278;

Soles

178-180.

37. Warren court

II; Betancourt

Davaras 1994, pp. 394-396; 1996, 35. Banou

pp.

19-20.

1965,

1967-1968;

1924,

pp. 25-26;

40. Demargne Betan

Levi

for Drakones, pp. 76-79.

Chapouthier,

1945,

Demargne,

pp. 1-12; and Dessene

1962. 41. Hall 1904-1905,1908.

2003b.

38.Momiglianol991. 39. For Patrikies,

and

756; 1995a,

over the land, agricul

see Bonacasa 1976, vol. 1, pp. 747 see Xanthoudides

42. Hawes 43. Hood

et al. 1908, 1971,

p. 39.

44. Momigliano 1983. 45. Kanta

1991,

no. pi. 6, p. 219.

1.

CONCLUSIONS

SURVEY

265

to inhabit tural production might have been increased,46 but decisions on are also to at have been least locations partly based likely dispersed cultural preferences. system of important characteristic of the Early toMiddle Minoan character. The sites were land use in this region is its open and undefended no traces of fortification walls. This situation is very tiny and exposed, with different from the one in several other parts of Crete, where walled towns An

and citadels,

fortification walls, suggest often with megalithic communities were fearful of attack.47 Elsewhere

that small

in eastern independent an attest to defense.48 road stations fortified Crete, may organized regional The relative safety of the region (perhaps imposed by a higher authority or elsewhere?) must have been a crucial factor at Gournia allowing the dispersal of the population. All of this activity implies

that new social and economic

factors were

the local area. Several

affecting

between

tionship For the Middle

theories have been suggested for the rela in Crete. this part of the island and higher authorities a case for a several have made Minoan authors period,

that included this region and was centered at Malia.49 An palace-state IA was alternative center was much closer at hand, however. EM III-MM a major period of expansion at Gournia,50 and this expansion coincided with the spread of the typical Gournia pottery tempered with rock in the series.51 granodiorite/diorite in EM III pottery was the majority fabric at Chrysokamino cave IA. The small workshop, the habitation and the all used site, and other ceramic products that had the same White-on-Dark Ware This

MM

same fabric as Gournia, Vasiliki, Pseira, Priniatikos Pyrgos, style and the and several other sites in this part of Crete.52 The ceramics had a regional a series of other ceramic classes distribution, replacing including Vasiliki and other burnished

Ware53

pottery. at the transition

the Early and the between developments Bronze Age involved cooperation between sites, indicating a greater awareness of settlement patterns changed. The regional goals. Traditional ore from abroad and metallurgy workshop expanded greatly, importing its Several

Middle

sending its copper to locations in the inland mountains were

away from the workshop. New foundations surely derived from the coastal population, have shared communal goals. A new, large, and widely dis

and they must tributed class of pottery replaced the Vasiliki Ware of EM IIB. Some of these new regional initiatives would seem to imply a directing organization. The authority must have had strong social (and perhaps religious) power, because there are no local fortifications, in sharp contrast with other parts of

46. versus

See

comments

intensive

on extensive

agriculture

by Halstead

1992. 47. Watrous

1982,

pp. 42-43,

62

63; Brown 1993, p. 65; Nowicki 1996a; 1996b, pp. 35-42; Schlager 1999. 48.Tzedakis

et al. 1989;Tzedakis

eastern

Crete.54

et al. 1990.

52. Betancourt

1984.

49. Cadogan 1995; Knappe? 1999. 50. Hall 1904-1905,1908; Betan

53. Betancourt

1979.

court

1984.

51. Haggis andMook fabric

2.

1993, p. 273,

54. Alexiou 1979,1980;Tzedakis et al. 1989; Tzedakis

et al. 1990; Man

ning 1994; Nowicki 1999,2002.

1996a; Schlager

266

CHAPTER

22

cluster of small sites probably managed a territorial region that was not too different from the one used earlier, except that itmust have been The

at the south, the Kambos, would have been larger at the south. The land in itwould have been attractive as pastureland, the but rainy season, marshy use was if its of the inland hills especially complemented by exploitation for transhumance. At the same time, the patterns of sherds on the surface, with

their evidence an

to for manuring, indicate that agriculture continued role in local subsistence. The division between gardens

important play and fields suggested in Chapter 21 was probably already in place, and the area site (Fig. 21.1) would have been used for adjoining the habitation care. and other vegetables, vines, plants requiring periodic at the end of the Early Bronze The evidence suggests that this period Bronze Age was an affluent time, of the Middle Age and the beginning with population growth, expanded craftwork and manufacturing (including at the metallurgy workshop), and the advent of new ways to exploit the into the hills beyond the coast and the larger landscape. The expansion use of this for transhumance, timber, and other resources, territory probable to significant advances in the region's economy. would have contributed If the stability of the earlier system implies a balance between subsistence the expansion of flocks and herds involving transhumance and population, on a scale may suggest new surpluses, either for local trade or, more larger a likely, to support rapidly expanding

MIDDLE

MINOAN

state centered

at Gournia.

IB-IIB

affected the influences must have increasingly IA, outside a new at follows destruction of The major territory Chrysokamino. phase were of when White-on-Dark Ware Gournia, deposited large quantities north of the town.55 New pottery styles, now using the potter's wheel more

After MM

IB. than previously,56 signal the beginning ofMM at and Phaistos this time, and magnificent Malia, Knossos, expanded were into a palatial class built. The polychrome pottery developed palaces called Kamares Ware,57 and the other arts were encouraged aswell. Cretan on many fronts. Palatial art objects had little society and economy advanced or no

on life in the territory, but the rise of palatial Chrysokamino impact culture must have resulted in a new economic situation for all of Crete. at the end of EM III The metallurgy workshop was abandoned the ultimate cause, was part of the MM IA. Its abandonment, whatever in Crete at this time. broader social and economic changes happening the founding of theMiddle Minoan palaces, metals may have been a smelted or available, or more centrally controlled, imported in condition. In this new era, there was no room for the small independent

With

more

workshop

II, metals were much more available By MM the island than they had been during the Early Bronze Age,

at Chrysokamino.

throughout and competition

from these new sources, surely involving imports from eastern ports ,meant that the older ways of obtaining metal would never on now again be viable.58 Chrysokamino depended primarily agriculture and the tending of animals for its subsistence.

55. Hall 1904-1905,1908; court

56. Warren 1985,

Betan

1984. 1980,

p. 492;

pp. 77-78.

57.Walberg 58. Betancourt

1976. 1998.

Betancourt

SURVEY

The

CONCLUSIONS

Site

267

Clusters

Minoan territory and its period, the Chrysokamino a to the south had of sites cluster small that included the adjoining region main habitation site aswell as some tiny locations near Lakkos Ambeliou small establishments persisted until the end of the (Chap. 19). These the Middle

During

phase. This scattered settlement pattern seems to have been of the period throughout the Kavousi region.59 The individual house typical sites are not contiguous, but they are so close together that a relationship seems between them, possibly kin-based, likely. They probably managed the land as a group, with cooperation for large projects affecting the whole

MM

IB-IIB

was well located both for region. The cluster farming access to the the seaport at Agriomandra.

and for controlling

Comments in the Chrysokamino IB-IIB developments region coincide with the growth of the palaces. The more centralized economic system and its success in of the obtaining imported copper surely led to the abandonment new use local metallurgical installation. The and the greater system of land

MM

exploitation of the land at Lakkos Ambeliou by the cluster of small sites a to have been the demise of local activities, may response manufacturing forcing an intensification of the agricultural efforts. Itmight also have been to a larger related to a need to produce a surplus for contribution political unit, one that was based at Kavousi, Gournia, or even Malia. Several authors have discussed the theory of a large palatial state evidence from the Malia hiero by Malia.60 The administrative aswell as the size and tablets,61 glyphic importance of theMiddle Minoan Malia palace, leaves little doubt that the site was supported by a substan tial hinterland, but the boundaries of that hinterland have been difficult governed

59. Haggis 2005. 60. Poursat 1987; Cadogan 1990, 1995; Knappe? 1999. 61. Poursat

1990.

62. Knappe? 1999, 63. For discussions,

pp. 632-634. see Pendlebury, and Money-Courts 1935

Pendlebury, 1936,

p. 62; Coldstream

and Huxley

[1972] 1973, pp. 94-95; Walberg 1976, pp. 150-151, form 47; 1983, pp. 190 forms

192,234-241,

mata);

no. 11 pi. 2, and Silverman

Betancourt

(Gournia); 1991, wicki

234-241.

et al. 1908,

64. Hawes

No p. 25, no. 412 (Gournia); no. 5 34, 13, (Katali p. fig. no. 8657 1945, 33, pi. Demargne

2002,

(Malia); Knappe? 1999, p. 631, fig. 19 (Malia);

Betancourt

p. 145, nos. BR

and Davaras

28-30

(Pseira);

1999, Knap

pe? 1999, p. 629, fig. 12 (PyrgosMyr tou); Seager 1904-1905 (Vasiliki); Betancourt

(Vasiliki).

1983,

p. 274,

nos.

257-260

to establish

because of lack of clear evidence. In the case of pottery, for one cannot know if connections also imply political control. The example, case for a site under the control ofMiddle Minoan Malia most compelling a town on the southern coast. For eastern may be found at Pyrgos Myrtou, Crete,

the situation from Malia

pottery somewhere

is less clear. Although many of the parallels between are apparent in vases made ofMirabello

and the Gulf

in the Gulf area, imports from the Phaistos area also occur at and elsewhere within the proposed Malia state, and some imports itself,62 suggesting that pottery was appear in the hinterland but not atMalia a not simply sent from a subsidiary political unit to central one. Evidently, across moved it difficult to pottery routinely political boundaries, making use it for those boundaries. defining

Malia

MM IIB ended with widespread in eastern Crete. The destructions use of the thin-walled final phase was characterized the carinated by eastern In the to of from Malia this cup Crete, Palaikastro, parts cup.63 no over often has white paint its dark slip, and it is decorated only with the cup grooves on the upper part. Although disappears hundreds of MM IIB have excavated. been The cup III, examples to the Gulf of Mirabello.64 is present throughout the region from Malia

thin horizontal inMM

268

CHAPTER

22

IIB in the Isthmus of Ierapetra that the end of MM a crisis period when people fled to high refuge sites, in coincided with Katalimata.65 The use cluding the easily defensible height of Monastiraki

Nowicki

has noted

of such an inaccessible the widespread

refuge site must architectural destructions

surely have been correlated with in the region. The disruption is

was buried with its contents,66 especially evident atVasiliki, where House A and at Pseira, where the town was destroyed and the cemetery abandoned.67 The extensive rebuilding after the period of the carinated cup indicates, occurred throughout the region. The most however, that the discontinuity important site with destructions is clearly visible the destruction

at this period is the palace inQuartier Mu.68

site atMalia;

It is tempting to associate the fundamental changes at the end of this period with the expansion of Knossos into eastern Crete. At Chrysokamino, is evident has been found, a transformation of the pattern of land use requiring a cluster of sites and the survival of only a single settlement, the habitation site built on the

where

no MM

architecture

in the abandonment

dolomite the MM

outcrop at Katsoprinos. A new house was constructed IIB period, and it survived until LM IB or early LM

MIDDLE MINOAN

III-LATE MINOAN

here after II.

I

the beginning of the Neopalatial III-LM I, the cluster period, inMM of sites southwest of the hill of Chomatas was replaced by a more nucle

At

pattern. For the remainder of the Bronze Age, until the was abandoned in LM IIIB, the habitation site was the complex only settlement within the Chrysokamino territory. The LM I farmhouse at Chrysokamino must have controlled the whole territory of Chrysoka ated settlement

domestic

mino, although the absence of sherds in an area south of AF 32 suggests that the territory included the land in Figure 20.2, but not the region south sites observed by Haggis south of Chrysokamino of it. Several tiny MM farmhouse, the vil (Fig. 19.8) must now have joined the Chrysokamino or some other settlement. The consolidation of the of lage Katsoprinos, a new sites into of land scattered use, system larger settlements represents and itmay imply a new bureaucratic management system organized under palatial control. is not com farmhouse The LM I architecture at the Chrysokamino seems a to it of have consisted single building. It pletely preserved, but rooms. Additional had an entrance on the east and a series of contiguous interior spaces were at the north fortifications were present. it seems

(destroyed

to have been

by the later architecture).

No 65. Nowicki 2002.

an

domestic independent complex Although at Chrysokamino LM I the situated in theMinoan countryside, building cannot be regarded as a villa or "country house" as defined by van Effenterre, of LM I, are large, and others.69 Minoan villas, a phenomenon Cadogan, architectural units with rich possessions, large amounts of stor that imitate the palaces.70 The architectural details and age, sophisticated imitations of the palaces are, as far as we know, completely architectural

unfortified

66. Betancourt

1977,

p. 346.

67. Pseira VI and VIL 68. Poursat

1996.

69. Cadogan 1971,1997; Hood 1983; Nixon 1987; Betancourt and Marinatos terre

1997;

Effenterre

1997.

70.Moody

1987b.

and Effen

CONCLUSIONS

SURVEY

269

at Chrysokamino. In LM I, the incomplete architecture leaves but the from this finest many questions unanswered, possessions period, stone vases and bronzes, are certainly not imitations of objects including found in the palaces, although they argue for amore complex social station isolated farmstead. Reused wedges for drilling stone than an impoverished, missing

vases (found in LM III levels) suggest an earlier industrial component may even have produced a surplus beyond its own needs.

that

Niemeier

has convincingly villa system of argued that the Minoan I had an earlier social and political history, and that it did not appear of any type.71 Before large manorial vil suddenly without predecessors las could exist in the Minoan imitating palatial architecture countryside,

LM

of peaceful and genteel, even luxurious, living, and creating microcosms the agrarian base of the rural countryside had to develop over a period of many years. The LM I farmhouse at Chrysokamino may represent a late survival of the type of small, semi-independent farmstead that helped craftwork traditions that develop the farming practices and small-scale of larger villages, towns, and splendid villas. would lead to the development was founded in the Final Neolithic That Chrysokamino and continued to III is a testament to the stability of the system of it used, but it is also proof of the limits of its immedi more ate landholdings, which were never able to support anything grand. the development of complex political systems inMinoan Whether Crete exist in LM

I and LM

land management

is regarded as evolution or revolution,72 the Chrysokamino domestic unit did not participate much in these dynamic developments, except as a recipi a stable ent of outside influences and directions. The residents developed pattern many

early in the regions other sites did.

history, but the farmhouse

did not evolve as

is not unique. During MM and LM I, the island of Chrysokamino a number of small, isolated domestic buildings Crete had that occupied the landscape between the larger villages and towns.73 Excavated examples in eastern Crete include Chalinomouri,74 Karoumes A,76 Chamaizi,75 are and Kokkino Phroudi.78 Unexcavated farmhouses Cheiromandres,77 also known Photia.79 Greece, larger

from other parts of Crete; six are found in the region of Hagia situations seem to have existed elsewhere in prehistoric

Similar

including estates

the mainland,

occupied

the

where

areas

both small, independent

between

urban

complexes.80

the nearest of these excavated Chalinomouri, close Like especially parallel for Chrysokamino.81 as well as in craftwork engaged farming, operating domestic complex that was less semi-independent Like Chrysokamino, 71. Niemeier

it survived

73. See Bevan

2002,

and Davaras

?g. 17;Mochlos IA. 75. Davaras

1973.

affords an itwas Chrysokamino,

farmhouses,

in the countryside as a a complex than village.

III. 76. Chryssoulaki 1999, pi. 7:?.

1997.

72. Branigan 1970a, p. 204; Cherry 1983. 74. Soles

into LM

farms and

p. 224, 1996,

fig. 5. p. 209,

77.

Chryssoulaki

1999,

pi.

7:ce.

78. Chryssoulaki 1999, pi. 7:y. 79. Tsipopoulou lou 1997, p. 206. 80. Palmer 2001, 81. Mochlos

and Papacostopou pp. 52-65. 103-132.

IA, pp.

CHAPTER

270

22

sites such as Chrysokamino and Chalinomouri do not fit some of the land use systems into Minoan for Crete. proposed perfectly as sites has that with isolated such MacGillivray suggested buildings Isolated

same category as as belong to the villages such Tylissos and Vasiliki,82 but the variability in the scale of these sites implies serious cultural differences. The Minoan settlement pattern was complex and multitiered.83 It probably encompassed many different types of habitation

Chalinomouri

sites, including isolated tiny houses, towns, and cities ruled by palaces. Several of the Cretan farmsteads

larger farms, villas, hamlets,

villages,

seem to have had defensive capabili to their agrarian functions. Pyrgos is an Chrysokaminou small fort at the top of the highest cliff east of Chrysokamino, unexcavated with a commanding view of the sea (Fig. 1.5). The small, isolated domestic that survived into LM I were all units that combined farming buildings ties in addition

a strategic location. Chiromandres, Kokkino Phroudi, and Karoumes seem to have been defensible farmhouses set on theMinoan road system;

with A

they have been

regarded

as

guard

are

stations for the roads.84 Chaimaizi

and

forts that dominate

their surroundings. hilltop Pyrgos Chrysokaminou is set so that it controls a small landing place for ships at Chalinomouri domestic unit is Plain. The Chrysokamino the eastern end of theMochlos the upper end of the ravine that also set at a strategic location overlooking sea. access to Its situation is better suited to controls the Kambos from the and control of the region than the nearby tiny sites that were one III toMM II, and this commanding present position could be of the reasons it survived into LM I and LM III. observation

in EM

As

have supposed a local palace at Gournia86 or that are coming from Knossos.87 Political alliances ultimately to prove for the local situation at Chrysokamino, but the basic

of aMalia decisions

the previous period, political allegiances are difficult for this era. Some writers have suggested the region was part

is the case with

to reconstruct

state.85 Others were

impossible occur inMM III/LM I indicate outside political forces are changes that more likely than simple reactions to the local topography or some other that took place at Chrysokamino internal factor, because the consolidation is similar to the nucleation visible elsewhere. At Pseira, for example, the town expanded to the adjoining hill after MM IIB, and it almost doubled size in a single period of construction, its geographic implying the arrival new come must from beyond the island.88 have of many residents who same time but in different at the Radical population changes, taking place locations, suggest someone was making the size and location of settlements. As LM

elsewhere

II and mature

decisions

regionally

in regard to

in this part of Crete, the period between LM IB-early LM II ended with a local destruction. The habitation

site was unoccupied for awhile, and the later architecture had a different were em techniques plan of larger scale. New architectural construction new settlement was ployed after the destruction, but the economy of the still based on agriculture and animal husbandry.

82. MacGillivray

1997,

p. 22.

83. Driessen

2001.

84.Tzedakis

et al. 1989;Tzedakis

et al. 1990; Chryssoulaki 1999. 85. Cherry 1986, 1995. Cadogan

p. 21, no. 2, fig. 2;

86. Soles 1991. 87. Betancourt 88. Pseira

2004c.

IV, areas C

and D.

SURVEY

CONCLUSIONS

LATE MINOAN The

LM

271

III

III period was

the Chrysokamino LM I predecessor.

of the high point for the economic exploitation new was its The much than territory. building larger It had an entrance on the east that allowed access to

an enclosed

were constructed of courtyard. Rooms for living and working some of them megalithic. found in the excavation Objects large stones, a from establishment: included everything expected semi-independent

stone tools of including vessels for cooking, serving, and storage; a of number of the used for grinding with many types, querns type large cattle, and grain; animal bones of several species, including sheep/goats, in the do marine and other and used swine; shells; loomweights objects its surrounding farmland and used it mestic economy. The site managed pottery,

successfully Behind

for crops and animal husbandry. all of these situations lies theMinoan

agrarian economy. One from this study is a better important resulting survives best of Minoan land evidence The management. understanding III period, but it is surely applicable earlier; it sug from the Late Minoan of the most

conclusions

divided the land into specific gests that the system in use at Chrysokamino were was used and maintained which parts, differently. The farmhouse I, and it surely had more people living in it.The mature system must be the basis for the productivity agricultural management that allowed the residents to support a larger population with diversified agriculture and animal husbandry involving several species. larger than in LM

zones the land into different system of dividing agricultural in Chapter 21 was fully mature by LM III. This agricultural system would have required land for fields, gardens, and pasture. Analogies with later Greek and Roman farming methods show that ancient practices The

described

treated vegetables and field crops very differently. Because of the effects in Chap. 2), the local of the local topography and climate (see discussion conditions would have been very dry, with greater variations in temperature, some humidity, wind conditions, and amount of precipitation compared to of the other parts of Crete. Plants grown in gardens would have needed and the addition of organic matter special care, including extra watering to the soil for nutrients and moisture retention. This extra care left traces in the archaeological record in the form of organic matter in the soil, the from the and sherds that were carried out to the land manure, chemistry with the household refuse used as fertilizer (Chap. 21). along a location south of the settlement has In the Chrysokamino territory, that suggest itwas reserved for gardens (Fig. 21.1). Here, crops that had to be tended individually and whose products were harvested frequently for fresh produce could be managed easily. This land all of the characteristics

is recognizable in the archaeological record because it is not separated from the settlement by any topographic barriers, because it is relatively flat, and because it has a heavy scatter of sherds on its surface. The sherds are different from sherds found in settlements; are small, worn, and they isolated from other fragments from the same vessel. They occur within the

89. Sarpaki 1992.

soil at a depth that suggests tilling. The crops would have included grapes, herbs, and other garden plants. Pulses, probably aMinoan vegetables, as well. staple,89 would also have been tended here

CHAPTER

272

22

Low hills with phyllitic soil constitute a separate class of land (Fig. 21.3). as they still are in this part They would have been used for dry farming, of Crete. The

phyllitic soil retains moisture better than the red soils, it has nutrients, and it iswell suited to the growing of olives and grain. At the phyllitic soils do not have the heavy concentration of Chrysokamino, worn on as sherds found land the identified indicat small, garden plots, were never used for crops like ing that they vegetables requiring repeated

more

deposits The

of manure

that were

excavation

of the habitation that meat

bones,

indicating were Sheep and/or goats in large numbers in LM

tilled into the soil. site yielded a large corpus of animal a substantial role in the r?sidents' diet.

played the dominant

species. Bones were already present in LM III, I, and they continued to be deposited that the occupants enjoyed a steady supply of meat. They would

indicating also have had access

to the hides, leather, and other skin products that are so useful to aswell as towool for early societies,90 weaving. They clearly flocks and and the herds, suggests that the eroded hill kept topography not for growing crops. It in of Chalepa suitable for but is, fact, grazing lacks the evidence, present elsewhere, for terraces, sherds on the surface, and deep soil useful for farming. It is likely that some of the flocks and herds were moved summer when

seasonally to upland meadows, particularly during the the low land covered with terra rossa loses much of its

moisture.

was a normal practice in Classical Greece.91 Analogy other prehistoric pastoral economies suggests itwas probably a regu lar part of ancient Cretan animal husbandry,92 andWatrous has made a Transhumance

with

convincing

case

ments

for

the

seasonal

movement

of

even

some

complete

settle

during period.93 It is unlikely, however, that the entire of would have moved. At least in the LM I and Chrysokamino population LM III periods, when the evidence ismost complete, the diversity of the the Minoan

economy, with craftwork as well as varied agricultural land use, indicates permanent habitation by at least part of the community. that out The evidence presented here supports the view of Halstead

domestic

engaged in highly diversified agricultural establishments towhat may have been more agriculture and animal husbandry, in contrast Animals included specialized palatial productions.94 sheep and/or goats, as as situation well fields. This and cattle, pigs. Crops required gardens animals could had many advantages for an isolated farmstead. Different subsist on different types of pastures, and diseases in one species would not lying Minoan

a not necessarily have spread to others. Crop failures in single harvest would have been disastrous if additional harvests could be realized at other times of the year. In addition, different crops required labor at different times of without the year, so the work force could be maximized, long periods of a The residents of the achieved territory inactivity. satisfactory system that a resulted in long-term, stable economy.

90. Driel-Murray 91.Whittakerl988.

92. Cherry 1988.

2000.

93. Watrous

1977.

94. Halstead

1992.

CONCLUSIONS

SURVEY

THE Like

273

ERA

POST-MINOAN other coastal

settlements

was

abandoned

Chrysokamino movement throughout patterns lations. Most

much

inmore

resulted

in this part of Crete, the habitation site at in LM IIIB. This was part of a general of the Aegean.95 This shift in settlement

inland sites and fewer undefended

coastal instal

the change to a disruption of the situation that prevailed during the height of relatively stable and peaceful economic the Mycenaean empire. The majority view is that pirates and marauders made the exposed sites too dangerous, and most people pre and Gournia were abandoned or nearly ferred to live elsewhere. Mochlos same at time. about the abandoned scholars would

attribute

New foundations and increased population at existing sites can be seen on the high hills south of modern Kavousi.96 The residents used Minoan pottery and other artifacts, but the population appears too small to account for all of the coastal residents. It has also been proposed that some people emigrated

to Cyprus.97

the 1st millennium b.c., the residents eventually moved back During down from the peaks south of modern Kavousi. They resettled the region between the modern village and the sea, although they would never again the Chrysokamino live on the coastal hillside within territory itself. The near the territory was at largest and most important village Kephalolimnos, on the inland side of the coastal hills.98 The site is not excavated, but it is is a prominent covered with pottery, and a Roman or Byzantine millstone local landmark (Figs. 19.5,19.6). The coastal territory of the Bronze Age farmstead must not have been during the Classical and Roman periods. Pottery from these is absent from the surface, a situation paralleled on the bare and periods of equally dryland nearby Pseira.99 The inland slopes of the hills, however, were certainly exploited. Small amounts of pottery attest to increased use of the region in used much

era. Several sherds from a were the Byzantine large Byzantine amphora on site (App. J, J-l). They suggest that found the surface of the habitation some of the ruins may have stood a little shelter; high enough to provide the walls were neither unfortunately, the construction of amodern mandra see Desbor

95. For bibliography,

ough 1964,1972; Drews 1993. 96. Gesell 1990; Coulson 1990; Haggis

1993b;

for more

recent

bib

liography, seeDay and Snyder 2004; Eliopoulos 2004; Klein 2004; Glowacki 2004; Nowicki 2004. 97.

Karageorghis

1992.

98. Boyd 1901, p. 156; Pendlebury 1939, p. 375. 99.

Seager

1910.

100. Faure 1975, p. 32; 1989, p. 343; Spanakis 1991, pp. 322-323. 101. Haggis 2005, p. 38.

of the stones were

photographed in the mid-20th

nor described

before

century when many architecture has been

no Byzantine robbed. Although this part of Crete was so heavily settled at this identified at Chrysokamino, time that people must have visited the territory regularly, whether they

farmed

there

or not.

slopes of the low hills were farmed during the Venetian period, was reached and the height of the post-Bronze Age activity during the are from Ottoman sherds these present in the occupation. Many periods The

territory. Probably most of the surviving terrace systems are from the 15th to 19th centuries, as suggested by the excavation of AF22b and its cave recent town small In the has been centuries, adjoining (Apps. G, K). at Kavousi,100 a settlement whose roots, like those of are Chrysokamino, in the Neolithic

period.101

CHAPTER

274

22

use of the land was very different from during these later periods the pattern in the Bronze Age. The changes have been documented through The

and description of manmade features in the landscape in is best understood by transformation land The management (App. G). occur in spatially related clusters.102 They reveal examining the tiny sites that a new settlement pattern for the local era.The territory in the post-Minoan the identification

lived on the coastal side of the hill, and they must have walked out to their gardens and fields and returned home with their harvest and tools. Because the later residents lived farther away, they needed field houses for storage of tools and harvests, threshing floors near the fields to avoid carrying to bulky harvests back home, and wells in the fields provide water.

Minoans

Terracing areas with territory has several Chrysokamino agricultural terraces are all made of locally avail supported by stone walls (Fig. 16.3). The walls in any way. The able stones, gathered from the vicinity and not modified

The

locations

are recorded

of the terraces were period of maximum The excavation

Surface pottery shows that most used in Ottoman times, as is to be expected for the usage at the end of a long tradition. one of of terrace AF22b provides detailed evidence for in Appendix

G.

the large terrace systems (App. K). To build this terrace, the land was cut back at the uphill side, and a one-faced wall of small stones was constructed on the downhill side. The space behind the wall was then soil. If this type of terrace construction began at the bottom of the hill, soil from each successively higher terrace could have been shov eled onto the adjoining lower level when the adjacent higher slope was cut mortar

without filled

inwith

the top of the system called AF 22, the highest terrace (AF 22a) is rather small, and itwould not have yielded enough soil to fill both itself and AF 22b, so soil from the excavation of the small cave (Apps. G and K, AF 9) was used for part of the fill.The system as awhole was not constructed evenly, but it followed the lay of the land, combining features of the stepped back. At

terrace and the braided

terrace as defined

and Moody.103 by Rackham comes from the Ottoman period, and this

Pottery from the excavation date is probably applicable for the present configuration of the entire system. It is possible, however, that earlier terraces may have been modified when the final system was built. A small amount of Venetian pottery comes from within the soil of terrace AF 22, and itmust attest to earlier use of the hill side. An

occasional Minoan

sherd also occurs on the surface of this terrace

system. The Ottoman pottery from the lower level of terrace AF likely providing the date of the wall, is from the 18th century.

Land

22a, most

Use

thatMinoan land use at Chrysokamino project clearly demonstrates use in later from the local land different fundamentally periods. The no evidence for boundary walls or any other indica Minoan period provided This

was

tions of private ownership. All of the territory seems to have been farmed so well that by people who knew each other physical markers identifying boundaries

were

unnecessary.

From

more

recent

times,

the

remains

of

102. On the application of this see Russell 1985, p. 30. methodology, 103. Rackham and Moody 1992.

CONCLUSIONS

SURVEY

Cluster Figure 22.1. Map of Clusters 1-3, with closely related anthropogenic features

1

Cluster

275

Cluster

2

3

was divided between several eight field houses showed that the territory own each of which required its households, permanent fixtures for storage or other functions. Short walls also existed, giving a clear sign that boundary the limits between fields could be the subject of disputes if they were not in a permanent way. A basic change in land use is thus apparent. features from the post-Minoan Several of the anthropogenic periods occur in groups that suggest individual ownership/usage (App. G). Clus ters often consist of a group of agricultural terraces, a small building, and

marked

a

awell occurs nearby. threshing floor. Sometimes A brief discussion of three clusters shows the pattern

Cluster 1

(Figs. 16.2,22.1): (Fig. 22.1,

AF

5 and AF

7. Field houses

AF

4. Threshing

AF

6. Agricultural

left)

(consecutive?)

floor terrace walls

1 is a group of closely spaced and obviously related features. The presence of a threshing floor (App. G, AF 4) indicates a use with grains or some other crop requiring threshing, and the latest phase (mid-20th

Cluster

on some of the other suggests century) of dry farming slopes of Chomatas as a candidate. Field houses for storage and for shade in used likely barley structures with a rectangular this part of Crete are usually single-roomed of of the two examples Because their different degrees preservation, shape. here are more

likely to be successive

AF

13 and AF

AF

18. Threshing

16. Field houses

rather than contemporary. Cluster 2 (Fig. 22.1, (consecutive?)

center)

floor

AF 12.Well AF

11 and AF

17. Agricultural

terrace walls

2 has features that are similar to those of Cluster 1.As in Cluster 1, the two field houses were probably not used at the same time. The addition of awell, one of three in the territory under examination, shows that the

Cluster

labor expended water

from

in digging was preferred

a distant

to the difficulty

Cluster 3 AF AF AF

of transporting

source.

(Fig. 22.1, right)

of stones, probably a field house floor 20. Threshing terrace walls 11 and AF 17. Agricultural 10.Mound

CHAPTER

276

22

a certainly collapsed build one of rather small dimensions. It probably served as the

of stones

The mound

in this cluster

is almost

ing, apparently field house for threshing floor AF 20. The threshing floor ricultural terrace walls (App. G, AF 11 and AF 17), where

is close to ag itwould have

at harvest time. These terrace walls extend to the vicinity a and 2, boundary between the two parts of the landscape may have existed, but it is not visible archaeologically. been convenient of Cluster

In addition to these clearly associated groups of features, other groups also exist (Fig. 16.2). The small cave (App. G, AF 9) is near a large group of agricultural terraces (App. G, AF 22), and itwas probably excavated for use as a field house. This uncommon in cases where mar practice is not land is being cultivated.104 A well boundary wall (App. G, AF 3). ginal

(AF 2) is associated with

a

nearby

territory also has features associated with animal husbandry. The site (App. G, AF 32) and its successor south of the habitation were to AF intended hold flocks and herds. Some of the field 28) (App G, houses could have been used as shelters for shepherds. The

oval pen

clusters and the more

These

of land use was

isolated features show that the later system than the situation during the Bronze

different

completely was divided between many users, with individual sections Age. The territory on that accommodated dry farming agricultural terraces by nonresident farmers who traveled out to their tiny plots of land from Kavousi and its like Kephalolimnos. The eight field houses suggest that the predecessors land was subdivided into about six separate plots (assuming that the nearby pairs

are

successive

rather

than

contemporary).

of the post-Minoan features are associated with small-scale ag riculture and animal husbandry. They furnish evidence for small individual not farm was quite large holdings, large agrarian estates. The Late Minoan Most

in comparison. The later features probably also represent multiple periods of use, and often only the most recent period is clearly visible. It seems was farmed and likely that the coastal strip grazed during many periods after the abandonment of the Minoan settlement.

The

Situation

21st

Century

at

the

Beginning

of

the

Since the end of the 20th century, the Chrysokamino territory has been used mainly for recreation, especially fishing, and for grazing animals. At the end of its long period of use, the coastal strip between Chordakia and sea an assortment of the has been extensively overgrazed. Even phrygana, small spiny shrubs and thistles, is restricted in distribution. The coastal so bare of that a contrast with the inland area is visible strip is vegetation as from the air.The satellite photograph Figure 1.3 was taken published in 1988 before the construction of the road from the earlier track to the excavated habitation is a

prominent

104. Halstead

site.The boundary between wasteland

feature.

2000,

p. 118.

and olive groves

SURVEY

CONCLUSIONS

277

? Z <

Less

? z

More

^??mmmm^^

Population

< z o a: <

<

Resources

o

o LU (0 D LU

LU CO CO CO LU

Figure 22.2. Pressures influencing greater

or lesser

use

of

marginal

land

O

Available Land

for agriculture

Comments use occurred when the farm Chrysokamino change in land to higher in LM IIIB, and most of the residents moved locations away from the coast. The coastal territory at this location would a permanent never It would, settlement. instead, alternate again have

A fundamental was

abandoned

as in the present day, periods of almost complete abandonment, and periods of use by farmers who traveled out to their fields from a village located away from the coast.

between

Like

the more

recent abandonment

not to farm the seacoast several factors, including tween size of population,

of the coastal strip, the decision after the Bronze Age must have been related to relative prosperity. A three-way relationship be amount of available land, and other components

of the region s economy has profoundly affected decisions about farming to is land. Land that difficult farm successfully, yielding only marginal a small return for substantial investment in time and effort, will not be used in every period. The principles are illustrated in Figure 22.2. More fewer outside resources, and a decreased availability of better population, land exert pressure to increase the exploitation of marginal land. If these factors

are

reversed,

the

pressure

to use

marginal

land

decreases.

In the late 20th century, external contributions to the economy (outside resources) and the availability of better land have caused the coastal strip to be abandoned. Being awaiter in a restaurant, for example, or raising olives in the Kambos is financially and socially more attractive than spending or animals in a desolate long hours trying to raise crops landscape. the early 1st millennium and b.c., the decrease in population During more land was availability of probably responsible for the coast. With from Crete to Cyprus and away from the migrations other places in the East, the people who remained chose better farmland than the exposed and dangerous coastal strip with its dry and marginal soil. the concomitant

move

change that caused the land to become the process of abandonment.

Climatic

drier must have intensified

It was probably an increase in population that prompted reuse of the area in the Venetian and Ottoman all the late periods periods. During

CHAPTER

22

out to the fields

from

278 the land was

when homes

farmed, inland. These

located

the farmers

traveled

new

continued

to be used

farming practices the early 20th century. As Lloyd observed in the case of the Roman to the fields "There is little doubt that the peasant commuting Empire, until

from

town,

or hamlet

village,

was

a common

sight."105

In conjunction with this new strategy, the farmers needed to construct field buildings and threshing floors (alonia) in the countryside. All of the in the present study were built after installations permanent recognized one of them, along period (App. G), and the excavation of with the closest part of its nearby terrace wall system, shows that the final era of use the Venetian period (App. K). postdated The several examples of the small field installa survey provides the Minoan

tions used

as

facilities in the former territory of agricultural processing a small building or Chrysokamino (App. G). They typically consisted of in shallow cave, a threshing floor, and a series of terrace walls. Elsewhere were a alonia used for of Greece, crops, including figs processing variety and grapes that needed to be dried.106 In this part of Crete, however, the flat areas were used primarily for threshing grain, and all of the examples noted at Chrysokamino have vertical stones at their edges to keep the grain contained during the threshing process. Walls made of rough stones were used as boundary markers before the modern fences were built (App. G). They divided the public land at the coast from the inland private land, but they were so flimsy that they a only survive in few places. These early boundary markers show that the a governmental regulations that currently operate and provide for balance between open land and private land has a long history in the region, but no evidence survives to suggest it operated during the Bronze Age. Like the agricultural-processing facilities, the boundary markers seem to be a post-Minoan

phenomenon.

use of the coastal hills for growing half of the 20th century at Chrysokamino. The

started theMarshall after the destruction

into the first grains continued In 1947, the United States

Plan, which provided funding to help rebuild Greece II. A part of the of its economy during World War

plan provided money for deep wells in the Kambos to permit irrigation.107 As a result, the land was planted in olives, a crop that was gaining in favor throughout Crete because of its nutritional value and other aspects of its production.108 After the transformation of the land to support olive culture, and its associ the coastal side of the hills, where the Bronze Age workshop cave were was ated settlement and burial located, increasingly neglected the coastal the irrigation system did not reach this far. Although recre at of from used the close the 20th aside century, strip longer one ation and animal husbandry the inland shepherd), (by only slopes of

because

was no

called Chordakia, were completely planted in olives and a few Chomatas, other trees. Chordakia and the Kambos, divided into small individually owned plots, supported olives and garden vegetables that were tended by of Kavousi, who traveled out to their fields from their homes in the village. By this time, the land management system that favored the in the midst of the gardens and fields had ceased location of settlements residents

to

exist

for many

centuries.

105. Lloyd 1991, p. 235. 106. Lee 2001, p. 61. 107. Allbaugh 1953, pp. 258-263. 108. Riley 2002.

APPENDIXES

and

Petrography

X-Ray

of

Diffraction

Analyses

and

Chimneys

Furnace

Slags

er and byGeorgeH\ My Philip P. Betancourt

were of the slags and furnace chimneys from Chrysokamino in the analyzed by optical thin section and by X-ray powder diffraction of Mineralogy Laboratory, Department Geology, Temple University. The sec D. M. Organist Laboratory (Newark, Delaware) prepared the thin tions. The methodology followed standard practices for optical thin section Samples

analysis.1 The Mclntosh,

and

presentation

of the results follows

the format used byMyer,

Betancourt.2

a analyses by X-ray diffraction used Rigaku XRD powder dif with a copper X-ray tube source, horizontal fractometer model DMAX/B (radius 185 mm) with total accuracy within 0.02 degrees, and goniometer The

curved crystal graphite monochronometer analysis system. The continuous scan mode, which counts integrated intensity per specified sample interval, scans run at 10 a diffraction produced degrees per minute, with sampling interval of 0.05. All data were recorded on magnetic disc, analyzed by use. Rigaku software, and printed and plotted for hard copy

PETROGRAPHY OF THE SLAGS are dark masses with a slags at Chrysokamino glassy appearance (see seenwithout Because that similar when many pieces appear Chap. 10). magni fication have very different compositions, the class as awhole can be described

The

as

in detail but similar in overall appearance. The differ heterogeneous ences are sufficient to suggest variations in the as well original materials as different in the furnaces during the pyrotechnological conditions process.

classes of slag can be distinguished. One class consists of on the inside of chimney walls, and the other class consists of

Two different condensation

solid pieces of dark vitreous material of four samples are presented here.

1. Bambauer,

1979.

Taborsky,

and Trochim

that formed from the melt. Analyses

2.

Myer,

1995.

Mclntosh,

and Betancourt

A

APPENDIX

282

Slag sample CHR 97.

Figure A.l. P?trographie black dendritic in a

glassy

thin

matrix;

section

crystals

showing of magnetite

plane-polarized

light, width of field 2.5 mm. 97 (vitreous coating in the interior of a chimney fragment, a Fig. A.l), slag containing opaque dendritic inclusions (mag netite) occurring as branching dendritic crystallizations, with no

1. CHR

or

fayalite

2. CHR a

pyroxene.

in the interior of a chimney fragment), of magnetite, with radiating crystallizations

93 (vitreous coating slag containing

minor

olivine.

87 (slag fragment, Fig. A.2), a slag containing opaque den dritic inclusions (magnetite) occurring as branching dendritic masses of along with green crystal points and crystallizations,

3. CHR

pyroxene and masses of cuprite and fayalite. 4. CHR 99 (slag fragment, Figs. A.3, A.4), a slag fragment consist a considerable percentage of ing of two dissimilar bands, with in the glassy matrix along with inclusions of an clinopyroxene opaque mineral, fayalite, cuprite, and unidentified phases.

with

Slag Number

Black

of samples examined:

of a chimney Matrix

Inclusions

Dendritic 1 (CHR

97, slag deposited

on the interior

fragment) Groundmass)

(Glass

A. Optical properties and color 1.Under plane-polarized light: mottled,

with

color variation

tan to yellow brown to dark brown; some of the color variation around the opaque inclusions. 2. Under

cross-polarized

light: isotropic.

grain size and modality from under 62 micrometers

B. Overall

vary widely, under 62 micrometers C. Overall

from

is zonal

of inclusions

and voids:

inclusions

to over 2 mm; voids vary widely, from to over 1 mm; ca. 50% inclusions; ca. 10% voids.

preferred orientation

of inclusions

and voids: random.

analyses

Figure A.2. Slag sample CHR 87. P?trographie

thin

section

show

ing magnetite (black, thin curved crystals), fayalite (pale, circular, at upper center), and pyroxene (dark rounded masses) in a glassy matrix; plane-polarized

light, width of field

800 micrometers.

Figure A.3. Slag sample CHR 99. P?trographie thin section showing pyroxene crystals (radiating), magne tite (black, at top left of center), and cuprite (circular, upper center) in a glassy matrix; cross-polarized light, width of field 800 micrometers.

Figure A.4. Slag sample CHR 99. P?trographie thin section showing pyroxene crystals (dark) and cuprite (black triangular mass with rounded corners

at center)

plane-polarized 2.5 mm.

in a

glassy

matrix;

light, width of field

of

slags

and

furnace

chimneys

283

A

APPENDIX

284 D.

Inclusions 1.Magnetite dendrites. Size range: from under

up to ca. 10 x 250

1 micrometer

micrometers.

orthogonal branching growth patterns with Shape: multiple initiated along both primary and secondary the crystallization branches; usually like a two-sided comb, with isolated cross-sec tions visible as well. opaque dendritic quench crystals with many as identified branches; isometric; magnetite by X-ray diffraction; in some locations the rows of long dendrites are at right angles to one another, creating a gridlike pattern; sometimes iron alteration Comments:

opaque

as red brown colorations

are visible

products

at the margins

of the

crystals.

2. Quartz.

Size range: from ca. 10 micrometers to rounded. Shape: subrounded

up to ca. 2 mm.

some grains single grains; undul?se extinction; are some rimmed with have acicular inclusions (rutile?); grains the opaque dendritic crystals (i.e., the quartz acted as a nucleation Comments:

surface on which

the crystals grew).

3. Chert. Size

range:

Shape: Comments:

from

ca.

subrounded

10 micrometers

up

to ca.

1 mm.

to subangular.

microgranular

texture.

Voids A. Size range: up to over 500 micrometers. B. Shape: mostly C. Comments:

oval. voids from entrapped

gas.

Modality Matrix Voids:

(glass groundmass): ca. 10%. Quartz: under

ca. 50%. dendrites: Opaque 1%. Chert: under 1%.

ca. 38%-40%.

Discussion on the interior of a chimney a Sample CHR 97 consists of glassy deposit a abundant dendritic crystal includes mottled The matrix, glass, fragment. as that crystallized when the lizations, which may be identified magnetite indicates the slag cooled quickly (see App. B). The abundant magnetite are unaltered which conditions. under formed Quartz grains, reducing slag in of Chalcolithic noted have also been relics from the original charge, slags date from Israel.3 The X-ray diffraction (the analysis of this sample shows magnetite and dendrites) quartz (the quartz grains and the chert grains). Other peaks No fayalite or other members are too ambiguous for certain identification. of the olivine group can be detected.

3. Bachman

1980,

pp.

108-110.

analyses

with

Slag

and

slags

of

furnace

Radiating

chimneys

Magnetite

of samples examined: 1 (CHR the interior of a chimney fragment).

Number

Matrix

(Glass

285

Inclusions 93, vitreous

on

coating deposited

Groundmass)

A. Optical properties and color 1.Under plane-polarized light: mottled,

with

color variation

from

tan to yellow brown to dark brown. 2. Under cross-polarized light: crystallized (devitrified), consisting of interlocking blades with some pleochroism (most likely a postburial an not characteristic the of reaction, original slag). B. Overall grain size and modality of inclusions and voids: inclusions vary to over 2 mm; voids vary widely, from widely, from under 62 micrometers to over 1mm; ca. 15% inclusions; ca. 10% voids. under 62 micrometers C. Overall D.

orientation

preferred

of inclusions

and voids: random.

Inclusions 1.Magnetite dendrites. Size range: from under 1micrometer

up to ca. 60 micrometers.

Shape: small radiating groups of crystals. Comments: opaque dendritic quench crystals consisting mostly of small radiating groups; identified asmagnetite by X-ray diffrac tion; sometimes Fe alteration products are visible as red brown at

colorations

the

margins.

2. Chert. Size

ca.

range:

30 micrometers.

subrounded.

Shape: Comments:

texture.

microgranular

3. Carbonate.

Size range: up to ca. 100 micrometers. Shape: films. Comments: 4. Olivine Size

amorphous;

postburial

depositions.

group. range:

under

10 micrometers.

masses. Shape: spherical Comments: alteration of the glass (devitrification); is a postburial alteration. sible that the devitrification

it is pos

Voids A.

Size range: up to over 500 micrometers.

B. Shape: mostly C. Comments:

oval. voids from entrapped

gas.

Modality Matrix Chert:

(glass groundmass): under 1%. Carbonate:

ca. 85%. Voids: under

ca. 10%. Magnetite: 1%. Olivine: under 1%.

ca. 4%-5%.

286

a

appendix

Discussion (CHR 93) consists of a glassy deposit on the interior of a amottled a few inclu glass, contains only chimney fragment. The matrix, that crystallized when the sions, which may be identified as magnetite

This

sample

slag cooled from

quickly. The

glassy matrix

is extensively

altered

(devitrified)

weathering.

(the analysis of this sample shows magnetite chert the devitri (in (the opaque crystals), quartz grains), hedenbergite fied glass), and a hydrated copper chloride (not present in the thin section in the X-ray prepared from this sample). No olivine could be detected diffraction

The X-ray

analysis, present

the optical examination shows that a tiny amount a too to in bulk analysis). little show up (apparently although

with

Slag

Dendritic

Black

is

and

Inclusions

Pyroxene Number

of samples

Matrix

1 (CHR

examined:

87, fragment

of loose slag).

Groundmass)

(Glass

A. Optical properties and color to brown. 1. Under plane-polarized light: pale green 2. Under cross-polarized light: isotropic where pale in color and anisotropic where

brown.

B. Overall

grain size and modality of inclusions and voids: inclusions up ca. 5%; aniso voids vary widely; micrometers; isotropic glass ca. 80%; inclusions ca. 10%-15%; voids under 2%. tropic glass

to ca. 20-30

C. Overall D.

preferred

orientation

of inclusions

and voids: random.

Inclusions 1.Magnetite dendrites. Size range: mostly ca. 5 x 10-20 micrometers. a two-sided Shape: branching dendrites, usually like visible as well. with isolated cross-sections

comb,

opaque dendritic quench crystals with many as isometric; identified magnetite by X-ray diffrac

Comments: branches; tion. 2.

Pyroxene. Size

range:

ca.

10 micrometers.

masses the dendritic enveloping Shape: crystal points and in the the is nucleated sometimes needles; pyroxene magnetite. Comments:

green.

3. Fayalite. Size range: ca. 10 micrometers

up to ca. 50 micrometers.

masses.

Shape: spherical Comments: yellow crystalline

structures.

to yellow redmasses with

internal radiating

analyses

of

slags

and

4. Cuprite. Size range: under Shape: Spherical Comments:

red

furnace

chimneys

287

10 micrometers. masses. in color.

Voids A. Size range: up to over 100 micrometers. B. Shape: mostly C. Comments:

oval. voids from entrapped

gas.

Modality ca. 80%. matrix (glass groundmass): Magnetite Anisotropic ca. 10%. ca. 5%. Voids: matrix (glass groundmass): Isotropic Pyroxene: under 5%. Fayalite: under 5%. Cuprite: under 1%.

dendrites: under 2%.

Discussion a of magnetite inclusions sample consists of glassy matrix with and The forms curved dendrites, pyroxene, cuprite, magnetite fayalite. acicular crystals within the glass matrix. The sample was consumed in the so that itwas not available for examination preparation of the thin section,

This

diffraction.

by X-ray

with

Slag

Pyroxene,

and

Cuprite,

Fayalite

Inclusions Number

of samples examined:

Matrix

(Glass

1 (CHR

99, fragment

of loose slag).

Groundmass)

A. Optical properties and color 1.Under plane-polarized light: isotropic glass with between the bands). (with great differentiation 2. Under cross-polarized light: black.

flow banding

of inclusions and voids: inclusions grain size and modality to over 1 cm; voids vary widely, from 62 micrometers under vary widely, to over 1mm; ca. 60%-70% inclusions; under from under 62 micrometers B. Overall

1% voids. preferred orientation

C. Overall

of inclusions

and voids: usually parallel

to the banding D.

Inclusions

Band

1 1.

(pale

colored glass)

Pyroxene.

Size range: average ca. 100-200

micrometers

in diameter.

A

APPENDIX

Shape: euhedral twinned crystals in radiating clusters. Comments: green to dark green to black or brown; pleochroic to brown in disseminated (weak to moderate pleochroism) green the glass; quench crystals from rapid cooling; differentially different interference colors at the crystal margins. 2. Fayalite.

zoned,

with

Size

range:

average

Shape: spherical. Comments: yellow

ca. 20-40

in diameter.

micrometers

to yellow redmasses with

internal radiating

structure.

crystalline

3. Cuprite. Size range: average ca. 20-100 Shape: spherical to cubic. Comments:

red; interstitial mineral.

4. Opaque Size

range:

average

ca.

1-12

micrometers

x 250 micrometers

isolated grains composed not identified.

Band 2 (brown glass, stronglyflow-banded; 1. Opaque mineral. Size

range:

average

ca.

1-12

in diameter.

iron (or tenorite?).

elemental

Comments: anisotropic;

in diameter.

to the pyroxene.

Shape: nearly spherical. Comments: opaque; possibly 5. Rock grains. Size range: 1 mm Shape: rounded.

micrometers

to 1 mm

of interlocking

dense w/microscopic

micrometers

crystals;

inclusions)

in diameter.

Shape: nearly spherical. Comments: similar to band 1 but much more 2.

x 1.25 mm.

abundant.

Pyroxene.

Size range: average ca. 100-200 micrometers Shape: euhedral twinned crystals in radiating

in diameter. clusters.

similar to band 1 but less abundant.

Comments: 3. Fayalite. Size

range:

average

ca. 20-40

micrometers

in diameter.

Shape: spherical. Comments: similar to band 1 but less abundant. 4. Cuprite. Size

range:

average

ca. 20-100

micrometers

in diameter.

Shape: spherical to cubic. Comments: similar to band 1 but less abundant. IDS A.

Size range: up to over 100 micrometers.

B. Shape: mostly C. Comments:

oval. voids from entrapped

gas.

analyses

slags

of

and

furnace

chimneys

289

Modality Band 1 ca. 70%. Matrix

Pyroxene: Opaque

mineral:

under

ca. 25%. Voids: under 2%. (glass groundmass): 1%. Cuprite: under 1%. Fayalite: under 1%. Rock

1%.

grains: under Band 2

ca. 70%. ca. 15%. Pyroxene: Opaque mineral: (glass groundmass): ca. 5%. Voids: under 2%. 1%. 1%. under Cuprite: under Fayalite:

Matrix

Discussion a slag with two dissimilar bands. The matrix Sample CHR 99 consists of contains pyroxene, fayalite, cuprite, and other unidentified inclusions. a X-ray diffraction analysis of bulk sample from this slag indicates that a sufficiently crystallized to yield patterns: clinopyroxene, red cubic crystals), probably augite (the pyroxene crystals), cuprite (the fayalite (brown radiating spheres), and elemental iron (probably the opaque inclusions). in this slag, one with Two types of glassy bands can be recognized are

four minerals

crystals and the other with fewer crystals but with clinopyroxene abundant, tiny spherical inclusions of an opaque mineral, possibly iron. The slag, which was produced under strongly reducing elemental includes tiny inclusions of fayalite (the Fe end member of the conditions, olivine group). Red crystals of cuprite (20-100 micrometers across) are or as a part of the also present, but whether they formed original process in postburial oxidation is uncertain. abundant

more

of fayalite and pyroxene provide an indication of the degree of reduction during the smelting process. The presence of large amounts of fayalite is a characteristic of slags formed under highly reducing The

relative amounts

as

to the occurrence

of pyroxene, which would be typi substantial amounts of pyroxene oxidizing as in this and the oxide cuprite are visible in slags from Chrysokamino, slag, and fayalite are present, suggesting formation under more less magnetite than the slag recorded as no. 3 above (CHR 87). oxidizing conditions conditions, cal of more

opposed

conditions.4 When

CHRYSOKAMINO FURNACE CHIMNEY FABRIC Number

of samples examined:

6 (CHR 91, CHR

92, CHR

93, CHR

94,

CHR 97, CHR 98) Matrix

(Groundmass)

A. Optical properties and color 1. Under plane-polarized light: mottled. 2. Under cross-polarized light: anisotropic. 4.

Hauptmann,

Maddin

Bachmann,

1994, p. 6.

and

B. Overall grain size and modality of inclusions and voids: inclu to over 5 mm; voids vary sions vary widely, from under 62 micrometers

APPENDIX

2?O

to over 1 cm; ca. 50%-60%

widely, from under 62 micrometers ca. 10%-20% voids. C. Overall preferred orientation to the surface of the furnace. D.

Silt-sized

inclusions

1. Quartz,

of inclusions

A

inclusions;

and voids: usually parallel

(under 62 micrometers)

type 1.

Size range: up to 62 micrometers. Shape: subangular to angular. single grains; the smaller size range of the larger 1 type quartz grains. 2. Quartz, type 2. Comments:

Size range: up to 62 micrometers. Shape: subangular to angular. Comments:

multiple grains; the smaller size range of the larger 2 type quartz grains; derived from the phyllite.

3.Mn Oxide(s)(?). Size range: up to 62 micrometers. Shape: irregular. Comments: opaque films and borders on grains and voids. 4. Chlorite. in section. Size range: ca. 5-6 x 2 micrometers Shape: flakes. colorless under plane-polarized Comments: light; first-order colors under interference gray cross-polarized light. 5. Carbonate. Size

range:

ca.

10 micrometers.

Shape: spherical. Comments: very few grains. Voids

Type 1 Size range: up to over 2 cm. B. Shape: varies widely: irregular;

A.

cylindrical;

various

plant

forms.

C. Comments:

voids

from burned

out organic matter,

including

chaff (Fig. A.5).

Type 2 A.

Size range: up to over 1 cm.

B. Shape: elongated. C. Comments: shrinkage

Inclusions

above

cracks.

Silt-Size

(above

62 Micrometers)

1. Carbonate.

Size range: up to 3 mm Shape: irregular. Comments: biogenic

in size. carbonate

rock; altered by heat; may

ANALYSES

Figure

A.5.

Furnace

chimney

OF

SLAGS

AND

FURNACE

CHIMNEYS

291

sample

CHR 97. P?trographie thin section showing clay fabric with voids from burned

out

chaff;

cross-polarized

light, width of field 6 mm. have oolitic contains

structure with one

foraminifera;

quartz grain consisting fluid inclusions.

spherical to ovoid to irregular oolites; mm grain inCHR 98 is attached to 1.5

of low-grade metamorphic

quartz with

2. Shale. Size range: up to 5 mm Shape: subrounded. Comments:

in size.

contains grains of quartz; some fragments

altered by heat. 3. Quartz, type 1. Size range: up to 0.5 mm

highly

in size.

Shape: angular. Comments: single grains;

clear; uniform

extinction;

precise

boundaries.

4. Quartz,

type 2.

Size range: up to 1 mm Shape:

in size.

angular.

Comments:

multiple

grains; derived

from the phyllite.

5. Chert. Size range: up to 500 micrometers Shape:

rounded,

subangular,

and

in size. angular.

a subrounded granular texture; CHR 94 contains with Fe-rich rhombic chert grain included crystals (siderite?). Comments:

6.

Ferruginous

siltstone.

Size range: up to 5 mm in size. Shape: subangular to rounded. Comments:

sedimentary

rock;

consists

of quartz

grains

cement

ed by hematite; may contain small fragments of shale. 7. Phyllite. in size. Size range: up to 600 micrometers Shape: subrounded. consists of quartz, feldspar, mica, and chlorite; extinction and multiple grains; CHR 89 has a has undul?se quartz a subhedral of crystal of epidote with light grain phyllite containing in brownish green pleochroism light and typical plain-polarized Comments:

A

APPENDIX

292

"stained glass window" colors in cross-polarized

appearance

of second-order

interference

light.

Modality Matrix

(groundmass): 20%-30%. Voids: 10%-20%. inclusions: 1. Quartz, Silt-sized type 1:10%. 2. Quartz, 3.Mn Oxide(s)(?): under 1%.

type 2: under 5%.

15%. 2. Shale: 15%. 3. Quartz, type 1: Larger inclusions: 1. Carbonate: under 5%. 4. Quartz, 2: under 5%. 6. 5. Chert: under 5%. Siltstone: type under 5%. 7. Phyllite: under 5%. Discussion fabric is consistent with

The

the Gulf

ofMirabello.

The

a local source in the region of the east side of quartz and phyllite inclusions are petrographi

oxide inclu calfy similar to those found in this part of Crete. Manganese sions are also present in clay deposits from the region, including deposits near the site of are common, and All of the constituents Chrysokamino. no distinctive would

elements

allow a precise

ANALYSIS

OF

have been recognized source to be identified.

SLAGS

BY X-RAY

by this methodology

that

DIFFRACTION

was on four Analyses by X-ray powder diffraction performed samples of were slag, two of which angular broken pieces of glassy slag containing mi two of which were vitrious coatings on the interior inclusions and croscopic were surfaces of chimney fragments. The following mineralogical phases sufficiently well crystallized to provide X-ray patterns in bulk samples:

CHR 96 (glassy slag) Akermanite Diopside Delafossite

Dolomite (?) CHR 99 (glassy slag) Augite Cuprite Fayalite Elemental CHR

iron(?)

93 (coating

on the interior of a chimney

fragment)

Magnetite Quartz Hedenbergite Unnamed hydrated CHR

copper chloride

97 (coating on the interior of a chimney

Magnetite Quartz

fragment)

SEM/EDAX

Analysis

by Yannis Bassiakos

This

study presents analyses of metallic prills and slag by SEM/EDAX. of Archaeometry, Institute analyses were made at the Laboratory of Materials N.C.S.R. Science, "Demokritos," Hagia Paraskevi, Attikis,

The

Greece.

Table B.l presents the analyses of copper prills arranged by increasing arsenic content. This set of measurements differs from the PIXE results shown

in Appendix

in that these measurements

D

were made

on clean, in this set

arsenic contents are substantially higher than they are in the values obtained from PIXE analysis to 26%, with an average of 5.51%. (App. D); they range from 0% were sawed, mounted, and analyzed with SEM/ Samples of the slag EDAX. The results are shown inTable B.2. uncorroded

metal. The

of measurements

TABLE B.l. SEM/EDAX ANALYSES OF COPPER PRILLS ARRANGED BY INCREASING ARSENIC CONTENT SEM/EDAX Analysis 24

No.

Arsenic in

Weight

Content Percentage _0.00 0.20

26 34 22

_0.50 _1.50 _1.10 4.10

11 _5.30 12 _5.50 45

_7.37

16

10.00

48

26.00

TABLE B.2. SEM/EDAX Analysis

No.

Al

Mg

ANALYSES

OF METALLURGICAL K

Si 3.4

nd

nd

nd

1.6

7.4

41

0.8

0.9

7.5

42.1

0.9

nd

nd

0.2

0.7

nd

0.5

0.1

nd

0.3

2.8

2.6

0.6

24.5

47.6

0.6

5.7

0.3

nd

nd

Ca

0.1

REMAINS Ti

Cr

Mn

Fe

nd

nd

nd

4.6

1.6

12

0.5

0.1

0.2

33

1.9

12

34

0.4

0.1

0.1

0.1

0.2

0.1

0.1

0.06

2.4

0.7

nd

0.1

nd

0.04

2.7

0.1

0.2

0.2

0.7

1.3

29.5

0.15

0.1

0.2

0.06

nd 22

0.1

0.2

0.4

0.1

nd

nd

18

1.3

0.7

0.1

0.05

56

0.1

nd

2.7

10

0.5

5.9

0.7

11

2.5

nd

0.3

nd

0.4

nd

0.4

nd

nd

1.9

12

2.4

nd

0.2

nd

0.1

nd

0.1

nd

nd

0.2

13

0.4

1.1

1.3

0.05

0.2

14

0.5

6.8

37.7

0.8

0.1

1.4

24

15

1.3

5.8

37.5

0.8

0.1

1.2

26

16

nd

0.4

0.4

nd

0.6

nd

17

nd

0.1

1.4

0.7

24

0.1 2.1

37

18

10

36.5

1.2

0.1

19

17.1

1.5

58.4

0.03

0.1

20

0.8

7.7

36.7

26

0.1

0.1

0.9

0.1

21

0.5

6.9

34.5

0.6

0.2

22

0.3

0.2

1.3

0.1

0.4

23

0.4

2.7

4.7

0.08

24

nd

nd

0.2

0.1

0.1

25

0.2

0.4

3.4

0.5

24

0.02

0.2

nd

0.6

nd

0.7

0.5

0.06

0.6

0.3

95

0.8

0.1

0.2

27

0.6

0.1

0.2

25

0.1

nd

nd

nd

nd

0.2

2.1

1.1

6.4

0.5

0.2 nd

14 29 34

0.9

20

0.7

nd

0.4

nd

nd

0.02

0.7

4.6

0.2

0.4

0.1

0.1

0.1

0.1

0.05

nd

4.8

0.1

nd

nd

0.1

0.6

0.2

4.3

39.2 1.4

nd

nd

nd

6.2

33.7

0.7

0.2

1.2

0.2

0.6

24

nd

1.3

32

0.2

nd

<0.1

0.7

0.4 14

<0.1

0.1

nd

0.06

0.4

nd

0.2 0.03

95

nd

0.4

24

0.6

nd

nd

nd

2.9

0.1

nd

0.06

0.8

<0.1

0.4

1.6

0.8

<0.1

13

1.2

0.6

18

0.3

2.3

18

30

nd

nd

31

nd

0.2

32

0.8

6.2

37.9

33

3.3

4.8

0.4

nd

0.1

nd

34

nd

0.1

1.4

<0.1

<0.1

nd

35

1.1

19.7

54

<0.1

0.1

2.1

36

1.2

28.9

47

0.3

0.1

37

1.7

15.1

51

0.3

0.1

38

1.2

2.7

0.07

0.2

0.1

39

1.5

20.4

60.7

nd

0.2

40

1.1

19.5

64.5

nd

nd

2.8

41

2.0

15.3

53.5

nd

nd

2.34

0.6 30

18

0.7

23

0.3

0.2

89

nd

<0.1

nd

0.2

0.1

<0.1 0.3

0.2

0.3

9.1

0.3

9.2

0.2

7.9

1.6

1.2

12.8

0.97

0.06

0.27

11.4

0.14

0.04

10.6

22.8

53

nd

nd

5.05

0.71

20.4

63.2

0.03

nd

4.00

1.54

0.84

0.04

0.15

44

3.5

14.5

22

nd

nd

0.9

1.48

1.1

0.65

0.33

0.17

nd

0.11

0.15

0.05

nd

0.05

0.86

0.31

0.94

0.45

0.23

0.26

47

nd

0.08

1.93

nd

48

nd

nd

3.06

nd

49

nd

nd

0.32

nd

25.4 3.28 24.5

69

1.3

1.78

0.32

3.5

2.3

6.37

42

4.9

0.1

42

nd

8.7 33

43

4.5

5.7

0.3

1.03

nd

4.2 37

0.1

0.5

0.86

1.9 11

0.7

0.2

29

46

85

0.6

28

45

28

0.06

21

0.8

26

2.3

0.1

nd

27

26

32.6

7.5 55 1.25 14.4

nd

1.53

0.07

0.01

0.01

4.18

nd

2.19

0.08

nd

nd

2.01

nd

0.34

0.04

nd

0.02

1.93

B.2

TABLE

No.

Analysis

(ana.) Cu

Au

nd

nd

0.6

0.2

0.2

nd

_90_ nd

0.3

0.1

0.1

nd

nd

2.3

12

Pb

1.6

_91_ 3.3

11

As

0.8

73

10

Zn

Sb

Cl

Na

nd

nd

~nd~

nd

nd

nd

nd

nd

nd nd

nd

nd

nd

nd

nd

0.3

1.4

nd

nd

nd

nd

nd

nd

nd

0.3

4.1

nd

nd

0.5

nd

nd

nd

nd

0.2

nd

nd

nd

nd

0.2

nd

nd

nd

0.2

1.6

nd

nd

nd

nd

0.3

nd"

nd

nd

nd

nd

nd

nd

nd

nd

0.1

nd

nd

nd

nd

96

0.6

0.2

nd

nd

nd

1.6

0.2

0.3

nd

nd

nd

nd

0.1

nd

nd

0.2

5.3

nd

nd

nd

nd

nd

nd

nd

0.5

5.5

nd

nd

nd

nd

nd

nd

nd

0.4

nd

nd

nd

nd

nd

nd

nd

0.2

0.2

nd

nd

nd

nd

nd

nd

nd

0.2

nd

nd

nd

nd

0.2

nd

nd

nd

nd

nd

89

13

_91_ 0.1

14

0.8

15 16

84

nd

17

63

nd

JO_ 3.7

nd

0.2

nd

nd

1.3

nd

nd

nd

nd

18

0.1

nd

0.1

nd

nd

nd

nd

0.1

nd

nd

19

1.1

nd

0.1

nd

nd

nd

nd

0.1

nd

nd

20

0.5

nd

0.1

nd

nd

nd

nd

21

0.6

nd

0.1

nd

nd

nd

nd

_22_ 23

nd

1.1

nd

nd

nd

0.2

nd

nd

nd

nd

nd

nd

nd

0.1

0.5

nd

nd

nd

nd

nd

0.1

nd

nd

24

97

0.2

nd

nd

0.1?

nd

nd

0.5

nd

nd

25

52

0.3

nd

nd

nd

nd

nd

4.2

nd

nd nd

26 2.5

_?z_ 28 29 30 31

0.4 _68_ J3_

nd

0.5

nd

0.5

nd

nd

5.3

nd

0.5

0.4

nd

nd

nd

nd

0.1

nd

nd

0.3

0.1

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

0.3

nd

nd

0.8

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd nd

32

1.3

nd

nd

nd

nd

nd

nd

0.2

0.3

33

0.1

nd

nd

nd

nd

nd

nd

<0.1

nd

nd

nd

0.5

nd

nd

nd

nd

<0.1

nd

nd nd

34 35

_91_ 0.6

nd

nd

nd

nd

nd

nd

nd

nd

36

0.1

nd

nd

nd

nd

nd

nd

nd

0.2

nd

37

1.5

nd

0.3

nd

nd

nd

nd

nd

nd

nd

38

0.1

nd

nd

nd

nd

nd

nd

nd

nd

nd

39

0.3

nd

nd

nd

nd

nd

nd

0.6

0.2

nd

40

0.3

nd

nd

nd

nd

nd

nd

0.7

0.1

nd

41

1.2

nd

0.26

nd

nd

nd

nd

nd

nd

nd

42

0.03

nd

nd

nd

nd

nd

nd

0.24

nd

nd

43

0.44

nd

nd

nd

nd

nd

nd

0.2

nd

nd

44

0.36

nd

nd

nd

nd

nd

nd

nd

nd

nd

45

89.5

46

0.4

nd

7.37

nd

nd

nd

nd

0.06

nd

nd

0.04

nd

nd

nd

nd

nd

0.15

nd

nd

3.73

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

47

62.7

0.27

48

63.7

0.18

49

70.5

0.26

26_ 1.93

are as indications. The of less than 0.1% are not reliable, and Concentrations given only they long calcareous-silicate phases The of Si. Minute skeletal crystals of magnetite contain mainly Fe but (Ca>Si) are fayalite-like. long-dark phases consist predominantly also Si, Ca, and Ti. Cu prills coexist with mattes. Limited and Si, S, Ca, Ba, and Fe are idiomorph phases exist inwhich Ti predominates, are not determined. also included. Single grains of barite "nd" means The abbreviation apparent. (BaS04)

Notes:

APPENDIX

296 The

49 analyses

Sample M

listed inTable

18-3-(B).

B.2 were

Slag containing

performed

on 10 samples. and one prill of

prills of copper

gold 1. Gold

grain by elements 2. Bulk analysis by oxides 3. Glassy matrix by oxides 4. Sulfide rim surrounding copper prill (by elements) 5. Copper prill by elements 6.Magnetite by oxides 7. Glassy matrix by oxides 8. Fayalite by oxides 9. Copper

prill by elements

Sample M 18-3-(A). Slag containing 10. Bulk by oxides 11. Copper 12. Copper

copper prills

prill by elements prill surrounded by matte elements by 13. Magnetite crystal by oxides 14. Glassy matrix

rim, denser,

by oxides

Sample M 18-3-(C). Slag containing 15. Bulk by oxides 16. Copper prill by elements

copper prills

17. Copper matte 18. Glassy matrix

by oxides oxides by by oxides containing

19. Fayalite 20. Fayalite

Sample M 19-1-(B) 21. Bulk analysis by oxides 22. Copper prill by elements 23. Magnetite by oxides 24. Copper prill by elements 25. Matte by elements Sample M 19-1-(C) 26. Copper prill by elements 27. Bulk by oxides 28. Magnetite

by oxides

Sample M 19-1-(A) 29. Bulk analysis by oxides 30. Matte by elements 31. Matte

by elements

Sample M 19-1-(D) 32. Bulk analysis by oxides 33. Magnetite crystal by oxides 34. Copper

prill by elements

high Ca

B

"clearer" Cu prill

SEM/EDAX

ANALYSIS

Sample N-17-2-(A) 35. Area of slag composed analysis by oxides

297

of glassy matrix

and "fayalites," bulk

36. "Idiomorph" fayalite by oxides, elongated 37. Glassy matrix, bulk analysis by oxides 38. Magnetite crystal by oxides 39. Ceramic 40. Ceramic

part by oxides part by oxides

Sample N-19-2-(D) 41. Glassy matrix, bulk analysis by oxides 42. Ceramic area, bulk analysis by oxides 43. Vitrified area, bulk analysis by oxides 44. Skeletal magnetite, spot analysis by oxides 45. Copper prill by elements Sample N-19-2-(C) 46. Bulk analysis by oxides 47. Matte A, analysis by elements 48. Copper prill, analysis by elements 49. Matte

B, analysis by elements

crystals

C

APPENDIX

Lead

and

Isotope of

Analyses

Slags

Chemical from

Chrysokamino byZofia S tos andNoel Gale

are not uncommon in the Aegean, Although slag heaps of various sizes are in the found of relevant mineral deposits. The they generally proximity mode of occurrence of copper slag at Chrysokamino thus seems unusual. most the of copper slags has been reported Perhaps frequent appearance ores are also found. relatively rich copper copper slag heaps are scattered in Many undated (but probably Archaic) in Phtiotis, where copper was mined the region of the Othrys Mountains not so long ago in Limogardion and Sfaki. On the Cycladic islands of and both ancient and copper slag heaps Seriphos, copper ores Kythnos can still be found.2 However, are present on the scatters of copper slag from northern

Greece1 where

island of Kea both at Kephala3 and below the church of Hagios Simeone,4 and copper ores on this island are rare to nonexistent, with some iron ores less than 5% copper. The copper extraction site at Chrysoka containing mino

seems,

contrary

to Mosso's

as we

claim,5

to have

no

nearby

confirmed when we first visited

occurrences

of

the site in 1983.

copper minerals,6 The general position of the site, which is set on an elevated coastal cliff, is quite similar to the locations of the slag heaps on Kythnos and Seriphos, but its location seems not to have been selected for its proximity to the source

of

copper

ore.

THE NATURE AND CHEMICAL ANALYSIS OF THE SLAGS AT CHRYSOKAMINO The slag heap at Chrysokamino appears first to have been described by Hawes7 and independently Mosso.8 Mosso, however, does not seem to by have visited it, basing his account on firsthand observations by Hazzidakis. in his Later, Mosso repeated his earlier account, almost word-for-word, on book Mediterranean archaeology.9 Davies had little to report about ore

1. 1986. Papastamatak? et al. 1985; Gale 2. Gale

4. and Stos

Gale 1989; Gale 1989; Stos-Gale 1989. 3. Keos

I.

Caskey 5. Mosso 6.

court

et al. 1988. 1910,

pp. 289-291.

Betan 1968, pp. 50-51; Branigan et al. 1999, see also the p. 352;

report in this volume. geological et al. 1908, p. 33. 7. Hawes 8.Mosso 1908, pp. 518-521.

9.Mosso

1910, pp. 289-291.

APPENDIX

3??

C

on Crete10 and presented hardly any information of substance about in his it in passing in book,11 though he did also mention Chrysokamino an earlier article,12 inwhich he cave as aworked-out the copper accepted deposits

mine. Although Fiedler described the Skouries site on Kythnos,13 he did not include Crete in his admirable and comprehensive survey of matters was in since and botanical Crete then under Egyptian Greece, geological rule and did not become

part of Greece until 1913. in 1983 as part of our long-standing We first visited Chrysokamino work on the archaeometallurgy ofMinoan Crete. We were guided to it by an inhabitant of the elderly nearby village of Kavousi. The site, described at length elsewhere in this report, is located on a cliff top exposed to the wind, much like the Bronze Age copper smelting site of Skouries on Kyth amount of copper slag with small nos.14 The site contains a considerable in In the Oxford, we ascertained through electron copper prills. laboratory some of the copper analysis that prills microprobe percent of arsenic, a feature not inconsistent with the in the Early Bronze Age and directly comparable to for some of the copper prills in the Early Bronze Age

up to several arsenical copper used the analyses reported

contain

copper slags at Skou in the copper prills of slags from

ries on Kythnos.15 The arsenic content is confirmed by the independent Chrysokamino analyses reported in this volume by Bassiakos A feature of the Chrysoka (App. B). distinguishing mino slag heap, first published by Mosso in 1908,16 is the omnipresence of thick coarse pottery sherds perforated with holes 1.5-2 cm in diameter, often fused and slagged on the interior surfaces, which are clearly the broken remains of furnace chimneys, not crucibles asMosso thought. at the Mosso that he believed dated slag heap Chrysokamino implied a sherd of EM II to Minoan times.17 Schachermeyr18 reported finding Faure believed the site postdated "mottled ware"19 at this site. However, was the Middle probably relatively recent, perhaps related to Ages and the work of gypsies.20 He based his opinion on the state of vitrification of the slags, the use of a siliceous/calcareous flux, and the assumed use of ore (for which he had no evidence). chalcopyrite Branigan, who was perhaps influenced by Faure, visited the site in 1967 and briefly described both the cave and the slag site, wrongly accepting the coarse ware sherds as crucible fragments.21 His discussion veered between as the of the simple melting characterizing operations at Chrysokamino ores. no at He found decorated sherds copper and the smelting of copper sherd found by the slag site and was unaware of the EM II Vasiliki Ware so he had no conventional means of and archaeological Schachermeyr, the metallurgical operations. Chemical analyses of the slag from made by Sargent for Branigan, suggested a high recovery of Chrysokamino, copper metal, revealed the presence of calcium, and suggested to Sargent dating

that a temperature of about 1150?F (621?C) or more was attained "during the smelting operations."22 All these features suggested to them a relatively modern date for the slag, "not earlier than about the 12th century AD." was a site used was led to conclude that Chrysokamino by "itin Branigan erant bronze-workers which copper they probably using ready-smelted carried with them," and that they may have belonged to aminority group, or thereabouts." Since the Chalkiades, "operating in the 14th century AD

10. Davies

1935,

pp. 266-268.

11. Davies

1935, pp. 7,113,264, 270 (n. 1 gives an analysis of the slag the calcium which misses by Desch, content). 12. Davies

1932,

13. Fiedler

1841,

p. 987. p. 97.

14. Fiedler 1841; Gale et al. 1985; Gale

and Stos-Gale

1989;

Stos-Gale

1989,1998. 15. Gale

et al. 1985,

p. 90.

16.Mosso

1908, p. 519, fig. 24; 1910, p. 291, fig. 164. 17.Mosso 1908,1910. 18.

Schachermeyr 19. Vasiliki Ware;

1938, p. 473. see Betancourt

1979. 20. Faure 21. Branigan 22. Branigan

pp. 47-48. 1968.

1966,

1968,

p. 50.

LEAD

ISOTOPE

AND

OF

ANALYSES

CHEMICAL

SLAGS

3OI

the melting point of copper, we suspect that 1150?C that itwas judged too high a temperature to be attained and intended, in Bronze Age furnaces. At the time when Faure and Branigan wrote, the study of ancient cop

1150?F

iswell below

was

per smelting and slag was in its infancy. The dating to the Early Bronze on Kythnos,23 however, along with the Age of the Skouries slag heap proof that slags from Skouries had melting points between about 1050?C calcium,25 shows that evidence based on the did not physical properties of the slag from Chrysokamino a rule out Minoan date for it.Our examination of the Chrysokamino slag a few ores samples of copper heap produced lying intermingled with the and 1200?C24 and contained and chemical

ores were all oxidized ores, mostly malachite, and suggest that slags. These to the smelting of oxidized copper ores. This this slag heap bears witness examination of suggestion is strengthened by the petrological microscopic a number of mala these slags by Bassiakos (Apps. B and F), who found chite inclusions in the thin sections examined. A sample of ore found by on the beach

near Chrysokamino to be of malachite, in Oxford, examination and may suggest sea. ores arrived at Chrysokamino by

Betancourt

at Agriomandra

on proved, that copper

SOURCES OF COPPER ORES ON CRETE IN THE ARCHAEOLOGICAL LITERATURE in European cop sources but the of this metal b.c.,26 and axes dating to the EM period are not

has shown that the island of Crete

Branigan

led the way

in the 3rd millennium

per metallurgy used for the numerous immediately copper mine

daggers obvious. Mosso made

a definite

but ill-founded

claim that a

existed on the island of Gavdos

and in a cave near Chrysoka either site, he wrote that ore containing somewhere there; at the same time he admitted that none of the rock samples sent to him by Hazzidakis from the walls of the cave contained any copper ore. Faure, from his personal observations, concluded that the cave was natural, rather than being amine,

he never visited mino.27 Although was found of "protoxide copper"

and

it never

that

contained

copper

minerals.28

30 years ago, several papers appeared implying that theMinoans were mining copper ore on Crete. Faure listed 22 sites on Crete where he asserted that copper minerals were found.29 Typical descriptions by Faure of such occurrences were similar to the one that he gave of a "copper min About

eralisation" near Kritsa At

(Mirabello

one and a half hours' walk

Bay): to theWSW

of Kritsa

one finds a

great variety of eruptive rocks, bordering the high plateau of Katharo_In such a context the geographer Raulin is the first to in a Creta have described in 1845 grains of pyrites disseminated ceous contact deposit... 23.

Stos-Gale

24. Gale 25. Gale

1998, et al. 1985, et al. 1985,

pp. 719-720. pp. 86-89. p. 87,

table 2.

26. Branigan

1968,1974;

I have seen traces of nickel

cf. Pendle

bury 1939, p. 280, and Caskey 1964. 27.Mosso 1908, pp. 518-521; 1910,

pp. 289-293. 28. Faure 29. Faure

1966,

sulphide. For

pp. 47-48.

1966,1980.

APPENDIX

302

C

the presence of many years local prospectors have mentioned carbonates of copper to the south of Katharo. For 80 years the or sent peasants of Selakano and of Kritsa have sold to foreigners to the Archaeological and votive axes Service bronze weapons site, probably to the NE of Selakono. coming from an unknown One can hardly doubt the existence of one or many centres for the treatment of metal between Selakano and the Katharo whilst the ... mountain is rich in toponyms describing furnaces, generally of charcoal; but everyone knows that at all times one has roasted and reduced

impure minerals

with

this fuel.30

not describe any real features of speculative and does or real, as such relatively galleries, spoil heaps, exploitation or ore minerals observed indeed of of any copper minerals, large, outcrops by Faure other than those of iron and nickel.

All

this sounds highly

ancient mineral

Branigan reassessed the hypothesis of the existence of Cretan copper ore sources in the early seventies when he conducted several seasons of several other survey in south central Crete and described archaeological occurrences of copper ore deposits.31 Apart from the occurrence at Chryso stomos near Laseia, which and azurite (as does indeed contains malachite are in west Crete),32 other mineralizations the occurrence at Sklavopoula as consisting of "copper minerals in schist." In claiming mostly described the discovery of an Early Bronze Age metal source in Crete, Branigan at the site of Fournou some describes pebbles" found "copper-bearing were made and of these Korifi. Apparently, chemical analyses pebbles "the analysis of two of the pebbles produced 1% and 2.2% of copper."33 Oh we were unable to find any copper-bearing pebbles visiting this site in 1983, at the location described, nor did we see any in the conglomerate cap of to a Brani the nearby hill, though heap of pebbles otherwise answering or was s furnace No found. crucibles, gan description pieces fragments, of slag were found. It seems probable that, if a copper source ever existed small and of no importance here, it was extremely except for strictly local

needs.

the basis of their surveys, both Branigan and Faure concluded that there are enough copper ore deposits on Crete to have satisfied Minoan and Rapp in the Early Bronze Age. Muhly for this metal demands on Crete in 1974, which discounted many of the al conducted a survey On

leged copper deposits reported by Faure; they stated that they found that in the literature are repre "a number of the copper deposits discussed are some on and rocks stains sented only by green serpentine rather than that Crete probably should not be considered copper."34They concluded a of copper in the Bronze Age. Becker, in the course of major producer extensive fieldwork on Crete conducted primarily to establish soft stone that many alleged copper sources on Crete have sources, also demonstrated been incorrectly identified.35 Finally, according to a Cretan commercial we have corresponded: "The occurrences mining engineer with whom to small samples our Faure refer mutual friend Professor cited by mostly found as float. In many cases the alleged site has produced no further evidence."36

30. Faure

1966,

p. 49

(translation

by

the authors). 31. Branigan

Blackman

1971;

and

Branigan 1975,1977. 32. This is confirmed in both cases our own by later.

in 1983

observations

33. Branigan

p. 12.

1971,

34.Wheeler, Maddin, 1975, p. 32. 35. Becker

1976.

36. This

information

a letter Diallinas, Aegean

is well

andMuhly

comes

in our archives who

and

from

from Michael known

archaeologists,

Gale inOctober 1977.

in

to many

reply

to Noel

LEAD

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

303

THE GEOLOGISTS' VIEW OF COPPER DEPOSITS ON CRETE of southeastern European mineral deposits does any copper ores at all on Crete,37 while the Greek section of the later UNESCO map of Europe mentions metallogenic only the cop on Crete at note mineral of The per deposit only deposits Sklavopoula.38 are current in the literature mentioned gypsum and iron. The geological from IGME working on Crete in the 1990s Greek economic geologists

The

standard handbook

not mention

denied

the existence

However,

itmust

of all but traces of copper ores on the island. be admitted that there is often a certain lack of un

derstanding among modern economic geologists concerning the relatively of five thousand years ago. The limited requirements of the metallurgists interest in the nonferrous metallic mineral deposits of the world modern on the deposits of Australia, Siberia, America, and the Far East. Over fifty years ago, Bateman wrote: tends to be focused

Africa,

was widely early times until 1800 copper produced in small was to 1801 1810 From the annual world quantities. production one to than less month's production only 18,200 tons, equivalent The tremendous growth in the use of some present-day mines....

From

of copper

is indicated by the fact that of the total world copper in in the the last 100 years about 80 percent was mined produced last 25 years-Annual from about production normally ranges 2 to 2.5 million

tons of metallic

copper.39

are quite mean copper deposits of the whole of the Mediterranean one thinks on such a scale. when modern economic ingless Consequently,

The

some cases be somewhat geological handbooks and maps may in misleading comes to it ancient the search for metal sources. when With

the help ofMichael Diallinas, Kostas Zervantonakis, and many we a in from the IGME office conducted Chania, helpful geologists on occurrences in of the mineral Crete 1979, relatively systematic survey 1983, and 1987. We were especially interested in the occurrences described

most

by Faure and Branigan, and thank both for their generous correspondence with us. Our main aim was the collection of ore samples for lead isotope on Crete are on two sites analyses. Copper minerals definitely present one to is that of Chrysostomos, known the archaeologists: southwest of on the central south coast, and the other is inwest Andiskari Sklavopoula inmetamorphic Crete. The copper carbonate mineralization rocks near the ruined chapel of St. John Chrysostomos is quite interesting. Faure men a in tions that 1952 Greek firm attempted to mine copper ore there, and 37.

Dunning,

Mykura,

and Slater

1982, with report byMarinos p. 237. 38.

1982,

and Koukouzas

Anastopoulos

1984. 39. Bateman

1946,

p. 479.

40. Faure 1966, p. 52; Branigan 1968,

p. 51.

and azurite, aswell as limonite, Branigan confirms the report.40Malachite are still visible in the walls of the large open pit that remains from this venture of the 1950s. It is difficult to assess how much copper mineraliza tion might have been present on this site some four thousand years ago. It was ever very extensive, but some is very unlikely that the mineralization oxidized copper ores might have been visible here and might have been used to a small extent in antiquity. The occurrence is surrounded byMinoan

APPENDIX

304

C

us around and later sites; on one occasion, J.A. MacGillivray, walking with several noticed Minoan sherds and the remains of Early Chrysostomos, a tholos tomb on a field no traces above the of copper just pit. However, in smelting, slag, or ancient galleries were found. The copper occurrence west

at Sklavopoula is high in the (Selinou), south of Rethymno, a station. A below radar mountains, just military comparatively recent min ores in here left of 196241 ing activity spoil heaps mainly iron/manganese with some dispersed oxidized copper minerals. The site has no evidence Crete

of any ancient exploitation of these minerals, and the geochemistry of the ores suggests that copper here is a minor iron of component only relatively ores. No were found slags anywhere in the region. Iron deposits have also to a small extent southwest

been exploited copper

also

collect

some

minor

occurs

as a minor

of

copper

of

component

samples of copper minerals

occurrences

of Chania.

minerals

the

ore,

In some of the mines, and we

at Kambanos.42 also

occur

relatively

were

able

to

Some other very near

the mine

of Chrysostomos, south of Hagia Triada and Phaistos on the south at Lasaia, Lebena, and Miamou. The results of our survey, combined with all reliable geological mation about mineral deposits on Crete, indicate that a significant extraction (copper, lead, and silver) from the local ores in Bronze Age is very unlikely. To widen our survey for possible metal extraction

coast, infor metal times

slags on Crete, we enlisted the of the Nowicki, help archaeologist Krzysztof in Crete. He showed us the who surveyed large areas of the mountains site near Kato Zakros, in the Valley of Death,43 but no copper slags were no copper present. Moreover, slag heaps have been located other than course that at Chrysokamino, though of slag layers have occasionally been sites such asMalia.44 It may be worth found in the excavations ofMinoan as Kato Zakros, which pointing out here that the structures at such places some have as Bronze are to sought identify Age copper furnaces, clearly not designed for this purpose.

CHEMICAL ANALYSES AND DATING OF SLAGS FROM CHRYSOKAMINO is the only site found on Crete so far where copper smelt Chrysokamino ing, the extraction of copper from ores rather than the making of artifacts from copper metal, was practiced. In 1994, we were able to obtain two thermoluminescence (TL) dates on two different examples of the charac teristic perforated furnace sherds that we had collected in 1983 and in the b.c.: 1990s from Chrysokamino.45 Both dates fell into the 3rd millennium 2420 ? 345 b.c. and 2710

? 365 b.c.46 These

dates substantiated

of the antiquity of Chrysokamino, which of Vasiliki Ware found there by Schachermeyr.47 indication

the earlier

level, similar to copper prills in the Bronze Age

1966,

see

iron;

1935,

1985.

dates

M.Tite,RLAHA, 46.

for

p. 47, n. 1.

1966,

44. Poursat 45. The

was

in Davies

42. Kambanu

47.

slags from

mining p. 57.

p. 268. 43. Faure

came from the sherd

sections of many of the Chrysokamino P?trographie slags revealed the numerous Chemical of metal presence copper prills.48 analyses of these an electron in fitted with wavelength made Oxford microprobe using prills showed that many of the prills contained arsenic at the dispersive XRF several percent

41. The Faure

Stos-Gale Schachermeyr

are courtesy Oxford. 1998,

of

p. 721. 1938,

p. 473.

48.We first had a sample of the slag

from Michael

Diallinas

when

we

visited him in his offices inHerakleion in 1978, and we subsequently more we visited samples when

in 1983.

collected the

site

LEAD

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

305

Skouries on Kythnos.49 This finding has now been confirmed and extended in this volume (Apps. B and F). by Bassiakos The pioneering work of Bachmann showed that useful information about the copper smelting techniques used at a particular site can be obtained from bulk chemical analyses of the slags and their interpretation by ther modynamic phase diagrams, though such inferences should, if possible, be aided by mineral identification based on petrography of the slags.50 Experi also allows mental work on silicate melts of various chemical compositions inferences to be made about slag melting points and furnace temperatures, aswell as about slag viscosities and densities, in turn permitting estimates of the efficiency of separation of metal from the slag phase in the furnace.51 Several slag samples collected from the Chrysokamino heap were, therefore, 22 and for elements, using the technique chemically analyzed powdered of inductively (ICP-AES). coupled plasma atomic emission analysis This work was undertaken at the Royal Holloway College, London. Analy ses for arsenic content were by wavelength dispersive X-ray fluorescence also undertaken by the analytical laboratories of the British Geological were checked Survey, Keyworth. AH analyses against concurrent analyses of standards. The chemical analyses are presented international geochemical in Table C.l. It is evident that, as noted by Faure and Branigan,52 these slags

are

quite rich in calcium, although also contain calcium at medium

Kythnos The bulk chemical

the EBA

slags from Skouries levels.53

on

high especially the relatively high calcium seem to be contents, and the ternary diagrams into which they classify, in accord with the p?trographie studies (Apps. A and B), which reveal compositions,

to be present in the (including augite), fayalite, and magnetite samples of loose slag from Chrysokamino. Bachmann54 has shown how such chemical analyses of slags may be

pyroxene

terms of the quaternary phase system interpreted in CaO?FeO?A1203? to include other elements, he expanded to: Si02, which, CaO{+BaO+Na20+K20]?FeO{+MnO+MgO}?A1203?Si02 basic system plots in space as a tetrahedron, which projects into four in order to include wollas and Zacharias, ternary systems, but Bachmann tonite, found itmore useful towork in terms of five ternary systems, which

The

include

the minerals

and anorthite, gehlenite, wollastonite, A1203, Si02, are neces with the that others.55 FeO, along They specify computations sary to interpret chemical analyses of slags in terms of these five ternary have programmed these computations diagrams. We using MathCad, the calculation

of the basicity number, and have further used to program the computations necessary to use Shaws method as a to compute slag viscosities function of temperature.56

including MathCad

49. Gale

et al. 1985,

50. Bachmann

p. 90,

table

3.

1980.

51. Shaw 1972; Bottinga andWeill 1972; Bottinga,Weill, and Richet 1982; Bottinga, Richet, andWeill 1983. 52. Faure 1968,

p. 50.

1966,

p. 48; Branigan

53. Gale

et al. 1985,

54. Bachmann

1980.

55. Bachmann

1980,

p. 87,

pp.

table 2.

120-131.

56. Shaw 1972. Subsequently we found that Ford (1992) has published on the Internet

a convenient

program

(CHEMCAST), which computes a

number of basic chemical/petrological and includes viscosities parameters smethod and densities using Shaw

following Bottinga,Weill, and Richet 1982 and Bottinga, Richet, andWeill 1983.

C

APPENDIX

3?6

TABLE C.l. CHEMICAL SiO^

Sample

^A

OF CHRYSOKAMINO

ANALYSES MgO

?A

CaO

NaO

SLAG

K<0

TiO^

MnO

Ba

CHR2

33.91

5.37

19.68

1.33

30.53

0.87

0.85

0.26

0.19

5960

CHR3 CHRYA CHRYB

42.37

5.51

34.64

0.70

11.88

0.72

0.53

0.23

0.21

920

38.06

5.38

23.08

1.22

25.70

1.19

0.89

0.25

0.22

6580

32.84

7.45

34.40

1.42

17.75

1.52

0.89

0.36

0.34

6460

CHRYC

34.05

7.06

34.16

1.53

17.30

1.29

0.96

0.33

0.32

4660

CHRYD

27.59

5.50

40.08

1.15

17.93

1.41

0.63

0.27

0.34

8180

CHRYE

41.56

6.14

19.53

1.09

23.44

0.83

0.94

0.28

0.12

1521

CHRYF

28.44

3.76

37.08

0.81

10.65

0.56

0.62

0.16

0.08

>20000

CHRYG

47.53

6.22

15.64

1.31

22.32

0.95

0.95

0.30

0.25

14240

CHRYH

45.48

6.49

15.46

1.55

23.44

0.99

0.97

0.31

0.27

14280

CHRYf CHRYK CHRYL

35.22

7.49

30.16

1.36

16.24

1.57

0.90

0.35

0.85

4570

31.59

5.47

42.04

1.19

14.12

0.33

0.80

0.26

0.14

532

46.62

6.32

15.33

1.35

22.70

0.95

0.94

0.30

0.25

13630

CHRYM

35.93

5.04

31.56

1.23

18.71

0.48

0.78

0.26

0.16

6020

of the most useful parameters that can be obtained from assign to the via their chemical compositions, ing slags, appropriate ternary phase as a reasonable estimate of the lower is the temperature, diagram liquidus One

of a suite of attained in the production that the application of the crude Si02? FeO?CaO ternary diagram, much used in the past for this purpose,57 predicts far too high liquidus temperatures, sometimes by many hundreds of degrees, and that Bachmanns approach produces liquidus temperatures limit of the furnace temperature slags. It has been shown

ancient

very much

nearer the truth.58 Another

equations, as derived

in terms of the fractional

method for estimating the liquidus one is and Van temperatures of slag silicate melts developed by Nathan Kirk as an initial part of their attempt to devise a computer model for frac in geological magmas.59 They derived empirical tional crystallization from

its chemical

of a number

cation

composition for the estimation

analysis, in a multicomponent of minerals In comparing the predictions

of the magma of the liquidus

silicate liquid temperature at atmospheric of the Nathan pressure. and Van Kirk equations with an analyzed set of 975 glasses produced at known temperatures and oxygen fugacities, Ford found that the equations to better than 40?C, which predicted the measured liquidus temperatures is perhaps adequate should bear in mind

one for archaeometallurgical purposes.60 However, and Van Kirk equations that the empirical Nathan

are based on data sets directly to geological melts, and applicable chiefly more iron-rich in of for the stand need recalibration composi they really tions of typical archaeometallurgical the minerals treated slags. Moreover, and Van Kirk do not include some that are more applicable to by Nathan in slags.61 Table C.2 shows the results of applying the Bachmann tions to the analyzed The table slags from Chrysokamino. those found

57. E.g., Conophagos

1980;Milton

et al. 1976. 58. Gale

et al. 1985.

59. Nathan and Van Kirk 1978. computa shows the

60. Ford

1981.

61. Bachmann

1982.

LEAD

TABLE

C.l

Sample

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

307

(cont.) Co

Cr

Cu

Li

Ni

Se

CHR2

16

90

5980

28

47

CHR3

21

127

12380

17

21

CHRYA

87

101

5820

38

60

CHRYB

46

137

19080

30

523

CHRYC

35

120

11660

28

258

CHRYD

31

130

14150

21

353

CHRYE CHRYF

24 40

109

23230

30

60

16470

CHRYG

23

107

CHRYH

18 55

114

Sr

Zn

Zr

Pb

465

104

30

127

63

18

231

540

26

1915

58

721

596

191

29

194

62

50

12

304

170

25

329

82

66

12

289

110

26

142

81

36

11

270

127

20

246

63

61

83

250

215

24

589

62

92

19

43

1225

101

17

861

42

56

16940

30

51

497

214

30

675

72

47

7870

31

32

550

230

32

582

74

35

152

22850

28

602

206

154

28

335

78

45

13

83

7480

26

36

233

282

39

267

67

40

CHRYL

20

110

10670

30

40

502

218

30

603

75

37

CHRYM

29

87

15630

47

169

365

141

39

700

70

96

CHRYJ CHRYK

Values

for oxides

are percentages;

values

for elements

12

15

are ppm.

calculated

selection

quotients Q^, Q^2, Q^3, the phase diagrams selected appropriate for each slag, the ages of the appropriate principal mineral phases, the basicity numbers, the densities using the approach of Bottinga and colleagues,62 the viscosities computed using Shaw's method, and the the method devised Nathan and liquidus temperatures computed by by as

Van Kirk. The

Bachmann approach shows that the Chrysokamino slags two into ternary phase diagrams, with the majority falling into the classify Anorthite?Wollastonite?FeO diagram and three falling into the Anor diagram. Figure C.l shows the 12 slags that plot in the thite?Si02?FeO or the Anorthite ternary region Hercynite region of the AN?WO?FeO it and shows that the diagram, slag liquidus temperatures probably fall in ? the 1200-1250?C FeO region. The three slags that fall in the AN?SiO diagram (Figure C.2) suggest liquidus temperatures of about 1200?C. On the other hand, the calculations of liquidus temperatures using the ap and Van Kirk, reported in Table C.2, suggest slightly proach of Nathan lower liquidus temperatures of about 1120-1170?C, though these estimates are based on the minerals

it is probable that augite and olivine. Because the operating temperatures of ancient smelting furnaces were typically at seem least 100?C above the melting temperature of the slags,63 itwould that at Chrysokamino the furnace temperatures were in the vicinity of

62. Bottinga,Weill,

and Richet 1982; Bottinga, Richet, andWeill 1983. 63. Freestone

64.Merkel and Wincierz

1988.

1983,1990; Bamberger 1990.

65. See Craddock

66. Lang 1978.

et al. 1985.

this agrees with experimental of copper reconstructions 1250-1350?C; in shaft furnaces.64 smelting simple a minimum Even assuming furnace temperature of 1200?C, the are of the order of 20 poise, slightly higher than the Rio slag viscosities lead slags,65 but still not very high. Combining the calculations of in Table C.2 and the of for the density of 7.8 cm"3 g slag density figure one can use at to molten Stokes' law 1200?C,66 copper compute the rate

Tinto

TABLE C.2. RESULTS OF APPLYING FROM CHRYSOKAMINO

COMPUTATIONS

THE BACHMANN

TO THE ANALYZ

Phase Sample

Q,

Q2

Qs

CHR2

0.093

0.092

1.012

CHR3

0.077

0.235

0.326

Density*

Diagram 3?

AN

No.

Basicity

FeO

SiO,

1.383

38.39

0.9

WO

FeO

55.02

23.93

21.05

2.98

23.60

2.96

3.01

34.43

27.18

at 1200?C

AN

CHRYA

0.083

0.107

0.775

1.185

50.32

26.08

CHRYB CHRYC

0.134

0.206

0.649

1.332

42.56

22.57

34.87

3.12

0.122

0.202

0.604

1.278

41.81

23.42

34.77

3.10

CHRYD

0.117

0.152

0.772

1.733

39.48

19.45

41.07

3.28

CHRYE

0.087

0.136

0.639

0.951

50.38

29.17

20.45

2.87

CHRYF

0.078

0.174

0.447

1.366

CHRYG

0.077

0.141

0.547

0.78

51.31

32.24

16.45

2.77

CHRYH

0.084

0.14

0.599

0.84

52.35

31.01

16.64

2.79

CHRYJ

0.125

0.224

0.558

1.138

42.94

24.89

32.17

3.03

CHRYK

0.102

0.202

0.506

34.97

22.25

42.78

3.25

CHRYL

0.08

0.141

0.565

0.797

51.80

31.87

16.33

2.77

CHRYM

0.083

0.14

0.591

1.207

42.02

25.36

32.62

3.07

by the methods as Loge(Viscosity)

given as

*Density **Viscosity ^Liquidus

computed quoted

temperatures

computed

from

3? 2?

by Bottinga, computed

43.67

1.421

Weill,

and Richet

by the methods

the formulae

39.72

given

given

by Nathan

43.23

17.05

13.54

42.79

1982; Bottinga, by Shaw

Richet,

1972.

and Van Kirk

1978.

3.24

andWeill

1983.

Ln(Vi at 1

LEAD

C.l. The Figure Wollastonite-FeO diagram 12 slag

and samples

the

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

309

Anorthite ternary locations from

phase within

it of

Chrysokamino Wollastonite [WO]: CaS?03 Anorthite [AN]: CaAI2Si208 Hercynite: Fe2AI204 W?stite: FeO Olivine (Fayalite) Fe2Si04 Melilite: Ca2AI2Si207

FeO

WO

AN

C.2. The Anorthite-SiO Figure FeO and ternary phase diagram the location it of three within slag samples

from

Chrysokamino

APPENDIX

3io

C

of separation of molten copper metal from molten slag within the furnace. For a spherical copper droplet of 1mm in diameter, Stokes' law gives a rate cm per minute, which for a copper of settling through the molten slag of 8 mm in diameter to of 0.1 0.08 cm per minute. Given the droplet drops furnaces (Chaps. 7 and 14) and the relatively small size of the Chrysokamino fact that ancient copper smelting furnaces were operated for several hours,67 copper droplets of 1mm in diameter and above could sink through the slag many times over, while very small copper droplets of the order of 0.1 mm in diameter or lesswould be trapped in the slag, as observed (seeApps. B and F). Such very small copper prills could not easily be broken from the slag by mechanical means, nor would this be necessary, since separation of copper mm in diameter from the molten slag by gravitational droplets larger than 1 settling during operation

of the furnace would

have been very efficient.

LEAD ISOTOPE ANALYSES OF COPPER AND LEAD ORES FROM CRETE ores collected our surveys isotope analyses of copper and lead during in Crete are listed inTable C.3, while lead isotope analyses of the Chryso small number of kamino slags appear in Table C.4. The comparatively

Lead

analyses of minerals from Crete, 28 in total, reflects directly the scarcity of and Sklavopoulou, copper ores in the few occurrences. At Chrysostomos on as is the surface of the host rock, though visible oxidation green copper and azurite are present on localities richer ores bearing malachite ore the rocks the dumps from 20th-century exploitation. At Sklavopoulou, some are iron with with and minerals, manganese copper impregnated host rock contains some the metamorphic carbonates. At Chrysostomos, in both

at lead ore occurrence and dispersed copper carbonates. The is relatively small, with exposed veins of fine-grained Ano Varsamonero neutron We galena. analyzed two samples of this galena using instrumental activation: the silver content in the samples was 150 ppm and 240 ppm, limonite

the gold content in both around 0.1 ppm. The silver content at this level would have been too low for extraction in the Bronze Age,68 so that the cannot have been aMinoan silver source. In deposit at Ano Varsamonero some of the iron mines in western Crete, it is possible to find relatively and a few such copper-rich minerals were collected rich copper minerals, from Kambanos

as well

as

Sklavopoulou. not able to find equally good copper minerals Unfortunately, was found at La in any other mines, though weak copper mineralization on near the central southern coast of the village of Lenta saia and Lebena Crete. The currently available lead isotope data for Cretan ores are plotted on two mirror diagrams in Figure C.3. we were

ores show that the copper ores in south isotope analyses of these and western Crete have very central Crete and those from Sklavopoulou Lead

1990. 67. Bamberger and Wincierz 68. Gale and Stos-Gale 1981a,

1981b;

Gale,

1984;

Pernicka

and Davis Stos-Gale, et al. 1983.

LEAD

ISOTOPE ANALYSES

TABLE C.3. LEAD FROM CRETE Sample

ISOTOPE

Occurrence

AND

CHEMICAL

OF

SLAGS

)7Pb/206Pb

206pb/204pb

CST6A

Chrysostomos

Malachite

2.08110

0.83980

18.618

CSTA1

Chrysostomos

Malachite

2.08187

0.83880

18.652

CST1A

Chrysostomos

Malachite

2.08014

0.83992

18.616

Chrysostomos

Malachite

2.08280

0.83828

18.715

CST3 CST4

Chrysostomos

Malachite

2.07765

0.83746

18.682

CHR4

Chrysostomos

Malachite

2.07895

0.84035

18.580

CHR4A

Chrysostomos

Malachite

2.08103

0.84077

18.599

CHR3

Chrysostomos

Malachite

2.07834

0.83913

18.573

CST2

Chrysostomos

Malachite

2.08253

0.84080

18.568

CST1

Chrysostomos

Malachite

2.08180

0.84320

18.483

CST5

Chrysostomos

Malachite

2.08381

0.83862

18.692

Kam200

Kambanos

Copper

ore

2.07179

0.83242

18.923

Kam201

Kambanos

Copper

(

2.06447

0.83253

18.950

Kam21g

Kambanos

Copper

ore

Kamdl

Kambanos

Copper

LAS1

Lasaia

Copper

LEB1

Lebena

Copper

MIAMI

Miamou

LAQ_ MLB skavl

2.06773

0.83455

18.879

(

2.06751

0.83312

18.947

ore

2.08554

0.84286

18.544

ore

2.08067

0.84110

18.540

Galena

2.07745

0.84033

18.614

Miamou

Galena

2.07676

0.84019

18.591

Mia

Galena

18.579

Sklavopoulou

Copper

skav2

Sklavopoulou

Copper

skav3

Sklavopoulou

Copper

2.07275

0.83998

ore

2.06482

0.83356

18.935

ore

2.06982

0.83479

18.897

ore

2.06450

0.83396

18.981

2.07143

0.83510

18.833

VARS1

Ano

Varsamonero

Galena

2.09771

0.8511

18.399

VARS2

Ano

Varsamonero

Galena

2.09393

0.85058

18.436

AN1

Ano

Varsamonero

Galena

2.09589

0.85083

18.405

ANV100

Ano

Varsamonero

Galena

2.09519

0.85033

18.394

skav4

Sklavopoulou

Copper

311

OF COPPER ORES AND GALENAS 208Pfr/206p?

Description

ANALYSES

ore

different

lead isotope geochemistry. The galena from Ano Varsamonero is isotopically different from both of these groups. Our earlier lead isotope copper oxhide ingots from Hagia Triada and Kato analyses of Minoan

Zakros69

show clearly that all of these ingots have lead isotope analyses completely different from those of any copper ore from Crete yet analyzed, so that present evidence shows that these copper oxhide ingots were not

made

from copper extracted from any known Cretan known Cypriot copper ores.70 69. Gale

1999.

and Stos-Gale

1986; Gale

copper ores, nor from

70. See Gale 1991,1999; Stos-Gale 1993.

C

APPENDIX

312

2.11

2.10

O

Copper ores, Crete Lead ores, Crete

O

Slags,

Chrysokamino

i

2.09

2.08

-*^

a. CO o

M

2.07

o

~^ 2.06 o

*> o ?O.1%

20Pb/206Pb

2.05

X 0.830

0.840

0.835

0.845

0.850

**?

18.9 18.8

^^r

4

a.
0.855

t 0.1%

*^

18.7 18.6

n a.

?* ?k

18.5

?to

18.4

*

18.3 18.2 207pb/206pb

18.1

0.830

0.835

0.840

0.845

0.850 C.3. Lead Figure tions of ores from

0.855 isotope Crete

composi compared

with those for slags from Chryso kamino

LEAD

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

313

TABLE C.4. LEAD ISOTOPE RATIOS OF PIECES SLAG OF CHRYSOKAMINO Sample

208pb/206pb

207pb/206pb

206pb/204pb

CHR2*

2.05868

0.83189

18.807

CHRYE

2.05793

0.83024

18.860

CHRYA

2.06281

0.8327

18.829

CHRYC

2.05539

0.82963

18.856

CHR3*

2.06808

0.83722

18.640

CHRYD1

2.06829

0.83438

18.754

CHRYB CHRYI CHRYH

2.06776

0.83403

18.809

2.06632

0.83309

18.812

2.06751

0.83432

18.784

CHRYM

2.06715

0.83276

18.868

CHRYJ CHRYG

2.06332

0.83196

18.833

2.06358

0.83228

18.830

CHRYD2

2.06896

0.83454

18.769

CHRYL

2.06953

0.83479

18.876

CHRYK

2.06326

0.83234

18.830

Pieces

were

remaining

in August

collected

marked

Samples

are Isotrace numbers numbers. Laboratory in 1985 with lesser accuracy than the analyzed on the new mass in 1995. spectrometer analyzed 1983. The

an asterisk were

with

samples, which

were

LEAD ISOTOPE ANALYSES OF SLAGS FROM CHRYSOKAMINO In 1983 we mino with

some

slags from the surface of the site at Chrysoka for chemical analysis and for lead isotope analyses to be compared the Cretan ores. The isotopic data obtained for some of these slags collected

is listed inTable C.4. were made in 1985 and published by Stos analyses Gale.71 As Figure C.3 shows, the lead isotope analyses of these 14 slag pieces form quite a compact group, clearly different isotopically from the ores. The analyzed Cretan superficial similarity of the isotopic composition of ores from western Crete with the slags from Chrysokamino, apparent on the 208Pb/206Pb can be quite definitely rejected in view of the diagram, different 206Pb/204Pb ratios of these two groups of ores. The copper ores Two

71. Stos-Gale kamino

1998,

p. 724, Chryso

1 and 2.

72. For Stos-Gale, 73. Gale

the data

for these

and Annetts Gale, et al. 1985; Gale

Gale 1989; Stos-Gale 1998.

ores,

see

1996. and Stos

of these

from Chrysostomos and the occurrences near Lasaia and Lenta are very different in all three lead isotope ratios. On the presently available evidence, some of the slags from Chrysoka mino show identical lead isotope ratios with copper ores from Lavrion,72 while

other Chrysokamino slags have lead isotope compositions match ores from the EC II smelting site of Skouries on the ing copper slags and island of Kythnos73 and copper ores or slags from Seriphos (see Cycladic

Table C.5). The

lead isotope plot for the ores and slags from these sites is

APPENDIX

3H

TABLE C.5. CYCLADIC ORES AND COPPER WITH SLAGS FROM CHRYSOKAMINO Island

Sample Orkos 3 Orkos 1

Orkos

Kea

SLAGS ISOTOPICALLY 208Pb/206Pb

Type of Sample

Region

Kea

C

Copper

Orkos

Copper

CONSISTENT 207T>L/206 Pb/206Pb

mineral

in a rock

2.06922

0.8338

mineral

in a rock

2.07010

0.83384

mineral

in a rock

2.06919

0.83400

Orkos 2

Kea

KYT 10 publ

Kythnos

Skouries-H.

Ioannis

Slag

heap,

slag

2.06796

0.83364

KYT 52b

Kythnos

Skouries-H.

Ioannis

Slag

heap,

slag

2.06998

0.83367

KYT 3 publ

Kythnos

Skouries-H.

Ioannis

Slag

heap,

slag

2.06896

0.83373

KYTOl

publ

KYT02

Orkos

Copper

Kythnos

Skouries-H.

Ioannis Ioannis

Slag

heap,

Ioannis

Slag

heap,

Slag

heap,

oxidized

Cu

ore

oxidized

Cu

ore

2.06892

0.834

2.06949

0.8345

2.0697

0.83461

2.06972

0.83397

Kythnos

Skouries-H.

KYT 20

Kythnos

Skouries-H.

Zogo2 publ

Kythnos

Zoghaki

KYT 17

Kythnos

Skouries-H.

Ioannis

Slag

heap,

slag

2.07011

0.83421

Kythnos

Skouries-H.

Ioannis

Slag

heap,

slag

2.07012

0.83404

Kythnos

Skouries-H.

Ioannis

slag

2.07015

0.83420

Zogo3 publ

Kythnos

Zoghaki

2.07016

0.83426

KYT03

Kythnos

Skouries-H.

2.07024

0.83415

2.06836

0.8345

2.06893

0.83459

2.07016

0.83408

2.06841

0.83499

publ

KYTT30 KYT 19

publ

SerAvy prill KEF 3

Seriphos Seripho

Kefala

Seriphos

Kondouro

ASPYR3

Siphnos

Aspros

Slag

heap,

oxidized

Mine, Slag

heap,

Cu

Cu

Cu

ore

ore

ore

Slag heap, Cu prill from slag

Avyssalos

KONDO1

oxidized

Mine,

Ioannis

slag

Slag

heap,

Mine, Fe-Cu

Pyrgos

slag

oxidized

Cu

ore

ore

on samples of slags from Chrysokamino compared with the data obtained seems in Figure C.4. This result very surprising at first, but it needs to be examined in terms of the paucity of copper ores in Crete and the known in the Eastern Mediterranean. The Aegean islands, copper mineralizations and Attica are the lands with copper ore deposits that are the P?loponn?se, closest to the island of Crete. suggested that the Minoan as a base for of exploitation stone from Laconia and perhaps also for

Starting from the northwest, settlement of Kastri on Kythera marble

and other

murex

decorative

shells.74 Laconia

copper,

it has been

was

used

has also been mentioned

lead, and silver. Together with and Mineralogical Exploration,

Geology and lead ores in 1987. Some several locations. However, small; there are no mines,

as a

source of possible from the Institute of

geologists we surveyed Laconia for copper mineralization is, indeed, present in copper in all cases is quite the amount of the minerals

but only isolated outcrops in the mountains, trial/survey trenches. No slag heaps that would provide evidence for smelting of metals were found during the survey or reported ores from the copper outcrops and from the by the geologists. Samples of

with

small, modern

deposit inMolai have been analyzed for their lead isotope not match any of the Bronze Age copper or and they do composition, lead artifacts. The largest copper deposit is in the northeast part of the large lead-zinc

74. Coldstream

and

Huxley

1984.

LEAD

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

315

2.08

2.07

2.06

2.05 0.828

0.830

0.832

0.834

0.836

0.838

0.840

0.828

0.830

0.832

0.834

0.836

0.838

0.840

18.9

18.8

18.7

18.6

Figure tions of and

C.4.

Lead

slags ores from

Cyclades

from

isotope

composi

Chrysokamino and the Lavrion

APPENDIX

3i6

C

at Ermioni. The main mineralization is deep and accessible P?loponn?se a number of surface modern but mining only by expressions of techniques, ores to in the Bronze Age miners. would, principle, have been accessible the lead isotope analyses of ores from the Ermioni deposits do However,

Figure C.5 (opposite). Lead isotope

those for Bronze Age copper-based artifacts, so that these ores not used for production of Bronze Age copper objects. on the Attic coast, is the ore Northeast from Ermioni, polymetallic near the port of Lavrion (Laurium, modern Lavrio). South from deposit islands where, on Kythnos and Lavrion, toward Crete, are the Cycladic ores are and ancient slags. Silver, lead, and possibly copper Seriphos, there

slags ores from

not match

were

also copper were mined on the Cycladic island of Siphnos. No copper ores are known from the Dodecanese coast of Turkey. islands or the Aegean The sea distance from Crete to Cyprus going toward the island of Rhodes to Lavrion. is approximately twice as long as the sea route from Knossos islands and Lavrion were for theMinoans Therefore, indeed, the Cycladic the closest sources of copper and lead-silver ores. of slag at Chrysokamino is quite small, and the lead iso of the analyzed pieces show compositions that would tope compositions in result from using the Lavrion and Cycladic ores separately, or mixed The

amount

was a smelting charge. It is possible that the whole smelting operation ore on in the search for result of only a few expeditions the shores nearest to the northern coast of Crete.

one

COMPARISON OF LEAD ISOTOPE ANALYSES OF SLAGS FROM CHRYSOKAMINO WITH EM-MM We

have

COPPER lead isotope

Minoan

ARTIFACTS and elemental artifacts. On

copper-based are the dominant and the Cyclades MM II Crete.

analyses of 118 Early and Middle the basis of lead isotope data, Lavrion sources of copper and lead in EM

lead isotope ratios closest to the analyzed pieces of slag from are found among EM-MM artifacts from several sites. Chrysokamino The

since very similar lead isotope ratios appear also among the However, LM artifacts, copper from Lavrion clearly was used on Crete over a long somewhere else, period of time and must have been extracted mostly not at Chrysokamino not in Crete at all), since it would have (perhaps of these objects been possible to transport copper metal to Crete. Many were cast in or settlement are made of tin bronze. If they palace workshops, the tin would have been added to copper there. Artifacts showing the best lead isotope match with Table C.6.

the slags are plotted

on

Figure C.5

and listed in

analyses

some

of

Protopalatial artifacts from

from

and

Prepalatial

copper-based Crete

alloy

compared

Chrysokamino, and

Lavrion,

slags from the Cyclades

with

copper ores

and

LEAD

ISOTOPE

AND

CHEMICAL

ANALYSES

OF

SLAGS

317

2.08

2.07

2.06

2.05

0.830

0.828

0.832

0.834

0.836

0.838

0.840

0.838

0.840

18.9

18.8

A

18.7

D 0>

Seriphos, Kythnos, Ores from Lavrion

Siphnos

Slags from Chrysokamino EM-MM copper artifacts

207pb/20* Pb 18.6

0.828

0.830

0.832

0.834

0.836

APPENDIX

3i8

C

TABLE C.6. EARLY MINOAN COPPER-BASED ARTIFACTS CONSISTENT WITH THE SLAGS FROM CHRYSOKAMINO

HM

Site

No.

Artifact

Chronology

Description

EM-MM

ISOTOPICALLY 208pbf206pb

206pb/207pb

2.05966

0.83147

2.05789

0.83197

2.066

0.83203

1557/St 9376

Mochlos

dagger,

HM

1556/St 9375

Mochlos

dagger

LM?

HM

1265

Hagia Triada

dagger,

triangular

EM IIA-MM

HM

1930/St 9383

Pl?tanos

dagger,

tongue

EM?

2.06137

0.83211

Hagia Photia Hagia Photia

chisel

EM I-IIA

2.06278

0.83217

chisel

EM I-IIA

2.06624

0.83295

ANM

4660

ANM

4667

HM 2041/St 9368

triangular

IA

EM I-II

2.06944

0.83302

2.0661

0.83305

2.06807

0.83344

2.06576

0.8336

Pyrgos

dagger,

Krasi

dagger with 3 rivets

ANM 4671

Hagia Photia

dagger

HM1195/St9425

Koumasa

dagger,

long

EM IIA-MM

dagger,

triangular

EM IA-MM IA

2.06572

0.83403

2.07083

0.83414

St 9447

HM 2005/St 9432

Marathokephalo

long

EM IIA-MM

IA

EM I-IIA IA

Hagia Photia

spearhead

LM I (?)

Pyrgos

punch/awl

EM I-II

2.0709

0.8344

Hagia Photia

chisel

EM I-IIA

2.07006

0.83444

HM 2040/St 9369

Pyrgos

dagger,

EM I-II

2.07085

0.83448

ANM

Hagia Photia Hagia Photia

fishhook

EM I-IIA

2.06915

0.83455

dagger,

long

EM I-IIA

2.06771

0.83464

long

EM I-IIB

2.06912

0.83468

2.07006

0.83476

EM I-MMII

2.06814

0.83477

4669

ANM

HM 2047/St 9366 ANM

4656

4662d

ANM

4670

long

HM

1852/St 9399

Pl?tanos

dagger,

HM

1182/St 9423

Koumasa

dagger with midrib

Koumasa

dagger,

HM1167/St9427

EM IIA-MM

long

IA

ANM

4674

2.07439

0.83484

4655

Hagia Photia Hagia Photia

EM I-IIA

ANM

awl (?)

EM I-IIA

2.06907

0.83497

ANM 4663

Hagia Photia

chisel

EM I-IIA

2.07241

0.83514

HM

1434/St 9429

HM 318/St 9364 MP/70/146

Porti

dagger,

long

EM-MM

2.07464

0.83529

Palaikastro

dagger,

triangular

EM III-MM I

2.07330

0.83534

EM III-MM I

2.07234

0.83551

Myrtos,

Pyrgos

pin

HM

1498/St 9443

Kalathiana

dagger,

fragment

EM IIA-MM

IA

2.07133

0.83552

HM

1932/St 9381

Pl?tanos

dagger,

long

EM IIA-MM

IA

2.07132

0.83563

Hagia Photia

ANM HM

4675 1552/St 9374

HM1294/St HM

9411

1188/St 9428

C-70.22; C1009B

chisel

EM I-IIA

Mochlos

dagger,

Hagia Triada Koumasa Kea

2.07242

0.8357

long

EM IIA-MM

IA

2.07177

0.83576

dagger,

long

EM IIA-MM IA

2.07275

0.83578

dagger,

long

EM IIA-MM

2.06923

0.83589

EC

2.06169

0.83108

EC

2.05964

0.83172

EC

2.06519

0.83239

EH

2.06353

0.8325

EC 2

2.06344

0.8326

EC-MC

2.06635

0.83261

Hagia

Eireini,

copper

IA

lump 16159

Naxos

Unknown

KPH.9

Kea

5807

Euboia

Manika,

AE 234 (1893.64)

Amorgos

Amorgos,

C-9.100; C775-C

Kea

Hagia

Kephala chisel dagger Eireini,

Cu

slag

LEAD

TABLE Artifact

C.6 No.

Site

ANALYSES

OF

SLAGS

319

Amorgos,

7206a

Lemnos

Poliochni,

16125

Syros

Chalandriani,

Amorgos

Amorgos,

Euboii

Manika,

Cyclades,

Kythn Euboia

Chronology

208pb/206pb

Pbl207Pb

206pb/204pb

EBA-MBA

2.06536

0.83365

18.823

EH?

2.07097

0.83371

18.856

punch

EC 2

2.06494

0.83372

18.805

axe

EC 2

2.06563

0.83394

18.794

EH

2.06982

0.83416

18.827

EC 2

2.0667

0.8344

18.811

EC

2.07005

0.83455

18.8

EH?

2.06595

0.83462

18.793

Description

Amorgos

1969/12.31.5

CHEMICAL

(cont.)

1927.1358

AE236

AND

ISOTOPE

Syros

dagger chisel

flat chisel

tweezers

Chalandriani, Kythnos, Manika,

flat

axe

spatula

ACKNOWLEDGMENTS like to thank very warmly the Director and members in of the Institute of Geological and Mineral Athens and Exploration Chania for their great help in surveying Greek copper and lead/silver de

The

authors would

we were and posits. On Crete, particularly assisted by Michael Diallinas Kostas Zervantounakis. Research in the Isotrace Laboratory was financed by the Science ment Research

and Engineering Research Council, the Natural Environ the British the Leverhulme Council, Trust, and Academy, are to of Oxford. We INSTAP for funding very grateful

the University our work between Department British Geological

For chemical analyses, we thank the at the Royal and the Holloway College

1995 and 2001.

of Earth

Sciences

Survey.

Arsenic A

Prills: by

Study

C. Ferrence

Susan

of

Content

Copper

Applying

and Charles

PIXE

P. Swann

study presents analyses by PIXE of copper prills included in slag at Chrysokamino, to Early Bronze Age workshop from the Final Neolithic on the arsenic content as measured from corroded surfaces in focusing

This

more pure comparison with the values obtained from the prill interiors. The as is survive interest because corrosion of artifacts many study only products, and the degree towhich these corrosion products reflect the original arsenic content

is important in the estimation of the original composition (i.e., is the surface arsenic enhanced because of the removal of other elements, does it remain stable, or is it depleted?). The study addresses the magnitude of the arsenic content

within

individual

of the prill surfaces, the degree of heterogeneity both prills and between prills, and the differences between

corroded prill surfaces and the more

pure prill interiors.

INTRODUCTION The many analyses of metal objects from the Aegean Early Bronze Age show that the metallurgists of this period made extensive use of arsenical scholars have Some copper.1 suggested that arsenical copper has advantages were aware over unalloyed copper and that early Aegean metallurgists, who of these advantages, deliberately produced alloys.2 Others have challenged the concept that arsenical copper offered advantages and the as cases of copper with high arsenic content were produced sumption that all exist in Uncertainties the arsenic content intentionally.3 regard towhether

both

or accidental, whether itwas present in the copper ores or or added later at the smelting similar situ casting stage, and whether ations existed in all Aegean workshops. The debate has been hampered by

was

intentional

was

several factors, including the minimal number of analyses from anything other than finished products. In particular, only a few analyses have been published

from Aegean

copper

1. Renfrew 1967; Charles 1967; Gale

and Stos-Gale 2. Charles

1967.

1989.

smelting

sites.

3. Budd andOttaway 1991; Budd 1991.

APPENDIX

322

1

D

C c 0) c o ?

<

53

55

28

101

Sample Numbers Figure D.I. PIXE analyses of copper

PIXE MEASUREMENTS Samples of prills from the Chrysokamino Bartol Research Institute at the University

prills arranged content

by

increasing

arsenic

were slag heap analyzed at the of Delaware. The analysis used

air (helium) proton-induced X-ray emission spectroscopy,4 an analytical tool for the study of archaeological artifacts that has been employed suc since about 1975.5 With the additional use of appropriate X-ray cessfully filters, the observation of all the elements from sodium through antimony and the heavy elements such as lead is possible with good detection limits. The great advantages of such systems are their basically nondestructive limits for most nature, the good detection character, their multielemental conditions, and the rapidity of data collection; the fact that PIXE is a tech can be both a nique for surface analysis positive on the intent of the study. depending

and negative

characteristic

RESULTS AND DISCUSSION 4. For

Figure D.l increasing conducted

presents the analyses of copper prills by PIXE, arranged by were arsenic content of the samples. The PIXE measurements on the corroded surfaces of tiny masses and lenses of copper

5. Swann Fleming

on

the methodol

Ferrence,

and Betan

comments

ogy, see Swann, court 2000.

and Fleming and Swann 1992.

1988,1990;

ARSENIC

CONTENT

First As Content Third As Content Figure

D.2.

showing

PIXE

OF

H

COPPER

PRILLS

323

Second As Content Average As Content

measurements

the variation

within

vidual prills

indi

included within glassy masses of slag.Where the size of the prill permitted, three PIXE measurements were made on each prill, with the measurements 100 micrometers an

average

of

apart. Arsenic

contents vary from <0.03% to 3.42%, with

0.72%.

as shown Figure D.2 illustrates the variations within individual prills by in the PIXE measurements D.I. arsenic The three values for given Figure each prill are shown as the first three bars in each group, with the average for each prill presented as the last bar in each group. Variation can be as little as 0.01% or as great as 1.29%. some of the copper The measurements presented here show that prills to Early Bronze at obtained from Final Neolithic Age smelting operations Chrysokamino

between

contain

arsenic

2% and 6% have been

invariable

amounts.

as

Percentages

amounts

of

arsenic

because

regarded significant qualities during casting and its ability to improve in tiny prills at the hardness.6 The amounts recorded from Chrysokamino cannot to level the gross amount in larger necessarily be related microscopic masses of copper because of the high variability detected. of arsenic's antioxidant

recon amount of arsenic is compatible with the methodology for Chrysokamino, where the copper was formed as tiny prills a disseminated have shown that incor through glassy slag. Experiments as an arsenic of within smelted copper occurs during the poration impurity This

structed

6. Budd

and Ottaway

1991.

APPENDIX

324

D

are present in the smelting process if arsenic minerals original charge and if the temperature is not high enough to result in amolten metal that would some conditions, be removed by tapping the furnace.7 During including those suggested for this site, some of the prills would be trapped within the slag even if some of the copper collected at the base of the furnace. As elsewhere, under the conditions present at Chrysokamino only the larger prills would migrate to the bottom of the furnace and collect as molten metal, leaving the smaller ones within the slag. indicated

The

lower arsenic values obtained

from corroded

with clean surfaces (see the results obtained by Bassiakos that the corrosion tends to deplete the arsenic content

surfaces compared inApp. B) shows contained within

is presumably a postburial action that occurred the prills. This condition corrosion. it is important the of In terms of methodology, processes during means to recognize this result because it that values obtained from corroded copper surfaces are useful asminimum the occurrence of arsenical copper.

values and can be used to document 7. Pollard

et al. 1991.

APPENDIX

E

Slag

by Wavelength

Analysis

Dispersive

Spectrometry

by Christine M. Thompson

was carried out at the Center forMaterials Research following research inArchaeology and Ethnology and the Center for Geochemical Analysis at theMassachusetts Institute of Technology. The samples were analyzed on aJEOLJXA-733 byWavelength Dispersive Spectrometry Superprobe with a beam current oflOnA, probe diameter of 1micrometer, and count of times 40-80 seconds element. per ing The

Five pieces of metallurgical slag from the smelting workshop at Chryso kamino were analyzed (numbered CHR 55,48,50, 8, and 74). They were to all dark, vitreous specimens from the Final Neolithic Ill Early Minoan Middle Minoan All the IA slag pile. samples contained small, visible prills of copper. In the following

charts, "nd"means

"not detected"

and "na"means

"not

analyzed."

SAMPLE CHR 55 were made at two Analyses opposing points of the samples matrix, and results indicate a nearly homogeneous composition, which is a silicate high in iron and barium (Points 1 and 2). Both the light and dark "needles" are same elements as the matrix. The dark comprised of the crystals within (Point 5, shown in Fig. E.l) are an Fe-oxide. The dark crystals in the backscattered electron image (Fig. E.l) appear light in the photo micrograph (Fig. E.2). the matrix

TABLE

E.l.

CHR

Point

55, POINTS

SiO^

(Figs. E.l, E.2)

&2<>s

*,o3

CaO

BaO

Cu

Total

Notes matrix

35.75

2.44

32.39

3.24

23.83

trace

97.65

37.05

3.04

33.42

3.69

19.60

trace

96.8

note

1

73.57

1.10

0.32

trace

103.81

note

2

28.83

Notes:

1-5

nd

29.80

2.62

35.15

1.08

31.16

trace

99.81

note

3

0.06

0.44

101.15

0.03

0.12

trace

101.80

note

4

1. Same matrix

opposite

Point

. 1; 2 Dark

"needles";

3. Light

"needles";

4. Dark

crystals.

E

APPENDIX

326 E.2.

TABLE

CHR

Point

Notes:

55, POINTS

Pb

Fe

Cu

nd

3.1

70.48

nd

0.54

99.13

inclusion;

2. Prill.

1. Circular

6, 7 (Figs. E.l,E.2)

nd 0.82

0.01

nd

25.75

0.23

nd

0.30

99.34 101.02

Figure E.l (left). Backscattered electron image of CHR 55 (200 x) with large copper sulfide inclusion,

SAMPLE CHR 48

"needles,"

inner regions. The Sample CHR 48 is comprised of three concentric most region contains Cu-As-oxides (Points 1 and 2). The surrounding with is of Cu-oxides high levels of Cl (Point 4). Point 3 region comprised is primarily exterior the region, which represents sample's dominant, Cu sulfide (Cu2S).

TABLE E.3. CHR 48, POINTS Point

Figure

E.3.

Notes

Total

Cl

Sb

As

Photograph

dark

crystals, by C. Thompson

and

prills.

light Figure E.2 (right).White photomicrograph of CHR 55 (200 x) with "needles" and small prills. Photograph

by V. Pigott

1-4 (Fig.E.3) As

Sb

Total

Ca

Si

26.67

0.88

0.13

92.65

26.88

0.88

0.13

90.34

Cl

Pb

Fe

Cu

nd

0.10

44.91

18.82

0.37

0.73

0.04

nd

0.20

42.28

18.48

0.33

1.12

0.04

nd

0.01

78.98

nd

0.04

nd

19.68

Na

98.71

nd

nd

58.73

nd

0.04

14.91

0.01

Na

90.63

Backscattered

electron

image showing the core of CHR 48 (40 x). Photograph

by C. Thompson

DISPERSIVE

BY WAVELENGTH

ANALYSIS

SPECTROMETRY

327

SAMPLE CHR 50 50 is primarily an Fe-Ca-Al-Silicate (Point 1). Ba-sul in the remainder of the is associated with Cu-chloride fate (BaS04) (CuCl) amounts of aluminum matrix. The dendrites contain iron oxide with small

The matrix of CHR

(Point 2). Analyses of the prills reveal high levels of copper, small amounts of iron, and trace amounts of arsenic and antimony (Point 3).

Figure E.4. Photomicrograph of CHR 50 (400 x)with dendrites and prill field. PhotographbyV. Pigott

TABLE E.4. CHR 50, POINTS Point

SiO,

*2?3

n2o3

CaO

BaO

Cu

Total

Notes

44.77

6.39

32.65

14.86

0.18

nd

98.85

matrix

0.94

3.93

96.42

0.42

nd

trace

101.71

dendrite

TABLE E.5. CHR 50, POINT Point

1, 2 (Fig.E.4)

Fe

Pb

3.52

nd

3 (Fig.E.4)

Cu

As

Sb

Cl

97.53

0.57

0.18

0.11

Notes

Total

101.97

0.06

prill

SAMPLE CHR 8 silicates comprise the matrix and needles of CHR 8. Cu-sulfide was detected at Point 4 and is representative of the circular inclusion shown an inclusion (not shown) contained in Figure E.5. Another Cu-sulfide Fe-Ba-Ca

inner region of metallic

TABLE E.6. CHR 8, POINTS Point

&"0,

*P3

copper with

arsenic

(Point 1).

1-3 (Fig.E.5) K2?3

CaO

BaO

Cu

Total

Notes

34.17

2.64

36.99

8.06

15.73

trace

97.59

matrix

29.56

1.17

64.02

4.60

2.82

trace

102.17

needles

0.20

0.36

98.20

0.09

0.39

trace

99.24

dendrite

APPENDIX

328

TABLE E.7. CHR 8, POINTS Point

Notes:

E

4, 5 (Fig.E.5)

Pb

Fe

Cu

As

Sb

Cl

nd

0.56

80.41

0.21

nd

nd

20.06

101.24

nd

1.27

97.87

2.97

nd

nd

nd

102.39

1. Circular

2. Circular

inclusion;

Total

Notes

inclusion.

Figure E.5 (left). Photomicrograph of CHR 8 (400 x)with "needles," dendrites, and circular inclusion of sulfide copper by V. Pigott

SAMPLE CHR 74 of CHR 74 was a Ca-Fe-Al-silicate, and the dendrites con were iron oxide with measurable amounts of aluminum. The prills

The matrix tained

comprised

of copper with minor

amounts

TABLE E.8. CHR 74, POINTS Point

Si09

^A

by C. Thompson

1, 2 (Fig.E.6)

FeO

CaO

BaO

Cu

Total

8.77

15.54

35.48

0.07

trace

97.12

1.19

5.67

93.41

1.74

0.13

trace

102.14

Pb

Fe

nd

0.14

Cu

102.3

3 (Fig.E.6) As

Sb

0.93

0.30

Cl

Total

103.67

Photograph

Figure E.6 (right). Backscattered electron image of CHR 74 (lOOOx) with matrix and dendrites. Photograph

37.26

TABLE E.9. CHR 74, POINT Point

of arsenic and antimony.

at center.

Notes

prill

Notes matrix dendrite

F

APPENDIX

Smelting the

Analysis

the

of

Reconstruction

Process

Based

of Ore

and

Copper

on Slag

Samples by Yannis Bassiakos andMihalis Catapotis application of scientific techniques in the study of Early Bronze Age metal artifacts and metallurgical finds has a very long history in the Ae for early metal production has been evidence gean.1 However, although our available since the middle of copper smelting 1980s,2 understanding technologies during the Early Bronze Age is still limited. This is due to The

sites and to the preoc the lack of systematic excavations of metallurgical in of studies the cupation archaeometallurgical Aegean with questions of as at the of considerations. Therefore, provenance expense technological the only systematically excavated smelting site in the Aegean, Chrysoka a for a detailed technological unique opportunity provides study of a copper smelting process. This report presents the results of analyses of ore and slag samples from conducted at the Laboratory of Archaeometry, N.C.S.R. Chrysokamino

mino

"Demokritos." By integrating archaeological and analytical data, we propose a was provisional reconstruction of the copper smelting process that taking at site. the place

MORPHOLOGY OF THE SITE

AND GEOLOGICAL CONTEXT

The metallurgical site of Chrysokamino is located on a promontory that is part of a small isolated peninsula3 consisting of terraced hillsides on the eastern edge of the Gulf ofMirabello. 1.

E.g.,

Lamb

1936; Keos

I; Crad

dock 1976; Gale and Stos-Gale 1981a; and Ioannou Mangou et al. 1990; Stos-Gale 2. E.g.,

Gale

1984;Wagner 3. This horst;

is a

Flemming,

Pernicka

1999; 1993.

et al. 1985;

Spitaels

andWeisgerber submerging

1985.

tectonic

Czartoryska,

Hunter 1973; Fortuin 1978.

and

Most of the typical geological formations of eastern Crete are present in this area (for the geology, see Farrand in Chap. 2). The stratigraphically lower autochthonous unit, constituting approximately half of the peninsula, is a carbonate formation known as Plattenkalk, present also in Epirus, in the P?loponn?se, and in other Cretan areas. It consists of a solid, rela or limestone, tively thinly bedded slabby gray to bluish semimetamorphic to brown calcitic veins. conspicuous white Sporadically in the same formation are some folds, either coarsely crystalline or intercalations of marly and green limestones. The first overthrusting

interlaced with visible with

APPENDIX

330

F

sheet is the unit of the Phyllite-Quartzite nappe, a formation exhibiting more intense in its is which low-grade metamorphism, stratigraphically constituents of this phyllitic unit are powdery (silty) and lower parts. Most mineralogically

altered as a result of in situ weathering of the calcite-rich, less extended in the site is the dolomite of the

rock. Much

fine-grained second successive Tripolitza

to the sheet, which belongs geologically overthrusting to and is predominantly dark gray and nappe (Figs. 2.1,2.3) gray bedded. It occurs in the area of the habitation site and between

massively it and the smelting location. An interesting geological feature of the area is aweathering-resistant carbonate ledge traversing the local phyllites with a strike of N40-55E and a dip of 40-50S.4 In our visits to the area, we noted that the selective of the weathering phyllites has often resulted in the exposure of the ledge,

In those cases, iron (hydr)oxides can be forming localized protuberances. seen on the lower contact between the carbonate ledge and the phyllites. Similar iron (hydr)oxide crusts can be seen frequently on rock fragments among the tumble covering the slope on the east of the promontory. Those thin crusts appear to develop locally into aweak iron mineraliza contact. The existence of tion, mainly hosted in the Plattenkalk/phyllite or mineralizations is, in calcite-accompanied quartz-accompanied uncommon in the Aegean HP/LT experience, not polymetamorphic in in the formations Attica, Cycladic complex, and elsewhere) where (e.g., the repeated metamorphic action causes mobilization and condensation of

weak our

relocated by means of ledge components, formerly disseminated metallic as a source formation. The possible role of the local iron mineralization is ad of fluxing material in the copper smelting process at Chrysokamino dressed later in this report. In contrast, no copper mineralization has been identified in the vicin site, suggesting that copper ores were brought to ity of the metallurgical The original interpretation of the nearby cave from elsewhere. Chrysokamino a as and others of Theriospelio copper mine5 has been rejected by Diallinas no evidence of mineralization.6 who have examined the cave but found

METALLURGICAL

REMAINS AND SAMPLES

at the site revealed many tons of dark-colored slag present, cm in size)* or completely pulverized in small (see mostly pieces (under 2 the finds were tens of thousands of perforated ceramic Chap. 10). Among identified as chimneys for furnaces (Chap. 7), remains of pot fragments bellows (Chap. 8), tuyeres (Chap. 9), and a small number of stone tools ore (malachite, azurite, and (Chap. 6). A few minute fragments of copper were revealed during the excavation. Our study concentrated chrysocolla) Excavations

of ores and slags from the site;Myer and Betancourt the furnace chimney fragments (App. A). Three surface finds of copper ore were collected by Bassiakos after the were used for labora of the excavations completion (August 2000) and on raw materials used to in information the order tory analysis provide

on the examination studied

et al. 1999,

4. Betancourt 5. Mosso 6. Branigan in this volume.

1910,

p. 352.

p. 219.

1968,

p. 50;

see

Chap.

18

RECONSTRUCTION

OF

THE

COPPER

SMELTING

PROCESS

331

in the smelting process. The fragments were of very small size (less than 2 cm). Macroscopically, all three appeared to consist of green copper ores to The green copper mineral was identified in the field quartz. adhering as malachite, Iron hydroxides were visible on two of CuC03*Cu(OH)2. the ore fragments (ORE-1 and ORE-2), both as individual particles and as

stains

on

the

surface

of

quartz.

were selected for this Forty-six slag samples analytical project, mostly deriving from passes 3 (sample series B) and 4 (sample series A) in trench N 20. One sample (sample series C) derived from the surface material. The were small, samples irregularly shaped fragments rarely exceeding 2-3 cm in size. Porosity was limited and of small size (<1 mm), whereas vitrification crusts varied both among and within individual slag fragments. Carbonate most derived from the calcareous environment coated thickly fragments (with the crusts up to ca. 1mm thick). This epigenetic material had often the penetrated to their interior through open porosity and cracks. Despite state of the fragmented slag and the presence of epigenetic material, both of which prevented a detailed macroscopic examination, many of the largest still had visible textural features that could be associated slag fragments with

slag tapping, i.e., the removal of liquid slag from the furnace through a hole. Minute copper prills (<1 mm) and small silica inclusions (<2 mm) could be seen occasionally with amagnifying lens. ore was iron An included in the slag samples fragment inadvertently

collected for laboratory analysis. The size of the fragment was about 3 to 4 cm, and macroscopically it appeared to consist of yellow orange iron was considered to be an This hydroxides. important find as it could have use been associated with the of fluxing agents in the process.

ANALYTICAL OBJECTIVES, METHODOLOGY, AND INSTRUMENTATION aim of this study was to understand the technology employed for the production of copper at Chrysokamino. Breaking down the smelting process into its constituent parts, we set the following questions as guidelines for

The

the analytical work: types of copper ores were processed, and what were the associated gangue minerals? 2. Is there evidence for the use of fluxing agents or other additions in the charge? 1.What

3.What

was

the nature of the smelting process (e.g., reduction etc.)? conditions (temperature and redox

smelting, matte smelting, 4.What were the operational

conditions) during smelting? 5.Was the slag tapped or was it left to solidify 6.What were the copper losses in the slag? 7.What

was

chemical

inside the furnace?

of the smelting process (i.e., what was of the smelted composition copper)?

the product

the

F

APPENDIX

332

state of the material did not allow detailed the fragmented macroscopic examination of the samples, more emphasis was placed on laboratory tech was conducted niques. Mineralogical analysis using optical microscopic methods. Polished specimens were examined on a LEICA DM polarizing

As

from x25 to xlOOO.Quantitative microscope using a range of magnifications chemical analysis of ore and slag samples was conducted using a Philips 500 (SEM) coupled with an Energy Dispersive Scanning Electron Microscope detector (EDS) (see App. B). In most cases, the chemical Spectroscopy was obtained area scans on composition by multiple specimens pelletized into resin. In cases where the size of the sample was small, impregnated chemical composition was obtained by multiple area scans on the polished was also used for the specimens used for microscopic analysis. SEM-EDS analysis of individual phases within samples. Finally, the Fe27Fe3+ ratio of five slag samples was obtained by M?ssbauer spectroscopy.7 chemical

ANALYTICAL RESULTS: THE ORES Most

macroscopic

observations

the

concerning

copper

ores

were

gener

as the copper minerals were ally confirmed microscopically, predominant malachite with some azurite, while the gangue minerals Cu3(C03)2(OH)2, were and quartz with small amounts of iron hydroxides predominantly of sample ORE-1 under the polarizing phases. Examination also revealed small quantities of sulfidic minerals inside the microscope oxidized matrix. The identification of those minerals with the aid of siliceous

the predominant sulfide to be pyrite small quantities of barium, covellite (CuS), (FeS2), occasionally containing some chalcocite and mineral chalcopyrite (CuFeS2), (Cu2S) (Fig. F.l).The texture and of those indicates that inclusions constituted residual ogy they SEM-EDS

hypogene oxidation

point

The by EDS O RE-3 fragment,

showed

sulfides, which had survived the weathering process inside the zone of a copper (vein?) deposit. No sulfidic inclusions were

in the

found

analysis

other

two

ore

samples.

of the three samples was obtained composition area scans) of the sections. Sample polished (multiple a a quartz of very fine flake of malachite adhering to

bulk chemical analysis consisted so two

separate

sets

of

and one for the gangue mineral. support the phase identifications

analyses

As

were

shown

conducted,

in Table

one

for

the

the EDS

ore

data

F.l, on microscopic analysis. In the of first place, the generally low sulfur contents reflect the predominance in the ore samples. In the case of ORE-1, oxidized minerals the however, residual sulfides identified microscopically slightly raise the sulfur content. based

quartz is shown to be the predominant gangue mineral, followed and aluminum, while the levels of lime are extremely by iron-hydroxides never content of arsenic and nickel in low, exceeding 0.3%. The average Secondly,

all ore samples iswell below 0.2%, although occasionally the percentages reached slightly higher levels (0.5%) in individual area scan analyses. ore examination confirmed Turning to the iron fragment, microscopic that the dominant

mineral

in the ore is goethite,

-FeO(OH),

showing

7.We poulos

are grateful from N.C.S.R.

this analysis conducting the analytical results.

to Dr. A.

Simo

"Demokritos" and processing

for

OF

RECONSTRUCTION

THE

COPPER

SMELTING

PROCESS

333

Figure F.l. Copper ore (ORE 1) from Chrysokamino showing resid ual sulfides (1: pyrite; 2: covellite) in amatrix of oxidized copper/iron minerals (3) and grains of quartz (4); reflected

light.

Scale: 30 micrometers

Figure F.2. Iron ore from Chryso kamino showing rhomboid crystals of calcite (dark gray) embedded in amatrix of iron (hydr)oxides (light gray);

reflected

light.

Scale: 0.1 mm

texture

in cross-polarized light, locally altered Small rhomboid (Fig. F.2). crystals of calcite (Fe203.nH20) were (ca. 100-300 micrometers) dispersed throughout the ore, raising the CaO content to 25% (Table F.l). The silica, on the other hand, was just characteristic

botryoidal

to "limonite"

5%.Mineralogically, therefore, the iron ore is different from the copper ores found at the site. also that the iron ore contains no Considering traces of copper, it follows that the iron ore fragment does not represent gangue mineral discarded during some process of in situ beneficiation but rather a material assemblage.

that was

separately

introduced

to the metallurgical

TABLE F.l. EDS ANALYSIS Sample

SiO^

(AVERAGE COMPOSITION)

FeO

CaO

^A

MgO

OF COPPER ORE AND

MnO

Na^O

K20

CuO

IRON ORE S

so^

cip

As203

NiO

ORE1

57.60

6.53

0.14

1.51

0.92

0.01

0.76

0.01

29.93

1.52

0.15

0.05

0.08

ORE 2

54.14

1.41

0.25

1.52

0.52

1.09

0.54

0.10

39.17

0.17

0.18

0.08

0.01

1.53

0.29

0.15

0.92

1.27

0.08

3.37

0.00

91.31

0.03

0.67

0.00

0.10

97.21

0.09

0.00

0.47

0.28

0.00

0.66

0.07

0.04

0.00

0.04

0.00

0.00

4.86

67.36

23.25

0.47

1.29

0.20

1.32

0.12

0.04

0.09

0.23

0.06

0.16

ORE 3 (ore) ORE 3 (gangue)

IRONORE Results

normalized.

RECONSTRUCTION

OF

THE

COPPER

PROCESS

SMELTING

335

Si02

a

copper-ore iron-ore slag

?* **

#**>

F.3. The

Figure

from

slags ternary

position

of ores

Chrysokamino

system

and

in the

FeO

CaO

CaO-FeO-Si02

Si02 + A1203

Chrysokamino

Other sites

* van

f?tf

F.4. The of Figure position slags from Chrysokamino and other EBA sites from the south smelting ern in the ternary system Aegean

CaO-FeO-Si02

CaO

(+A1203)

FeO

ANALYTICAL

RESULTS:

THE

SLAG

to the results of the EDS analysis of pelletized According samples, the slag to from Chrysokamino the and it also con system CaO-FeOx-Si02, belongs tains aluminum and smaller quantities ofMgO, MnO, and alkalis (Table F.2). area of olivines and pyroxenes, to Most samples fall in the corresponding in this slag system (Fig. F3). the lowest liquidus temperatures It should be noted that the high levels of lime characterizing the are uncommon slags from Chrysokamino come of with copper slag examples high lime levels small number of Near Eastern smelting sites, including the sites F2 and 39 atTimna,8 the EBA site of Shahr-i-Sokhta

copper Known 8.Merkel

and Rothenberg

9. Hauptman, Strecker

Rehren,

and Schmitt

the Roman

2003.

10. Bachmann

1999.

smelting

1980.

Aegean

site of Beer Ora

Early Bronze Age

atTimna.10 They (Fig. F.4).

are

in prehistory. from only a Chalcolithic

in Iran,9 and certainly unique in the

TABLE F.2. EDS ANALYSIS Sample

OF SLAG PELLETS

SiQ2

FeO

Al2Q3

20.49

5.93

**P

CuO Kp_

so*

Cl20

*Ps

MO

1.73

0.81

0.41

0.38

0.16

0.15

0.01

0.33

1.76

0.73

0.78

0.18

0.09

0.24

0.04

1.53

0.66

2.72

0.76

1.29

0.21

0.29

0.07

0.03

5.45

1.61

0.30

2.35

1.10

0.52

0.49

0.10

0.05

0.00

6.55

1.70

0.10

2.14

1.20

1.51

0.20

0.20

0.18

0.06

0.08

0.02

A01

28.46

36.34

A02

28.98

45.12

12.04

5.45

1.81

A03

11.00

39.45

33.04

7.60

A04

20.76

31.29

34.58

A05

18.64

33.78

31.55

1.59

0.27

A06

27.98

37.89

16.03

7.02

3.38

0.86

2.66

0.96

0.67

0.30

0.14

AQ7

19.39

36.00

30.78

5.30

1.16

0.20

2.90

0.78

0.55

0.17

0.13

0.09

0.00

A08

19.36

35.53

32.97

5.33

1.54

0.21

1.21

0.98

0.92

0.19

0.10

0.07

0.03

A09

20.50

34.39

31.01

5.53

1.52

0.18

2.15

1.04

1.35

0.25

0.22

0.20

0.06 0.03

1.51

0.27

2.61

1.13

0.63

0.21

0.12

0.16

10.76

1.59

0.20

4.41

1.10

1.15

0.12

0.17

0.13

0.00

11.31

7.24

2.43

0.40

1.99

1.11

0.98

0.16

0.20

0.37

0.01

47.89

18.23

5.95

0.95

0.35

2.25

0.70

1.14

0.17

0.16

0.17

0.04

22.47

38.43

22.01

7.68

2.38

0.40

2.40

1.00

1.11

0.02

0.09

0.00

0.13

A15

11.92

39.38

29.52

8.85

1.58

0.34

3.26

0.73

2.30

0.20

0.32

0.07

0.09

A16*

13.70

40.24

31.22

5.93

1.42

0.21

2.46

0.57

1.70

0.19

0.06

0.66

0.00

A17*

18.33

34.85

31.84

7.35

1.76

0.34

1.79

0.91

0.92

0.33

0.03

0.13

0.05

A18*

9.85

38.97

33.82

8.62

1.75

0.30

3.71

0.93

0.92

0.02

0.07

0.06

0.04

38.99

28.03

8.58

2.34

1.15

1.22

0.78

0.07

0.05

0.00

0.06

1.52

0.39

1.97

0.68

2.57

0.13

0.18

0.63

0.03

24.83

43.28

16.49

All

17.94

45.81

12.67

A12

25.11

46.77

A13

19.71

A14*

A10

A19*

14.66

6.79

2.68

A20

17.38

40.36

24.03

7.22

A21

27.42

32.45

27.68

4.85

1.31

0.24

2.07

0.66

0.89

0.22

0.11

0.24

0.05

A23

9.25

40.12

32.02

9.17

1.92

0.52

2.87

1.29

0.97

0.30

0.13

0.00

0.04 0.01

A25

20.35

42.98

20.99

6.40

1.43

0.15

1.99

0.91

1.75

0.14

0.24

0.21

A26

21.41

42.20

24.63

3.90

1.23

0.16

1.77

0.64

2.02

0.11

0.20

0.07

0.06

A28

19.76

41.95

25.31

5.69

1.33

0.15

1.65

0.66

0.98

0.19

0.22

0.05

0.01

A29

18.83

36.88

29.11

6.99

1.60

0.29

2.76

0.85

0.75

0.22

0.09

0.07

0.00

TABLE

F.2

(cont.) CaO

Sample

Si02

FeO

AIP3

M#0

MnO

Na20

CuO Kp_

so*

cip_

^P3

MO

BOl

24.05

35.31

28.43

5.09

1.49

0.23

1.45

1.11

0.44

0.30

0.10

0.12

0.04

B02

22.62

38.57

22.63

4.92

1.34

0.20

1.21

0.64

2.23

0.15

0.12

0.44

0.01

B03

25.88

43.88

15.14

5.53

1.46

0.31

1.96

0.80

1.58

0.03

0.13

0.34

0.00

B05

22.29

34.65

27.10

5.71

2.70

0.32

2.58

0.81

1.09

0.14

0.17

0.36

0.07

B06

15.34

35.76

33.02

6.29

2.06

0.16

3.04

1.06

0.97

0.39

0.16

0.03

0.03

B07

14.56

34.86

35.85

6.13

1.56

0.17

2.93

0.85

1.06

0.34

0.24

0.03

0.08

B09

31.05

38.34

13.25

5.73

5.86

0.29

1.61

0.81

0.67

0.19

0.14

0.21

0.04

Bll

26.87

41.40

17.42

5.70

1.48

0.25

2.29

0.74

1.11

0.07

0.17

0.35

0.07

B12

14.34

38.69

28.75

9.21,

2.47

0.38

1.89

1.02

1.20

0.21

0.21

0.04

0.01

B13

20.71

35.86

30.49

5.49

1.64

0.13

1.55

0.89

0.97

0.24

0.13

0.08

0.03

B14

14.97

34.73

34.53

7.02

1.62

0.28

3.00

0.83

1.51

0.30

0.13

0.03

0.00

B15

13.98

36.72

31.11

8.02

1.61

0.52

2.72

0.64

2.85

0.19

0.25

0.17

0.04

B17*

19.97

40.53

24.95

6.37

1.25

0.48

2.77

0.80

0.64

0.27

0.03

0.18

0.05

B18

17.34

38.36

28.01

7.45

1.58

0.37

2.25

0.96

1.01

0.31

0.18

0.18

0.04

B20

23.97

33.46

27.56

5.64

1.82

0.26

2.89

0.90

0.85

0.25

0.14

0.05

0.00

B22

25.98

43.43

16.66

5.06

1.14

0.21

2.29

0.58

1.31

0.09

0.14

0.47

0.00

B23

15.36

36.31

30.38

8.00

1.85

1.39

3.25

1.05

0.44

0.50

0.16

0.11

0.04

B24*

19.46

33.37

32.09

7.76

1.41

0.56

1.96

0.96

0.68

0.42

0.03

0.20

0.00

B26

22.38

40.90

23.87

5.24

1.36

0.14

1.86

1.06

0.71

0.26

0.14

0.04

0.06

COI

17.58

35.77

31.31

7.30

1.48

0.30

3.06

0.81

1.00

0.09

0.01

0.00

0.05

with

an asterisk

Analyses

marked

have been

obtained

by multiple

area scans

(see text). Results

normalized.

F

APPENDIX

338

The average copper content of the slag is ca. 1%. This low amount is indicative of a very efficient smelting process with low copper losses. Arsenic and nickel, which are generally concentrated in the metallic phase are rather than the slag,11 also found, therefore, at very low levels (<0.5% on

average).

Examination izing microscope and intergrowths, individual

of polished sections of slag samples under the polar showed a wide range of mineralogical phases, textures, notable both within and between heterogeneity being

samples. A brief description

of the major

identified

phases

is

given below.

and

Silicates

Glass

to fully vitrified, the slag textures range from fully crystallized is usually dominant. Silicate crystalline formations are glassy component a fine dendritic texture indicative most in them present slag samples, giving

Although

of rapid precipitation from a liquid solution. Their chemical composition to the pyroxenes and olivines, most samples approximating corresponds the composition of the compounds and hedenbergite CaO.FeO.Si02 in a glassy matrix olivine (Table F.3). The silicate crystals are embedded that often shows red internal reflections due to the presence of bands of should note finely dispersed copper and (possibly) cuprite inclusions.12 We that a considerable part of the identified "glass" matrix might be in fact as indicated by the chemical similarity between the glassy cryptocrystalline, and siliceous phases

Iron

in the slag (Table F.3).

Oxides

are (FeO) and magnetite frequent in most slag samples, (Fe304) is always present in the latter usually being the dominant oxide. Wustite are as dendritic form, whereas magnetite crystals generally present large

Wustite

skeletons

near

the

center

of

the

slag

fragments,

changing

to

numerous

aligned dendrites with excessive growth along a single direction near the is clearly associated with differences in the cool surface. This difference areas in the rate to of cake: closer the various surface, the drop in ing slag is with the steeper, temperature rapid crystal growth producing fine-armed shape and size is primarily determined by kinetic factors. to the results of M?ssbauer analysis of five slag samples, the According state iron is of Fe2+, with the Fe27Fe3+ ratio ranging valency predominant dendrites whose

from 80:20 to 90:10 (TableF.4). A notable feature of many slag samples is the presence of thin bands of magnetite present at the interior of the fragment (Fig. F.5). In many are cases those bands run through the sample, but they uninterruptedly more

The microstructure of the slag, however, commonly discontinuous. is always different on the two sides of the band. A similar textural feature has been identified by Okafor13 in early iron smelting slags from Nigeria and has been of those bands

shown to reflect successive

the presence tap cycles. Whether is questionable, indicates a similar process at Chrysokamino

ll.Yazawal980,p.381. 12. Cf. Hauptman, Schmitt-Strecker 13. Okafor

Rehren,

2003. 1993,

p. 446.

and

RECONSTRUCTION

OF

THE

COPPER

TABLE F.3. EDS ANALYSIS OF MAIN COMPONENTS AND GLASS PHASES IN SLAG SAMPLES FeO

SMELTING

OF SILICATE

Sample

Phase

CaO

Si02

A04

silicate

27.63

33.93

34.71

1.01

1.99

0.23

A04

silicate

25.66

36.33

28.94

5.69

1.26

0.85

A05

silicate

22.44

42.19

19.80

1.01

2.32

A05

silicate

28.36

34.18

33.27

0.47

2.76

0.57

A13

silicate

20.85

47.32

21.21

5.29

2.41

0.42

A14

silicate

32.72

40.78

12.15

6.63

3.07

3.92

A14

silicate

33.14

40.15

12.00

6.63

3.77

3.39

A14

silicate

33.12

40.56

12.44

7.08

3.09

3.12

A14

silicate

23.14

40.43

21.64

7.97

2.64

1.50

A15

silicate

20.69

43.04

25.36

7.84

2.12

0.64

A18

silicate

10.53

32.97

46.34

7.33

1.09

0.35

A19

silicate

13.61

36.75

35.58

8.97

2.94

1.02

A19

silicate

13.39

36.58

33.90

8.61

3.43

1.22

B01

silicate

25.76

34.02

32.29

3.13

B01

silicate

26.08

34.12

31.94

2.85

1.91

1.24

B01

silicate

26.44

33.97

32.65

2.28

2.58

1.03

MgO

Na20

1.24

B01

silicate

34.57

41.01

13.87

4.38

3.46

1.87

B01

silicate

34.35

40.65

14.08

4.85

2.78

2.22

B01

silicate

34.76

40.77

13.07

4.18

4.22

2.13

B12

silicate

12.30

30.95

39.97

11.08

3.76

0.48

B13

silicate

33.94

40.64

15.91

4.26

2.46

2.02

B13

silicate

33.14

40.84

15.93

4.33

2.40

2.10

B17

silicate

20.77

42.52

27.21

6.28

0.89

0.96

B20

silicate

28.45

36.24

25.33

5.71

1.43

1.40

A05

glass

22.28

39.97

28.48

6.27

1.82

0.86

A13

glass

27.05

49.48

12.88

6.02

1.11

1.41

A14

glass

21.96

39.91

26.15

8.26

2.61

0.34

A14

glass

21.46

41.76

24.52

8.31

2.07

0.70

A14

glass

33.49

40.22

12.38

7.23

2.85

2.99

A14

glass

23.07

39.20

23.15

7.69

2.37

2.21

A17

glass

20.96

39.47

24.77

7.25

1.93

3.35

A18

glass

10.33

41.58

33.17

8.75

1.63

3.06

A19

glass

15.33

39.92

28.84

8.79

2.03

3.13

A19

glass

15.40

39.00

28.83

8.21

2.64

2.81

B05

glass

24.70

38.07

25.31

6.71

2.20

2.09

B12

glass

15.53

38.92

27.81

8.95

1.53

4.20

B13

glass

15.31

41.02

30.11

6.69

0.92

1.45

B13

glass

15.00

41.46

30.95

7.78

0.55

0.55

B17

glass

21.09

40.68

26.76

5.92

1.47

2.31

B24

glass

22.36

37.54

24.83

7.91

1.91

3.26

Results

not normalized.

PROCESS

339

APPENDIX

340

TABLE F.4. RESULTS OF M?SSBAUER FE+2/FE+3 RATIOS OF SLAG SAMPLES Solidified Ferrous

Sample

Hedenbergite

A17

40

A18

80

A19

40

B17

80

B24

40

phases

iron

of

Valencies

Intermediate

(unidentified)

35

0

40

AND CALCULATED

SPECTROSCOPY

(unidentified)

50

F

Magnetite

25

Fe'2

Fe*

Fe+2/Fe+J

83.3

16.7

5.0

20

80.0

20.0

4.0

10

90.0

10.0

9.0

20

80.0

20.0

4.0

86.7

13.3

6.5

20

Figure F.5. Sample A04: magnetite "bands"

forming

Reflected

light.

three slag "layers." Scale: 50 micrometers

as the thickness of individual mm. However, "slag-layers" rarely exceeds 2-3 as the presence of a bands is strong indication of solidification magnetite in contact with the air, it seems very plausible to argue that the bands formed by the "folding" of a relatively viscous slag as itwas tapped out of is considered the furnace. The possibility of slag tapping at Chrysokamino more in detail in the discussion of the analytical results.

Copper

Prills

are common inmost Copper prills slag samples. Their diameter ranges from a few micrometers to the results of to 1mm in some rare cases. According analysis, their chemical composition corresponds to the quaternary are Cu-Fe-As-Ni Arsenic levels (Table F.5). alloy highly fluctuating, often or even the level of solid solubility in copper, which reaching exceeding is ca. 8%.14This results in the formation of the arsenic-rich Cu3As phase present inmany prills in the slag (Fig. F.6).15 The nickel content of copper

EDS

14. Northover

1989.

15. Cf. Lechtman and Klein 1999, p. 521, fig. 24.

RECONSTRUCTION

TABLE F.5. EDS ANALYSIS OF COPPER PRILLS IN SLAG SAMPLES Sample

Cu

A05

OF

THE

COPPER

TABLE

F.S

As

93.66

2.81

2.65

0.88

0.00

B05

88.96

A05

90.36

1.30

6.99

1.25

0.10

B05

A05

94.74

2.34

2.57

0.35

0.00

A05

94.82

1.36

2.85

0.74

A05

95.67

0.71

2.90

A09

96.62

2.00

0.83

A09

96.05

1.01

2.46

A09

95.06

0.65

A09

96.43

All

89.26

A14

87.94

A14

Ni

As

Ni

2.30

7.79

0.89

86.70

1.95

10.47

0.81

B05

72.49

1.79

22.84

2.76

0.23

B05

91.36

1.23

6.82

0.53

0.72

0.00

B05

67.11

1.31

31.00

0.38

0.37

0.17

B17

64.82

1.78

32.61

0.67

0.32

0.16

B17

74.15

0.74

24.75

0.27

4.04

0.09

0.15

B17

67.07

3.85

26.26

2.72

0.43

2.75

0.24

0.15

B20

88.62

2.78

7.47

1.12

1.65

7.78

0.41

0.90

B20

3.76

3.64

6.62

0.95

2.20

6.59

0.77

2.50

B20

85.99

4.89

8.19

0.88

90.13

1.28

7.78

0.78

0.03

B20

88.00

2.86

8.09

0.99

A15

97.52

2.29

0.00

0.02

0.16

C01

95.81

2.07

0.96

1.05 1.01

Sample

A15

97.30

2.31

0.09

0.00

0.30

COI

94.53

3.09

1.22

A15

95.91

2.48

0.93

0.45

0.23

COI

95.20

2.69

0.94

1.07

A15

95.33

3.33

0.36

0.90

0.08

COI

94.23

3.37

1.55

0.79

A15

89.52

1.23

8.44

0.81

0.00

COI

93.96

3.28

A15

88.83

0.30

9.83

0.97

0.07

COI

95.43

2.39

1.40

0.67

A15

93.34

1.17

3.65

1.72

0.12

COI

93.26

4.71

1.07

0.89

A15

90.78

1.84

4.00

0.50

2.89

COI

95.71

1.77

1.45

0.92

A16

96.15

3.54

0.22

0.04

0.05

COI

94.75

2.11

2.13

0.84

A16

96.97

3.03

0.00

0.00

0.00

COI

95.11

2.35

1.44

0.90

A16

98.24

1.66

0.10

0.00

0.00

COI

94.97

3.39

0.45

1.10

A16

98.57

1.34

0.00

0.08

0.00

COI

95.94

0.35

2.18

1.29

A17

90.69

2.21

4.94

1.97

0.19

COI

94.89

2.77

1.55

0.71

A18

90.20

2.47

5.24

1.99

0.10

COI

93.99

1.48

3.32

1.03

A18

92.42

3.08

3.19

1.22

0.10

A18

91.23

2.46

A19

61.71

0.45

4.86 23.11

1.25

0.20

14.70

0.03

Results

341

(cont.) Fe

Fe

PROCESS

SMELTING

0.48

normalized.

prills is also variable, but most values are between 0.5% and 2%. Iron content has amean value of 2.2% but reaches 5% in some cases. No -iron to their dendrites were identified microscopically, possibly due extremely small size.16 The only exception is a large prill (Diam. = ca. 1 mm) found in sample A18 containing massive -iron dendrites and sulfidic inclusions in a copper matrix. The composition embedded of this prill was found by EDS analysis to be 69.7% Fe, 23.7% Cu, 3.9% As, and 0.8% S. Finally, small matte particles are often embedded in copper prills, but this phase is 16. Roeder,

Sculac,

and Notis

1984.

usually present

as "rims" surrounding

the metallic

prills (Fig. F.6).

APPENDIX

342

F

Figure F.6. Sample B18: Copper massive showing prills of a second arsenic-rich

precipitation phase

(Cu3As). The prill is surrounded by matte. Reflected light. Scale:

Liquid

10 micrometers

'blister' copper

Liquid Cu

Figure F.7. The position of matte inclusions

wt % S

Matte

present

Chrysokamino Cu-Fe-S

in in the

slags

from

ternary

Inclusions

Although matte inclusions are common, they are significantly less common than the copper prills with which they are often associated (see above). Most inclusions have characteristic Widmast?tten structure resulting from solid-state

precipitations from the sulfide solution.17 As shown by the results analysis (Table F.6), the average iron content is 4%, while nickel and arsenic are usually below 1%. The matte compositions deviate from equilibrium compositions having a slightly higher copper content (Fig. F 7). This is due to the fact that matte compositions were acquired using area of EDS

17. Cf. Doonan

1996,

p. 57.

system

OF

RECONSTRUCTION

THE

COPPER

TABLE F 6. EDS ANALYSIS IN SLAG SAMPLES

SMELTING

OF MATTE

PROCESS

INCLUSIONS

Sample

Cu

Fe

As

A04

77.42

2.69

0.60

A05

79.18

1.83

0.30

0.17

A05

78.68

2.20

0.40

0.21

A05

79.37

2.57

0.24

0.00

A09

79.08

2.48

0.17

0.00

A17

72.43

6.33

0.40

0.30

A18

74.26

5.98

0.34

0.25

A18

68.58

9.37

0.05

0.09

A18

75.26

5.44

0.08

0.00

M

A19

79.39

1.65

0.70

0.53

A19

79.64

1.40

0.30

0.28

A19

78.17

1.01

1.66

1.16

A19

77.85

5.18

0.28

0.30

B01

75.55

4.60

1.02

0.07

B01

79.51

1.53

0.96

0.24

B01

78.70

1.30

2.10

0.33

B05

76.85

4.32

1.63

B05

77.90

3.14

0.90

B05

77.11

3.11

2.23

BQ5

76.78

4.27

1.27

B05

78.94

2.08

1.35

B12

80.39

0.95

0.11

0.03

B13

75.66

4.53

0.07

0.04

B17

75.43

1.87

4.44

0.25

B17

78.13

2.38

1.25

0.42

B17

79.53

1.46

0.60

0.53

B17

75.46

4.84

0.94

0.19

B20

79.02

3.20

0.17

0.10

B20

75.72

6.78

0.00

0.01

B20

72.38

9.17

0.77

0.14

B20

75.00

0.29

0.06

B24

68.53

9.24

0.30

0.46

B24

72.47

6.61

0.68

0.69

B24

73.25

6.43

0.44

0.45

Results

normalized

("na"

= not

analysed).

343

APPENDIX

344 scan analyses,

F

en and mechanically incorporating precipitated are visible and trapped phases. Such inclusions in the matte microscopically include globules of metallic copper and a light blue copper-arsenic phase.18 If those phases were not included in the analysis, the copper content of matte

therefore

inclusions would

Undissolved

range from ca. 75% to almost

80%.

Materials

raw materials that were not dis Very few inclusions in the slag represent solved in the slag matrix during smelting. Small silica grains (0.5-2 mm) are with signs of severe fragmentation occasionally present, and their suban were not in the process of dissolution. gular shape suggests that they Only a was identified. In rare of silica with single grain signs ongoing dissolution cases, fine dispersions of green and red copper minerals, probably reflect ing residual copper ore (malachite and cuprite?), were associated with the quartz grains. Remains of iron ore are also present, either asminute, heavily altered ore fragments or,more often, as agglomerations of iron oxides, which have clearly resulted from the decomposition of larger inclusions.19

Other

Inclusions

a mineral associated with slightly Cuprite (Cu20), in the smelting furnace, is uncommon in the slag it in is When found present, usually clearly corroded a corrosion it is that suggesting product rather than

reducing conditions from Chrysokamino. areas of the samples, a

phase formed dur the Delafossite another mineral as process. ing smelting (Cu2O.Fe203), is uncommon in slags from sociated with slightly reducing atmospheres, two Chrysokamino, being present in only samples. Small fragments of are present in four slag samples, from the furnace walls was dissolved

ceramic material the material

suggesting that some of in the slag.Minute gold

are in the identified, usually as dispersions grains (2-5 mm) occasionally a matrix with EDS of copper prills. glass aligned analysis gold grain gave 90% Au, 4.6% Fe, 3.4% Si, 1.0% Ca, 0.8% Cu and 0.6% Pb. Itwas not possible to formulate a coherent system for the classification of the slag samples on the basis of the microstructural features described most to due the of above, mainly heterogeneity samples and mineralogical to facilitate the the absence of obvious, distinct "groupings." Nonetheless, were se interpretation of the analytical results, six typical microstructures textures lected to represent the range of mineralogical and identified phases (Fig.F.8). a correlation between chemical Although composition and microstruc ture exists, it should not to indicate the presence of dif be taken necessarily ferent types of slag in the assemblage. Instead, as the chemical composition of the slag samples falls in an area of the system with three CaO-FeOx-Si02 areas even the iron and olivines, oxides), precipitation slightest (pyroxenes, chemical differences

could have resulted in the formation

of different phases (or different relative abundances of the various phases) during solidification. To be specific, higher levels of iron in the molten slag favor (a) the forma tion of olivines 2(Ca, rather than pyroxenes (Ca, and Fe)O.Si02 Fe)O.Si02

18. Cf. Lechtman and Klein 1999, p. 515, fig. 19. Cf.

14.

Rehren, Hauptman, Schmitt-Strecker 2003, p. 204,

and fig. 6.

RECONSTRUCTION

Figure F.8. Typical microstructures found in slags from Chrysokamino. Reflected light. Top left:A04, A07, A08, A09, A17, A20, A21, A25, A29, B01, B05, B06, B07, B12, B13, B14, B15, B18, B20, B24, COI. Top center: A03, A15, A16, A18, A23. Top right: A01, A06, B09. Bottom left:A14, A19, A28, B17, B26. Bottom center: A02, B03, B22. Bottom right: A10, All, B02,

Bll,

B23.

A12, A13,

Scale: 50 micrometers

OF

THE

COPPER

SMELTING

PROCESS

345

and/or wustite. Conversely, higher levels (b) the precipitation of magnetite of silica and/or lime promote the formation of pyroxenes, the suppression of the formation of free iron oxides, and the presence of unfused silica grains due to the supersaturation

of the liquid slag.

DISCUSSION OF RESULTS: MASS BALANCE investigation of the mass balance of a smelting process is a powerful to visualize the nature and relative quantities of materials going in and out of the furnace. The approach followed here is based on the simple

The

means

premise that all elements present at the output of the process (i.e., in the slag and metallic phases) should be accounted for by materials present at the input (raw materials). Therefore, through the comparison of copper it is possible either to confirm their ores, slag, and metallic inclusions, chemical similarity or to postulate the use of additional materials that do not appear in the archaeological

Copper

and

Ores

record.

Slag

A

the chemical composition of the gangue striking difference between in the copper ores and the slags from Chrysokamino is that the minerals latter contain significant levels of iron oxides and lime (Fig. F.3). Unless that all three copper ore samples included in this study are from those actually smelted at Chrysokamino, different the con entirely must iron in levels of and lime the be accounted for by sistently high slag it is assumed

present in the furnace during the process. The role of fuel ash as a source of lime in early copper smelting slags has been emphasized in the literature but can only account for some 6%-8% of CaO in the another material

bulk composition.20 20.Merkel

1990;Merkel and 1999; Rothenberg Shugar 2003.

during but

only

smelting might a minor

one.

clay from the furnace wall have been an additional source of lime in the slag,

Erosion

of the calcareous

APPENDIX

346

F

A plausible samples derives

for the chemical composition of the slag interpretation from a consideration of the function of the fragment of iron ore found at Chrysokamino. As already mentioned, the presence of iron ore at a copper smelting site could reflect the use of such ores as fluxes, since the predominant to the results of EDS

especially

gangue mineral in the copper ore is silica. ore analysis, the iron sample contains 67%

According FeO and 23% CaO, so its use as a fluxing agent in the smelting process some material from the erosion (together with the fuel ash and perhaps iron and lime levels in the of the furnace wall) would increased explain the slag. Another is that the fragment of goethite has interesting hypothesis derived from the local weak iron mineralization hosted in the Plattenkalk/ contact

(see above). Detailed comparative analyses using geological are necessary in order to examine this samples possibility.

phyllite

Metallic

Phase

and

Copper

Ores

A

the chemical composition of the metallic comparison between prills found in the slag and the composition of copper ores reveals yet another discrepancy. All copper prills (except for the prills in sample A16) contain high levels of arsenic and ca. 1% nickel. Yet those elements were not pres ent at significant levels in the three copper ores analyzed. Although the limitations of the EDS technique for the analysis of elements at levels near or below 0.5% should not be ratios in the ignored, the Cu/As and Cu/Ni ores are far too low to have resulted

in the percentages of As and Ni found conclusive results would derive from the analysis

in the copper prills. More or like AAS, of more copper ores and the use of techniques ICP, NAA, for chemical analysis. On present evidence, however, we suggest WDS that the arsenic and nickel found in the copper prills did not come from the copper ores but from another material that was added to the charge. is the of lead isotope analysis also results suggestion supported by of slag samples from Chrysokamino show significant (App. C), which ores of various sources. the from variability, probably reflecting smelting it is assumed that the copper ores from the various sources were Unless these elements all characterized by similar arsenic/nickel concentrations, This

were

probably introduced into the smelting furnace separately. It should be noted that the suggested addition of arsenical minerals and iron fluxes could have influenced the isotopic composition of the slag (and the smelted ore rendering any comparison with deposits misleading. However, ore the source(s) maybe problematic, although identification of the specific a strong is from the isotopic variability characterizing Chrysokamino slags

copper),

for the smelting of ores from multiple deposits. present, no evidence from the site indicates the type of arsenical

indication At

stud that might have been used at Chrysokamino. Experimental ies on the production of arsenical copper have shown that various types of arsenic-bearing minerals (including arsenides, arsenates, and sulfarsenides) can be co-smelted with copper ores to produce copper-arsenic alloys.21 In

minerals

fact, that early Aegean metallurgists cal minerals has been established

and exploited such arseni Doonan and colleagues,22 by

recognized recently

21. Lechtman and Klein 1999; Shimada andMerkel 1991; Rostoker and Dvorak 22. Doonan

1991. et al. forthcoming.

OF

RECONSTRUCTION

THE

COPPER

SMELTING

PROCESS

347

near presented evidence from the EM I-II site of Poros Katsambas Knossos suggesting production of arsenical copper by addition of roasted l?llingite (an iron arsenide) in melted copper. in the arsenic content of prills results from the uncon The fluctuation

who

trollable formation of the volatile compound the process and As203 during does not, therefore, reflect variable arsenic content in the raw materials. Nickel, on the other hand, is usually collected in the metallic phase during so its low level in the prills suggests a similarly low consistently smelting, level in the raw materials.

THERMODYNAMIC PROFILE AND OPERATIONAL PARAMETERS of the complex physicochemical Investigation phenomena taking place a at the is process necessary step for the during smelting Chrysokamino reconstruction of the technology for the extraction of copper employed from its ores. For this, it is necessary to make accurate estimations of the during the process and to explain how they would of cupriferous ores into a copper-rich promote both the transformation and its from the product separation slag phase. prevalent

conditions

Temperature The

standard method

for the estimation

of the operating temperature of the calculation of the liquidus tem

ancient

involves smelting processes of the The perature slag. liquidus temperature is, in turn, obtained from to the the phase diagram corresponding slag under investigation. For this a Bachmann23 has purpose, proposed procedure that allows the selection of the most appropriate subsystem in the system based Al203-CaO-FeO-Si02 on the relative abundances of the various oxides in the present slag. this is a useful approach, it should not be Although employed uncriti cally in the study of early copper smelting processes. First of all, as Bach mann himself argues,24 it is necessary to examine whether unfused materials common in (very prehistoric slags) are present in the slag samples because their inclusion in the calculation of the liquidus temperature will produce overestimated

in the moderately reducing characterizing early copper smelting practices, the first phase to the liquidus temperature) precipitate from the liquid slag (thus determining is very often magnetite. This compound, however, is not always included in the ternary diagrams employed in archaeometallurgical studies. This smelting

temperatures.

Secondly,

conditions

is because modern

phase diagrams usually describe slag systems in equi iron and therefore under strongly reducing conditions =10-11-10"12 atm). Under such conditions, however, iron is present in (p02 the slag in its ferrous state, and no magnetite a is precipitated. Moreover, recent study of the of liquidus temperature Al203-MgO-CaO-FeO-Fe203 and Yazawa25 has shown that the effects of common Si02 slags by Kongoli or CaO) on the oxides (e.g., liquidus temperature of the slag A1203, MgO, librium with metallic

23. Bachmann 24. Bachmann 25. Kongoli

1980. 1980, p. 110. and Yazawa 2001.

may be different

under reducing

and oxidizing

conditions.

F

APPENDIX

348

use of of the liquidus tem phase diagrams for the determination case in is valid the of the from because of the perature slags Chrysokamino limited presence of unfused materials. Given the abundance of magnetite in the slag samples, we employed a phase diagram26 that corresponds to = 10"8 atm). The solid phases pre moderately reducing conditions (p02 dicted by this diagram are in good agreement with the mineralogy of the The

our selection. From the slag samples, therefore confirming phase diagram, the liquidus temperature of most slag samples is found to lie between 1150 and 1230?C. In silica-rich samples, the liquidus exceeds 1300?C, but this value is probably an overestimation due to the increased presence of unfused silica grains in them. Overall, raising the temperature to ca. 1300?C would have produced a slag of sufficiently low viscosity.27 Such temperatures were attainable in the Early Bronze Age.28

Redox

Conditions

(i.e., the reducing or oxidizing capacity study of the redox conditions of the furnace atmosphere) during smelting is usually based on the inves tigation of elements exhibiting multiple valency states in typical copper

The

the most important being copper and iron.29 Copper conditions, = ca. 10"5 atm associated with phases moderately reducing conditions (p02 as are rare at 1300?C), such cuprite (Cu20) and delafossite (Cu2O.Fe203), in slags from Chrysokamino (see above), suggesting that conditions were smelting

more

reducing.

At more

estimations of the partial pressure of conditions, temperature can be deduced from the relative abundance

reducing

oxygen at a given of ferrous and ferric cations, taking into consideration activities by other oxides present in the slag.30 Using

the effects on their the Fe27Fe3+

ratio

and extrapolating from the data spectroscopy by M?ssbauer in the literature,31 a range of values from 10"75 to 10"95 atm p02 at 1300?C is derived. of the Similar redox conditions can be deduced from a consideration

provided available

copper content in the matte, which ranges from 75% to almost 80%, ex on matte cluding the precipitated phases (see section analysis). Such matte (i.e., when co-present with metallic ca. 10"9 to 10"7 atm.32 from values Finally, the copper) correspond p02 an content in copper prills, with iron average value of 2.2%, is also compat redox conditions.33 ible with such intermediate compositions

under metal

saturation

to

26. Kongoli and Yazawa 2001, p. 587, fig. 11. 27. Cf. Kresten 1986, p. 44, fig.

28.Merkel Thermodynamic

Profile

of

the

Smelting

Process

values calculated above are compatible with the phases identified p02 in the slag, so they probably describe accurately the local redox conditions at the area in the bottom of the furnace where the slag was collected. However, the numerous perforations on the ceramic shaft, on the one hand, and the combined action of the pot bellows and the strong winds, on the other,

The

would

have created extremely complex furnace during the smelting process.

and variable

conditions

inside the

29. Hauptmann, 1989. Bachmann

and

Weisgerber,

30.Turkdoganl983,p.237. 31. Timucin and Morris p. 3196, figs. 11,12. 32. Sridhar, Togurl,

1970,

and Simeonov

1997, p. 195, fig. 5;Yazawa 1980, p. 379, fig. 2. 33. E.g., cf.Merkel Shugar2003.

1.

1990.

1990

and

of

reconstruction

the

copper

smelting

process

the complete profile in the p02 /temperature can be investigated through experimental only Chrysokamino safe to assume that in contrast to a typical shaft furnace, the more subjected the descending charge to oxidizing perforations that resembled typical conditions of ore-roasting processes. In Although

349

furnaces work,

at it is

presence of conditions

fact, similar furnaces have been used for the roasting of copper perforated cylindrical no on the cop ores.34 Such conditions would have had significant effect per carbonates or the iron hydroxides (apart from their calcination due to the increasing temperature), but they would have burned off much of the sulfur present in the charge and removed much of the arsenic as As203.35 Near the bottom of the furnace, the atmosphere was probably more reducing and the temperature much higher, enabling the formation of copper. The reducing slag and the reduction of copper oxides to metallic a cover conditions and the presence of of liquid slag then allowed most in the metallic of the remaining arsenic to be collected copper. Nickel

was

in the liquid copper. A small amount of entirely concentrated the determined redox iron, conditions, also entered the metallic mainly by phase. Finally, since iron sulfides readily react with copper oxides producing metallic copper and/or copper sulfides,36 the formed matte phase would gradually become copper-rich. almost

to this model, the output of the smelting process According probably a and consisted of a slag phase, a Cu-As-Fe-Ni alloy phase, high-grade matte. Those are the phases that were identified in the smelting slags from Chrysokamino.

Slag-Metal

Separation

indications

that liquid slag was

tapped from the furnace the during smelting process. Firstly, many slag fragments have characteristic flow features that are typically associated with tapped slag. Secondly, the bands inside the slag samples reflect "foldings" occurring during magnetite Strong

suggest

slag in contact with the air. Finally, the texture phases present in the slag samples is indicative of a fact that could be associated with solidification

the viscous flow of molten of most mineralogical

rapid crystallization, outside the furnace. It should be noted, however, that a significant amount of furnace slag (up to 50% of the total amount)37 probably remained inside the furnace after the completion of the smelting process. 34. Craddock 1995, p. 169;we thankOli Pryce for drawing our to the attention between the similarity furnaces and the fur Chrysokamino naces from Yeke, Katanga. reported 35. Cf. Lechtman and Klein 1999, p. 510. 36. Biswas

and

p. 81; Rostoker, 1989, p. 73.

37.Merkel 38. See Chap.

1976, Davenport and Dvorak Pigott,

1990. 10.

Although was

much

of the copper produced during the smelting process collected at the bottom of the furnace where it formed an

probably a ingot, significant part remained trapped in the slag (both tapped and furnace slag), requiring crushing to achieve further separation. This is state of the slag found at the site, slag indicated by the highly fragmented cakes being absent and slag pieces rarely exceeding 2-3 cm in size.38

The very low copper content found in all analyzed slag samples suggests that the entrapped metal was not finely dispersed in the slag matrix but was instead concentrated in sizable nodules that could be easily collected by crushing the slag into small pieces (1-2 cm) and handpicking. However, the evidence shows that this might not have always been the case. The

APPENDIX

35o

F

at Chrysokamino amount of is probably too large to pulverized slag found have derived from the simple crushing of the slag, and it could suggest that smelters often needed to grind the slag to a very fine size to expose minute copper inclusions. Such fine metal particles would have been impossible to collect by hand; instead, a "secondary" beneficiation process (probably washing) would have been required to separate the metal from the slag. It is not possible, on present evidence, to explain the different degrees of comminution of the solidified slag for the recovery of entrapped copper. process was used for the more Perhaps the laborious grinding/washing viscous furnace slag or was undertaken after a less successful smelting of the pulverized operation. Examination slag from the site (despite the it is chemi obvious analytical difficulties) may help to determine whether different from the larger slag fragments. cally and/or mineralogically are that could There various types of evidence from Chrysokamino include not only be associated with the processes described above. They of stone tools found at the site, but also the two shallow, carved on the bedrock that might have served as depressions

the small number

cup-shaped "mortars" for this purpose.39 These finds, however, could equally have served other purposes including the preparation of the raw materials or even (in the case of stone tools) the destruction of the furnace after the completion of the smelting process.40 it should be emphasized that the overall copper losses To conclude, were at the process extremely low. Chrysokamino during metallurgical Based on the average copper content in the slag samples (ca. 1%), the total amount of copper entrapped in ca. one ton of slag would be approximately

10 kg. Product

The

The main product of the smelting process was a quaternary copper-arsenic iron-nickel alloy. The arsenic content was probably highly variable due to the uncontrollable

This argument is As203. supported by the results of the EDS analysis of copper prills, which show a single slag in the arsenic content even within considerable fluctuation was relatively constant because this ele content The nickel fragment. ment

formation

of the volatile

copper, while iron probably due to the variable redox conditions. As high showed minor fluctuations iron contents make copper very brittle,41 an additional refining stage was necessary, although there is no evidence, at present, to indicate that this process was conducted at the site. Some arsenic was also removed in the is generally

concentrated

in the metallic

fumes during the refining process, although if a reducing atmosphere was retained the loss would have been insignificant. The chemical composition in the Aegean. is not unknown of the copper prills from Chrysokamino of arsenic above 2% and nickel at about 1% in copper has also in the copper prills of a slag fragment from the EBA smelting site of Skouries on Kythnos.42 A similar chemical pattern is found inmany

Co-presence been found

Early

Minoan

Whether

artifacts.43

similar raw materials this reflects the use of mineralogically of the very same sources is impossible to say on present

or the exploitation

39. Betancourt

et al. 1999,

pp. 354

40. Betancourt

et al. 1999,

p. 366.

366.

41.

2001. Papadimitriou et al. 1985, p. 90, 42. Gale 43. Branigan

1968,

p. 48,

table table

3. 15.

OF

RECONSTRUCTION

THE

COPPER

SMELTING

PROCESS

351

In any case, this similarity suggests that technologies comparable to that employed at Chrysokamino might have also been used in other areas in the Aegean for the production of arsenical copper during the Early Bronze Age. evidence.

was secondary product of the smelting process at Chrysokamino on present evidence, we cannot estimate matte, although, high-grade in each smelting operation. If sulfur was mainly the amount produced A

a

introduced into the charge by the residual sulfides present in the copper ores, then based on the analysis of the ore samples, the amount of matte If, however, the arsenical produced would have probably been minimal. mineral added to the charge was a sulfarsenide/sulfarsenate, then more matte would have been produced. Further work is required to investigate presence of matte in the output of the smelting remains. absence in the metallurgical

the possible its apparent

product

and

CONCLUSIONS was

the locus of highly sophisticated smelting activities raw careful selection and of and ma materials mixing by in the furnace to enable the production of the conditions of

Chrysokamino characterized

nipulation arsenical copper and its separation from the gangue materials with minimal copper losses in the slag. The copper ores used in the smelting operations were copper carbon ates, occasionally containing some residual sulfides, associated with quartz and small amounts of iron (hydr)oxides. Calcite bearing iron ore, similar to was as a flux. It is fragment found at the site, probably used possible that this material was collected from the weak iron mineralization hosted in the

a

contact, which we located during a geologi nearby Plattenkalk/phyllite an arsenical cal reconnaissance of the Chrysokamino peninsula. Finally, mineral containing low levels of nickel was probably added to the smelting the nature of this mineral cannot be established on charge. Unfortunately, present evidence. However, arsenical copper at the EM

the use of iron arsenides site of Poros Katsambas44

for the production of could suggest that a

similar mineral was

at Chrysokamino. employed effects of air penetrating from the perforations on the shaft, the air blasts from the bellows, and the fuel supply created a very complex p02/temperature profile inside the smelting furnaces at Chryso kamino. The maximum temperature attained (near the combustion zone) was as deduced from the chemical 1300?C approximately composition, The

combined

and texture of the slag. The redox conditions were probably mineralogy, was ca. 10"8 atm, but at the bottom of the hearth, the variable; highly p02 more at conditions oxidizing probably prevailed higher levels in the furnace. as the significant amount of arsenic and sulfur was, therefore, removed charge descended in the shaft. The output of the smelting process consisted a a of aCu-As-Fe-Ni alloy, lime-rich ferrosilicate slag, and seemingly small a amount of high-grade matte. The slag was periodically tapped to enable was concentrated at the bottom continuous smelting process. Most metal

A

44. Doonan

et al. forthcoming.

of the hearth

forming

an

ingot, but some remained

entrapped

in the slag,

352

F

APPENDIX

Arsenical

mineral

nickel

Fuel (charcoal?)

Iron ore +

carbonates +

Copper

+

residual sulfides

calcite

SMELTING

Air

G

Tapped

slag

Furnace

Matte

ingot

Copper

slag

ases

t 9

Crushing +

Grinding +

handpicking

Crushed slag

washing

Copper nodules

Pulverized

Copper prills

Cu-As-Fe-Ni

slag

alloy

F.9. Figure tion of the process

at

Schematic proposed Chrysokamino

representa copper

smelting

RECONSTRUCTION

OF

THE

COPPER

SMELTING

PROCESS

353

to be crushed and/or pulverized to enable further sepa a simple remelting in relatively reducing atmosphere would have been necessary to remove the iron present in the smelted copper without no evidence of this process has causing significant arsenic losses. However, been found at the site.

which

then needed

ration. A

the technological questions set at the beginning of this report Although have been explored in varying degrees of detail, it is possible to propose a of the smelting process at Chrysokamino that provisional reconstruction could serve as aworking hypothesis for future research (Fig. F 9).More ar analytical, and experimental work is essential for an enhanced of the metallurgical processes taking place at the site. understanding chaeological,

APPENDIX

G

of

Register

Anthropogenic

Features byPhilip P. Betancourt, Lada Onyshkevych, and William B. Hafford

A

careful

between

of the Chrysokamino territory was conducted It included an instrument survey as well as a

examination 1995

and 1999.

groups of terrace walls), of the natural landscape. The workshop

visible

a sherd scatter, or some other human

alteration

the territory that included the metallurgy was gen primarily of low bushes (Fig. G.l). Visibility and features were clearly visible. A total of 41 Anthropogenic

vegetation consisted

erally good, Features were

feature (AF) was anthropogenic terrace walls or architecture (including

features. An

recording of anthropogenic defined as a location with

within

the Chrysokamino territory. Many of them noted the previously by Kavousi-Thriphti Survey.1 The close view of the landscape presented here may be regarded as an extension of the previous survey, focusing on a small part of the originally surveyed

were

recorded within

at locations

territory in greater detail. It differs from the earlier survey in that itmakes note of features from all not just from the Bronze periods, Age. The fol were recorded (for their locations, see features lowing Fig. 16.2). AF

1.Terrace walls

AF2.WelP AF AF

3.Wall segments, perhaps 4. Threshing floor

a

boundary wall3

AF 5. Field house (?) AF

6. Terrace walls

AF

7. Field house

AF

8. Sherd

AF

9. Venetian/Ottoman

AF

10.Mound

AF

scatter4 to modern

of stones, probably 11. Terrace walls

AF 12.Well 1.Mook 1992,1993b, 2. Haggis 3.Haggis 4. Haggis 5. Haggis

andHaggis 1990;Haggis 1995,1996b, 2005. 2005, locus 49. 2005, locus 49. 2005, locus 52. 2005, locus 49.

AF

13. Field house

AF

14. Terrace walls

AF

15. Field house

AF

16. Field house

AF

17. Terrace walls

(Chordakia)

cave (see App. K) a field house5

APPENDIX

356

G

Metallurgy Workshop

Figure G.I. The metallurgical work shop

and

its

nearby

from the habitation AF

18. Threshing

AF

19.Wall, perhaps a boundary wall 20. Threshing floor

AF AF

territory,

floor

AF

21. Boundary walls 22. Terrace walls

AF

23. Metallurgy

workshop6

AF 24. Field house AF

25. Terrace walls

AF

26. Boundary 27. Boundary

AF

AF28. AF

wall wall

Sheepfold (Mandraf

AF

29. Chrysokamino 30. Terrace walls

AF

31. Terrace walls

AF

32. Oval

AF

33. Field house

AF AF

34. Theriospelio 35. Terrace walls

AF

36. Terrace walls

AF

37. Terrace walls

AF

38. Terrace wall

habitation

site8

enclosure9 cave10

AF 39.Well AF

40. Road

AF

41. Metal

and house walls fence

The methodology devised in Chapter 15. The special were dictated by the project as the need to determine beneficiation

for the recording of these features is discussed circumstances for this highly specialized survey

s goals, the topography, and other factors, such or if any tiny features associated with mining ore were in of the vicinity of the smelting workshop. present

the main considerations that determined the special methodology Among were the small size of the to be examined and the existence of a territory surface The survey. prior, traditional, regional previous survey had already

as seen

site (AF 29)

6. Haggis 2005, locus 88. 7. Haggis 2005, locus 50. 8.Haggis 2005, locus 50. 9. Haggis 2005, locus 52. 10.Haggis 2005, locus 55.

REGISTER

OF

FEATURES

ANTHROPOGENIC

357

pottery, and few artifacts remained on the surface. Consequently, pottery collection for this project was limited to a few specific features, a as selective rather than com sampling strategy that could be described collected

that the pottery is selectivity decreases the probability representative.11 On the other hand, since all pottery collecting has value,12 some of the evidence permits different conclusions from those reached by This

prehensive.

the survey of Haggis, and the new collection be illustrated with specific ceramic evidence.

allows

to

these conclusions

was designed tomeet the specialized needs of this recording system were means with aTotal Station Locations recorded survey. by electronic a in the established for 1995. feature was Each manmade using region grid a number, and groups of terrace walls were lettered given sequentially from often defined the highest to the lowest on the slope of the hill. Locations The

as "sites" in traditional

at the surveys were divided into several components were later joined stage of investigation.13 Their designations preliminary reasons ease in it. For when warranted several study (including together were in and all references manmade features map digital recording), given ones such as the metal fence, AF 41). (including modern Dates for theMinoan and post-Minoan pottery were assigned by Betan court and Poulou-Papadimitriou respectively. feature numbers

REGISTER OF ANTHROPOGENIC

FEATURES

1 Terrace walls

AF

Fig. G.2 Grid coordinates: E5459 N4851

H.:

(at center

120 masl

(at center of area).

of area).

Visibility: good. ca. 70 m

Dim.: Date:

north-south

Pottery: Comments:

none

All

ca. 7,500

east-west;

m2.

collected. Four

parallel

to the Chrysokamino fence.

x 108 m

unknown.

the walls

are west

terrace walls

agricultural

of the modern

gate

area, uphill from the road and downhill from the modern

here

are constructed

of

irregularly

blocks

shaped

of carbonate,

ranging in size from 0.06 x 0.06 x 0.06 to 0.3 x 0.3 x 0.3 m, with the largest stones positioned the walls Wall the

at the bottom

Soil

away. on the wall highest three walls. It is the least well

AF

lower

bedrock

of the walls.

has

over

eroded

the walls,

and parts

of

have washed la,

the

slope,

was

preserved

made of

with the

smaller

four walls,

stones and

than it uses

in its construction.

Wall AF lb is 27 m downhill (north) ofAF la. It is preserved to amaximum height of 1.1 m, and it is up to 0.2 m thick. It is founded on bedrock. Wall AF lc is 16.2m north ofAF lb. Its greatest preserved height is0.6 m (0.2m 11. This authors,

has been

e.g., Binford

noted 1968a,

by many p. 13;

Hope Simpson 1983; Cherry, Gamble, and Shennan

1978.

thick). Bedrock is used only in small parts of the wall. Wall AF Id is 16 m north of AF lc. It is the same thickness as the others, with a preserved height of up to 0.7 m. Well

AF2

12.Whallonl983.

13.On the difficulties of site defini tion, pp.

see the comments 17-21.

byWells

1996,

Grid coordinates: E5363.5 N4896.6. H.:

124 masl.

Visibility: good.

Fig.G.3

APPENDIX

358

G

Figure

G.2.

Location

of

anthropo

genic features AF 1,AF 2, and AF 3, uphill from the road at the right of the photograph. View looking north from the intersection marked AF 40 on Figure 16.2.

1.45 m

Diam.: Date: Pottery:

Byzantine none

Comments:

interior, or

2.2 m

exterior;

H.

m.

0.45-0.80

later.

collected.

The

southwestern

slope

of Chomatas

has

two

peaks.

A

well

is located on the saddle between them, uphill from the habitation site (AF 29). The wellhead is built of unworked fieldstones without mortar, forming a circular It includes

feature. x 0.15

both

and flat

blocky

stones

of various

sizes,

up

to ca. 0.3

x 0.4

m.

AF 3 Wall

segments Fig. G.4 Grid coordinates: E5362.4 N4890.7 H.: 124.4 masl (at center of wall).

(at center of wall).

Visibility: good. Dim.: Date: Pottery:

ca. 82.2 Byzantine none

m

north-south or

(length

of reconstructed

wall).

later.

collected.

Comments: Three segments of wall are located near thewell recorded asAF 2. They The

cannot stones

be are

terrace walls very

irregular,

because and

they the wall

are on can

a flat never

area have

at the ridge stood very

of high.

the hill. Wall

AF 3a meanders along the saddle for 23 m. Its height rangesfrom 0.07 to 0.17 m, and

REGISTER

Figure

G.4.

Segment

of a wall,

OF

ANTHROPOGENIC

FEATURES

359

pos

sibly a boundary marker (AF 3) its thickness is 0.1 to 0.3 m. It ismade of unworked fieldstones laidwithout mortar. Wall AF 3b is 4.2 m south ofAF 3a. It consists of seven stones in a Une.Wall AF 3c, 2.7 m long and 0.2-0.4 m thick, is northwest of AF 3b. It ismostly buried. The most likely explanation for these sections of wall is that they form a boundary marker, dividing the public land on the coastal side of the ridge from the private land inland. The well (AF 2) is just barely inside the wall (land side). The boundary today,marked by amodern fence (AF 41), makes all of the hilltop public land, including this feature. AF 4

Threshing floor Fig. G.5 Grid coordinates: E5296.9 N4993.2.

H.:

121.8

masl.

Visibility: good. Diam.: Date: Pottery:

ca. 7.9 m. probably none

Venetian

or Ottoman.

collected.

Comments: This oval or circular threshing floor nearAF 5 has a floor of hard packed soil and a raised soil perimeter with vertically placed phyllite slabs, 0.25 0.6 m in length. Slabs remain standing in place at the north and west. They are all unworked.

Figure G.5. Destroyed floor

with

a few

standing

the perimeter (AF 4)

threshing stones

at

APPENDIX

360

G

Figure G.6. Collapsed field house

(AF5) AF5

Fig. G 6 Grid coordinates: E5269.3 N4983.6. Field house

H.:

117 masl.

Visibility: good. Dim.:

9.1

Date: Pottery:

x 9.7 m.

Comments:

or Ottoman.

Venetian

probably none

collected. remains

The

of a one-roomed

structure

of drywall

built

masonry,

primarily of flat phyllite slabs, lies between AF 4 andAF 7. Stones are small in the north and west walls (average size 0.15 x 0.10 x 0.05 m) and larger in the eastern portion of the north wall (average size 0.9 x 0.25 x 0.4 m). Although it is not certain that

the

north

preserved

wall

meets

the

natural

bedrock

outcrop

on

the

it

east,

is likely that this outcrop was used as part of the east wall. This bedrock outcrop jogs

in to the west,

but

a

it maintains

primarily

line for a distance

north-south

of

5.9 m. It is 1.2 m high at its highest point, but no evidence of awall on the bedrock survives. Although the north wall is the best preserved one, even this wall is not more than two courses high inmost places. It is 9.1 m long (height between 0.29 and 0.70 m). The poorly preserved west wall, which is 9.7 m long and 0.7 m thick, is laid on bedrock. The other walls have collapsed. The field house is near a threshing floor (AF 4) and terraced fields (AF 6), suggesting

that

this group

of features

Fig. G 7

AF 6 Terrace walls

(at center of area).

Grid coordinates: E5257.0 N4968.1 H.:

(at center

109 masl

a unit.

forms

of area).

Visibility: good. Dim.: Date:

ca. 22 m probably none

Pottery: Comments: built

m

east-west;

ca. 600 m2.

or Ottoman.

collected.

West

of unworked

x 33.3

north-south Venetian

of AF

fieldstones,

terrace walls. 5 is an area of agricultural are about 8-10 m are apart. They mostly

The

walls,

carbonate

blocks, up to 0.3 x 0.3 x 0.2 m in size, and they follow the contours of the hill, occasionally

incorporating

bedrock

into

their

construction.

This

system

of terrace

walls, inwhich parallel walls are placed on the slope of a hill, is called a check-dam. The highest wall (AF 6a) stands 6-7 courses high in some places (H. 0.8 m).

REGISTER

OF

ANTHROPOGENIC

FEATURES

361

Figure G.7. One of the terrace walls atAF 6 AF 7

Field house Figs. G.8, G9 Grid coordinates: E5253.8 N4982.2

H.:

111.2

(at center

masl

(at center of area).

of area).

Visibility: good. ca. 4.4 m

Dim.: Date: Pottery:

east-west.

or Ottoman.

Venetian

probably none

x 3.6 m

north-south

collected.

Comments: The lower parts of thewalls of a small building with one room are downhill from AF 5. The structure is built of drywall masonry using flat phyllite It is oriented

slabs.

northwest-southeast.

The

southeast

wall,

4.15 m

long,

contains

a doorway. The interior is filled with stones fallen from the walls, obscuring the floor. The northwest wall is 4.16 m long, and the northeast wall is 3.45 m long. Wall thickness varies from 0.51 to 0.69 m. Only the lower parts of the walls are preserved. The preserved wall height is up to 0.7-0.9 m. It is possible thatAF 5 succeeded this poorly preserved structure. Sherd scatter

AF 8

Fig. G.10 Grid coordinates: E5390 N4530 (at center of area). H.: 112 masl (at center of area). Visibility: good. ca. 85 m

Dim.:

x 170 m

north-south

east-west.

Date: FN-LM III. Pottery: recorded with AF 32. Comments: A scatter of EM-LM sherds (recorded asAF 8) occurs on the red soil south of the habitation site (AF 29). Except for AF 32, an oval structure near The the form

the

southern

sherds oval a

side

are

present structure. No

fairly

even

of

this

across

specific across scatter

no

scatter, the

concentration the

entire

or other

architecture area

entire

from

feature

the habitation

seems area. Their

to be present, fabrics

range

is visible. to south

site as

the

in date

of

sherds from

FN to LM III.The discussion inChapter 21 suggests that this area is the garden for

the habitation

AF9

site.

Cave Grid coordinates: E5411 N4702 H.:

111.5

masl.

(at center of interior).

3o2

APPENDIX

G

Figure G.8. Collapsed field house (AF 7) on the slope uphill from the metallurgy site. View looking west toward the metallurgy workshop

Figure G.9. Collapsed field house (AF 7). View looking south

Figure G.10. Looking downhill (south) from the habitation site (AF 29) to the location of a scatter of sherds (AF 8) and an oval enclo sure (AF 32)

OF

REGISTER

FEATURES

ANTHROPOGENIC

363

Visibility: good. Date:

Venetian none

Pottery:

x 6.6 m to modern.

collected

A

Comments: 1.5 m

entrance

north-south to Ottoman

ca. 9.2 m

Dim.:

from

the

high,

facing

surface.

cave

small manmade west

east-west.

cut

the soil has a semicircular phyllitic sea. The interior is oval. It is 1.8 m

into

the

toward

m deep. The floor is loose soilwith a high near the entrance, 9.2 m wide, and 6.6 few

scattered

The

stones.

phyllite

excavation

The

K).

(App.

angular

cave was excavated in 1997 under the supervision of Brigit Crowell to terrace

proximity

AF

discovered 22b

that the

suggests

it was cave

out

carved

is contemporary

in

stages, this

with

and

the

feature

to Ottoman).

(Venetian

of stones

AF 10 Mound

Grid coordinates: E5483.5 N4760.7. H.:

139.8

masl.

Visibility: good. ca. 6.7 m.

Diam.:

seeAppendix J, J-6 and J-12. Comments: An artificial mound of stones 1.2 m high is found on thewestern ridge of Chomatas, uphill from the habitation site (AF29).The mound seems to be a collapsed field house, and a portion of one wall (0.8 m long) is visible at the Date: Ottoman:

south/southwest.

AF

near

from

Pottery

the feature

is Ottoman.

11 Terrace walls Grid coordinates: E5504 N4766 H.:

(at center

137 masl

(at center of area).

of area).

Visibility: good. Dim.: L. 1.3 m (wall 11a); 2.3 m (wall lib). unknown.

Date:

none

Pottery:

collected.

Comments: Agricultural terracewalls in the public land on Chomatas (outside the modern fence, AF 41) show that the boundary has not always been where it is now. Two

of terrace walls

fragments

are

preserved

is 0.25-0.3 m high and 0.15 m thick; wall AF lib and

high)

m

0.2-0.35

thick. The

walls

are not

at this

location.

Wall

AF

11a

is 0.3-0.4 m high (2-3 courses

completely

preserved.

Well

AF12

Grid coordinates: E5513.5 N4800.1. H.:

128.5

masl.

Visibility: good. ca.

Diam.: Date: Pottery:

1.0 m.

probably none.

Venetian

or Ottoman.

Comments: A well is 5.1 m north of the field house called AF 13.The south ern part of the wellhead is partly preserved, with a height of 1.05 m, enclosing an area 1.0 m in diameter. The walls are 0.55 m thick, built of fieldstones in a drywall technique.

AF

13

The

stones

measure

0.1

x 0.1

Field house

Grid coordinates: E5515.8 N4793.2. H.:

130.5

masl.

Visibility: good.

x 0.05

m

to 0.6

x 0.4

x 0.18

m.

APPENDIX

364 Dim.:

m.

4.2x5.1

Date: Pottery:

G

probably none

Comments:

or Ottoman.

Venetian collected.

The

a small,

of

ruins

one-roomed

are 5.1 m

building

south

of

AF 12.The building is oriented from north to south, with an opening 0.9 m wide east wall.

in the

The

rubble

are 0.55

walls

m

thick,

is filled

AF

from

of a terrace

part

ably

stones

with

the

wall,

east-west

An

walls.

collapsed is near the

to three

up

preserved

high (0.55 m). Stones are up to 0.3 x 0.3 x 0.2 m in size.The

stones

interior of the room 6.2 m

wall

long,

prob

building.

14 Terrace walls (at center of wall).

Grid coordinates: E5665 N5047 H.:

84.7

(at center

masl

of wall).

Visibility: good. ca. 91 m

Dim.: Date:

Pottery: Comments: ravine

x 15 m

northwest-southeast

1,200

northeast-southwest;

m2.

unknown. Minoan

downhill

collected.

terraces

of nine

group

from

none

seen;

pottery

A

the modern

a

forms

to

gate

the

of

system of

territory

in the

check-dams

Walls

Chrysokamino.

range from 4.5 to 11.4 m in length, and they are placed at intervals of 3.2 to 9.15 m in horizontal distance and 0.6 to 2.6 m in vertical distance. The lowest few curve

walls

southwest,

slightly in range height x 0.15 m in size. All

walls

or other

houses

AF15

from

0.5

seen

sherds are

features

of the direction following to 1.5 m. Most of the stones on

the

ravine's

are small,

area

in this

surface

the

The slope. ca. 0.2 x 0.2

are Minoan.

No

field

nearby.

Field house

Fig. Gil Grid coordinates: E5636.8 N4843.0.

H.:

113 masl.

Visibility: good. ca. 3 x 4 m.

Dim.: Date:

Pottery: Comments:

Ottoman

Chrysokamino orientation

small Ottoman with

comparison

collected.

or later field the

house

with

less well-preserved

a

single small

x 0.1

0.1

the entrance

with

x 0.1 m

to 0.4

x 0.7

top with

sticks,

brush,

plastic

sheet,

and

gravel.

The

with mud and plaster. The floor is soil and bedrock with back. Wooden

AF

16

buildings

in the northeast

x 0.5 m. The

provides in the

are

beehives

sometimes

placed

wall.

floor measures

Rubble walls 0.5 m thick support a flat roof built of wooden on

room

territory. It is built on bedrock, using the drywall technique. The

is northeast-southwest, from

vary

none

seen;

pottery

A for

information

sizes

or later.

Ottoman

interior

Stone 3 x 4 m.

crossbeams layered is

mostly

covered

a bedrock bench at the

nearby.

Field house

Grid coordinates: E5545.9 N4796.2. H.:

126.7

masl.

Visibility: moderate. Dim.: Date:

ca. 3.9 m

x 4.0 m.

Ottoman.

Pottery: seeAppendix J, J-13. Comments: A poorly preserved building with only the north and west walls still partly in place is built of unworked stones in the drywall technique. Stones range from 0.1 x 0.1 x 0.1 m to 0.45 x 0.5 x 0.15 m. A sherd from the feature is

REGISTER

Figure G.ll.

OF

ANTHROPOGENIC

FEATURES

365

Standing field house

(AF15) Ottoman. Terrace walls on Chordakia

(AF17)

and a threshing floor (AF18)

are

nearby.

AF 17 Terrace walls (Chordakia) Grid coordinates: E5508 N4620 (at center of area). H.: 110 masl (at center of area). Visibility: poor to moderate to good. ca. 296 m

Dim.: Date:

x 141 m

north-south

east-west;

m2.

42,000

to modern.

Minoan

Pottery: Appendix H, AF 17.1 toAF 17.3 andAppendix (EM III to LM IIIB, Venetian, Ottoman). Comments: show many

east

The of

stages

slope

and

rebuilding

m apart. The name Chordakia of

on

sherds

the

terraces

since theMinoan the

is covered

of Chomatas reuse. Most

of

terrace

with

is used for this terraced slope. A

indicates

that

the

region

has

been

period (or earlier?). Sherds published

used

which

walls,

are about

terraces

the

and J-14

JJ-3J-9,

5 to 15

light scattering for

agriculture

in this volume illustrate

range.

The area is planted in olive trees.The olives here and in the adjacent plain called the Kambos were planted after theMarshall Plan provided funds in 1947 for

wells

deep

for

irrigation.14

Before

then,

the

terraces

were

used

for grains.

Wall AF 17a is closest to themodern fence (AF 41). It is preserved to a height of ca. 0.5

to 1 m, with

an average

thickness

of ca. 0.4 m.

It is

into

the wall

in some

of dolomite

mostly

blocks (from ca. 0.15 x 0.15 x 0.1 to 0.8 x 0.8 x 0.2 m). Bedrock

is incorporated

places.

Wall AF 17b is ca. 12-13 m downhill from AF 17a.The preserved part of the built portion of this wall ranges from ca. 0.6-0.8 m. It abuts bedrock cliffs, some of which drop asmuch as 5 m. Dolomite blocks are used in the construction, with stones

up

to ca. 0.3

Wall AF and

14.Allbaugh 1953, pp. 258-263.

up

in the

x 0.25

x 0.15

m.

17c is ca. 6 m down the hill from AF

to ca. 0.3 m construction.

thick.

Dolomite

blocks

up

17b. It is ca. 0.6-0.9 m high

to ca. 0.3

x 0.3

x 0.25

m

are used

APPENDIX

366

G

Wall AF 17d is from 0.7 to 3m high, with a thickness of 0.3-0.4 m. Dolomite blocks

part

than

this. The

is washed

away.

Walls AF more

in of

walls

AF18

x 0.25

to 0.7

up

smaller

height. this

x 0.2 m

and

and bedrock.

phyllite

are

techniques

are

stones

The

northern

to those

similar

used

in the

other

region.

Threshing floor Grid coordinates: E5540.9 N4803.2

H.:

but most

in its construction, some

17f consist of several separate segments, up to 1m or

17e andAF Materials

are used end uses

southern

125.9

(at surviving walls).

masl.

Visibility: moderate. Dim.: 4.9 m (length of surviving walls). Date: Pottery:

AF

probably none

Ottoman. collected.

Comments: A threshing floor is 4.7 m north of the field house recorded as 16. The shape is unknown (not preserved). It has a soil floor. The perimeter, survives

which

in two

only

of

consists

sections,

an average height of 0.5 m and awidth

slabs with

thick,

set

vertically

of 0.35 m. The

carbonate

slabs are not

worked.

AF19 Wall (at center of wall).

Grid coordinates: E5724.9 N4903.3 H.:

129.4

masl.

Visibility: good. Dim.: ca. 28.5 m (length of wall). Date: Pottery:

Byzantine none

or later. collected.

Comments: A longwall, 28.5 m in length, is constructed with many vertically a

set stones,

used

technique

in

and

Byzantine

later

times

in this part

of Crete

for

walls thatwere not parts of buildings. The date for this technique isprovided by sev eral Byzantine walls on Pseira, including awell-dated wall from the 5th to 9th cen tury

AF

a.D.

that

ran

is 0.8 m. The

thickness

across

to west

east

the Minoan

plateia.15

The

orientation

for

to east/southeast. The height is 0.15 to 0.85 m, and the

19 iswest/northwest

wall

is

a

probably

wall.

boundary

AF 20

Threshing floor Grid coordinates: E5495.5 N4736.1. 136 masl.

H.:

Visibility: poor. not measurable.

Dim.: Date: Pottery:

probably none

Venetian

or Ottoman.

collected.

Comments: A threshing floor, mostly stones,

is not

sufficiently

exposed

concealed under a pile of loose field

to be measured.

Its

perimeter

stones

are

vertically

set phyllite slabs. AF 21

Boundary walls

Grid coordinates: E5601 N4840 H.:

G.12-14

119 masl.

Visibility: Dim.:

Figs. (at center of area).

good.

ca. 39 m

east-west.

15. Pseira W,

p. 139.

REGISTER

OF

ANTHROPOGENIC

FEATURES

367

Figure G.12. Sections of wall (AF 21), probably a boundary wall

Figure G.13. View of AF 21 from above

Date:

or

Byzantine

later.

Pottery: Appendix H, AF 21.1 andAF 21.2. Comments: Two walls are west of the field house called AF 15. They are roughly parallel with the modern fence, and they are on the crest of the hill, so they cannot

be

constructed The

terrace

walls.

as double-faced

wall

sections

One

section

appear of wall

have They walls with

an

and

faces,

rubble

a

to be has

two

are the preserved parts clearly the two courses of stones.

fill between

predecessor corner. oblique

of

the modern The

part

fence. that

is east

of

the

corner

is 15.8 m long, 0.1-0.35 m high, and up to 0.3 m thick. It is constructed of car bonate fieldstones up to 0.25 x 0.3 x 0.2 m in size.An additional 15.9 m of wall is west-northwest

of

the

corner.

The

thickness

double stones; its height is up to 0.25 m.

in this

section

reaches

0.55

m, with

APPENDIX

368

G

'"" *>^l^-r The

section

second

Figure G.14.

~?1$*\&?!^;I?^^ of wall

is oriented

It

north/northeast-south/southwest.

is 11.4 m long (as preserved). Stones range from very small (0.05 x 0.05 x 0.10 m) at the south/southwest end to very large (0.7 x 0.8 x 0.3 m) at the north/northeast end. Its thickness is 0.3 m, and it has a height of 0.35 to 0.8 m. Fig. 2.16 Grid coordinates: E5366 N4667 (at center of area). H.: 105 masl (at center of area). Visibility: moderate.

AF 22

Terrace walls

ca. 172 m

Dim.: Date:

Venetian

not

east-west;

m2.

21,000

J, J-4, J-5, and J-7 (from the excavation of AF 22b); a

Pottery: Appendix few Minoan

very

x 165 m

north-south to Ottoman.

and Ottoman

sherds

were

seen

elsewhere

in the

for a check-dam

system

area but were

collected. A

Comments:

between length,

series

15 terrace walls

of

is in a ravine

the cave called AF 9 and the sea.They vary considerably in preserved and

height,

spacing.

While

some

of

the walls

curves

continuous

form

extending to both sides of the ravine, many of them extend only 30-40 m along one

side

spacing

of

the

between

run

ravine

and

walls

can be

into as much

other

terrace

same

in this

walls

as 20 m. The

are

walls

generally

system. The over 2 m

high, although some of them risemuch less (only a single course). Thickness var ies from 0.2 to 0.8 m. One terrace wall (AF 22o) has between 15 and 20 courses and

preserved

stands

1.5 m

high. Where

preserved,

the stones

are

mosdy

carbonate

blocks that range in size from 0.1 x 0.1 x 0.1 to 0.5 x 0.3 x 0.3 m. An excavation was made in terraceAF 22b in 1997 under the supervision of Crowell (seeApp. K). The excavation discovered that the terracewas built in the Venetian or Ottoman period, and at least some of thewall was definitely Ottoman. No

evidence

for

earlier

activity

was

found.

Samples

of

the

are discussed inAppendix L. AF23

Fig. Metallurgy workshop Grid coordinates: E5015 N5001

H.:

1.7 (at center of area).

38 masl.

Visibility: good. Dim.: 34 x 10 to 15 m. Date: FN to EM III-MMIA. Pottery: see Chapter 5. Comments:

see Part

II in this volume.

soil

from

this

terrace

Section of wall (AF 21)

ANTHROPOGENIC

OF

REGISTER

FEATURES

369

Field house

AF 24

Grid coordinates: E5286 N4544. H.:

masl.

118.5

Visibility: good. Dim.:

ca. 2.1

Date:

Ottoman

x 2.1 m. or modern.

none

Pottery:

collected.

of the habitation site (AF29), two rough stonewalls are built

Comments: West up

windows

x 0.3

ca. 0.5

a small

to form

the bedrock

against

x 0.2 m

shelter.

The

two walls

in size. The

northern

wall

are about

1m

built in drywall technique using carbonate fieldstones 0.1x0.1

mto0.3x0.3x0.3m.

The

is from

shelter

two

has

They

high.

small are

ranging in size from 0.1 x

the Ottoman

period

or later,

and people tending crops or animals probably used it as a field house. AF 25

Terrace walls (at center of area).

Grid coordinates: E5747 N4918 H.:

(at center

129 masl

of area).

Visibility: good. ca. 70 m

Dim.: Date:

x 35 m

north-south

ca. 2,000

east-west;

m2.

unknown. none

Pottery:

collected. East

Comments:

15 is an area with

of the road near AF

three

terrace walls,

the

highest of which is close to the probable boundary wall called AF 19.The highest wall (AF25a), mostly washed away, is preserved only two stones high (0.2 m).The building

material,

carbonate

mostly

varies

blocks,

from

0.1

x 0.1 x 0.1 m

to 0.2 x 0.4

x 0.25 m. The thickness is 0.25 m.Wall AF 25b is lower on the hill, to thewest, and it is also poorly preserved. It ismostly composed of large, yellow, irregular blocks of limestone, with a few smaller stones (range 0.15 x 0.15 x 0.1 m to 0.55 x 0.65 x 0.25 m). Two courses are preserved (height 0.55 m) with a thickness of 0.25 m. Between 15-20 m lower on the hill iswall AF 25c, also mostly washed away. It is 2-3 courses high (0.3-0.65 m) with a preserved thickness of up to 0.5 m. AF 26

Boundary wall? Grid coordinates: E5662.8 N4838.4 H.: 109.2 masl (at center of wall).

(at center of wall).

Visibility: good. Dim.: 17.7 m (preserved length of wall) Date: Ottoman (?). Pottery:

Ottoman

pottery

seen;

none

collected.

Comments: A poorly built rubble wall runs perpendicular to the modern road near AF 25. It is only one course high (0.15 to 0.40 m), and its thickness or three rows of stones in some parts (0.35 to 0.70 m thick). up to two incorporates to be The wall is above ground, and it appears fairly recent, part of it has although a washed is It wall. away. probably boundary

AF 27

Boundary wall

Fig. G.15 Grid coordinates: E5677.4 N4868.3

H.:

(at center of walls).

117 masl.

Visibility: good. Dim.: Date: Pottery:

ca.

19.5 m

Byzantine none

east-west

(length

of

preserved

walls).

or later. collected.

Comments: Walls constructed of limestone blocks and vertical slabs of phyl lite in two rows with a fill of rubble between them (compare AF 19 andAF 21).

APPENDIX

37?

G

Figure G.15.

Section of boundary

wall (AF27) iswest of themodern

The main section of wall continues the line ofAF 21, which road. The

the hill,

down

is 0.6 m.The

thickness

and what

runs

wall

appears

primary to be a cross-wall

an east-west

along lies 5 m

from

Une up and easternmost

the

cross-wall is only 1.5 m long. It consists of a single row of (from 0.15 x 0.15 x 0.1 m to 0.3 x 0.3 x 0.25 m). Farther west,

end of the wall. The large fieldstones the main

east-west

meets

wall

another

wall

running

northwest

and

southeast

for

a distance of 7.5 m. This wall also uses two stones for itswidth (up to 0.8 m), and although fewer vertical stones are present, it is probably related to the uphill wall. Two

courses

are

preserved

to a

height

of 0.4 m.

Stone

sizes

vary

from

0.2

x 0.1

x

0.1 m to 0.5 x 0.35 x 0.2 m.

AF28

Sheepfold

Grid coordinates: E5384 N4560 H.:

(at center of feature).

120 masl.

Visibility: moderate. Dim.:

ca. 7.2 m

north-south

x 20.9

m

east-west.

Date: middle of the 20th century. Comments: About the middle of the 20th century, a shepherd built a sheep fold over the south side of the Chrysokamino habitation site (AF29) using stones from the archaeological site.The sheepfold has been recorded by Haggis.16 Itwas dismantled

AF 29

during

the

course

of excavation.

site (Chap. 17) Figs. G.16-18 Grid coordinates: E5388 N4569.

H.:

Habitation

123 masl.

Visibility: good. Dim.: southwest.

ca. 20 m

north/northeast-south/southwest

x 26 m

east/northeast-west/

16.Haggis 2005, p. 115, locus 50.

REGISTER

Figure G. 16. General view of the habitation site (AF 29) before exca vation

Figure G.17. The habitation site (AF 29) before excavation, showing the east

Figure

side,

G.18.

looking

The

main

south

entrance

to

the habitation site (AF 29), looking north from outside the building

OF

ANTHROPOGENIC

FEATURES

371

APPENDIX

372 Date: FN-LM IIIB. Pottery: see Chapter 17; see also Appendix see

Comments:

AF 30

G

J, J-2 and J-3 (Venetian).

17.

Chapter

Terrace walls

Grid coordinates: E5222 N4663. H.:

50 masl.

Visibility: good. ca. 30 m

Dim.: Date:

Minoan

and Ottoman

area AF

terraced

large

seen;

pottery

none

collected.

downhill from AF 22 andAF 39 are probably an extension

Comments: Walls the

ca. 700 m2.

east-west;

unknown.

Pottery:

of

x 26 m

north-south

22. Two

are visible.

of wall

sections

section

One

is

built inside a steep ravine, and it prevents the sides from eroding (compare AF 31, which is similar), and the other wall terraces additional land south of it.The two walls are well preserved. They are made of dolomite fieldstones using drywall and

masonry,

AF 31

as much

stand

they

as 1 m

or more

in

height.

Terrace walls (at center of area).

Grid coordinates: E5463 N5073 H.:

(at center

25 masl

of area).

Visibility: moderate. ca. 105 m

Dim.: Date:

northwest-southeast

of

(length

system).

unknown. Minoan

Pottery: Comments:

and Ottoman from

Fishermen

seen;

pottery Kavousi

none

collected.

use

routinely

the

at this

ravine

location

towalk down to the sea (for sport fishing). Rubble walls here are built of drywall to create

masonry

a series

of

terraces.

are made

They

with

carbonate

blocks

and

stand over 1m high. Some sections of wall are at the edge of the ravine, designed to

its walls

prevent

sion

of

terrace

the

from

eroding

(compare as AF recorded

system

AF

30). The

be an exten

may

terracing

36.

Oval enclosure

AF 32

Grid coordinates: E5397 N4491. H.:

106 masl.

Visibility: good. ca. 42 m

Dim.: the

Date: tury

north-south

x 69 m

east-west;

2,300

m2

of

interior

in

space

enclosure. Minoan

activity

leaving

a sherd

oval

scatter;

structure

isOttoman,

cen

18th

A.D.

J, J-8 (X 1345, from within

Pottery: Appendix of the oval

enclosure);

additional

sherds

form

a

the construction of the wall

pottery

scatter

that

extends

across

the structure and uphill and slightly downhill from it; to 32.1 AF AF 32.53 (FN to LM III). Appendix H, Comments: An oval enclosure with the east side open is located downhill the hillside both within

from it, but

the

settlement

a sherd

found

(AF in the

29). This stones

entire used

area has

as the

lowest

a scatter course

of Minoan and

foundation

sherds

on

for the

wall (Appendix J, J-8) provides an 18th-century date for this architectural feature, a is almost The enclosure above ground. pen certainly completely mortar stones is of without It built any goats. sheep (drywall masonry). over the area scatter that form the pottery The sherds this late feature including a are not related to it. The area is of for the Minoan the part garden apparently

which for

is almost and

habitation site (see Chap. 21).

OF

REGISTER

AF 33

ANTHROPOGENIC

FEATURES

373

Field house

Grid coordinates: E5471.8 N4514.5. H.:

94.8

masl.

Visibility: good. Dim.: The eastwall is 5.5 m long, thewest wall is 4.8 m long, the north wall is 2.9 m

long, and the Ottoman. Date:

Pottery: Appendix seen

pottery

(LM I to LM

and J-33.2

J, J-33.1

on

(all measurements

long

the

exterior).

IIIA/B); Ottoman

collected).

(not

small

A

Comments:

is 3.1 m

south wall

one-room

is constructed

building

carbonate

of uncoursed

fieldstones without mortar (drywall masonry). The upper parts of the walls have collapsed into the interior, covering the floor. The building is oriented north-south with a doorway 0.38 m wide in the east wall. A niche is built into the north end of the west wall, in the interior of the building. It is 0.37 m wide by 0.38 m high m by 0.3 the wall

The construction technique deep. stones and smaller up. The higher

uses

stones

larger

northeast

corner,

in the

lower

where

part

the walls

of are

best preserved, is 1.4 m high. AF34

Theriospelio cave Fig. 1.11 Grid coordinates: E5180.7 N5146.9

H.:

44 masl

(at entrance).

(at entrance).

Visibility: good. L.:

ca. 60 m.

III or later.

Date: FN-EM see

Pottery: Comments:

AF 35

Chapter see Chapter

18. 18.

Terrace walls (at center of area).

Grid coordinates: E5437 N4966 H.:

80 masl

(at center

Visibility: moderate ca. 50 m

Dim.: Date:

none

AF 36

x 8m

east-west

(estimated

ca. 400 m2).

collected. A

of carbonate

curately

to good.

north-south

unknown.

Pottery: Comments: walls

of area).

because

area north

terraced fieldstones.

of

the poor

of

extent

The

the modern of

this

road

system

has

cannot

poorly

preserved

be measured

ac

preservation.

Terrace walls

Grid coordinates: E5497 N5085. H.:

40 masl.

moderate. Visibility: ca. 31 m north-south Dim.: Date:

x 17 m

east-west;

ca. 400 m2.

unknown.

Pottery:

none

collected.

Comments: This area is probably related toAF 31 andAF 35.Walls are built of carbonate fieldstones. Two walls can be seen.AF 36a is 31 m in length, andAF 36b is approximately the same length. AF 37

Terrace walls

Grid coordinates: E5204 N4880 H.:

75 masl

(at center

Visibility: good.

of area).

(at center of area).

APPENDIX

374 12 m

ca.

Dim.: Date:

x 84 m

north-south

ca. 800 m2.

east-west;

unknown. none

Pottery:

collected.

Comments: Terrace walls uphill from the metallurgy carbonate

and

blocks,

phyllite

of phyllitic soil.Walls AF 38

G

a check-dam

creating

system.

location are built of The

terraces

are all

are not well preserved.

Terrace wall

Grid coordinates: E5221 N4599 H.:

(at center

75 masl

(at center of area).

of area).

Visibility: good. Dim.:

ca. 4 m

Date:

unknown. none

Pottery:

x 5m

north-south

16 m2.

E-W;

collected.

Comments: A terrace wall south of AF 30, built of carbonate fieldstones, as probably part of the same system recorded AF 30 andAF 22.

is

Fig. G19

AF39 Well

Grid coordinates: E5313.8 N4676.5. H.:

86 masl.

Visibility: good. Diam.

ca. 2 m.

of exterior:

Date:

or later.

Ottoman

Pottery:

none

collected.

Comments: A circular well consists of a shaft lined with stones. It is dug into the

soil

covering

the bedrock

on

this

runs downhill from near AF 9. The well is partly filled with stones. AF 40

Road and house walls

Fig.

Grid coordinates: E5336 N4821 H.:

115 masl

(at intersection

hillside,

on

a terrace

interior is ca. 0.80-0.90

above

the

ravine

that

m in diameter. The

G20 (at intersection on west slope of Chomatas).

on west

slope

of Chomatas).

Visibility: good. Dim.:

average

width

ca. 5 m.

Figure G.19. Opening

(AF39)

of awell

REGISTER

OF

ANTHROPOGENIC

FEATURES

375

Figure G.20. The modern road (AF 40) showing the section con structed inMay 1996, as seen from the habitation site (AF 29)

Date:

is modern;

road

none

Pottery:

walls

are Minoan.

collected.

Comments: A graded but unpaved road is one of the most easily visible landscape features of the Chrysokamino territory.The portion of the road from the intersection at E5336 N4821 to the habitation site (AF 29) was constructed inMay of 1996. The other parts of the road are older (graded in the mid-20th or earlier,

century

traditional following new section of the scarp

in the

visible

AF 41 Metal

access of

the

routes

to the area). House

walls

are

road.

fence

Grid coordinates: E5497 N4744 (at the highest point on Chomatas). H.: 135 masl (at the highest point on Chomatas). Visibility: good. Dim.: Date:

not

applicable. modern.

Pottery: Comments:

none

collected. A

metal

fence

with

two

construction

periods,

both

after

the

middle of the 20th century, is important for the interpretation of earlier stone walls that follow the same general line (AF 3,AF 26, andAF 21/27). The fence is also a useful orientation feature for those using the maps prepared by this project because,

like

the modern

road,

it is

easily

visible.

The modern fence demarcates the private land (which is inland) from public land (the coastal strip, owned by the village of Kavousi). It is the latest in a series of

boundary

markers

at

approximately

the

same

place,

but

the earlier

walls

are not

continuous. The latermetal fence clearly defines the border between the territory used by individuals and the public land.

APPENDIX

H

Minoan

The

Pottery

the

from

Survey

byPhilip P. Betancourt

The Minoan

pottery

presented

here

is used to illustrate both

the date of

several of the anthropogenic features recorded in Appendix G and the character of the preserved ceramics associated with these features (Figs. H.l to H.4). It is all collected from the surface. Many of the sherds are very on is often based worn, and their identification shape and fabric rather than

decoration.1

AF 17 Early AF

Minoan

17.1 Vessel, Max.

III-Middle

(X 1669; from AF

Minoan

17d surface)

Fig.

IA H.l

sherd.

body

4.2.

dim.

Mirabello Fabric (yellowish red, 5YR 5/6). Comments: EM III-MM IA; the fabric of X 1669 is typical of EM III, but it continues

Late AF

into

the next

Minoan 17.2

I

(X 1299; from AF 17a surface)

Cooking Diam.

period.

H.l

Fig.

dish, rim sherd.

of rim not measurable;

max.

dim.

5.3.

A phyllite fabric (red, 2.5YR 5/8). Thickened Comments:

LM

I; the piece

is an

ordinary

rim.

cooking

dish,

a broad,

open

shape

used in cooking from the EM until the LM IIIC periods (for the shape, see Betan court 1980). The fabric used here is the phyllite fabric that ismost common in LM I (Haggis andMook 1993; for the petrography of this fabric, see Betancourt andMyer

1. For

fabrics

in the region,

see

Haggis andMook 1993; Betancourt andMyer 1995;Myer, Mclntosh, and Betancourt

1995.

Late AF

1995).

Minoan 17.3

I-III

(X 1668; from AF 17d surface)

Tripod cooking pot (?), body sherd.

Fig.

H.l

H

APPENDIX

378 Max.

dim.

3.9.

A phyllite fabric with white stone inclusions (red, 2.5YR 5/6). Comments: In this part of Crete, the phyllite fabric with white sions is usually LM III.

stone inclu

AF21 Late

I

Minoan

AF 21.1

(X 1739; from AF 21 surface) Fig. H.l

Pithos, body sherd. Max.

dim.

9.0.

A coarse fabric (dark yellowish brown, 10YR 4/4, with a light reddish brown surface, 5YR 6/4). on

Ropework

Late

exterior.

Minoan

IIIA-B

AF 21.2

(X 1725; from AF 21 surface) Fig. H.l Tripod cooking pot, leg sherd. dim.

Max.

5.3.

A phyllite fabric containing other stones (red, 10R 4/8). Comments: For the use of the cooking pot, see Betancourt fabric with white stone inclusions is usually LM III.

1980. The phyllite

AF29 Late

IIIA-B

Minoan

AF 29.1 Kylix, Diam.

(X 326; from the surface at E5397.284 base

Fig. H.l

N4588.443)

sherd. 6.

of base

Fine fabric (light reddish brown, 2.5YR 6/4). Traces of dark slip on exterior. Comments:

AF

surface

32 and AF

worn.

For

shape,

see

Popham

1969.

8

pottery from AF 8 comes and a rocky ridge just south of minus post quern for the pens sherd scatter extends over the

The

the

from the area between

the habitation

site

a

large oval animal pen (AF 32). The ter is in the 18th century. The construction a few meters down the other and ridge

side of the hill, but it does not continue all the way down the hillside to scatter contains a heterogeneous mixture of small, Lakkos Ambeliou.The worn sherds. Pieces do not most cases join with other fragments, and in the corners

are rounded, and the paint is missing. The character of this is from ceramic pieces found in and near houses, and it different pottery is interpreted in this volume as a deposit representing casual sherds tilled

THE

MINOAN

POTTERY

379

i i

i

i i

/

t

AF 17.3 (X 1668)

AF 17.2 (X 1299)

AF 17.1 (X 1669)

i

i

i

'? 11 AF 32.2 (X 1793)

AF32.1(X1785)

er*?j

i

AF 29.1 (X 326)

AF21.2(X1725)

AF21.1(X1739)

i

'

AF 32.3 (X 1770)

AF 32.4 (X 1672)

I I

o

I

AF 32.5 (X 1778)

? ! I

AF 32.6 (X 1766)

I I

AF 32.7 (X 1765)

AF 32.8 (X 1763)

D-'? t AF 32.9 (X 1764) H.l. Figure the surface,

Minoan

from

AF

Scale 1:3

pottery 17 to AF 32.

AF 32.10 (X 1798)

AF 32.11 (X 1803)

AF 32.12 (X 1800)

H

APPENDIX

38o

into the ground along with manure. The interpretation is supported by near in the results of chemical analysis related deposits Lakkos Ambeliou2 and on Pseira Island.3

Final

Neolithic

AF 32.1

(X 1785; from AF 32 surface) Fig. H.l vessel,

Open Max.

dim.

sherd.

body 6.2.

A dark-surfaced fabric (mostly black) with phyllite inclusions; burnished. vessel

Open

with

are an unusual

rior of the vessel comments

Early

in

interior.

element.

For

the FN

in this part

in the inte

of Crete,

see

the

5.

Chapter

Minoan

III-Middle

Minoan

AF 32.2

in the

dots

impressed

Probably from the Final Neolithic. The punctations

Comments:

II

(X 1793; from AF 32 surface) Fig. H.l dish, body sherd.

Cooking Max.

dim.

3.0.

Mirabello

Fabric (dark reddish brown, 5YR 5/4). Comments: This is a typical cooking dish fragment of the type used in

EM

III and

later

times.

For

the

shape,

see Betancourt

1980.

(X 1770; from AF 32 surface) Fig. H.l Cooking dish, rim sherd.

AF 32.3

Max.

3.0.

dim.

Fabric (dark reddish brown, 5YR 3/4). Comments: Similar toAF 32.2.

Mirabello

(X 1672, from AF 32 surface) Fig. H.l

AF 32.4

Small vessel, handle sherd. Max.

3.0.

dim.

Mirabello

Fabric (brown, 7.5YR 5/4). It is

Comments: have

similar

Early

possible

Vessel, Max.

this

is a handle

from

a cup, but

some

small

jugs

appendages.

III-Late

Minoan

AF 32.5

that

Minoan

I

(X 1778; from AF 32 surface) Fig. H.l body dim.

sherd. 2.0.

Fabric (brown, 10 YR 5/3). Comments: Mirabello Fabric is typical of theMiddle Bronze Age; its use comes to an end at the beginning of LM I (Haggis andMook 1993, type 1). Mirabello

AF 32.6 Vessel, Max.

(X 1766; from AF 32 surface) Fig. H.l sherd. body dim. 3.5.

Mirabello interior).

Fabric (brown, 7.5YR 5/4 exterior with very dark gray, 7.5YR N3

2. Morris

pp. 187,197. and Hope Simpson and Betancourt Evershed,

3. Betancourt 1992; 1999.

Bull,

1994,

THE

POTTERY

MINOAN

381

Middle

Minoan-Late

AF 32.7

(X 1765; from AF 32 surface) Fig. H.l

Vessel,

Minoan

sherd.

body

A coarse fabric (between brown and dark brown, 10YR 4/3). (X 1763; from AF 32 surface) Fig. H.l Closed vessel, body sherd.

AF 32.8

Max.

dim.

2.7.

A coarse fabric (yellowish red, 5YR 5/8 exterior with 4/3 interior). AF 32.9 Closed Max.

reddish brown, 5YR

(X 1764; from AF 32 surface) Fig. H.l rim

vessel, dim.

sherd.

2.6.

Rolled rim. A coarse fabric (red, 2.5 YR 5/6). (X 1798; from AF 32 surface)

AF 32.10 Vessel, Max.

H.l

sherd.

body dim.

Fig.

2.5.

A fine fabric (reddish yellow, 5YR 6/6). AF 32.11 Vessel, Max.

(X 1803; from AF 32 surface) Fig. H.l sherd.

body dim.

2.2.

A phyllite fabric with a few white

(X 1800; from AF 32 surface) Fig. H.l

AF 32.12 Vessel, Max.

inclusions (reddish brown, 5YR 5/4).

sherd. body dim. 1.4.

A coarse fabric (yellowish red, 5YR 5/6). (X 1794; from AF 32 surface)

AF 32.13 Vessel, Max.

Fig.

H.2

sherd.

body

2.7.

dim.

A coarse fabric (red, 2.5YR 5/8). AF 32.14 Vessel, Max.

(X 1799; from AF 32 surface) Fig. H.2 sherd. body dim. 1.5.

A coarse fabric (reddish brown, 5YR 5/4). AF 32.15

(X 1760; from AF 32 surface)

Tripod cooking pot, body sherd with Max.

dim.

Fig.

H.2

leg.

4.5.

A phyllite fabric with many white inclusions (between red, 2.5YR 5/6 and reddish brown, 5YR 4/4). Leg with thick oval section with three vertical slashes on outside

AF 32.16

of

leg.

(X 1786; from AF 32 surface)

Bridge-spouted

jar,

rim

Fig.

H.2

sherd.

A phyllite fabric with many white Thickened rim.

inclusions

(yellowish

red, 5YR 4/6).

APPENDIX

382 Middle

Minoan-Late

I

Minoan

(X 1780; from AF 32 surface) Fig. H.2

AF 32.17

handle

Vessel, Max.

H

dim.

sherd.

3.2.

A coarse fabric (yellowish red, 5YR 5/6). (X 1759; from AF 32 surface) Fig. H.2

AF 32.18 Jar, rim Diam.

sherd.

Mirabello

Middle AF 32.19 Vessel, Max.

ca. 24-26.

of rim

Fabric (light red, 2.5YR 6/6). Outturned

Minoan-Late

rim.

III

Minoan

(X 1768; from AF 32 surface) Fig. H.2 sherd.

body dim.

3.6.

A coarse fabric (red, 2.5YR 5/6 exterior with yellowish brown, 10 YR 3/4). AF 32.20 Vessel, Max.

(X 1769; from AF 32 surface) Fig. H.2 sherd. body dim. 3.7.

A coarse fabric (yellowish red, 5YR 5/6 exterior with 3/2 core). AF 32.21 Vessel, Max.

(X 1772; from AF 32 surface) 2.7.

A phyllite fabric with a few white AF 32.22 Vessel, Max.

H.2

sherd.

body dim.

Fig.

dark brown, 7.5YR

inclusions (red, 2.5YR 4/6).

(X 1797; from AF 32 surface)

Fig.

H.2

sherd. body dim. 3.0.

A coarse fabric (yellowish red, 5YR 5/6). AF 32.23 Vessel, Max.

(X 1773; from AF 32 surface) Fig. H.2 sherd. body dim. 3.0.

A coarse fabric (red, 2.5 YR 5/6). AF 32.24 Vessel, Max.

(X 1771; from AF 32 surface) Fig. H.2 base dim.

sherd. 5.4.

A phyllite fabric with a few white AF 32.25

(X 1782; from AF 32 surface)

Closed

vessel,

base

Diam.

of base

32.

H.3

inclusions (reddish brown, 5YR 5/4).

(X 1783; from AF 32 surface) Fig. H.3

Closed vessel, body sherd. Max.

Fig.

sherd.

A phyllite fabric with a few white AF 32.26

inclusions (reddish brown, 2.5YR 4/4).

dim.

6.8.

THE

MINOAN

POTTERY

383

h

1

Q-?

I I

AF 32.14 (X 1799)

AF32.13(X1794)

AF 32.15 (X 1760)

AF 32.16 (X 1786)

\\

a-i

F AF 32.17 (X 1780)

AF 32.18 (X 1759)

AF 32.19 (X 1768)

i 1 1

i AF 32.20 (X 1769)

AF 32.21 (X 1772)

*

AF 32.22 (X 1797)

V AF 32.24 (X 1771) H.2. Figure the surface,

Minoan AF

32.

pottery Scale 1:3

from

AF 32.23 (X 1773)

APPENDIX

384

H

AF32.25(X1782) AF32.26(X1783)

7 ( AF 32.28 (X 1671)

AF 32.27 (X 1795)

\

Y

\

'

AF 32.29 (X 1805)

\ i

" ? AF 32.32 (X1787)

AF 32.31 (X 1762)

AF 32.30 (X 1804)

/

11 AF 32.34 (X 1776)

AF 32.33 (X 1761)

I

i !

AF 32.36 (X 1779)

?-!

II

AF 32.37 (X 1788)

o-#;

AF32.35(X1789)

i i

?-*

AF 32.38 (X 1777)

AF 32.39 (X 1775)

Figure H.3. Minoan the

surface,

AF

32.

pottery from Scale 1:3

THE

POTTERY

MINOAN

Convex

385

profile.

A phyllite fabric with many white

Middle

III-Late

Minoan

AF 32.27 Conical Diam.

inclusions (yellowish red, 5YR 5/6).

(X 1795; from AF 32 surface) rim

cup,

I

Minoan Fig.

H.3

sherd.

of rim 7.

A fine phyllite fabric (brown, 7.5YR 5/2).

Late

I

Minoan

AF 32.28

Vessel,

(X 1671; from south of AF 32)

Max.

H.3

sherd.

body dim.

Fig.

4.1.

A phyllite fabric (yellowish red, 5YR 5/6). AF 32.29 Cup Diam.

(X 1805; from AF 32 surface) or tumbler, of rim

rim

Fig.

H.3

sherd.

10.

A fine fabric (yellowish red, 5YR 5/6). AF 32.30 Vessel, Max.

(X 1804; from AF 32 surface) rim dim.

sherd. 2.1.

A phyllite fabric with a few white

inclusions (yellowish red, 5YR 5/6).

AF 32.31

(X 1762; from AF 32 surface) Fig. H.3 Tripod cooking pot, leg sherd.

Max.

dim.

3.4.

A phyllite fabric (red, 2.5YR 4/6). Leg with thick oval section.

Late

I?III

Minoan

AF 32.32 Closed Max.

(X 1787; from AF 32 surface) Fig. H.3 vessel, dim.

body

sherd.

3.6.

A phyllite fabric (yellowish red, 5YR 5/6). AF 32.33

(X 1761; from AF 32 surface)

Fig.

H.3

Closed vessel, body sherd with handle. Max.

dim.

2.7.

A phyllite fabric with a few white AF 32.34 Closed Max.

inclusions (red, 2.5YR 5/6).

(X 1776; from AF 32 surface) Fig. H.3 vessel, dim.

base

sherd.

6.3.

A phyllite fabric with a few white

inclusions (red, 2.5YR 5/6).

APPENDIX

386

H

i

i

i

AF 32.42 (X 1796)

AF 32.41 (X1802)

AF 32.40 (X1801)

-

\ /

\

?i

AF 32.43 (X 1676)

/

T*

AF 32.46 (X 1767)

AF 32.45 (X 1674)

AF 32.44 (X 1677)

*

AF 32.47 (X 1784)

i

n i

\1 I

i # 1^1

P

AF 32.50 (X 1790)

AF 32.49 (X 1774)

AF 32.48 (X 1791)

V

iII AF 32.53 (X 1792)

AF 32.52 (X 1678)

AF 32.51 (X 1673)

11

Ii

i

AF 33.1 (X 1724)

l

AF 33.2 (X 1723)

THE

Figure H.4 (opposite).Minoan pottery from the surface, AF 32, AF 33. Scale 1:3

MINOAN

AF 32.35 closed Max.

POTTERY

387

(X 1789; from AF 32 surface) Fig. H.3 vessel> handle sherd> dim.

3.7.

A phyllite fabric with many white

inclusions (brown to dark brown, 10YR

4/3). AF 32.36 Vessel, Max.

(X 1779; from AF 32 surface) body dim.

Vessel, Max.

inclusions (red, 2.5YR 4/6).

(X 1788; from AF 32 surface)

Fig.

H.3

Fig.

H.3

sherd.

body dim.

H.3

2.8.

A phyllite fabric with a few white AF 32.37

Fig.

sherd.

2.2.

A phyllite fabric (red, 2.5YR 5/6). AF 32.38 Vessel, Max.

(X 1777; from AF 32 surface) sherd.

body dim.

1.4.

A phyllite fabric (yellowish red, 5YR 5/6). AF 32.39 Vessel, Max.

(X 1775; from AF 32 surface)

H.3

Fig.

H.4

sherd.

body dim.

Fig.

2.2.

A phyllite fabric (red, 2.5YR 4/6). AF 32.40 Vessel, Max.

(X 1801; from AF 32 surface) sherd.

body dim. 3.1.

A phyllite fabric with a few white AF 32.41 Vessel, Max.

inclusions (yellowish red, 5YR 5/8).

(X 1802; from AF 32 surface)

H.4

Fig.

H.4

sherd.

body dim.

Fig.

2.6.

A phyllite fabric (red, 2.5YR 5/6). AF 32.42 Vessel, Max.

(X 1796; from AF 32 surface) sherd.

body dim.

2.9.

A coarse fabric (yellowish red, 5YR 5/6). AF 32.43

(X 1676; from AF 32 surface)

Fig.

H.4

Cooking dish, rim sherd. A phyllite fabric with white stone inclusions (yellowish red, 5YR 4/6). Comments: This is often a Late Minoan III fabric in this part of Crete. AF 32.44

(X 1677; from AF 32 surface)

Fig.

H.4

Tripod cooking pot, leg sherd. Max.

dim.

5.4.

A phyllite fabric with white Comments: See AF 32.43.

stone inclusions (red, 10R 4/8).

APPENDIX

388 (X 1674; from AF 32 surface)

AF 32.45

Fig.

H

H.4

Tripod cooking pot, leg sherd. Max.

dim.

4.7.

A phyllite fabric with white stone inclusions (yellowish red, 5YR 4-5/6). Leg with thick oval section. Comments: See AF 32.43. (X 1767; from AF 32 surface)

AF 32.46

Fig.

H.4

Tripod cooking pot, leg sherd. Max.

dim.

3.4.

A phyllite fabric with a few white Leg with thick oval section. Comments: See AF 32.43.

inclusions (red, 2.5YR 5/6).

(X 1784; from AF 32 surface) Fig. H.4

AF 32.47

Tripod cooking pot, leg sherd. Max.

dim.

3.8.

A phyllite fabric with a few white Leg with thick oval section. Comments: See AF 32.43.

inclusions (yellowish red, 5YR 4/6).

(X 1791; from AF 32 surface) Fig. H.4

AF 32.48

Tripod cooking pot, leg sherd. Max.

dim.

6.0.

A phyllite fabric (between red, 2.5YR 5/6 and dark reddish brown, 5YR 3/4). Leg with thick oval section. (X 1774; from AF 32 surface)

AF 32.49

Fig.

H.4

Tripod cooking pot, leg sherd. Max.

dim

5.2.

A phyllite fabric with a few white Leg with thick oval section. Comments: See AF 32.43.

(X 1790; from AF 32 surface) Fig. H.4

AF 32.50

sherd.

Jar, handle Max.

dim.

6.5.

A phyllite fabric with many white Comments: See AF 32.43.

Jar, rim

sherd. of rim

ca. 40.

A phyllite fabric with white Comments: See AF 32.43. AF 32.52

inclusions (red, 2.5YR 4/6).

(X 1673; from AF 32 surface) Fig. H.4

AF 32.51

Diam.

inclusions (yellowish red, 5YR 5/6).

stone inclusions (reddish yellow, 5YR 6/8).

(X 1678, from AF 32 surface)

Fig.

H.4

Jar, body sherd with part of handle. Max.

dim.

5.2.

A phyllite fabric with white Comments: See AF 32.43.

stone inclusions (red, 2.5YR 4-5/8).

THE

MINOAN

Late

POTTERY

III

Minoan

AF 32.53 Closed Max.

(X 1792; from AF 32 surface) vessel, dim.

389

body

Fig.

H.4

sherd.

4.3.

A fine fabric (reddish brown, 5YR 5/6). Comments: Dated on the basis of the fabric.

AF33 Late

I

Minoan

AF 33.1

(X 1724; from AF 33 surface) Fig. H.4

Vessel, body sherd. Max.

dim.

2.6.

A phyllite fabric (yellowish red, 5YR 5/6).

Late

Minoan

AF 33.2 Vessel, Max.

IIIA-B

(X 1723; from AF 33 surface) Fig. H.4 sherd. body dim. 5.2.

A phyllite fabric containing other stones (red, 2.5YR 6/6). Comments: Possibly a cooking pot? Dated by the fabric.

I

APPENDIX

for

Evidence

Beekeeping

by Susan C Ferrence and Elizabeth B. Shank

Evidence

for possible beekeeping within the geographic territory of the was discovered summer of 1999. farmstead the during Chrysokamino A pottery sherd measuring 5.0 cm long by 4.3 cm wide by 1.4 cm thick was found in a survey of the south of the habitation site landscape came from a (Figs. 1.1,1.2). The sherd hilltop that is lower than the eleva tion of the habitation site. Both hills overlook the Gulf ofMirabello. The sherd was

on the surface among thyme and small pine trees. No sherds from pottery vessels were observed nearby. of its original appear The sherd is too small for any reconstruction a to it but class of vessels ance, may belong resembling large basins with on the interior. The scoring possibility that such vessels might have served found

as beehives

in Crete has been proposed several times,1 although it has also as suggested that the objects could have been used graters.2 The in as beehives has been discussed in detail of the class terpretation by Melas, who cites many parallels for incised beehives from Classical Greece and been

Scoring in the interior would have enabled to adhere to the internal surface of the beehive.

elsewhere.3

the honeycombs

suggests that most beehives were located to the south of, but near to, houses on hillsides covered with "aromatic shrubs and stunted pines."4 In is significant because this context, the sherd discovered at Chrysokamino of its distant location from the habitation site, indicating that beekeeping Melas

was not always limited to the immediate a or farm. surroundings of house The area inwhich the sherd was found would have been ideally suited for and casual pasturage. beekeeping sherd from Chrysokamino is important because most similar one and this is of the better fragments have been found in settlements, use as the of such incised vessels beehives.5 Scored ves arguments against sel fragments were found within the rooms of buildings at several places, The

1. See, pp. 98,168,

II, among others, Kommos nos. 467,1483; Kommos

III, p. 25, no. 439; Floyd 1998, p. 75, no. 263,

from

Pseira.

2. Floyd 1998, p. 180;Davaras 2001, pp. 81-82. 3. Melas 4. Melas

1999. p. 487. 2001, p. 82. a list, see Melas 1999. 1999,

5. Davaras 6. For 7. Melas

1999,

p. 490.

including Kommos, Kato Syme Viannou, Pseira, and sites on Karpathos.6 These occurrences have been interpreted to represent the storage of beehives when they were not in use. Scored vessel fragments have also been found within

the habitation

site at Chrysokamino. that Minoan honey production

was Melas suggests largely oriented toward domestic use and local trade.7 Isolated farms needed bees for crop

APPENDIX

392

habitation

I

site

Agnomandra?.

Figure 1.1. Location fragment settlement

interior

I

I

was Chrysokamino bees would lage. Raising

and its by-products a self-sufficient

away

of the beehive from

any

I

I

X1807 pollination. The honey within the household.

found

Figure 1.2.The beehive fragment would

have been

consumed

farmstead isolated from any vil have been necessary for ensuring the pollination of the crops and running a successful private farm. The farmer could also take advantage of the resulting honey and wax production, making life for his family that much sweeter.

APPENDIX

J

to

Byzantine

The

from

Pottery

Ottoman

the

Survey

byNatalia Poulou-Papadimitriou

a Although only few pieces of pottery from the Byzantine and later periods are present at are interesting for the information Chrysokamino, they they about the region during these periods (Figs. J.l, J.2). No signs of a provide of permanent settlement are present, and no spot has a large concentration sherds. The pottery fragments occur in isolation, and they probably attest area. Several of to farming, herding, hunting, or casual travel through the the sherds are from closed vessels that would have been useful in carry ing water open

and other

a few pieces

liquids, while

are from dishes

and other

shapes.

CATALOGUE The

catalogue follows the format used for the Minoan pottery, with the sherds arranged by periods. Colors are described using the Munsell system.1

Byzantine J-l

Period

(X 1485, from the surface of unit XB-42-1) Large

amphora,

many

body

Fig. J.l

sherds.

Diam. of body (restored) ca. 30. A fine fabric (reddish yellow, 5YR 6-7/6). Comments: of

the Minoan

12th-13th habitation

century,

probably

13th

century.

Found

on

the

surface

site.

Venetian Period J-2

(X 1657, from AF 29 surface) Fig. J. 1 Open bowl, body sherd. Max.

dim.

5.9.

A fine fabric (reddish yellow, 7.5YR 7/6). Creamy off-white glaze on interior and upper part of exterior; greenish yellow drips evenly spaced on exterior. 1. ration

Kollmorgen 1992.

Instruments

Corpo

Comments: production.

Venetian,

15th-16th

century Glazed

Painted Ware;

a Cretan

APPENDIX

394

J

(X 1300, from AF 17 surface) Fig. J.l

J-3

rim sherd. Amphora, Diam. of rim 14.

A fine fabric (light red, 2.5YR 6/8) with Thickened,

flat

Comments:

Venetian,

Venetian

rim; groove

under

15th-16th

or

Period

inclusions; without

century.

Ottoman

Period

(X 1167, from the excavation of terraceAF 22b,

J-4

unit

or large jar, body sherd.

dim.

4.

A fine fabric with mica

inclusions (light red, 2.5YR 6/8).

or Ottoman.

Venetian

Comments:

(X 1235, from the excavation of terraceAF 22b,

J-5

unit

J.l

Fig.

Xgamma-1-9)

Closed Max.

1

J.

Fig.

Xgamma-1-9)

Amphora Max.

slip.

rim.

vessel, dim.

body

sherd.

3.4.

A coarse fabric (red, 2.5YR 5/6, to reddish gray, 5YR 5/2).

Ottoman

Period

(X 1656, from AF

J-6

or Ottoman.

Venetian

Comments:

10 surface) Fig. J.2

Amphora,

body sherd.

Max.

5.2.

dim.

A coarse fabric (reddish yellow, 5YR 6/6). Almost straight profile. Comments: The shape has parallels from Eleftherna from as early as the 14th century, but the hard-fired paste suggests this piece is from the 17th to 18th century.

J-7

(X 1168, from the excavation of terraceAF 22b, Fig. J.2 unit Plate, Diam.

Xgamma-1-10)

base

sherd. 8.

of base

A fine fabric (light reddish brown, 5YR 6/4), slipped, with a yellowish glaze over

the "random" base. Glazed with slip. Ring Comments: Cretan Ottoman. production.

drip pattern. A trace of the

tripod

trivet

used

in

stacking is visible in the glaze in the interior of the vessel. J-8

(X 1345, from AF 32 surface) Fig. J.2 Flask, body sherd. Max.

dim.

11.3.

A fine fabric with

inclusions (light red, 2.5YR 6/6). Comments: Found within thewall of the oval structure atAF 32.18 th century.

Undecorated,

J-9

without

slip;

(X 1307, from AF Small Diam.

lekane,

rim

of rim 28.

from

a flask with

17a surface) sherd.

two

handles

(a water

container).

Fig. J.2

THE

BYZANTINE

TO

OTTOMAN

POTTERY

395

I J-l (X 1485)

\ /

I J-3 (X 1300)

J-2 (X 1657)

i i

/ J-4 (X 1167) to Ottoman

Figure J.l. Byzantine pottery from the surface of the territory (J-l to J-5). Scale 1:3

J-5 (X 1235)

APPENDIX

396

J

?f?v?">

?V,?.*"

?.r.v**-*** ?vil w?<

J-7 (X 1168)

J-6 (X 1656)

7 J-9 (X 1307)

J-8 (X 1345)

/ J-1KX946)

J-10 (X1594)

i i

> i

/

J-12 (X 1655)

/

J-13 (X 1654)

I

I

J-15 (X 1669)

J-14 (X 1670) Figure J.2. Ottoman the

surface

of the

pottery from

territory

(J-6 to J-15). Scale 1:3

THE

BYZANTINE

TO

OTTOMAN

POTTERY

397

A fine fabric (light red, 2.5YR 6/6), with yellow glaze (5Y 8/8) on the interior and

J-10

exterior

of

rim.

the

Thickened,

outturned

rim.

Comments:

Ottoman,

19th

century.

(X 1594, from the surface of unit XB-33-0)

Fig. J.2 from of sherd the the base. Large lekane, beginning Max.

dim.

9.8.

A fine fabric (dark reddish gray, 5YR 4/2) with lowish brown, 10YR 4/6) on interior. Comments:

Glazed Ware

Ottoman.

Found

on

(greenish glaze without

the

surface

of

Large Diam.

base

lekane, of base

the Minoan

habitation

site.

slip).

(X 946, from the surface of unit XB-12-1)

J-ll

inclusions; glaze (dark yel

Fig. J.2

sherd.

ca. 10.

A fine fabric (reddish yellow, 5YR 6/6); yellow glaze (2.5Y 7/6) in interior. Comments:

on

Found

the surface

of the Minoan

habitation

production.

(X 1655, from AF

J-12

Amphora, Max. dim.

body 5.

10 surface) Fig. J.2

sherd.

A coarse fabric (yellowish red, 5YR 5/8). Straight profile. Comments:

J-13

Ottoman.

(X 1654, from AF

16 surface) Fig. J.2 Closed vessel, body sherd. Max.

dim.

4.2.

A coarse fabric (yellowish red, 5YR 5/6). Incised bands. Comments:

Ottoman.

(X 1670, from AF

J-14

17d surface) Fig. J.2

Flask, body sherd. Max.

dim.

5.3.

A fine fabric with Comments:

inclusions (yellowish red, 5YR 5/6).

Ottoman.

(X 1669, from AF 17d surface)

J-15

Flask, Max.

site. Glazed

Ware

slip; this very rough ware is found until the 19th century; local Cretan

without

Fig. J.2

sherd.

body dim. 5.1.

A fine fabric with Comments:

Ottoman.

inclusions (red, 2.5YR 5/6).

The

of Cave

Excavation

and Terrace

AF

AF

9

22b

byBrigit Crowell and Philip P. Betancourt

cave (AF 9, small manmade Figs. K.l, K.2) and the second highest ter race in a series of step-terraces (AF 22b, Figs. K.3, K.4) were excavated in order to better understand the date and character of the agricultural

A

near the habitation site. No sherds were present on the surface of plots either feature. The investigation revealed that both the cave and the ter race were used in Venetian theMinoan

or (more likely) Ottoman

times rather than in

period.

THE EXCAVATION OF CAVE AF 9 A

trench was

excavated

in 1997 in a small cave located near a large group 9, and Fig. K.4). The goal of the

of agricultural terraces (see Fig. 16.2, AF excavation was to ascertain the function cavern

and phases of use for the small the feature s shape, size, and other characteris cavern is the cave within the borders of the largest manmade

and to describe

tics. This

habitation site, so the Chrysokamino territory, and it is near theMinoan use was of in its the Minoan question possible period important to the area. of the interpretation cave faces north-northwest.

The

It is carved out of soft phyllite, which into brown soil. The entrance is carved

has almost completely decomposed into a low cliff (running roughly north-south),

and it opens just above ter terrace system. The width from the 22b, providing easy passage of the entrance is ca. 3.45 m, and the height is 1.53 m from the modern surface inside the cave to the upper center of the entrance. The sides of the cave mouth are ca. 0.95 m thick at the base, and they rise in an arch above the entrance. Inside the cave, the rock is roughly carved into a dome that race AF

is less than 2 m high. Areas of the interior appear to have been cut back to form an irregular oval area of floor, with two areas extending deeper on the north and south sides. The dimensions of the interior are 5.98 m from the to the back wall, with aminimum width of 8.21 m north-south. floor and the land just outside the entrance slope downhill from east

entrance The

to west.

A

1 x 3 m was laid out on an east-west trench measuring axis, with east into the cave and 1 m to west the outside extending

2 m extending

APPENDIX

400

K

Figure K.l. Cave AF 9, looking south-southwest (1999)

Figure

an initial debris, the trench clearing of modern in 11 passes of 10 to 15 cm levels. The first pass deep an elevation of 111.91 masl, and soil was removed to a at began depth of 1.79 m. the entrance.

was

After

excavated

soil was fairly consistent and homogeneous in all passes. Itwas a soft, fine-grained, powdery soil, pale yellow (2.5Y 7/4) when dry and light brown (2.5Y 6/4) when wet. The only indications of intrusion yellowish were found fairly close to the surface. In passes 2 and 3, at 111.80 masl, a The

soil was present. Itwas found only in the eastern part a lower level (111.52 masl), a reddish, burned area with small fragments of charcoal was present. Natural soil was found directly above and below these two deposits. Despite their small size, the deposits are useful because or animal demonstrate that human they activity may be cave. within the environment of the The easily recognized only artifacts recovered were two fragments of thin, corroded iron found near the lower lens of darker soil. lens of darker brown

of the trench. At

The

two lenses of different

soil indicate

that the cave was

carved out

in stages. In view of the consistently natural soil found in all low levels, it may be concluded that the cave is fairly recent, having nothing to do with the Minoan period. Its construction was probably associated with one or more of the periods of use of the nearby terrace AF 22b.

K.2.

The

cave's

entrance

THE

Figure

K.3.

Terrace

group

looking north-northwest

AF

OF

EXCAVATION

CAVE

AF

9

AND

AF

TERRACE

22B

4OI

22,

(1999)

Figure K.4. Terrace AF 22b, looking north-northwest (1999)

THE EXCAVATION OF AGRICULTURAL TERRACE AF22B habita AF 22b is part of a terrace system near the Minoan tion site, on the southwest slope of the hill of Chomatas (see Fig. 16.2, AF 22). No olive trees grow on any of the terraces in this system, and there is no visible indication of recent agricultural use. It is the second highest Terrace

terrace in the system, and it is adjacent to the small cave AF 9 (Figs. K.1, K.2). A trench was excavated in this terrace to investigate the terrace wall's

manmade

construction to the

technique

terrace

wall

on

and its period of use. Itwas

an east-west

axis,

4m

east-west

laid out at a right angle by

1m

north-south,

across the level area of the terrace. The

extending were missing, on

the

close

before

excavation.

trench was

excavated

surface

The

so the first few passes excavated

upper stones of the wall were only soil. No artifacts

in 10 cm passes from the surface to a level excavation exposed both sides of the rows of stones of various sizes and irregular

to the base of the wall. The

terrace wall.

It was built with

was mostly dolomite, which shapes, often having flat faces. The material on the surface occurs in just south of this location. large quantities

APPENDIX

402

The

K

Soil

In the upper levels of the trench, the soil was similar to the soil on the surface of the terrace. It was hard and dry, and it appeared brown when first excavated. Itwas a compact mixture of soil, phyllite pebbles, and larger stones. Modern snails and roots were found in the upper levels. The soil was

pale yellow (2.5Y 7/4) when dry and light yellowish brown (2.5Y 6/4) when wet. It was easily distinguished from the natural soil of the hillside, it. which below The and lay description analysis of the soil is published inAppendix L. the natural soilwas cut back to form a depression for the browner soil that the wall supported, or the slope of the hillside was very shallow here before the wall was constructed. The bottom of the fill was higher at the east, where the natural soil was found at 107.75 masl (just over 1 m below the modern surface in the northeast corner). At the west, near the Either

terrace wall,

The

the natural

of

Date

the

surface was

at 107.50 masl.

Terrace

upper courses of the preserved stones of the terrace wall were first excavation ended encountered in pass 7, at an elevation of 108.10 masl.The at 107.50 masl. No artifacts were recovered below a sherd found in associa

The

tion with were

found

Ottoman),

the upper preserved course of the wall (J-7). Only three sherds in the excavation: J-4 (Venetian or Ottoman),J-5 (Venetian or and J-7 (Ottoman). This pottery indicates that the terrace was

or Ottoman times. Ottoman construction probably constructed inVenetian rested just of at least part of the wall is attested by the sherd J-7, which behind the face of the wall below its uppermost preserved course.

Conclusions use of the terrace behind wall AF 22b. helped clarify the It established the extent and depth of fill that the wall supports, as well as the characteristics natural soil of the terraces soil and the underlying

The

excavation

hill may have been cut back to form a depression for the agricultural soil, which was carried in from another location, possibly the cave. The relative absence of intrusive material suggests that little nearby of the hill. The

or activity occurred at any depth. If the soil had been repeatedly fertilized more sherds or other types of artifacts would have it is that redug, likely sherd (App. J, J-7, been found in the soil.1 The presence of an Ottoman 1168) provides a date for at least the upper part of the terrace wall. It appears that this system of terraces, which is one of the largest in the was constructed at least Minoan partly in conjunction with Otto territory, X

terrace and the nearby cave (AF 9) can probably farming practices. The be associated with the same phase of agricultural activity in the region.

man

1. One

can contrast

the

terraces

on Pseira,

two Minoan which

had

ity within

extensive the

soil;

evidence see Betancourt

Hope Simpson 1992.

in

situation both

of

for activ and

L

APPENDIX

and

Soils Natural

from

Sediments Deposits

at

Chrysokamino byEleni Nodarou

study investigates selected samples of soils and sediments from the site of Chrysokamino.1 low-technology analytical archaeological Using in it examines particular physical properties of the sediments, methods,

This

organic material content, and magnetic cluding particle size distribution, of aim the The analysis is to investigate the properties of susceptibility. the natural deposits and provide comparative data for the archaeological at the Minoan farmhouse. The comparative study on of the soil material and the provenance insights archaeological an assessment of factors in soil formation processes. The anthropogenic thesis.2 work presented here forms part of my masters sediments

excavated

offers

MATERIAL AND METHODOLOGY used in the present study consists of 12 samples of ca. 500 g each collected in situ from an agricultural terrace (phyllitic sediment), a road exposure (terra rossa), and an alluvial deposit. Field observations were recorded for all locations.

Material

that were

Processing was conducted at the INSTAP Study Center for East Crete where the soil samples were dried, weighed, and sifted through a 2 mm sieve. All gravels were removed and collected, and the fine fraction was were labeled, sealed, and subsequently subsampled.The samples transported to the laboratory of the Department at the University of Archaeology 1.1 would

like to thank

of Sheffield.

P. Betan

to carry out this for permission to study and for the invitation publish part of it in this volume. This work has court

benefited advice owed ments

from guidance greatly from C. Frederick. Thanks

and

com for useful McCoy on an earlier draft of the paper.

to F.

Antiquities

and

the Institute

of Geology andMineral Exploration (IGME). 2. Nodarou

1998.

amortar

Analytical

carrying out any analysis, all soil lumps were gently and pestle and sifted through a 2 mm sieve.

Procedure

are

Sampling permits were kindly provided by the 24th Ephorate of Prehistoric and Classical

crushed with

Before

were analyzed for particle size distribution, samples content, and magnetic (Table L.l). susceptibility

All

Particle

Size

organic material

Analysis

textural properties of a deposit are produced by several factors, includ contributions from the parent material, alteration during transport, ing

The

L

APPENDIX

404

TABLE L.l. ANALYTICAL Mean Depth (cm)

Samplet

DATA FROM NATURAL LOI0,

Susc.

Magn.

SI

Mz

AT CHRYSOKAMINO

DEPOSITS GravelsL,

Silt% Clay %

Sand%

Textural

Class

Agricultural Terrace 1/S ample

1

7.5

3.9

18.32

2.13

4.94

32

25.0

30

13

loam

Zone

1/Sample

2

22.5

4.0

21.79

0.80

5.27

45

19.0

25

11

loam

Zone

2/S ample

3

40.0

3.0

26.44

1.86

5.23

36

24.0

25

15

loam

Zone

2/S ample

4

60.0

2.2

27.74

1.36

5.27

41

23.0

21

15

clay loam

Zone

2/S ample

5

77.5

4.4

28.70

2.20

5.50

38

17.0

27

18

clay loam

3/Sample

6

97.5

3.7

10.24

6.76

2.49

7.5

64

28

clay

Zone 3/S ample 7

120.0

2.1

7.71

6.50

2.11

5.0

73

22

silt loam

Zone

4/S ample

140.0

4.9

13.18

0.96

4.8

43

29.0

15

13

sandy clay

Road

Exposure

Zone

1

12.0

9.3

193.16

2.90

5.03

29

28.0

24

19

Zone

2a

38.5

4.5

134.72

4.06

4.66

16

38.0

23

23

sandy clay

Zone

2b

38.5

6.4

179.40

2.50

5.66

40

23.0

17

20

clay loam

7.9

248.38

8.03

3.02

34

57

clay

Zone

Zone

8

Red Alluvium

0.5

9.0

loam-clay

loam

and postdepositional alteration. environment, are and and they modified, culturally deposited as of these different be treated artifacts. The deciphering

aspects of the depositional sediments Archaeological can therefore

us to inputs helps interpret the sedimentological of human activity and to develop the period during of the history of the area under study.3

conditions a better

prevailing

understanding

about the percentage of size analysis provides information mean size in the and and sediments. The sand, silt, sorting clay particle of grain size distribution) are indicative of the nature (i.e., the homogeneity Particle

reflects forma of the depositional environment.4 Particle size distribution tion processes and transport modes, energy or velocity of the transporting Based on and energy or turbulence in the basin of deposition. medium, was at of the tell of size distribution Davidson Sitagroi, analyses particle rate to in and breaks the of buildup able draw conclusions about the telTs continuity of occupation.5 The limitations of the method

should also be considered:6

must

take into account the small size of the sample to the detail of the analysis. Archaeological levels cannot always be fully sampled, and therefore the sample may not be statistically representative of the entire level. This might

1. Conclusions

in relation

result in a slightly 2. Incomplete that must 3. Artifacts often

skewed average.

disaggregation, be considered

are factors organics, and carbonates in the results of the analysis.

and ecofact particles contribute

(ceramic sherds, bones, significant weight fractions.

etc.) can

In this research project, particle size analysis has been employed for the description and comparison of the different sediments deriving from differ ent locations and environments. The samples were homogenized through were discovered in the samples presented here. sieving. No ecofacts

3. Gladfelter 1977, p. 522; Aitken 1992. 4. Hassan

1978,

p. 206; Aitken

1992. 1973.

5. Davidson 6. Canti

1995,

p. 184.

AND

SOILS

SEDIMENTS

FROM

NATURAL

DEPOSITS

405

The hydrometer method was used for grains of silt and clay size, and the dry sieve was used for sands and gravels. The entire procedure followed of Soils (SMPSAS). The the Standard Method for Particle Size Analysis mean

particle size and the standard deviation of size distribution calculated and plotted against depth in each section.

were

Material:

Organic

Loss

on

(sorting)

Ignition

in sediments derives mainly from the decay of plant Organic material and animal residues. It comprises compounds of carbon and nitrogen, along with carbonates and charcoal. People are sedimentary agents, and human a soil. Therefore, in activity changes the physical and chemical traits of contexts, high quantities of organic matter may indicate archaeological significant contributions from plant and animal sources, horizons of human or an area of occupation, specific human activity. This organic material can be used for diet, climate, and Vegetation, landscape reconstruction. are the most common foodstuffs, housing, and roofing material organic imports.7

the limitations

Among 1. Possible

confusion

of the method, can

exist

between

it should be mentioned natural

organic

matter

that:8 accu

as part of soil-forming processes and organic matter to the site by people during occupation. brought 2. Organic matter can be lost due to postdepositional alterations. mulated

In the present research, the organic material was treated as a significant in the differentiation of the sediments under study and as a crite

factor

rion for their provenance. The method used for the analysis was loss on involved measuring the loss of weight after a 10 g split of ignition, which the silt and clay fraction of each sample was placed in amuffle furnace for 24 hours at 400?C.

Magnetic

Susceptibility

are susceptibility measurements Magnetic widely used in paleoenviron mental studies.9 Sediments often contain magnetic features resulting material, which can be used for studies of and reconstruction correlations, provenance, lithostratigraphic of environmental in the conditions. Human intervention and occupation can to the fluctuations of natural environment also be detected according from the influx of detrital

magnetic

allochthonous

values.10

of magnetic susceptibility was undertaken using the a sensor meter with Bartington Bartington dual-frequency an mm. was to set internal diameter of 36 It low Type and

Determination standard MS2 of the MS2B

7.Aitken 1992; Stein 1992, p. 195; Ball 1964. 8. Stein 1992. 9.Mullen 1977. 10. Gale

and Hoare

1991.

frequency in the SI system, with the range multiplier knob set at 1.0. The were a 2 mm sieve, and part of the fine fraction was passed through samples to fill the standard cubes for the measurement. used Four measurements were

taken for each sample, and the average of these values was calculated. was derived from the susceptibility for each sample equation of sample/0.01). Average/(Mass The magnetic

L

APPENDIX

4o6

THE AGRICULTURAL TERRACE studies on agricultural terraces have proved significant Geoarchaeological in providing information on ancient agricultural systems and practices as well as artifact distribution.11 An agricultural terrace (AF 22b, Fig. K.4) was

at Chrysokamino in 1997 under the supervision of Crowell a at is site at Kat distance from the habitation It situated short (App. K). to in the stepped type the classification of Rackham soprinos and belongs and Moody.12 The slope is fairly steep, and the purpose of the extensive was to retain the soil, preserve moisture, and make the cultiva terracing excavated

successful. Terracing was a common soils of the region under study.

tion of the area more

for the

practice

phyllite-quartzite The terrace was

to record the succession of strata investigated in order interference. Archaeological finds and to discern potential anthropogenic secure were scarce dates for the topo (only three sherds), but they provide feature. graphic terrace consisted

The

of four zones. Each one was

subdivided

into two

or three subzones The

and sampled separately according to field observations. criteria for separating the zones and subzones were soil color (using

color charts), uniformity in color and texture, structure, and bound differentiation. The samples were sifted through a 2 mm sieve, and the ary on was fine fraction ignition), magnetic analyzed for organic material (loss

Munsell

and particle size distribution. (magnetic susceptibility), properties Field observations are reflected on the stratigraphie column (Fig. L.l). Zones 1 and 2 consist of gravelly silty loam of pale yellow color. The indi and the boundary of Zone 1 are indicative of an Ap vidual characteristics horizon, a humic surface with mixed gravels and organic matter. In regard the abundance of coarse 2, it is important to investigate whether or a of anthropogenic interfer is result of colluviation processes material ence. The analytical data clearly indicate the existence of three different to Zone

types of sediment: 1. An anthropogenic 2. A natural stratum 3.Weathered

bedrock

stratum

(0-0.85 m)

(0.85-1.3

m)

(1.3-1.7

m)

of a cultivated A horizon. 1/S ample 1 bears all the characteristics for values include high These characteristics organic content and magnetic of fine and extremely poorly sorted and the predominance susceptibility sand (Mz = 2.13 phi). In comparison with lower levels, particle size distribu tion shows a greater quantity of gravels, similar quantities of sand and silt, and less clay. Lower values of organic content and magnetic susceptibility, Zone

occur sorting, deeper than the humic surface of for Zone analysis 1/Sample 2 and Zone 2/Samples 3, deposit.The a content of organic matter, amean 4, and 5 show general decrease in the coarse to medium sand, and an extremely poor sorting. At particle size of the bottom of Zone 2/Sample 4 (ca. 70 cm below the surface), magnetic at the same an increase, with a culminating point depth and properties same. increase in the percentage of gravels, while the sorting remains the aswell

as

seemingly

better

a natural

indicate that the sediment was not the product these properties its formation was due to anthropogenic situ weathering; activity.

All

of in

11. Betancourt 1992;

Bull,

and Hope Simpson and Evershed

Betancourt,

1999,2001. 12. Rackham

and Moody

1996.

SOILS

AND

Loss-on-lgnition (percent)

Stratigraphy

NATURAL

0

5

10 152025

0

30

2

4?1-lit!

O

DEPOSITS

Mean Particle Size (phi)

Magnetic Susceptibility

Sample Numbers

Zone

FROM

SEDIMENTS

6

4

8

407 Sorting (phi)

0

10

12

3

4 5 6

4?I?I?hH?I

-I-(-1-1-f

a 5 G

1

1

Gravels (percent)

Sand (percent)

0 1020 3040 50

0 10 20 30 40 50 0

Figure

L.l.

Analytical

data

for agri

0

4?,?,?I?,

\?I?I?I?I?I

The

cultural terraceAF 22b

Silt (percent) 40

60

80

0

10 20 3040

though

50

4-1-1-1-1

H-1-1-j

evidence,

archaeological

20

Clay (percent)

scarce,

also

favors

an

anthropo

genic interpretation for the sequence. The excavation of the terrace started at the depth of 108.82 masl and was terminated at a depth of 106.86 masl. At the depth of 107.75 masl (excavation unit XD-1-9), which corresponds to Zone 2/S 4 excavation of the the discovered ample stratigraphie column, a

an body fragment from unpainted closed vessel (App. J, J-4). It is from the Venetian or Ottoman period. In the next lower stratum (XD-1-10), a base from a was found plate dating to the (App. J, J-7). It is Ottoman, use cannot be excluded, 18 th century. Although the possibility ofMinoan in the absence of any relevant archaeological evidence, the Venetian to Ot toman finds afford a better date for the construction and use of the terrace.

This means construction

that all the inputs and disturbances of

the

terrace

were

Venetian-Ottoman

of the natural or

soil for the

later.

The boundary between Zone 2/Sample 5 and Zone 3/Samples 6 and 7 marks a considerable change in the nature and properties of the deposit. strata constitute the soil formed in situ from the weathered These bed rock. This

level consists

organic matter

of silty and well-sorted phyllitic material. Little are is present, and the values of magnetic susceptibility

APPENDIX

4o8 low. Little

is present,

gravel

but the amount

of clay is higher

L

than in the

strata.

other

lowest stratum (Zone 4/Sample 8) consists of partially weathered veins of calcite. It constitutes the base on which

The phyllite

and intercalating

activity took place, and it is close to bedrock. This iswhy anthropogenic there is an increase in the content of gravelly material at the expense of sand and silt. The sorting is extremely poor, and the mean particle size is that of coarse sand (Mz = 0.96 phi).

THE ROAD EXPOSURE Between

Ambeliou and the top of the hill where the Minoan is located, the landscape is dominated reddish sediment by over dolomite bedrock (Fig. L.2). The bed is relatively solid (terra rossa) Lakkos

farmhouse in some

parts

and more

to examine

weathered

in others.

the characteristics

An

of

exposure

sediment

was

and to provide from the Minoan

of this sediment

sampled sediments comparative material for the archaeological farmhouse (Figs. L.3, L.4). The upper zone (Zone 1,0-0.24 m below surface) consists of an A hori zon that includes amount

(such as roots) and a fair expected, the analyses show high and magnetic susceptibility, while the particle

remains of modern

of coarse

lithic material.

vegetation

As

values for organic content size distribution indicates extremely poorly sorted sediment with a mean = 2.90 size of sand, and very fine to fine sand (Mz particle phi). Gravels, silt

share

almost

the

same

percentage,

whereas

clays

are

present

in a smaller

percentage.

The second zone (Zone 2, ca. 0.24-0.53 m below surface) was divided into two parts that were sampled separately because field observations indi cated an area of different soil color. The first sample (Zone 2a) has a dark red color, a low content in organic material, and a low value of magnetic susceptibility. The hydrometer analysis shows that the sample consists of

Figure Lakkos

L.2.

The

sediment

Ambeliou

the habitation

(terra site

at

deposit rossa),

in the distance

with

SOILS

AND

SEDIMENTS

m

0123456789

NATURAL

Magnetic Susceptibility

Loss-on-ignition Zones

FROM

10 0

100

DEPOSITS

Mean Particle Size

(phi)

409 Sorting 2

150 200

3

(phi) 4

0 ,::;.?x

Gravels (percent)

Sand (percent)

Silt (percent) 30 20 )

Types of Soils Dark red hard I Stones

Brown hard ?^

o

Bedrock

Figure L.3. Analytical road

data for the

exposure

THEAGRICULTURA^TERREE

L.4.

Figure

Locations

of the

two

sets

of analyzed soil samples. Aerial photo graph 1988.

taken ?

at 6:52

1999 Aerial

SovInformSputnik

a.m. Images,

on

June

5,

Inc., and 1999

THE ROAD CUT

Clay (percent)

5

APPENDIX

4io

an

a mean extremely poorly sorted sediment with particle = coarse to fine sand silt (Mz 4.06 phi). The percentage of particles is lower in this part of the exposure, while sand-sized Silt-sized and clay-sized particles share similar predominant.

L

size of very gravel-sized are particles values.

part of the zone (Zone 2b) has a yellowish brown color and higher values of organic content and magnetic susceptibility. The examination of particle size distribution shows that the sample is very The

differentiated

sorted, with amean particle size in the fine sand-sized range. The percentage of gravels varies up to 40%, while sand, silt, and clay fractions vary up to 23%, 17%, and 20% respectively. poorly

Zone 2 lies the partly weathered bedrock. Sample 2a may is the products of weathering. 2b related to the Bt horizon, Sample explains its higher values.

Beneath contain which

THE RED ALLUVIUM a Vertisol, alluvial deposit at Lakkos Ambeliou (Fig. L.2) constitutes and its properties have been discussed in detail by Morris.13 The present

The

was undertaken to compare the properties of this natural deposit with study those of a thick deposit of earth excavated in one of the trenches of the Minoan farmhouse (XB-9). The red color and the clayey texture ofthat deposit contrasted markedly with the white phyllitic sediments found in the other trenches. The aim of the analysis was to elicit information about its provenance and function. The alluvium at Lakkos Ambeliou

was

the most

obvious

source to

since it displays the same macroscopic the red characteristics, to texture. and Due the textural of the the color, fine, clayey homogeneity alluvial deposit, a single sample was taken. The analysis shows high values of magnetic and organic content, very poor sorting, and a susceptibility

be tested

mean

near the boundary between very fine silt and coarse particle size clay = 8.03 The different size classes are dominated (Mz phi). by clay-sized an aeolian source. to particles, which excludes According Pye,14 soils that are not are silty, containing predominantly deposited by aeolian action

sand and up to 40% clay, which approximates the case for the water content at of the Lakkos Ambeliou. The deposit high clay particles makes them too heavy to have been carried in large quantity by the wind. 10%-25%

analysis indicates that the properties of the clayey alluvium from Lak are similar to those of the sediment of the kos Ambeliou archaeological Minoan farmhouse. The concentration of this red soil in the archaeological The

context has been

interpreted

as a

collapsed mudbrick

wall.15

COMMENTS study of the natural deposits from Chrysokamino tion of macroscopic observations and low-technology on the properties analysis offers important information

The

deposits.

By comparing

the values

using

a combina

sedimentological of various natural

of these samples with

those from the

13. Morris

2002.

14. Pye 1992. 15. Nodarou

1998.

AND

SOILS

SEDIMENTS

FROM

NATURAL

DEPOSITS

4II

sediments excavated at theMinoan farmhouse, it has been archaeological to differentiate between natural and anthropogenic deposits and possible to elucidate

the provenance of the latter. case In the of the agricultural terrace, the analysis permits the distinc tion between natural deposits and levels affected by human activity. The a presence of a pottery fragment of the Venetian/Ottoman period provides indication for the construction of the terrace. The analysis is chronological informative regarding the provenance of the archaeological material, since the gray white earth excavated from most areas at the Minoan farmhouse shares similar properties

with

the phyllitic

sediments

of the agricultural

terrace.16

The analysis of two red deposits, the red alluvium from the road expo sure and the Vertisol at Lakkos Ambeliou, had a twofold aim: to examine the natural properties of those deposits and to compare them with the red at theMinoan

earth excavated

farmhouse. The

former has been

shown to

be a natural deposit, devoid of human interference and somewhat different from the clayey deposit at Lakkos Ambeliou. Another study has indicated an that the latter contains anthropogenic component,17 but the present to its connection demonstrates the Minoan It appears farmhouse. study to have provided the raw material for the construction of amudbrick wall that collapsed and disintegrated when occupation ceased.18 The use of basic sedimentological analytical techniques has proved to be an efficient

tool in the combined

and comparative study of natural and on the provenance This archaeological deposits. approach provides insights of the archaeological sediments, the use of space in archaeological contexts, and the interference of the human agent in natural deposits.

CATALOGUE OF SAMPLES are listed in the Samples catalogue by zones, with Zone are first. Colors recorded first as dry, then moist.

Agricultural Zone

Terrace

1 Depth:

m.

0-0.33

Munsell

colors: 5Y 7/3 (pale yellow); 5Y 5/3 (olive).

Texture:

silt-loam.

slightly hard.

Consistency: Lower

boundary: Structure: medium

Roots: Horizon:

few

fine,

clear

smooth.

coarse, common

subangular

blocky,

moderate.

large.

Ap.

Comments:

35%-40% coarse fragments (<2mm).

Zone 2 16. Nodarou 17. Morris 18. Nodarou

1998. 2002. 1998.

Depth:

Munsell Texture:

0.33-0.85

m.

colors: 5Y 7/3 (pale yellow); 5Y 5/3 (olive). gravelly

silt-loam.

1 (the upper zone)

APPENDIX

412

slightly hard to hard.

Consistency: Lower

boundary:

Structure:

L

abrupt.

medium

coarse,

subangular

blocky,

moderate.

Roots: few fine, few very fine. Comments:

coarse

50%

fragments

(<2 mm).

Zone 3 m.

0.85-1.3

Depth:

Munsell

colors: 5Y 8/3 (pale yellow); 5Y 6/3 (pale olive).

Texture:

silt-loam.

slightly hard.

Consistency: Lower

platy common

Roots:

the

fine.

<1% coarse fragments (<2mm); bedding; in situweathered phyl

Comments: lite,

abrupt, wavy. coarse to very coarse.

boundary:

Structure:

structure

breaking

along

phyllite

formation

planes.

Zone 4 m.

1.3-1.7

Depth:

Munsell

colors: 5Y 7/3 (pale yellow); 5Y 5/3 (olive).

Lower

boundary: none. Structure:

Comments: is visible;

Road Zone

bedrock.

none.

Roots:

tion

approaching

Saprolite, integral phyllite weathered product. The phyllite folia

there

are zones/veins

of calcite

and

large

aggregates.

Cut 1 m.

0-0.24

Depth:

Munsell

colors: 7.5YR 4/4 (brown); 7.5YR 3/4 (dark brown).

Texture:

silt to

clay.

hard. Consistency: clear. Lower boundary: Structure: grains. Roots:

medium

moderate

subangular

blocky

to smaller

breaking

granular

common. A.

Horizon: Comments:

40%-50%

coarse

fragments.

Zone 2 m.

0.24-0.53

Depth:

colors: Sample a, 2.5YR 4/6 (dusky to dark red), 2.5YR 3/6 (dark red); Sample b, 5YR 4/6 (yellowish red), 5YR 3/4 (dark reddish brown). Munsell Texture:

clay. hard.

Consistency: Lower

boundary:

Structure: Roots:

very few,

Horizon: Comments:

abrupt, fine

fine,

wavy. subangular

blocky,

moderate,

strong.

fine.

Bt. 25%-35%

coarse

fragments;

some

parts

are

rocky

up

to 70%.

Residue

Organic of

Pottery

Analysis

Sherds

from

Chrysokamino byRuth F. BeestonyJoe Palatinus, Curt Beck, and Edith C. Stout

We from

report on the results of organic residue analysis of twelve pottery sherds the Early Minoan Ill-Middle Minoan IA site of Chrysokamino. sherds were

These

excavated

from in and near a small structure located

in

close proximity to copper smelting furnaces where, according to evidence described elsewhere in this volume, arsenical copper was produced. Ves sel types represented in this group of sherds include jars, shallow bowls, a rounded cup, and ajar or bellows. residue analysis is a destructive method in which pottery Organic sherds are ground and samples for analysis are extracted with solvents capable of dissolving molecules that were porous matrix The technique

a wide

range of organic Organic compounds. and protected from degradation by the of the ceramic material are thereby released into solution. absorbed

of gas chromatography-mass is (GC-MS) spectrometry for the separation, detection, and identification of the minute quantities of often numerous organic compounds present in serves to separate the the sherd extracts. Gas chromatography complex

an effective method

mixture

of chemicals

stream of helium time

to traverse

by passing gas. Different

the column

it through

compounds and therefore

a heated

in a flowing require different lengths of at emerge and are detected column

as retention times). A plot of signal versus or TIC) time (the total ion chromatogram, is generated. As they emerge are from the column, molecules into in the mass spectrometer, passed are which ionized and broken collisions with apart by they high-energy characteristic

times

(known

electrons. A collection of fragment ions is produced, and their masses are amass spectrum, structure displayed in giving clues about the molecular to spectra of of the compound. The mass spectrum can be compared thousands of reference compounds using a computer-searchable database. In some cases, an exact determination of the identity of a compound can a be made based on pattern of peaks in the mass spectrum. matching In other cases, one can readily determine the class of organic molecules a to which acid ester, straight chain alkane, compound belongs (fatty methyl ketone, etc.), and that information, along with the retention time, a certain times of known compounds provides identity. The retention can be measured same conditions under in order to the independently a a for the identity of If peak cannot provide confirmation compound. be assigned to a particular compound, its mass spectrum is still recorded

M

APPENDIX

414 in the hopes

that additional

information

allow an assignment

will

in the

future.

in pottery sherd extracts found by GC-MS compounds to three sources. The first source, and the one of primary that came in contact with the pottery vessel interest, are those molecules use in ancient times; these were absorbed into the during its compounds The

organic can be attributed

surface and interior of the porous ceramic material where they remained careful interpreta preserved and protected from light and oxygen. With tion, these compounds provide clues about the original contents and use of the vessel. Unfortunately, human contact during excavation, cleaning, and storage of the sherd can contaminate the sample by intro handling, a second source of These include organic compounds. ducing compounds (from contact with plastic buckets, tools, and bags), chemical cosmetic products, and cleaners, all of which can contribute to

plasticizers sunscreens, the mixture

of molecules

the laboratory

detected

during organic residue analysis. Finally, in preparation involved of the sample extract procedures source of contaminants. introduce an additional Although

unavoidably in most cases these contaminants

are

and discounted, readily recognized their presence complicates the analysis, and it could mask the presence of molecules of archaeological significance. to be present From the list of organic compounds revealed by GC-MS in a sherd, the chemist can attempt to draw conclusions about the nature of

the original vessel contents. This difficult step involves examining known and combinations of botanical (and other) sources of specific compounds much is known about the chemical composition compounds. Although substances (wine, olive oil, plant extracts, etc.), not all botani cal species have been analyzed, and the nature of long-term changes in one must is still poorly understood. Therefore, be cautious composition residue analysis can the chemical results. Organic about overinterpreting

of various

on the use of certain "ingredients" based provide evidence for the likely the absence of that have been detected. However, collection of molecules for

evidence

factors, of

a

a

specific

including

compound

extremely

does

not

rule

low concentrations,

out

its presence;

can prevent

various

the detection

compound.

In an effort residue

to assess

analysis, sherds in this study were analyzed ratories equipped with different GC-MS three

sections,

subsequent showed

some

of organic the information obtained, in separate labo in duplicate

the reproducibility to enhance and ultimately

the collection

and limitations

instruments.

of molecules

As

confirmed

discussed

in

to be present

variation.

EXPERIMENTAL PROCEDURES Sample

Preparation

A 3-5 gram sample of each pottery sherd was obtained. Table M.l provides of ceramic material, and the the sherd number, vessel type, description some mass of sample for each of the 12 sherds in this In study. analyzed cases the entire sherd was used as provided;

for sherds weighing

>5 g, an

TABLE M.l.

OF

ANALYSIS

RESIDUE

ORGANIC

OF SHERDS ANALYZED

DESCRIPTIONS

POTTERY

SHERDS

FOR ORGANIC

415

RESIDUES Wt.

Vessel

EUMNo.

bucket 521 522

jar 523

shallow524 or jar 525

reddish

jar

red on outside,

brownish

jar

side, black

inner

brown

red, darker

side

layer 4.11

layer; white

2.82

side

on other

reddish

grainy side, red on other

on one

red, darker

527 jar 528

(ifknown)

with

brown,

pieces; brown on one

basin

on one

and white

two

bowl

(or bellows?) jar 526

bucket

Appearance

of Sherd

brown, grainy3-5 g thick, light brown with thin light layer on one side 3-5 g

jar

bucket

of

Description

Type

Ceramic

on one

3.08 one

g

side

3-5

g

g

and black

particles

3.10

g

g side

3.95

g

547 jar redwith gray sections 3.98 g 548 550

rounded cup 3-5 g 549 bowl shallow gray inmiddle, light red on one side 4.21 g a 3-5 g jar,probably bridge-spouted jar was obtained by breaking the sherd. To prevent appropriately sized sample were were in sherds handled with contamination, forceps and wrapped paper during breakage. Each sample was ground for five minutes

in a stainless steel mill, and the a teflon-lined cap. an in vial unused placed glass powder equipped with To the powder, 5 ml of dichloromethane (Aldrich 99.9+%, PRA grade) and 5 ml of diethyl ether (Aldrich anhydrous 99+%, ACS reagent grade) were was

in the vial was sonicated in an extraction, the mixture bath for 10-15 minutes at intervals during a two-day period. The

added. To facilitate ultrasonic

bath was kept below 40?C. The transferred to a clean glass centrifuge tube and centrifuged for 3-5 minutes. The clear liquid extract was removed with a glass disposable a new vial. The volume was reduced to about 1 ml pipette and placed in the vial in a 30-35?C water bath and directing a by partly immersing

temperature mixture was

stream

of

of the water

nitrogen

treated with

gas

in the ultrasonic

onto

the

surface

of

the

extract.

an ether solution of diazomethane

The

extracts

were

from Diazald

generated containing treated extracts were loosely capped and kept in a ventilated hood for 24 hours. The volume was again reduced under a stream of nitrogen gas until approximately 0.1 ml remained to produce a

and KOH. Vials

sufficiently

concentrated

Instrumental

sample for GC-MS

analysis.

Analysis

12 sherds were

analyzed atDavidson College; three of these (EUM 521, were and also analyzed at Vassar College. Table M.2 shows the 524, 526) instrument parameters in both laboratories.

All

were identified by mass College, peaks comparison of the mass to in those the NI ST contained database spectrum using the spectral library search capabilities of the SatView software. In addition, authentic saturated methyl esters, aldehydes, al samples of normal hydrocarbons, At Davidson

ketones, and others were injected under the same conditions for mass spectra. These standard comparison of retention times and compound cohols,

spectra were added to the searchable library.We as contaminants plasticizers, (including recognized

mass

do not report peaks modern synthetic

APPENDIX

4i6

M

some chemicals, and impurities introduced in the laboratory). However, are unusual molecules likely to have been introduced through handling mentioned.

At Vassar College, of Mass Spectral Data, (Palisade Corporation), standards of authentic

identification

was made

using

theWiley

Registry 6th ed., with the search system Benchtop PBM as well as an in-house MS DataBase listing samples and spectra from the literature. The re

ports of the Vassar analyses include some unusual extraneous compounds from plastic bags and from labora (but not the ubiquitous contaminants in whose identification chemicals) tory may be helpful to other workers the field. Because

the instruments

at Davidson

and Vassar

are

equipped with mass of retention times types analyzers, of individual compounds differ, as do some features of their mass spectra. in the three sherds analyzed in both laboratories were Some compounds detected on both systems, while other compounds were detected and iden tified on only one of the instruments. It appears that more slowly eluting such as long chain fatty acid esters and higher alkanes, were compounds, more on the instrument at Vassar. a detected readily Clearly, if compound different

columns

and different

on both

systems, that substance is certain to be present in the the comparative results between our laboratories indicate that in one laboratory should not be failure to detect a particular compound as an indication that the sherd did not contain that interpreted compound. is detected

sherd. But

that different portions of the same sherd, even when ana same the instrument and identical conditions, may differ quite lyzed using in the organic residue components that can be detected. We dramatically found in either laboratory are believe, therefore, that all the constituents

We

have found

actually present and

in the three sherds that were

analyzed

both

at Davidson

at Vassar.

RESULTS spectrometry allows Organic residue analysis by gaschromatography-mass the detection and identification of compounds that are soluble in the non polar extraction to pass through and nonvolatile,

solvent (ether/dichloromethane) and sufficiently volatile the gas Chromatograph. Polar, water-soluble compounds molecular molecules (for weight organic example, high are not detected. In this and other macromolecules) salts, carbohydrates, section we provide a summary of the findings for each of the 12 sherds from

Data including retention time and mass spectra (molecular Chrysokamino. ion, base peak, fragment ions) will be included in a later publication.1

EUM 521.Washed sherd from a bucket jar,100 (X 2) Seven compounds were identified in the extract of sherd EUM 521, includ and triacetin. A ing three alkanes, camphor and two related compounds, is evidence that this sherd contained, or was exposed large camphor peak in modern

times,

to a medicinal

preparation. Well

over 100 occurrences

1. Beeston

et al. forthcoming.

RESIDUE

ORGANIC

TABLE M.2. GC-MS

INSTRUMENT Davidson

GC-MS

ANALYSIS

OF

SHERDS

Vassar

College

College

ion trap

Spectrometer

GC Column

6890

Hewlett-Packard

5973 MSD MSD

x 0.25 mm

0.25

Injection

Injector Carrier

Helium;

Temperature

Program

flow

electron 2.5 min.

Base-Line

Recorded Correction

Scans Averaged

Helium;

for 19 min.

held

flow

rate 1.2 ml/min.

initial 50?C; immediately ramped to 250?C

at 5?/min.

40 min.

60 min.

Ionization

Masses

rate 1.2 ml/min.

initial 50?C; held for 1min.; ramped to 250?C

Filament Delay

poly(dimethylsiloxane), 0.25 \i 1:10 and splitless luL 250?C

at 5?/min.;

Total Run Time

x 0.25 mm

Lt

300?C

Temperature Gas

Column

15 m

1:10 luL

SplitMode Sample

Hewlett Packard HP-1 MS

ZB-5

poly(5% diphenyl; 95% dimethyl-siloxane),

Phase

Stationary

ChemStation

quadrupole

Phenomenex 30 m

GC

Hewlett-Packard

Hewlett-Packard Mass

417

PARAMETERS

Varian Saturn 2000 GC/MS/MS Varian 3800 GC

System

POTTERY

impact;

70 eV

electron

impact;

70 eV

1.5 min.

35-500

40-550

yes

yes

3 to 5

3 to 5

of camphor are reported by Duke,2 and this list contains many herbs, in cluding varieties of rosemary, savory, thyme, lavender, basil, hyssop, sage, Index lists several modern industrial uses licorice, and mint. The Merck of camphor, as well as its use as a topical anti-infective and antipruritic. Smaller peaks were identified as the related compounds, campholenal and in has been reported endo-isocamphonone. Campholenal Thymus funkii annuus (sunflower), a native of North America, (thyme) and Helianthus are listed3 or revealed a but no occurrences of endo-isocamphonone by Chemical Abstracts

search.

can be based on the presence of the three alkanes, tet radecane (C14), heptadecane (C18). Plant sources (C17), and octadecane are of these hydrocarbons known,4 but these compounds might also have been introduced by contact with a mineral oil or other petroleum-based No

lotion

conclusions

or ointment.

in this sample is likely to be due to con source reported for triacetin is the fruit of does not occur in theMediterranean region. It is an

presence of triacetin tamination. The only botanical The

Carica papaya, which as Enzactin it and Fungacetin). Therefore, antifungal agent (also known in EUM 521 was introduced through is a possibility that this constituent handling. As these results

2. Duke 2002. 3. Duke 2002. 4. Duke 2002.

illustrate, a common problem arising in the inter of residue about a organic pretation analysis is the lack of information com sherd's recent history. While plasticizers and other modern synthetic can be pounds easily recognized and ignored, natural ingredients introduced

via lotions, perfumes,

insect repellents,

and sunscreens

cannot

always be

APPENDIX

4i8 from compounds distinguished minimal the need for handling

M

arising from ancient sources. This points to of sherds to be analyzed by archaeologists, and for avoidance of products that might

and illustrators conservators, introduce contaminants through

contact with

sherds.

EUM 522. Sherd from a bucket jar,101 (X 169) In this sherd, eight components were identified, including camphor, two alkanes, four aldehydes, and a 2-ketone. No fatty acid methyl esters were It is possible that incomplete detected in the sample analyzed atDavidson. as the Vassar re occurred during the methylation procedure, sults (described below) revealed that fatty acids were present. As in EUM 521, was detected, but in much lower occurs in quantities. Camphor camphor cannot to aromatic its medicinal and be used herbs; many presence pinpoint esterification

a certain decanal

source. The plant aldehydes nonanal, decanal, undecanal, and do can each be attributed to numerous sources (28,23,5, and 9 plant

occurrences

are listed by Duke).5 Several plants, most notably respectively Coriandrum sativum (coriander) and several species of Citrus are reported to contain at least three of the four aldehydes observed in EUM 522. Since

citrus fruits were

not

introduced

in the Mediterranean

until

the

these can be ruled out; coriander was known and used as Ages, a flavor agent undecane during the Bronze Age.6 The two hydrocarbons, and dodecane, are common to Glycyrrhiza glabra (licorice root), Pimenta dioica (allspice, a native only of central America and theWest Indies), and

Middle

has ten known (Skhabar).The ketone, 2-undecanone, Cymbopogonparkeri rue and or?gano. exact botanical sources Although plant sources, including cannot be pinpointed, these results are consistent with the use of aromatic in the preparation detected at Davidson.

herbs

Sherd EUM

of flavored wine.

522 was

nents,

16 could be identified

nants,

and

3 remain

Pine

resin acid esters were

not

also analyzed at Vassar College. Of 24 compo as ancient constituents, 5 asmodern contami

unassigned.

is a group of 12 largest group of ancient constituents nine acids include the saturated fatty caprylic acid (C8), They acid (C9), capric acid (C10), undecanoic acid (Cll), tridecanoic acid (C14), pentadecanoic acid (C15), palmitic acid myristic The

fatty acids. pelargonic acid (C13), (C16), and

stearic acid (C18).The shorter, and especially the odd-numbered, members of this series are likely to be degradation products of the larger ones. More significant is the presence of the three unsaturated fatty acids, palmitoleic acid (C16:l), oleic acid (C18:l eis), and elaidic acid (C18:l trans). They are a certain indication of but plant oils. Olive oil has this composition, other plant oils cannot be excluded. The presence of two diterpene resin acid esters, methyl dehydroabi etate and its 7-keto-derivative, is a certain indication of pine resin. Its only known use is as an additive to wine to produce retsina. trace of borneol, identified with probability but not certainty, is a to assign amore specific origin. Duke plant product that is too ubiquitous lists no fewer than 141 plants containing this compound, including many A

conifers, but also herbs, e.g., or?gano.7 Biers and colleagues vases they studied.8 in two of the Corinthian "plastic"

found borneol

5. Duke 2002. 6. Mabberley

1997,

p. 146.

7. Duke 2002. 8. Biers, 1994,

p. 26.

Gerhardt,

and Braniff

ORGANIC

RESIDUE

ANALYSIS

OF

POTTERY

SHERDS

419

certain identification of acetanilide poses a puzzle. This compound or listed as a natural plant product in the compilations of Duke Karrer.9 A complete Chemical Abstracts search yields only one animal source (the sea urchin Temnopleuros toreumaticus) and one plant source (Peganum and Asia; it contains harmala). The latter is native to the Mediterranean an as alkaloids10 and has been used intoxicant possibly since hallucinogenic The

is not

to treat b.c.; it also has a long history in medicine and Parkinsons disease.11 acetanilide However, eye diseases, rheumatism, has also been reported in numerous analyses of modern waste water, into the 5th millennium

it is probably leached by the biod?gradation of a number of modern herbicides.12 Since the presence o? Peganum alkaloids cannot be confirmed our method of analysis, it remains uncertain whether acetanilide is an by

which

ancient

or a contaminant. We

constituent

have found

it previously

in a

tripod cooking pot from Chania, Crete. Of five certain and uncommon contaminants,

three are degradation a of structure. with plasticizers products 2,2,4-trimethl-l,4-pentanediol a is known ingredient of suntan lotions, and ditolyl 2-Phenoxyethanol ether is certainly amodern synthetic compound, albeit of unknown origin. suggest a resinated wine with aromatic or was ingredients that protected from air by a layer of oil (probably olive oil) to prevent its oxidation to vinegar.

Together, medicinal

these constituents

EUM 523. Sherd from ajar, 103 (X 211) The

most

the

alkanes

constituent found in the extract of EUM 523 is prominent As numer described this medicinal has camphor. previously, compound ous plant sources, but may also arise from contact with a modern topical preparation. Other organic compounds found in smaller amounts include undecane,

dodecane,

tetradecane,

pentadecane,

and

octadecane.

saturated hydrocarbons provide no relatively weight information. diagnostic Glycyrrhiza glabra (licorice) and Pimenta dioica are the (allspice) only plants inwhich all of these hydrocarbons occur,13 but

These

low molecular

is not indigenous to Crete. Other the latter possibility possible origins for the alkanes include contact with plastic or a petroleum product (mineral oil in skin lotion). Two aldehydes and a ketone were also found to be present. Nonanal, decanal, and 2-undecanone (all of which were also detected in EUM 522) each have numerous plant sources, including citrus fruits and herbs. Citrus can be ruled out based on its much fragrant and medicinal later introduction

9. Duke 2002; Karrer 1958-1981. 10. Harborne,

Baxter,

and Moss

1999, pp. 204,213. ll.Mabberleyl997,p.437. 12. andTuorinen Stampert

1998.

13.Duke 2002. 14.Duke 2002. 15. Beck coming.

et al. 2001;

Beck

forth

into the region.

EUM

524. Sherd from a shallow bowl, 94 (X 270)

EUM

524 contained

compounds,

not

a rich collection

of organic molecules. A total of 19 known contaminants, were identified. Many of including the alkanes tetradecanae and octadecane, and camphor,

these, including the aldehydes nonanal and decanal, have been discussed previously. The esters of acid hexanoic octanoic acid (caprylic methyl (caproic acid, C6), were and nonanoic acid identified. These acid, C8) (pelargonic acid, C9) are a occur acids in licorice root, hot pep present in variety of plants (all per, and cinnamon),14 and the C6 and C8 acids occur in the fat and milk of sheep and goats.15 Nonanoic acid may arise from the decomposition

M

APPENDIX

420

of oleic acid, an important constituent of olive oil, but this is an unlikely were not de source in this case, since higher molecular weight fatty acids are those of benzoic tected. Other methyl esters identified in EUM 524 acid is sometimes found in our acid. Benzoic acid and 2-ethylhexanoic in "blank" samples, indicating that this may be a laboratory contaminant a case. of sources16 has it number and this also has However, plant large been previously noted in pottery sherds as amarker for Aleppo pine,17 used to flavor retsina. 2-Ethylhexanoic acid has a number of possible sources, wine. and including fruits, herbs, 524. of interest were identified in EUM Several other compounds source listed by Duke is and its identity was confirmed by a an authentic study of saffron extracts, analysis of isophorone sample. In was found to be the second most abundant volatile component isophorone this spice was it cannot be determined whether after safranal.19 Although or a at colorant used as a flavor agent, a medicinal agent, Chrysokamino,

for which the only known Isophorone, Crocus sativus (saffron),18 was detected,

botanical

the presence of saffron in this vessel is very likely. Another compound of a known in this sherd is verbenone interest detected (2-pinene-4-one), constituent of 28 plants, including many varieties of rosemary, hyssop, mint, in oil of verbena, from Verbena germander, and sage.20 It is also present Verbena officinalis?2 The latter variety, also called vervain, and triphylla,21 was used for a variety of medicinal purposes in ancient Greece and is still used today for treatment Two

lactones,

of liver ailments.23

gamma-octalactone

and

gamma-nonalactone,

were

of each are listed;24 licorice identified with certainty. Four occurrences root is the only plant reported on both lists. Other possible sources are dis or cussed below. Another peak was identified as either 1,3-diacetylbenzene both of which are present in tea leaves.25 This com 1,4-diacetylbenzene, a more is likely modern contaminant. pound resin Finally, three peaks that clearly indicate the presence of pine ester of dehydroabietic were acid identified. The methyl (C21H30O2) was detected; two other peaks were identified as and C21H3602, C21H3402 with the same skeletal structure as the abietic acid, but compounds with one and no double bond, respectively. These compounds are certain in dicators of pine resin. Coupled with the observation of benzoic acid (detected as methyl benzoate) mentioned earlier, the resin ismost likely from pine. identified in this sherd reveal that the resinated The collection of molecules a herbs. variety of fragrant and medicinal was at Vassar College. Of 29 compounds also analyzed Sherd EUM 524 one is an be ancient 22 constituents, identified, may unassigned mixture, Six saturated fatty acids are unusual in and six are modern contaminants.

wine was flavored with

that they are of low molecular acid (Cll), (C10), undecanoic

acid (C9), capric acid weight: pelargonic lauric acid (C12), myristic acid (C14), and acid is acid (C15). The branched anteisopentadecanoic pentadecanoic a and hence action bacterial very likely the product of postdepositional contaminant.

The

absence of palmitic common

these are the most

striking; and plant fats.

acid (C16) and stearic acid (C18) is saturated fatty acids in both animal

16. 49 occurrences

reported

in Duke

2002. 17. Beck

et al. 2001.

18.Duke 2002. 19. Cadwallader, 1997,

Baek,

and Cai

p. 68.

20. Duke 2002. 21. Budavari

et al. 1989,

22. Duke 2002. 23.Jashemskil999,p.92.

24. Duke 2002. 25. Duke 2002.

p. 1565.

OF

ANALYSIS

RESIDUE

ORGANIC

POTTERY

SHERDS

421

Equally curious is the presence of six straight-chain, primary alcohols: 1-nonanol (C9), 1-decanol (C10), 1-dodecanol (C12), 1-tetradecanol and 1-octadecanol combination (C15) (C14), 1-pentadecanol (C18).The of straight-chain fatty acids and straight-chain primary alcohols is usu esters in epicuticular waxes, but the ally the result of the hydrolysis of the chain length of both the acids and the alcohols is here too short to allow that interpretation. The alcohols found occur in the free state in numerous so that no are assignments plants,26 possible. Two alkanes, heptadecane to invite assignment (C17) and octadecane (C18) are also too ubiquitous to a

single

source.

plant

with 8, 9, and 10 carbon atoms are known to gamma-lactones a occur fruits, according to together in wide range of Old and New World a Chemical Abstracts search. are also as They reported degradation products of plant oils (including olive oil) and animal fats, but the absence of palmitic and stearic acids in this sherd rules out these sources. Three

may be derived from Peganum harmala, or itmay be the product of modern herbicides, as discussed in the case of

Acetanilide

biod?gradation EUM 522 above.

are six occurrences of 1,4-diacetylbenzene in the literature are two in Abstracts of which found fruits {Chemical search), (peach and are found in industrial air and water the four other blackberry); pollutants. The antiquity of this compound is therefore in question. There

with

evidence that the vessel once contained wine resinated Unequivocal resin is and pine given by the presence of methyl dehydroabietate

13-ethyl-16-norpodocarpa-8,ll,13-triene.

are three the less common contaminants Among degradation products of 2,2,4-trimethyl-l,3-pentanediol from plasticizers, 2-phenoxyethanol suntan lotions, and 2(2-butoxyethoxy)-ethanol. In summary, the vessel must have contained resinated wine flavored or medicated with a range of plant materials that cannot be identified from the compounds found in the sherd. 525. Sherd from ajar or basin, 111 (X 525)

EUM Ten four

compounds alkanes

were

in the extract of EUM

identified

dodecane,

tetradecane,

and

pentadecane,

525, including octadecane.

the

These

to any number of plant sources or to contamination amodern product. As in EUM 524, both isophorone and verbenone were detected and confirmed by with authentic samples. These comparison are in saffron and verbena (aswell present respectively fragrant compounds as other aromatic herbs). another Camphor, compound of medicinal value a was derived from variety of plants, detected in EUM 525; its presence in most of the sherds from Chrysokamino suggests either a common ingredi may with

ent

be attributed

or

a common

The common 26. Karrer

1958-1981,

Duke 2002. 27. Duke 2002.

pp. 43-45;

source

aldehydes occurrences

of

contamination.

nonanal of these

and decanal two

were

also

identified.

Twelve

are

compounds reported,27 including of Citrus and the spices coriander, dill, ginger, saffiower, and lemon verbena. The only ester detected in EUM 525 was the methyl es six varieties

ter of octanoic

acid (caprylic acid), 17 occurrences

of which

are listed by

APPENDIX

422

Duke.28 The

vessel from which

amedicinal

herbal preparation,

this sherd originated most a flavored wine. perhaps

526. Sherd from ajar or bellows,

EUM

The most was

M

likely contained

63 (X 685)

significant result from the analysis of EUM 526 at Davidson the presence of a cluster of at least eight peaks for which the

College mass spectra revealed molecular

ions at 272. The

are consistent with

mass

for which

spectra of these numerous isomers

abietadiene, compounds are is related to abietic acid, an important constituent possible. Abietadiene of pine resin.When pine resin is heated (or perhaps allowed to decompose over a to can be long time) produce pine tar, the acid group of abietic acid lost through decarboxylation, resulting in compounds with the formula In addition, two peaks revealed mass spectra with molecular ions C2QH32. at 274 and 270, most likely due to related compounds with one fewer and one additional

double bond, respectively. We did not attempt to identify exact structures of these compounds, but their presence is a clear indication of pine resin or pine tar. include methyl compounds detected and identified atDavidson esters and methyl decanoate of (the methyl caprylic and capric Several peaks with spectra characteristic acids) and gamma decanolactone. of alkanes, but with retention times that do not match the unbranched Other

octanoate

alkane

reference

were

compounds,

also

noted.

Sherd EUM 526 was also analyzed at Vassar College. Of 35 compounds one is an identified, 27 are ancient constituents, unassigned mixture, and seven

are modern

contaminants.

sherd is rich in straight-chain, saturated fatty acids.The 11 members of this group identified are caproic acid (C6), enanthic acid (C7), caprylic acid (C8), pelargonic acid (C9), capric acid (C10), undecanoic acid (Cll), lauric acid (C14), pentadecanoic acid (C15), palmitic acid (C16), and stearic The

acid (C18). There are no unsaturated fatty acids, suggesting an animal fat rather than a vegetable oil as the source. The presence of anteisopentanoic acid most likely results from postdepositional microbial activity and is hence a contaminant. The branched acid has been fatty acid 2-ethylhexanoic reported

in wine;29

including wine meat and milk

a Chemical Abstracts

and Old

and New World

search yielded 135 occurrences, fruit, but it also occurs in the

of sheep and goats. The latter source is suggestive, since three of the saturated fatty acids (caproic acid, caprylic acid, and capric acid) in this sherd are named after their isolation from the genus Caper, or acid in 14 of the 24 goat. Biers and colleagues identified 2-ethylhexanoic vases "plastic" they analyzed.30 of low molecular weight, alcohol single

Corinthian A

1-nonanol

mon

(C9), is too com amounts of five

to be assigned to a particular plant source. Trace saturated hydrocarbons (alkanes) with 26, 27, 28, 29, straight-chain, and 30 carbon atoms were detected. These can be attributed either to the waxes

that coat the leaves, petals, and fruits of virtually all higher or to the insect waxes, of which beeswax is the most familiar. How plants, plant ever,

neither

wax

esters

nor

their

and fatty acids with more in this sherd.

alcohols

degradation

products,

i.e.,

long-chain

than 20 carbon atoms, have been found

28. Duke 2002. et al. 1999,

29. Danzer 30. Biers, 1994,

Gerhardt,

pp. 37-57.

pp. 26-34. and Braniff

RESIDUE

ORGANIC

A

single

natural

ANALYSIS

OF

POTTERY

SHERDS

423

(C10), has more

lactone, gamma-decanolactone

than 100

occurrences.31

The

presence

of pine resin is shown by seven diterpenoid

compounds: the hydrocarbon 13-ethyl-16-norpodocarpa-8,ll,13-triene, along with the a its 7-keto-derivative, resin acid esters methyl dehydroabietate, hydroxy and three other identified oxidation derivative, only partially products. The identification of methyl benzoate makes it possible to narrow the source of is indigenous to the the resin to the Aleppo pine, Pinus halepensis, which eastern

Mediterranean.

The

modern

2,2,4-trimethyl-l,4-pentanediol, i.e., the same contaminants and EUM

four plasticizers derived from and ditolyl ether, 2-phenoxyethanol, that have been found in sherds EUM 522

contaminants

include

524.

In summary, the ancient contents of this vessel are identified as a res inated wine with unspecifiable additions of botanicals that may have been or medicin?is. flavorings

EUM 527. Sherd from ajar, 110 (X 1658) were in EUM identified Eight compounds been discussed previously. The finds included of nonanoic

527, most

of which

have

ester camphor, two to and three In alkanes. addition acid, lactones, these, was detected. Of 17 occurrences of anisaldehyde reported in the methyl

anisaldehyde Duke, the most likely are Foeniculum vulg?re (fennel), Cuminum cyminum of these three plants (cumin), and Pimpinella anisum (anise). Cultivation in the Mediterranean

region for flavoring andMycenaean times.32 Fennel

toMinoan

and medicinal

purposes dates is still commonly used to treat

uses in ancient times, as recorded its medicinal digestive problems; by treatments include for Dioscorides, stomach, eye, and bladder problems, as well as for fever and animal bites.33 sources

Possible

detected

of

and

gamma-octalactone

gamma-nonalactone,

also

in EUM

524, include Glycyrrhiza glabra (licorice root) and several a of fruit. types Camphor may originate from any of number of botanical sources.

an oxidation nonanoate, presence of methyl product of oleic no be indicative of olive oil other acid, may although fatty acid esters were detected. The absence of peaks for other saturated and unsaturated fatty com acids may arise from a failure of our methylation procedure. The parative results reported above for sherds analyzed in both laboratories The

indicate

that the method

used at Davidson

was not as effective

at detect

ing fatty acids. EUM 527 was found to contain heptadecane, octadecane, and eicosane, saturated alkanes with 17,18, and 21 carbons, respectively. As previously these are of little diagnostic value since there are a number of mentioned, known plant sources, including the epicuticular waxes of many fruits, leaves, and flowers, aswell as modern products, that could have introduced these as contaminants. 31. Chemical

Abstracts

32.Mabberley

search.

1997, pp. 159,230,

454.

33Jashemski

1999, p. 52.

The

compound diethyltoluamide sherd. Clearly amodern contaminant, products.

in this (DEET) was also detected is present in insect repellent DEET

APPENDIX

424 The

of the organic composition of aromatic and/or medicinal

presence cannot be identified.

residue

in EUM

plants,

although

M

527

suggests the specific botanicals

EUM 528. Sherd from a bucket jar,34 (X 1850) interesting compounds were identified in EUM 528. One is anisal was also detected in EUM 527 and is discussed above. This dehyde, which use of fennel, cumin, or anise most compound likely originates from the as either a flavor agent or a medicinal extract. was also Ethyl salicylate found in this sherd. Duke lists only two occurrences for this compound,

Three

orientalis (hyacinth),34 known in the northeastern Mediter Hyacinthus to and ulmaria (meadowsweet), is indigenous which ranean, Filip?ndula as Eurasia.35 The third compound, identified is not isomethyl b-ionone, listed by Duke. The mass spectra of isomethyl b-ionone and isomethyl a-ionone ent. The

are

so either isomer could have been pres quite similar, however, a Abstracts results of Chemical search for occurrences of isomethyl is used in the perfume that this compound industry of the leaf oil of Peumus boldus,36 a tree native component and therefore not a possible source. Isomethyl a-ionone is also

b-ionone

indicate

and is aminor to Chile used

in

perfumes,

and

cosmetics,

hair

products.

Three

occurrences

are

a recent paper for this compound, listing isomethyl including as a component or ground elder,37 of Aegopodium podagraria, a a use since at least Roman times.38 plant with history of medicinal Ground elder was used for the treatment of gout, an inflamation of the reported a-ionone

ionone that the isomethyl joints of the feet and hands. It is possible found in EUM from an ancient plant product, but the 528 originated a or cannot be possibility of its introduction via modern cosmetic perfume eliminated. Other compounds identified in sherd EUM 528 include nonanal and occurrence two alkanes, dodecane and tetradecane. Due to the widespread none of these can be considered of these compounds, for a diagnostic particular

plant.

EUM 547. Sherd from ajar, 102 (X 144) sherd, like EUM 524, was found to contain a rich selection of organic molecules. A total of 22 compounds were identified in EUM 547. Finds include three alkanes (C14, C17, and C18) and four methyl esters (those of 2-ethylhexanoic acid, octanoic acid, nonanoic acid, and decanoic acid).

This

and octadecane were also detected together in Tetradecane, heptadecane, can be based on the presence of these com EUM 521. No conclusions as to a variety of plant sources or to a they might be attributed contaminant. The methyl esters of the C8, C9, and C10 acids have a are found in variety of plant sources; all Glycyrrhiza glabra, Cinnamomum aromaticum, and Capsicumfrutescens. Of these, only Glycyrrihiza glabra (lico rice) is indigenous to Crete. Other possible sources of these acids include

pounds, modern

of olive oil (C9). The fact that animal fat (C8, C10) and the degradation no were detected does not molecular acids weight fatty necessarily higher rule out the presence of oil and fat, as our experimental methods may not have enabled us to detect

these.

34. Duke 2002. 35.Mabberley 1997, pp. 227,283. 36. Bruns

and K?hler

1974.

37. Paramonov

et al. 2001.

38.

1997,

Mabberley

p. 11.

ORGANIC

RESIDUE

The

ketones

ANALYSIS

2-decanone,

OF

POTTERY

2-undecanone,

SHERDS

and

2-dodecanone

425

were

also

547; their retention times were compared with known to confirm the identifications. These compounds are almost certain samples to the medicinal plant rue (Ruta graveolens, Ruta chalepensis). ly attributable in EUM

detected

to standard by comparison samples) were the and dodecanal. Several occur corresponding aldehydes, decanal, undecanal, rences for each of these are listed but Coriandrum sativum by Duke,39 only

Also

present

(as confirmed

and two species of Citrus (lemon and petitgrain) are known to contain all three aldehydes. Citrus is ruled out based on its absence from Bronze Age Crete; coriander, however, was known very early in the Mediterranean. Additional in EUM detected 547 include camphor, compounds verbenone,

Each

gamma

anisaldehyde,

of these has been

identified

octalactone,

in other

and

gamma

nonalactone.

sherds as well. These

might have originated from aromatic herbs such as verbena, fennel, cumin, anise, and licorice root, as discussed previously for sherds EUM 524, 525, 527, and 528. Four

additional

1,2-benzisothiazole, compounds, N-methylphtha and diethyltoluamide lidimide, 3-bromothieno[3,2-c]pyridine, (DEET) were identified in EUM 547. These are almost contaminants; certainly is present

DEET compounds

inmost

insect repellents, but possible

sources of the other

are unknown.

The variety of compounds identified in EUM 547 is consistent with the presence of various herbs and medicinal plants. The absence of marker com or us from pounds for pine resin fatty acids prevents determining whether a was an or this oil infusion, flavored wine, another herbal extract. EUM

548. Sherd from a rounded

cup, 78 (X 210)

were detected and identified including compounds camphor, dodecane, octadecane, nonanal, decanal, and the methyl esters of octanoic acid and decanoic acid, all of which have been discussed previously. In this sherd, nine

to the number of plant sources known for these compounds, none offer information. The remaining two compounds found in diagnostic were EUM 548 tentatively identified as isomethyl b-ionone and isomethyl a-ionone lists four (one of which was also found in EUM 528). Duke occurrences for a-ionone under the (listed synonym a-cetone): isomethyl

Due

definitive

cuneifolia (Turkish savory), Nepeta racemosa (catmint), Origanum sipyleum, andMicromeria myrtifolia. Ground elder is another possible source, as discussed above in are used in regard to EUM 528. Both compounds the perfume industry (isomethyl a-ionone is used in soaps with scents of

Satureja

raspberry, cranberry, strawberry, and blueberry, air fresheners, perfumes, and potpourri); therefore, contact with amodern scented product cannot be ruled out as a source of these finds.

EUM 549. Sherd from a shallowbowl, 89 (X 168) of this sherd revealed a cluster of peaks, very similar to those Analysis found in EUM 526, for which the mass spectra showed amolecular ion at 272. This 39. Duke 2002.

is consistent

that these compounds abietic acid derivatives

with the formula and the spectra suggest C20H32, are isomers of abietadiene, degradation products of found in pine resin. The abietadienes are consistent

M

APPENDIX

426

the presence of pine tar (produced by heating or decomposition of was in from the vessel which this sherd obtained. pottery pine resin) In addition to the abietadienes, this sherd was found to contain many of the same compounds that are found in other sherds from Chrysokamino,

with

one ester (methyl including five alkanes (C12, C14, C15, C17, and C18), are too These and decanoate), camphor. widely distributed compounds in nature to allow any definite conclusions regarding the contents of the vessel.

were also found together in EUM 524 Isophorone and verbenone, which were use of saffron and verbena or and 525, also identified. They suggest the another aromatic herb, as verbenone is rather widely distributed. The three gamma-lactones,

and

gamma-nonalactone,

gamma-octalactone,

gamma

of these include a variety of fruits octalactone, none of which (including apricot, blueberry, papaya, pineapple, and peach, are native to theMediterranean), licorice root, and peppermint. are also present. Occurrences

Traces

of 2-decanone

occur

These

together

and Ruta graveolens Age Mediterranean.

and 2-undecanone

were

also noted

in this sherd.

tree native

to Chile) only in Peumus boldus (boldo (rue), amedicinal plant known and used in the Bronze

found in EUM 529 suggests the presence variety of compounds of pine tar (or aged resin) aswell as a variety of plant extracts. A resinated wine flavored with aromatic or medicinal botanicals is the most likely source of this collection of molecules. The

EUM

550. Sherd from jar, probably

a

jar, 82 (X 149)

bridge-spouted

in 550, we identified 25 organic compounds, 3 3 2 2 and esters, ketones, alkanes, lactones, aldehydes. cluding inwhich methyl esters of fatty This is the only sample from Chrysokamino acids longer than 10 carbons were detected atDavidson. The esters of both In the extract from EUM 8 methyl

acid (C14) and palmitic acid (C16) were identified, along with myristic decanoic those of (C7), (C10), nonanoic (C9), octanoic (C8), heptanoic and hexanoic (C6) acids. For each of these acids, several known botanical occurrences are listed; two plants contain all of these: Glycyrrhiza glabra triloba (pawpaw), a native of eastern North root) and Asimina source of these acids, Another possible particularly nonanoic, and palmitic acids, is olive oil. Oleic acid, the most prevalent myristic, was not detected. However, the nonanoic acid may fatty acid in olive oil,

(licorice America.

arise from the decomposition

of oleic acid. The

other ester found

in this

has many

occurrences, including wine, sample, methyl 2-ethylhexanoate, fruits, herbs, fish, and cheese. Dodecane (C18) were the (C12), heptadecane (C17), and octadecane are too to permit in identified EUM 550. alkanes These widespread only are sources. the assignment of specific The longer alkanes present in epi on leaves, fruits, and flower if such waxes petals. However, the source of these hydrocarbons, we would expect to find additional

cuticular waxes were alkanes The

containing lactones

nonalactone

were

more

than

20

gamma-hexalactone,

all found

carbon

atoms. gamma-octalactone,

in this sherd. Licorice

and

gamma

root is the most

likely

RESIDUE

ORGANIC

OF

ANALYSIS

source of these three lactones. The

POTTERY

combination

SHERDS

427

of the ketones 2-decanone

in EUM

550 (and also in EUM 547 and 549), is consistent with the presence of rue, amedicinal herb. Nonanal and decanal each have more than 20 occurrences; no additional diagnostic information may be obtained from the presence of these compounds. and 2-undecanone

found

Additional

compounds

identified

in this sherd

include

isophorone, 3,3,5-trimethylcyclohexanone (dihydroisophorone), is too ubiquitous (or the a isomer). Camphor methyl b-ionone

camphor, and iso in nature

is amajor constituent of provide any certain identification. Isophorone as discussed previously. The related compound dihydroisophorone saffron, is not listed in Duke, and a Chemical Abstracts search does not reveal any occurrences for this a or botanical likely compound. Isomethyl ionone (the b isomer), as noted in the cases of EUM 528 and 548, may be a modern to

or a constituent

contaminant Two most

additional

likely modern

of ground elder, catmint, or Turkish triacetin and N-methylphthalimide, compounds, contaminants.

savory. are

CONCLUSIONS have been found to contain a variety of can that synthetic compounds only have been introduced by recent contact with skin care products insect repellents, (medicinal ointments, sunscreens, etc.). This raises the possibility that some of the natural prod ucts found could either be ancient residues or modern contaminants. It is sherds from Chysokamino

The

modern

highly are

unlikely,

however,

that all of the many

natural products

identified

contaminants.

the conclusion that emerges from the numerous compounds Thus, found in the 12 Chrysokamino samples is that these vessels were not used for the preparation of ordinary food; cooked meals would have left much larger quantities of animal fats and vegetable oils. Rather, these vessels to have been used

in the making and storing of complex herbal included resinated ancient concoctions, wine, awell-known to more Greek make medicines strategy palatable. The organic residue are results from this of 12 sherds group analysis entirely consistent with the idea that the Chrysokamino hut (or the original location from which the sherds found in its soil originated) served as aworkshop for the preparation of herbal extracts from plants such as rue, saffron, coriander, fennel, anise, were and licorice root. Since the workers of Chrysokamino engaged in the ores to of arsenical make arsenical decidedly unhealthy activity smelting to it is were infer that these concoctions herbal intended copper, plausible to alleviate the symptoms of arsenic that i.e., poisoning, they were, in fact, medicinal preparations. appear

several of which

Finally, it is clear from the duplicate analyses of sherds EUM and 526 in two different laboratories that the organic compounds

522,524, detected

in two pieces of the same sherd can vary substantially. Reasons for the at results and obtained in Vassar Davidson include differences contrasting the methylation

procedure

and in the GC-MS

conditions

(which resulted

APPENDIX

428 in the detection sible introduction masked

M

of fewer fatty acid methyl esters at Davidson), the pos of different laboratory contaminants could have (which

the presence

of some marker

compounds

in each laboratory), and in different parts of a sherd.

the varying composition of organic compounds These observations indicate that negative results, the apparent absence as in an organic residue, should not be of a compound interpreted proof that the compound did not exist within the sherd. In order to obtain an list of nonpolar, volatile molecules present in a sherd, that sample have to be analyzed multiple times and under varying experimental a and is certainly conditions, costly time-consuming proposition. There a to a much information be gained from single analysis of pottery sherd; exhaustive

would

the presence of many compounds can be detected with certainty, especially to those when both the mass spectrum and the retention time are matched of an authentic sample. The list of molecules detected provides evidence for the use of various sel. However,

one

substances must

assume

(or types of substances) that

other

compounds

in the original ves are

also

present

but

due to low concentrations, peak "masking" by other compounds emerging from the column at the same time, or the to traverse the column and enter the mass inability of those compounds case in the of (as spectrometer fatty acids that have not been converted to methyl esters). Continued work in the of organic residue development undetected. This

absence maybe

and sample handling protocols will allow organic analysis methodology to an residue analysis to make further contributions of hu understanding man behavior in civilizations. early

APPENDIX

N

Petrographic

Two

Final

Neolithic

the

Chrysokamino

from

of

Analysis

Sherds

Location

Metallurgy byEleni Nodarou

were location at Chrysokamino pottery sherds from the metallurgy are both from the Final Neolithic selected for petrographic analysis. They period, and they derive from open vessels. The results of the analysis sug

Two

gest a broadly both vessels.

local provenance

DESCRIPTION

and similar manufacturing

techniques

for

OF SHERDS

Information

General

Fabric: Semicoarse

with metamorphic

rock fragments

Samples: sherds 1 (X 410) and 10 (X 611) Matrix

(Groundmass)

A. Optical properties and color some color differentiation 1.Under plane-polarized light: there is the core and the margins of the section. The margins are dark and the core is orange brown. are dark red brown 2. Under cross-polarized light: the margins and the core is dark orange brown; the micromass is optically active.

between brown

of inclusions and voids: coarse grain size and modality mm 1.9-0.1 from dimension. The fine fraction is vary long <0.1 mm and up to 60 micrometers in its longest dimension. Ratios for are ca. 35:43:8 (X410) and 15:82:3 inclusions and voids (coarse:fine:voids) (X 611). Voids range in size from 1.14 to 0.07 mm in longest dimen B. Overall

inclusions

sion.

C. Overall preferred orientation of inclusions and voids: in sample X 410, the voids display preferred orientation parallel to vessel margins; in are sample X 611, they randomly oriented. In both samples, the nonplastics are oriented. randomly

APPENDIX

43? D.

N

inclusions (under 62 micrometers) 1.Monocrystalline quartz Size range: up to ca. 62 micrometers Shape: equant, subangular to subrounded

Silt-sized

Comments:

straight extinction

2. Phyllite Size range: up to ca. 62 micrometers Shape: elongate 3. Biotite mica Size range: up to ca. 62 micrometers Shape: laths Comments: cases

Dark mica

it is also a component

is present

in the matrix, but in some and quartzite frag

of the phyllite

ments.

4.White

mica

Size range: up to ca. 62 micrometers Shape: laths mica ismost likely muscovite. It is present some cases it constitutes a component of the but in

Comments: White in the matrix,

phyllite and quartzite. 5. Polycrystalline quartz Size range: up to ca. 62 micrometers Shape: equant; subrounded Comments: (with evidence 6. Chert

In some cases, the quartz grades of metamorphism).

into quartzite

Size range: up to ca. 62 micrometers Shape: equant; subrounded. Comments:

fine-grained.

Voids A. Size range: Voids vary in size from 1.14 to 0.07 mm

in their longest

dimension.

B. Shape: Vugs are rare; the majority are mesoplanar and macroplanar are to close The in voids, vugs double-spaced. irregular shape; the planar voids

are

Inclusions

elongate.

above

Silt-Size

1.Monocrystalline quartz Size range: 0.77-0.1 mm

in longest dimension

Shape: equant; angular to subangular Comments: straight extinction 2. Chert Size range: 1.14-0.2 mm in longest dimension or Shape: equant slightly elongate; subrounded Comments: fine-grained

of

analysis

PETROGRAPHIC

fn

two

sherds

431

rock fragments 3.Metamorphic Size range: 1.9-0.76 mm in longest dimension (for the quartz (for the phyllite) ite); 0.77-0.4 mm in longest dimension to the is equant slightly elongate; angular quartzite Shape: subangular. The phyllite is elongate. constituent The main metamorphic of the grains have undergone alteration

Comments:

is quartzite.

or deforma A majority tion. In some cases, there is evidence of shearing. There are also rare fragments of phyllite, composed of biotite mica, quartz, and inclusions suggest an environment white mica. The metamorphic The scarcity and the small size of low-grade metamorphism. the metamorphic rock fragments lead us to characterize them as detrital in the fabric.

with

5.Micrite Size range: 0.5-0.2 mm in longest dimension Shape: equant; subrounded Comments:

merging

boundaries

6. Sandstone Size range: 1.14-0.38 mm in longest dimension to slightly elongate; subrounded Shape: equant Comments: there are two types; one is composed of unevenly distributed, equant quartz grains with very little clay matrix, oc (quartzarenite). The other is composed of small, equant quartz grains and biotite mica laths with more matrix (greywacke). casionally

"cemented" with

calcite

7. Alkali

feldspar Size range: ca. 0.12 mm

in longest dimension

Shape: slightly elongate Comments: simple twinning 8. Plagioclase feldspar Size range: ca. 0.15 mm

in longest dimension

Shape: elongate Comments: twinning polysynthetic 9.Mica Size range: ca. 0.08 mm in longest dimension Shape: laths Comments: white

mica,

10. Textural

mainly

biotite mica,

concentration

features

but there are some laths of

also.

Size range: 0.70

to <0.1 mm

(clay lumps) in longest dimension to subrounded

Shape: equant; subangular Comments: These features range in color from dark red to very dark brown or black. In some cases, they contain very fine quartz in others they have no inclusions. A few frag inclusions, while ments in both samples might be grog; their size ranges from 0.76 to 0.2 mm

in longest dimension; their shape is equant, angular to in and the color subangular; cross-polarized light is greenish gray. have boundaries and voids around the margins. merging They

APPENDIX

432

N

Modality Matrix

ca. 40%-50%

(groundmass): 3%-8%

Voids:

inclusions

Silt-sized

quartz: ca. 3% 1%

Monocrystalline Phyllite: under

under

1%

mica: under

1%

Biotite mica: White

Polycrystalline Chert: under

quartz: under 1%

1%

inclusions

Larger

Monocrystalline Metamorphic Sandstone: 5%

quartz: 20%-30% rocks: 5%-10%

under 2%

Micrite:

5%

Chert:

feldspar: under 2% Plagioclase feldspar: under 2%

Alkali Mica:

3%

Textural

concentration

features: 3%-7%

Discussion is a semicoarse, noncalcareous, red-firing fabric, characterized by the rock fragments. Its composi and detrital of presence quartz metamorphic seems that a red (alluvial?) deposit tion is not indicative of provenance, but it

This

of these vessels. The detrital quartzite and phyllite fragments point toward a red soil deposit, while the angular These shape of most of the grains indicates short distance transportation. area is around Chrysokamino observations, along with the fact that the dominated by the presence of the Phyllite-Quartzite series, suggest a local to not It is for this pottery. assign specific provenance, but possible origin the red alluvial deposits of the north coast are very likely to have provided has been used for the manufacture

the raw material

of this pottery. for the manufacture similar two samples examined present textural and compositional an shows similar the matrix of color The firing conditions, consisting of ity. the low similarity firing temperature. Moreover, oxidizing atmosphere and such as the in the texture of some of the plastic and nonplastic components, inclusions, lead us to suggest chert, the micrite, and the other nonplastic in terms of raw material that the two samples share the same provenance used and technology of manufacture. The

REFERENCES

P. 1997.

Acheson,

Territory of Halieis

in the

JMA 10, pp. 165-190. of

Khok Phanom D, Central Thai (diss. Univ.

land"

New

of Otago,

Andreou,

"npc?toinivc?iKai xoccpoci Hpcx 7tocp? xo KavXi-KacT?XXi

k?,??od," CretChron -.

5, pp. 275-294. "Te?%r| Kai ocKporc?taic

1979.

OTT]MlVCOlKT) KpT)TT| (0 UUGo? ir|? uivcoikt|?

8,

KprfToAoyia

?ipr|vr|?),"

pp. 41-56. -. 1980.

pp. 9-22. 1994. Kpr?Tr?:Die Samaria

A.

Alibertis,

schlucht

Athens,

A',

Ivve?piov

und

ihre Pflanzen,

Herak

leion. Alimov,

Boroffka,

M.

J. Cierny,

Bubnova,

J. Jakubov,

E. Pernicka, Parzinger, V. Radililovskij, V. Ruzanov, T. Siri nov, D.

bergbau

inMittelasien,"

Eurasia

137-199.

an

Underdeveloped

Sheet 8," inM?moire explicatif de la m?tallog?nique

de

l'Europe

and Greece,"

in A Review, Aegean Prehistory: ed. T. Cullen, Boston, pp. 259-327. G. Cadet, Delibrias, J.-P. J., Ang?lier, M. Gigout, M. Guil C. Lalou, and Hogrel, 1976. "Les d?formations

J. Fourniquet, lemin, M.-T. Perre.

du Quaternaire

marin,

n?otectoniques:

Quelques exemples Revue de g?ographie

m?diterran?ens,"

indicateurs

et de physique g?ologie dynamique, 2nd ser., 18:5, pp. 427-448. and L, V. Tolstikov, Antonova, 1996.

The Gold

Homers

Searching for London.

Fabled

of Troy: City,

M. 2000. "Human Atherdon, Impact on the of Southern Vegetation and Problems Greece of Palynologi

Aurenche, 1999.

Area,

J., and C. Koukouzas. Anastopoulos, 1984. "The Metallogeny of Greece, Carte

and K. Kotsa

of Northern

Age

Frederick

Allbaugh, L. G. 1953. Crete:A Case Study of Princeton.

Sampson,

A Case cal Interpretation: Study in Halstead from Crete," and

"Pr?historischer

4, pp.

Antiqua

and G. Weisger Zinn

Starsinin,

1998.

Kai to

"The Neolithic

M.Treister. K., N.

J. Burjakov, J. Lutz, H.

ber.

Bronze

G.

aicpo "npoocvocKTopnc?? in Uenpayii?va TioXei? Trj? Kprjiric," xov A'AieOvov? Kpr\ToXoyiKov

ed. A.

pp. 435-437. S., M. Fotiadis,

kis. 2001.

S. 1951.

Alexiou,

%<x?,kiv?v avxi

o oikioji?c,

VEKpoxacpe?o, Athens,

Zealand).

1993.

Gappou Keijx?vcov arc? xr\ Ekot?ivt| v: vicov," in iKoreivrf Oappovvi To onr\Xaio,

J. J. 1992. "Archaeological as Artifacts. The Dirt Sediments

Aitken,

Dunedin,

E.

"Xr||iiKri av?X\)cr\

Period,"

pp. 411

Paris,

limitrophes,

Andreopoulou-Mangou,

in the Late

Hellenistic

Classical-Early

pays 427.

Settlement,

and Erosion

Agriculture,

des

the Economic

"Does

Work?

Explanation

et

pp. 62-78. and S. K. Kozlowski.

2000, O.,

La

naissance

Proche

Orient,

Bachmann,

H.-G.

du

n?olithique

au

Paris. 1980.

"Early Copper

Smelting Techniques in Sinai and in theNegev asDeduced from Slag Investigations,"

in

Scientific

Studies

REFERENCES

434 in

and Extractive Early Mining ed. P. T. Craddock,

Language

London,

-.

ed. P. Bellwood

103-134.

pp.

1982. The Identification of Slags Sites (Institute from Archaeological Publi of Archaeology Occasional cation

Baumann,

and F. Lo

B., and

and Congress of Prehistoric Italia Protohistoric Sciences, Forli,

Baumann,

as

Ball, D. F. 1964. "Loss-on-ignition an Estimate of Organic Matter

Soils," Journal

of

Soil Science

H.

Bambauer,

D.Trochim.

Determination Minerals

by M.

Bamberger,

15:1,

Archaeology lar and Site

in Craddock

and

Bamberger,

"Ancient Smelting of Oxide Copper in The Ancient

Ore,"

Metallurgy

ed. B. Rothenberg, Copper, 123-157. pp. E.

Banou,

S. 1995a. House

(Seager's in Pseira -.

I, pp.

1995b. The

Pottery,"

London,

AD

108

G.

1985.

Prehistoric

Cambridge. and J. Lloyd,

Europe, Barker,

G., Roman

Landscape, 0.2001.

Bar-Yosef,

Farming eds.

Philadelphia.

in

1991.

London. "From

Sedentary to Hierarchies: Foragers Village Institu The Emergence of Social tions," Social

Human in The of Origin Institutions (Proceedings

of the British Academy 100), ed.W. pp. -.

G. Runciman,

London,

1-38. 2002a.

Paleolithic

"The Upper Annual Review

Review,"

Anthro of

pology 31, pp. 363-393. -.

2002b.

"The Natufian

Culture

and the Early Neolithic: Social and Economic Asia,"

Trends in

Examining

in Southwestern the

Farming

M.

C. W. R. F, J. Palatinus, Beeston, and E. C. Stout. Forthcoming.

"The Pottery, Building in Pseira II, pp. 13-26.

AC,"

Inferred

JFA 3, pp. 361-374. J., and P. P. Betancourt. 1997. Richard Berry Pioneer Seager: and Gentleman, Proper Archaeologist

1998.

Barker,

Aloupi.

and of Cy

on Crete,"

124. -.

K. M.

Stout,

V. Karageorghis, 2001. "The Uses

prioteWhite-Slip Ware

Becker,

Center.

I, pp.

E.

and

from Organic Residue Analysis," und Levant 14, pp. 13-43. ?gypten Sources Becker, M. J. 1976. "Soft-Stone

of

"Building AA A). The Pottery," 19-21.

"Building in Pseira

E. H.

C. W.,

in Bronze

Investigations ed. Y. Tzedakis Oxford.

Woukulich, 1990.

and P.Wincierz.

M.,

Age H. Martlew, Beck,

"Certainty

in Remains," Organic Science: Biomolecu Meets

Greece,

1985, pp. 151-157.

Hughes

inArchaeol

in the GC-MS Analysis

of Ancient

of Rock-Forming W. E. Tr?ger, Stuttgart. 1985. "The Working

Process,"

Statistics

2003.

alpidischen

Ag?is," 66, pp. 492

Rundschau

andDoubt

Conditions of theAncient Copper Smelting

"Die

s?dlichen

ogy, London. Beck, C. W. Forthcoming.

and F. Taborsky, 1979. Optical

U.,

1977. der

Geologische 522. Baxter, M.

84-92.

pp. H.

and

in Non-calcareous

Carbon

pp. 33-40. and H.

30:3/4,

Stockwerke

"The

Crete,"

Best,

G.

A.,

Wachendorf.

8/14 September 1996, Forli.

1976.

of Eastern

Tectonophysics

national

Organic

Pile

Nappe

Inter

Thirteenth

Europe:

New York. Deposits, G. Best, W. Gwosdz,

A.,

and H. Wachendorf.

eds.

Schiavo,

1996. The CopperAge in theNear East

and C. Renfrew, pp. 113-126. Cambridge, A. M. 1946. Economic Bateman, Mineral

6), London.

Bagolini,

Hypothesis

Dispersal

(McDonald InstituteMonographs),

Metallurgy,

Beck,

New "Organic Residue Analysis: at in Finds Chrysokamino," Meets Science: Biomolecu Archaeology lar and Site in Bronze Investigations Age Greece, H. Martlew,

ed. Y. Tzedakis

and

Oxford.

F, S. Schmitt-Strecker, Begemann, 1992. "The and E. Pernicka.

Metal Finds from Thermi III-V: A Chemical Study," 239.

Studia

and Lead-Isotope Troica 2, pp. 219

and F., E. Pernicka, Begemann, 1995. "Thermi S. Schmitt-Strecker.

on Lesbos: A Case Study of Chang ing Trade Patterns," 123-136.

OJA

14,

pp.

Beile-Bohn, and K.

M.,

C. Gerber,

Schmidt.

1998.

sche Forschungen

im

M. Morsch, "Neolithi

Obermesopo

REFERENCES

tamien:

und G?bekli G?rc?tepe IstMitt 48, pp. 5-78. Tepe," S. 1983. "D?couvertes Bernus, hypo et preuves: th?ses, reconstitution Le

cuivre m?di?val

-.

P. P. 1977.

in

(Aegaeum R. Laffineur, -.

Chrono

logical Problems in theMiddle Minoan

Dark-on-Light

Eastern

Crete,"

Pottery 81, pp. 341

AJA

of

1979. Bronze

Age

Early in Crete

Style

{SIMA 56), G?teborg. -. 1980.

Cooking Kommos: A

Minoan

Vessels

from

from Vasilike, Pseira, Sphounga and Other ras, Priniatikos Pyrgos,

Sites

Museum

(University

Dark

Ware

The History Princeton.

Pottery, -.

1994-1996.

Betancourt,

Minoan of

"MivcoiK?

Kpr?rncr? Ecrr?a

-.

1995.

?U7i?pio

5, pp. 47-70. for Trench

"Handbook

Supervisors,

Chrysokamino

Study

Center

Pacheia

Crete),

Ammos. -.

1997. Site

"The Copper

Smelting

at

Crete," Chrysokamino, Greek Proceed Technology: 1st International ings, Conference, Thessaloniki 1997, Thessaloniki,

Ancient

-.

pp. 51-54. 1998.

"Middle

Minoan

and

the Orient b.c.:

Proceedings

50th

Anniversary cinnati, 18-20

Cline

-.

in the Second

Symposium, April

1997,

and D. Harris-Cline,

pp. 5-12. 1999.

FN/EM

"What

and

Day, Mook,

of the Cin ed. E. H. Li?ge,

isMinoan?

I in the Gulf ofMirabello

and R. A.

'Plastic'

Investiga Vases by

Chromatography-Mass L. R.

Spectrom

1968a.

Archaeology, L. R. Binford, 1968b.

"Archaeological inNew in Perspectives ed. S. R. Binford and pp. 5-32. Chicago, "Post-Pleistocene

Adaptations," in Archaeology, L. R. Binford,

inNew

Perspectives ed. S. R. Binford

and

pp. 313

Chicago,

341.

Ecria

P. P., C. R. Floyd, "Excavations

1997.

sokamino,

1990.

Environ Bintliff, J. L. 1977. Natural ment and Human Settlement in

Prehistoric Greece (BAR-IS 28), Oxford. Bintliff, K.

P. Howard, J. L., E. Farinetti, and K. Sbonias. 2002.

Sarri,

"Classical

8,

Crete,

and J. D. at Chry

1996," AJA

101,

Hidden

Farms,

Prehis

3, pp. 47-54. P. P., and N. Marinatos. 15-19. P. P., J. D. Muhly, W. R. L. Steams, Onyshke

B. Hafford, and D. Evely. vych, W. at 1999. "Research and Excavation Crete,

Hesperia

Chrysokamino,

Crete,

AJA 102, p. 391.

in 1997,"

and A. Snod J. L., P. Howard, 1999. "The Hidden Land

scape

of Prehistoric

12, pp.

Greece,"

JMA

139-168.

J. L., and A. M. Snodgrass. "Off-Site Pottery Distri butions: A Regional and Inter

Bintliff,

regional Perspective," pp. 506-513. "Les D. 2002. Bischoff,

pp. 28-37. A. K.,

Biswas,

and W.

1976. Extractive

1995-1998,"

68, pp. 343-370. P. and Betancourt, P., J. D. Muhly, C. R. Floyd. 1998. "Excavations

grass.

CurrAnthr

29, de

megaliths

G?bekli Tepe," Arch?ologia 393,

C.

Chrysokamino,

JMA15, pp. 259-265.

1988.

1997. "TheMinoan Villa," inH?gg pp.

and Greek Rural Landscapes, A Response and Update," Survey:

Bintliff,

pp. 374-375. P. P., and R. Hope Betancourt, Simp son. 1992. "The Agricultural System of Bronze Age Pseira," Cretan

Farrand,

Scents:

toric

Betancourt,

Betancourt,

Objects in theNear East," in The Aegean Millennium

of

Gulf VIII,

pp. 9-22.

1997,

Corinthian

Perspectives,"

pp. 20-37. P. P., and C. R. Floyd. Betancourt, 2000-2001. "AvaGKacpri gto Xpu

Betancourt,

of

Binford,

and Knossos:

KprjTiiaj

Lost

etry (JMASCAP11), Philadelphia.

I??Pio?oi 1987-1989," Ajua??eia

GOKauivo,"

1994.

Gas

82-85,

Studies

pp. 217-255. Biers, W. R., K. O. Gerhardt,

in Pseira

?p?i)V?? GTT|v Ydpa:

"Ap%aioXoyiK??

Landscape

in the

P. P., and C. Davaras.

Muhly.

Excavations"(INSTAP for East

in

"The Rural

tion

of an East

Seaport,"

2002.

A.

The

Muhly 2004, pp. 21-28.

X?Ood: Ta oTOi%??a arc? vc\ vtjgo ^E?pa,"

"Pseira

Slags AJA

of Neopalatial Kythera,"7M^ 15,

-.

Transformation

White-on

51), Philadelphia.

1985.

The

Cretan

Museum

(University

Monograph -.

Cretan

-.

Crete,"

Philadelphia. Bevan,

Lime

in Pseira

Area,"

Smelting

P. P., and J. S. Silverman. (Uni Pottery from Gournia 72), versity Museum Monograph

Braniff.

Century,"

in the Eastern

pp. 99-104. 2004c.

Mono

graph 47), Philadelphia. -. 1984. East

pp. 91-97. "Traditional 2004b.

Mirabello

Objects Exca

vated

VIII,

Production

Preliminary

1983. Minoan

Early

Twentieth

-.

Report {UCLAPap 7), Los Angeles. -.

Archaeometry 2004a.

1999. "Copper from Chrysokamino, 103, p. 273. Betancourt, 1991.

and

Eastern Gulf ofMirabello

Vasilike Ware: An Pottery

-.

Bronze Aegean Age ed. K. P. Foster and

Archaeology 3, pp. 3-12. "Coastal Trade:

Faenza,

Heritage, pp. 395-399.

pp. 117-121. Li?ge, 2003c. "The Impact of Cyclad on ic Settlers Crete," Early Minoan

Mediterranean

-.

353. -.

the 24),

Cultural

ed. P. Vincenzini,

inMETRON:

Crete,"

in The Ceramics

Crete,"

Ceramic

"Interpreting

Measuring

1985,

"Some

and

Petrography: The Special Product Model, aNew Model for the Distribution of Pottery in Early Minoan

P. P., and G. H. Myer.

Betancourt,

1995. "Phyllite Fabrics in Eastern

"Interpretation in Pseira VII,

-. 2003b.

pp. 71-80. Betancourt,

1,

123-139.

pp.

"MetalWorking in theAgadez Region (Niger):An Ethno Approach,"

2003a. Conclusions,"

dAzelik-Takedda

Craddock andHughes

vol.

inMELETEMATA, Region," pp. 33-40.

in (Niger)," Metallurgies africaines, ed. N. Echard, Paris, pp. 153-171. 1985. Bernus, S., and N. Echard.

Archaeological

435

at

G.

Davenport.

Metallurgy

of Oxford. Copper, Bittel, K. 1959. "Beitrag zur Kenntnis anatolischer der Metallgef?sse

zweiten H?lfte des dritten Jahr tausends

v. Chx.,"JdI74,

pp.

1-34.

REFERENCES

436 Blackman,

D.

1975.

"An

J., and K. Branigan.

the South

-.

Survey between

Archaeological of Crete, Coast

-. 1994. ronment

the Ayiofarango and Chrysosto mos," BSA 70, pp. 17-36. 1977. "An Archaeological of the Lower

Survey

Catchment

of

Blitzer,

Md,

13-80. 1993.

H.

"Olive

Bottinga, 1983.

Cultivation

and Oil Production inMinoan in La

Crete,"

production enM?diterran?e.

de Vhuile

nean Area, -.

Oil and

163-175.

Social Cycles of ofPhaistos: in theMesara Region Complexity Crete (Monumenta Archaeologica and H.

Vallianou, Angeles, Bogaard, and G.

Charles,

pp. 909-919. H. A. 1901. Boyd,

of

Kavousi, pp.

P. Halstead,

"The Scale Jones. 2000. Evi of Cultivation: Intensity in dence from Weed Ecology,"

1.1976.

"On

stazione

e Festos,"

Haghia n.s. 29-30,

J. Cierny,

zinger, E. Pernicka, "Bronze ber. 2002.

in Eurasia

(McDonald

Levine,

E.-M.

tigations

and Bronze

Copper

Bronze

Early

Crete

Age

The Foundations

1970a. Palatial

1970b.

of

London.

Crete,

The Tombs

ofMesara,

London. -.

1971. Source

Metal pp.

"An Early Bronze Age in Crete," SMEA 13,

10-14.

-. Early Oxford.

1974. Aegean Metalwork and Middle Bronze Age, 1984.

"Early Minoan

of

the

Society:

The Evidence of theMesara Tholoi 1967. ?ber

der pr?historischen ArchDelt 22, A, Bottema,

-.

-.

135-159.

Vorbericht

Bronze

(SIMA 19), Lund.

C. Ren

Cambridge,

in

Working

Institute

ed. K. Boyle,

Monograph), frew, and M.

pp. 211-239. 1968.

-.

Notes," Preliminary Interactions: East and West

in Ancient

Bossert,

-.

ASAtene,

J. Lutz, H. Par and G. Weisger Tin from Age

"The Early

Age Daggers of Crete," BSA 62,

Asia:

Central

pp.

fra

1967.

K.

Branigan, in

pp. 7-54.

N.,

Boroffka,

Report

Exploration at Gournia, Excavations

1904," University of Pennsyl Transactions the of Department Free Museum of ofArchaeology, Science and Art 1, pp. 177-188.

the Ori

medio-minoica

Triada

"Gournia:

vania

Europe," ments: S. ed. J. V. Piggott, Festschrift S.Megaw, London, pp. 69-76. N. 1967-1968. "Patriki?s? Bonacasa, Una

1904-1905.

Society's

129-134.

Copper-Processing in To Illustrate theMonu

5,

Crete,

Halstead and Frederick 2000, Bogn?r-Kutzi?n, gins of Early

in 1900," AJA

of the American

and

pp.

at

"Excavations

Crete,

125-157.

-.

111-217.

pp. A., M.

1972.

Compositions," et Cosmochimica Acta 46,

Geochimica

Los

Blitzer,

Y.,

Silicate Liquids 1:Revised Method

D. Hadzi

ed. L. V. Watrous,

min?

for Aluminosilicate

Plain

23),

Bulletin

pp. 129-138. and D. F.Weill.

Science 272, Journal of pp. 438-475. and P. Richet. Y, D. Weill, Bottinga, 1982. "Density Calculations for

in The

Mesara,"

Liquids,"

American

and Sub "Agriculture in the Late Ottoman and

Post-Ottoman

F Weill.

and D.

of the Density Coefficient Expansion

"TheViscosity ofMagmatic Silicate Liquids: A Model for Calculation,"

and

pp. 403-535. 2004. sistence

the

106,

ralogique

Implements in Kommos 1:1,

Industries,"

of the

"Calculation

Bottinga,

Amouretti

J.-P. Brun, Paris, pp. 1995. "Minoan and

-.

ed. M.-C.

A Review

pp. 45-68. Y, P. Richet,

of Silicate

in theMediterra

Production

Wine

Envi

"The Prehistoric

of Greece:

and Thermal et

du vin

31,

Palynology

in Record," Palynological Beyond in the Site: Studies Regional Aegean ed. P. N. Kardulias, Area, Lanham,

theAyiofarango Valley," BSA 72, pp.

and

Paleobotany pp. 193-217.

on

"Kastri auf Syros: eine Untersuchung

in Aux

nisme: La Cr?te

Inves

S. 1980.

of

origines et la Gr?ce.

van Effenterre, let, Paris, pp. 29-37. -. 1988. Pre-Palatial:

?Henri

Siedlung," pp. 53-76.

"Palynological on Crete," Review

Reviewed,"

dations

of

Palatial

de l'hell?

Hommage ed. C. Nico

The Foun

Crete: A

Survey

of

REFERENCES

in the Early

Crete

Bronze

1991.

"Mochlos?An

-.

Early

'Gateway Community'?" Aegean et la in Thalassa, VEg?epr?historique mer: Actes de la troisi?me Rencontre

of the

Society

ed. P. Halstead,

Greece,

Sheffield,

Future Metal

pp. 57-65.

Broodbank, C. 1992. "TheNeolithic Social

Change the Bronze Age,"

Labyrinth: sos before

-.

A.

Brown,

Evans's

Island

K.,

JMA

5,

Archaeology

1974.

K?hler.

"?ber

and P. E. Heckelman,

A.

Smith,

eds.

1989.

and

-.

Budd,

Copper:

Heat

Biologicals,

Treatment

and

the

P., B. Chapman, C.Jackson, R. Janaway, and P. Ottaway, eds. 1991. Sciences 1989: Archaeological on the Proceedings of the Conference

Application of ScientificTechniques to

Bull,

I. D,

Implications of Eneolithic

dence Manuring Pseira,

for a Structured Regime Crete during

on

R.

the Minoan

M.

8), pp. 60-63. Suppl. (Hesperia 1995. "AMixed-Method to in

Archaeological

Sediments and

I. Macphail,

London,

pp.

and 183

State"

(diss. Univ.

Los Angeles).

of California,

et 1946

"Early Arsenical View,"

150-181.

-.

and

"Determinative

the

Origins

Mineral

of Metallurgy,"

inCraddock andHughes

1985,

pp. 21-28. 2000.

Furnace:

Out

of the Fiery and Meditations

Recollections

Geoarchaeological

rapport: Explo

1980. "The of Copper Coming and Copper-Base and Iron: Alloys A in The Metallurgical Sequence," the A. ed. T Iron, Coming of Age of Wertime and J. D. Muhly, New

ogy

Carothers, J. J. 1992. "The Pylian Kingdom: A Case Study of an Early of

-?.

Leslie

of Theodore

and ex?cut?es

Metallurgical

Haven, pp. -. 1985.

190.

Evi

Agricultural the Island

Region," inHonor

Aegean Studies

Review

AJA 71, pp. 21-26.

in the Copper in Commemorative

Interpretation, Analysis, ed. A. J. Barnham Management,

and P. P.

"Chemical

of Arsenical

1967.

J. A. Bronzes?A

and C. Ho, Washington, pp. 66-79. E. R. 1949. "On the Prehistoric

Analysis," and Soils:

Metallurgy," pp. 132-142.

et al. 1991,

1999.

Caley, Use

Quatri?me

Charles, of

Flavor

Approach

Copper for the Devel

R. P. Evershed,

Betancourt.

"The

and

"Characterization

(ACS Symposium Series 660), ed. D.C.,

CAJ11, pp. 105-121. 1973. Documents inMyce J.

ration dupalais (1929-1935 1960) (?tcr?t 12), Paris.

Society,"

Baek,

of Agriculture:

Feature,"

?Mallia.

of the Pyrgos

S. J. Risch

Canti, 1991.

1997.

Origins

Greek, Cambridge. F, P. Demargne, Chapouthier, 1962. Fouilles A. Dessene.

pp. 97

inMinoan

Cambridge. G. O. Rollefson, J., I. Hodder, and T. Watkins. 2001. Bar-Yosef,

Chadwick, naean

Extract by Aroma in Flavor Dilution Analysis," Spices: andAntioxidant Chemistry Properties

Shear

September

Budd, P., and B. Ottaway. of Arsenical Properties

"The Role

trans.

ofAgriculture,

"The Birth of theGods and the

Pal

in Ilenpayii?va

K. R., H. H.

Cai.

Saffron

Budd,

opment in Budd

Country

M.

of Interpretation in Processes," Manufacturing Pernicka and Wagner 1991, pp. 35 44.

Archaeology, Bradford, 1989, Oxford.

Cauvin,

Kprfro?oyiKov

House

Cadwallader, Arsenical

Metallographic

Alloys:

1997.

the Origins

T Watkins,

inH?gg 1997, pp. 99-103.

11th

ed., Rahway. P. 1991. "Eneolithic

a

Rethymnon,

1739

pp.

2000. The Birth of theGods and

37, pp. 172-174. "Mallia and Lasithi:

A',

IvveSpiov 104.

TheMerck Index:An Encyclopedia of Drugs,

-.

BICS

Palace-State," Tov Z'AieBvovq

and

Rotterdam,

des divinit?s, J. 1994. Naissance naissance de Le r?volu l'agriculture: au tion des Paris. symboles n?olithique,

O.

A

P. Marinos

ed. G.

Cauvin,

in

in the Old

the

1745.

Baltimore.

"Lasithi

ace Period," -. 1995.

und Kosmetik

Parfumerie 55, pp. 225-227. S., M. J. O'Neil, Budavari,

ical Sites,

145-148. -. 1990.

and

Bronze

Cambridge. L. Mendoni,

G. C. Koukis,

Minoan Landed Gentry?" BICS 18, pp.

J. 1964. Greece, Crete, in the Islands Early

A. Papasta 1988. and N. Neloyannis. in Keos: A First "Metals Approach," in The An of Engineering Geology cient Works, Monuments, and Histor

Kim,

"Was There

Aegean 102,

AJA

mataki,

Smelting

pp. 331-338. Land and Labor

1971.

G.

Cadogan,

of

des Boldo

die Zusammensetzung bl?tterols,"

Chemicals,

1993.

Age,"

M.,

Caskey,

Korea,

Gyeongju, Burford, A.

Cyclades, Cambridge. 1993. Knossos?Arthur Before Travels in the Balkans and

and M.

Age, 2002.

K.-W.Yi,andH.-TKang,

at Knos

Crete, Oxford. Bruns,

Caskey, Aegean

ed. G.-H.

Age,

the Greek World,

pp. 39-75. 2000. An the Early

Copper

in the Southern

Inequality

at Feinan/Jordan," in International the Proceedings of Fifth on the the Beginnings of Conference Use Metals and of Alloys: Messages to the Metals theHistory from of

in

Is Power: "Knowledge and Social Specialization 1998.

T.

Early Bronze pp. 414-415.

b.c.

nium

of East

Uplands inNeolithic

ern Crete,"

Geochemi "An Organic of cal Investigation of the Practice on at aMinoan Site Manuring Pseira Island, Crete," Geoarchaeology

Technology from the 3rdMillen

and L. Basch, Li?ge, pp. 97-105. -. 1999. "Late Neolithic Colo nization

Carter,

Craft

"Wind-Powered

7), ed. R. Laffineur

(Aegaeum

1,

pp. 69-74. 2001.

16, pp. 223-242. Bunk, W. G. J., A. Hauptmann, S. K?lschbach, and G. Woelk.

internationale de l'Universit? ?g?enne de Station de recherches Li?ge, et sous-marines oc?anographiques avril Corse, 23-25 Calvi, (StaReSO), 1990

vol.

inMELETEMATA,

Period,"

Age,

Amsterdam. -.

437

of

a

Cambridge. Metallurgist, kcci ?oravia I. n.d. 0vra

Chavakis,

Kpijrrj?,

Athens.

E. Chernykh, I metallurgija Sofia.

N.

1978. v

Gornoe

drevnejsej

delo

Bolgarii,

Revolu J. F 1983. "Evolution, of and the Origins Complex inMinoan in Crete," Society

Cherry,

tion,

rr?c

REFERENCES

438 and Nixon

Krzyszkowska pp. 33-45.

"Polities

Polity and Socio-Political

C. E.

Conophagos,

State

in Peer

Formation,"

-.

and Palaces:

inMinoan

Problems

action

Stockholm,

antique, W.

Inter

TIenpayii?va

Cambridge, 1988.

Kpr\xoXoyiKov

of Animals

Protohistoric

in the PreEconomies

in Pastoral

Aegean," Classical

Chania,

and

in

ed. C. R. Whit

Antiquity, taker, Cambridge, pp. 6-34. J. F, J. L. Davis, A. Demitrack, Cherry, E. Mantzourani, T. F. Strasser, and

Greece,"

-.

Cherry,

and M.

AJA

Patterns Gypsum: of a Prestigious 102, pp. 391-392.

S. 1999.

"Minoan

Houses:

Huelva,

pp. 75-85. H. H. on

Roads

1972.

"Some

and Bronze,"

Copper pp. 93-104.

Creutzburg, of 52,

J. N., and G. L. Huxley, 1973. Excava eds. [1972] Kythera: tions and Studies Conducted the by Museum University of Pennsylvania and the British School at Athens, repr. Park

Ridge, 1984.

N.J. "The Minoans

ra," in The Minoan Myth

and

Reality,

and

1985,

1994. J. R., and D. J. Vaughan. Craig, Ore and Ore Microscopy Petrography, New York.

Coldstream,

-.

in Craddock

Spain,"

London.

Reflec

Working ArchAustr

1985.

nology inAntiquity (BMOP 48),

War

the Prehistoric

Tite.

N. H.

B. Rothenberg, "The Investiga

pp. 199-218. P. M. and eds. Craddock, T, J. Hughes, 1985. Furnaces and Tech Smelting Hughes

Regained," in POLEMOS: Le contexte guerrier en ? du Bronze Eg?e l'?ge (Aegaeum 19), vol. 1, ed. R. Laffineur, Li?ge,

Coghlan, tions

I. C. Freestone,

tion of a Small Heap of Silver Smelting Debris from Rio Tinto,

Archaeology, S. 1998. "The Archaeol

Chryssoulaki, and Guard

in the

Pal?orient

151-165. P. T,

N. D. Meeks,

Gale,

Oxford.

ogy of Minoan of Consumption

for the Earliest

Mediterranean,"

Craddock,

and

to Fur

"From Hearth

Smelting Technologies

26, pp.

and S. Shen J. F., C. Gamble, in Contem nan, eds. 1978. Sampling

and Roman

Evidence

Eastern

Cherry,

Stone,"

2001.

Metal

13-35.

Chlouveraki,

"The Composi

Etruscan,

nace:

1991,

British

1976.

4, pp. 93-113. Civilisations,"./^ 1995. Metal Early Mining D.C. Production, Washington,

-.

Los Angeles. E. Mantzou J. F, J. L. Davis, 1991. rani, and T. M. Whitelaw.

Times,

porary

P. T

Greek,

Cherry,

pp.

Age

tion of Copper Alloys Used by the

and E. Man J. L. Davis, Cherry, 1991. tzourani. Landscape Archaeol as Northern ogy Long-Term History: Keos in the Islands Cyclades from Earliest Settlement until Modern

Survey Methods," and Mantzourani

Bronze

bridge.

J. F,

Davis,

during

and M. E. Bailey, Palmer, pp. 93-108. P. and R. I. Courty, M.-A., Goldberg, 1989. Soils, Micromor Macphail. and Cam phology Archaeology,

Craddock,

in

A',

Peiser, T

159-176.

"The

Ivve?piov 185-193.

pp.

Oxford,

92,

AJA

in

tov ET' AieOvov?

Catastrophes

scape:A Middle Neolithic Example pp.

"The Proto

at Kavousi,"

Civilisations (BAR-IS 728), ed. B. J.

L. E. 1988. Talalay. "Archaeological in an Artifact-Rich Land Survey from Nemea,

1990.

1998. "The Soil Record Courty, M.-A. at 4000 b.p. of an Exceptional Event in the Middle in Natural East,"

of the

Economies

LeLaurium

Period

geometric

and J. F. Cherry, pp. 19-45. "Pastoralism and the

E.

D.

Coulson,

Changes

1980.

Athens.

ed. C. Renfrew

Role

107

pp.

112.

-. 1986. Some

N. Marinatos,

1983,

Thalassocracy: ed. R. H?gg and

et al. 1977.

Dall'Aglio, 1991.

P. L.,

General

Greece.

Geological Map Scale: 1:200,000,

of Athens.

Crete

Island.

and G. Marchetti.

"Settlement

Patterns

and

Structures of the Roman Agrarian in the Period of Piacenza," Territory

in Barker and Lloyd 1991, pp. 160 168. Danin,

of Kythe

N.,

A.,

1983. Hard

R. Gerson, "Weathering Limestone

and J. Garty. on Patterns and Dolomite

by Endolithic Lichens and Cyano

REFERENCES

bacteria:

Rossa

Soil Science

Soil,"

pp. 213-217. Danzer, K., D.

de

G. Thiel,

and M.

1999.

for Social Structure in LM IIIC

Evidence Supporting Contribution of Terra

for Eolian

la Calle

136,

ganic American

Trace

-.

1997.

of Inor

the Basis

1971.

-. 1973.

"N?ai

?pEDvai ?i? tt|v in oiK?av Xauai?iou," rov T'A leOvov? Kpr?ro

?XX?i\|/0?i8r| Ilenpayfi?va ?oyiKov

A',

Ivve?piov

atzi.

Guide

1989b. 2nd

uities,

rrj? E??rjviKTJ?

to Cretan

on

"Comments in Pseira

teia Building," 88.

V, pp. 79

Review,"

Tomb

7, pp. 355-416.

and Architecture,

Near

Copper

11,

0.1932.

Muhly, Palaces:

New

Rethymnon, De

of York.

Prehistory

S. Mook,

eds. 2004. Proceedings

and J. D. Crete beyond of

Eastern

New

the Crete 2000

Diamond,

Evidence

Ships, Natural

the

and

Ivve?piov 197-212.

in

A',

and F. Adams.

"Microanalytical

Metal

Bronze

Ages,

York. J. 1995. "Blueprints, Bloody and Borrowed Letters,"

104:3, pp. 16-21. History 1997. N. Dimopoulou, "Workshops in the Harbour and Craftsmen

Technology, I. Shaw,

and

J. 2001.

and Hier "History Observations Preliminary

Driessen,

on

of Minoan

Pattern

in Urbanism

in the Bronze

ed. K. Branigan, Shef 51-71. field, pp. Duke, J. A. 2002. Dr. Duke's Phyto chemical and Ethnobotanical Data

Md.

Beltsville,

http://www.ars-grin.

gov/cgi-bin/duke. F W., W. Dunning, D.

Mykura, 1982. Mineral

Slater.

2: Southeast

of Europe London.

and Deposits

Europe,

M. K., and C. J. Lister. 1983. Durkin, "The Rods of Digenes: An Ancient in Eastern Marble Crete," Quarry

Easton,

78, pp. 69-96. F 2002. Schliemanns

D.

at Troia

Excavations

from Tell

The Greek Dark

and

pp. 299-319.

Cambridge,

BSA

pp.

I., A. Adriaens,

1972.

the

Karphi:

and

Beydar," 45, pp. 579-590. Archaeometry V. 1964. The Last Desborough, Myce naeans and Their Successors, Oxford. -.

Conference, Philadelphia. L. P., and L. M. Day, Snyder. 2004. at Vronda "The House' and 4Big at

Ryck, 2003.

Istanbul.

"Neotectonic

Products,"

of Agriculture, bases, Department Research Service, Agricultural

Technology Study of Ancient Near

"Bronze

'Great House'

World

of Crete," rov Z' AieOvov?

KprjTo?oyiKov

pp. 143-152. 1943. The Tomb

Age Mining the Aegean," Nature 130, pp. 985-987. in 1935. Roman Mines Europe, Oxford.

Day,

1993.

V. Karakitsios,

1995.

nenpayp?va

round

L. P., M.

of

Lagios. Recent Deformation

15:1,

at Thebes,

Rekh-mi-re

n?cropoles

inAnatolia,

M.,

E.

D. A. 1973. "Particle Size Davidson, and Phosphate Analysis?Evidence for the Evolution of aTell,"

Davies,

Trade

Dermitzakis,

101-111.

Archaeometry N. de G. Davies,

?

Dercksen, J.G. 1996. The OldAssyrian

Philadelphia. C. J. 1979. "Some Ancient Davey, Levant Eastern Pot Bellows," pp.

Journal

"Leatherwork in Ancient

Materials Egyptian ed. P. T Nicholson

Crete,"

(?tcr?t 7), Paris. and C. Perl?s.

J.-P,

of the Bronze

Age Aegean,

"TheGreek Neolithic: A New

2004.

TheHagia Photia Cemetery I:The Groups

des

2000.

archy: the Settlement

1998. KoajuijjLiara O Neo TIpo?oTopia?

Exploration

(1921-1933) D?moule,

and P. P. Betancourt.

C,

Demargne, Mallia.

the Pla

The End

C.

Driel-Murray, and Skin

Athens. ?rjaavpo?, P. 1945. Fouilles ex?cut?es

?iOiko?

Antiq

ed., Athens.

2001.

Davaras,

Demakopoulou,

p. 387.

1993.

Princeton.

Age,

Photia, Siteias," Aghia in the and Society Aegean ed. K. Branigan, K.

R.

Drews,

Sheffield, pp. 133-149.

"ApxociOTr|T?? Kai uvr| ji??a AvaToXiKT|? Kprirnc," ArchDelt

in Aegean Archaeology, Day, Oxford.

ed. P. M.

"Pots, Labels, at

ing Complexity in Early Bronze Age Crete: The Metallurgical Finds from Poros Katsambas," in Sheffield Studies

vol. 2, pp. 275-290. and

1997,

Cemetery in Cemetery Bronze Age,

X?ov, -. 1989a.

37, B'1(1982),

and E. Kiri

People: Burying Ethnicity in the

Niko

Ayiov

E. Wilson,

1998.

Athens.

-.

D.

Age"

Day, N. Dimo and V. Kilikoglou. Forth "Lame Excuses for Emerg

poulou, coming.

1997.

-.

Kai uvr|

Kpr|Tr|?," ArchDelt

Movaeiov

Relaki.

inMochlos

Pottery,"

Smelting Valley,

R. C. P., P.M.

Doonan,

"Reassessing Specializa in Prepalatial Cretan Ceramic in Laffineur and Bet Production,"

27, B' 2 (1972), pp. 645-654. [1981].

and M.

"Old Flames,

(diss.Univ. of Sheffield).

tion

Athens,

"ApxaiOTT|T??

ji?ia AvaxoXixTi?

-.

P.M.,

Day,

ancourt

pp. 46-53. -. 1977.

-.

L. Joyner,

Neopalatial IB, pp. 13-32.

4, pp. 392-397.

l^?a^AAA

-.

P. M.,

the

Oama?

Ayia?

v?Kpoxa(p??ov

Exchange and Outlying in East Neopalatial

2003. "AP?trographie Analysis of

"IIpcoTouivcoiK?v

R. C. P. 1996.

Doonan,

and Society: Copper in the Ramsau Technology Austria the Bronze during

Crete," inH?gg 1997, pp. 219-227. Day,

C.

1997,

Slags,

"Ceramic

Settlements

Analysis," pp. 26

34. Davaras,

(diss. Univ.

Crete"

and

vol. 2, pp. 433

438.

Towns

between

31:20,

Laboratory

Betancourt

of Cambridge).

ofWine

on Variety and Organic

East

Neopalatial

in Laffineur

Katsambas,"

Approach to the Study of Pottery in

Garcia,

at Poros

of Knossos

Town

and Muhly

2004, pp. 63-79. 1991. "A P?trographie Day, P. M.

Samples According toOrigin and Grape

in Day, Mook,

Crete,"

Reichenb?cher.

"Classification

439

1870?1873

(StudiaTroica Monographien

2),

Mainz. Efe, T.,

and A.

Hash.

1997.

"Pottery

Links between theTroad and Inland Northwestern Trojan Poliochni neUEgeo Doumas

Anatolia

in

Settlement," e l'antica et? del Bronze

ed. C. G.

settentrionale, and V. La Rosa,

pp. 596-609. H. van, Effenterre, terre.

the

during

Second

1997.

van Effen

and M.

"Towards

Athens,

a

Study

of

REFERENCES

440 Words

'Villas': Modern

Neopalatial

"H Tcpoiaxopncrj

1977.

of Anthrosols

Examination

2004.

T

Eliopoulos,

R. W. Fairbridge, Sedimentation nean

A Natural

of Interrelated

"Gournia,

Prehistoric

Final Z.

Eres,

Los Angeles. Report, 2003. "Traditionelle im

chitektur

M.

and H.

Ozdogan,

Mainz,

Esin, U

-. 1966.

Dorfar in

Z. Eres,

of

coveries J. D. Neolithic

Evans,

at Knossos, 1964.

cr?toise, Amsterdam. -. 2000. "Ports

Cr?te,"

at Knossos,

1971.

Growth

Continuity

The

Knossos: PPS

Five

37,

and N. Momigliano,

2002.

Pigments Bronze Age

Theile

pp.

and P. P.

of "Analysis on from Pottery Minoan Archaeo

des

K?nigreiches

5,

Griechenland

in denJahren 1834 bis 1837 2,

in Knossos:

London,

rov H'

logical Sites," Aegean Archaeology pp. 47-54. K. G. 1841. Reise durch alle Fiedler,

Farming Pre A Labyrinth ofHistory: Papers sented inHonour Sinclair Hood, of ed. D. Evely, H. Hughes-Brock, 20.

mouillages nord de la

in Uenpayii?va

Betancourt.

"The Early Millennia: in a and Change

Settlement,"

et

la c?te

S., C. P. Swann,

Ferrence,

White

of a Settlement,"

pp. 5-117. -. 1994.

de

Kpr?ToXoyiKOV Ivve?piov Herakleion, pp. 457-469.

A', Part

and Conclusions,"

"Neolithic

cr?tois,"

AieOvov?

BSA 63, pp. 267-276. -.

"Hydronymes

Ivve?piov

16-19, Kpr?ToXoy?a pp. 30-61. -. 1989. Recherches de toponymie

in the

"Excavation

1,

Kprjro?oyia

Kpr?ToXoyiKOV pp. 150-168.

Athens, -. 1984.

minoens

BSA 59, pp. 132-240. 1957-60," -. 1968. "Knossos Neolithic, II: Summary

in?dites),"

Dis by the 4 vols., London.

Settlement

la

1966, pp. 45-78. et de la villages entre 1577 et 1629

orientale

AieOvov? A,

Cretan

Stages of the Early as Illustrated

de

sous "Eglises Cretoises 9, pp. 53-83. roches," Kprjro?oyia -. 1980. "Les mines du roi ' rov A in Llenpay^?va Minos,"

in of Aksaray, Hauptmann Turkey," et al. 1999, pp. 23-30. The Palace Evans, A. J. 1921-1935. of Minos: A Account of the Comparative Successive

"Les minerais

pp. 28-36. -. 1979.

As^kli (KizilkayaVillage), Province

Civilization

pp. 13-26. Guide to the

antique," -. 1975. "Villes

(listes

Site

Field

RA

Cr?te

Parzinger,

Neolithic

VIII,

2001.

Herakleion. Geology of Crete, P. 1964. Fonctions des cavernes

Cr?te

1999. "CopperObjects from

the Pre-Pottery

Pa.,

Cretoises, Paris.

155-173.

pp.

Sea: Laboratory,

Faure,

in

Istranca-Gebirge," in

1, Arch?ologie A?agi Pinar Eurasien 15, ed. N. Karul,

C. G.

Fassoulas,

2: The

1968-1970

in Pseira

Pseira,"

eds. 2003.

Excavations

Sitagroi: Greece,

Sedimentation

2004. "The Bedrock Geology of

of the Ex

Shrines

on the Isthmus Age in and Day, Mook,

S., and C. Renfrew,

Northeast

by Tec and Sea

Stanley, Stroudsburg, pp. 99-113. Farrand, W. R., and C. H. Steams.

Muhly 2004, pp. 81-90. E.

Controlled

ed. D. H.

Vronda

Kavousi, Kephala Vasilikis: A Triad piring Minoan of Ierapetra,"

"Quaternary in the Mediterra

Paleoclimates, in The Mediterranean

Level,"

197,

1972.

Region

tonics,

by

Science

Phosphate Analysis," pp. 1327-1333.

2002.

of the Neo Dating on lithic Settlement of Ftelia My in 2002b, konos," pp. 309 Sampson 315.

and

"Detection

I.Maniatis.

"Radiocarbon

7t?ipa|LiaTiKr) apxccio^oyiKri rcpoG?y Greek Technology: 1st International Proceedings, jThessaloniki 1997, Conference Thessaloniki, pp. 391-400. R. C.

and

G.,

Facorellis,

yiarj," Ancient

Elster,

Crafts:

Jonsered. 1997.

N.

7iaGGa^?7rr|KTr| oiko?ouikti Ttap? ?oGrj tod EXXa?iKot) xcopoi): Mia

Eidt,

2000. Minoan

Tools and Techniques 2 (SIMA 92),

pp. 9-13. Efstratiou,

R. D. G.

Evely,

forMinoan Things," inH?gg 1997,

Leipzig. Fimmen,

D.

1921. Die

nische Kultur, 1

First

Civilization Civilization

kretisch-myke

Leipzig. = The First in Europe in and the Oldest Europe

REFERENCES

in the World?Varna,

Gold

-.

Bulgaria,

Tokyo 1982. J. L. 1989. "Esse Quam Videre: A Reconsideration of the Kythnos

uk/glgsoft.

Fleming,

Physics Flemming, Tectonic Vertical

N.

in

B 64, p. 528. 1972. "Eustatic

C.

Displacement in The Mediterra Coast,"

Sargonic 39, pp. 31-43.

oke Collection

in the

Isthmus Hawes, Foxhall,

Floyd, C. R. 1998. The Plateia Building

H. A.

of Pastoralist Context ture

"The

1995.

Identification

Sites within

of Estate-Based

in Ancient

Greece,"

Agricul BSA 90,

struction,"

in

Contingent

side: Settlement,

of Construction

Recon Country and Land

Economy, in the Southern Since Argolid S. B. ed. 1700, Sutton, Stanford,

Use

-.

'Landscape Degradation' tainable Economy Debate Post-Mediaeval Southern

versus

the Sus

1999.

Metallurgy

in Antiq

-. 1972.

Studies

inAncient

Leiden. nology 9, Ford, C. E. 1981. "Parental

Tech

Liquids Intrusion Cumu

the Skaergaard lates," Nature 291:7,

pp. 21-25.

Furness,

A.

Gale,

D.

Gale,

1991.

Assemblage

"The

Bronze

P. Budd,

Pollard,

Ixer, Oxford,

N. H.,

110

pp.

Gale,

Z. A. 1985.

N. H.,

and Z. A.

Stos-Gale.

1981a.

Lead and Silver "Cycladic BSA 76, pp. 169 Metallurgy," 224. -. 1981b.

14,

Pot

48, pp. 94

244, -.

Artefact Cop

A New pp.

"Lead

Aegean,"

and Silver Scientific

Age Copper in the Mediterranean:

Approach," 11-19.

Science

in the

American

pp. 176-192. 1982. "Bronze

Sources Surface

Papastamataki, and K. Leonis.

and Copper Sources "Copper in the Bronze Metallurgy Aegean in Craddock and Hughes Age," 1985, pp. 81-93.

Amorgos,

for a Prehistoric

A.

Stos-Gale,

Metallurgy

"The Neolithic BSA

A. M.

Young, and R. A.

Ancient

tery of Knossos," 134.

of

for Late

121.

Geoarchaeology

1953.

of Copper

M.

Micromorphological

Perspective," pp. 151-189.

Leiden.

on Stra

Agricultural and Site Formation

atMarkiani,

113-120.

Isotope Charac of the Ore Deposits of and Its and Sardinia Appli to the of the Discovery

Sources

and T. M. Whitelaw.

The

Greece:

15, pp.

"Lead

Age Oxhide Ingots," inMetals in Antiquity (BAR-IS 792), ed. S.M.

"Soil Erosion,

Terracing Processes

inHalstead and Frederick 2000, pp. 95-109. Forbes, R. J. 1950.

Society C. A., French,

in the Argolid,"

Cyprus cation

Points "Melting of Ancient Slags: to the Discussion,"

Historical of the 22, pp. 49-51.

Journal

"Landscape Exploita tion via Pastoralism: Examining

OJA

1999. terisation

1988.

A Contribution

pp. 41-70. 2000b.

uity,

I. C.

S. Schmitt on Lesbos:

'Thermi

Patterns.'" -.

ick 2000, pp. 79-94. and Viscosities

in the

A Case Study of Changing Trade

and Archaeological tigraphy, Dating in Halstead and Freder Visibility,"

Freestone,

E. Pernicka,

Strecker,

in Classical

Method

in the

Trade in Trade

Mediterranean (SIMA Age N. H. Gale, ed. 90), Jonsered, pp. 197-239. on F. 1996. "Comments Bege

mann,

and J. Salmon, London, pp. 44-67. C. D., and A. Krahtopoulou. Frederick, 2000. "Deconstructing Agricultural Terraces: the Influence Examining

Econ "The Agrarian of the Ermionidha around Ethnohistorical

-.

Antiquity: ed. G. Shipley

the

pp. 325-338. 2000a. omy 1700: An

Boyd

and Culture,

Ingots:

Bronze

on the

the Earth "Feeling on Cultivation Techniques

Landscapes Environment

Age Metals

Mediterranean,"

the Excava

Li?ge. L. 1996.

inHuman

1991.

Bronze on Gournia:

Harriet by

Ivve?pwv

pp. 299-316. "Copper Oxhide

Their Origin andTheir Place in the

Pottery,"

Sites

or rov IT

Kprjro?oyiKov

Chania,

-.

of

Crete?Local

in IIe7tpay[i?va

AieOvov?

Steep Slopes inClassical Antiquity,"

Monograph

94), Philadelphia. Forbes,

Holy

"The Provenance

for EBA

Cycladic?"

1987,

"The Mount

oflerapetra

Move:

1990.

A,

of and Other

tion at Gournia

Areas in

and Wagner

pp. 55-64. N. H. 1989. "Lead

Metals

Age from the

and Marinatos

3, pp. 1-30. 1993. New Light Documents Unknown

pp. 47-65. Museum

Bronze

V.

Fotou,

of the Aegean Island Arc," Region in Science International, Diving ed. N. C. Flemming, London,

(University

Iraq

of Minoan

in Pernicka

Isotope Anal to Provenance Studies: yses Applied in A Brief Review," Proceedings of on the 25th International Symposium Y Maniatis, ed. Archaeometry, Athens, pp. 469-502.

7mS

for Ver

"Archeological tical Earth Movements

K. P. 1978.

Foster,

of Prehistoric

1991, Gale,

-.

Economy: in H?gg 11-16.

pp.

Pa., pp. Stroudsburg, N. C, N. M. G. Czarto Flemming, 1973. ryska, and P. M. Hunter. Evidence

pp. 1991. "Geo and B. Ottaway. and Arte Surface Survey physical

Austria,"

Activ

A View

et al.

in Budd

D.,

Im

and

"Commercial

"The Late

1987.

East,"

Stanley, 189-201.

Crete

Mesopotamia,"

Palace

Sedimentation

ed. D. H.

Laboratory,

"Late Cenozoic

1977.

in

ity

-.

of the

Aegean nean Sea: A Natural

B. R.

Foster, and

in the Relative

Factors

1991,

143-150.

fact Assemblage

for the Geology and plications of the Southern Geodynamics enMijnbouw Area," Geologie Aegean 451-464. 57, pp.

Pottery

to the Islamic World," Exported and Methods Nuclear Instruments Research

Austria,"

Copper Mining/Working

of Eastern

History II:

of Chinese

Versus

"Blowpipes

21, pp. 345-350. Fortuin, A. R. 1978.

"RecentApplications of PIXE Spectrometry Characterization

Gale,

per Mine,

Bellows inAncient Metallurgy,"/?i

1992.

in Archaeology

for Education

http://www.glg.ed.ac.

S. 1994.

Forenhaber,

Hoard of Early Cycladic Tools," AJA 93, pp. 31-39. S. J., and C. P. Swann.

"Software

1992. in Geology,"

Fitton,

-.

441

216,

REFERENCES

442 -.

1986. in Crete

"Oxhide

Age Metals 100. -.

Copper Ingots and the Bronze

and Cyprus

BSA

Trade,"

81, pp. 81

Glotz,

1989.

"Some Aspects

of Early

Z. A.

N. H.,

1984. Used

Hesperia N. H.,

Gale,

Stos-Gale,

"The Provenance

and J. L. of

1996. J. von, and J. Bojadziev. der "Zur absoluten chronologie Eurasia Urgeschichte," bulgarischen 105-173. 2, pp. Antiqua 1930. The W. Gowland, Metallurgy of theNon-Ferrous

143-173.

1991. S. J, and P. G. Hoare. Sediments: Quaternary P?trographie

Mo?keOov?gc,

Western

"Animal

Husbandry, in and Urbanism,"

pp.

An Island Polity: TheArchaeology of inMelos,

P. 1963.

Garelli,

Les

Assyriens

P. 1992.

inWells 1992, pp. 147-153. Gat,

1980. J. R., and M. Magaritz. in the Eastern "Climatic Variations Sea Area,"

Mediterranean

G.

IIIC

C.

1990.

Period

Haggis,

rov Z'

XoyiKov

A'

IvveSpiov

pp. 125-147. 1977. M. Gimbutas,

AieOvov?

Kx?pia in

-.

1987.

Palaces,

Cave

in an Area

of

in the Bronze

Crete

1993a. Burial

Age (diss. Univ.

"The Early Minoan Antonios Ayios

at

and

Some Problems in Early Bronze Age Chronology," SMEA 31, pp. 7

Kprfio

34. Sensa

-.

1977.

and

"Intensive Settlement

Survey, and

Patterns,

Dark Age Crete: The Case of Early

19:4,

Iron Age 174.

"Geoarchaeol

Geomorphologist

1993b. Traditional

of the Kara

Cemetery tionally novo Civilization," Expedition pp. 39-47. ogy: The

eds.

"The Kavousi

Iron Age" and Early of Minnesota).

1, Rethym

"Varna: A

Rich

B. G.

1992.

Patterns

Settlement

non,

Gladfelter,

C.

Eastern

A', Ivve?piov KprjToXoyiKov Chania, pp. 317-332. E. 1995. "UM Kai MM Giannouli,

riejtpay?eva

D.

of the

Thriphti Survey:An Analysis of

(Ierapetra),"

Kai 7i?TataSGxnuoc a\|/i?coTa Kai TCiGavq tod? GTinaG?a,"

E.Jones,

R., and N. Marinatos, H?gg, Minoan The Function of the Stockholm.

of

rov ET' AieOvov?

in ?ejtpay?eva

ed.J.

pp. 31-34. 1997. ed. The Function R., H?gg, " KMinoan Villa, Stockholm.

"The Late Minoan

at Kavousi

1986,

atBangor,

1983. Gesell,

and S. [J. A.]

O.,

Gwynedd,

Natur

67, pp. 80-87. wissenschaften Greece = Geologic Map Geologic Map of Greece 1:500,000, 2nd ed., Athens

130-132.

Island of Kythnos, Evi Archaeological in Ancient dence," Aspects of Mining a British and Acta Metallurgy: of School at Athens Centenary Conference

Cappa

of the Land,"

"Yield

in

48,

Antiquity

on the Aegean Part Two: The

doce, Paris. Garnsey,

Crete,"

1974.

MacGillivray. 1988. "AnEarly Bronze Age Copper Smelting Site

Cambridge, en

1997. NeoXi0iicr?

Athens.

Hadjianastasiou,

ed. C. Ren

Exploitation frew and M. Wagstaff, pp. 161-171.

B.

J. R. A., and P. M. Warren. Greig, "Early Bronze Age Agriculture

York.

Population

London.

Metals,

D.

Grammenos,

Methods for the Study ofUnlithified 1982.

Metal

pp. 951-963.

Gale,

C.

Chalcolithic

G?rsdorf,

Copper Alloy Artifacts," AnatSt 35,

Rocks, New

"Household

lurgy at Shiqmim, Israel,"JAS 28,

of Anatolian

Sources

K. T. 2004.

Concerning

at Irini, Keos," Ayia 53, pp. 389-406. and Z. A. Stos-Gale,

and Copper

Gamble,

?g?enne,

pp. 125-136. Golden, J.,T E. Levy, and A. Haupt mann. 2001. "Recent Discoveries

G. R. Gilmore. 1985. "AlloyTypes

pp.

civilization

Analysis inDark Age Crete," in Day, Mook, andMuhly 2004,

en las yMetalurgia Antiguas Mediterr?neas Civilizaciones y 1, ed. O Domergue, Europeas Madrid, pp. 21-37.

Lead

La

1923.

Glowacki,

Miner?a

Davis.

G.

42, pp. 519

Paris.

Cycladic Copper Metallurgy," in

Gale,

Amer Ant

Archaeology," 538.

-.

Kavousi,"yM/i

1995.

"Archaeological

6, pp.

131?

Survey

REFERENCES

at Kavousi,

Settlement

Crete:

Development and Late Minoan

inMiddle

1996a. in Eastern Roman tian

Crete

and

Horreum 'Corn Route,'"

N. Hammond,

-.

1996b.

"Archaeological Survey East Crete: Preliminary 65, pp. 373-432. Hesperia

at Kavousi,

Report," 1996c.

pp.

"The Typology

ronmental

of Form

1997, -.

in

pp. 291-299. 2000. "Coarse Ware

Age," rov H'AieOvov? A' 1, Ivve?piov

nenpayfieva KprjroZoyiKov 2002.

Crete,"

ited:

-.

2005.

1992,

Crete:

logical

Survey

Cultural

Region

Area,

1904-1905.

from

et al. 1908, p. 57. -. 1912. Excavations Crete:

pp.

in Hawes

in Eastern

Sphoungaras, Philadelphia. P. 1981. "Counting Sheep

inNeolithic and Bronze Age

"The Earliest

-.

2001. of Copper Research

in

fr?hen

Seville, Hauptmann,

"The Archaeometallurgy at Feinan, Jordan: Field and Analytical of an Work

Ore

and M.

pp. 419-436. A., H.-G.

R. Maddin.

E. Pernicka, eds.

and

1989.

Old World

(Der Anschnitt

Archaeometallurgy

Beiheft 7), Bochum. and A.,T Rehren, Hauptmann, S. Schmitt-Strecker. 2003. "Early

Bronze Age Copper Metallurgy Shar-I-Sokhta

T

inMan

and G.

St?llner,

at

(Iran) Reconsid

and J. Cierny,

Mining, G.

K?rlin,

Bochum,

ed. Steffens,

pp.

197

213. A., Hauptmann, F. Begemann, 2002.

S. Schmitt-Strecker, and A.

"Chemical

Palmieri.

Composition

the

Related

and Other 'Royal' Tomb at Finds Arslantepe,

Eastern

Anatolia,"

Pal?orient

28,

and A., G. Weisgerber, Bachmann. 1989. "Early in Oman," Metallurgy

Hauptmann, H.-G.

and

Metallurgie

in Fenan/Jordanien,

Respaldiza,

A., Hauptmann, G. A. Wagner,

Copper

in III Con District," de greso Nacional Arqueometr?a, ed. B. M. G?mez Tub?o, M. A.

Ancient

Beginnings ofMetallurgy (DerAn schnittBeiheft 9), Bochum.

in The Beginning of theUse ofMetals Zur

Kupfers Bochum.

1-30.

pp.

pp. 43-69.

119-135.

des

"Early Painted

IIIWare

1989.

in

2000.

Shipwreck of Ulubumn," BASOR

from

197

Jordan," Hauptmann, and Wagner Pernicka, 1989, -.

"Early Minoan the North Trench,"

A.

Feinan,

Crete,"

from Gournia, Crete," Pottery Transactions of the Department of Free Museum Science Archaeology, of and Art 1, pp. 191-205. -. 1908.

5, pp.

and the

and Lead Isotopy ofMetal Objects

Implica and

Periods of Copper Metallurgy

1993. A

AJA 97, pp. 265-293. E. H.

in Ar

"Sediments

Analysis,"//^

Hauptmann,

East

1978.

and A., R. Maddin, Hauptmann, M. Prange. 2002. "On the Structure

ered,"

213.

Bronze Age Chronology for Survey in the Mirabello

"Practical

Methods and chaeology: tions for Paleoenvironmental

16),

D. C, and M. S.Mook. Haggis, Coarse Wares: "The Kavousi

Griechen

161-166.

pp.

F. A.

Hassan,

I: The Archaeo

(PrehistoryMonographs Philadelphia.

die

and Crete, Amsterdam.

Oxford,

of the Kavousi

und

and G. P. J. B., H. Baxter, eds. 1999. Moss, Phytochemical 2nd ed., London. Dictionary, G. W. M. 1993. The Romans Harrison,

Revis Labyrinth "Minoan"Archaeol

Kavousi

1992.

Compo BASOR

T. Reh A., E. Pernicka, Hauptmann, eds. 1999. The ren, and ?. Yal?in,

in

Use

"Copper

1-17.

pp.

328,

Harborne,

in

Rethinking ogy, ed. Y Hamilakis, pp. 120-142.

and

Aspects of Grape-Growing and the Ideology of Greek Viticulture," inWells

A View from the Countryside in Eastern

V. D.

Hanson,

the Land

"Charting of Power in 'Minoan'

scapes

in

pp. 535-543.

Herakleion, -.

von Lavrion,

from Arad?Their Objects sition and Provenance,"

and Composition of Copper from Ingots Excavated

eds.

and Land

Landscape

and

Tin

15:2, pp. 1-80. land," Emser Hefte in the 1988. "Agriculture Hansen, J.M. versus Data Prehistoric Aegean: AJA 92, pp. 39-52. Speculation,"

Ceramic

in the Bronze

of Ierapetra

in Halstead

Effects,"

Mineralien

in theNorth Isthmus

Distribution

1992,

in Postglacial and Envi

Greece, Sheffield. Postglacial 1994. H. "Der Bergbau Hanke,

and

Implications Early Bronze Age Ceram and Betancourt ics," in Laffineur

Style

Causes

P., and C. Frederick,

2000.

of the

pp. 3-10. A., F

Hauptmann, Begemann, S. Schmitt-Strecker. 1999.

314,

Frederick 2000, pp. 110-128.

I Chalice and the

Early Minoan

"Land Use

Cultural

Halstead,

Cultural

Halstead,

in the

"Agriculture

2000. Greece:

at Kalo

"Excavations

pp. 645-681. 1997.

Hall,

107,

JHS

105-117.

-.

Khorio, East Crete," AJA 100, -.

?a change?"

Bronze Age Aegean," inWells

209.

-.

and Ancient

inMediterranean

Economy

-. 1992.

183

15, pp.

"Traditional

plus Europe: pp. 77-87.

the Egyp

OJA

1987. Rural

of a

the Role

Jordan," 94, ed. S. Demirci, Archaeometry A. M. Ozer, and G. D. Summers, Ankara,

-.

Evidence

at Feinan,

in

pp. 307

Cambridge,

New

Smelting: Copper from Excavations

339.

of Tholos

along

Past of the David Clarke, of Isaac, and

in Memory G. ed. I. Hodder,

I," in nenpay?eva

"The Port

in Pattern

Greece," Studies

Minoan

rov Z' A ie6vov? Kpr?roXoyiKov A' 1, Rethymnon, Ivve?piov 369-381. pp. -.

443

L.

Hawes,

1994. "Chalcolithic

H.

Seager,

Boyd, B. E. Williams, and E. H. Hall. 1908.

toric Sites petra, Hayden,

Bachmann,

and

Cam

pp. 34-51. bridge, Mass., C. H., and H. Boyd Hawes. Hawes, 1909. Crete: The Forerunner of Greece, London.

Gournia,

Rodr?guez,

ed. R. Maddin,

Alloys,

Vasiliki,

R. B.

and Other Prehis

on the Isthmus

ofHiera

Crete,

Philadelphia. B. J. 1995. "Rural

Settlement

of the Orientalizing through Early Classical Period: The Meseleroi

REFERENCES

444 Eastern

Valley, -.

Crete,"

Aegean

Aegean, pp. 31-68.

2, pp. 93-142. for "Evidence

Archaeology 1997.

-.

'Mega of Late Minoan

lithic Farmsteads'

1990a. Settlers?

Cadogan,

Leiden,

"Autochthons

or

Evidence

for Immigration

III through Early IronAge Date,"

at the Beginning of the Early

in La Cr?te

Bronze

pp.

ed. J. Dries

myc?nienne,

sen and A.

Farnoux,

Age jLi?va rov IT' Kov Ivve?piov

Athens,

195-204.

-.

2003a.

Vrokastro

Eastern

Area,

Crete,"

pp. 363-412. "The Final 2003b.

107, -.

-. 2004. Eastern

Area,

Neolithic

and

B. J., J. A. and Hayden, Moody, 1992. "The Vrokastro O. Rackham. Survey Project, search Design Results,"

Hesperia

Dates

1990.

in Austria,"

-.

Metal

Production

ed. P. T

Hiller,

Craddock

S,

Karanovo Hood,

1983.

Salzburg. The Minoans,

Jones, G. Greek

Tell

London.

'Country House' in Society," Krzysz and Nixon 1983, pp. 129

"The

135. 1986.

"Evidence

for Invasions

at the End in the Aegean Area in The the Early Bronze Age,"

15, pp. 267

OJA

1987.

of

End of theEarly Bronze Age in the

Practice

"Agricultural BSA Prehistory,"

in 115

82, pp.

123. 1992.

"Weed

Phytosociology

and Crop Husbandry: Identifying a Contrast

and Minoan kowska

in Neolithic

Change

Greece,"

Modern

1988,

S. 1971.

-.

through and J. Lang, 1988.

Pompeian

Austin.

1999. "Chronology of Greece in the Final and South-East Europe Neolithic and Early Bronze Age,"

-.

and the Ages,

pp. 69-75. and V. Nikolov.

London,

Day,

PPS 65, pp. 319-336.

Smelt

in Mining

F. 1999. A

W.

and Economic

ing Site in theMitterberg Mining Area

to the Present

Southern

and Hughes

Age

and Runnels, 1994. A Greek Coun

295.

32, 11," Archaeometry pp. 211-237. K. T. M., and J. E. Ericson. Hegde, 1985. "Ancient Indian Copper in Craddock

Prehistoric

1996. "Water, Animals, Johnson, M. and Agricultural Technology: A Study of Settlement Patterns

C. R.

Dateiist

1985, pp. 59-67. H. 2003. "Bronze Herdits,

1962.

pp. 135-146. M. H., C. N.

Herbal,

fordAMS System: Archaeometry

Furnaces,"

Baltimore.

Jashemski,

the Ox

from

R. W.

Crete,

Prehistory Stanford.

R. A. Housley, J. van Klinken.

and G.

"Radiocarbon

Hutchinson,

tryside:The SouthernArgolidfrom

and Preliminary 61, pp. 293

and Metal J. F 1978. Mining in the Greek and Roman World, lurgy London.

Bronk,

1981. Archae Smyth. Knossos Area, of the

Survey

Jameson, T. van Andel.

Re

Healy,

R. E. M.,

S., and D.

1992,

353.

Hedges,

2,

1992. "Agricultural Jameson, M. H. in Ancient inWells Labor Greece,"

and

Related Studies, Philadelphia.

1986-1989:

Studies

R. 1983. "The Limi Simpson, in tations of Surface Surveys," Keller and Rupp 1983, pp. 45-47.

Settlement

2: The

in Crete

"Settlers

Hope 3,

Archaeometry

on the Vrokastro

Reports Crete

pp. 367

Chania,

b.c.," Cretan

ological London.

Mi

Area,

Vrokastro Area of the

History

Hood,

Crete," Mediterra

rabello, Eastern nean Archaeology pp. 31-44.

Kpr\roXoyi

151-158.

pp.

of the Vrokastro

History

in Tlenpay

AieQvov? A',

-. 1990b. c. 3000

AJA

I/IIA Settlement

EarlyMinoan

in Crete,"

375.

"Final Neolithic-Early in the I-IIA Settlement

Minoan

-.

ed. G.

Professor

between Practice," van Zeist

Palaeobotany ed. J. P. Pals, M. pp.

and

Festschrift

(Review

for

of 73),

Palynology

and

J. Buurman,

van der Veen,

and

Ancient in

Amsterdam,

133-143.

Jones, G., A. and J. G.

Bogaard, Hodgson.

M.

Charles,

2000.

"Dis

tinguishing the Effects of Agri to Fer Practices Relating A Func tility and Disturbance: in tional Ecological Approach

cultural

REFERENCES

1073

27, pp.

Archaeobotany,"%/^? 1084.

P. Halstead, Jones, G., A. Bogaard, 1999. M. Charles, and H. Smith.

"Identifying the Intensity of Crop on

Husbandry Floras, BSA Jones,

167-189.

94, pp. I. Smith.

and

G.,

"The Plant

ofWeed

the Basis

ingsAF South andAF North," in 1978.

B.

"The Oldest

Cop

perMetallurgy in the Balkans: A Study of theDiffusion of Copper to Southeastern

from Asia Minor

-.

21:1, pp. 9-17. Europe," Expedition 1995. of the "Continuity in the Central Prehistoric Mining in Ancient

and Mining in Southeast ed. Metallurgy Europe, B. Jovanovic, 29-35. pp. Belgrade, Balkans,"

Juleff, G. 1998. Early Iron and Steel in Sri Lanka, Kanta,

A.

The Late Minoan

in Crete: A

III

of Sites,

Survey

Pottery and theirDistribution (SIMA 58), G?teborg. -.

1983. Crete:

"Minoan

Some

in

1983,

pp.

and Traditional

Parallels

two Cultures ment,"

between

in the Same

Krzyszkowska 155-162.

Environ and Nixon

12th Century b.c., ed.W. A. Ward and M. Joukowsky, Dubuque, pp. 79-86. E. 1996. Le Bronze Ancien Karantzali, Cyclades

et en Cr?te

(BAR-IS

631), Oxford.

Basel. 1997.

"Copper

Alloys

and

Cycladic Metallurgy during the Bronze

Age,"

Ancient

Greek Tech 1st International

nology: Proceedings, Thessaloniki Conference, saloniki, pp. 73-83. Katsarou,

S., and D. U

"Emerging Neolithic

Thes

1997,

2004

Schilardi.

and Early

in Paros: Settlements Cycladic and Sklavouna," Koukounaries BSA Keller,

99, pp. 23-48. D. R., and D. W.

1983.

Archaeological Area Mediterranean Oxford.

in

Kraus,

pp. 29-42. P. 1986. "Melting of Ancient Viscosities

Bristol. eds. R., and P. P. Betancourt, TEXNH: Craftsmen, Crafts women and in the Craftsmanship Bronze 16), Aegean Age (Aegaeum 2 vols., Li?ge. 1929. Greek and Roman Lamb, W. Laffineur,

1997.

Survey (BAR-IS

155),

Princeton

at Thermi

1936. Excavations Lesbos,

Surface

2001.

of FeO-Fe

in

Cambridge.

Partial and Materials

2001.

'Schatz

Schatz

"Neue

des Priamos':

von Troia,

A

seine

und

ungsort Troia-Traum

tion B257.

32B,

Leach, in

tural pp.

pp. 373-383.

ed. T. A. Wer

Slags

Cyprus,

Hungarian Krahtopoulou, Terraces Greece,"

R. Maddin, Nicosia, ed. 1994.

b.c., ed. J. D. and V. Kara pp. 117-141. Treasures the of

Bronze A.

Age, Budapest. 1997. "Agricultural

in Livadi, in

Thessaly,

Archaeological

Sciences

Central Coastal

Horticul

Systems?A

Perspective," 135-148.

Andes:

71,

Antiquity

"The Production in the

Alloys

Highland

Ores

and

18, pp. 43

Smelters?"/^

76. -.

1996.

"Arsenic

Bronze:

Dirty

Copper or Chosen Alloy? A View

Washing

of Cyprus," inEarlyMetallurgy 4000-500

Copenhagen. "The Terminology

1997.

H. 1991. Lechtman, of Copper-Arsenic

F. L., and A. 1982a. Koucky, Steinberg. "Ancient Mining and Mineral on in Dressing Cyprus," Early Pyro The Evolution technology: of the First

ton, D.C., pp. 149-180. -. 1982b. "The Ancient

its Colonies,

H. M.

Production

Stuttgart,

Industries, Fire-Using time and S. F Wertime,

and

of Agricultural Origins and Food

sein Auffind

und Wirklichkeit,

Korfmann,

Cara

1976. The OldAssyrian City State

Aspekte Der

Datierung,"

ed., sec

1967. Old Larsen, M.T Assyrian van Procedures, Istanbul.

-.

pp. 583-592. M. Korfmann, zum

Transactions

of

and 58th Physics, Chemistry ed. R. C. Weast, Cleveland,

Metallur

Pressures,"

in CRC Handbook

Elements," -

O

SiO-CaO Slag Containing ?LO. MgO and Cu O at Intermediate

georghis,

eds.

London.

Bronzes, -.

The Late

F., and A. Yazawa.

Kongoli,

Muhly,

in the

and

Lang, G. 1978. "Density of Liquid

Kov?cs,T, Rupp,

Points

theCambridge Colloquium 1981,

on

1992.

gical

ein Organ der

"'Kamm,'

Brussels,

the South Coast Crete. of 1:1 = J.W.Shaw and M. C. Shaw, eds. The Kommos Region, and Minoan Industries, Ecology, Princeton 1995. II = P. P. Betancourt, The Final

Pottery,

Slater,

pp. 249

Slags: A Dis Historical cussion," Journal of Society 20:1, pp. 43-45. Metallurgy eds. O., and L. Nixon, Krzyszkowska, 1983. Minoan Society: Proceedings of

N.Y Newburgh, = Kommos: An Excavation Kommos

Age

1982.

Kresten,

Instmments Kollmorgen Corporation. Soil Color Charts, 1992. Munsell

Bronze

E.

Oxford,

Selbstverwaltung in Les pou Zeit," altbabylonischen et dans voirs locaux en M?sopotamie les ed. A. Finet, r?gions adjacentes,

639.

1990. Princeton Pottery, III = L. V. Watrous,

F. R.

Sinclair,

st?dtischer

a "Assessing Polity Crete: The Malia

Protopalatial

ed. M.

Konstitution und 1958-1981. Karrer, W. der Vorkommen organischen Planzen stojfe, Katsa, L.

of

"The Architecture

the Late Minoan IIIC Shrine (Building G) atVronda, Kavousi," inDay, Mook, andMuhly 2004,

Oxygen

The

les

L. 2004.

"Liquidus

V. 1992. "The Crisis Karageorghis, in The Crisis Years: Years: Cyprus,"

dans

ed. A. (1995), and J. Gowlett, 259.

Middle Minoan III Neolithic through

Mainz.

1980.

Period

N.

Late

1977.

Princeton Klein,

Kephala:A and Cemetery,

Lasithi State,"AJA 103, pp. 615

Pseira X. Jovanovic,

I = J. E. Coleman, Settlement Neolithic

Keos

pp. 91-101. C. 1999. Knappe?,

Forthcoming. from Build

Remains

445

in

from the Americas,"/i?i pp. 477-514. and S. Klein. Lechtman, H., "The Production

23, 1999.

of Copper-Arsenic

Alloys (Arsenical Bronze) by Co smelting: Modern Experiment, Ancient 26, pp. 497 Practice,"./^ 526. Lee, W.

E. 2001.

Archaeological and Change

"Pylos Regional Part Project, the Human

IV,

Landscape

REFERENCES

446 in aModern

Greek

inMes

Village

2 vols.,

Rome.

at Kastelli Crete,"

Country

inWells

and "Agriculture in Classical Athens,"

Lyon,T. The Nature

and H. and

O.

York.

D. Mabberley, A Portable

The Plant-Book:

J. 1997.

C.

the Vascular

Dictionary of 2nd ed., Cambridge. J. A. 1997. "The Cretan

Plants, MacGillivray,

L. H. Sackett, J.A, and S. Hemingway.

Driessen,

at Palaikastro

"Excavations

J.M. 1992.

1991,"

and C. Guilford. Rocks Igneous New York.

of

Textures,

Prehistoric Artefacts

-.

and P. V.

H.,

Composition

W.

My

p. 237. E. Petrascheck.

London,

Slater,

P., and W.

Research 4:1), Athens. Geophysical E. G. 1961. An Outline Mariolopoulos, the Climate of Greece, Athens. of F. and G. Scan Matthiae, P., Pinnock, d?lia

eds. civilt?

1995. Ebla:Alle

urbana:

Trenfanni

di scavi in Siria delVUniversit? di Maxwell,

" Milan. Sapienza, V. 2002. "Metalworking "La

in Sampson

Ftelia,"

2002b,

at 147

pp.

149.

of

Greek

V.

McGeehan-Liritzis, and

Development the Late Neolithic

1996.

The Role

in ofMetallurgy and Bronze Early

Age ofGreece (SIMA-PB 122), Jonsered.

"On

the Chemical

of Prehistoric

Composition Artefacts Copper-Based BSA Mainland Greece,"

V.,

McGeehan-Liritzis, Greek

from 94, pp. 81

100.

Gale. Isotope

1994.

Age (Sheffield Archaeological Mon and 8), ed. C. Mathers ographs S. Stoddart, Sheffield, pp. 221-270.

1988.

"Chemical

Analyses

and N. H. and Lead

of Greek

Late

and Early Bronze Age

Neolithic Metals,"

"The Emergence Manning, of and Divergence: Development on Bronze and Decline Age Crete in and the Cyclades," Development Decline in theMediterranean Bronze S.W.

and D. G.

Roma

Copper-Based from Crete," BSA 93,

pp. 91-102. 1999.

Deposits of Europe ed. F W. Dunning,

origini

1998.

Ioannou.

inMin

"Greece,"

2: Southeast

done Matthiae,

in Bago

and T. S. Maddin, R., J. D. Muhly, 1999. "Early Metalworking Stech. et al. at in Cay?n?," Hauptmann 1999, pp. 37-44. the Chemical

12,

1956. Lavrion (Geological and

lini and Lo Schiavo 1996, pp. 9-15.

Mangou, "On

1982.

G.

kura,

"Archaeological

tion of Early Metallurgy,"

ArchDelt

YleSi?doc;," 102-141.

Europe,

"Some Metallurgical in the Reconstruc

Considerations

sur les

"?tude

33, pp. 377-383. S. 1929. "np xouivco?KO? x??poc rcap? to %cop?ov

GoWco?

eral

Field Survey in the Vicinity of Monte Irsi, 1996-1997," AJA 102, pp. 416-417. R. 1996. Maddin,

179-193.

pp. 1958.

KpaGi

Marinos, 1998.

M.

MacKinnon,

19%2. Atlas and Their

Industries Marinatos,

Marinos,

F, C. H. Donaldson,

W.

J.-R.

pp.

BSA 87, pp. 121-152. Mackenzie,

Isik, Bonn,

Mar?chal,

des cuivres propri?t?s m?caniques ? l'arsenic," M?taux Corrosion

Countryside in the Old Palace Period," inH?gg 1997, pp. 21-25. MacGillivray,

zur in Studien und Europa," Religion Kultur Kleinasiens und des ?g?ischen

Bereiches:Festschriftf?r Baki Og?n (AsiaMinor Studien 39), ed.

of Soils,

Properties

in Eastern

87, pp. 103-120. "Das ?g?ische Chalko

lithikum und das erste Silber in

1929.

Buckman.

Kaukasien," aus Iran

Phournis

BSA

J. 2000.

Maran,

1992, pp. 29-56.

L.,

New

1992.

'K?nigs

Manteli, K. 1992. "TheNeolithic Well

240.

Life

nach

Beziehunger

in Barker and Lloyd 1991, pp. 233 H.

"Die

gr?ber' von Alaca H?y?k und ihre Arch?ologische Mitteilungen und Turan 33, pp. 19-61.

Settle J. 1991. "Forms of Rural Lloyd, ment in the Roman Early Empire,"

Lohman,

2001.

G.

Mansfeld,

70, pp. 49-98. senia," Hesperia e la civilta minoica, Levi, D. 1976. Festes

Archaeometry

30, pp.

199

225. R. 1982. Trees and Timber Meiggs, the Ancient Mediterranean World,

in

Oxford. Melas,

M.

Minoan

1999.

"The Ethnography and Mycenaean Beekeep inMELETEMATA, vol. 2,

ing," pp. 486-491.

of

REFERENCES

= P. P. Betancourt,

MELETEMATA

MATA.

IB = K. A

R. Laffineur,

V. Karageorghis, W.-D. Niemeier.,

and

Enters His 65th Year (Aegaeum20),

Mellink, M. J. 1956. "The Royal Tombs atAlaca H?y?k and theAegean in The Aegean

World,"

and

theNear

Goldman, Festschrift Hetty ed. S. S. Locust Weinberg, Valley, N.Y., pp. 463-479. 1986. inWest

"The Early

Bronze

Anatolia:

and C. ?lu?re, eds. du m?tal, Paris.

Mohen,

and P.Walter.

Age

Aegean in The End Correlations," in the the Early Bronze Age Aegean, ed. G. Cadogan, Leiden, pp. 139

de

1990.

struction

Evidence

Copper in The

2000.

pp. 78

J., and B. Rothenberg.

Merkel,

in theWestern

Arabah,"

A.,

4th

Meteorology, Milton,

C,

Finkelman, "Slag torical

pp.

an Ancient

Khania

Metallurgy

Society

Katsanos.

Dimitra's

Metallic

Grammenos 1997, = Excavations Mochlos

Samples," pp. 91-94. atMochlos,

of

in

Culture

IA = J. S. Soles and C. Davaras, eds. Period III. Settlement Neopalatial on the Coast: The Artisans' Quarter at Chalinomouri. and the Farmhouse

7),

-.

C,

Haquet, Evidence

at Neolithic

of Cotton Pakistan:

of Analysis from a Copper

Fibres

in

Karpatenbeckens von Hajd?s?mson

Depotfundhorizonte und Kosziderpad

Ids, Budapest. as

Palace

The Distribution

H?gg

Trade

-.

Landscape,"

Tin,"

-. and

for

in Bronze

in The

Search for Ancient Franklin, J. S. Olin,

pp. 43-48. 1985a.

"Beyond

and

Typology:

in itsHistorical

in Contributions

gean Archeology:

Age Tin,

D.C.,

Washington,

Aegean Metallurgy Context,"

1990.

in

"New Evidence

T. A. Wertime,

Md.,

Walls

Hamden,

Age,

of and Trade

ed. A. D.

Kardulias

Climate

and Enclosure

1978. Sources

on the

in Halstead

in the Bronze

Conn.

Studies

Lanham,

Re ofMineral Metals theNature the of

sources and

and

and A. T. Grove.

in the Cretan

des

Bronzefunde

Muhly, J.D. 1973. Copper and Tin:

ed. P. N.

"Holocene

in Crete,"

1967.

A.

Mozsolics,

Neo

Frederick 2000, pp. 52-61. J.A, Moody, "Terraces

M. J.-F Tengberg, and B. Mille. 2002. "First

Moulherat,

III"

Aegean Strategies: and Environment

pp. 61-79. 2000. Change

Minoan

from

A.

Mehrgarh, Mineralized

or Just Misunder

European Fringe, and M. T. Shutes,

Crete.

The Sites (PrehistoryMonographs Philadelphia 2003.

in

and Archaeol

ser., 12:6, pp. 479-582. -. 1910. The Dawn Mediter of ranean Civilisation, London.

pp. 235-240. "The Cretan Environ

Abused

stood?"

Soil Science Test Cases

5th

1987,

1997. ment:

K. Nigge, and E., U. Zwicker, 1997. and "Analysis of Examination Metallographic

Mirtsou,

A. A.

149-271.

Crete:

"The Minoan

-.

33.

Crete,

of Tennessee,

Philadelphia. 1908. "Le armi piu antiche di ramo e di bronzo," MemLinc,

Mosso,

"The Environmen

Artifact,"

Prestige Marinatos

of Eastern

Crete,

of Minnesota).

-.1987b.

10, pp. 24

and

Cambridge,

(diss. Univ.

ogy: Three

1935-1936.

ofWest

Region

a

Metals of

Bead" JAS 29, pp. 1393-1401. 1987a.

(diss. Univ.

His of the

Sites

Knoxville). -. 2002.

Ivve?piov

lithic through Late Minoan

at

Smelter

rov IT

tal and Cultural Prehistory of the

1980.

"Early Metal in The

pp. 28-33. M. W. 1994.

ological Greece"

102-109. J. A.

Moody,

ed., Columbus.

Israel," Journal

Develop

1902,1905:

B.

M.

S. 1988.

"A Pedological of Catchment Basins Investigation below Late Minoan Period Archae

Morris,

Pendlebury, Pendlebury, and Money-Coutts 1935-1936,

E. J. Diwornic, R. B. and P. Toulmin. 1976.

from

Timna,

Mass.,

in

"Metal,"

ed. A. E. Pernicka, Hauptmann, T Rehern, and U. Yal?in, Bochum, pp. 149-165. and R. A. Anthes.

"Le

and

Money-Coutts,

in The Beginnings ofMetallurgy,

P. R.

in Mesopotamia," the Use Beginnings of ed. R. Maddin, Alloys,

Prepalatial Protopalatial from the Room of the Jars Deposits in the Royal Stores," BSA Pottery 95, pp. 65-105.

1999.

1990.

Haggis. Survey,

lurgy

in?dite

86, pp.

"Knossos

The

"The Earliest Steps to Copper Metallurgy

BSA

Reassessment," -.

London,

"The

C.

Kavousi-Thriphti

1988-1989,"4/^94,p.323. Moorey,

A', Chania, pp. 477-478. 1991. "MM IA from Pottery at Knossos: Evans' Excavations A

-.

of Copper,

ed. B. Rothenberg, 122.

1990.

pp. 273-297. M. S., and D.

"The

europ?en,"

KprjToXoyiKov

AieBvov?

Recon

from Timna," Metallurgy

invention

in Flenpayiieva

Knossos,"

Smelting Based on Archaeological Ancient

Mook,

ment of the Footed Goblet ('Egg Cup') from EM II toMM III at

of Exper "Summary for Late Bronze

"Experimental of Bronze Age

ogy of Prehistoric NW Crete," JFA 23,

103-110.

pp. N.

Momigliano,

-.

une

du Bronze

l'Age

Technel,

Age Copper Smelting and Refin ing,"MASCA]2:6, pp. 173-178.

Miller,

J.-P,

four-creuset,

of

152. J. F. 1983. imental Results

and G. Rapp J., O. Rackham, 1996. "Environmental Archaeol Jr.

Moody,

1991.

1994.

183-191.

pp.

8), Phila

Mohen,

J.-R, D?couverte

and

Asiatic

Merkel,

Mediterranean Landscape, ed. S. Bottema, G. Entjes-Nieborg, van Zeist, Rotterdam, and W.

The Small Finds (PrehistoryMono graphs 9), Philadelphia 2004.

East:

-.

Eastern

IC = J. S. Soles and C. Davaras, eds. Period III. Settlement Neopalatial on the Coast: The Artisans' Quarter at Chalinomouri. and the Farmhouse

London.

CatalH?y?k,

Set

Neopalatial on the Coast: The Artisans'

(PrehistoryMonographs delphia 2003.

1999.

Li?ge 1967.

MellaartJ.

III

at Chali and the Farmhouse Quarter nomouri. The Neopalatial Pottery

in Aegean Archaeology asHe H Wiener

toMalcolm

3 vols.,

Mans Role in the Shaping of the

and T. M.

Barnard

Period

Brogan, tlement

eds. MELETE

Studies

Presented

447

Studies

toAe inHonor

of

William A. McDonald (Publications inAncient Studies 1), ed.N. Wilkie

REFERENCES

448 E. Coulson,

and W.

D.

lis, pp.

109-141.

-.

of Tin

of Bronze

and

the

1989.

Metallhandwerk

"Cay?n? in

World,"

and

Tepesi

the

1991.

Earliest

1

pp.

in

the

"Copper Cyprus: inMohen and

Ware

and

"Early Bronze

the Taurus,"

97, pp. 239

4//4

in the Aegean," Schiavo 1996,

Metals and Lo -.

"The First Use

Archaeometallurgy," pp. 771-773. -.

"Copper and the Eastern in The

ranean,"

in

Mediter

Archaeometallurgy

of

theAsian Old World (MASCAP16), ed. V. C. Pigott, pp. 15-25.

Philadelphia,

"Early Metallurgy in Anato and Cyprus," lian Metal 2 (Der Anschnitt Bei 15),

ed. U. Yal?in,

pp. 77-82. 2004a.

pp. 1-6. 2004b. "Chrysokamino

-.

Beginnings and Crete

Mook,

Look

at

8, pp. 1997.

"The Role

and

the North

of Metal in the

and

Technology in Day,

K.

the on

andMuhly 2004, pp. 283

L.

1992. J. D., and E. Pernicka. and "Early Trojan Metallurgy inHeinrich Metals Schlie Trade," mann: und Grundlagen Ergebnisse 100 Jahre nach moderner Arch?ologie Schliemanns Tod, ed. J. Herrmann, pp. 309-318.

Muhly, J.D, andT. Stech. 2003. "The Metallurgy of Ninevite 5," in The Origins ofNorth Mesopotamian Ninevite 5 Civilization: Chronology,

Belgium, G. E. 1977.

in Soil

Science,

Kirk.

Journal

of Petrology

I. A.

1936. Zvorykin. of Red Soils of Soil Science

1997. 'Villa

41,

of "The Origins in System," H?gg

pp. 15-19. 1987. "Neo-Palatial

Minoan

Outlying of the the Function

and

in H?gg

Palaces,"

Marinatos

and

pp. 95-98. Noble, J. V. 1988. The Techniques rev. ed., of Painted Attic Pottery, London. 1987,

E.

Nodarou, ments

1998.

from

"Soils

at

Chrysokamino, A Geoarchaeological (master's

and Sedi

the Habitation East

thesis, Univ.

Site Crete:

Study" of Sheffield).

1982.

Noll, W.

und "Mineralogie der keramiken Altkretas,"

Technik

Turn

Neues Jahrbuchf?r Mineralogie 143, pp.

Sus

ceptibility of the Soil and Its Signif icance

Athens,

9),

pp. 417-428. "Magnetic

ed. C. G.

ofMagmatic

Greece,"

Settlements

Muhly,

in

Metallurgy," et? del bronzo

and C. K. Van

I., and

the Minoan 1997,

in the Devel

and V. La Rosa,

pp. 397-421. W.-D. Niemeier,

Nixon,

of Poliochni

Aegean

"Investigations

Aegean,"

Economy, Society (Subartu and H. Weiss, ed. E. Rova

in

A,"

Metallurgy,"

settentrionale,

nelVEgeo Doumas

Attica,

289.

Berlin,

Aegean 1-32.

of Aegean e Vantica

opment Poliochni

Nevros, and

Renfrew 2003, BMCR, June 21, 2004,

New

Crystallization," 19, pp. 66-94.

Bochum, of Elster

Review

and P. P.

"Appendix

JMA

1978. "AModel

Greece

heft

pp. 51

143-153.

pp. 634-648. H. D., Nathan,

in

2002.

Ceramics,

London,

Nakou, G. 1995. "The Cutting Edge: A

101,

and Bronze

inAncient

1995.

I, pp.

-.

1999. Cypms

Pseira

in

AJA

of the

K. G. Mclntosh,

G. H., Myer, Betancourt.

of

Bagolini pp. 75-84. in "Recent Works

1997.

of Vasilike

the Production

55.

253. -. 1996.

and

Studies Scientific ed. M. J. Hughes,

Tin

Age

p. 6.

1979,

and P. P. Betancourt.

Mottled Colours of the Slip," in

Phase," 1993.

Mullen,

G. H.,

Myer,

?lu?re 1991, pp. 357-374. -.

hout,

in Betancourt

Sections,"

1989,

Powder

"X-Ray

1981. "The Composition

-.

-.

1979.

Diffraction and P?trographie Thin

11.

-.

G. H.

Myer,

in the

Hauptmann,

and Wagner

Pernicka,

(Offa-B?cher 75),

Neum?nster.

Beginnings ofMetallurgy Ancient

28, pp. 223

1994. Altanatolisches

A

M?ller-Karpe,

Metallurgy,"

AJA 89, pp. 275-291. -.

Soil Science

of

246.

"Sources

1985b. Beginnings

Journal

Minneapo

a Review,"

150-199.

Noll, W.,

R. Holm,

"Chemie tischer

and L. Born.

und Techniken Vasenmalerie

1971.

altkre vom

Kamares

REFERENCES

Die Naturwissenschaften pp. 615-618. Northover, J. P. 1989. "Properties Use

of Arsenic-Copper

1989,

pp. K. 1996a.

Nowicki,

"Report X: Studies

in

1993 and 1994,"Archeologia 46 (1995), pp. 63-70. -. 1996b. Hundred

"Lasithi of

Years

Research,"

pp.

Archaeological

Aegean

pp. 27-47.

-.

2002. "AMiddle Minoan Deposit at the Refuge Site of Monastiraki

II

Archaeology 5 (2001), pp. 27-45. -.

2004.

of Kavousi,

"South

Siteia

at the End

of the

Mountains Bronze

in Day, Mook,

Age," 1996.

W.

O'Brien,

Bronze

in Britain

Mining

and

Age

Copper

Buckinghamshire. E. E. 1993.

"New Evidence

for

Iron Smelting in South-East in The Nigeria," Archaeology of Early

ed. T.

Africa, B. Aandah, don, Olsen,

and A. Okpoko, Lon pp. 432-448. F L. 1973. The Kiln Book, Bas

se?, Calif.

B. S. 2001.

Ottaway,

"Innovation,

in Specialization Prehistoric Metal Early Copper lurgy," EJA 4:1, pp. 87-112. 1999. and A. ?zdogan. M., ?zdogan, "Archaeological Early Metallurgy in Hauptmann 22. 1999.

R.

on the

Evidence at

Cay?n?

et al. 1999, "Perishable

Tepesi," pp. 13

Goods

Palmer, in in Floreant Texts," Mycenaean Studia Mycenaea. Akten des 10. Inter nationalen

Mykenologischen Kolloqui ums in vom 1.-5. Mai 1995, Salzburg ed. S. Deger-Jalkotzy, S. Hiller, O.

Panagl,

Vienna,

-.

and G.

Nightingale,

pp. 463-485.

torical

the Mycenaean Period,"

in

"Craft

Anatolia,"

lithic

Rome,

Revealing

Ancient of Athens.

G.,

Papathanassopoulos, Culture Neolithic

V. N. 2001. Activity pounds Insects

E. A., A.

of Volatile Isolated 3:

to the His Prehistory

and

"On

ochni

Pare,

C. 2000.

"Bronze

and

the Bronze

OJA

9, pp. 263

Eurasia

3, pp. 41-180. Antiqua S. Schmitt E., F. Begemann, 1993. and G. A. Wagner. Strecker, and Early Bronze Age from the Balkans

"Eneolithic

Artefacts

Copper and their Relation PZ

68, pp.

to Serbian

Ores,"

1-54.

C. Seeliger, G. A. E.,T. F. S. Schmitt Wagner, Begemann, S. Eibner, ?. ?ztunali, Strecker, and I. Baranyl. 1984. "Arch?ometal

Pernicka,

lurgische

Untersuchungen

Nordwestanatolien,"

and

Compounds

on Lemnos,"

Pernicka,

Khalilov.

26:6, pp. 584-586.

and Prove the Composition of Metal from Poli Artefacts

S. Schmitt E., F Begemann, and I. Kuleff. Strecker, H. Todorova, 1997. "Prehistoric in Copper Bulga ria: Its and Provenance," Composition

ed. 1996.

podagraria,

Metallob

pp. 607-714.

Pernicka,

JRGZM

in 31,

pp. 533-599. K. Assi E., G. A. Wagner, F. Begemann, C. Doumas, and W. Todt. 1983. "An Analyti

Pernicka, "

ofAegopodium Natural Chemistry of

Unter

zur Herkunftsbestim

298.

Chromatography-Mass Com of Volatile

pounds

Geochemische

arch?ologischer

nance

in

Plants

in

in

S. Schmitt E., F. Begemann, A. P. 1990. and Grimanis. Strecker,

and Biological Com Organic

from

York.

Pernicka,

Z. Khalilova,

"Identification

the Role

"Erzlagerst?tten

jekte,"/?GZM34,

Metallur

and L. M.

1987.

E.

suchungen

in Greece, Athens.

Odinokov,

New

Civilization,

mung

gical Technology, J., et al. 1959. Geological Papastamatiou, Greece: Athens. Map of Ierapetra, G. 1981. Neolithic Papathanassopoulos, trans. and Cycladic Civilization, A. Doumas, Athens.

Paramonov,

J. 1989. A Forest Journey:

Altertum.

A. 1986. A Study of Papastamataki, Ancient Slags and their Contribution to the

Society Sheffield,

der?g?is und ihreAusbeutung

1207

2, pp.

ed. P. Halstead,

ofWood in theDevelopment of

rr|? MaKe?ovia?,"

MocKe?ovia

Specialization of Greece," inNeolithic

108-120.

pp. Perlin,

1999.

in the Neo

in Greece,

A., and A. Papasteriou. Papanthimou, 1993. "O TtpO?GTOplKO? OIKIGUO? GTOMav?a?o: Nea GToi%e?a gttjv

Apxocia 1216.

of Trapeza,"

Cave

36, pp. 5-131. and K. D. Vitelli. Perl?s, C,

pp. 713-733.

Athens,

I:The

BSA

Spectrometry

2001. "Bridging the Gap: The Continuity of Greek Agriculture from

Facorellis,

The Archaeol

J. D. S., H. W. Pendlebury, B. Money-Coutts. 1935 in the Plain of "Excavations

Lasithi

G. 2001. "Simulation Papadimitriou, of Ancient Their Bronzes: Study Mechanical and Metalworking in Issues Properties," Archaeometry in Greek and Prehistory Antiquity, and ed. Y. Bassiakos, E. Aloupi,

and

Production,

and E. Cher

Pernicka,

rcpo?GTOpia

in the S. 1999. Eden Oppenheimer, East: The Drowned Continent of Southeast Asia, London.

J. D. S. 1939. London.

Pendlebury, and M.

Ana Archaeologica Alba Palmieri tolica etMesopotamica ed. M. Frangipane, Dedicata,

Y

P. Sinclair,

Shaw,

at

ogy of Crete,

pp. 573-599.

and Ireland,

Okafor,

pp.

Rome,

H. Hauptmann, M. Liverani, P. Matthiae, and M. Mellink,

Muhly 2004, pp. 265-280.

Age Europe. Held

Conference

1-38.

Pendlebury,

and Over

the Rivers

a

ed. C. F. E. Pare, Oxford,

1997,

"From Arslantepe to Arsenical Copper in Eastern

of

Proceedings

Go

and Circulation

theUniversity ofBirmingham inJune

Anal

theMountains:

East

ofMochlos: The West

1993. nykh. Metalwork

Supply in Bronze

1936. K. Sertok,

A. M.,

Technology in Between

Katalimata,"^^?

2000.

and F. Pinnock,

Peyronel, 1259-1272.

Palmieri,

3,

Archaeology

and A. Hauptmann. from Ebla: Chemical

A.,

the World

Make

The

Metals of

from the Artefacts yses of Metal in Proceed and Iron Bronze Ages," Con the First International ings of on the the Ancient gress of Archaeology Near East, ed. P. Matthiae, A. Enea, L.

One

(Crete):

Palmieri,

"Metals

on Investi

in Greece

gations

in

inMetals

Age," Round:

History:

and

Alloys," and Wagner

Pernicka, 111-118.

Hauptmann,

Class and Political Ethnicity, Mon ed. D. W. Tandy, Economy, treal, pp. 41-84.

58,

Typ,"

449

menos,

cal Study of Prehistoric Lead and Silver Objects in Proceedings

from

the Aegean," Interna

22nd of the

REFERENCES

45o on tional Archaeometry, Symposium ed. A. Aspinall and S. E. Warren, Bradford, pp. 292-302. and G. A. Wagner, 1991. 90, Basel. Archaeometry

Pernicka,

"Experimental

1991, Pollard,

D.

K. 2001.

-.

F. 1988.

Poursat,

on

"Observations

in

B.c.,"

ed. H. Waetzoldt

the

Heidelberg, P. A. 1988.

A',

Wirtschaft

ofCoastal Changes (BAR-IS 404),

from Holocene

Shorelines:

and Antikythira

(Greece),"

86, pp. 27-43.

Tectonophysics

Pitty, A. F 1978. Geography and Soil London. Properties, L. 2002. "The Political Platon, Cultural Palaces Wider

on

and

of the Zakros

Influence

Sites

Nearby

and

in a

inMonuments

Context,"

ofMinos: Rethinking theMinoan Palaces

ed. J. Driessen, (Aegaeum 23), I. Schoep, and R. Laffineur, Li?ge, pp. 145-155.

Plog, S. 1974. The Study ofPrehistoric Change, Podzuweit, formen

New C.

und

anzugrenden zur verglei Mainz.

Beitrag

Stratigraphie,

Seals,

and

Sealings 5),

(Aegaeum

ed. T. G.

pp. 25-32. Li?ge, ex?cut?es ?Mal

Palaima, -. 1996. Fouilles

III Artisans lia: Le Quartier Mu Les maisons-Ateliers du Mu B.

(Etcr?t 1976.

32), Paris.

"The Analysis of Soil Phos

Proudfoot, and Interpretation

in Contexts," phorus Archaeological in Earth Science, Geoarchaeology, and the Past, ed. D. A. Davidson and M. L. Shackley, Boulder, Colo., pp. 93-113. at Pseira, Crete, Pseira = Excavations

Philadelphia

I = P. P. Betancourt

varas,

eds. The Minoan

and C. Da Buildings

on

theWest Side ofArea A (University Museum Monograph 90), 1995. II = P. P. Betancourt

C. Davaras,

eds.

Building

and AC

(the "Shrine")and Other Buildings in Area A

York. 1979.

Trojanische Gef?ss inAnato der Fr?hbronzezeit

lien, derAg?is Gebieten: Ein chended

sis," inAegean Administration

Quartier

P. A.,

Crete

Docu "Hieroglyphic and Sealings from Malia Mu: A Functional Quartier Analy ments

Y. Thom J. Thommeret, L. F. and J. Laborel, Montag 1982. "Crustal Block Move

pp.

Period,"

1987,

pp. 75-76. 1990.

minoens:

Oxford,

at

and Palace

"Town

in the Protopalatial

157-184.

Raban,

gioni. ments

-.

in

of Archaeological in Data," Archaeology

and Historical

meret,

xov E'

inH?gg andMarinatos

"Sea-Level Movements

1987. Malia

and H. Haupt pp. 107-110.

and Crustal

Contribution

Pirazzoli,

et artisans

"Ateliers

in ?enpay^i?va

Kpr\xoXoyiKov Ivve?piov Herakleion, pp. 297-300.

-.

und

theHellenic Arc (Greece):The

ed. A.

1985.

J.-C.

minoens,"

von Ebla (Heidelberger Gesellschaft Studien zum Alten Orient 2),

Changes

"The Late Minoan

AieOvov?

Millennium

mann,

1969.

62,

pp. 299-304.

Trade of Lapis Lazuli in the Illrd

Pirazzoli,

BSA

Goblet and Kylix," BSA 64,

14, pp. 189-209.

175-184.

Pinnock,

"Late Minoan

1967.

Pottery, Summary," pp. 337-351.

G. J. Varoufakis, and Phelps, W. W., 1979. "Five R. E.Jones. Copper BSA 74, Axes from Greece," pp.

R.

Exper of Smelting et al. in Budd

a

Assessing and Signature of Rural in Greek Landscape

Settlement

M.

Popham,

and

"Some

the

Concerning A?rsenical Copper," 1991, pp. 169-174.

the

Archaeology/'/M^

1991.

iments

the

"Chasing

Farmstead:

Formation

in

and Wagner

pp. 127-136. A. ML, R. G. Thomas,

P. A. Williams.

their Regional Significance,"/o?r72^/ of the Geolog ical London 142, pp. 501 Society of 513.

Pettegrew, Classical

in Pernicka

Britain,"

and

Crete

Implications

for Early Bronze Age Metallurgy

S. R. Troelstra, and D. van Peters, J.M., Harten. 1985. "Late Neogene and in Vertical Movements Quaternary Eastern

of Secon

Smelting Minerals:

dary Copper

P.

D. 1991.

and P. A. Williams.

Ware,

eds.

E.,

R. G. Thomas,

A. M.,

Pollard,

Museum

(University

Mon

ograph 94), 1998.

IV = P. P. Betancourt

C. Davaras, B, C, D,

eds.

and F

Monograph

Buildings (University

105), 1999.

and inAreas Museum

REFERENCES

=

V

= P. P. Betancourt

eds. The Pseira

varas,

= P. P. Betancourt

C. Davaras,

(PrehistoryMonographs and R. Hope

varas,

Ceme

The

Archaeological Island, Part 1

eds.,

(Prehistory

IX = P. P. Betancourt,

C. Dava

Simpson, Pseira Island, Archaeological Survey of Part 2: The Intensive Surface Survey = P. P. Betancourt

X

eds. Pseira

varas,

and C. Da

2001.

O.

in Greek

Timber

L. Myres

tieth]. Oxford. Rackham,

Plant

"Terraces,"

1996. Landscape, A. Rapoport,

Making Manchester. 1969. House

Form

of their Use of the North

Bronze

Age sus," Bulletin

Read,

and Effectiveness,"/^ 491.

lurgy -.

A. and

Fish

the

Aegean

Age,"4//f71,pp.l-20. 1969. "The South-East

Early

Autonomy

European

PPS 35, pp. 12-47.

Copper

atMinoan

"Fishing of a Bronze Formation

Age

5, pp. Archaeofauna M. 1994. "Hallan ?emi Rosenberg, Some Further Observations Tepesi: and Mate Stratigraphy Anatolica 20, pp. 121

140.

Bronze

of the Age,"

Pro

"Copper

in the Northeastern

A.

Sampson,

Balkan

-.

in Papa pp. 73-74. "The Neolithic Period

1996.

"Euboea,"

1996,

thanassopoulos 2002a. in the Aegean," pp. 161-167.

in Sampson

ed. 2002b.

A., Sampson, Settlement

Sanders,

from Crete,"

Assemblage

Concerning rial Culture,"

Metal

1999.

2002b,

The Neolithic

at Ftelia, Mykonos,

Rhodes.

-.

1998. Chairs:

"Cycladic

of

and

Indenbaum,

Eneolithic Culture," JAS 26, pp. 1059-1068.

R.

135-140.

Pro

13, pp. 477

1967.

C.

Ores

beam

1996.

Pseira:

Cauca

Population

duction from Polymetallic Sulfide

and

246.

Problem

Renfrew,

inMicro

Rose, M.

in the

"Sampling

Processing

and Medieval

N., G. Ryndina, V. Kolossova.

Nutritional Methods

Analysis ed. A. D. Romig 1984, Jr. and J. I. San Francisco, Goldstein, pp. 243

and

Ae

for Domestica

(TAPS 48:3), Philadelphia.

Iron in Early Smelted Copper from

for Regional Surveys: of Representativeness

cedures A

1986.

D. W.

Crete:

and M. J. F, J. F. Sculac, 1984. "The Notis. Precipitation

of theMetals Museum, Institute Metals 23, of

The Japan pp. 1-18.

Age

Roeder,

Cliffs, N.J. Englewood I. G., and N. V. Ryndina. 1995. Ravich, and "Early Copper-Arsenic Alloys the Problems

Ancient

Storage Life ofMinoan Olive Oil," OJA 21:1, pp. 63-75.

Culture,

Evidence

1986.

5:3, pp. 299-308.

2003,

in P. M. Warren,

Remains,"

Timna,"

and J. Hansen. in the Prehistoric

Russell, J. C. 1985. Control ofLate

Properties,

Cretan of the

The

1985, pp. 123

tion in the Early Bronze Age," OJA

and Metallurgy,"

and Renfrew

on Bronze

Lecture,

C. N.,

gean: The

Riley, F. R. 2002. "OliveOil Production

130. -.

2003.

Myrtos:An Early Bronze Age Settle ment in Crete, London, pp. 315-317.

and J. A. Moody. 1992. inWells 1992, pp. 123

O.,

Slater.

Renfrew, J. 1972. "AppendixV: The

The Twen

History: Memorial

Runnels,

in Crad

Evidence,"

"The Olive

and E. A.

C,

Perspective,

Cambridge,

Fur Smelting Israel: The Ar

"Copper in the Arabah,

chaeological

pp. 301-318.

and

Trees, Wood,

1985. naces

150.

in Elster

Parts of the Mediterra Adjacent nean Sea," Earth Processes Surface 271-288. 17, pp. andLandforms Rackham,

-.

Valley of London.

Copper Mines,

Commodities

"Metal Artifacts

coming. Pye, K. 1992. "Aeolian Dust Transport over Crete and Deposition and

3, pp. 69-87. B. 1972. Timna:

Rothenberg, the Biblical

in Prehistoric

in Cultural

"Cheating

Territoriality

atMusical and Sedentism

in an Context," Evolutionary CurrAnthr 39, pp. 653-681. M., R. Mark Nesbitt, Rosenberg, Redding,

andT

F

Strasser.

I. F

Sarpaki, A. botanical

1992.

"The Palaeoethno

inWells

1992, pp. 61-76.

2004. Sarpaki, A., and J. Bending. "Archaeobotanical Assemblages," inMochloslC, pp. 126-131. Schachermeyr, ?ber eine kreta,"^

1995.

Crete,

The Mediter Approach: ranean Triad or Is It a Quartet?"

Schickler, R. W.

1982. Roman

Warminster.

F. 1938.

"Vorbericht

nach Ost Expedition 1938, cols. 465-480.

H.

1981.

Zinnbronzen,"

"'Neolithische' in Studien

zur

Bronzezeit: Festschriftf?r Wilhelm

"Hallan ?emi Tepesi: Some Prelim

Albert

inary Observations

Mainz,

Concerning

to

Archaeo

Interaction,"

materials

dock andHughes

Renfrew,

forth

Mineral

Europe," in The SocialLife ofThings: ed. A. Appadurai, pp. 141-168.

12), 2005.

X, Block AF,

ofWealth

Emergence

eds., The

(PrehistoryMonographs

pp. 147-222. Angeles, "Varna and the

1986b.

1989.

and J. R. Pigott, "Direct Reduction

Copper Metal by Oxide-Sulfide

and

Gimbutas,

5, pp. 5-20.

V. C.

W.,

Dvorak.

Sitagroi in Excavated Areas," at A Prehistoric Sitagroi: inNortheast Greece, ed.

-.

ras, and R. Hope

Rostoker,

Se

"The

1991.

Smelting to Copper Alloys," Archaeomaterials

London.

Europe, 1986a.

and J. R. Dvorak. with Co Experiments

W.,

"Some

Civilization: The Before Revolution and Prehis

Village C. Renfrew, M. E. Elster, Los

Pseira of Mono

graphs 11), 2004.

in

quence: The Excavations

C. Da

Simpson, Survey

toric -.

6), 2003.

= P. P. Betancourt,

VIII

-. 1973. Radiocarbon

of the Tombs

Subsistence Early Neolithic in Eastern Anatolia," Behaviors Anatolica 21, pp. 1-12. Rostoker,

sation: The

and

eds. The Pseira

tery 2: Excavation

the Pre

The Emergence of Civili and the Cyclades Aegean the Third Millennium b.c., London.

1:

The Surface Survey (Prehistory Monographs 5), 2002. VII

14 and

"Carbon

-. 1972.

and C. Da Cemetery

1971.

of Europe," history Scientific Ameri can 225:10, pp. 63-72.

109), 2001.

Monograph VI

-.

The Archi J. C. McEnroe, Pseira of (University Museum

tecture

4SI

v. Brunn,

ed. H.

pp. 419-445.

Lorenz,

REFERENCES

452 N.

Schlager,

"A Town

1999.

in the Far East

Nero

Aspro

in POLEMOS:

Crete,"

en

guerrier

?

Eg?e

of

du Bronze

252:2, neolith

"Pleistoz?ne,

Shugar,

im fernen

und Ruinen

Sikka, and L. Dollhofer.

N.,

1998.

und Bleiisoto

Naturwissenschaftliche und lithische Industrie

ed. M. pp.

R. Behm-Blancke, 1904-1905. 1904,"

the

Department Free Museum

-.

pp. 207-220. 1909. Island

of

Unter

ammos, Seidel, H.

-. 1992.

on the Island

in the Island

"Eo-Alpine

in the

and System?Petrology 2: Part Geochronology, Synopsis

-.

Ophiolites," to Mineralogy 76, pp. 351-361. Wild

Flowers

pp. and

of Crete,

Spanakis, R.

1972.

of

Method Journal

pp. 870-893.

of Prediction," of

Science

272,

Hesperia

atMoch 65,

S. G.

P. 1993.

"The Chro

'Kastri

Recon Group' 88, pp. 5-20. 1991. Il?Xeic Kai%ojpi? erro n?paona tcov aico

rr/? Kpr?TJ]c vcov 1, Herakleion.

Silicate Liquids: An

Magmatic Empirical American

"Viscosities

Quarter: IA, pp. 7

175-230.

Sotirakopoulou, of the nology sidered," BSA

Athens. Shaw, H.

63, pp. 391-436. "Excavations

1996.

los, 1992-1993,"

Metamorphics

Contributions

inMochlos

1994. "Ex J. S., and C. Davaras. atMochlos, cavations 1990-1991,"

Hesperia of

"The Artisan's

Soles,

Nappe

Petrology Sfikas, G. 1987.

on

Ceme The Prepalatial and Gournia,

A,"

Building 40.

1981.

Haare.

Metamorphism Unit of the Cretan

High-Temperature and Associated

The

Work

1997. "A Community of Craft atMochlos," in Laffi Specialists neur and Betancourt 1997, vol. 2, pp. 425-431. -. 2003.

ofPachy

Philadelphia. H. Kreuzer, Okrusch,

Uppermost

a New

Princeton. -.

The Cemetery

and W.

Particle

AJA 95, pp. 17-78.

the

Crete,

Raschka,

for

Crete," Expedition 20, pp. 5-15. -. 1991. "The Gournia Palace,"

in 1908,"

Explorations Boston.

E., M.

Archaeological WorldArch 4, pp. 5-18.

teries atMochlos

1910. Excavations

ofMochlos, -. 1916.

in Pop

"Changes in

Museum's

University 1,

ofPseira, Philadelphia. -. 1912.

1972.

Pressure

J. S. 1978. "Mochlos, at Old Excavations: Look

AJA 13, pp. 273-303. -.

Copper and Renfrew

Soles,

of

ofArchaeology, Science and Art

Crete,

with

SizeAnalysis of Soils (ASTM D 422 63), Las Vegas 1972.

"Excavations

on

P. E. L.

ulation

T?bingen,

"Excavations

"Sherds in Elster

Explanation," SMPSAS= Standard Method

Transactions

of Mochlos,

xpDG?ccpi gtt| December

2003, p. 303.

108-123.

B. Seager, R. at Vasiliki,

E. 2003.

Smith,

38),

Forschungen

"NeoA,i9iK? KaOrijuepivrj, p. 15.

Deposits,"

suchungen

(Istanbuler

(paper,

454. Slater,

der Metallfunde penverh?ltnisse vom Hassek inHassek H?y?k," H?y?k:

I. 2002.

24,2002,

S., F. Begemann, 1992. "Chemische

Zusammensetzung

Israel"

Sjoberg, A. 1976. "Phosphate Analysis of Anthropic Soils,"JFA 3, pp. 447

67, pp. 4-35. and E. Pernicka.

at Abu Matar,

M?kovo,"

"Dieminoische Grabh?hle Chosto imKastri Goudourou Sitias," ?Jh Schmitt-Strecker,

"Reconstructing

Milan 2003).

157-220.

pp.

Scientific

pp. 39-47. A. N. 2003.

Process

osten Kretas," ?Jh 70 (2001), Schlager,

Peru,"

American

the Chalcolithic Metallurgical

rezente

und

in

"Copper-Alloy Metallurgy Ancient

1, ed. R. Laffi

bronzezeitliche

Befunde

Catalogue of Cycladic in theAshmolean Museum: Antiquities The 2 vols., Oxford. Captive Spirit, 1991. Shimada, I., and J. F. Merkel.

171-177.

pp.

S. 2000.

Sherratt,

contexte

Le

l'?ge

19), vol.

(Aegaeum neur, Li?ge, -. 2002. ische,

of Castles':

Fortified Site at

An MM/LM

-.

1993. n?Xeic rr?c Kocixoipi? tcov ai?vcov 2, Kprjrrj? oro nepaafia Herakleion.

REFERENCES

Sperl, G.

1990.

and other Cycladic Islands in the

des

"Urgeschichte

Bleis," Zeitschriftf?r Metallkunde 81, pp. 799-801. P. 1984. Spitaels, Mine

Period

3 (Theatre

of Early Minoan Metal in lurgy," History Kea-Kythnos: ed. L. G. Mendoni Archaeology, Origins

Helladic

"The Early

A. Mazarakis-Ainian,

Sector),"

8, pp. 151-174. A. B. 1865. Travels

Thorikos Spratt,T London.

-.

in Crete,

R.,

Copper Materials

D. H.,

and Propachlor," in Microbiology

Reviews

Meto

T.

-.

Critical 24:1,

pp.

1

pp. 325

and N. H.

"Aspects in Mesopotamia in The Archaeometallurgy

tolia,"

pp. 59-71. and V. C. Pigott. Stech, T,

1986.

Millennium

b.c.," Iraq 48,

pp. 83-101. Z. A., N. H. 1996.

and

Gale,

"Lead

1991.

"Sources

from

of Metals

and Trade

1992.

Archaeological in Archaeology,

in Age Aegean," Trade in the Bronze Mediterra Age nean (SIMA 90), ed. N. H. Gale,

Bankoff. 1993-1995," 1:

pp. 255

Sofia,

ed. V. T. Holliday,

D.C., pp. 193-205. Washington, Z. A. 1989. "Cycladic Cop Stos-Gale, in per Metallurgy," Hauptmann, and Wagner Pernicka, 1989, pp. 279-292. 1992.

of Thermi OJA

of Metal

Origin

on

the Island

11, pp.

of

155-177.

1993. "TheOrigin ofMetal Used forMaking Weapons in Early andMiddle Minoan Crete," in and

in Prehistoric

Exchange

Europe (OxbowMonograph Scarre

and F.

Healy,

33), Oxford,

115-129. 1998.

on Euboea,"

"The Role

of Kythnos

Aegean

of

Archaeology

Implications

from Knossos

and

for the Aegean

3,

1994.

Neo

Pat

"Settlement

Settlement

Perceptions:

the Site: Regional Studies in Beyond the Aegean Area, ed. P. N. Kardulias, Lanham, Md., pp. 313-335. Swann,

C. P., S. Ferrence, and P. P. Bet ancourt. 2000. of Minoan "Analysis

from

Pigments Used on Pottery Palaikastro,"

ments

and Methods

Nuclear in

Instru

Physics pp. 714-717. 1988. Swann, C. P., and S. J. Fleming. "Proton-Induced Emission X-Ray Research

B

164-165,

R.

Apacprivi,"

J. V. Thirgood, Mediterranean of Resource

1955.

Prakt

ev

"AvaGKaqrri

1952,

pp. and

1981. Man

129-151. the

Forest: A

History New York.

Depletion, F. C. 1958.

"The Early and Bronze," Copper 1-7.

Thompson,

of

Metallurgy Man 58, pp. C. P., C. C. Thornton, M.

2002.

Young.

Lamberg-Kar and S. M. N.

Liezers,

"On Pins

the Evolution

Tracing Base Alloying via

at

ICP-MS

Place

Tepe

Analysis

and Needles: of Copper Iran, Yahya, of Common

29, pp.

Items,'7^S

1451

Varro: The Farmer,

inQuaternary Alluvial Crete"

Timucin,

and A.

M.,

(diss.

Knoxville). E. Morris.

1970.

and Thermody

Equilibria

namic Studies in the System CaO their

Rethinking the Greek Village," in

White

D.

Theocharis,

Deposits Univ. of Tennessee,

lithic,"AJA 106, p. 251. S. B.

53:2,

Archaeology

pp. 70-73.

"Phase

Axes

Delhi.

"Lost Continent:

of Eastern

Strasser,T. F 2002. "The Polished

Sutton,

Civilizations,"

Developed

pp. 49-62. Stone

1964.

M. L. 1992. "An Timpson, Investiga tion of the of Soils Pedogenesis

and

of Manika

Cemetery

New

Tropical Asia, J. E. 2000.

Tilly, B., ed. 1973. London.

Metal Artifacts from the Early

terns, "The

Objects from the Early Bronze Age Lesbos,"

pp. 248-280. Z. Stos-Gale, A., A. Sampson, E. 1996. Mangou. "Analyses

and R. L. Donahue.

1460.

Jonsered,

Helladic

in "Organic Matter in Soils Contexts,"

in

lovsky,

in the Bronze

and H. Hauptmann, 338.

ed. C.

Isotopic in the Isotopic Research inMETRON: Aegean," Measuring the Aegean Bronze Age (Aegaeum 24), ed. K. P. Foster and R. Laffineur,

Isotope the Aegean," Analyses 38:2, pp. 381-390. Archaeometry Z. A., and C. F. Macdonald. Stos-Gale,

"The

in In the Gaul Steps ofJames Harvey in Gaul, Memoriam, James Harvey ed. M. H. Todorova, Stefanovich,

Trade

and

of Ores

and H. A. Stefanovich, M., 1998. "Kamenska Cuka

Site

"Lead

N. Annetts.

pp. 39-64.

J. K.

2003.

Stos-Gale,

Metals Trade in Southwest Asia in the Third

pp. 59-63.

Stockholm,

Li?ge,

of

theAsian Old World (MASCAP16), ed. V. C. Pigott, Philadelphia,

P. Bali,

Soils:Their Chemistry and Fertility

An Alternative View of the Rise of

Gale.

Other

of Early Met and Ana

1999.

allurgy

-.

Y in

8.3,"

"The Minoan

natos,

and Methods in Physics B 49, p. 65. R. V., D. P. Motiramani,

Tamhane,

Thalassocracy in the Aegean Metal Trade," and The Minoan Thalassocracy: Myth ed. R. and N. Mari Reality, H?gg

22.

pp.

Z. A.,

1984.

and O. H. Tuorinen.

"Biodegradation Herbicides Alachlor,

lachlor,

-.

from

Terrell,

Stos-Gale,

191

of the Aceta

nilide

-.

of Metals

"Origin

in

Instruments

329.

and

Smelting," Metallurgical Transactions 28B, pp.

in Spectrometry Archaeology," 2, p. 197. Scanning Microscopy -. 1990. "Selective Filtering PIXE Nuclear Spectrometry," Research

and

Stampert, 1998.

Stein,

and

Athens,

Isotope Analysis. Appendix Elster and Renfrew 2003,

200.

Stech,

pp. 717-735. 2003.

and

Sitagroi asDetermined by Lead

and S. Simeo J.M. Togurl, 1997. Losses and "Copper in Considerations Thermodynamic

Sridhar, nov.

453

Metallurgical FeO-Fe203-Si02," Transactions 1, pp. 3193-3201. H. 1978. The Eneolithic Todorova,

in

Bulgaria (BAR-IS 49), Oxford. and A. Papacosto M., 1997. "'Villas' and Villages poulou. in the Hinterland of Petras, Siteia,"

Tsipopoulou,

inH?gg 1997, pp. 203-214. E.T.

Turkdogan,

1983.

Molten of

Properties London.

Physicochemical and Glasses,

Slags

of J. 1978. "The Vegetation Greece Prehistoric Times: during The Palynological in Evidence,"

Turner,

Thera

and

Presented Scientific

World: the Papers Aegean at the Second International

Congress,

Santorini,

Greece,

REFERENCES

454 ed. C. Doumas,

1978,

August

Lon

Greece," Argolid, pp. 103-128.

don, pp. 760-773. Turner, J., and J. R. A

van Andel,

1975. Grieg. Pollen Diagrams Review from Greece," ofPalaeo and 20, pp. 171 botany Palynology 204.

Veenhof, K. R. 1972. Aspects ofOld

-.

Tuyeres," 1991. Alloys:

in

W.

and ?lu?re 1991, pp. 213 and J. F. Merkel. Tech Smelting

R. F,

1974.

W.

1985,

Y Veni Y., S. Chryssoulaki, 1990. "Les eri, and M. Avgouli. routes minoennes: Le poste de Xoi et le contr?le

BCH114, pp. 43-62. S. S. Vou Y, Chryssoulaki, 1989. "Les and Y. Venieri.

tsaki, routes minoennes: naire:

"Tracce neolitici

insediamenti dell'antica

ed. G.

in Onore

di Doro

and G.

P. Carratelli

l?tico di Festos,"

tion

Wagstaff, 133.

-.

1996.

neo

-. 1983.

34-35

Neolithic:

theWider

World,"

5, pp. 29-39. and P. Belli. 1978.

L.,

and Problems

Chronology," 43. -.

125-163.

Vagnetti, Y. Tzedakis.

and

neolitici,"

Christopoulou, 1989. "Saggi negli in Scavi aNerokourou,

Rome,

Kydonias,

1991.

J.-P.

Productive and Lloyd van Andel, 1986.

Difficult but

Relationship," 1991, pp. 10-17.

T., C. Runnels,

and Abuse

Age

Middle

Provincial

of

Mi

in Barker

and K. Pope. of Land

Years

in the Southern

1965.

Desertification

"The First Minoan

and Early Minoan CretChron 19, pp. 7

1980.

"Problems

of Chronol

ogy inCrete and theAegean in the Third and Earlier Second Millen nium

"Survey Archaeology

"Five Thousand

106

Bronze 1972. Early Myrtos:An in Crete, London. Settlement

strati -.

pp. 9-97.

and Rural History?A

Vases

Stone

of the Final

L., A.

P. M.

Warren, "Char

inCrete," SMEA 19,

Neolithic

and pp.

Cambridge,

Mainz. Pottery, Walls, J. 1982. Combating in China, Nairobi.

Studies

Cretan Vagnetti, acters

ed. C. Renfrew

noan

"The Final

Enters

1982. and S. Augustson. in An Island Land Use,"

G. 1976. Kamares:A Walberg, Study Palatial Middle the Character of Minoan Pottery, Uppsala.

Rizza,

(1972-1973), pp. 7-138. Crete

inMelos,

M. Levi,

n.s.

ASAtene,

3. Jahrtausends,"

des

Polity: TheArchaeology ofExploita

territorio

Catania, pp. 1-9. -. 1975. "L'insediamento

aus dem

1-18.

M., Wagstaff, "Traditional

inAntichit?

Gortina," Studi

Cretesi:

und

pp. G. A., and G. Weisgerber. Wagner, 1985. Silber, Blei, und Gold auf Sifnos, Bochum.

di due nel

in

und Arsenbronzen aus Ebla

0^23,

circulation de la d?fense?"BCH 113, pp. 43-75. L. 1973. Vagnetti,

Ost Zeitschrift

1975.

Mesopotamien

Rapport pr?limi ou la circulation

de

d?fense

"Zinn

den Texten

munications," Tzedakis,

Schneider.

'Phyllite-Serie' M?lange,"

Gwosdz.

1984.

com

des

Baumann,

"Geodynamische Geolo Ostkretas," Interpretation 728-750. Rundschau 64, pp. gische Bachmann. and H.-G. Waetzoldt, H.,

Tzedakis,

Use

"Die

A.

and W.

Gesellschaft Geologischer 25, pp. 237-251. Wachendorf, H., G. Best, and

pp. 3-20.

pojiav?pe?

Terminology,

der Deutscher

1985.

and Future,"

inCraddock andHughes

Vallat,

Gwosdz,

kretas?Eine

"Experimental niques: Achievements

pp.

H.,

Wachendorf,

221. Tylecote,

its

Greece, Urbana.

Base

"Early Copper or Man-Made?" Natural

Mohen

and

Vickery, K. F 1936. Food inEarly

to

107-118.

13, pp.

Trade

Assyrian Leiden.

Pot Bellows

"From

and Histori

17, pp. 379-396.

cal Greece,"/^

London.

Levant

id Prehistoric

Erosion

Tylecote, R. F 1976. A History of Metallurgy, -. 1981.

and A. De T., E. Zangger, 1990. "Land Use and Soil

mitrack.

"Some Holocene

55,

Hesperia

-.

84, pp. 487-499. of the "The Genesis

b.c.," AJA 1987.

Minoan

Palace,"

natos

1987,

Watrous, ments

L. V.

pp. 2-6.

in

H?gg

pp. 47-56. 1977. "Aegean

and Transhumance,"

and Mari

Settle TUAS

2,

REFERENCES

-. 1982.

of History Plain in Highland

logic Industries Deforestation

Princeton.

Crete, -.

2001. in

1999," A Review,

1994

"Addendum:

Aegean ed. T

and H.

Wheeler,

Age World,

"Spatial Organi at and Smelting Archaeol Jordan: Mining

through and J. Lang,

Craddock

pp. 76-89. H. 1997. Weiss,

London,

in Third Millennium

ed. N. Dalfes, Collapse, and H. Weiss, Berlin,

G. Kukla,

of Ancient

1993. Courty. and Collapse of the in Akkad, Akkadian the Empire," First World Structure, Ideol Empire: ed. M. Liverani, ogy, Traditions, Padua, B.

pp. 1996.

Survey, pp. 9-22. Wells,

"Introduction,"

ed. B. Wells,

B., C. Runnels,

Wells,

1990.

-.

Werner,

K.

the Aegean

inAn

Sherd

Scat

of Land Tenure "Aspects Cretan Studies 2, Crete," 1985.

"The Pottery

Court

House

at Knossos,"

BSA

and

80,

pp. 281-364. D. E., Wilson,

pp. 207

"Ceramic

1994. and P. M. Day. in Prepalatial Regionalism Crete: The Mesara Imports

from EM I to EM IIA Knossos,"

1993.

The Megaron during and Anatolian Bronze Age

BSA -.

89, pp. 1-87. 2000. "EM I

Chronology

neue

(Der

8), ed. T. Rehren,

pp. 279-289.

pp.

13),

ed. U. Yal?in,

and

Bochum,

17-30.

1999. and E. Pernicka. Yal?in, ?., "Fr?hneolithische Metallurgie in Hauptmann A?ikli H?y?k," 45-54. 1999, pp. 1980.

A

Various

von et al.

of

"Distribution

Elements

between

Copper, and Slag," Erzmetall33:7/8, pp. 377-382. Yener, K. A. 2000. The Domestication Matte

of

Metals: The Rise ofComplexMetal Industries Zohary, M., Outline

inAnatolia,

Leiden.

and G. Orshan. of the Geobotany

Israel Journal

of Botany

"An

1965.

of Crete," 14, pp.

1-49

and Zois,

suppl. A. A. 1968.

"Yjcapxei

e7io%f|;," in nenpay?eva

Architecture of the EM IIAWest

Central

(677kW108),Jonsered.

"Extensive

pp. 221-230. D. E. Wilson,

chaeological Survey:The 1988 OpAth

1990. in Dorian

and E. Zangger. Ar

18:15,

Sherd-Sampling 9, pp. 323-333.

niques,"/^ 1989.

Some Intensity: 16, pp. 31-46. Results,"//^ R. F 1955. Aristocratic Willetts, Society inAncient Crete, London.

"The Berbati-Limnes

Season," 238.

by Means Tech

ters and Land-Use

Stockholm,

Agriculture Stockholm.

cient Greece,

Zones

Manured

von

and J. D. Muhly,

A. Hauptmann,

Yazawa,

Recent

in

Archaeological

ed. 1992.

B.,

-.

131-155.

The Berbati-Limnes

pp. 251-259.

of Extensive

and M.-A.

"The Genesis

Beiheft

Beiheft

Cambridge,

of

der Metal "Anf?nge in Anatolien," in verwendung 1 (Der Anschnitt Anatolian Metal

Wilkinson, T. J. 1982. "TheDefinition

pp. 711?

Tombs

The Vaulted

und Untersuchung in Interpretation," Metallurgica inHonour Hans-Gert Antiqua, of and Robert Maddin Bachmann

Bochum, -. 2000.

"Population, in Prehis sources, and Explanation in in Pattern Past: Studies tory," of the David ed. I. Hod Clarke, Memory of der, G. Isaac, and N. Hammond,

b.c.

4,

London.

1998. "Der Keulenkopf Yal?in, ?. Can Hasan (TR). Naturwissen

Re

P. 1981.

Wilkinson,

723. H.,

and

vol. 2, pp. 265

Cambridge. Millen

"Late Third

1924.

Anschnitt

C. R., ed. 1988. Pastoral Whittaker, in Classical Economies Antiquity,

ed. P. T.

ArchDelt

schaftliche

17:4,

in Laffineur

Ttpcoto

"M?ya?

Tacpo? n?pyoD,"

136-170.

Mesar?,

274.

Climate Change and Old World

Wells,

1997,

pp. -.

Day, E. Kiriatzi, 1997. "Ceramic

Korifi,"

Betancourt

nium Abrupt Climate Change and Social Collapse inWest Asia and Egypt,"

and

and J. D. and the

Expedition

P. M.

Fournou

and Metal the Ages,

uiv?hko?

Traditions in EM IIBMyrtos,

ogy Beyond theHistory of Tech in Mining

"Ingots

and V. Kilikoglou.

2003.

of Mining

nology," Production

in Ar

pp. 31-39. Whitelaw,T,

Feinan,

Weiss,

1975.

Mediterranean,"

Leiden. Weisgerber, zation

of Con

in Keller

Survey,"

inAmer

Wines,

ica, Philadelphia. S. 1918. Xanthoudides,

Bronze Age Copper Trade in the

of thePersian Gulf: Technology,

G.

"Methods Collection

S., R. Maddin,

T

Muhly.

MATA,vo\.3,pp.905-911. L. R. 2003. Early Metallurgy the Bronze

95, pp. 21-63. R. A. 1985. Fertilizer

10,

Antiquity,"/f?4

Rupp 1983, pp. 73-83.

in the

Weeks,

and

1983.

Surface

chaeological

inMELETE

Period,"

Neopalatial

1999.

Blitzer.

of Gournia

"The Region

Trade,

R.

Whallon,

Boston,

trolled

pp. 216-223. L. V., Watrous,

in

from the Pottery at Knossos," BSA Tests

Practice: Palace

Early

and Mediterranean

445-452.

Prehistory: Cullen,

Social

T A. 1983. "The Furnace Wertime, versus the Goat: The Pyrotechno

A

Lasithi: on a

Settlement

455

UM

III

rov B' Aie

A', IvveSpiov ?vov? Kpr\xoXoyiKov Athens, pp. 141-156. -. 1993. "AvaGKacpri BaGi?,iicr|? Prakt (1990), pp. 315 Ieparcetpa?," 341. 1991. "Natural Copper and Smelted Arsenic Alloys in Bronzes Produc Early Metal and Elu?re tion," inMohen 1991,

Zwicker,

U.

Arsenic

pp. 331-340.

INDEX

Abietadiene,

422,425-426

abietic acid, 420,422,425 acetanilide, 419,421 a-cetone. See isomethyl-a-ionone acid

esters,

416,418,423

Aegopodium

podagraria. 169-170

Afghanistan, or deer, agrimi

212

Agriomandra,

8-9,21,

See

ground

30-31,

elder

33, 37,

41,100,142,181,185,198,200 201,203-204,221-223,226,235, 237,239-241,243,249,251,253 255, 301 Agriospelio, 13-14 292

akermanite, Akkadian

empire,

162

Aksaray, 161 Alaca H?y?k, 170,175 aldehydes, 415, 418, 419,421,422, 425-426 cave,

Alepotrypa

157-158,165

Aleppo pine, 420,423 alkanes, allspice, almonds,

413,416-419,421-424,426 418-419 258

anthropogenic

features

AF 1,199,355,357-358 AF2,199,276,355,357-358 AF 3,199,355,358-359 AF 4,199,275,355,359-360 AF 5,199,275,355,359-361 AF 6,199,275,355,360-361 AF 7,199,275,355,361,362 AF 8,199,355,361-362,378 AF 9,39,199,274,276, 355,361, 363,368,399-402 AF 10,199,275,355,363,394,397 AF 11,199,275-276,355,363 AF 12,199,275, 355, 363-364 AF 13,199,275,355,363-364 AF 14,199,355,364

AF 15,199,355,364 AF 16,199,275,355,364-366, 397 AF 17,199,275-276, 355,365-366, 377-378,394,397 AF 18,199,275,356,366 AF 19,199,356,366 AF20,199,275-276,356,366 AF21,199, 356, 366-368, 370,378 AF 22,39,199,251,273-274,276, 356,363, 368,372,374,394, 411-412 399-402,406-408, AF23,199,259,356,368 AF 24,199,356, 369 AF 25,199,356,369 AF 26,199,356,369 AF 27,199,356, 369-370 AF28,199,276,356,370 AF29,199,259,356,361-363, 370-372,378 AF 30,199,251,253,356, 372,374 AF 31,199,251,253,356,372 AF 32,199,201,243-246,249,268, 276,356,361-362,372,378 389, 394 AF 33,199,356,373,386,389 AF 34,199,259,356,373 AF 35,199,356,373 AF 36,199,356,372,373 AF 37,199,356,373-374 AF 38,199,356, 374 AF 39,199, 356, 372, 374 AF 40,199,356, 374-375 AF 41,199,356, 365, 375 aluminum, 328, 332, 335 Alvania,

149,151

Alykomouri, 217,225,236-237 Amnissos

cave,

Amorgos,

319

amphoras, Anatolia, Anavisos,

157

73, 393-396 61 157

INDEX

458 bladelet, 108 blowpipe, 112

Andiskari, 303 Andros, 157 animal

Chania, 303-304, 319,419

anisaldehyde, 423-425

blueberry, 425,426 boldo tree, 426

anise,

borneol,

fats, 420-422,424,427

423-425,427

anorthite, 305, 307, 309 Ano

310-311

Varsamonero,

acid, 422

anteisopentanoic

apricot,

426

apsidal

structure,

main

brick making, 38

65,145-147,157,163-169,

97,381-382,415,426 Britain, 142,163 bronze

166-167,171

Arslantepe,

161-162

A?ikh H?y?k,

See pawpaw See Pyrgos.

Aspros Atomic

176 bucket jar, 79, 81,87, 93, 97,415,416, 418,424 buckle, 212

145

arsenopyrite,

Emission

Bulgaria, 157,159

Siphnos

Spectrometer,

38

See

axes,

capric

302-303,

110,153-154,184,223,233,

241,247,250,260,262,275 basil, 417 basins,

79, 93-94,391,415,422

basket

impression,

beads,

158,161-162 161

bears, Bebonas, Beer

Ora.

71

beeswax,

83-84,

351,413,415,422 420

166-167 Beycesultan, 430-432 biotite,

Bir Safadi, 142 birds, 212,254 bismuth, 172 blackberry, 421

75-76

160,162

425,427

146,175

221

cerith,

422

acid,

vessel,

176

?atal H?y?k,

cedar,

See Timna

86, 94, 96,112,

125-132,177,183-184,186-187,

benzoic

open

Cay?n? Tepesi, 160-161

227

61, 68,

slags, 285-286,

221

catmint,

8-9,197-198,200-201,

203-204,221-223,226,237,239, 243,248-249,251,253,255,272 70,216-220

Chalinomouri, 229,269-270 Chamaizi,

Chordakia,

9-10,12,14,16,203-204,

221-232,235,238-239,278,365

4,269-270

330

142,184-186,

chrysocolla,

territory,

303-304, Chrysostomos, in ravine, 204,254 church

area of, 204 310-311

See travertine

Chylopittes. Cinnamomum

aromaticum,

formation 425

419

185 282,289

coal deposit, 40 combed

149,151-152 71

decoration,

110,

furnaces,

349

conical cup, 72, 88, 97,213, 385 cooking dishes, 65, 68, 72-73,79, 84, 377, 87, 91, 93, 97,183,209,212, 380,387 65-67,

pots,

cooking

73,183,209,

377-378,381,385,387-389,419 coriander,

418,421,425,427

Coriander

sativum.

Cornwall,

143

See coriander

321-324

studies, 332-333

covellite,

Chalandriani, 165,319 chalcocite, 141, 332 chalcopyrite, 141, 300, 332

chalice,

226,229

Chondrovolakes,

corrosion

149

Cerithium, 149,151 chaff, 110,113,133-135,153-154, 180-182,184,233,241,250,262, 290-291

Chalepa,

3,5,9-10,14,47,198,200,

202-203,205-213,221-233,235, 237-239,243,251,253,255,261, 278,358,365,374

Columbella,

in

cattle, 212,234,248,271-272

bell cups, 208 bellows,

carinated

cassiterite,

146

chlorite, 144,290-292

Congolese

cup, 208,267

Caucasus,

391-392

bees,

424

carinated

carob,

discussion,

chisels, 165,172,174,189,212, 318-319 chloride, 146, 327

clinopyroxene,

caprylic acid, 418,419,421,422

290-292,331

Balomir, 159 barium, 325, 327,332

main

fragments,

109-123

clinoclase,

419,422

Capsicum frutescens, carbonate inclusions 155-160,164

84,

citms, 418,419,421,425

418,420,422

acid,

269-270

284-286,291-292,430-432

cinnamon,

acid,

caproic

142,144,183-186,

330,332

barley,

65,416^419,421,423,

campo. See Kambos

174-175,301,319

Balkans,

Theriospelio

425-427

159

101,155,159,164,170-171,

azurite,

See also

3,15.

Camphor,

Seriphos

awls, 156,159,161,189,318 axe-adzes,

cave,

79

Chrysokamino

Byblos, 170

augite, 289,292, 305,307 Avgo, 227-229,231 Avyssalos.

burial

pot",

Chomatas,

at farmstead, 213 jug discussion use, main of, 169?

bronze

145

arseniosiderite,

"cheese

Cheiromandres,

chlorine, 83, 87, 89, 94,

jar, 78,

bridge-spouted

171-177,185-186,188,300,304, 321-324,327-328,332,338,340 344,346-347,349-352,413,428

Asimina.

174

bracelet,

discussion,

221,426

chimney 78-79,

72-73,75-76,

cheese,

chert,

87,91, 93,97,209,212,216-217, 393,413,415,419,425

55-66 arsenic,

68,70,

bowls,

327-328

322,

antimony,

fields,

4,197,275-276,278,355-359, 366-367,369-370,375

acid, 420

anteisopentadecanoic

for agricultural

walls

213

Charonia,

418

boundary

212

cups,

champaign

crab, 149,152 425

cranberry, cristobalite, crucibles,

309 14,16,

63-64,113,137,158,

163,165-166,173,177,182,184, 263,300,302 cumin, Cuminum cuprite,

423-425 cyminum.

See cumin

141,144,146,282-283,287

289,292,338,344,348 cups, 65,70,72-73,77-78,87-89,97,

183,208-209,212,380,385,413, 415,425

459

INDEX

in bedrock,

cuttings

99,101,104-105,

350 70-71,

Cyclades,

79, 99,142,145-146,

Fascio

lar ia, 149,152

247,255,299,313-316,319,330 311 Cyprus, 16,175-176,188,273,

208-209,301,318 Debla, 154 decanal, 418,419,421,425,427

289,292,296-297,305,309

deep water fish, 255 dehydroabietic acid, 420 delafossite, 146,292,344,

fields

348

cultural

Fournou

farming

methods,

iron, 288-289,292 157 422

entrapped epicuticular

waxes,

140

421,423,426

epidote, 291-292 Ermioni,

316

Eskiyapar, 170 ethyl salicylate, 424 Euboia, 157,164,319 Euripides,

cave of, 157

Faience,

137

farmstead

at

3,5,

Chrysokamino-Chomatas,

7,12-13,16-17,33,181-184,

193,196-198,200,201-202,205 213,215,221-256,259,262,267 274,330,356, 369-372,375,391 393,397,408,410-411

chromatography,

hares.

See

Lepus

210-211

hearth,

horn

413-432

discussion,

22-36

Microprobe,

See licorice

road,

215

bones, 160

hyacinth, 424 orientalis.

Hyacinthus

See iron oxides/hydroxides 157

near modern

walls

Hungary, 16

420

glabra.

149,151

161

human

Electron

70, 75-76

vessel, shell,

horses, house

gold, 296,310,344 Gournes,

barley jar, 75-76

House of theTiles, 171

G?bekliTepe Ziyaret, 161 goblet, 71, 76-77 goethite.

See

horned

Giali, 158 ginger, 421 Glazed PaintedWare, 393 Glazed Ware, 397 Glycyrrhiza

acid, 419,426 See leather

horned

gehlenite, 305 main

338

arc, 23

Hordeum.

420,423,426

301

germander,

170

sites,

hieroglyphic tablets, 267 hook, 212

423 426

149,151

gastropod,

geology, Geoscan

8

Ioannis,

Hagios

Hagios Onouphrios, 168 Hallan ?emi Tepesi, 161

hides.

420,423,426 gamma-octalactone, 243-248 discussion, gardens, main

Gavdos,

70,226-227,229,

231,236

hercynite, 307, 309

421,426

gamma-nonalactone,

gas

gas bubbles,

304, 311,

417,421,423,424,426 heptadecane, herbal medicines, 65-66

162

beads,

gamma-lactones,

acid,

Antonios,

Hagios

Hellenic

133-134,136

lining,

gamma-hexalactone,

423

elemental

cycles,

Galena

165-168,171,217-218,

hedenbergite, 141,286,292, 138

gamma-decanolactone,

elaidic acid (C18:l trans), 419 Emporio, enanthic

Photia,

262,269,318 Hagia Triada, 70,166,168, 318-319

hexanoic Ware

Ebla, 169-170 Egypt, 125 eicosane,

Fournou

Myrtos

White-on-Dark

SeeWhite-on-Dark

Ware.

See farmstead

Hassek H?y?k, 167

trees, 247,258

furnace Cretan

location.

Chrysokamino-Chomatas See Kea Eireini.

Harrapan

Ftelia, 156

164

Durankulak,

See fennel

See

Korifi.

freeze/thaw fruit

at Hagia

Korifi

21

157

culture,

gypsum, 36-37, 303

for agri

See meadowsweet

vulg?re. 291

foraminifera,

dove shell, 149,151-152 drill, 174

East

walls

Final Neolithic period. SeeNeolithic period Fine GrayWare, 217 fish and fish bones, 149-151,255,426 fishhooks, 156,161,165,318 fishing for sport, 276 fish tanks, 28 flux, 143,145,183-186,330,346 Foeniculum

418,419,421,424-426

dodecane,

Filip?ndula

162

Gumilnitsa

Hagia ulmaria.

248

discussion,

discus

figs, 278

418,425

lands, main

Habitation

See boundary fields

field walls.

ditolyl ether, 419,423 Diyarbakir, 161

Dorian

main

for agriculture,

grazing

ground elder, 424,426-427

sion, 250-253

diethyltoluamide (DEET), 423-425 dihydroisophorone, 427 DikiliTash, 156 dill, 421 Dimini, 164 Dimitra, 156 diopside, 292 Dioscorides, 423

dodecanal,

used

391

Greenland,

291-292

siltstone,

graters,

250

field houses, 197,252,274-276,27%, 355-356,360-365,369,373

420,421

diacetylbenzene, 4 Diaskari,

142-143,177,180,188

ferruginous

acid, 424-426

decanoic

247-248,252,258,262,266,

271,278

feldspar, 291-292,431-432 fennel, 423-425, 427

159,163-166,170,172,189,

71, 73, 76,100,127,

181-182,215,218,228,231,256, 259,264-267,270,273 grapes,

140-141,188,282-284,286

fayalite,

Feinan, Daggers,

149

152

163,166-168,171,177,179,181,

4-5,18,

Gournia,

fatty acids, 413,416,418,420-426,428 faunal remains, main discussion,

See

hyacinth

hydrated copper chloride, 286 hydroxy-derivative,

423

hyssop, 417,420 root

Ierapetra,

4,22-23,26,28,223,226,

255 Ilipinar,

171

Indus River, 163,170 ingots, 187-188, 349, 352 Ionian

Sea,

19

374-375

INDEX

460 167

Iran,

isomethyl

a-ionone,

424,425,427

220,252,259,264,266,268,270, 316, 347 Kokkino Phroudi, 269-270 Kolonospilio, 14 Kommos, 69,125,165, 391

isomethyl

b-ionone,

424,425,427

Kondouro.

iron oxides/hydroxides,

328,

143,

330-333,341,344-346,350-352 irrigation pipe, 51

l?llingite, 347 loomweights, 65,209,213,245,271

70, 76,155-156,159,166,

Knossos,

Ireland, 142

See

Macedonia,

297,305,338,340,345,347-349 malachite, 318-319

76,160,165,168,

Koumasa,

Krasi, 170-171, 318 170

Jadeite,

301-302

Kritsa,

kylix, 212, 378 Kythera, 171-172,314

81, 83-84,

jars, 61, 65, 70, 75-79,

96-97,128,209,

86-89,93-94,

212, 382, 388,413,415,419,421 424,426 Jordan, 142-143,177,180,188 jugs, 65, 81, 87, 89,97,208,212,220

Mani, 157 Manika, 171-172, 319 of East

map marble

Kalathiana, 168, 318 kalathoi, 209 Kalo Chorio, 24, 70,220,228 Kamares

310-311

304, 9-13,

Kambos,

30,221-224,226,229

251,253-256,260,264,267,378, 408,410-411 lapis lazuli, 169-170

75

Cuka,

Kanli Kastelli, 218 kaolinite, 184,245

93

larnax, Lasaia,

157

laurel,

221

Karoumes,

269-270

lauric

acid,

lavender, 182

Lavrion,

Kastelli Phourni, 70 See Kythera See Monastiraki Katalimata.

Kastri.

Syme,

Kato

Zakros,

304,

4,125,

311

4-7,12,14,17-18,22,27,

70,179-182,197, 34,40,47,52, 205,217,236,249,254,256,258, 262,264,267,273,276,375 Kavousianos Kavousi sion,

River,

survey, main

regional 221-232

Kephala.

See Seriphos See Kea

Kephalolimnos,

9-10,198,221-224,

226,238-240,243,249,276 ketones,

Levant,

415,418,419,425-427

knives, 165,172,212

155 24-25,161-162,167,169 24-25 420,423

17,41,142,145,175,181,

methyl dehydroabietate, 418, 421,423

analysis,

304

168,175-176,

422,426

methyl

esters,

methyl

ketone,

methyl

nonanoate,

methyl

2-ethylhexaoate,

415,418,419-426,428 413 423 426

Miamou,304,311

272

430-432

mica, 310-311 70

319

Micromeria

425

myrtifolia,

milk, 419, 422 minimum

number

128

of bellows,

417,420

mint,

Mirabello

Fabric,

70-71,

73, 78-79,

425

83-84,

See Lebena

110,127-128,130,132,134,136, 181,377,380

Lepus, 42,149-150,212,254

licorice

Kea, 165,299, 314, 319 Kefala.

discus

19-20

decanoate,

304,

Lerna,

in Crete,

Mehrgarh, 161 mellite, 309 Melos, 100

methyl

171

Lesbos, 166,171,173-174,176

12

424 meadowsweet, mean temperatures

417

Ware,

Lenda.

345

balance,

benzoate,

Lebena

lemon,

65,213,235,

methyl

Lebena,

Lemnos, 33

mass

Mesopotamia,

420,422

Lefkandi, 171-172 lekane, 394, 397

33,198,200,

43,52,

metacarbonate,

isotope

leather,

391

203,222-223,229,238-239,243, 249,251,253,268 Kavousi,

313

310-319

9-10,12,16,

Katsoprinos,

310-311,

185,255,263,313,315-317 lead, 145,168,175-176,186, lead bead, 166 lead

301-302

Katharo, Kato

Katalimata furnaces

See Congolese Katanga. 70 Katharades,

resources,

245,271 Marshall Plan, 278, 365

Mesara, 304,

302,

Karanovo,

391 Karpathos, Kamm Kanesh,

34, 36,

9,12-13,17,

237-239,243-244,246,248-249,

231,234,248,250,254-256,266, 277-278, 365 Kamenska

Ambelliou,

38,42, HO, 113,184,197-198,201, 203-204,221-223,226,229,235,

157

vases,

marine

420,423,426-427

Lakkos

168, 318

169-170

Mari,

ladle, 213

266

Ware,

Kambanos,

lactones,

4

Crete,

Marathokephalo,

314

Laconia,

See sheepfold 145

manganese,

167-168,177,179-181,185,255, 263,299-301,305,313-317,319, 350

127

Kalami,

142,144,183-187,254,

301-303,311,330-332,344 Malia, 264-268,270, 304 M?ndalo, 158,164 Mandra.

17,20,41,112,142-143,145,

Kythnos,

139,282-287,289,292,296

magnetite,

Seriphos

Konya basin, 162

420,421,426-427 isophorone, 4-5 Isthmus of Ierapetra,

158,164 15

Magnesia,

167 root, 417-420,423,424-427

lime, 137,139 limekiln, 204 Limogardion, 299 limonite, 303, 310 limpet, 149-152,213 Linear B, 10,14,247,263 lions, 161

Mn

86-87,

290,292

oxides,

Mochlos,

96-97,

91,93-94,

4,26,28,

70,170,186,217

218,229,231,258-259,264,273, 318-319 modern

toponyms,

main

discussion,

8-14

Molai,

314

molds,

63-64,113,172

mollusks,

43

Monastiraki

Katalimata,

Monodonta,

149-152,213

268

81,

INDEX

186

ware.

mottled

299 Othrys Mountains, See AF 32 under oval structure.

245

montmorillonite, mortars and

461

pestles, See Vasiliki

Ware

features

pogenic

Mount Holyoke College, 17 daub

mud murex

Pacheia

62

structures, 314

shells,

176

Mushiston,

Palaikastro,

acid, 418,420,426

myristic

Fournou

Myrtos

Korifi,

171

Neolithic

62, 67

3, 8,21,42,

period,

70, 72, 79, 87,109,142,149,155 166,179,183,193,205,209,217, 225,233-234,236-238,244,248, 250,257-262,273,321,323,325, 380,429-432 racemosa.

Nepeta

111,

furnaces,

Nisyros, 158 N-methylphthalidimide, nonanal,

Pyrgos

198,237-238,270 Pyrgos Myrtos, 267

426

pepper, 420 peppermint, 426

426

acid, 419,423,424,

obsidian,

of slags, main

99-101,105,107-108,209,

octadecane,

423-426

417,419,421,

octanoic

166

Petromagoula, Peumus boldus,

oleic acid, 418,420,423, 426 olives (olive oil), 110,143,153-154, 235,241,246-248,250,252,255,

pigs,

258,260,262,276,278,365,414, 418-421,423, 424,426

pile-based

335, 338, 344 176

Oman,

421

1-decanol,

421

1-dodecanol,

1-pentadecanol, 1-tetradecanol,

421

421 425

1,2-benzisothiazole, oolitic or?gano,

structure,

resin,

pine

tar, 422,

pine

trees,

291

Royal

418,420-423,

425,426

426

halepensis.

425

See

pine

cup, 77,

Cemetery

87-89, of Ur,

97,413,425 169

rue, 65,418,425,427 See rue

Saddle-shaped

querns,

212-213,

241,271 tanks

safflower,

157

166,

422

saffron, 65,420,421,426,427 sage, 221,417,420 Salamis, 157,170 S argon 1,162

Platys River, 12, 37,229-230,255 Poliochni, 170-174,176, 319 Katsambas,

420

318

Aleppo

pithos, 378 PIXE, 321-324 Pl?tanos, 165,168, 318

Poros

417,

graveolens. rutile, 284

149,152 See fish

73

Ruta

391

Platomagoulia,

418

Origanum sipyleum, Orkos. See Kea

Pinus

piscinae.

421

rosemary, rounded

pineapple, 426

Pisania,

421,423

1-octadecanol,

62

See anise

pins, 156-158,172-174, 421

1,4-diacetylbenzene, 1-nonanol,

anisum.

419,421,423,426,428

159

Romania,

construction,

pine

125-126 wine,

ritual vessel,

212,234,271-272

Pimpinella 307,

resinated

Rethymno, 304 Rhodes, 316 ring-idols, 157,159

185

pharmacosiderite, 246 phosphorus,

Photolivos, 158 Phtiotis, 299

141,282,284-286,289,

425

Rekh-mi-re,

186,197,204,221,223,229,234

olivine,

153 171-172

raspberry,

Phaistos, 75,155,259,266-267,304

acid, 419,421,424,425,426

Rachis, Raima,

Redox Conditions, 348

426

424,

26

phacoids,

422

octanoate,

querns

discussion,

70

Petrokorio,

245

saddle-shaped

quicklime, 42,112

petrography

154

See

querns.

281-292 Oats,

331

332,344,351,430-432

4

Petras,

Mu, 268 183,284,286,290-292,

quartz,

petitgrain, 425

425,427

70,170,217-219

289,305,335,338,344 pyxis, 67, 70-73, 76,208-209 Quartier

419

harmala,

Perganum Peru, 163

9,14-15,197

140-141,188,282-283,286

pyroxene, 422

acid, 419,421-422

pentadecanoic

318

pyrite, 141, 301, 332-333

23 419,

63-66

structure,

Chrysokaminou,

Pyrgos Ware, 419-422

acid,

P?loponn?se,

418,419,421,424-425,427

nonanoic

149-152,213

pelargonic

338

apsidal

165,170,218,

157

peach, 421,426

346

cave,

Pyrgos

Patras,

pentadecane,

Nevali ?ori, 161 nickel, 186, 332, 338,340-343, 347,349-352 Nigerian

149,152

Patella,

pawpaw,

See catmint

of the

purpose

Paracentrotus,

157,164

2,245

pulses, 246-247,252,271 punch, 174, 318-319

228

Papoura,

terrace G,

Pseiran

161

papaya, 417,426

157,168,319

4, 8,18,21-22,26,28-29,

32,40,43, 70, 79,127,154,181, 215-217,231,245,251-252,258, 264-265,268,270,273,391

418

acid,

Panagia Skali, 229 panthers,

needles,

Pseira,

142,171

palmitoleic

Mishmar,

See bellows pot bellows. vessels. See jugs pouring ore 177 Precambrian deposits, in 19-20 Crete, precipitation

palmitic acid, 418,420-422,426

88,181,220,

Myrtos Pyrgos, 318

Naxos,

Porti, 168, 318

Priniatikos Pyrgos, 4, 73,265 318

4,125,155,267,

Palestine,

252,302

Nahal

7-8,17-18,

161

Pakistan,

156

Mykonos,

4-5,

Ammos,

179,181-182,233,254,258

65

hearth,

portable anthro

347

Satureja

cuneifolia. saws, 165, 318 sea level, 21,199

See Turkish

savory

INDEX

462 212-213

sealstone, sea shells. sea urchins,

Taurus

Semitic

See

289

terraces,

3-4,49,197,200,202-204,

Congolese 335

furnaces

Tharrounia

shale, 291-292

Thebes, Egypt, 125 61, 72, 78-79,

bowls,

84, 87,

91,97,413,415,419,425 and goats,

sheep

silver, 157,159,166-167,169,172, 175,304,310,314,316

403-412 319

spatula,

318

spearhead, Sphoungaras,

whorls,

spindle Sri Lanka

176

stannite, stearic

acid,

222,225-227,233,243,249,251, 253,257,260-262,276,330,356, 373

418,420,421,422

storage

discussion,

99-108

65

vessels,

67-68,

213 strawberry,

425

Strophilas, 157 sunflower,

417

416-417, decoration,

Syros, 162,165,167,169,171,

tankard,

176 171

acid,

cooking shells,

212

weasels,

234,272

wedges for drilling, 269 316

wells,

4,197,256,274-276,278,

355-358,363,374 wheat, 153,234,241,247,250,260,262

9,11,

34-37,

40,

White-on-Dark

Ware,

427

wild boars^l61 wine,

418

65, 71-72,

88-89,126,181,264-266

88

414,418-423,425-427

wollastonite,

305,

307,

309

wustite, 141, 309, 338, 345 See cooking

pots. 213

pots Xerambela,

227,229 See Lefkandi

Xeropolis.

157

Yannitsa,

Dryas,

Younger

162

127,133-134,136,172,183 Zaire.

184,187 tweezers

319

422

esters,

77-78,

tuyeres,

See pigs 167 swords,

swine.

Tajikstan,

wax

Spectrom

Dispersive 325-328

weaving,

Troy, 160,169-176 tulip shell, 149-152 tumbler, 208, 385 Turkish savory, 417,425,427

72,183,212

161-162

425

thyme, 41,391,417 Timna, 142, 335 tin, 143,163,167,169-176, top shell, 149,213

tripod triton

70,258 226

Vronda,

etry,

tridymite, 309

209

tools, main

Vrokastro,

Wavelength

stone

vases,

Vouliagmeni, 157 33,181,

4,197,252,274-276,

floors,

269

system,

278,355-356,359-360,365-366 Thriphti, 4, 70 Thriphti Argira, 228

threshing

420,421,425,426 420

vultures,

tridecanoic

stone

villa

62

triacetin,

stone

verbenone,

17

14

420,422,425-426

vervain,

3-bromothieno[3,2-c]pyridine,

trickle

71, 76-77,127,

422,427

period,

verbena,

7,11-12,22,27,

beach,

stirmp jars, 212 bowl,

dating,

oils,

vegetable Venetian

203

177

15-16,

300,304

3, 7-9,11-12,15

17,34,193,197-198,203,215-220,

cave, 165 Trapeza travertine formation,

70,215-218,258 65

furnaces,

cave,

Theriospelio

Thrace,

discussion,

Ware,

208-209,217-218,220,264-265,

229,231,233,254-255

166 main

Vasiliki

9,12,198,200,203-204,215

Tholos

See Kythnos

soap, 221 socketed spearhead, soils and sediments,

157

cave,

157

Varna,

Thessaly, 62,164,166

Sklavopoula (Sklavopoulou), 302-304, 310-311

171

Vasiliki, 4,70,181,217,258,265,268, 270

thermoluminescence

Siphnos, 179,314-317 Sitagroi, 156-159,164,171,404 Siteia, 24 Skhabar, 418

Skouries.

157,164,166

Thermi, 166,171,173-174,176

166

pendant,

418,420,422

acid,

Usatovo,

220

212

silver zoomorphic

cave,

Theopetra Therio,

212,234,248-249,

253-254,271-272,276,420,423 sheepfold, 200-201,211, 356, 370 Shiqmim, 142,177 shrew,

undecanoic

Ur, 169-170 Urfa, 161

417,419,421,424

Shahr-I-Sokhta,

shallow

418,425 418-419

undecane,

tetradecane,

140-141

minerals,

undecanal,

tenorite,

224,246,251,274-278,355,357 358,360-361,363-365,369,372 374,399-402,406-408,411-412

314,316-317 Sesklo, 164,166 Sfaki,299 Shaba.

Unaltered

89,175

Tell Brak, 170

20,142-143,179-181,299,

Seriphos,

419,425-427

Tylissos, 270

23,161,167,169

65, 73,

teapots,

See Sklavopoula loan word, 10,14

Selinou.

418,

2-undecanone,

Mountains,

170

43,149,152,254,419 302

Selakano,

171

Tarsus, resources

See marine

2-decanone,

Zakros. Zas

425-427

2-ethylhexanoic 418 2-ketone, 2-Phenoxyethanol,

acid,

420,422,424

See Congolese furnaces See Kato Zakros

cave,

zoomorphic 419,

423

157,164,168

Zoghaki. SeeKythnos morphic

pendant. pendant

See silver zoo