Swartkrans: A cave's chronicle of early man

wartkrans ACave's Chronicle of Early Man

c. K.Brain

November 1993 Transvaal Museum Monograph Transvaal Museum Monografie Transvaal Museum, Pretoria

N8 O.

Transvaal Museum, Pretoria 1993

e

First edition, first impressicn Eerste uitgawe, eerste drul<:

ISBN O 907990 142 rSSN 0255-0172

© Al! righls resarved. Apart trom citatlons for the purposes ot resaarch or review, no part of this book may be reproduced in any form, mecnanlcat or electronic, including phatocopying and recording, without permission in writlnq trom the publisher.

© AlIe regte voorbehou. Geen qedeene van hlerdie boek mag gereproduseer ward op enige manier, meganies 01 elektronies, insluitende plaat- 01 bandopnames en fctokcplérlnq. sondar die skrtttelike toestemming van die uitgewer nie, behalwe redelike aanhalings vir navorsinqs- en resesensiedoelelndes.

Copies ot this Monograph are omeeiaere fram the Transvaal Museum Bookshop, P. O. Box 413 Pretoria, 0001 South Atrica Kopieé van hierdie Monogralie is verkrygbaar by die Transvaat Museum Boekwinkel, Posbus 413, Pretoria, 0001 South Africa

Transvaal Museum EditorlRedakteur: Mrs A. Dreyer

Layout & design, tecnncat editing and crccucncn byl Uit/eg & ontwerp, tBgniese redigering en produksie deur Dr Nico J. Djppenaar, Ist8g Scienñttc, Irene Imageset bylBeelds9twerlt: deur Photoprint, Cape Town Printed bylGedruk deurCTP Bcok Printers, Cape

For Laura

Contents Acknowledgements. .

vi

Foreword - F. Clark Howell

. . . . . . . . . . . . . . . . . . . . . . . . vii

Introduction - C. K. Brain A Pictorial Overview of People and Events at Swartkrans 1948-1992 - C. K. Brain .

7

Chapter 1. Structure and Stratigraphy of the Swartkrans Cave in the Light of the New Excavations - C. K. Srain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23

Chapter 2. Composition of the Swartkrans Bone Accumulations, in Terms of Skeletal Parts and Animals Represented - Virginia Watson

. .. 35

Chapter 3. Description and Preliminary Analysis of New Hominid Craniodental Fossils from the Swartkrans Formation - Frederick E. Grine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..

. 75

Chapter 4. Hominid Postcranial Remains from Swartkrans - Randall L. Susman . . . . . . . . . . . . .

117

Chapter 5. Additional Fossil Equidae from Swartkrans - C. S. Churcher & Virginia Watson . . . . .

137

Chapter 6. New Fossil Carnivore Remains from Swartkrans - A. Turner .

151

Chapter 7. Stone Artefact Assemblages from Members 1-3, Swartkrans Cave

.

-J.D.O~

167

Chapter 8. The Swartkrans Bone Tools - C. K. Srain & Pat Shipman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Chapter 9. The Incidence of Damage Marks on Swartkrans Fossil Bones from the 1979-1986 Excavations - Rosemary Newman

.

217

Chapter 10. The Occurrence of Burnt Bones at Swartkrans and Their Implications for the Control of Fire by Early Hominids - C. K. Srain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

229

Chapter 11. Chemical Characterization of Burnt Bones from Swartkrans - Andrew Sillen and Thomas Hoering .

243

Chapter 12. Stable Carbon Isotope Studies of Swartkrans Fossils - Julia A. Lee-Thorp and Nikolaas J. van der Merwe

.

251

Chapter 13. A Taphonomic Overview of the Swartkrans Fossil Assemblages - C. K. Srain

257

Index . . . .

265

Acknowledgements The longterm Swartkrans project was undertaken while I was Director 01 the Transvaal Museum ano, consequently, it had to be accommodated among other responsibilities. This accommodation would not haya been possible without the support 01 my famiJy, who helped with all aspects 01the operaticn. In the early years, my wife, Laura, extracted tossils from tretr matrix in a makeshift, backyard Jaboratory. Our sons Tim and Conrad helped with the excavatian process and did many experiments with bane tools and the effects 01 tire on modern banas. 01 our two daughters, Aosemary Newman, prepared, numbered and cataloqued a vast number 01 fossil banas and did an SEM study 01 damage marks on them (Chapter 9), while Virginia Watson helped with the excavatlon of Member 3 and did a complete analysis of all the fossils from the 1979-1986 excavations (Chapter 2). Thia willing family participation made the whole project a pleasure. At the Transvaal Museum, my load as Director was lightened by the help of many people, but it was particularly the support of Miems Erasmus that provided me with time for research at Swartkrans. During my directorship, I had the support and encouragement of the Museum Board, and my particular gratitude is due to its then Chairman, Fritz Etoff, and Vice-Chairman, Gerrie de Graaff. I am fortunate to still enjoy such support and encouragement from our new Director, Naas Rautenbach, Assrstent Directors lna Plug and Len Prinsloo, and the new Council chaired by Nok Frick. The project would not have developed at all had it not been for the early support of the Wenner·Gren Foundation for Anthropological Research in New York, and for the personal interest of its then Director of Research, Lita Osmundsen. Her Foundation also funded the 1976 Burg Wartenstein symposium on Taphonomy and Vertebrate PalaeoecoJogywhlch consolidated the discipline of taphonomy in Africa - an issue central to the Swartkrans investigation. Once started, the project was given annual support first from the Council for Scientific and Industrial Research and latterly from the Foundation tor Research Development. People involved here who I would Iike to thank particularly are Willie Weideman, Rein Arndt, Gideon Louw and Gudrun SChirge. When we were three years into the project, Swartkrans became the property ot the University of the Witwatersrand. but my colleagues there were very generous in alJowing the work to continue and in giving me every encouragement. My grateful thanks are due to many people, but particularly to Robert Charlton, Frtedel Sellschop, Peter Tyson, Mike Raath (who has sinee moved to the Port Elizabeth Museum), Bruce Rubidge, James Kitching, Phillip Tobías. Alun Hughes, Ron Clarke. and Maeiej Henneberg. At Wits I have much appreciated my conneetion with the Bernard Price lnstitute for Palaecntoloqical Research, the Department 01 Zoology, and the Palaeoanthropology Research Unit. I have received mueh hetp from the unrt's Director, Phillip Tobias, and severa' of its members, particularly TIm Partridge and John Voger. At the Transvaal Museum I have benefited from my contact with a series of Curators of Palaeontology, starting with John Robinson, who introduced me to Swartkrans and its rlches. Eüsabeth Vrba added a great deal to the proper understanding of the Swartkrans faunas and to my enjoyment of those early excavation days, Our current Curator, Francia Thackeray, started his Swartkrans connections in the early 1970s and has been very supportive ever stnce, For many years, David Panagos has been conecten Manager in the Palaeontology Department, and he has assisted with the Swartkrans work in a variety of ways. Gilt Turner helped a good deal with the time-consuming work of cataloguing tosslls, as did Ellzabeth Voigt, while the stratigraphic studies of Karl Butzer were invaluable to our understanding of the complex sequence of deposits. For over 20 years we have had a field team based at Swartkrans and this was supervised Initially by Absalom Lobelo and Obit Sepeng. In 1970 George Moenda [oined the group and has been with us ever since. He has helped with all the excavations and built the extensive stone walls and steps that now make the site accessible. In fact, we have now arranged a raute for visitors around the whole site, with 25 beacons featured in an iIIustrated guide (Brain and Watson, 1992. Annals ofthe TransvaaJMuseum 35: 343-365). In September 1992, Swartkrans was declared a National Monument and, for assistance with this development, I am indebted to Janette Deacon and Johan Loaek. Particular thanks are due to all the collaborators in the Swartkrans project whose detailed studies have led to chapters in this book (in arder of appearance): Virginia Watson, Fred Grine, Randall Susman, Rufus Churcher, Alan Turner, Desmond Clark, Pat Shipman, Rosemary Newman, Andrew Sillen, Tom Hoering, Julie Lee·Thorp and Nick van der Merwe. The time thal they have so willingJy spent on their various investigations is very mueh appreciated. lt is a pleasure to thank Clark Howell for agreeing to write a Foreword to this book. His association with Swartkrans is a long one, extending aven befare his visit to the Transvaal Museum in 1970, when he brought his then graduate student, Don Johanson, with him for a study of our hominid specimens, and co-authored a paper on early Horno at Swartkrans. His support and encouragement is very much appreciated, while Don and Lenera Johanson have recently been baek to Swartkrans in the eourse of making a doeumentary on human origins. Jn conclusion, my thanks are due to Nico Oippenaar, Anita Dreyer and Liz Herheldt for their very competent seientific editing and production of this volume.

vi

Foreword Almost six years ago I happily agreed to a raquest from Bob Brain to produce a Foreword for a planned new volume to be devatad to sorne aspects 01 continuing, intensified investigalions at the Swartkrans hill palaeontologicallocality. One thing and another haya delayed the volume that is now on the ave 01 publication. I haya no doubt that workers in many palaeoendeavours, and not only palaeoanthropology, will fínd that it has been well worth the walt. The outlines and sorne preliminary concíustons were given by Brain ano seven collaborators five years ago (e. K. Brain el al., South African Journal ot Sclence 84(10): 828-835. 1988). whereae Braln and Watson (Annals ofthe Transvaal Museum 35(25): 343-365. 1992) only last year produced an informativa, well-illustrated guide to the Swartkrans locality, with extensive documentation 01 its geologieal hlstory, complexities and organic residues, on the occasion 01its declaration as a National Monument. I have known Bob Brain only atad short 01 forty years. In 1954, in the course of a seven-month study tour in sub-Saharan Afriea, 1spent upwards of a month at the Transvaal Museum, Pretoria. In my time there, and before with eolJeagues lrom Johannesburg, I was enabled to visit all the australopithecine-bearing loealities, and to obtain sorne firsthand appreciation of their settings and ínfflls, and particularly 01 the difficulties 01 unravelling thetr respective stratigraphic successions and taphonomies due to tneír extensive exploltatíon and attendant disruption by Iime-quarrying activities in every ínstance. Such activities are encouraged by the preclpitatlon, subsequent to cavern dissolution, ot calcite-rich flow and dripstones within motst, air-filled cavities through evaporative processes of lime-charged waters in the vados e zone and henee befare subsequent infill by external sediment derivatives through enlarged [oints. bedding planes and attendant roof taus. Braln had shortly befare set up requisite laboratory facilities, through wenner-Gren Foundation support, to carry out sedimentological analyses in addition to his stratigraphic studies of all the Transvaal localities. (The lamentad Frank Peabody's substantive stucty of the Thabaseek travertine sequence, near Buxton, Cape Province, in the Harts valley was soon to appear, and would cast sorne light on the Taung infilling, the first such fossiliferous locality to be recognized). Brain's efforts, satisfying his doctoral (1957) requirements at Cape Town, and soon published as a major monograph (The Transvaal Ape-man Bearing Cave Deposits, Pretoria. 1958), constituted a significant, indeed essential, step lorward in elucidating not only stratigraphic history of sueh eavern infills, but particularly in the clarification of sourees and processes of sedimentation from an actualistic perspective. His perspective on solution processes and cavern formation drew substantially on the researehes (in the Ozarks) of J. H. Bretz, of ehannelled-scabland fame, to whose ideas and work I had been exposed in student days at the University of Chicago. At this time John Robinson was wholly engaged on his own doctorate, focussed on the australopithecine dentition

(The Dentition of trie Australopithecinae, Pretoria. 1956). 1, thus, tor the most part studied cranial and postcranial morphology of the Transvaal Museum australopithecine collections: beyond an abstract, my own observations in the last respect remained unpublished as he contemplated a major volurne on all such materlats then known, the publicatícn of whieh I subsequently strongly encouraged by my own University's press (Earfy Hominid Posture and Locomotion, Chicago. 1972). I have never forgotten mose days at the Transvaal Museum, when new friendships were made, my own budding interests in Africa were set in motion, and the pcst-bellurn expansion ano intensification of palaeoanthropological studies were patently evidenced. A few years later, John Robinson, Phillip Tobias, Kenneth Oakley and Ronald Singer, all principal figures in African palaeoanthropology, were to be visitors to the University of Chicago, and elsewhere in the United States, for substantial intervals. From 1961 to 1965, Brain was on the staft of the National Museums of Rhodesia (in Salisbury, now Harare, Zimbabwe), tocare ot his birth in 1931, where he served as Keeper of Zoology and Deputy Director. He returned to the Transvaal Museum in 1965 as Curator ofthe Palaeontology Department, and assumed the museum directorship in 1968 (serving until 1991). We mel again, after a ten-year gap, in July 1965 al the massive and prolonged Burg Wartenstein (Austria) conferance (published as Background tor Evo/ution in Aldea, Chicago. 1967). He was eager to institute a rnajor research eftort at the Swartkrans locality; I was, everlastingly hopeful and persistent, seeking approval in Ethiopia (with help from Louis Leakey) to set an amo Research Expedition (ORE) underway (finally aulhorized in mid-1966). I made three subsequent visits to South Atrica, in 1969, 1970 and 1975, all stimulated by the ORE endeavour. My research goals were focussed on early hominid craniodental morphology, and thus a continuation and expansion of concerns of sorne fifteen years earlier. In eaeh instance I became apprised of the renewed eftorts at Swartkrans under Brain's guidance, and (ultimately) the programme under Phllip Tobias' direction, and the late Alun Hughes' supervision at Sterkfontein. The consequences of each were to transform previous and then prevailing coneeptions of these locales, their geologic histories, infillings, faunal compositions and hominid representatives. In the tirst instanee R. J. Clarke and I made the trip together, from Nairobi via Dar-es-Salaam. Elisabeth Vrba, recently arrived at the Transvaal Museum and initiating work on extant and extinct Bovidae, discussed her work and intentions with me at sorne length. From my friendship and association with Alan Gentry, and experience in the ORE field programme, I felt jt only approprjate to observe that she was undertaking a task of uncommon magnitude, and one demanding of extraordinary morphological experience and comprehension. Her subsequent dissertation and monograph (The Fossil Bovidae of Sterkfontein, Swartkrans and Kromdraai, Pretoria. 1976) damonstrated beyond any doubt both the vii

FOREWORD

substance 01my own observations and her own sxtraordtnary capabilities to produce results 01exceptional quality, originality, -,-o uncommon scientific rigour and 01broad importance in raspeet of palaeobiology, bioslratigraphy and palaeoecology. On lhis occasíon Ron Clarke (initially) and I recognized the exact conjoining as a single individual (from the Hanging Remnaot, Member 1), of the SK 80 (considered Horno) and SK 847 (considerad A. robustus/crassidens) hominid specimens. Ultimalely published by Clarke, Howell and Brain (and further expanded on, and defended by Clarke and myself, and analysed in detailin the former's dissertation) this was to be the death knell 01an iII-conceived 'single-species hypothesis' which confounded and obfuscated hominid evolutionary studies al that particular momento (At the annual meeting 01 the American Anthropological Assoclañon in late 1969 in New Orleans, I was severely berated by several conlirmed and insistent adherents 01 this ill-conceived, and unlruitful schema upon presentation 01 this singular evidence; it required just snort of a decade for the issue to be finaily slified as a consequence 01 accumulating, incontrovertibly supportive evidence from lhe (upper) Koobi Fora Formalion, Easl Turkana Basin.) I considered then, as 1 did earlier on the occasion 01 the subsequent objections to the proposal 01 Horno habilis spec. nov., that there prevailed substantive differences (mind sets) between tne perceptual worlds 01 sorne students 01ancient Hominidae and those 01 others. Evidently the normal science of one did nol at all malch up with lhe hypolhesized expectations ot another and the consequence was an essential tension leading to controversy, acrimony and misdirection of research goals. This did not then, nor does it now seem to me to be a fruittul and positiva way to conduct supposedly scientific endeavour since possible, even probable alternatives were being arbitrarily excluded out of existence by the assumptions and precepts mandated by sorne (fortunately few) workers. In a different torm, and in another context this same problem is very much with us still today. Surely, efforts have to be made and ways sought to escape the constrtctíons posad by this uncommon and exclusionary way of thinking. Upon returning to Nairobi, en route back to the field, I was informed by radío-tetephone of a helicopter crash and lsolation, on the opposite side of the amo, of a substantial group of expedition scientists. The cool-headedness of a superb woman pilot, ferrying a few individuals at a time from a miniscule, makeshift field strip al the Yellow Sands (Mursi Formation) locality, saved the occasian from one of bereavement to one 01 rejoicing. Here, as on numerous other occasions with ORE, Frank Brown lackled the problem head-on, crossing lhe amo in a native's dugout, reaching a radio-equipped vehicle to contact camp, and, ultimately, by Single Side-band radio to inlorm Nairobi of the mishap. Karl Butzer carne atthis time to South Africa, afterthree field seasons wllh lhe ORE, as did al so Richard Klein, one a colleague and close friend and the other a former student. Each was lo begin highly successful, prolonged and muchneeded research programmes within the frameworks of geochronology and faunal analysis, respectively. The former conlributed signilicantly to early hominid researches through his expanded geomorphological and sedimentary studies in the Harts valley area, investigation of formation and ¡nfill viii

==z;:.

'.C:_c

.

processes at Sterkfontein, as well as to the elaboration of a formalized stratigraphic framework at Swartkrans, within the emerging vistas afforded by Brain's research programme at the site. The next year D. C. Johanson accompanied me to the Transvaal Museum and to the Witwatersrand University in Johannesburg. Over sorne weeks we made an exhaustive morphometric study 01 all early hominid dentitions in these institutions. These results, and similar studies 01 various collections at lhe National MuseL!m of Kenya, Nairobi, were never published as such, but were employed subsequently In comparative studies 01 hominids botn from the Omo Group formations and from Hadar. I continued, and expanded on this work in 1975, both in Nairobi and at the Transvaal Museum and the University of the Witwatersrand, after a visit to Addis Ababa to return amo fossil hominids, on loan for study, to the National Museum ot Ethiopia. The work 01 the ORE was presented in lectures al trie first two institutions; 1recaü sorne unease at the latter when the possible affiliation 01 the Taung individual to a 'robust' australopithecine Iineage was provocalively, if hypothelically, proposed. F. E. Grine had only recenlly arrived at the Witwatersrand University and was soon to begin his extensive studies of the deciduous dentition of modern and extínct African hominids. Eiisabeth Vrba had jusI completed her doctoral thesis, with wonderfully significant results (to be presenled lhe following year al lhe vast UISPP congress in Nice). The work at Swartkrans, now in Brain's second 7-year phase, had moved ahead rapidly and fruitfully, with extensive overburden removal underway, clarification of stratigraphic aspects (including recognition ofthe fossiliferous 'pink breccia' as a Hanging Remnant of the lower infill), and elaboration ot programmes of faunal and taphonomic analyses. Similar progress was evidenced at Sterkfontein, under the watchful eye of Alun Hughes, particularly in the old 'Exlension pit,' and where the stratigraphic succession was being rapidly elaborated, and the STW-53 cranium, a Homosp. with artefact associations lrom Member S, was only a herald of the unexpected and diverse hominid abundance emanating lrom Sterkfontein in subsequent years and until his reeent sad passing. Brain and 1 participated together at five subsequent gatherings ot palaeoanthropological focus. The first in 1974, a colloquium organized by Clifford Jolly under the auspices of the Wenner-Gren Foundalion, in New York City (published as Early Hominids 01 Alrica. 1978). The second in 1975, an annual meeting and atiendant symposium of the L. S. B. Leakey Foundation, in Washington, O.C., following which Brain visited several institutions, including the University of California, Berkeley. The third, the Firsl Internalional Congress on Human Palaeonlology, held in Nice, autumn 1982, whlch initiated a new organizatian 01 international scope and membership under the remarkable vigour and foresight of Henry de Lumley. The fourth, Ancestors. The Hard Evidence (published in 1985), held al the American Museum of Natural Hislory, New York City, in 1984, a once in a lilatime exhibition of hominid fossils and simultaneous symposjum of palaeoanthropologists and affiliated scientists, masterminded by Eric Delson, lan Tattersall and John van Couvering. And lastly, Evolutionary History 01 the Robust Australopithecines (published 1988), a symposium/workshop held al the State Universjty of New York, Slony Brook, spring 1987, of supreme importance

FOAEWOAO

and exceptional impact with respect to cntícal analysis and comparativa evaluation of African early hominid structure, diversity, systematics, geochronology and palaeoecology, meticulously organizad and directed by Fred Grine. In The Hunters or the Hunted? An Introduction to African Cave Taphonomy(Chicago, 1981), Brain brouqht toqemer tbe results 01 more than a quarter century of personal research into the problem 01 investigating, analysing and deciphering sotutíon cavities, their origins, infillings and fosslliferous contenis. This volume uniquely revealed the resolving power 01 a combination 01 actualistic and histortcal approacnes, each 01 whích he had pioneered in Southern Africa. With the publicatíon a deeade later of Peter Andrews' Owls, Caves and Fossits, focussed largely on north temperate latitude situations, eoncerned students and investigators in palaeoanthropology, and a host o, associated endeavours, have at hand two resources of uncommon clarity, sccpe and íncísiveness from which to view such important sources for the vertebrate fossil record, including of eourse hominids. These studies are in the tradition of Buckland, Weigelt and Efremov, and refleet the remarkable impact of broadly conceived and executed taphonomic ínvesfiqañons. Brain's volume carried us through the second 7-year phase ot his renewed study ot the Swartkrans locality. Swartkrans: A Ca ve's Chronicle of Early Man here reveals the outcome of the third 7-year phase (it is thus appropriately read in conjunction with the chapter on Swartkrans in the 1981 volume). Thirteen contributors in an equal number of chapters discuss in detail significant additions to knowledge of this very impressive fossiliferous locality. The sequence of infills at Swartkrans has been a maln research tocus sinee the first year-Iong fossil exploitation there by R. Broom and J. T. Robinson in 1948-1949. This problem was exacerbated by subsequent commercial lime-quarrying activity until over ayear later, following Broom's death, when palaeontological and geological studies were resumed by Robinson late in 1951. Brain's first intensive work on infills were carried out between 1953 and 1957. The presenee of a large fallen roof block, around which a basal travertine had developed on ¡ts southwest margins, dividing the cavern into Lower, and Inner and Quter caves, was early recognized, but the spatial distribution and temporal relationships of infill units remained difficult to decipher until miners' rubble was c1eared, enclosing walls revealed, and external openings either evideneed or inferable. Thus, there is a long history and substantial publication record of initial1y Robinson's, and there· after Brain's efforts (eoupled with those of Butzer in the mid1970s) to resolve this fundamental issue. Once it was recognized that the hominid-rieh lower 'pink-breccia' (Duter Cave) was an infilled remnant (Hanging Remnant, Member 1) the ultimate resoJution was in sight, ¡t not yet in hand. (The fossil content of this unit was discussed in extenso in Brain's 1981 book, and is not really treated in detail again herein.) The earlier ¡nfill thus comprises a Basal Travertine (K.W. Butzer's lA) and Member l. Member 1comprises a Lower Bank Unit (LB) more recently recognized, representing a very steeply dipping talus eone trom an external shaft whieh infilled the southeast chamber area with orange·coloured, poorly calcified fossiliferous, stony breccia; a major erosional interval and flushing out of earlier infiU via an external connection between south and north walls; and a Hanging Remnant Unit

(HR; KWB's lB) 01 lossil-rich pink breccia inlilling the Outer . Cave and derived from a north wall shaft The Lower Bank untt has been extensively and very carefully excavated, and yielded abundant fauna, hominids, lithic artefacts and utilizad bone. It and later infill units are a major focus of this velurne. The LB---?HR temporal relatianship appears mast probable on stratigraphic and related grounds, but Brain eonsiders that sorne (?slight) faunal evidenee might even suggest a reverse arder of accumulation; probably, for the moment, (he evidence is insuffieient to satisfactorily and completely resolve this issue. Member 2, unconformably separated from its anteeedents, eomprises brown breccia (Outer) and stratified brown breccia (lnner) infilling both cavern eomponents. This is a heavily calcified infill, rieh in fauna, including hominids, and yielding artefacts, derived from an external opening between the south and north walls. Member 3 is a partlally preserved (c. 20 x 5 m) steep-sided gul1y. with some 6 metres of infill situated along the west wall ofthe cavern system. eroded into and underlying wíth sharp unconformity the earlier Members (1, 2). lt is quite fossiliferous, yielded sorne (not many) hominids, evidence ot hominid presence and even occupation in the form ot Iithic artefacts and utilized, cut and burnt bone. Two other Members testify to later Pleistocene infilling. Member 4 occurs in the northeast corner of the cavern, and ls rich in Middle Stone Age llthic artefacts, but may lack bone; it remains unexcavated. Member 5 is an Infllled channel, adjacent to the cavern's northwest waü, and termed the 'Antidorcas bondi channel' fiU due to the abundance of that extinct springbuck; it has a e 14 age 01 c. 11 000 years. The faunal representation of individual Members is a malter of major concern, and three chapters deal specifically with this issue. The summary table here compiled from Virginia Watson's chapter and Brain's earlier (1981) treatment of the Hanging Remnant assemblage, affords data on minimum numbers of individuals (MNI) and number of ldentified specimens (NiSP) values by family or, il approprlate, higher taxonomic category. Primates (including hominids) are substantially represented throughout, least frequent in Member 3 (13,6 %), between 17-20 % in two other Members, and uneommonly abundant (46.7 %) in HR where baboons (four species) are about as common as hominids (two species), an altogether unique eircumstanceto my knowledge. Such a situation. coupled with other data, led Brain to explore and elaborate a hypothesis of carnívore, particularly leopard predation as a major factor in producing such an accumulation (and as witnessed by distinctive damage here analysed by Rosemary Newman). Among the Cercopithecoidea, papionínes always predominate, theropitheeines are usually present but not eommon, and eolobines are absent or at best very rare. Carnivores (five families) are well represented, lesser forms usually less commonly than larger taxa, but (except for Member 3 = 18,8 %) have lower colleetive MNI percentages (though higher NISPs), between 12-16 %, than do primates. The relative numbers offelids, hyaenids and canids varies from Member to Member, with a rather inordinate number of caníds (both MNI and N1SP, of three species) in Member 3. The HR of Member 1 yields high numbers 01 both felids (three species) and hyaenids (four species), but in fact has the lowest overal1 relative percentage (12,2 %) of this

FOREWORD

arder 01 all the Members. The number 01 leopard relativa to (all) hyaenid Individuals is equal or almost so (LB = 4:5, and in which coprolites were very abundant;

Man:~~lían

taxa (families) ene thetr frequencies et Swartkrans.

laxa (number of specíes)

MNI (NISP) Member 1

Member 2 Member 3 Member 2 =3:3: Member 3 =5:6) except LB HR in HR (12:18) in which hyaenids predomínate. Except for leopard there is an PRIMATES Hominidae interesting differential representatian 01 Paranthropus crassidens (1) 13 (19) 17 (26) 87 (218) 9 (12) large cats - thus, Iion (only LB), cheetah Horno sp. (1/2) 1(1) 3(4) 2(6) (only Member 3), Dinofelis (only HR) Pcstcranfa -(16) -(10) -(9) -(12) and Megantereon (missing ir¡ LB and Colobinae (1) 1 (1) Member 2), all of which are represented Papionim (3) 9 (66) 103(176) 9 (75) 11 (95) Theropithecini (1) 1 (1) 17 (31) 1(2) 4(4) by occurrences of single individuats. CARNIVORA Homotherium, interestingly, ís unknown Felidae (8) 7 (50) 14 (32) 2 (41) 8 (58) at Swartkrans. Three largar hyaenids Hyaenidae (3) 4 (27) 19 (47) 4 (44) 7 (91) Canidae (4) are representad - brown hyaena 4 (58) 6 (11) 5 (50) 9 (128) Mustelidae (1) 2 (6) 1 (1) 2 (5) \hroughout (1-3 individuals), extinct Viverridae (5) 2 (7) 4 (6) 7 (28) Chasmaporthetes (absent in Member 2, PHOLlDOTA (1) 1 (1) and abundant in HR) and spotted 1 (1) TUBULlDENTATA (1) 1 (3) 1 (1) PROBOSCIDEA (1) 2 (4) hyaena (absent in LB. and numerous in 1 (2) HVRACOIDEA (2) 20 (271) 26 (62) 13 (221) 16 (249) HR). Aardwolf is represented (rarely) in PERISSODACTVLA all Members. Among canids, bJackEquidae (3) 7 (10) 3 (30) 4 (20) 7 (75) backed jackals occur consistently (3-5 ARTIODACTYLA Suicae (2) individual s), faxes are not common, and 2 (4) 7 (8) 4 (18) 2 (41) Hippopotamidae (1) 1 (15) 1(1) a large indeterminate canid (not necesGiraffidae(1/2) 1(2) 1(2) sarily African wild dog) occurs only in Bovidae Members 2 and 3 (and also al KromAlcelaphini (4) 12 (102) 50 (188) 17 (143) 32 (113) 18 (41) Antilopini (6) 15 (23) 18 (66) draai). The clawless otter is documented 35 (140) 2 (3) Bovini (1) 3 (13) 2 (7) 3 (9) in three Members (absent in HR of MemReduncini (1) 1 (1) ber 1), but otherwise mustelids are Tragelaphini (2) 4(12) 1 (2) 2 (9) poorly known. Hippotragini (1 +) 1 (2) 1 (5) 3 (10) Pelini (1) 2 (2) Ungulates (sensu lato) represented 2 (6) Ovibovini (1) 3(10) by a single perissodactyl lamily (EquiSubtotal (Bovidae) 32 79 77 39 dae) and tour families of artiodactyls Bovidae r (0-23 kg) (146) (67) (36) (56) generally dominate the several assemBovidae II (23-84 kg) (330) (731) (372) (230) Bovidae 111 (84-296 kg) (128) (520) (300) (361) blages (between 33-51 % 01 the total), Bovidae IV (296 kg) (17) (30) (13) (59) except ofcourse in HR in which primates Total (Bovidae) (2381) (3429) (2529) (3561) are overall more numerous (46,7 % ve LAGOMORPHA (1) 9 (74) 4 (28) 7 (70) 34,5 %). Equids occurthroughout. in low RODENTIA (Large) Pedetidae (1) 1 (5) 1 (2) 1 (3) to moderate numbers, with Hipparion Hystricidae (1) 2 (14) 3 (6) 1 (3) 2 (5) (single individuals) and Equus capensis (more common) in all Members, but Burchell's zebra appearing only in MernMember 3) Connochaetes is uncommonly numerous (19 and ber 3 (a single individual). Suids (two species) occur throughout, but are few in individuals (except HR with seven). Glraffids 17 individuals, respectively). The only other substantial representations of another alcelaphine is Member 2 (Damalisare known only in the later two Members, single indivlduals only, and presumably may include Sivatherium. Hippopotacus « 6) and, especially, in Member 3 (Damaliscusand another mus is recorded only in LB and Member 3, again as single alcelaphine :::: six each). With respect to Antilopini, the total number of individuals is closely similar in the three earlier individuals. Members, but about doubled in Member 3. Species of Gazella Of the eight bovid tri bes represented, most exhibit little or of Antidorcas (two species) clearly domínate, but in different diversity, mostly single species, and in fact always occur in low combinations according to Member. Member 3 stands out not frequency. Exceptions are the alcetaphines (?six species) and only in the high number of antilopines (rather more than antllopines (si x species) which accordingly dominate the analcelaphines, and thus the reverse ot HR Member 1), but al so telope assemblages, but differently so according to Member. in the simultaneously elevated numbers ot individuals of Except in the HR unñ (13:28), antilopines are more numerous larger-bodied alcelaphine taxa. In lact, the Member 3 bovid than alcelaphines, although there are a nearly comparable assemblage has a different aspect, in composition and trenumber of species (6 vs 5) ot each, but more alcelaphine (six quencies, comparad with such assemblages trom earlier probably, ot which tour are extinct) trian antilopine (four) genMembers. Megalotragus, a large extinct alcelaphine, occurs in era. The maximum diversity of alcelaphines in any Member is three upper Members (lacking in LB), with three individuals four (LB - Member 1 has only three), and in two (HR and

x

FOREWORD

each in HR and Member 3. Other very larga bovids, Syncerus, Taurotragus, and Hippotragus niger are also somewhat unusual in occurrence. The first occurs throughout, with three individuals each in HR and Member 3. However, the last two species are both absent from LB and HA, and occur only in Members 2 (1, 1) and 3 (1, 3), respectíveíy. An ovibovine (ct, Makapania), several taxa 01which are known in Miocene· Pliocene faunas af Atrtca. is only documentad (three individuals) in the HR unit, Finally, it should be mentioned that Elephas sp. (presumably E. reckl) occurs in both LB and in Member 3, but represented only by ivory fragments. And, lastly, hyraeoids (two speeies) are represented in all four Members, and most abundantly in HR (6) and LB (20), although there are numerous skeletal elements in other Members as well. The ages 01the successive mfllls at Swartkrans can only be inferrad, until now, on the basis of macrovertebrate biostratigraphy. Bovidae, and Suidae to an extent, have been most informative in this regard. E. Vrba's initial estimate was between 2 and 1 m.y.a., tor the older inlill (HR) as then known, and was subsequently reduced to c. 1,8-1,5 m.y.a., by closar comparisons with radiometrically dated fossil successions in eastern Africa. The discovery and fuller exploitation of additional intill units, all of which share a number of large mammal species, continues to offer bread support far this lnterpretañon such that a time-span between 1,8/2,0-1,0 m.y.a. is a reasonable approximation. Conceivably this might be confirmedl refined by palaeomagnetic sampling and comparison with the Geomagnetic Reversal Time Scale, although the mode of inñll and sedimentary matrices may preclude obtaining magnetic irnprints. Other radiornetric/lsotopíc methods have proved unsuitable or inapplicable in these circumstances, although recent refinement of electron spin resonance age assessment may enable that method to be applied in the future. The overal1 rnacroteunal spectrum, as Virginia Watson here indicates, is suggestive of highveld grassland among low rocky hills adjacent to which riparian woodland savanna and vlei-like situations centred on a (periodically) enlarged more active Blaaubank stream in the valle y floor. Ultimately, a fairly detailed documentation of habitats may be expected to emerge from full identificatian and analysis of the abundant microvertebrates and avian remains represented in the several Members. Brain considers that the infills were coincident with interglacial intervals and that major episodes of tlushing out and erosion through the cavern system coincided with glacial intervals. If so, these would have reflected those short·cycle (40 000 years) intervals of the earlier Pleistocene, before c. 1 m.y.a., as exemplified in the astronomical scale, and as well documented in sorne pollen successions in (mostly) northern latitudes. Ultimately, further investigation of such Iinkages might well be pursued in respect of the whole known suite of cavern infilJings in the Transvaal. Although no! the type loeality (whieh is Kromdraai B), Swartkrans has produced the largest single sample of hominid remains representative of the distinctive Paranlhropus clade within the australopithecine greup. This occurrence is considered by F. E. Grine, and by me (as once al so by 8room and Robinson). as a distinct species, P. crassidens. It clearly differs in sorne dental morphometrics, and to an incompletely k.nown extent in craniofacial morphology from its counterpart at Krom-

. draai B and type species. P. robuslus. The distinctions from, 'but cladistic affinities with other species in eastern Africa, P boisei and the less well-known P. eemiopicus, are now well known, and are generally recognized and broadly accepted by investigators of early hominid palaeontology who appreciate their systematic meaning and phylogenetic significance. The largest, most diverse sample of P. crassidens is that obtained from the HR unít, early in its exploitation, and this has subsequently grown to at least 90 individuals. Additional cranialJgnathic/dental remains of the same identical species are now known and described here by Grine (and 1988, 1989) from the LB (some 14 indlviduals), Member 2 (upwards 01 19 individual s) and Member 3 (nine individual s) units. Postcranial elements, largely but not only ofthe manus and pes, have also been recovered from all these units, described here by R. L. Susman (and 19a9), and add to an earlier small sample (ten specimens) known, largely, from the HR unit (including írnportant peívic and femoral elements). Craniodental rematns attributable to Homo (aft. ergasler') were previously known trom the HR unit (three individuals) and are now also known from LB (one cr two) and from Member 2 (two) but strangely not from Member 3. A few postcranial elements from the former units may arsc be reasonably attributed to Horno. Artefactual evidence, but generally ot insecure provenance. first ca me to light at Swartkrans in the late 1960s as part of Brain's renewed research programme, the intñal descrfpñcn of which was given by Mary Leakey in 1970. Subsequently, more extensive documentation has been obtained from the LB, Member 2 and Member 3 units from which excavated, wellprovenanced collections are now known. Desmond Clark affords a very thorough, critícal evetuanon ot the available corlections, which are not very large in LB (62 specimens of just over 400 fractured pieces) and in Member 3 (32 specimens of 72 fractured pieces), but much more numerous (sorne 400 fractured pieces) in Member 2. The raw material is almost exclusively chert and quartz, the latter as cobbles and the former as blocks, chunks and tabular pieces; there are corresponding problems in distinguishing hominid-induced from naturally-fractured examples in such materials and in such derived sedimentary contexts. Clark's analysis is thus a model of the requisite procedures to be followed in such a situation. The assemblages constitute pieces with sharp, fresh edges but lack the fuI! ranga of by-products (1-2 cm size range) and conjoinable elements to be expected and encountered in hominid activity situations, including flaking and workshop circumstances. The basie categories inelude retouched and/or modified pieces, both heavy- and light-duty, modified and edge·damaged pieces, and cares and polyhedrons, and sorne related unmodified waste by-products. Overall, the occurrences approximate the Developed Oldowan of eastern Afriea, although the possibility of there being an (earlier) Acheulean in Member 3 cannot be excluded (several bifaces, a pick-like piece and a c1eaver-ended specimen, obtained from dump breccia residues, were reported earlier by Leakey). Clark has sought to examine distributional patterns, if any, in the several infill units, and although several spatially concentrated occurrences are evidenced in the LB and Member 2 units, horizontal concentration is weak or largely absent, and most suggestive of episodic input from external surface occurrences aJong with the intiJI. However, sporadic occupation is more likely in XI

FOREWORD

Member 3, in which specimens occur within the gully·infill confines, and sorne horizontal-depth concentration is discernible. In addition to lithic artefacts, utilizad bane 'tools' occur in all Members, rarely in HR (2) but more numerously in LB (17), Member 2 (11) and, especially, in Member 3 (40). Using experimental, actualistic studies and replication, and analysis by SEM, as discussed in detall by Brain and P. Shipman, il is apparent that these mostly exhibít distinctive signs of usage in digging activities, presumably including the procurement of underground storage argans of plant resources. In addition to the high Irequency 01 such objecls in Member 3 it is the only unillo yield specimens 01 bone (16) with cut or chop marks, largely on diaphyseal segments, which lestlly lo butchery practices, and perhaps particularly to scavenging activities on the part of one or another homínido That unit contains also abundant burnt bana, which occurs repeatedly through much 01 lhe infilled channel. As discussed by Srain and by A. Sillen and T. Hoering, distinctive evidences 01calcining, 01histologi-

12 June 1993 Ankara, Turkey

xii

...:z:w

cal changas and the production 01 char are readily and thcroughly demonstrable on such osseous resídues. This strongly supports the conclusión that the utilization 01 tire was both well-established and habitual over the course 01 the occupa-

lion interval(s). This adds subslantially lo the progressively emerging knowledge 01 this distinctively human practice, still known fromvery tew situations in eastern Africa01comparable or perhaps greater antiquity. This volume bears witness to the highly successful outcorne 01 an uncommonly well-designed, bread-based research programme 01immediate palaeoanthropological significance, but 01 aven broader relevance to Quaternary studies in general. Brain's endless patience and inquisitive bent, in addition to longterm participation 01 his wíte Laura and their children, is reflected lhroughout the monograph. They and associated scientists have afforded us a work 01 exceptional breadth and one 01 major interest to students 01 the hominid past and its contextual settings.

F. Clark Howell Laboratory for Human Evolutionary Studies nepartment of Anthropology University of California. Berkeley

Introduction C. K. Brain TransvaaJ Museum, Pretoria

The purpose of thls monograph is to present a comprehensive overview 01 the results of the final seven years 01 excavation in the Swartkrans cave, which ran from 1979 till 1986. In the evaluation 01 thase results, I haya had the valuad cooperatlon 01 a number of colleagues, both locally and in various parts 01 the world, whose contributed chapters have rnade this boa k possible. The excavatíon activity, which led to the results presentad here, was, far me, the culrnination 01 a long and personal involvement with the Swartkrans cave.lt started in 1951 when, as a young graduate, I worked as a geologist in the Building Research lnstitute 01the C.SJ.R. in Pretoria. With my interests centrad on natural history, I was a frequent visitar to the Transvaal Museum where I had discussions wtth John Robinson, who had just succeeded Robert Broom as the Museum's palaeontologist. He was busy wtth an excavation at Swartkrans and invited me to make a geological study of the cave deposit, as he had observed that the stratigraphic sequence visible at the time did not conform to that proposed by Lester King for australopithecine-bearing caves in generaL Preliminary results of that study appeared as my first scientific paper, co-aumored by John Robinson, in the proceedings 01 the Vth International Geological Congress, held in Algiers during 1952. The following year, Kenneth Oakley 01 the British Museum visited South Africa, and I had the opportunity of showing him the new stratigraphic information from Swartkrans that we had been gathering. On the strength of this, he lent his support to an application which John Robinson mace to the WennerGren Foundation for Anthropological Research in New York, the outcome ot which allowed us to set up a laboratory at the Transvaal Museum ter the analysis ot cave sediments. It also provided me with financial support for three years, while I completed a Ph.D. project, results 01 which were published as a Transvaal Museum Memoir, The Transvaal Ape-man-bearing Cave Deposits (Brain, 1958). On the completion of this work, I was not able to find employment in palaeontology, so took a position as Curator of Lower Vertebrates and Invertebrates at the Transvaal Museum. This allowed me to develop my zoological interests, working closely with the previous Curator in that Department and then Director 01 the Museum, Vivian FitzSimons, who introduced me to the reptile fauna of Southern Atrica, and provided me with a great deal of pleasure tor tour years. I then moved to the National Museums of Rhodesia (now Zimbabwe) as Keeper ot Zoology at the Oueen Victoria Museum in Salisbury (now Harare). While 1was pursuing a zoological career in Zimbabwe, John

Robinson left the Transvaal Museum to take a position at the University of Wisconsin and, sorne time later, Vivian FitzSimons asked me to consider returning to the Transvaal Museum as palaeontologist. The idea appealed to me for one particular reason - that I would have the opportunity ot turther investigating the Swartkrans cave which, in the course of my earlier geological work there, had exerted a curious influence on me. It ts a place with a powerful aura, which for me was irresistibly attractive. 1 have never been able to explain this influence but, whenever 1 am at Swartkrans, I experience a sense of peace, overriding all other feelings. But in addition to thls, I had another, intellectuaJ reason for wanting to work again at Swartkrans. It had to do with Raymond Dart, for whom I had a great respec!. In 1955 I had attended the 3rd Pan-African Congress on Prehistory, held in Livingstone and, at this meeting, Dart presented the evidence for his concept of an Osteodontokeratic Culturethat he thought had been practiced by Australopithecus. He gave an analysis of a sample ot 7000 tossn bones from the grey breccia at the Makapansgat Limeworks cave in the northern Transvaal, where he and his co-workers had found remains of Australopithecus africanus. The grey breccía was an amazingly rich source ot lossils and the skeletal-part analysis that Dart provided of a sample of these was the first of its kind from any early hominid assemblage in Africa. Raymond Oart's particular concern was to explain how this vast hoard ot bones, including occasional hominid remains, had come to be in the Limewarks cave. He was, in Iact, attempting a taphonornic reconstruction, befare the principies of taphonomy had been established, and in so doing, he played an important róle in the crystallization ot the taphonornic discipline. The great majority of the tossü bones came from antelope but, among them, not all parts 01 the sksteton were equally represented - in fact sorne parts occurred to the virtual exclusión 01others. Mandibles were found in abundance, as were distal ends 01 humeri, while proximal humeral ends and caudal vertebrae were absent. I~ thinking 01 an explanation for such discrepancies, Dart conceived the idea that early hominids were powerful hunters who used many 01 the bones of their prey as tools and weapons. He visualized the Limeworks cave as having been an australopithecine living place, to which useful banes were brought back. In his paper on the Cultural Status 01 the South Alrican Man-apes, Dart (1956) wrote: 'The fossil animals slain by the man·apes at Makapansgat were so big that in 1925 I was misled into believing that only human beings 01 advanced ¡ntelligence could have been responsible tar such manlike hunting work as the bones

2

INTROOUCTION

revealed ", these Makapansgat protomen. fike Nimrod long after them, were mighty hunters. They were also callous and brutal. The mast shocking specimen was the fracturad lower jaw 01 a 12·year-old son 01 a manlike apeo The lad had been killed by a violent blow delivered with calculated ac~uracy on the point 01 the chin, either by a smashing fist or a club. The bludgeon otow was so vicious that it had shattered the jaw on both sides 01 the lace and knocked out all the front teeth. That cramatic specimen impelled me in 1948 and the seven years following to study further their murderous and cannibalistic way 01 lile: The dramatic conclusions that Raymond Dart drew about early human nature, on the basis 01 his palaeontological studies, provoked a great deal or discussion. They stimulated the dramatist Robert Ardrey to write a succession of wldety read books on the roots of human nature, starting with African Genesis, and it was the numerous discussions I had with him and with Aaymond Dart that led me to crave for an opportunity to study a different bone assemblage from an australopithecine cave site, and to see what light tbts could throw on Dart's concept of the 'predatory transition from ape to man.' Swartkrans could provide me with precisely thts opportunity and so the Brain family returned in 1965 from a comparatively idylfic Rhodesia ta the poJitically turbulent Transvaat, settling once again within easy reach af Swartkrans. My immediate objective at Swartkrans was to assemble a sizable cotlectíon of fassils, fully representativa of what occurred in the cave, and from it to reconstruct the behavíour of hominids and theír associated fauna. The first obvious task was to restore order at the site after the chaotic Iime-mining episode tnat had interrupted the Broom/Robinson excavations between 1949 and 1951. We started on the dump below the cave, sorting fossiliferous blocks ot breccta from other rubble. This yielded a large number of tosslls, but unfortunately without stratigraphic information. lt sao n became apparent that Swartkrans had been extensively minad for lime befare the 1949 episode, and that to clear up the results of this operation would invol ve a major eftort. Funds and resources were needed and, once again, the Wenner·Gren Foundation for Anthropological Research carne to the rescue, with a grant al10wing the Swartkrans Palaeontological Research Project to be set up, which was to run for an uninterrupted 21 years. Throughout, it has been very much a family enterprise, in which I have had the continual support of my wife, Laura, and of each of our children, on tasks varying from the excavation of fossils, their preparation, cataloguing and identification, to experiments wíth bones in camp-fires. The project had been going for three years when the farm on which the Swartkrans cave is situated carne up for sale. It was fortunately bought by the University 01 the Witwatersrand and added to their existing property surrounding the Sterkfontein cave on the other side ofthe valley. Al about thattime also, the longterm Sterklontein excavation started, directed by Phillip Tobias and undertaken on a day-to-day basis by Alun Hughes, whose company 1greatly appreciated on my weekly visits to Sterkfontein to fiU up the water-trailer usad on our excavation. I have elsewhere raferred to the first phase ofwork as 'Savan year's hard labour at Swartkrans' (Srain, 1973). It involved the removal and hand-sorting of miners' rubble from the entire

cave area to a depth of 15 metres below the natural surtace ot the billside. and gave useful new insights into the original form of the cave. It also provided a sample of 14 000 tossü bones, among which were 15 hominid tossiís. Progress on the Swartkrans project was summarized in a series of papers in Nature, in which new specimens were announced and carnivare activity was postulated as the major bone-accumuJating agency at Swartkrans (Brain, 1970), two new early hominid vertebrae were described (Robinson, 1970), more evidence of an advanced hominid al Swartkrans was presented (Clarke, Howell and Srain, 1970), and stone artelacts from the outer cave breccia were described (Leakey, 1970). In the late 1960s we were joined at the Transvaal Museum by Elisabeth Vrba who was looking lor a Ph.D. topic and accepted my suggestion that the fossil antelope remains from Swartkrans and the other australopithecine caves wouíd be worthy 01 study, as indeed proved to be the case. She made very significant contributions, not only to the proper understanding of the bovid assemblages, but aíso to the correlation and faunal dating of me deposite. to palaeoecological reconstructions and evolutionary theory. By sorne coinciden ce, the Swartkrans project has progressed through four seven-year periods, each characterized by a different activity. The second of these ran from 1973 to 1980 and was dominated by the systematic removal of natural overburden obscuring tne extent of the cave system. While this was being done by our field team, under the supervislon ot George Moenda, I had the opportunity 01 investigaling a variety of taphonornic questions. But the most immediate result ot the site-clearing was the discovery that the breccia partially removed during the Broom/Robinson excavation was, in fact, an isolated mass, clinging to the north wall, and undercut along its entire length by an erosional space, usually partially lilled by younger sedimenls. It became known as the Hanging Remnant, but it was olear that it must have originally rested on a lower and older deposit, which I lermed the Lower Bank. It seemed very unlikely to me that this deposit could have disappeared complelely, so a seemingly interminable search began for the elusive Lower Bank, which had something of the quality of a religious quest. At conferences, correagues would ask me 'Have you faund it yet?.' and I would reply 'no, but I know it's there!' As newly exposed sections of the Swartkrans cave filling became available far inspection, the time had come for a delaited stratigraphic re-evaluation. It was fortunate that Karl Butzer was busy at that time with a sedimentological reinvestigation of the South African australopithecine breccias, and was able to give detaited attention to Swartkrans. On the basis of an analysis of 61 samples from six key profiles, he was able to formally designate Members 1 and 2 of the _Swartkrans Formation (Butzer, 1976), further particulars of which are given in Chapter 1. An interpretation of the very extensive fossil samples available by that time lrom Swartkrans and the other australopithecine~bearing caves, led me inevitably into the emerging discipline of taphonomy. In the mid-1970s I was not atone in this interest, there being several other people concerned with the taphonomy 01 East Alrican bone assemblages. Together we approached Lita Osmundsen, Research Director of the Wenner-Gren Foundation for Anthropological Research,

INTRODUCTION

about the possibility of holding a symposium on taphonomy. As aJways, wholehearted support was forthcoming, and a Burg wartensteín Symposium on Taphonomyand Vertebrate Palaeoecology: with Special Reference to the Late Cenozoic of Sub-Saharan Africa was held in July 1976, co-organizad by Kay Behrensmeyer. Andrew HiII and myself. 1t could be said that, at this meeting, vertebrate taphonomy carne 01aqe. and the symposium led to the publication 01a voJume Fossils in the Making (Behrensmeyer and Hill, 1980), Just as Raymond Dart had experienced in the case 01 hts pioneering investigation 01 the Makapansgat bone accumulation, skeletal disproportions were something that had to be addressed at Swartkrans. Shortly after starting work there in the mlddle 1960s, I had found the reason Ior the striking disproportional representation of skeletal parts in any assemblage worked over by carnivores and scavengers. lt was a very obvious concluslon, once it had been articulated, that certain elements in a vertebrate skeleton are more robust than others, ano are consequently better able to withstand whetever destructlve Jnffuences they might be subjected lo. The evidence for this ccnctuston carne to me when I made a collection of goat bones that had been discarded by Nama people living in small villages in the central Namib Oesert of Namibia. The goats were ñrst eaten by the villagers and the bones were then gnawed by their dogs, before being discarded onto the desert surface. To my surprise and delight, the pattern of survival and disappearance of skeletal parts matched that found by Dart at Makapansgal very close/y. No longer was it necessary to invoke deliberate selection of skeletal parts by hominids to explain disproportions in the overafl pattern. This did not mean that hominids had not used tools at all, but they need not have been responsible tor the selection of every piece in the Makapan grey breccia assemblage. Before publishing anything on that very informa tive colleetion of goat bones frorn the Namib, I took lt over to Johannesburg and showed it to Raymond Dart. Understandably he was taken aback, beca use the implications of it were that one aspect of his osteodontokeratic culture had fallen away. But within a few minutes, as he handled the goat bones and compared the pattern of survival with that at Makapansgat, he became more and more excited. He was genuinely delighted (hat we had new insight into the interpretation of fossil assemblages, and immedialely nominaled me lor an award - a gesture ¡ appreciated very mucho Raymond Oart's enthusiasm for a new inslght, challenging one of his cherished be!iefs, was to me a measure of his greatnes8 in science. A decade Jater, renewed problems with skeletal dispropor· tions at Swartkrans again ¡nvolved fossil hominids. By thattime we had remains 01 about SO individuals 01 Australopithecus robustus from the cave, but these were represented very largely by cranial pieces. For sorne curious reason, there were almost no body parts to go with the skulls, but the pattern of damage on sorne of the specimens suggested that predation had been the cause of death. When ene compared the pattern of skeletaj-part survival in hominids to that of antelope, however, an interesting difference was apparent. Bovids were much better represented skeletally than were primates. Could this mean that skeletons of primates were more susceptible to destruction than were those of bovids? 1 went back to the Namib to find out, where my friend AtilJa Port, Ihe well-known

3

wildlife enthusiast, made it possible for me to do so. On his farrn bordering the desert we investigated the contents 01 leopard lairs, captured a leopard and a group of cheetahs, and did a long series of feeding experiments on them. We cornpared the damage done to skeletons 01 baboons by the feeding of large cats, with that done to antelopes of similar Hve weight, and the results were significant. They showed that, fer structural reasons, the skeleten 01 a baboon is far more vulnerable to damage than is that of an antelope, and that this kind of taphonomic bias is likely to be encountered in fossil assemblages. By 1980 we had assemb/ed a good dea! of information on bone-collecting agencies, and their influences in the African context, and I was aote te put together an introduction to Atrican cave taphonomy. This was published under the title The Hunters or the Hunted? and included interpretations of bone accumulatíons of Swartkrans, Sterkfontein and Kromdraai. At Swartkrans itself, by this time, we had at last uncovered the elusiva Lower Bank - the sedlment base on which the Hanging Remnant had originally accumulated. We had also exposed the entire perimeter wall of the cave, finding that the cavern had extended tar further to the southeastthan had been anticipated. The time was now ripe for the systematic excavation of in situ sediments, and the third seven-year span of Swartkrans activity began. The first requirement was to set up a permanent metal grid over the excavatron area and then to proceed with meticulous excavation in which every fragment of bona was recovered, prepared and catalogued. In fact, during the period 1979-1986, the span covered by this book, we handled 350 000 pieces 01 10ss!1 bone, 01 which 49000 were individually catalogued. The extractton of this number of bone preces from the deposit was made pcssible by the fact that the Lower 8ank of Member 1 was relatively lightly calcified, and thus tairly easy to work, particularly when wet. From the start it was obvious that the sediment composing this lower bank had entered via an entran ce different from that whích had admitted the Hanging Remnant material. lt had been situatad above the southeastern wall of the cave and had bui!t up a talus con e below it in which the stratification was inclined towards the northwest. Whereas stone artefacts are ver¡ rare in Ihe Hanging Remnant breccia, they proved to be abundant in the Lower Bank sediment and were accompanied by sorne very striking bone tools, characterized by smooth. tapering points. The pattern of wear and scratching on these bonas suggested that they had been used as digging tools - a conclusion confirmed by digging experiments with modern bane flakes undertaken particularly by Conrad Brain on the modern Swartkrans hillside. By 1987 we had 60 of these bone tool8 from the excavation of Members 1-3, and 1was able to take them, logether with experimental equivalents, to Pat Shipman's laboratory al the Johns Hopkins University Medical School in Baltimore. By that time, Pat Shipman had considerable experience with the study of East African bone tools and she showed me her technique of making replicas of the bone surfaces tor examination in the scanning electron microscope. We were able to confirm that the f05s11 bone pieces had, in aH probability, besn used for the digging of edible bulbs from the stony dolomite hillsides in the vicinity of Swartkrans, as further elaborated in Chapter 8.

4

INTRODUCTION

It now seerns likely that su eh behaviour may haya besn important for hominid survival on the open grasslands 01 the" Transvaal. A tew 01 the bane tools had a striking polish superimposed on the characteristically worn surfaces 01 the specimens. This we interpretad as evidence that the tools had besn rubbed on a smooth, soft substance such as leather, suggesting the intriguing possibility that these early hominids were warking with animal skins, perhaps as bags ter carrying their gathered lood, or lor clothing. One 01the pleasures 01having found the bane tools was the opportunity 01 showing them to Raymond Dart who, though almost blind al that staqs, retained a kaen ínterest in the use 01 borre tools. He ran his fingers over each 01 the tapering points and then said; 'what do you think the ape-men used these for.' I replied that they seemed to have been used for digging. Dart was taken aback - 'that's a most unromantic explanation' he said and, holding one of the points against my ribs, he continued '1 could run you through with this!' I fear that several of my taphonomic reconstructions struck this gentle, yet strangely bloodthirsty man, as hopelessly prosaico In addition to bone tools, the excavation ot the Lower Bank and of later Members yielded a consistent scatter ot stone artefacts throughout the sediment. The presence of these, together with that 01 the bone tools, suggests that the outer cave at Swartkrans was regularly visited by hominid bands who brought their possessions with them. This may have been in contrast to the situation in the Hanging Remnant deposit, where artetacts are rare, and where at least some of the fossils could have been bony lood-remains that dropped into the fossilization site from trees aboye in which leopards fed. Our Swartkrans excavation policy of recovering every scrap of bone, however insignificant it might appear to be, had some useful consequences. One was the finding of a number of hominid hand and foot banas - elements previously rare in the Transvaal cave assemblages. These have been described in detall by Randall Susman (Chapter 4) who also conciuded that some 01 the 22 hand bones, which he attributed to Paranthropus robustus, indicated that the hand of this hominid was adapted tor precision grasping (Susman, 1988), and that tools could well have been used by both kinds of early hominid present at Swartkrans. In February 1984 we were excavating aboutfourmetres east of the cave's west waU when a piece of fossil bone turned up that showed every indication of having been burnt. This is not something that we had seen in the very large collection of bone from elsewhere in the cave but, as we continued, more and more pieces, suggestive of having bean heated, carne to light. We then realized that the sediment containing these specimens was not continuous with the Member 2 mass to the east, but that there was a near-vertical unconformity between the two deposits. The later deposit, now designated Member 3, filled a steep-sided gulley that had been eroded into the existing Members 1 and 2, along the west wall 01the cave. The Member 3 excavation of the next two years yielded nearly 60000 pieces olfossil bone, olwhich 270 appearto have been heated to various extents (Chapter 10). Confirmation of heating came both from my own histological examjnations and from chemical tests done by Andrew Sillan, results of whjch were published in Na/ure (Srain and Sillen, 1988). The evidence suggests that fires were repeatedly tended in the Member 3

gully, during the accurnulatíon of about six vertical metres ot sediment. According to Fred Grine's study (Chapter 3), hominld remains from Member 3 can be attributed to níne individuals of Paranthropus robustus, although Horno erectus must surely have been in the vicinity and was probably responsible ter the fire-tending activities reflected there. The excavation was done in 10 cm-deep spits, in 25 cm squares and burnt bone was tound in up to 20 consecutive spits through a vertical profile of six metres. This suggests that tires were repeatedly made on the tloor 01 the gulley over a period of some thousands of years and that fire-management was mastered in the interval between the accumulation of Members 2 and 3. The age of Member 3 is likely to be about one million yaars, and it seems very likely that lire was gathered Irom naturallightninginduced blazes lor a very long time - perhaps 500 000 years - befare it could be mada at will. But from the inception of fire-management, hominids would have had a measure of pratection from prsdatlon by the ever-present cats that stalked them, and was presumably one of the first steps on the long road to human domination of our natural world. The realization that Member 3 occupies an eraded space within the cave highlights the conclusion that the key to the understanding of Swartkrans stratigraphy lies in the fact that the deposit ls the product of repeated cycles of deposltion and erosiono Following the Initial Member 1 infilling, which took place through two entrances, a major erosional episode occurred in which a six-metre-wide space was created between the top 01 the Lower Sank and the bottom 01 the Hanging Remnant. This was subsequently lilled with Member 2 sediment that washed in through a reactivated entrance aboye the south wall. Once again the filled outer cave was subjected to erosion, this time cutting a gully along the west wall, later to be filled with Member 3 sediment. The pracess was probably repeated many times, but we only have evidence ot two other infillings - a rnajor one in the northeast comer of the cave that dates to Middle Stone Age times and ls now designated Member 4, and the most recent 01 the deposits, lilling a space clase to the northwest waJl, known as the bondi channel, on account of the fact that the most abundant tosslls in lt are from a smalJ extinct springbuck, Antidorcas bondi. Its bones have given a radiocarbon date 01 11 000 years, showing that this most recent infilling took piace at the height 01 the current intergiacial, and probably happened lairly rapidly. In tact, lt now seems likely that each ot the Swartkrans Members are interglacial deposits and that the alternating glacial periods are represented by the erosional interludes. The factthat this cave was such a sensitive indicatcr of climatic change, is due to the fact that the outer cave there was rather like a shelf, with a succession of openings aboye it as well as conduits below leading to lower caverns. Sediments rested temparally on this shelf, to be partly carriad away again when the erosional force of the water passing through the cave became too great. The reallzation thatthe Swartkrans deposits were reflecting c1imatic changes, probably based on global temperature variations, led me in the late 19705 to thoraughly investigate what had been published on temperature variations during the Cainozoic Era, and to correlate these with hominid evolutionary events. I presented the results to the Geological Society 01 South Alrica in 1979 as the Aiex L. du Toit Commemorative Lecture (Brain, 1981). which

INTRODUCTION

emphasized the coincidence 01 striking low-temperature avents in Antarctica with habitat changas in Africa, and with significant evolutionary developments. The theme was greatly developed in subsequent years by Elisabeth Vrba, Tim Partridge, George Denton and L10yd Burckle who,jointly, haya organizad a series 01 very productiva conferences under the general title 01 Pa/aeoclimale and Evo/utian, with the last meeting in the series giving emphasis to human evolution. Despite íts very long duration, the Swartkrans project has never besn without its pleasures and surprises, and these are by no means exhausted. But I have a stronq inclination now to widen my scientific horizons to encompass other forms 01life and longer geological tirne-spans. As recreation from the interminable stream 01fossil bones of recent years, I have had much pleasure from watching and photographing microscopic invertebrates. parficularty rotifers, that are ubiquitous in ponds and mosses wherever one may travel in the world. Rotifers are supero models of relatively simple metazoans, made up of a

5

rnanaqeable number of ceJlsin transparent bodies, so that ene may see what is gOlng on inside. They have led me naturally to an interest in the origins of multícellular!ty, that must have occurred in the immense time-span of the late Precambrian Era. Greatly encouraged by the positive comments ot Bil! Schopt, Gerhard Germs, Malcolm Walter and many olhers, I have started looking tor such a retord in Southem African Proterozoic sediments, once again supported by Laura who grinds the thin seclions 01 hopelul-Iooking cherts. Where this enterpríse wiU lead us I do not know, but wherever it does, I will rernatn grateful for the insights that Swartkrans has provided in so many directions. In September 1992 the Swartkrans cave was declared a National Monument, giving it an extra measure of protection that it deserves. It remains the property of the University of the Witwatersrand and I have no doubt that, in due time, its riches will result in new chapters, written by other palaeo-enthusiasts, in this remarkable cave's chronicle ot early mano

AEFEAENCE5 AADREY, R., 1961. Alrican genesis. A personal ínvestiqetion into tñe animal ongins and nature al man. conns. Landon. BEHRENSMEYER, A. K. and HILL, A. P., 1980. Fossils in themakinq. Vertebrate taphonomy and paleoecology. The unfversrtv of Chicaga Press, Chieago. BRAIN. C. K., 1958. The Transvaaf ape-man-bearing cave aeaosus. Transvaal Museum Memoir No. 11, Transvaa! Museum, Pretoria. BRAIN, C. K., 1970. New finds at the Swartkrans australoptthecine site. Nature225: 1112-1119. BRAIN. C. K., 1973. Seven year's hard labour at Swartkrans. Trenevaal Museum Bulletin 14: 5-6. BRAIN, C. K., 1981. The Hunters or the hunted? An introduction to African cave taphonomy. The University ot Chicago Presa, Chicago. BRAIN, C. K., 1981. The evelution of man in Atrica: was it the consequence of Cainozoic cooling? Alex. L. du Toit Memorial Lecture, 17. Proceedings ot the Geologicaf Society 01 Southern Alrica, Annexure to Vol. 84: 1-19. BRAIN, C. K. and ROBINSON, J. T., 1953. A geological note on the austraíopithecine-bearinq depcsft at Swartkrans. Congrés Geologique tntemetionet. Seco V, Alger, 1952: 55-56.

BRAIN, C. K. and SILLEN, A., 1988. Evidence from tne Swartkrans cave ter the earuest use of fire. Nature 336: 464-466. BUTZER, K. W., 1976. Lithostratigraphy of the Swartkrans Formation. $outh African JournaJ 01 Scíence 72: 136--141. CLARKE,-R. J., HOWELL, F. C. and BRAIN, C. K., 1970. More evidence ot an advanced hominid at Swartkrans. Nature 225: 1219-1222. DART, R. A., 1956. Cultural status of the South Afriean man-apes. Smithsonian Repor1424D: 317-338. DART, R. A., 1957. The osteoaontakeretic culture ot Australopitheeus prometheus. Transvaal Museum Memoir No. 10, TransvaaJ Museum. Pretoria. LEAKEY, M. D., 1970. Stone artetacts trom Swartkrans. Nature225: 1222-'225. ROBINSON, J. T., 1970. Twc new early hominid vertebras trorn Swartkrans. Nature225: 1217-1219. SUSMAN, R. L., 1988. Hand of Paranthropus robustus from Member t, Swartkrans: fossil evidence for tool behaviour. Science 240: 781-784.

Chapter 1

Structure and Stratigraphy of the Swartkrans Cave in the Light of the New Excavations c. K. Brain

Transvaaf Museum. P. o. Box 413. Pretoria, 000 1 South Africa

The stratigraphy 01 Ihe ancient cave filling at Swartkrans ts remarkably complex, bUI it has only been as site-clearing and excavation of the deposit proceeded that these complexities carne to jight. The process of discovery al Swartkrans is by no means over and I am convinced that as further work ís done there in the future, new surprises will surlaee. with the result that revisions 01 the interpretation presentad here will be needed. In fact, surprise has been a consistent feature 01 the Swartkrans project, and it has served as a stimulus and incentive.

tm

Breccia Dripstone [Gil Dolomile

g

THE HISTORICAL DEVELOPMENT OF STRATIGRAPHIC CONCEPTS AT SWARTKRANS The palaeontological exploration of Swartkrans started in November 1948, when Robert Broom and John Robinson moved part of their field team from Sterkfontein to investigate an outcrop of fossiliferous breccia visible clase to the crown ot the Swartkrans hill (Fig. 1). This initial work was partiyfinanced by the University of California's Africa Expedition, and confinued till November 1949. By this time, a large stalagmitic boss (Fig. 2), and an inciined seam 01 white travertine (Fig. 3) had been uncovered along the north wall of the excavation and a local lime-miner moved in while the museum team was away ter the Christmas holidays. The mining operation continued far more than ayear and created a large space immediately to the north of the active excavation area, shown in Fig. 4, as documentad by Robinson (1952). As is indicated in this diagram, the Iimited excavation area 01that time (se e also Fig. 5) was a very rich source ot australopithecine specimens ano it was otear then that the fossiliferous breccla lay above and around the basal stalagmite in the northwestern corner 01the cave. This sltuatíon did not agree with the stratigraphic sequence which Lester King (1951) had proposed for all the known australopithecine cave deposita in the Transvaal at that time. He had visualized a sequence of two wind-borne red sand deposits, separated by a travertine layer, and had inferred a climatic sequence of dry-wet-dry, with the australopithecine remains having accumulated during the earlier dry phase. This interpretation was clearly not applicable to Swartkrans, and we were also able to show that the sediment component of the Swartkrans 'pink' breccia had a particle size grading typical of hillside dolomite soil, rather than an aeolian sand (Bram and Robinson, 1953). At the time 01 Robinson's 1952 assessment of the

Fig.1

A vertical sectíon, running north-scuth, through the eastern end of the Swartkrans sita, as visualizad by Rabinson (1952) at the start of excavation in 1948.

Swartkrans stratigraphy, all the hominid fossils had been found in typical pink breccia, with the exception of a mandible, SK 15 (Fig. 6), initially referred to Telanthropus capensis - a form 'somewhat allied to Heidelberg man and intermediate between one of the ape-men and true man' (Broom and Robinson, 1949). but latertransferred to Homo erectus (Robinson, 1961). As shown in Robinson's 1952 diagram (Fig. 7), this was found in a pocket of chocolate-brown sediment at the southern edge of the breccia mass as lt was then exposed. In this connection, Broom and Rabinson (1949) wrote: 'Though lhis [lhe mandible SK15] was discovered in lhe same cave as the large ape-man, it ís clearly of considerably younger date. In the main bane breccia 01 tne cave depasit there has been a pocket excavated and refilled by a darker type ot matrix. The pocket was of very limited extent, being only about 4 ft by 3 ft and about 2 ft in lhickness. The deposil was remarkably barren, there being no other bones in it except the human jaw and a tew remains of very small mammals. We are thus at present unable to give an age of the deposit except to say that it must be considerably younger than the rnain

24

SWARTKRANS STRUCTURE & STRATIGRAPHY

deposit. 1I the main deposit is upper Pliocene, not improbably the pocl<et may be lower Pleístocene.' My geological study (Brain, 1958) 01 the Swartkrans deposits, as they appeared then, showed that there were two distinct masses 01 breccia, separated by a sleep unconlormity. These were termed the pink breccia, occupying the Outer Cave, and the stratified brown breccia, partially lilling the Inner Cave (Fig. 8). It was appreciated then that the structure of the northwest comer 01 the cavern system was determined by the presence 01 a massive floor block that had subsided Irom lhe rool, partitioning the chamber ¡nto Inner, Outer and Lower caves. The stages 01 lormalion, visualized at the time, are shown in Fig. 9 and it will be seen Fíg.2 that, after the lilling 01 the Outer A massive stalagmitíc boss in the northwest comer 01 lhe Swartkrans cave, exposed by excavalion Cave with pink breccia, the lurther in 1949. The arrow poin\s lo a fossil hominid child skull, SK 27, in íts original posilion. subsidence of the floor block rePholograph by J. T. Robinsen. opened an entrance to lhe Inner Cave, allowing the entry 01 the stratilied brown breccia, and the creation 01 an unconlormable contacl belween the two deposils. This scheme was accepted by Kenneth Oakley (1954a,b) in his writings about the sile, and seemed to stand up reasonably well until1973 when, as a result 01 the removal 01 miners' rubble, we opened a connection belween the base 01 the Inner Cave and the Lower Cave. Immedialely the signilicance 01 the Lower Cave in the depositional history 01 the Outer Cave breccia became apparent and it seemed likely tha!, throughout the cave's history, the Lower Cave has served as a space through whích water has passed on its way lrom the surface to subterranean reservoirs. It was clear lhat, in passing through the Outer Cave lilJing, rain water had carved a series of irregular channels which served, and continue to serve, as stormwater drains. Some of these channels have remained open, others are choked with loose or partly calcilied fillings. This addítion to the suggested sequence 01 events is shown in Fig. 10, as proposed at a Wenner-Gren Foundation symposium in New York in February 1974 (Brain, 1978). At that time I assumed that the Outer Cave at Swartkrans Fig,3 was originally filled with a single deposit, the 'pink' breccia, The basal travertine layer in the northwest comer 01 lhe Swartkrans through whích lhe channels 01 various ages ramilied. Bul as cave, as it appeared early in 1950, before ils removal by lime-miners. Elisabeth Vrba's sludy 01 lossil bovids progressed, she In lhe pholograph is Daniel Mosehle, a Technical Oflicer al lhe expressed her doubts that we were dealing wilh a single faunal Transvaal Museum, who was associated lor many years wilh lhe early Broom/Robinson excavations at Sterkfontein and Swartkrans. assemblage. The Outer Cave assemblage at that time conPholograph by J. T. RoblOson. sisted 01 the excavated sample Irom the Broom/Robinson operations, as well as a very large colleclion 01 lossils from the early lime-mining operations at the cave. 1I became more and progressed, we made the unexpected discovery that the pink more obvious that the Outer Cave used lo contain breccias 01 breccia that had produced the Swartkrans hominids was an isolated mass, c1inging lo the north wall 01 the cave, and more recent age, in addition to the pink breccia that occurred clase to the north wall. As our site-clearing operations undercul along its entire length by an erosional surface. Below

SWARTKRANS STRUCTURE & STRATIGRAPHY

25

Dolomite ~

o

4

8

O

Pink breccia

reel

"

..

. .

'

.:

..

'

"

: ..

::

::~•. \.:.:-:.~~

:

.

. ': .

.'

.

Fig.5 The Broom/Robinson excavation in the Hanging Remnant in the northwest corner 01 the Swartkrans cave, as it appeared in 1951. Pholograph by J. T. Robln.on.

Fig.4

I

A plan 01 the Swartkrans excavation, in the northwest corner of the cave, as drawn by Robinson (1952), on the resumplion 01 palaeontological work late in 1951 lollowing the Iime-mining interlude. The space resulting Irom the removal 01 the travertine layer is shown, while 'A' indicates lhe area Irom which some hominid specimens were obtained during the mining operations, including the crested skull, SK48, an innominate, SK 50, and a large mandible, SK51. 'B' indicates the position 01 the large stalagmilic boss shown in Fig. 2. The localions 01 various olher hominid specimens in the original Broom/Robinson excavalion isalso shown: 1: badlycrushed skull; 2: juvenile jaw, SK61; , 3: massive mandib/e and snout, SK12; 4: palale, SK11; 5: palale, SK13; 6: crushed child skull, SK27; 7: Telanlhropus mandible. SK15; 8: child mandible, SK62; 9: mandible with chin; 10: jaw fragment, SK45; 11: Iype mandible, SK6; 12: adolescent palale, SK55; 13: complete mandible, SK23.

this surface there was generally a space and then, furlher out, browner and younger breccia deposits occurred. It was clear [hat the mining operations 01 [he 1930s had removed a large volume of these laler deposits. and lhat their fauna was differenl trom that ot lhe 'pink' breccja itself. Making use of new sections available in the Outer Cave, Karl Butzer (1976) formally defined Members 1 and 2 of the Swartkrans Formation. Member 1 look in the whole ot

Fi9·6 The mandible. SK15, originally designated Telanthropus capensis, now referred to Horno erectus, from a pocket 01 later breccia withln the Hanglng Remnant deposit, as shown in Fig. 7.

SWARTKRANS STRUCTURE & STRATIGRAPHY

26

Soll overlying doIomite

e

A

~ DoIornite

IZlD_ G Pink breccia

SCALE

Artificial retaining

o

wals

~ Chocolate breocia

....

I

20 [

40

50

[

[

FEET

I

E

Fig.7

LJ Q

lnner cave brown breccia

[[]]]

Travertine

...

A diagram provided by Rabinson (1952) ct the excavation area in the northwest comer of the Swartkrans cave, as lt was in April1949 when the 'Telanthropus'mandible, SK15, was discovered al posltion 7 within a pocket 01 darker breccia. Other numbers on the diagram refer to specimens tisted in Fig. 4.

•

Outer cave pink breccia

Dolomita

Fig.9

Fig.8

Vertical north-south sections through the Swartkrans cave, as visualizad by Brain (1958), showing the suggested slages 01 developmenl 01the cave and lts deposits al that time.

Vertical sections through the Swartkrans cave as visualizad by Brain (1958), showing the relatlve poaiticns ot the Inner. Outer and Lower caves, and the contact between tne pink and brown

breccias.

STAGE 1

STAGE 2

STAGE 3

STAGE 4

STAGE 5

STAGE 6

STAGE 8

STAGE 7

O

20

I

!

&ale In teet

40

[]]]

•

Travertine

C] :.>::'; Brown

Pink breccia

Dolomite

~

•

Travertine

mm

breccia

Dolomite

O

20

I

!

Scale in feet

40

27

SWAATKAANS STRUCTUAE & STAATIGAAPHY

a

,

Outtl'Cave

I Lower Cave

Inner Cave

D

m

Qulerc.... bt
,

,

Mcmbor 1 - .

O

¡'¡ombor 1 bfwciI

[3]j T... _

~ >- ,_...... •

"""'""

Fig.l0 Vertical south-north sections through the Swartkrans cave, as visualized in 1974 by Brain (1976), showing the suggested stages 01 development 01 the cave. The Outer Cave is shown to have contained a single deposit only, through which solution channels ramify.

Fig.11 Vertical south-north sectlons through the Swartkrans cave, as visualized by Brain (1976), showing suggested stages 01development 01the cave, foHowing the discovery of the Hanging Remnant, and Ihe assurnption that the Lower Bank must exist. The Outer Cave was then assumed to contain deposits o, both Members 1 and 2.

the former 'pink' breccia, but included the basal travertine (Member 1A) and the thin deposil 01 'orange breccía' in Ihe Inner Cave, interbedded between travertine layers (Member 1e). which had been accommodated by further subsidence 01 the Iloor block. Member 2 was undivided bul comprised both the extensive breccia deposit in the Quter Cave as well as the 'stratiñed brown' breccia 01 the Inner Cave. The presence of later deposits was recognized, but they were not formally designated. A sequen ce of suggested stages in the fermation

of the Swartkrans deposits, as visualized at thattime, is shown in Fig. 11 (Brain, 1976). The realization that Member 1 in the Outer Cave represented a Hanging Remnant, stuck to the north wall, led to the search tor its stratigraphic counterpart, the Lower Bankon the cave traer, as described in the Introduction to this volume. This involved the systematic rernoval 01 natural overburden obscuro ing the walls and contents 01 the entire Outer Cave and the excavation 01selected parts of the deposit.

28

SWARTKRANS STRUCTURE & STRATIGRAPHY

THE CURRENT INTERPRETATION OF SWARTKRANS,. STRATIGRAPHY A plan of the Swartkrans cave, with the exception of lhe Lower Cave, but indicating the position of the Hangíng Remnant ís shown in Fig. 12. Also visible is the permanent metal grid thal has be en erected oVer the main excavation area (see also Fig. 13). Since Ihe time 01 the formal recognition of Members 1 and 2. lhe excavation has provided further insighls into the complexity of lhe filling, now known to have resulted from repeated cycles 01 deposition and erosion within lhe cave. Al present, five Members are recognized, each separated from ils older counterparts by an erosional discontinuity. There is no question lhat many olher separate deposits have fitled erosional

spaces in the Swartkrans cave fill during its 1,7-million-year history, but lhese have been lost by erosion or mining. The importanl point is that the Swartkrans cave has funclioned as a sediment trap, by virtue of its slructure, with openings lo the surface tlbove and olhers leading downwards to lower caverns. Under the inlluence of alternating ciimatic cycles, almosl certainly of a worldwide glaciaJ-interglacial nature, episodes of deposition in the Swartkrans cave have been followed bythose of erosion, creating a depositional mosaic 01 considerable complexity. It now appears lhat Members 1 and 2 occupied much of the GUler Cave, soulh of lhe Hanging Remnant, but the positions in which Members 3, 4 and 5 have been found are indicated in Fig. 14. A vertical section through lhe filling 01 lhe Guler Cave, as it is currently visible, is shown in Fig. 15.

Fig.12 A plan 01 the Inner and Outercaves al Swartkrans, as currently exposed, showing the position 01 the Hanging Remnant 01 Member 1, and the permanent excavation grid.

smMl

hanging remnan!

N

1 o

5

t:-=::.::J

m

Fig.13 An overview 01 the .Swartkrans cave, laken Irom aboye the north wall, looking soulh over the permanent excavalion grid.

29

SWARTKRANS STRUCTURE & STRATIGRAPHY

Fig.14

[:+./\1 Member

A plan 01 the Swartkrans cave, as currently exposed, showing the positions 01 lnñllinqs desiqnated Members 3, 4 and 5.

f.;:·.:.~

Member 4

•

Member 3

N

1 o1 _ . 15 m

Fig.15

s

SECTION A - B

N

A vertical sectlon running northsouth through the filHng of the Outer Cave, as it ls currentty visible, alonq the secticn line as shown in Fig. 12.

~MEMBERI

~ LowerBank

~!;fl~§.1 TRAVERTINE _ MEMBER4 M5A Unit

t:::·:::·:l

MEMBER 2

~

DOLOMITE

o IlwwI

5

.....

6 """"

SWARTKRANS STRUCTURE & STRATIGRAPHV

30

SUGGESTEO STAGES IN THE fORMATION Of THE . CAVE ANO ITS OEPOSITS

OOlOMITE WATER

The suggested formational staqes 01the cave and its contents are presentad as a series 01 reconstructed and somewhat generalized sections (Figs 16-24) running diagonally acrcss the cave trorn the southeast to the northwest corners. Many 01 the features described in these suggested stagas can be seen by visitors lo Ihe site and, fer the benelil 01 such people, a reute leading past 25 points 01interest has recently been constructed, while details 01what can be seen are given in an illustrated guide (Brain and Watson, 1992),

TRAVERTINE

MEMBER 1 Lower Bank MEMBER 1 Hanging Remoant MEMBER 2 MEMBER 3 STRATIFIED MEMBER 2

Fig. 16. Slage 1. A cavern has been dissolved in the dolomitic country rack at a considerable depth below the surlace, in the phreatic zone, below the level 01 standing water. lts form has been determined by joints and planes ot weakness in the dolomite. lt has connections to lower caverns and the tloor ot the main cavern ls strewn with dolomite blocks. The age of the cavern íe probably Miocene.

NW

SE

~'Nr:~_~--~=---:;

_;

~

-=-.'""""'

__

~r:o~ ~ ..~~--"=~ -. - "'=~~-_ ~--=:.....: ==-= ~~ ~ =---=::-:ra...= -: ......-:.-..,¡;:~ -:;-=..;: _ ~;::i:>lOo.:: •~ _

-_...:lIo:

'it..::=u:--_

f-r.l--T::-:o..;

=-w-

~

......:.-=-.:.

~ -......

=->o.::oo....~ .

:-r-:

.

~=-_-....:~-d!J.

:o.:: ~ :-:-= ~ -:::-ir...:: - - ....:: ~~"""~-:..:; _ =--~..=o.....:-----=-=--=-~ _

=-=-

""",..::lo
=_

~

~ ~ ~~--j _ -: =-...:-

~

-:""3IL...:_

-

Fig. 17. Stage 2. The level of standing water in the cavern has dropped as a result of vatley lncisicn. The cave is now air-fiñed and a very large block ot dolomite, the Floor Block, has fallen lrom the roof in the northwest comer of the cavem. Stalactites and stalagmites are now forming in the cavern, through the evaporation of Hme-richwater, and a massive boss ot travertine is being deposited over the southwest síde of the Floor Block. Joints are being enlarged in the dofornite roof of the cavem. through the downward passage 01 water, particularty above the southeastern wall. These events are probably Püocene in age.

~4··

~

-~

~ V.

~

.'-'

~

c,

~

SWARTKRANS STRUCTURE & STRATIGRAPHY

31

Fig. 18. Slage 3. Joints in the roo' o, the cavern haya been further enlarged and, over the southeast wañ, the first shaft-like opening to the cave has developed. tr 15 admitting surface-derived sediment which forms a taíus cone below the shatt. The aqe 01 this first sediment, designated the Lower Bank 01Member 1, is probably in excess 011,7 mllllon years.

.~

Fig. 19. Stage 4. The original entrance to the cave has been further enJargad, as haya other joints ln the root. A substantial deposit 01 Lower Bank sediment, lightly calcitied by drip from the root, nowalmost fiUsthe cave. This first inlilling process may haya taken about 20 000 years.

"

Fig. 20. Stage 5. The original entrance to the cave has now been choked by Lower Bank sediment and a new shatt-llke opening has formed above the north wall. This ls admitting sediment destined to become the Hanging Remnant of Member 1 which, being heavily catctñed in comparison to the Lower Bank, suggests that a rieh source of lktte-bearlnq solutions was present along the north wall, or that deposition rates were slow, or both. The abundanee of fossils in the Hanging Remnant breccla indicares that the environs of the shaft-Iike entrance to the eave at that time favoured the coneentration of hominid cranial rernains, as discussed in Chapter 13.

32

SWARTKRANS STRUCTURE & STRATIGRAPHY

Fig. 21. Stage 6. A period 01 eresion within the cave occurs, with

a goad deal of water entering through a new snatt-ooenína, midway between Ihe soutn and

north walls. An irregular gap, severa! metras wide, is eroded between Ihe too of the lower Bank and Ihe lower surtece ot tne Hanging Hemnant. as deca.lcified sediment la trickled away to lower cavems in Ihe dolomita.

Fig. 22. 5tage 7. The gap created by tha erosión in Slaga 6 is now tillad with sediment 01 Member 2, which has enterad through the midway shatt. 1I ls rich in bones and artefaets, resulting trom hominid and cther animal acñvity in this new entrance area

01Ihe cave.

Flg. 23. Slage 8. A renewed oyere 01erosion has created a gulley alonq the west wall 01 the cave, in the postncn indlcated in Fig. 14, removing parts of the earlier Member 1 and 2 deposíts. It has tiüed with Member 3 sediment. in which normal and bumt bones, as wel1 as artefacts, are preserved. lt appears tnat aarly hominids used the entrance area to this new gulley for shelter, and it was here that pioneering expertments with the management ot 'ire took place (Brain and Sillen, 1988). For oetass. tne reader ls referred to Cttapter 10. Yet another minor shaft has admltted sediment, destined to become the stratiñed brown component o, Member 2 in the extreme northwest comer al the cavem. This oeposlt is probably broadly contemporaneous with the main body of Member 2.

Fig. 24. Stage 9. The contemporary situation befare mining or excavatlon took place. Most ot tne dolomite root of the cave has been removed through hillside eroslon and channels. and larqer erosíonat spaces have ramified through the older deposits. Two of these now cantain sediment reterred to Members 4 and 5, while several others have presumably been íost through weatherinq and mining.

'C'

SWARTKRANS STRUCTURE & STRATIGRAPHY

THE AGE Of THE SWARTKRANS DEPOSITS Repeated attempts to obtain absoluta dates for the Swartkrans deposita and fossils haya so far been fruttless, while results 01 a palaeomagnetic study preved ambíouous (Brock, McFadden and Partridge, 1977). An attempl by Tim Partridge (1973) to date the first opening 01 each 01 the australopithecine caves by measurements 01 cyclic nick-point migration and vañey flank regression gaya a figure 012,57 million years S.P. for the first opening 01 the Swartkrans cavern. Although this figure is interesting, there ts disagreement among specialists on whelher a geomorphological method 01this kind is reliable. During the 19605 and 19705, Elisabelh Vrba made a detailed and comprehensive study 01 the fossil bovids from Swartkrans and the other australopithecine cave deposits. On the basis 01lhis work she suggested (Vrba, 1975) that fossils from the Hanging Remnant ot Member 1 could be referred to the Swartkrans Faunal Span, with an age indicaíion of between 2 and 1 million years. She later refined this esñmare (Vrba, 1982) to place Ihe Hanging Remnanl deposit al between 1,8 and 1,5 mllllon years. At that time she also had at her disposaJ a large collection of 'Member 2' fossils which, however, carne from a variety of stratigraphic sources that had been disturbed by mining. Age estimates on thls collection were understandably ambiguous. Fossil assemblages from the recently excavated stratigraphic units now designated Members 2 and 3 have been sludied in detail by Virginia Watson (Chapter 2, this volume) and other specialists, while cultural remains have also be en investigated (Chaplers 7 and 8). It now appears lhat Members 2 and 3 do not difter significantly in fauna or culture from Member 1, and that Australopíthecus robustus remains are found in all three units, whose age is assumed lo be between 1,8 and 1 million years. Member 4 is a largely uncalcified deposit, rich in Middle Stone Age artelacts that accumulated under the overhanging north wall in the northeast comer ot the cave. It has not provided any fossil banes to date and has not yat been excavated. Member 5, on the other hand, is a tour-rnetre-thíck, tightly calclfied deposit that partially fills an erosional space

33

surrounding a massive stalagmitic boss on the narthwest side

al the cave. rts enclosed lossils (Chapter 2) are dominaled by remains 01 the extinct springbuck Antídorcas bondl, banes of which have been dated by"C lo 11 000 years B.P. (J. C. Vogel, personal comrnurucation).

CLlMATIC IMPLICATIONS Of THE DEPOSITIONAL SEQUENCE It seems likely that the particular conliguration of the Swartkrans cave has made it a sensitiva indicater of envirenmental change within the catchment area 01 the cave's entrance. The most recant of the sediment infilJings, Member 5, is an instructive deposit, whose four-metre depth probably accumulated lairly rapidly, apparently during the period 900012 000 years S.P., and suggests a time when storm-water entering the cave carried a good deal 01sediment eroded from the hillside. This in turn indicates an incomplete mat of vegetatíon on the Swartkrans hill, as might be expected on the basis of the low 'moisture index' values, inferred for the early Holocene at both the Wonderkrater peat tecality in the central TransvaaJ and the Wonderwerk Cave in the northern Cape (Thackeray and Lee- Thorp, 1992). The erosional space into which the Member 5 sediment was deposited, on the other hand, seems to have been formed under different conditions, when abundant water entering the cave was both devoid of a sediment load, and acidic. Such conditions may have prevailed during the long preceding 'glacial' interval when low temperatures led to reduced evaporatíon and to more water passing through a dense vegetation mat, rich in humic acid. If this was, in fact, the sequence of events at Swartkrans during the most recent glacial-interglaciaJ cycle, then a similar pattern is likely to have been repeated many times during the last 1,7 million years, some products of which appear to be reflected in the earlier Swartkrans Members. ACKNOWLEDGEMENT I am very gratefuf to Virginia Watson fordrawing many ofthe illustrations for this Chapter.

REFERENCE5 BRAIN, C. K., 1958. The Transvaal epe-men-beerinq eave deposits. Transvaal Museurn Memoir No. 11. Transvaal Museum, Pretoria. BAAIN, C. K., 1976. A re-interpretation of the Swartkrans site and its remalns. South Alriean Journaf ot Seienee 72: 141-146. BRAIN, C. K., 1978. Sorne aspects ot the South African austra'opithecine sites and their bone aceumulations. In: JOLLY, C., ed., Early hominids ol Afriea, pp. 131-161. Duckworth, London. BAA1N, C. K. and ROB1NSON, J. T., 1953. A geoJogical note on the australopithecine-bearing deposit at Swartkrans. Congrés Geolo· gique Intemational, Seco V, Alger, 1952: 55-56. BRAIN, C. K. and SILLEN, A., 1988. Evidence from the Swartkrans cave for the earliest use of fire. Nature 336: 464-466. BRAIN, C. K. and WATSON, V., 1992. A guide to the Swartkrans early hominid cave site. Annals althe rrBnsvaal Museum 35(25): 343-365. BROCK, A., McFADDEN, P. L. and PARTA1DGE, T. C., 1977. Preliminary palaeomagnetic results from Makapansgat and Swartl<;rans. Nature 266: 249-250. BROOM, R. and ROBINSON, J. T., 1949. A new type of fossil mano Nature 164: 322. BUTZER, K. W., 1976. Lithostratigraphy of lhe Swartkrans Formation. South Alrican Joumai 01Scienee 72: 136-141. KING, L. C., 1951. The geology of Makapan and other caves. Transaetions ol the Royal Soeiety 01South Aln"ca 33: 121-150.

OAKLEY, K. P., 1954a. The datinq ot the australoplt'iecinae of Africa. American Journal 01Physical Anthropology 12: 9-28. OAKLEY, K. P., 1954b. Study tour of early hominid sites in southem Africa. South Alriean Archaeofogical Bulletin 9: 75-87. PARTRIDGE, T. C., 1973. Geomorphological dating of cave opening at Makapansgat, Sterkfontein, Swartkrans and Taung. Nature 246: 75-

79. ROBINSON, J. T., 1952. The australopitheclne-bearing deposits of the Sterkfontein area. Anna/s ol the Transvaal Museum 22: 1-19. RD8INSON, J. T.. 1961. The australopithecines and their bearing on the origin of man and of stone tool~making. South African Journal 01Seience 57: 3--13. THACKERAY, J. F. and LEE-THORP, J., 1992. Isotopic analysis of equid teeth from Wonderwerk Cave, northern Cape Province, South Africa. Palaeogeography, Pafaeoclimate, Palaeoecology 99: ·141-

150. VRBA, E. S., 1975. Some evidence of chronology and palaeoeCOlogy of Sterkfontein. Swartl<;rans and Kromdraai from the fossH Bovidae. Nature 250: 19-23. • VRBA, E.S., 1982. Biostratigraphy and chronology, based partict,Jlarly on Bovidae. of southern African hominid-associated assembl~ges. Proc. Congres de Paléontologie Humaine ler Congres, pp. 707752. Nice.

Chapter 2

Composition of the Swartkrans Sone Accumulations, in Terms of Skeletal Parts and Animals Represented Virginia Watson Transvaal Museum, P.

o. Box 413, Pretoria, 000 1 South Aldea

Excavations at the Swartkrans cave in the Sterkfontein valley carrled out between 1979 and 1986 have yielded a substantial ccñectlcn of banas trom tour dlstlnct Members of different ages. These haya been analysed and were found fa represent 64 mammalian laxa (excluding microvertebrates) as well as representativas of birds, reptiles and amphibians. The results suggest a relauvely ccnsistent fauna throughout the depositlonal history of the cave. Carnivoras were rnost likely the agents responsible tcr the accumulation of prey species in the cave. The environment, basad on fauna) compoenon, appears to haya been essentially unchanged throughout the depositional history of the cave. It comprtsed what is thought to have been highveld grassland with a few rocky hills, and riverine woodland sevanna in the victnlty 01the Blaaubank stream, which itself is envisaqed as having been larger and more substantial than at presento

INTRODUCTlON The bone sample excavated from Swartkrans cave between 1979 and 1986 exceeded 350 000 individual elements (Table 1). These were individually numbered and catalogued, and details such as date, provenance and identification recorded. These elements were excavated from four stratigraphically lndependent Members ranging in age from 1,8 million years S.? (Member 1. Lower Bank) to 11 000 years B.P. (Member 5). Members 2 and 3 are believed to be c. 1,5 and 1 ,O million years old respectively. The microfaunal bonas were put aside ano counted but no attempt was mace at identification. They are currently being studied by M. Avery al the Soulh Alrican Museum. Of the lotal number ofbane píeces, they accounted for a larga percentage: Member 1,67 %; Member 2,52 %; Member 3, 65 %; Member 5,5 %. The rest of the bones were then sorted into identifiable and unidentitiable pieces. The unidentifiable bone consisted of bone that could not, with any confidence, be assigned to taxa. The unidentifiable pieces accounted for 30 % of the total number ot bane pieces in Member 1,43 % in Member 2,30 % in Member 3, and 74 % in Member 5. Unidentifiable pieces were assigned either to 'flakes' or 'misceJlaneous fragments.' Flakes, which were extracted first, were defined as pieces from shafts of long-bones, particularly of bovid origin, that lacked complete articular ends and did not preserve more than half the circumference of the long-bone shaft (Brain, 1981). The remaindar constituted the mlscenaneous fragments. Bone shafts are dealt with in the analyses of the vartous Members. The identifiable bone pieces were then assigned to skeletal part and where possible to genus and species. The cranial

bones were consistently the most likely to be identified to genus and species level. Once this had been completed, itwas possible to calculate the minimum number of Individuals of each taxon using the single rnost common skeletal part from the particular taxon. For example, if there were ten left lower third molars representing the rnost numerous identified rernains of Connochaetessp. in a particular Member, a minimum number (MNI) ot 10 individuals was assumed since each individual has only one such tooth. The results of the identiñable-bone analysis are summarized in Table 3 and iIIustrated tor the separate Members in Figs 1-4. The results for each Member are set out in detail in the sections below. The taxa are discussed according to taxonomic sequence in Brain's (1981) The Hunters or the Hunted?

THE MACROVERTESRATE SONE COMPONENT Macrovertebrate bones were analysed in detail and much of this chapter deata wlth them. Damaged and bumt bone pieces are included in the numerical figures in Tabla 1. 'Miscellaneous bone pieces' were numerous and most of them were less than 2 cm in size and could not be identified. 'Bone flakes' were measured and placed into size classes in Table 2. The percentages of each size class show that most of the flakes in all Members fall into the smaller size classes, indicating a similarity in the procedures used ter praducing the flakes, although Member 5 shows a larger prcportion ot bigger flakes when compared to the omer three Members. Stone artefacts and bone tools are discussed in Chapters 7 and 8 of this volume respectively.

COMPOSITION OF SONE ACCUMULATIONS

36

Table 1

Summary 01 excavated bone pieces from Swartkrans (1979 -1986).

Member 1

Number ot elemenrs

Member 2

% 01 total

Number 01 elements

Member 3

% of total

Number 01 elements

Miscellaneous bcne pieces Bane ñakes Microfaunal banas Identifiable macrovertebrate bonas Bane toots

40 110 6336 102371

26,1 4,1 66.6

22938 6919 36213

32,8 9,9 51,8

22602 9433 69850

4950 17

3,2

3768 11

5,4

< 0,1

< 0,1

6173 40

Total number af specimens

153784

69849

L1ST OF LARGE MAMMAL TAXA REPRESENTED IN SWARTKRANS MEMBERS 1-3 AND 5 Horno sapiens, man: Member 5. Horno sp., early man: Members 1. 2. Australopithecus robustus, robust apernan: Members 1-3. cf. Cercopithecoides sp: Member 1. Papio hamadrayas ursinus, chacma baboon: Member 5. Papio hamadrayas robinsoni: Members 1-3. Theropithecus oswaldi danieli: Members 1-3. Papio (Dinopitheeus) ingens: Member 1. Panthera pardus, leopard: Members 1-3, 5. Acinonyx juba tus, cheetah: Member 3. Felis caracal, caracal: Member 1. Felis el. tybiea, wild cal: Member 3. Felis servet, serval: Member 1. Panthera leo. lion: Member 1. Megantereon cultridens, dirk-toothed cat: Member 3. Hyaena brunnee, brown hyaena: Members 1-3. Chasmaporthetes nilidula, hunting hyaena: Members 1-3. Crocuta crocuta, spotted hyaena: Members 2, 3. Hyaenid Indel.: Member 5. Proteles sp., aardwolf: Members 1,3. Proteles cristatus, aardwolf: Member 5. Vulpes ehama, Cape fox: Member 3. Small eanid gen. & sp. indet.: Member 1,5. Canis ef. mesometss, black-baeked jaekal: Members 1, 2, 3, 5. Large canid gen. & sp. indet.: Members 2, 3. Aonyx capensis, clawless atter: Members 1,2,3. Atilax sp., water mongaase: Members 2, 3. Cynictis penicillata, yellow mongoase: Members 2, 3. Herpestes Ionneumon, large grey mongoose: Member 1. Surieata sur/caNa, suricate: Members 2, 3. 5. Mongoose lndet.: Members 1, 5. Genetta tigrina, large-spotted genet: Member 3. Viverrid indet.: Member 5. Manis sp., pangolin: Member 3. Oryeteropus afer, antbear: Members 1, 2, 3. ct. Elephas sp.: Member 1,3. Procavia transvaalensis, hyrax: Members 1-3, 5. Procavia antiqua, hyrax: Members 1-3. Procavia capensis, rack hyrax: Member 5. Hipparion Iybieum steytleri, three-toed horse: Members 1-3.

108098

Member 5

% 01 total

Number ot elements

% of total

20,9 8.7 64.6

13841 867 1077

69,3 4,3 5,4

5,7

4187

21,0 0,0

<:

0,1

19972

Grand total

351 703

Equus cepensis, gian! Cape horse: Members 1-3,5. Equus burchel/ii, Burchell's zebra: Members 3, 5. Phaeoehoerus sp., warthog: Members 1-3, 5. el. Tapinoehoerus meadowsi, large fossil suld: Members 1-3, 5. Hippopotamus sp., hippopotamus: Members 1, 3. Giraffid gen. & sp. Indel.: Members 2, 3. Megalotragus sp., giant hartebeesl: Members 2, 3, 5. Connoehaetes sp., wildebeest: Members 1-3, 5. Beatragus sp.lAleelaphu5 sp., medium-sized alcelaphine: Members 1-3, 5. Damaliscus sp., blesbok: Members 1~3, 5. Antidorcas australis and/or A. marsupialis, springbuck: Members 1-3, 5. Antidoreas recki, fossil springbuck: Members 2, 3, Antidorcas bondi, Bond's springbuek: Member 5. Antidorcas sp.: Member 1. ? Gazella sp., gazelle: Members 1-3. Oreotragus oreotragus, klipspringer: Members 1, 3. Raphicerus campestris, steenbok: Members 1-3,5. Syneerus sp., buffalo: Members 1-3, 5. Taurotragus oryx. eland: Members 2, 3, 5. Tragelaphus strepsiceros, kudu: Member 3 Hippotragus cf. niger, sable: Members 2, 3, 5. Pelea sp., rhebuek: Members 3, 5. Redunea cf. arundinum, reedbuck: Member 5. Lagomorph gen. & sp. indet., hare: Members 1-3, 5. Petietes sp., springhare: Members 1-3, 5. Hystrix afrieaeaustralis, poreupine: Members 1-3, 5.

Analyses 01 the macrovertebrate eomponents af each Member is given below and a summary is given in Table 3. MEMBER 1 LOWER BANK Figure 1 illustrates the macrovertebrate taxa found in this Member and the material recovered ís listed below. Class MAMMALlA Order PRIMATES Famlly HOMINIDAE The hominid material ts discussed and described in detail in Chapters 3 and 4, this volume.

COMPOSITION OF SONE ACCUMULATIONS

37

Table 2 Numbers of bone tlakes recovered from Swartkrans.

Member 1 ------

n

Flake

Jength

%

Member 2 -----n %

Member 3 ------

n

ot total

ot total

%

Member 5

n

of total

% of total

(cm) 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 Total

1749

2096 1214 583 274 134 70 29 11

8 3 6171

28,3 34,0 19,7 9,4 4,4 2,2 1,1 0,5 0,2 0,1 0,0

1844 2810 1251 607 187 90 27 7 8 3 1 6835

27,0 41,1 18,3 8,9 2,7 1,3 0,4 0,1 0,1 0,0 0,0

Horno sp. MATERIAL. Juvenile mandible: SKX 21204. Australopithecu5 robustus Twenty cranial pleces representing 13-16 individuals. MATERIAL. Maxilla: SKW 12. Mandible SKX 21204. 150lated teeth: SKX 3300 RI', 3354 LP', 3355 LM', 3356 LM', 3601 LM', 3890 RP', 5002 LM" 5004b L1" 5007 Lc, 5014 RM o, 5015 LM 3 , 5023 RM" 5024 LM?, 6013 Re, 6277 LM', 7781 3 Rp , 16060 Rdi', 19031 RI'. Hominid postcranial material:

MATERIAL. Fifteen pieces: Humerus pieces: SKX 3774, 34805. Radius pieces: SKX 3602,3699,12814. Ulna piece: SKX 8761. Melacarpal: SKX 5020. Metatarsal: SKX 5017. Phalanges: SKX 5018, 5019, 5021,5022,8963,9449,45690.

cf. Cercopithecoides sp. MATERIAL. A single spectmen possibly belonging to this species: tsolated RM', SKX 21257 + 21258. Papio hamadryas robinsoni Sixty specimens representing a minimum of one juvenile and five adults.

MATERIAL. Maxilla pieces: SKX 13261 + 13262. Mandible pieces: SKX 14029, 38043. Isolated deciduous teeth: SKX 14056, 14057, 9942. Isolated incisors: SKX 39899, 39694, 38434. Isolated upper teeth: SKX 8923, 4877, 8894, 8851, 38008,4601, 38269, 38553, 39272, 5801, 12049, 12050, 12051, 14571,21314. Isolated lowerteelh: SKX 40145, 8067, 38403,39212,8836,4807,8980,8635,6771,8375,6772, 37995, 13919A, 13932, 13939, 13940, 38378, 14397, 8395, 38138, 16246, 14146,8394,4810,40198,39793, 14145, 8690, 21749, 39275, 39276, 38556. Isolaled tooth pieces: SKX 40166,40132, 7181A, 6797. Theropithecus oswaldi daniefi

MATERIAL. Rlqht mandible SKX 9579 with P" M,-M 3 from an adult.

2854 2501 1683 1096

550 243 107 58 17 8 6 8923

29,7 28,0 18,9 12,3 6,2 2,7 1,2 0,7 0,2 0,1 0,1

125 182 117 119 106 66 23 13 4 1 1 757

16,5 24,0 15,5 15,7 '4,0 8,7 3,0 1,7 0,5 0,1 0,1

Peoio (Dinopithecus) ingens MATERIAL. An isolated LP" SKX 38376, from an adult.

Large papionid lndet, Eight specimens.

MATERIAL. Isolated moiars or motar fragments: SKX 4664, 13768,14074,37996,38432,38937,39610,39914.

Cercopithecoid indet. Ninety two cranial and 41 postcranial specimens from a minimum or five lndívlduals.

MATERIAL. Cranial pieces: SKX 38210, 39972, 40307 + 40308 + 40309. Maxilla piece: SKX 40423 + 40427-40433 + 40435. Mandible pieces: SKX 9617, 15329. lsolated canines: SKX 12105, 12001, 9532, 47076, 38266, 38265. lsolated incisors: SKX 40126, 39990, 39915, 38248,15151,14755, 14754, 13981A, 13719, 13542, 12732-12739, 12095, 11840, 8924, 4808, 40337, 11944, 39768, 38636, 38375, 38094, 21678,21373,18234,16229,13442, 12766A, 11943, 9605, 9028,8454. Isoiated teeth: SKX 5939,13711,12055. Molar fragments: SKX 13769, 39267, 39005, 38401,38262,38136, 38052, 14480, 13701, 10046, 8662, 7984, 6773, 40498, 38256b, 6278, 9519, 21309, 14548, 14075, 13935, 13934, 13933, 13767, 13127, 11841, 11652, 11623,9704, 8999, 8287, 6536, 6156, 5210, 5063, 4809, 4620, 11665. Atlas vertebra: SKX 14303. Axis vertebra SKX 39788. Scapula fragment: SKX 9634; Humerus piece: SKX 38029. Radius pieces: SKX 9113, 12751, 14802,40080,45621. Ulna pieces: SKX 8614, 13365, 39862, 40397. Pelvis piece: SKX 9906. Femur pieces: SKX 5041,8566,8731, 9773A, 14121, 21130a. Patella: SKX 14200. Tibia pieces: SKX 5732. Astragalus: SKX 13477. Calcaneus piece: SKX 14122. Metapodial pieces: SKX 38529, 38221. Phalanges: SKX 38491, 40125, 14582, 13894A, 13894, 13676, 13631, 13079, 8367, 6892, 6523, 5917,4935,4345, no number.

COMPOSITION OF SONE ACCUMULATIONS

39

Table 3 NISP =

Macrovertebrate taxa ideritltied from Swartkrans. number of specimens: MNI = Minimum number of individuals.

NISP

MNI

Memoer 3

Member 2

Member 1 Lower Bank %MNI

NISP

MNI

%MNI

0,6 8,1

6 31 9

2 17

1,5 13,0

70 2

8 1

6,1 0,8

NISP

MNI

Member 5

%MNI

NISP

MNI

%MNI

Class Mammalia Order Primates Family Horninidae

Theropithecus oswaldi danie/i

Papio (Dinopithecus) ingens Large papionid mdet. Cercopithecoid indet. Order Carnivora Family Felidae Subfamily Felinae Panthera pardus Acinonyx jubatus

FeNs caracal Fefis cf. Iybica Felis serval Panthera leo Subfamily Machairodontinae Megantereon cu/tridens Family Hyaenidae Subfamily Hyaeninae Hyaena brunnea Chasmaporthetesnitidula Crocuta crocuta Hyaenid mdet. Subfamily Protelinae Proleles sp. Prote/es cristatus Family Canidae vutpes chama Small canto gen. & sp. indet. Canis ct. mesame/as Large canid gen. & sp. indet. Canid indet. Family Mustelidae Aonyx capensis Family Viverridae Subfamily Herpestinae Atilax sp. Cynictis penicillala Herpestes ichneumon Suricata suricatta Mongoose indet. Subfamily Viverrinae cr. Genetta tigrina Viverrid indet. Carnivore tndet.

1 20 15

1

13

12 11

9

3,8

95 4

11 4

4,7 1,7

0,6 10 60 1 1 8 133

6 1 1

46

4

3,7 0,6 0,6

5 80

2,5

41

246

2

1,5

53 1

3

1,7

8

5 1

2,1 0.4

12

0,6

0,6

1 3

1 2

0,6 1,2

21

6

1 3

2

2

0.4

13 3 3 25

2 1 1

1,5 0,8 0,8

15 4 3 65

3 2 1

0,6 1,9

1,3 0,9 0.4 2

1,1

8

3

1,7

2 105

2 5

1,1 2,8

1 2

0,6 1,1

0.4

46 4

4 1

1,2

2 "1

17

2

0,9

95 14 1

5 1

2,1 0.4

0,8

5

2

0,8 0,6

3 4

1,5

6 14

3,0 0,8

2

0,6

1 3 192

2 67

88

6 0,9

0.4 0,4 5

2,1

Order Tubulidentata Family Orycteropodidae Orycteropus afer

0,6

2

1,2

3

0,8

2 37

0.4

2

2 5

0.4

0.4

15

2

0,6

2 5

0,4

0,6

2

6 53

2 0,6 0,6

1 2

Order Pholidota Family Manidae Manis sp.

Order Proboscidea Family Elephantidae el. Elephas sp.

0,6

5

Horno sapiens

Horno sp. Australopilhecus robustus Hominid pcstcraníal family Cercopittlecidae ct. cerooounecotaee sp. Papio hamadrayas ursinus Papio hamadryas robinsoni

0.4

0,6

COMPOSITION OF SONE ACCUMULATIONS

40

Table 3 - continued

Order Hyracoidea Famlly Procavlidae Procavia transvaalensis Procavía antiqua Procavia capensis Hyracoid indet. Order Perissodactyla Famlly Equldae Hipparion Iybicum steytleri Equus capensis Equus burchellii Equid jndet. Order Artiodactyla Family Suldae Phacochoerus sp.

ct. Tapinochoerus meadowsi Suid indet. Family Hippopotamidae Hippopotamus sp. Family Glraffidae Giraffid gen. & sp. indet. Family acvíeae Subramily Alcelaphinae Megafotragus ap. Connochaetes sp. Médium sized aíceiacmoe: Beatragus sp.lAlcelaphus sp. Damaliscus sp. Alcelaphine indet. Subfamily Antilopinae Antidorcas australis/marsupialis Antidorcas recki Antidorcasbondi Antidorcas sp. ?Gazella sp. Oreotragus oreotragus Raphicerus campestris Antilopini/Neotragini indet. Subfarnily Bovlnae

Syncerus sp. Taurotragus oryx TrageJaphus strepsiceros

Member 3

Member 2

Member 1

Member 5

NISP

MNI

%MNI

NISP

MNI

%MNI

NISP

MNI

%MNI

NISP

45 205

3 17

1,9 10,6

34 187

3 10

2,3 7,6

31 218

5 11

2,1 4,7

31

4

2,2

462

37

20,7

0,4 2,1 0,4

7 '9

2 6

1,1 3,3

0,4 0,4

12 4

2 1

1,1 0,6

MNI

%MNI

5

8 8

1 3

0,6 1,9

14

9 10

, 3

0,8 2,3

13 36 4 21

3 1

2,3 0,8

12 6 23

3

3 1

0,6 0,6

15

0,8

17 1

1 5 1

0,4 2

0,8

2

0,4

7 56

1 7

0,8 5,3

7 199

3 17

1,3 7,3

2 16

1 3

0,6 1,7

1,9 1,2

16 30 34

3 6

2,3 4,6

54 80 27

6 6

2,6 2,6

42 35 16

5 4

2,8 2,2

11

6,8

43 9

10 2

7,6 1,5

93 1

'8 1

7,7 0,4

29

6

3,3 25,7

2 5 1 1

1,2 3,1 0,6 0,6

337 5

46

2 9 1 4 3

7

2

1,1

3

2

1,2

54

7

4,3

13 14 18

3 2

22

13

5

25 4 9 8

14 1 3

6,0 0,4 1,3

1,5 0,8

9 7 2

3 1 1

1,3 0,4 0,4

2 1

0,8

5 5

3

1,3

15 2

3

1,7

6

2

0,9

13

3

1,7

1 139 1562 138 3 873

1 '4

0,6 2,2

"

0,6

0,8

2 7 2

3,8

2 1

0,6 0,6

Subfamily Hippotraglnl

1 4

Hippotragus cf. niger HippotraginilReduncini lndet. Subfamily Pelinae Pelea sp. Subfamily Reduncinae

Redunca

ct. arundinum

Bovid síze class I Bovid size ctass 11 Bovid size class 1I1 Bovid stze cíass IV Bovid lndet. cramal'" Order Lagomorpha Lagomorph gen. & sp. lndet. Order Rodentia Family Pedetidae Pedetes sp. Family Hystricidae

Hystrix africaeaustraJis Mammal indet.

67 372 300 30 2714

'1

74

9

5 14 103

2

56 330 128 13 2002

'1

5,6

28

4

0,6

0,6

1,2

0.8

146 731 520 57 2125

3,1

72

2

0,8

3

3

0,8

73

5 115

7

3,0

40

2,2

0,6

0,4 2

4

0,9

0,6 88

Class Aves

Struthio sp.

10

0,6

COMPOSITION OF BONEACCUMULATIONS

41

Table 3 - continued Member

Member 1

NISP Ibis FrancoHn Bird size class I Bird size crass 1I Bird size class 111 Bird size class IV Bird indet. (atl size crasses) Class Reptilia Chelonia joder. Beptile lndet.

2 3

214

MNI 1 2

10

26 11

%MNI

NISP

MNI

2'

Member 3

Member 5

%MNI

NISP

MNI

%MNI

1 5 8 30 5 3

0,4 2,1 3,4 12,8 2,1 1,3

0,6 1,2

NISP

MNI

%MNI

0,6

6,2

187

10

7,6

2 9 40 279 49 19 130

0,6 0,6

26 11

2 1

1,5 0,8

57 26

2 2

0,9 0,9

28 4

18,0

9

5

3,8

9

2

0,9

14

6

3768

131

6173

234

4187

178

6 29 7 8 3

2 3 1 2

1,1 1,7 0,6 1,1

0,6 0,6

Class Amphibia

frog/toad indet.

75

29

4950

161

3,3

• MNI where tne totat mínimum number of inowíouels in the indet. category exceeds that in the idenlifiable categoríes. •• Bcvrd horneare fragments and enamel tooth fragments.

Order CARNIVORA Family FELIDAE Panthera pardus Eight orante! pieces and 38 postcraníal pieces from a minimum number of cne juvenile and three adults. MATERIAL Mandible piece: SKX 45403 + 45689. Isolated canines: SKX 5099, 8113, 8528. Isolated premolars: SKX 9145,9389,9402,13053. Scapula piece: SKX 8069. Humerus pieces: SKX 6134A, 13229,45393. Ulna pieces: SKX 6345 + 6346, 8721, 13232, 45449. Patella: SKX 4203. Tibia pieces: SKX 8164, 11925,45395. Tarsal: SKX 14332. Metapodials: SKX 7122, 5471, 5736, 4922, 4168, 9715, 13227, 14701, 21337,45404,45405,45408,45450,45613. Phalanges: SKX 7225, 6747, 6746, 6745, 8109, 8110, 8896, 45399, 45402, 45537.

Feils caracal MATERIAL. Distal metapodial, SKX 5792.

Hyaenid indet. Four cranial and 17 postcranial pieces from a minimum of one juvenile and one adult. MATERIAL. Isolated incisors: SKX 4576, 14408. Isolated canine: SKX 9479. Premolarfragment: SKX 14481. Axis fragment: SKX 7854. Scapula piece: SKX 9571. Radius and ulna pleces: SKX 7822, 8673 + 8675, 8674, 45475. Tibia piece: SKX 6988. Calcaneus: 6990. Tarsal bone: SKX 5795. Metapodials: SKX 13664, 4673, 7971, 45515. Phalanges: SKX 7067,13321,13722,47091.

Subfamily Protelinae Proteles sp. One cranial and one postcranial piece from a minimum 01 one individual. MATERIAL. Maxllla piece: SKX 13280. Humerus piece: SKX 45394.

Family CANIDAE Felis serval MATERIAL Right astragalus, SKX 4116. Panthera leo Two postcranial elements from a minimum 01one individual. MATERIAL. Phalanges: SKX 8484A, 45401.

Family HYAENIDAE Subfamily Hyaeninae Hyaena brunnea MATERIAL. Premolar tragment: SKX 5845. Chasmaporthetes nitiaute Three cranial pieces from a minimum ot two individuals. MATERIAL. Left lower canine: SKX 13481. tsolated premolar pleces: SKX 5414, 9720.

Small canid gen. & sp. indet., fax size Six postcranial pieces from a minimum of one individual. MATERIAL. Tibia piece: SKX 45528. Tarsal bone: SKX 13575. Metapodial pieces: SKX 7135, 8605, 13538, 13557. Canis ct. mesomelas Twenty four craníal and 29 posteranial pieces from a rninimum of three individuals. MATERIAL. Maxilla piece: SKX 8833. Mandibie pleces: SKX 16228,16230 + 16256, 16231. Isolated incisors: SKX 7875, 4933,4934,5047, 5314, 8844, 14069, 14772. Isolated canines: SKX 4929, 4930 lsolated premolars and rnolars: SKX 4566, 4843, 4937, 4938, 4939, 4940, 6175, 7997, 8084, 10021, 13668. Radius and ulna pieces: SKX 4585, 9542, 47216. Carpal bone: SKX 12111a. Femur piece: SKX 9149. Tibia piece: SKX 13422. Astragalus: SKX 13421. Calcaneus:

42

COMPOSITION Of 80NE ACCUMUlATIONS

SKX 5864. Metapodials: SKX 13131,4793,8355,8356,893"1, 13384, 15044. Phalanges: SKX 6602, 5447, 7030, 7994, 8016;8289,8357,8358,8424,11866,12070,13323,18177, 18988. Family MUSTELlDAE

Order PROBOSCIDEA Family ELEPHANTIDAE ct, Elephas sp. Fifteen craniaí pieces from a minimum 01 two individuals.

MATERIAL. Ivory fragments: SKX 45691 + 45692a-k. Unerupted tooth fragments: SKX 7457, SKX 139816, SKX 14961.

Aonyx cspensis

Six cranial elements from a mínimum 01 twa Indivkíuals. MATERIAL. Mandible pieces: SKX 13508, 14686. Iso/ated canine: SKX 13123. Isolaled molars and premolars: SKX 8323,8719,9717.

Family VIVERRIDAE Herpestes ichrteumon

One crarual and one postcranial piece fram a minirnum of one individual.

MATERIAL. Upper premolar: SKX 14520. Metatarsal: SKX 6939. Mongoose mcet. Five postcranial pieces from a mínimum of two individuals. MATERIAL. U/na piece: SKX 9380. Tibia pieces: SKX 9737a, 21220, 45610. Calcaneus: SKX 21225.

Order HYRACOIOEA Procevie transvaaJensis Twenty nina cranial and 16 postcranial pieces from a mlnimum of three indlviduala. MATERIAL. Mandlble pleces: SKX 14178, 14547+14551. 14631, 14643. Isolated upper mclscrs: SKX 5085, 4619, 8056, 8861, 8861A, 8869, 8918, 13597, 14831. Isolated upper prsmolars and rnolars: SKX 4621, 7963, 12786, 8627, 8944, 8957.lsolated lower incisors: SKX 13411,7968,8640, 14687, 12266, 12787A. lsotated lower molars and premolars: SKX 4916,5045.8876,8636. Humerus pleces: SKX 14302,6287, 14774,15391. Radius pieces: SKX 8100, 9954,14598. Ulna preces- SKX 8039, 14264. Femur plece: SKX 15238. Calcanea: SKX 13320, 8863, 14644. Metapodlals: SKX 4724A, 5377,13879.

Procavia antiqua One hundred and forty slx cranial and 59 postcranial plecas

Carnivore indet. Thirty one cranial and 57 postcranial pieces. MATERIAL. Maxilla piecea: SKX 10021,6016. lsolated íncísors: SKX 8470, 891 0,13089, 17266a, 21124, 40434.lsolated canina pleces: SKX 4963, 5324, 5809, 8532, 8714, 8911, 9229, 11722, 11903e, 13042, 13065, 14605, 16267,45398, 47182. rsoratec tooth fragments: SKX 4711, 9300, 9524, 13380,14058,14111,14293,14730 Vertebrae: SKX 7149, 7799, 7880a. Humerus pieces: SKX 45436. Radius and u/na pieces: SKX 15050, 18148. 45552. 14752, 21750. 45573, 45594,47210. Pelvis piece: SKX 13269. Femur pieces: SKX 45603, 45643. Patella: SKX 4339, 9823, 13878. Calcanea: SKX 4926, 17330. Astragalus: SKX 4284. Metapodials: SKX 5344, 6858. 8200, 9420, 11720, 12023b, 13922, 15420, 15640, 45559, 45623. Pha/anges: SKX 4363, 4350, 5071, 6914,8683,8738,8899,9122,9193,9203,9268,9306,9323, 9381, 13220, 13898, 13988, 14033, 14973, 14982, 17982, 18093,18111,18156,40288.

carmvore moet.: coprolite pieces Flfty were recovered. SKX 4167, 5353, 5569,5605, 5672, 5685, 5734,5851, 5854, 5888, 5906, 5920, 5963a, 6002, 6010, 6023a, 6102, 6116, 6131. 6457, 6712,6799,6894, 6973,6998,7025,7812,7844,7947,8041,8383,8522,8601, 8688,8774.9019,9074,9090,9106,9200,9308,9668,9681, 9734,9825,9934,13163,14610,14963,18005.

Order TUBULlDENTATA Family ORYCTERDPIDAE Orycteropus afer MATERIAL. Right humerus piece: SKX 14261.

from a minimum 01 17 individuats, at least two of whlch were

juvenlte. MATERIAL. Maxilla pieces: SKX 4717, 5805, 5903, 6597, 9958,10054,12020 + 12035, 13503, 13568, 13787, 17169, 17190,21346. Mandible pieces: SKX 4101,4402,4761,4762, 5050, 5120.5254,5451,8008, 8483A, 9983, 11538 + 11539, 13235,13412, 17997A, 18088,21359,40360,40403,47096. Isolated upper incisora: SKX 4379, 4409, 4410, 4426, 4902, 4918, 4974, 5190, 5262 + 5263, 5284, 5755, 5936, 6381, 6587,6762,7133,7158,7167,7218,7795,8037,8148,8269, 8706,8770,8882,8928,9002,9038,9051,9060,9339,9478, 9507,9625,9645.9749,10004,11603,13115,13387,13410, 13479,13659, 13695,13786,13835,14002, 14256, 14283, 14739,14878,15081,18089,18090,18102,18110,18114, 18140,18180,18209,18240.18241.18643.40533 + 40534. Isolated upper premolars and molars: SKX 5072,5216,5365, 5740,8884,8933,9248,9982,13788,14609. Isolated lower incisors SKX 5101, 5102, 5151, 5346, 5853b, 6187, 6426, 6587,6922,7964,8066,8889,9434,9629, 13413, 13974, 14107 + 14108,14398,14773,14851,15063,18181.18837, 40S02a+b. laotated lower motare and premolars: SKX 4844, 5124,5699,5763,8046,8611,9604,13182,14262,14621, 14770,18096,18619. room fragments: SKX 8641. Scapula ptece: SKX 4312. Humerus pieces: SKX 5157, 5158. 5690, 7770,7820,8552,8801,11528,11670,11881,12795,13609, 14665, 17135, 18246, 40555, 45446. Radius pieces: SKX 4613, 5318, 5433, 7780, 11985, 18223. Ulna pieces: SKX 4261,6184,7800,11563,13313.13405,13480,45681. Fe· mur pieces: SKX 4387, 4821, 5364, 5407A, 8282, 13423. 40420. 45567, 45601. Tibia pieces: SKX 5486, 8707. caicanea: SKX 15061, 4721A, 6033, 6207, 11596b, 13567, 13975. Metapodials: SKX 7788,8011,6873,7771,13193, 14078, 14501, 15667, 18904

COMPOSITION OF BONE ACCUMULATIONS

Hyracoid indet. Flve cranial pleces MATERIAL. Maxilla pieces: SKX 7796, 8718. Mandible pieces: SKX 6185, 7893, 9488. Order PERISSOOACTYLA Family EQUIDAE

Hipparion Iybicum steytteti Two cranial ano six postcraniaJ pieces from a minimum 01 one adult. MATERIAL. Isolated teeth: SKX 4195, 5832. Metapodials: SKX 6778, 8972,13642. Astragalus: SKX 5967. Tarsal bone: SKX 8967. Phalanx fragmen/: SKX 9166.

Equus capensis Eignt postcranialpieces from a minimum ot one juvenile and two adults. MATERIAL. Radius piece: SKX 45561 Tlbia piece: SKX 9596. Metatarsal piece: SKX 15351. Astragalus pieces: SKX 6288,6314, 12760. Tarsal pieces: SKX 4791, 15075. Equid lndet. Eight cranial and six postcranial preces. MATERIAL. Tooth pieees: SKX 6407, 8559a, 9989,13494, 14222,14522, 18072a, 21313. Femur pieee: SKX 9662. Astragalus pieces: SKX 5985, 12757b, 15370. Tarsal fragmen!: SKX 6001. Phalanx piece: SKX 8501. Order ARTIODACTYLA Family SUIDAE Phacochoerus sp. Two cranial and one postcranial piece from a mínimum 01 one adult. MATERIAL. Isolated molar fragments: SKX 6041, 8782a. Second phalanx: SKX 45577

cí, Tapinochoerus meadowsi One cranial and one postcranial piece from a mínimum 01 one individual. MATERIAL. Molar fragmen/: SKX 33464. Left proximal metacarpal: SKX 47167

43

8892,8526,14394. Maxilla fragments and associated teeth: tour right maxilla fragments with Rp 3_M': SKX 21104-21107, 21110-21113; left maxilla fragments with M'_M 3 : SKX 4996ac. Isolated permanent upper teeth: SKX 17272, 4985, 5707, 7919, 8461A, 45553, 8530, 8891, 13630, 40561, 40387, 45499, 4560, 9353, 13507, 6200, 16291 + 16292, 45464, 21121, 9254, 4931. Assoeiated lower dentitlons: SKX 13821, 13822,13823,6194,6195. Isolated permanent lower teeth: SKX 14769a, 6197, 6608, 5464a, 5475, 8023A, 8893, 4670, 6232,7955,9411 a, 5843, 13159. Molarlragments: SKX 4397. 45547, 45636. Horncore fragments: SKX 45420, 45421, 45479,47132.

Medium-sized alcelaphine: Beatragus sp. IAlcelaphus sp. Thirteen craníal pieces from a rnlnlrnurn of one juvenile and two adults, MATERIAL. Right mandible fragment with two deciduous teeth: SKX 4932. lsolated deeiduous teeth: SKX 9644, 40342. Isolated permanent upper teeth: SKX 16247,40318,5688. Lett mandible with associated molars: SKX 4990, 4991.lower molar fragment: SKX 5844. Horncore fragments: SKX 4993, 6514,45441,47136. Damaliscus sp. Fourteen cranial pieces trom a mínimum of one juvenile and

one adult. MATERIAL. Deciduous teelh: SKX 11927, 5361 and 5367. Left maxilla fragments with associated prernolars: SKX 5312, 5331. Isolated upper permanent teeth: SKX 4829, 4299, 12839a. Lower permanent teeth: SKX 14151,5833,13389, 7207. Permanent molar fragmenls: SKX 6408, 7208, 12819. Atcelaphlne indet. Eighteen cranial preces. MATERIAL. Isolated deeiduous premotars fraqrnent: SKX 9570. Isolated permanent upper dentitions: SKX 4365, 7064, 7863,9752,9835,11674,14574,18054. Isolated permanent lower tooth: SKX 5311. Molar Iragments: SKX 14153, 4915, 6158, 6760. Horneore fragments: SKX 9692, 9824, 14863, 45429. Subfamily Antilopinae

euetretie and/or marsupialis Twenty two cranlal pieces from a minimum of five juveniles and six adults. (See Table 7 for age protile). MATERIAL. Deciduous premolar in rightmandible fragment: SKX 8293.lsolated permanent upperteeth: SKX 17239, 5821, 12067,12068, 14250. rsclateo permanent lower teeth: SKX 4191,4914,11602, 8455a, 8455c, 6331, 6432, 13571,4878, 7066. Molar Iragments: SKX 4842, 9784. Horneore fragments: SKX 5683, 6822, 7748,13610. Antidorcas

Family HIPPOPOTAMIDAE Hippopotamus sp.

Fifteen cranial pieces from a minimum 01 one individual. MATERIAL. Isolated premolars and molars: SKX 6840, 9684,11742. Tusk fragments SKX 4822, 9621, 9721. Tooth fragments: SKX 11736 + 11737, 11738, 11739, 18073, 14685, 11703,11552,14011,6847. Family BOVIDAE Sublamily Aleelaphinae Connochaetes sp. Fifty four cranial pieces from a minimum of two juveniles and five adulta. MATERIAL. lsolated deciduous teeth: SKX 21431, 11880,

Aottaorces sp. Two cranial pieces from two adutts. MATERIAL. Isolated permanent tower third molars: SKX 13337, 13570. ?Gazella sp. Nrne horneare preces from a minimum of tive individuals,

44

COMPOSITION OF BONE ACCUMULATIONS

MATERIAL. Horncore pieces: SKX 4921, 5978 + 9782, 9079,9884,11901,12759,13703,14689,45400+45411.

5173,6465,6316,6283,5668,18228,18200,18196. Metapodials: SKX 8854, 8929, 14653.

Oreotragus oreotragus

Order RODENTIA Family PEDETIDAE

A right mandible fragment with P4-M3 in situ from an adult.

MATERIAL. SKX 4988. Pedeles sp. Raphicerus campestris

Four craníal pieces all possibly Irom the same adult, MATERIAL. Isolated upper permanenl leelh: SKX 5285, 8494. Isolated lower permanenl teelh: SKX 8535A, 14150.

Five cheekteeth from a minimum of one individual.

MATERIAL. Isolated cheekteeth: SKX 7123, 18171, 18186, 7982,15062. Family HYSTRICIDAE

Antilopini/Neotragini índet. Three associated permanent lower cheekteeth.

MATERIAL. SKX 13353. Sublamily Bovinae

Syncerus sp.

Hystrix africaeaustralis Eleven cranial and three postcranial pieces trorn a minimum of one juvenile and one adult.

MATERIAL. Isolaled incisor Iragments: SKX 8051, 8760, 14535. lsolateo molars: SKX 4891, 4892, 5212, 7867,13142, 14286,21037,21040. Metapodials: SKX 6678, 7151, 8855.

Three cranial pieces from a minimum 01 cne juvenile and

cne adulto MATERIAL. Isolaled declduous teeth: SKX 5421, 1630B. tsoíated permanent lower prernolars: SKX 21032 + 21033.

Bovid size elass I Six cranial and 61 postcraníal pieces from a minimum 01 three individuals, one juvenile and two adults (Tabla 4c). Bovid size c1ass 1I Forty cne cranial and 331 postcranial pieces from a mínimum 01 cne juvenile and five adults (Tabla 4a).

Bovid size class 111 Twenty five cranial and 275 postcranial pieces from a rnini-

mum 01 two juveniles and seven adults (Table 4b).

Bovid size class IV Two eranial and 28 posteranial pieees from a minimum of

Mammal indet. Three cranial and 95 postcranial pieces.

MATERIAL. Maxilla piece: SKX 18204. Mandlble pieces: SKX 8327, 9279. Scapula pieces: SKX 13529, 11729, 5592, 5049,4156. Humerus pieces: SKX 47151, 7165, 5689,18824, 14950,8150,6764. Radius pieces: SKX 471 44, 8379a, 4561 9, 8307. Ulna pieces: SKX 13573, 13364, lB213, 18144,14561, 14065, 8268, 7819, 5213, 421B. Carpal bone: SK 11955. Pelvis pieces: SKX 13519, 9114, 8315, 16230, 14841, 14678, 14295, 13886, 13617, 13433, 11596c, 6661, 6115. Femur pieces: SKX 40463, 9212, 6032, 5659, 4739, 45687, 9683, 9048,5693,5427,4980,4709,4552,4285,4194,4184. Tibia pieces: SKX 47125, 45512, 45506, 8529A, 8751, 8519, 8061, 5286. Patella: SKX 21139, 9598, 9190, 6586. Calcanea: SKX 18166, 14077, 9452, 5347, 4221. Melapodials: SKX 45675, 15055, 14742, 13645, 13583, 13501, 13224,9643,8874, 8131, 8122, 6601, 6305. Phalanges: SKX 21659, 21382, 15207, 1470B, 14502, 9432,8768,524B,4941.

one juvenile and one adult (Table 4c). Bovid indet. cranial (all síze elasses) Enamel fragments: 2206. Isolated mcísors: 40. Premolars: 17. Horneore fragments: 397.

Order LAGOMORPHA Lagomorph gen. & sp. indel. Nineteen eranial and 55 posteraniaJ pieees from a minimum of nine individual s, at least two ot whieh were juveniles.

MATERIAL. Palatine: SKX 6368. Maxilla pieces: SKX 9821, 13038. Mandible pieces: SKX 2993, 21786, 14702, 6213, 5244,40472,8771,6631,5918,45660. Isolated cheekleeth: SKX 8960, 8846, 8767, 8119, 7965, 5348. Scapula piece: SKX 45595. Humerus pieces: SKX 5250, 6011, 8388, 40297, 45541,45686. Ulna pieces: SKX 9822,14691,5150. Pelvis pieces: SKX 5039, 5104, 8101,11627,11592,15330,5626, 13448, 45488, 45539, 4708B, 47089. Femur pieces: SKX 4553,9480,4425,6729,8089,8172,11798,45532,45632. Tibia pieces: SKX 8813, 7031, 45469, 45540, 45615, 45622, 45653. Calcan ea: SKX 531 9,6329,8010,9322,13373, 4723A,

Class AVES Ibis indet. Two bill pieces from a minimum of one individual.

MATERIAL. SKX 5894, 9737c. Francolin indet, Three spurred tarsometatarsi from a minimum 01 two indi-

viduals, MATERIAL. SKX 4510, 5913, 6209. Bird indet. (all size classes) Three cranial and 211 postcranial pieces from a mínimum ot ten individuals.

MATERIAL. Cranial pieces: SKX 1001, 2937, 3779. Keel pieces: SKX 880B, B990, 10056-59, 47090, 47084, 45679, 40333. Coracord pieces: SKX 47095, 45609, 40364, 8757, 18806D, 1803B, 15B58, 15714D, 15530B, 18053, 18021, 5046,5539,5678,5981,4965,6121. Scapula preces: SKX 5544, 4512. Humerus pieces: SKX 15797, 40365, 11258,

COMPOSITION OF SONE ACCUMULATIONS

6380,5234,4923,17148,5778. 15714c, 17256.5517, 14023, 11568. 11719, 5817, 5097, 9414, 47077, 18806E, 11191, 4214,11190,4908,6260,45542,45470,38186,4113,21473, 17210,11618,17313,4924,6183,5725,6071.6320,5573, 5698.6229.4166,4408. 11863. Radius pieee: SKX 47205. Ulna pieces: SKX 47092. 40156, 6774, 18806c, 17252, 15530e, 12432, 12833, 11591, 11515, 11213, 10261,5982, 5938,4411,4331. 4401, 5721, 5723, 5860, 5877, 11516, 17147,11993,11971,11721,11708. Carpometacarpus and wing digit pieees: SKX 14499, 12078d, 5174, 19004, 157146, 6321,5269,5697. Vertebrae: SKX 40215. 3864, 38217, 4769, 3253. Saerum: SKX40347a, 15491,2496,4545,5323 Femur pieees: SKX 7989, 45618. 45617. 45598, 45556, 45534, 40529,4437,19026, 15714A, 15530A, 12416, 11919, 11316, 6208, 6097, 4514. 4436, 2992. 2794, 4732, 4733, 11557, 11559, 17220, 12762, 12056, 11820, 4403, 5540 + 5543, 5574,4982. Tibiotarsus pieees: SKX 8485a. 47104, 45607, 45481,40343,18013,17309.16264,155300,11865,11558, 11233,10160.5561,4406,4743,4819,4820,5560,5765, 5780,17170,16256,12731,11920.4622,4785 Tarsometatarsus preces: SKX 9778, 47181, 45656. 45406. 40514. 21057c, 17969, 15530E, 11562, 11308. 11266.21255,5748, 4662. 5552, 5729, 5895, 21253, 5303. Phalanges: SKX 13228.6336, 15647, 13548,5780, 18625, 13612. Miseellaneous fragments: SKX 6893, 17802, 17838, 15833, 15748, 6064A, 5890, 5889,4326,4242.

Class REPTlLIA Order CHELONIA Family TESTUDINIDAE

Chelonia indet. Twenty stx carapace pieces from a mínimum of one individual. MATERIAL. SKX 4678, 4704, 5053, 6648, 6833a, 6948, 7774, 9160, 9762, 10053, 11968, 12096, 13039, 13061, 21195, 21272, 21273, 21281, 21282, 21284, 21286, 21697, 21698,21699,21700,21701.

45

18097, 14483, 13977, 13731,9990, 6686, 5541,5328. Pelvis piece: SKX 8254.

DISCUSSION Comparad with Member 1 material obtained between 1948 and 1975 and analysed by Brain (1981), the current material presents a similar picture. In raspeet 01 nurnbers, the australopithecines outnumber Homo sp., and Peplo h. robinsoni ls as well-represented as in the earlier sample. The specimen usted aboye as Cercopithecoides sp. bore certain resernblance to Paracolobus mutiwa (E. Oelson personal communication), but beca use this identification is uncertain it has been raferred to 'Cercopithecoides sp.' Felids are well represented in the more recent excavation, wttn caracal (Felis cereceñ, serva! (Felis servafJ, lion (Panthera leo) and leopard (Panthera pardus) recorded. Serva1 and caracal had not been recorded from Swartkrans before, although caraca¡ was found at Sterkfcntein in Member 2 (Turner, 1987). Elephant ívory was found in the present excavation, and teeth and ivory of a hippopotamus were recave red. The identification of the ?Gazella sp. in Members 1-3 was based on horneares that did not match the known Antidorcas specíes from the Sterkfontein valley caves and appeared to be rnost Hke those ot Gazella sp. The absenee of dental or other identifiable material of Gazella sp. is inexplicable. The frogs and toad s that have been recorded may have fcund their way tnto the cave by way of owl pellets. However, the amphibians may also have inhabited pcols in the vicinity of the cave and formad the meals of other animals that utilized the cave, or might have succumbed to adverse conditions and their bones washed into the cave.

MEMBER 2 Figure 2 i!lustrates the macrovertebrate taxa found in this Member and the material recave red is listed below.

Class MAMMALlA Order PRIMATES Family HOMINIDAE

Reptile indet. Eleven postcranial pieces from a mínimum of ene individual. MATERIAL. Vertebra: SKX 40304. Limb-bone pieees: SKX 4404, 9737b, 9936, 11697d, 12838b, 21766b, 40305,40527, 40480,40510.

Homosp. 8ix crarual pieces trom a mínimum of two indiviouals. MATERIAL. Isolated teeth: SKX 257 RM,I, 258 LM" 267 Rdm', 334 RM', 339 RI'. 610 RI'.

Class AMPHIBIA

Australopithecus robustus

Frog/toad tnoet. Seventy five postcranial pieces trom a minimum of 29 individuals. MATERIAL. Humerus pieees: SKX 18987, 18985, 18978, 18955,18943,18939,18929,18928,18915,18896,18870, 18865,18830,18820,18815,18745,18709,18703,18698, 18252,18243, 18240A, 18238, 18230, 18226, 18219, 18173, 18165, 18150, 18124, 18109, 18094, 18087, 18043, 13914, 13819,13698,13499,13256,13158,12785, 12023c, 9499, 9398,9345,9336,9115,6416,6240.6761,6804,6239,6812. Urostyle pieces: SKX 47174.18935,18930,18921,18860, 18673.18262.18259,18247,18178,18130,18119.18104.

Thirty one cranial pieces from 17-22 individuals. MATERIAL. Maxillae: SKX 162, 265. Mandibles: SKX 4446. 5013. lsolated teeth: SKX 163 Rdm', 240 L1', 241 R,. 242 RI'. 257 RM" 258 LM,. 267 Rdm', 268 RM', 269 R', 271 L1', 272. Tooth fragments, 310 R', 311 LP3, 312 L', 313 L1', 333 RM" 334RM',339RI'.341 RP?,610RI', 1016RP3, 1017L1" 1312 LM', 1313 la 1437b LP'. 1788 L1'. 3559 RI,. Hominid postcranial material Nine elements were recovered. MATERIAL. Thoracic vertebra: SKX 3342. Triquetral: SKX 3498. Metacarpals: SKX 2954, 3646. Patella: SKX 1084. Metatarsal: SKX 247. Phalanges: SKX 344, 1261, 3062.

COMPOSITION OF BONE ACCUMULATIONS

46

Pepio hamadryas robinsoni Seventy specimens from a mínimum 01 tour juveniles and

tour acutts. MATERIAL. Calvaria pieces: SKX 821, 820, 819, 818. Deciduous teeth: SKX 3558,2774,2775,2780,2781,1177 (in maxiila fragment) 169 (in mandible piece). Maxillapieces: SKX 3218,814,815. Mandible pieces: SKX 12440, 12443, 12444, 2907,2828, 2469d. Isolated canines: SKX 3220,3221.3223, 3444,2494,1348,306, 12389.lsolated lowerteeth: SKX 3401 + 3402, 2900, 2278, 2779, 2270, 2773, 1629, 1630 + 1631, 12442,12445,3950,2956,2778,611,80,1593+ 1594, 613+ 617, 616.lsolated upperteeth: SKX 2777, 608, 610, 618, 619, 2902,1627,197,3254,2252,1344,817,816,354A.lsolaled tooth fragments: SKX 3224. 3225, 2776, 1626, 1628, 1632, 612,614,615.

Family HYAENIDAE Sublamily Hyaeninae Hyaena brunnea

Savan crantal and six postcranial pieces from a mínimum 01 cne juvenile and cne adult.

MATERIAL. Mandible pieces: SKX 424, 590 + 591, 595 + 596. Isolated premolar pieces: SKX 50 + 53, 51, 365, 1892, 3236 + 3237. Carpal bones: SKX 2469c, 2469f. Metacarpals: SKX 2469b. Tibia pieces: SKX 577, 2394. Astragalus: SKX 2469a.

Chasmaporthetes nitidula Three isolated premolar preces from a minimurn 01 one individual.

MATERIAL. SKX 635, 637, 638.

Theropithecus oswaldi dan/el;

Crocuta crocuta

Two cranial specimens from a mínimum 01cne individual. MATERIAL. A crushed cranium with maxillary and nasal regio n SKX 1oaa-b, and a LM, SKX 2995.

Three isolated tooth fragments possiblyfrom a mínimum 01 ene individual. MATERIAL. SKX 957,1146,15771.

Larga papionid indet.

Hyaenid gen. & sp. indel.

Five specimens. MATERIAL. Canine pieces: SKX 1031,1 077.lsolated teeth: SKX 2800, 2960, 2962.

Thirteen cranial and 12 pcstcrantal pieces from a mínimum 01 two individuals. MATERIAL. lsolated tooth fragments: SKX 3055 + 3056, 3028+ 3030, 3027, 3008, 3004, 1262, 1201, 1034, 1032.959, 642, 640, 592. Axis vertebra fragments: SKX 621 a-t, 545. Humerus piece: SKX 2541. Radius piece: SKX 1921. Carpal bone: SKX 1920. Metacarpal: SKX 2687. Patella: SKX 46919. Astragalus: SKX 15815. Metatarsal: SKX 552. Phalanges: SKX 191, 497,1185.

Cercopithecoid indel. Forty eight cranial and 32 postcranial pleces.

MATERIAL. Calvaria piece: SKX 182. Maxiila piece SKX: 3416. Mandible pieces: SKX 2817 + 2818,15974. Isolaled incisors: SKX 274, 307, 1041,1042,1043, 1131,2274,2420 + 2419, 2452, 3522, 3443, 3400, 1474, 1129, 777, 13162a. Isolated teeth and tooth fragments: SKX 2272, 2996, 3181, 1036, 1037, 1038, 1039, 1040, 2269, 2280, 15548, 3965, 3623, 3418 + 3419 + 3420, 3253, 3032, 3029, 3026, 2957, 2855,2745,1675,1674,1672,1604,386,457 + 458. Humerus pieces: SKX 15812, 810, 406. Radius pieces: SKX 15429, 3244. Ulna pieces: SKX 585, 408, 407, 788, 2666/2. Femur piece: SKX 221. Tibia pieces: SKX 581 + 584. Carpal/tarsal: SKX 15864. Astragalus: SKX 1217,396. Calcaneus: SKX 395, 2766. Metapodials: SKX 3081, 164, 1513/1,773,2688,3082. Phalanges: SKX 15532, 3270, 2813, 1669, 1680/10, 1790/3, 2374/4,1259. Order CARNIVORA Family FELlDAE Panthera pardus

Twelve cranial and 29 postcranialpieces from a minimum 01 two individua/s. MATERIAL. Skuil fragments: SKX 544, 632. Maxiila pieces: SKX 628 + 629. Mandible piece: SKX 355c. Isolated incisors: SKX 600, 630, 639. Isolated canines: SKX 593, 3544. Isolated premolars: SKX 812,813,1574. Scapula pieces: SKX 431 + 431a. Humerus pieces SKX 1337, 674, 604 + 605. Radius pieces: SKX 586 + 770. Ulna pieces: SKX 2880, 580. Carpal bones: SKX 2887, 2890, 2886, 2885, 785, 599. Metacarpals: SKX 1552 + 1554, 769, 587, 554 + 779. Metapodial pieces: SKX 558, 578, 782. 3808. Phalanges: SKX 65n, 555, 556, 557, 559, 772, 3307.

Family CANIDAE Canis cf. mesomelas Thirteen cranial and 33 postcranial pieces from a mínimum

01 tour individuals. MATERIAL. Maxilla piece: SKX 1130 + 1136 + 1136a. lsolated teeth: SKX 52,749,925,1052,1085,1115,1436,2465, 2772, 3016, 3990, 2616. Atlas vertebra piece: SKX 2477 + 2478. Axis vertebra piece: SKX 62. Humerus piece: SKX 496. Radius pieces: SKX 2695,1652,855,753. Carpal bones: SKX 2716,2742/8. Metacarpals: SKX 2870, 2792, 2693, 2691 + 2692,2690 + 2710. Tibia piece: SKX 677. Fibula piece: SKX 2538/11. Calcaneus: SKX 1655. Metapodial piece: SKX 962. Phalanges: SKX 1045a+b, 1700/1, 2346, 2707, 2708, 2709, 2711,2712,2714,3989,18794,15926,46634,15926,15897. Large canid gen. & sp. indel. Four terminal phalanges all possibly from the sama individual.

MATERIAL. SKX 61, 1386,2271, 15785.

Family MUSTELlDAE Subfamily Lutrinae Aonyx capensis

MATERIAL. Worn left upper canine: SKX 3222.

COMPOSITlON OF SONE ACCUMULATIONS

48

Family VIVERRIDAE Sublamily Herpestlnae Atilax sp. MATERIAL. Humerus piece: SKX 2877. Cynictis peniciltete

MATERIAL. Left proximal ulna: SKX 15531. Suricata suricatta

Two mandible pieces from a minimum 01two individuals. MATERIAL. SKX 286, 662. Subfamily Viverrinae Viverrid indet,

Two mandible piecas. MATERIAL. SKX 366,117211.

Carnívore indet, Thirty nine cranial and 28 postcranial pieces. MATERIAL. Skull piece: SKX 2681. Maxllla pieces: SKX 641 + 793, 368i. Mandible pieces: SKX 625, 455. IsoJated teeth and tooth pieces: SKX 15497, 15399, 15398,3123,3106,3072,3071, 3064,3010,2939,2903,2353,2351/2,1857,1605,1577,1562, 1345,1343,1149,1148,1147,1117,1082,1054,1035,958,884, 643,636,548,547,309,104. Vertebra: SKX 31 + 31a. Seapula piece: SKX 3090. Humerus prece: SKX 2351/6. Ulna pieces: SKX 1918, 763m, 726, 501, Carpal bone: SKX 15564. Pelvis pieces: SKX 1656, Tibia pieces SKX 470, Patella: SKX 15538, Tarsal bone: SKX 1446/3, Metatarsal: SKX 1308, Phalanges: SKX 46711,15956,3104,2812,2715,2652, 1948a, 1761a, 1680/11, 1304,1260,1181,1049,748,

Procavía antiqua

One hundred and thirty two cranial and 55 postcranial pieces 1rom a rninirnum 01ten individuals. MATERIAL. Maxilla pieces: SKX 60, 137, 659, 841, 842, 933,939,1180,2570, 3021a, 287, 1179, 1683, 1894,2506, 15609,1893 + 1894, 190517, 2362, 2531,2681/6, Mandible pieces: SKX 15890, 15325,2658,1895,3915,3031,2770, 2769,2771, 2799,54,55, 660,706a,285,273,665, 1573,658, 1643,745,717, 706b, 12439, 12386, 2854, 775, 661, 288, 2852,2657,718, 711c, 2250, 979, 838,159,602,56,57,58, 149, 59, 657, 935, 936, 2493, 2639, 2906, 15849, 15942, Isolated upper incísor pieces: SKX 78, 400, 603, 656, 706p, 776,869,934,971, 1187, 1624, 1722, 1769/3, 1845d, 1993/1, 2660, 2731, 2784, 2899, 2958, 2959, 3017, 3525, 3526, 15290, 15392, 15469, 15503, 15593, 15728, 15761, 15846, 15945, 15948, 18779, 18804, 19016, 19023, Isoiated lower incisors: SKX 649, 650, 739, 740, 937, 1178, 2482, 2483, 2666/4, 15326, 15768, 17970, Isolated upper molars and premolars: SKX 2904, 2719,2331,2321,2312, 1905/11,1111, 654, 454, lsolated lower molars and prernolars: SKX 652, 15375, Vertebrae: SKX 723, 725. Scapula pieces: SKX 1680/9, 999/6, 702, 693, 685, Humerus pieces: SKX 255b, 690,700,727,756,1183,1668,2675/5,2873,15490,15717, 15739, 18774, 19020, Radlus pieces: SKX 682, 710b, 780, 1703, 15376, 15718, Ulna pieces: SKX 2500/4, 1369, 722, 710c, Pelvis pieces: SKX 1702, 1332b, 760a, 692, 678, 532p. Femur pieces: SKX 63, 131, 542, 931, 932, 2500/6, 2549/6, 15174, Tibia piece: SKX 681. Calcanea: SKX 15643, 15605, 1790/2, Astragalus: SKX 1472, Metapodials: SKX 456, 736, 737,751,3132,

Order PERISSODACTYLA Family EQUIDAE

Carnívora indet.: coproñte pleces

MATERIAL, Eight coprolites: SKX 706, 720,1094,15613, 15744,15746,15773,15841,

Order TUBULlDENTATA Family ORYCTEROPODIDAE Orycteropus afer Three phalanges possibly from the sama individual.

Hípparíon Iybícum steytlerí One cranial and eight postcranialpieces from a minimum 01 one individual. MATERIAL. Isolated tooth piece: SKX 1706 + 1708, Metacarpal piece: SKX 3206, Tarsal bone piece: SKX 15822, Melatarsal piece: SKX 453, 2265. Phalanx pieces: SKX 1059, 1058, 2626a+b, 3149,

MATERIAL. SKX 75, 498, 1199.

Order HYRACOIDEA Procevie transvaaJensís Twenty cranlal and 14 postcranial piecas from a mínimum cr three tndlviduals. MATERIAL. Maxilla pieces: SKX 136,597,799, Mandible pieces: SKX 714, 715, 1376, 15304, Isolated upper incisors: SKX 716, Isolated lower incisors: SKX 647, 651, 738,15685, Isolated upper molars and prernolars: SKX 655, 1373, 1665, 3170. Isoiated lower molars and premolars: SKX 653, 719, 15303, Molar fragmenl: SKX 664, Humerus preces: SKX 686, 699, 2867, Radius preces: SKX 684, 1653, 15205, 15206, Pelvis piece: SKX 675, Femur pieces: SKX 731, 732, Tibia pieces: SKX 676 + 691,679, 1654, Metapodial: SKX 733,

Equus cepensis Three cranial and seven postcranial pieces from a minimum 01twe juveniles and ene adulto MATERIAL, Isolated deciduous premolar: SKX 667 + 668, Iso/ated permanent teeth: SKX 892, 666 + 669, Carpal borre: SKX 15784, Tibia piece: SKX 2390, Tarsal bones: SKX 3466 + 3467 + 15650, Calcaneus piece: SKX 3415, Astragalus piece: SKX 3837 + 15840 + 15998, Metapodial pieces: SKX 394,2392,

Equid indet, Three postcranial pieces. MATERIAL Metacarpal piece: SKX 2549/2 Astragalus piece: SKX 995 + 996,

COMPOSITION OF 80NE ACCUMULATIONS

Order ARTIODACTYLA Family SUIDAE Phecocnoerus sp. Nine cranial and eight postcranial pieces from a mínimum 01 two juveniles and cne adult, MATERIAL. Isolated tooth fragments: SKX 217,626,627, 950,1441, 1823b, 2893, 2912,15618. Humerus piece: SKX 2835. Phalanx pieces: SKX 122, 302, 913, 1198, 1247,2505, 3140.

cf. Tapinochoerus meadowsi One unerupted molar fraqment from a juvenile.

49

1351, 804, 476, 415, 284, 953 + 954, 803, 801. Lower molar fragmenls: SKX 2630,1560,829. Horncore pieces: SKX 3200, 960, 768, 767. Alcelaphine indet. Thirty four cranía! pieces. MATERIAL. lsolated deciduous premolar: SKX 598. Maxilla pieee: SKX 1908. Mandible piece: SKX 467. Isolaled upper perrnanent teeth: SKX 830, 2618, 2484. Tooth fragmenls: SKX 2571,2964,3012,3172, 15301,2963,2831,2402,2284, 2283,2278,1352,1216,947,805,754,1010,988,414,413. Horneore pieces: SKX 2895,1423,1126,316,314,212,167, 115.

MATERIAL SKX 1926 + 1927.

Family GIRAFFIDAE

Sublamily Anlilopini

Giraffid gen. & sp. índet,

Antidorcas australis and/or marsupialis Forty three craniaí pieces from a minimum ofthree juveniles and seven adults.

One cranial and cne postcranial piece from a mínimum 01 cne individual.

MATERIAL. Isolated inclsor piece: SKX 1676. Carpal SKX 3226 + 3276.

Family BOVIDAE Sub/amily Alcelaphinae Mega/otragu5 sp. seven cranial pieces from a minimum 01 one individual.

MATERIAL Isolaled upper lhird molars: SKX 881,1326 Molar fragments: 868, 1029, 1243, 1349, 1606.

MATERIAL. Mandible pieces: SKX 2396 + 2397. 2847, 2853, 2725, 500. Isolaled deciduous teetn: SKX t 350, 1030, 2529,1050. lsolated upper permanent leelh: SKX 1055,2526, 12391,2495,283, 12390, 802, 2528, 2901,806,808. lsolated lower permanent teeth: SKX 477, 2487, 1476,2677, 465, 2736,2527,2735,198,909,908,466,3134,907,895 Toolh fragmenls: SKX 1050, 2347, 1896, 2994, 2908. Horncore pieces: SKX 222,106918,3092. Antidorcas recki Nine cranial pieces from a minimum ot one juvenile and one

Connochaetes sp. Fifty six eranial pieces from a mínimum of two juveniles and five adults. MATERIAL. Isolated deciduous pramotars: SKX 3013, 2965,2348,1151. Isolaled upper permanentteeth: SKX 4021, 3730,3729,3704,3250,2830,2738,2620,1491,1263,1264, 1265,1266,1157,1075.893,848,670,671,672,673,607, 357,2285,2829,3360,3907,3018/3,3018/4,3018/5,3018/6, 301817, 107. Upper molar and premolarfragmenls: SKX 2646, 2523, 1442. lsolated lower permanent teeth: SKX 4034, 3264, 2856,2518,2519,2520,2286,1469 + 1470,843,800,721, 452,183. Lower molar and premolar traqments: SKX 3011, 2401. Horncore fragmenl: SKX 98.

adult. MATERIAL Mandible prece: SKX 2721. Isolaled deciduous tooth: SKX 952. Isolaled upper permanenl teeth: SKX 2492, 2488,910,906, 811.lsolated lowerpermanentteeth: SKX 369, 370.

Medium-sized alcelaphlne: Alcelaphus sp. or Beatragus sp. Sixteen cranla! pieces trom a minimum 01 three individuals MATERIAL. Mandible piece: SKX 17336. lsolated upper permanent molars and premolars: SKX 3446, 3302, 2254, t 360 + 1361,281,280, 106. Isolaled lower permanenl molars and premolars: SKX 2404, 1707, 1697, 1230. Horncore pieces: SKX 3745 + 3746 + 3747, 2568, 1484.

MATERIAL Maxilla piece: SKX 3726. ísolated lower molar: SKX 3255.

Damaliscus sp. Thirty cranial píeces from a minimum ofone juvenile and five

adults. MATERIAL. Isolaled deciduous tooth: SKX 3731. Maxille piece: SKX 2717. lsolated upper permanenl teeth: SKX 2544, 948, 946, 506, 3301. Isolated lower permenent teeth: SKX 15243,2801,2727,2633,2281 + 2282,1558,1557,1375,

?Gazella sp. Eleven cranial pieces from a mínimum of five individuals. MATERIAL. Maxilla pieces: two unnumbered maxilla from the same individual. Horneare pieces: SKX 160, 196, 945, 1176,1278,1279,2360,2405,2472,3574,3836. Raphicerus campes tris Two cranial pieces fram a minimum of one individual.

Subfamily Bovinae Syncerus sp. Seven cranial pieces from one adult.

a minimum of one juvenile and

MATERIAL. Isolaled deeiduous tootn: SKX 282. Iso/ated leeth: SKX 827, 828, 831,834, 835, 1340. Taurotragus oryx Two cranial pieces from a minimum of ene individual.

MATERIAL. Isolated permanent leelh: SKX 521,4026.

COMPOSITION OF SONE ACCUMULATIONS

50

Sublamily Hippotraginae

........ 15727,15912. Radius pieees: SKX 689, 1845e, 1930/5, 3727 .

Hippotragus el. niger

MATERIAL. Molar fragment: SKX 1575.

Hippotragini/Reduncini indet. Four cranial pieces from a minimum 01 cne individual. MATERIAL.lsolated teeth: SKX 1110, 1347, 1462,3650.

Ulna pieees: SKX 680, 1481/1. Pelvis pieees: SKX 65e, 1636/2, 15567, 231. Femur pieees: SKX 65g, 683, 706i, 2292/8, 2681/2, 2868, 3124, 3916, 15606, 15541, 1097, 2410/11, 15943, 1535, 1513/3, 1501,2806, 19010, 3732, 15541. Tibia pieees: SKX 701, 763a, 2351/4, 2949, 3308, 12430,15560. Caleaneus: SKX 9001. Metapodials: SKX 1461, 1371, 697. Phalanx: SKX 750, 377. Patelia: SKX 15620. CarpalsfTarsals: SKX 763e, 711d, 5321.

Bovid síze class lindel.

Fourteen cranial and 42 postcranial pieces from a minimum 01one juvenile and one adult (Table 4e). Bovid síze class 11 indet, Eighteencranialand 312 postcranial pieces froma minimum 01two juveniles and six adults (Table 4a). Bovid size class 111 indet. Savan cranial and 121 postcranial pieces from a minimum 01three juveniles and three adults (Table 4b). Bovid size class IV indet. Thirteen postcraníal pieces from a minimum 01 three indi-

viduals (Table 4e). Bovid indet. cranial (all síze classes)

Enamel fragments: 1624. Horneare fragments: 350. Incisors: 28.

Order LAGOMORPHA Lagomorph gen. & sp. indel. Three cranial and 25 postcranial pieces from a minimum 01

tour individuals. MATERIAL. Mandible pieees: SKX 645, 1905/8, 2532. Seapula pieee: SKX 15798. Humerus pieees: SKX 687, 922, 1804a, 2374/3. Pelvis pieees: SKX 298, 546, 744, 873e, 1332a, 2608, 3406 + 3647, 15788. Femur pieees: SKX 279d, 824b, 2651, 12378, 1403 + 1407. Tibia pieee: SKX 3760. Calcan ea: SKX 1527,3129,3664,15533,15782.

Class AVES Bird indel. (ali size elasses) One cranial and 163 postcranial pieces from a minimum of ten individuals.

MATERIAL. BiII pieee: SKX 3717a. Vertebrae + Sacra: SKX 15787,724,72,1428. Keel pieees: SKX 3881,1418,2500/5. Seapula pieees: SKX 2374/5. Coraeoid pieees: SKX 18788b, 18785e, 18772,3991,3906,3091,2841,2840,2760,41, 132, 329b, 417h, 1273/4, 1356/9, 1446/7, 1481/6, 1611/4,2351/7, 2410/1,2458/3,45978. Humerus pieees: SKX 1288, 522, 47, 1408,2479,237,135, 388j,1383, 1611/3, 15749,3966,2836, 236, 1254/2, 1385, 1502, 18788e, 18788a, 2872, 2793, 528, 285,276,3117,112,17200,12379,15863,15796,113,2875, 2991. Ulna pieees: SKX 18806a, 15931, 4006, 3973, 2791, 706e, 1503 + 1504, 1948b. Carpometaearpus pieees: SKX 18806b, 18787, 15731, 3625, 1335, 110, 1382. Pelvis pieees: SKX 1356/5. Femur pieees: SKX 17134, 3997, 3235, 2953, 2808,2761,2689,2686,529, 148, 144, 143, 1381, 1487, 1513/6,1930/2,1930/3,1930/7,2500/10,12433. Tibiotarsus pleees: SKX 17177, 4054, 4046, 3365, 530, 523, 464, 441, 440. Tarsometatarsus pieees: SKX 18788d, 18785d, 18785b, 18785a, 15973, 15792, 15681, 15484,2810,2553, 760d,930, 1402,1543/3,2742/7. Phalanges: SKX 18771,18776,15970, 15480,3422,694, 103, 1078, 1079, 2410/12, 2545, 2713, 15964. Miseelianeous pieees: SKX 134, 601, 2374/2, 3496, 15515. Bird indet, (all síze classes). t.írnb-bone shafts: 23 in total,

Family PEDETIDAE

Pedetes sp.

Order CHELONIA Family TESTUDINIDAE

Two isolated cheekteeth from a minimum of one individual.

MATERIAL. SKX 1769/1,2624/11. Family HYSTRICIDAE Hystrix africaeaustralis Three isolated molars from a minimum of one individual.

MATERIAL. SKX 1673,2997,3685. Mammal lndet. Twelve cranial and 61 postcranial pieces.

MATERIAL. Maxilla pieees: SKX 3219, 15800. Mandible pieees: SKX 771, 926, 2750, 3014, 3148, 3727, 12382. Calvaria pieee: SKX 39. Isolated ineisors: SKX 646, 2785. Vertebra: SKX 15888, 15708, 15666, 1328,963,752,703. Seapula pieee: SKX 3141. Humerus pieees: SKX 695, 698, 991,15581,

Chelonia indet, Twenty six postcranial pieces from a minimum of two individuals.

MATERIAL. Humerus pieees: SKX 1410, 42. Limb-bone pieee: SKX 894. Pelvis pieees: SKX 125, 43. Carapaee and plastron pieees: SKX 15882, 15881,3843,3171,3683 + 3725, 3034,2948,2947,2915,2815,2511/5,2292/7,1642,1603, 1165/5,1046,1024/5,786,46,45,44. Reptile indel. Eleven postcranlal elements from a minimum of one individual.

MATERIAL. Vertebra, SKX 2351/8. Limb-bone pieees: SKX 15950, 15780, 15567a, 15535, 15528, 2952, 2871, 246ge, 1790/1,651.

COMPOSITION OF SONE ACCUMULATIONS

51

Class AMPHIBIA

Theropithecus oswaldi danieli , Four specimens trom a minimum ot one juvenile and three

Frog/toad indet.

adulls. MATERIAL. Isolated teelh: SKX 37323, 32148, 28490, 27586 + 27587.

Nine distal humerus pieces from a mínimum 01 five individuals.

MATERIAL. SKX 133, 15786, 15824, 18129, 18770, 18782, 18805,18807.

Cercopithecoid indet. One hundred and sixty nine cranial and 77 postcranial specimens.

DISCUSSION As Member 1, Member 2 ls characterized by a larga number 01 austraJopithecines coexisting with Horno sp. The bovids are representad by a variety 01 species, the minimum numbers 01 whrch indicate that they were the main source 01 prey for the

carnivores that contributed lo the bane accumufation in the cave. Of lnterest rs the presence 01 a giraffid and the giant hartebeest, Mega/otragus sp.

MATERIAL. Craniai pieces: SKX 19855, 31676, 34484. Maxilla pieces: SKX 345381, 37269. Mandlble pieces: SKX 36715,39572,26857,25881,22990,19632. Isoiated incisors: SKX 46729, 46380, 46375, 46070, 37321, 37209, 34675g, 34535,31462,39341,37824,37822,37707,37643,37506, 37465,37435,26900,26457,26452, 25057d + 25076, 22362, 26010,37373,37312,37150, 37073, 36965, 36882, 36721, 36704, 36702, 36647, 36619, 36348, 35819, 35422, 35330, 35295,35075,35065, 35024, 33238, 33124, 32222, 32215,

32064,32056/55,31882,31881,31880,31631,31370, MEMBER 3 Figure 3 lllustrates the macrovertebrate taxa found in this Member and the material recovered ;5 listad below.

Class MAMMALlA Order PRIMATES Family HOMINIDAE

AustraJopithecus robustus Twelve cranial pieces from 9-10 indlviduals. MATERIAL. Isolated teeth: SKX 19892 LM" 21841 RM 3 , 25296 Re, 26625 Lp4 , 26967 U" 27151 Ldm, 27524 U', 28724 Re, 32162 RP 4 , 32832 Rdm', 35025 RM" 35416 Rh. Hominid postcranial material Eleven fossils were recave red.

MATERIAL. Medial cuneiform: SKX 31117. Metatarsal: SKX 33380. Phalanges: SKX 19576, 22511 + 30220, 22741,27431, 27504,33355,35439,35822,36712. Papio hamadryas robinsoni

Ninety üve specimens trom a minirnum 01two juveniles and nine adults. MATERIAL. Maxilla pieces: SKX 37704, 37094, 32059 + 32060 + 32061, 30755 + 30756. Mandible pieces: SKX 33189 + 33191 + 33193-33198 + 33201 + 33203-33205, 20658 + 20659 + 20660 + 20662, 36607, 36095, 35315 + 35316. Isolated deciduous teelh: SKX 30523, 19595 + 19596, 32046, 37706. Isolated upper teeth: SKX 27023 + 28998 + 27802 + 27805 + 27808 + 27798, 34601, 34549, 25171, 25172,46069,37467 + 37172, 36904, 35791,30546,27041, 25472,35754,27797,28703-28710 + 28714-28720, 37956, 33397,33026,37507,35287,30333,31690,30452,37705, 33578, 33277, 33542, 33286, 29175, 48339. Isolated lower teelh: SKX 35423, 33757, 31662, 27664, 36703, 35817, 35453, 35430, 30178, 19902, 33752, 37259, 35953, 33756, 27367, 33723, 27430, 32022, 37647, 36896, 36663, 31574, 33759, 33765, 33755, 37708, 37275, 33761, 33425, 37319, 36280, 33341, 37318, 31798, 22663. 33758, 33760, 34552, 46163,22174,35236.

30915,30547,30490,30093,26863,29214,27965,22792a, 27481,20085,19942,26817, 26072.lsolated canines: 46834, 46662,29274,21539,20130,37823,35322,30526,35324. Isolated teeth and tooth Iragments: SKX 37713, 37711, 46702a+b, 46377a+b, 46277, 45999, 45932a+b, 25098, 16414,37663,37320,37315,37071,37058,36877,36755, 39344, 39435, 39453, 39542, 39543, 39600, 36754, 36753, 36743, 36648, 36315, 36267, 36104, 35753, 35863, 35742, 35719,35666 + 35667, 35424, 35417, 35386, 35385, 35381, 35328,35076,33701,33700,33699,33612 + 33613, 33595, 33562, 33432, 33354, 33237, 33006, 32960, 32958, 32893, 32762,32161,32147,32073,31651,31532,31517,31408+ 31409,30454,30351,30125,30027,30026,30025,30024, 30023,29879,29276,29175,27963,25529,22512, 19758, 19693,19655 + 19659, 25481, 32146, 28998, 36880. Axis vertebra: SKX 46098. Scapula piece: SKX 26739. Humerus pieces: 26262, 30536, 33044, 35163, 20117, 35964, 37838. Radius pieces: SKX 31815 + 31817, 36645, 33939, 31957, 30701. Ulna pleces: SKX 26450A, 31822, 19447 + 19453 + 19469,26435,28521 +28522,33186,31456,35705. Carpal bones: 27023, 48378. Pelvis piece: SKX 36089. Femurpieces: SKX 28941,32217 + 32218. Tibia pieces: SKX 22678, 34872, 46175, 35738. Metapodials: 26456, 33151, 46738, 23400, 35192,35959,37026,31372,31870,36669. Phalanges: SKX 33823,33842,35755,36683,37389,37390,37177,29161, 31616,35352,37391,21832,23401,25286,25482,25978, 26623,26709,27238,27239,27240,27675,28373,29161, 29376,20165,30110,30348,31048,31125,31391,33242, 33245,26526. Order CARNIVORA Family FELlDAE Sublamily Felinae Panthera pardus Thirteen cranial and 40 postcranial pieces trom a minimum of one juvenile and tour adulta.

MATERIAL. Skull piece: SKX 34693. Maxiíla pleces: SKX 34683,32203,26759 + 26760 + 26761. Isolated teeth: SKX 21979,27574,31654 + 31655, 32467, 35425 + 35426, 23635, 28098,34193,22742. Vertebrae: SKX 28870, 29651 + 29652,

52

COMPOSITlON OF BONE ACCUMULATIONS

32973, 37679. Humerus piece: SKX 38418, Radius preces: SKX 37364, 36644, 35801, 38337, 3B220. Ulna pieces: SKX 36825,30054,28067. Carpal bone: SKX 22922. Femur piece: SKX 39114. Tibia pieces: SKX 37552,32559. Calcanea: SKX 37594, 48093. Astiagalus: SKX 38424. Tarsals: SKX 36955, 35313, 35234. Metapodials: SKX 38145, 39038, 39112, 39854, 36484 + 26483, 33693, 33893, 34553, 35238, 35345, 36947,36407,37477,32419. Phalanges: SKX 38578, 37029, 32397.

Acinonyx juba tus MATERIAL. Canine Iragment: SKX 36273. Felis cf. Iybica MATERIAL. A left mandible piece with P,and M, In situ, SKX 39460, and a right proximal lemur and shatt, SKX 3941 O, nave been assigned to this species.

Subfamily Machairodontinae Megantereon cultridens MATERIAL. The tip 01 an upper canine: SKX 45391. An extremely large lelid right distal tibia fragment, SKX 39460, has been provisicnalíy assigned to this genus.

Family HYAENIDAE Subfamily Hyaeninae Hyaena brunnea Twelve craniat and three postcrantal pieces trom a minimum 01 three individuals. MATERIAL. Maxilla pieces: SKX 30820 + 30822 + 30824, Mandibie pieces: SKX 34159 + 30057. Isolated teeth: SKX 20182,20176 + 20180 +20181,30401,30831,31535,31858, 31969, 35418, 37303, 37433 + 38256c. Calcaneus: SKX 37178. Metatarsal: SKX 26268. Phalanx: SKX 36942, Chasmaporthetes nitidula Four ísolated teeth trom a mínimum ot two individuals. MATERIAL. SKX 21615, 22972 + 22992, 25755, 29205.

34887,33350.32397,29163,29031,28711,28181,27291, 27170, 22792.

Subfamily Protelinae ProteJes sp. MATERIAL. A right mandible piece: SKX 39109. Family CANIDAE VuJpes cneme Two cranial and 15 postcranial pieces from a mínimum of two individuals. MATERIAL. Mandible piece: SKX 36547 + 36620. Isolated tooth: SKX 19697. Scapula pieces: SKX 23006, 36211. Humerus pieces: SKX 36093, 36123, 36127, Radius pieces: SKX 31820, 35735. Ulna pieces: SKX 36300, 36335 Calcanea: SKX 20113, 31749. Astragalus: SKX 29670. Metapodial: SKX 33968. Phalanges: SKX 25520, 30273. Canis ct. mesometes Fifty ñve craruat ano 40 postcranial pieces from a mínimum

of five individuals. MATERIAL. sxuu piece: SKX 38340. Maxilla pieces: SKX 25973. 29986 + 29987 + 29988, 30057, 30825-30830 + 30832 + 30833, 36349, 45956. Mandible pieces: SKX 38049, 22503,21513-21515,24254 + 34255. lsolated teeth: SKX 38092,38135,38261,39277,19503,22385,22453,22514, 23060,26989,26991,27371,28329,28800-28801,28814, 29038, 29965, 30060, 30138, 30727, 30880, 31653, 31839, 31919,32602,32890,32892,32894,33445 + 33446, 33662, 34661,34822,34988,35770,36080,36352,37056,37434, 37642,37811,37815, 38392,46216,46389,Vertebrae:SKX 19556.22975,31851,35792,36124,37168. Humerus pieces: SKX 37751, 33412, 40151. Radius piece: SKX 21540. Ulna pieces: SKX 46724, 30498, 32140. 30483, 29535, 27955, 22448,19531,38147. Tibia piece: SKX 38152. Calcanea: SKX 20128,40266, 25450. Tarsal bone: SKX 27735. Metapodiais: SKX 38506, 38567 + 38571, 22964, 26193, 26912,27354 + 27447,27979,29490,30028,30300, 33689,34496d,34776, 35164,36429. Phalanx: SKX 39214,

Crocuta crocuta One cranial and two postcranial pieces from a minimum of cne individual.

MATERIAL. Tooth Iragment: SKX 38256a. Metapodials: SKX 35528, 37235. Hyaenid indet. Thirty five cranial and 30 postcranial pieces.

MATERIAL. Mandible pieces: SKX 21544 + 21545, 25447, 35073. Isolated tooth fragments: SKX 40043, 40029, 39974, 38574, 38400, 38399, 38335, 38334, 38093, 37994, 46264, 37314,36249,35325,35273,336888,33644,31117,31111, 30434 + 30575, 29750, 26618, 26190, 25906, 25780, 25759, 25445,24712,23063,21548,21547,19387. Radius pieces: SKX 20056, 37591. Ulna pieces: SKX 20011,33310,35848, 36842, 36657, 37148, 37376, Pelvis pieces: SKX 33739, 36898, 46050a-c. Tibia piece: SKX 36835. Calcaneus: SKX 39116. Tarsal bone: SKX 33400. Metapodials: SKX 35394, 22358 Phalanges: SKX 48500. 45997, 37652, 37067, 35977,

Large canid gen. & sp. indet. Four cranial and ten postcranial pieces from a minimum of one individual.

MATERIAL. Maxilla pieces: SKX 26258, 30108, 37455, Mandible piece: SKX 35314. Humerus piece: SKX 29958. Carpal bone: SKX 28659. Pelvis plece: SKX 31465. Tibia pieces: SKX 37548, 30763. Metapodial: SKX 46064, Phalanges: SKX 23059,19502,24881,27480, Canid índet.

MATERIAL. Proximal metacarpal: SKX 37843.

Family MUSTELlDAE Aonyx capensis Four cranral and one postcranial piece from a minimum ot two individuals.

54

COMPOSITION OF SONE ACCUMULATIONS

MATERIAL. Mandible: SKX 30361. Isolated teeth: SKX 35508,35790, 35954. Phalanx: SKX 39356. Family VIVERRIDAE Sublamily Herpestinae

Atilax sp. Three postcrarual pieces from a minimum 01one individual. MATERIAL. Humerus pieees: SKX 46015, 46020. Astragalus: SKX 20012.

Gynictis peniciltete Four cranial pieces from a minimum 01one individual. MATERIAL. Cranial pieee: SKX 48477. Right mandible pieee: SKX 36332. Maxilla pieees, SKX 48467, 48529.

Suricata suricatta Six cranial pieces from a minimum 01five individuals.

28834, 35237, 35251, 40233. Femur pieces: SKX 30053. 30469, 32955. 37136, 37299, 38439. Patella: SKX 30774, 27094, 45769. Tibia pieees: SKX 26445, 32696, 31628. 35498,40554. Fibula pieees: SKX 46116. 39732. Calcanea: SKX 25883, 32082, 37749. Astragalus: SKX 26457. Tarsal bones: SKX 38087, 40519. Metapodial pieees: SKX 39814. 39406,39407,39408,37999,36733,29086,36903,34561, 31987, 31872, 31601, 25293, 22966, 22634, 32279. Phalanges: SKX 38645, 38635, 38398 + 38393, 35094. 26141.22613,20064,46733,46108,39659,37377,36873, 36869,35991,35136,33696,33263,32093,31773,30646, 27783, 26447, 26096, 25582, 25569, 25264, 23340, 22647, 37759,37676, 37596, 35749, 35402, 35168, 33402, 31514, 22269. Drder PHOLlOOTA Family MANIOAE

MATERIAL. Mandible pieee: SKX 392S2, 35135, 26913, 37483,36961,27001.

Manis sp. MATERIAL. A third phalanx Irom a lronl 11mb: SKX 38192/2.

Mongoose índet, Three cranial and níne postcranial pieces. MATERIAL. Mandible pieees: SKX 19405, 27140, 23763. Humerus pieee: SKX 46891. Radius pieee: SKX 46263. Pelvis pieees: SKX 46549, 348801. Femur pieees: SKX 27264, 28934. Tibia pieees: SKX 45927, 46091. Caleaneus: SKX 46051.

Order TUBULlDENTATA Family ORYCTEROPODIDAE

Orycteropus afer MATERIAL. A third phalanx: SKX 37832. Order PROBOSCIDEA Family ELEPHANTIDAE

Subfamily Viverrinae cf. Genetta tigrina MATERIAL. Distal radius: SKX 35715. Viverrid índet. Three cranial pieces. MATERIAL. Maxilla pieee: SKX 32670. Mandible pieees: SKX 28065, 31561. Carnívore indet, Eighty eight eranial and 104 posteranial pieees. MATERIAL. Maxilla pieees: SKX 38188, 35855, 35835. Mandible pieees: SKX 39812, 34593, 31585, 27988. Isolated teeth and tooth Iragments: SKX 40144, 39270, 38552, 38309b, 38213, 38149, 37993, 46821a+b, 46062, 45996, 37714,37313 + 37325, 37307, 36963, 36933, 36875, 35717, 35382,35380,35321, 33688A,33672,33665,33582,31746, 30453 + 31968 + 32185, 29116, 28898, 28871,27323,27182, 26687,26310,26193,26188,26173 + 26181 + 26191, 25972, 25781,25749,25748,25173,22452,22008,37812,37350, 39274,38596,37304,37288,33891,33537,32468,32437, 31635, 30716, 30550, 29459, 27520, 26988, 26046, 25758, 25756,25754,25753,25752,25751,25750,23355,22968, 22559,22217,22170,22049,19470,19457,34853,33179, 28276, 26047, 21842, 20121. Vertebrae: SKX 22472, 20077 Seapula pieees: SKX 36250, 27227. Humerus pieces: SKX 34518, 38289, 39173. Radius pieees: SKX 19450, 21614, 30398,36305,37833, 46301a+b. Ulna pieces: SKX 27897, 34803, 46372, 27984, 36247, 46084, 39773, 36657. Carpal bones: SKX 35703, 35969, 46211, 40131. Pelvis pieees: SKX

el. Elephas sp. Two cranial pieces from a minimum 01 one juvenile individual. MATERIAL: An unerupted molar fragment: SKX 32051. Erupting dp' fragment: SKX 36082. Order HYRACOIDEA Family PROCAVIIDAE

Procavía transvaalensis Thirteen cranial and 18 postcranial pieces from a minimum of five individuals. MATERIAL. Upper ineisor pieces: SKX 38141, 22007, 22302 + 22303, 27502 + 27514, 35606. 36245, 36475, 37432. Lower ineisors: SKX 38572, 40180, 27515, 37036, 37149. Humerus pieces: SKX 35643, 36390, 36498, 37807, 37445, 38616,38242,39117,25595 + 25602 + 25603, 34572, 36503, 39787. Radius pieces: SKX 34562, 34675h. Ulna pieees: SKX 34517,30059,38048. Metapodial: SKX 37032. Procavia antiqua One hundred and forty four cranial and 74 postcranial pieces from a mínimum of 11 individuals. MATERIAL. Maxilla pieees: SKX 35570, 25761, 27539, 25760, 20063, 19429, 29900, 33011, 361 SO, 36706, 36950. 38009.27964.22557,33429, 34659a, 20105, 36756, 37826. 33900.35317,35067 + 35068. 30812. 34314, 27109,25979, 25969. Mandible pieees: SKX 32451. 39437, 36478. 34261. 25762.31434.39900,39188 + 39211,29781,27784,20018. 36649, 29145 + 29146, 30685, 26606, 36355, 22506, 22923 +

COMPOSITION OF 80NE ACCUMULATIONS

22924, 19656 + 19657,37667,36031, 31437, 25254, 36854, 37176,39865,36281,31977,38240,39800,39280,39231, 36353,37436, 35431,30743,23131,26604 + 26605, 30868, 24898,46883,35337,34538g, 33512,21825, 19745, 26287, 30770, 22510, 38236, 38098. Upper incisor preces: SKX 46014,38351,34214,33621,33481,32875 + 32874, 33367, 31339, 30117, 27368, 26793, 22560, 22616, 22288, 22218, 21928,40064,38212,39792,39883,40220,40231,24947, 26548,27369,31594,31595,31617,31896,31897,34989, 36476, 36876, 37589, 46038, 46752, 25425. Lower incisors: SKX 47996, 39225, 39790, 39791, 22724, 25371, 26023, 26280, 27215, 28557, 30879, 3152_, 33366, 37814, 46058, 46081. isolated teeth: SKX 22313, 22795, 24976, 28808, 32869,33168, 33287, 35569, 36250, 37567, 37568, 46600, 21812. Axis vertebra: SKX 30405. Scapula pieces: SKX 32604, 32560, 39133, 39009. Humerus pieces: SKX 48221, 39553,38490,39110.39833,22918,24971,25055,25604, 26622, 26683, 30704, 32495, 32573, 36086, 32698, 36263, 36502, 38057. Radius pieces: SKX 40193, 39353, 39454, 27065, 32876, 34918, 35078, 35777, 36425, 26538. Ulna pieces: SKX 48229, 39411, 38305, 39217, 39772, 26434, 28149,31956,32847,35061,35393,35642,46107. Pelvis pieces: SKX 46790, 32002, 38146. Femur pieces: SKX 39409, 38097,40219,21425,28037,28243,32464,33561,34215, 37827, 37382, 30157. Tibia pieces: SKX 28209, 25586, 25549,38505,36212. Calcanea: SKX 46197. 37095, 35338, 19610,19558. Metapodial: SKX 21510.

Order PERISSODACTYLA Family EQUIDAE

Equid indel. Ten craníal and 11 postcranial plecas. MATERIAL. Isoiated tooth fragments: SKX 46528, 46150, 25475, 19456,34326,33872,32976,37213,26353,37368. Radius pieces: SKX 29940 + 29960. Ulna piece: SKX 30852. Tibia piece: SKX 35007. Metapodial pieces: SKX 26745, 29524, 28238, 24952, 46908, 34676. Phaianx pieces: SKX 24951, 46722. Order ARTIODACTYLA Family SUIDAE Phacochoerus sp. Ten cranial and two postcranial pieces from a mínimum 01 one subadult. MATERIAL. Isolated tooth fragments: SKX 38244, 37808, 32035,30769,33221,35955,36602,37302,37310,37590. Calcaneus: SKX 38164. Astragaius: SKX 38496.

et. Tapinochoerus meadowsi Six crantal pieces from a rninimum 01one individual. MATERIAL. Isolated tooth fragments: SKX 19446,26214, 31528,33117,33464,37638. Suid indet. Eleven cranial and 12 postcranial pieces. MATERIAL. Maxilla piece: SKX 36716. Mandible piece: SKX 33376. Isoiated tooth tragments: SKX 29808, 22636, 22852, 22865, 33309, 35546, 35860, 46212, 46578. Axis vertebra: SKX 31742. Scapula piece: SKX 27545171. Femur piece: SKX 38618. Metapodials: SKX 27085, 35621,37137. Third phalanges: SKX 33447, 35455, 32481,32846,37637,46525.

Htpperion Iybicum steytJeri One cranial and 12 postcranta! pieces from a minimum 01

Family HIPPOPOTAMIDAE

one adult. MATERIAL. Mandible: SKX 22747-22750 + 29171 + 29388. This mandible has been modified into a tool. Radius pieces: SKX 22320c, 21971. Metapodial pieces: SKX 20024 + 20154, 28440,28975 + 28976, 29023 + 29027 + 29044, 3961 8, 22086 + 22126, 22936. Phalanx pieces: SKX 26812, 30180, 39183.

Hippopotamus sp. MATERIAL. Proximal first phalanx: SKX 29836.

Equus capensis Thirteen cranial and 23 postcrarua! pieces from a mínimum 01 two juveniles and three adults. MATERIAL. isolated deciduous teeth: SKX 28881,35429, 35156 + 35157.lsolated permanentteelh: SKX 37356, 36437, 36416 + 36417 + 36418, 31932, 27290 + 27285, 27230, 26048, 19802, 19658, 19616. Femur piece: SKX 35761. Ulna piece: SKX 48406. Tibia piece: SKX 30256. Calcaneus pieces: SKX 22935, 30447. Astragaius pieces: SKX 22729, 37479 + 37480, 37551, 38058. Tarsal bone: SKX 36273. Metapodial pieces: SKX 19563, 19806, 19808, 20099, 29344 + 29356, 33202. 34477, 39181, 39179 + 39180. Phalanges: SKX 39182,37548,39815,28313.

55

Family GIRAFFIDAE

Giraffid gen. & sp. indet. Two postcranial pieces from a minimum of one individual. MATERIAL. Calcaneus piece: SKX 22651. First phalanx piece: SKX 28313. Family BOVIDAE Subfamily Alcelaphinae Megalotragus sp.

Seven cranial pieces from a minimum ot ene juvenile and two adults. MATERIAL. Isolated upper teeth: SKX 28872, 27934, 27800,26229. Isoiated lower teeth: SKX 27823,19910. tsolated tooth fragment: SKX 27935. Connochaetes sp.

Equus burchellii

Four postcranial pieces from a minimum 01one adult. MATERIAL. Astragalus piece: SKX 22815 + 22824. Meta)odiai pieces: SKX 25704, 28495. Phalanx piece: SKX 39617.

One hundredand ninety nine cranial pieces from a minimum of tour juveniles and 13 adults.. MATERIAL. isolated deciduous teeth: SKX 37717, 37715, 36333,36081,34925,34891,34168,33240,32596,32125.

-------', - - 56

COMPOSITION OF SONE ACCUMULATIONS

31440,30488,29860,29540,29056,27801 + 27803. Isolatétt upper permanent teeth: SKX 38592, 39108, 39718, 38593, 39709,37639,37215,37055, 36720, 35820, 35428, 35420, 35318, 34586, 34492, 34490, 34292, 33558, 33284, 32886, 32597,32555,32553,32552,32160,31860 + 31906, 31706, 31611,31529,31065 + 31067, 30874, 30524, 29911, 29769, 29770,29768,29091,28175,27483,27275,27015,26997, 26049,24861,24502,23240,22814,22625,22285 + 22286, 22280 + 22284, 22252, 21526, 19696. Isolated upper molar fragments: SKX 38267, 46778a-d, 36543, 35720, 35509, 35465,35421,34821,33525,32640,31744,31589,31588, 31575 + 31576, 31460, 27894, 25872, 19691.lsolated lower permanent teeth: SKX 39540, 39601,38137,39208,38270, 39071, 39102, 39872, 38436, 39283, 40216, 40230, 37946, 37649,37608,37563, 37561,37289,37187a, 37102,36248, 35753,35748,35041,35040,35037,34973,34926,34293, 34291, 33506, 32888 + 32889, 32790, 32588, 32494, 32158 + 32159, 32124, 31660a, 30456, 30211,29807,29805,29440, 29325,29323,29279,28832,28803,28655 + 28656, 28653, 28027,28026,27822,27702,27701,27454,26899,26880, 26736 + 26737, 25905, 24830, 22913, 22546, 22374, 22352, 22239, 22088, 21944 + 21945 + 21946, 21807, 20048 + 20050, 19832, 19781, 19544. Isolated lower molar fragments: SKX 37710, 37090, 36438, 36346, 36057, 36052, 35042, 34806,34206,34016,33579,33328,32885,32497,31476, 30642,30412,29806,39392,28251,28099,27607 + 27609, 27084,26717,25902,25720,22626,20118. Molarfragments: SKX 37213, 31890, 29280, 28804, 28562, 22077. Horncore pleces: SKX 30121,22679,22356.

Medium-sized alcelaphine: Beatragus sp.lAlcelaphus sp. Fifty tour cranial pieces trom a minimum 01 cne juvenile and five adults. MATERIAL. Maxilla plece: SKX 391 01. Mandible piece: SKX 39451, 28055, Isolated deciduous teeth: SKX 31739, 22283, 21797. Isolated permanent upper teeth: SKX 39544, 39539, 38433, 38182, 38508, 40018, 37075, 35319, 34687, 33628, 32493,31990,30462,30124,29216,29110,25691,24831, 24717,21911. Upper molar fragments: SKX 31609, 25016, 22720. Isolated lower permanent teeth: SKX 39928, 40186, 39348,37701,36435,35177,34463,34294,34290,34211, 31771 + 31772, 30817, 29487 + 29488, 28654, 28028, 22281, 19540. Lower molar fragment: SKX 31459. Horneare pieces: SKX 36682, 30251,29922,29423 + 29528 + 29486 + 29881, 26949,26538,22355. Damaliscus sp. Eighty cranial pieces from a mínimum 01 two juveniles and tour adulta. MATERIAL. Isolated deciduous premolars: SKX 39767, 37272,34682, 31661, 30964, 27184, 25528, 21948. Isolated permanent upper teeth: SKX 39708, 38140, 39834, 40083, 37054,37041,35005,34949,33580,33560,33505,32639, 32095,30684,30489,30136,30135,30061,29489,29215, 28874, 28492, 28109, 27668, 27457, 27409, 25970, 22795, 22050, 19694. Upper molar fragment: SKX 29092. Isolated permanent lower teeth: SKX 38374, 38336, 38263, 38591, 39107, 37043, 34208, 32107, 32005, 31747, 30963, 29705, 29277,28467,28274,27623,25527, 19675, 19543. Lower molar fragments: SKX 31859,31664,31461,30807,29848,

29326, 26250. Molar fragments: SKX 32879, 22994, 22980. Horncore pieces: SKX 37634, 37550, 37097, 36398, 36309, 34310,32715,32645,28768,24906,22524 + 22525 + 22526, 21496.

Alcelaphine indet. Twenty seven cranial pieces. MATERIAL. Mandible plece: SKX 35537. Molar fragments: SKX 40063, 39763, 39157, 39106, 38562, 38551, 38438, 38435,38095, 38065a, 34551. Horncore pieces: SKX 39166, 38204,37840,37502,36206,36033,34966,34172,31865, 30413,29923,25690,23033,22276,21950. Sublamily Antllopini

Antidorcas australis and/or marsupialis Ninety three cranial pieces from a mínimum01fourjuveniles and 14 adults, MATERIAL. Mandible pieces: SKX 32176, 21826 + 21834 + 21835,46727,40010. Isolated deciduous teeth: SKX 21840, 22746,27391,32704,36351,32705,30252 + 30253, 39209. lsolated upper permanent teeth: SKX 28008, 29147, 29278, 21980, 46244, 37809, 36347, 32887, 27876, 27717, 19572, 19713,35320,30334,30332,28321,20143,37716,35066, 35326,25562,36545,35327,33839,28381,35388,35384, 37102,38183,32624,32703,22129 + 22130 + 22131 + 22134 + 22135, 22287, 25040, 39103, 39611, 39719. Upper molar tragments: SKX 30304,29418,28090,27624,27402,27211, 21929, 19646. Isolated lower permanent teeth: SKX 28491, 30875,29771,26844,28999,37821,28393,39908,38594, 37198, 35038. 34250, 20101, 324249/51, 30878, 29420, 30806, 22254, 25369, 27338, 36803b. Lower molar fragments: SKX 37597, 37641,35265,34655,25367. Horncore pieces: SKX 36312, 35293, 31123 + 31124, 29655 + 29666, 28809,23631,22933,20140 + 21504, 20130, 19735 + 19766.

Antidorcas recki MATERIAL. Lower molar: SKX 37508.

? Gazelfa sp. Twenty five horneare fragments from a minlmum 01 14 individuals, MATERIAL. SKX 22177, 23174, 23300, 23301, 24813, 25980,28080,28675,28949,28691,29207 + 28208, 29206, 29322, 29539 + 29678 + 29695, 29677 + 29679, 30156 + 30216,31943,32551,33680,35856,36420,36528,29207+ 29208,36309,37140.

Oreotragusoreotragus Twa upper permanent teeth and twe terminal phalanges from a minimum of ene individual. MATERIAL. Isolated teeth: SKX 29419, 30349. Terminal phalanges: SKX 19777,27021.

Raphicerus campestris Ninecranial pieces from a minimum oftwo juveniles andoneadulto MATERIAL. Isolated deciduous teeth: SKX 37813, 30974. Isolated upper permanentteeth: SKX 38091,30092. Mandible plece: SKX 22082. Isolated lower permanent teeth: SKX 33783,32601. Horncore pieces: SKX 24829, 20119 + 21583.

57

COMPOSITlON OF BONE ACCUMULATIONS

Antilopini indet. Eight tooth fragments. MATERIAL SKX 32123,29968, 29963 + 29967, 29861, 29541,27061,27022, 19645.

subtamily Bovinae Syncerus sp. Nine cranial pieces from two juveniles and one adult. MATERIAL. Isolated deeiduous teetn: SKX 39210, 40165, 31550, 29653, 28982. Isolated upper teeth: SKX 29602, 28176. Isolated lowerteeth: SKX 36878, 29933. Taurotragus oryx Seven cranial pieces from a mínimum 01 one individual. MATERIAL lsolated teeth: SKX 19474 + 19475, 19926, 25943,27960, 32595 + 32618, 34972, 34987, 36518. Tragelaphus strepsiceras One isolated tooth and ene horneare piece from a mínimum 01 ene individual. MATERIAL Isolated premolar: SKX 22561. Horneore pieee: SKX 32175. Hippotragus cf. niger Five cranial pleces from a mínimum oftwo juveniles and one

Bovid índet. cranial (all size classes). Horneare fragments:

344; eoame! tooth fragments: 1668; incisors: 113. Order LAGOMORPHA Lagomorph gen, & sp. indel. Sixteen cranial and 56 postcranial pieces from a minimum of two juveniles ano five adults. MATERIAL, Palatine bone: SKX 24968a, Maxilla pieces: SKX 39281, 22691, 33320, Mandible pieees: SKX 38214, 39901,40183,48049,22468,27186,33199,36148,36259, 36269, 36546, Scapula preces: SKX 46885, 46387, 32684, 31665,29890, Humerus pieees: SKX 36100, 32106, 26654, 26463,22723,22737,39612,38569, Ulna pieees: SKX29037, 26195, 26003, 39882, 39075, Pelvis pieees: SKX 36827, 36402,36391,34177,27014,40205,38224,28597. Femur pieees: SKX 38192/3, 37554, 37439, 37370, 36863, 36859, 36538,35845,35714,35675,32647,29810,27446,38003, 38617. Tibia pieees: SKX 30614, 30286, 23182, Calcan ea: SKX 33243, 32765, 32213, 32110, 32109, 31633, 29449, 26309,25316,25262,21963, Astragalus: SKX 22618. Order RODENTIA Family PEDETIDAE

Pedeles sp.

a minimum ot

subadult.

Two cranial and one postcranial piece from one individual.

MATERIAL. Isolated deeiduous teeth: SKX 34892, 37640. Isolated permanent teeth: SKX 37820, 37197, 37042.

MATERIAL. Isolated teeth: SKX 19836, 35779, Pelvis piece: SKX 24805.

HippotraginilReduncini indet. Five orante! preces, MATERIAL Isolated tooth pieces: SKX 24762, 28322, 36338, Horneore pieees: SKX 29857, 25090,

Family HYSTRICIDAE

Hystrix africaeaustralís Three cranial and two postcranial pieces from

a minimum

of

one juvenile and one adutt. Subfamily Pelinae Pelea sp. Six cranial pieces from a mínimum 01 two individuals. MATERIAL, Mandible pieee: SKX 38590. Isolated permanent teeth: SKX 39988, 36722, 36477. Horneore pieces: SKX 36781,20049 + 20052. Bovid size class I indet. Twenty tour cranial and 122 postcranial pieces from a mini-

mum 01four individuals (Table 4e), Bovid size class I1 indet. Eighty three cranial and 648 postcranial pieces from a minimum ot 14 individual s {Table 4a). Bovid size class 111 indet. Sixty seven cranial and 453 postcranial pieces from a minimum 01 13 individuals (Table 4b). Bovid size class IV indet, Nine cranial and 48 postcranial pieces from a minimum of three individuals (Tabte 4c).

MATERIAL Isolated teeth: SKX 29532, 40197, 45953, Ulna pieee: SKX 23168. Metatarsal pieee: SKX 38507, Mammal indet, Eight cranial and 107 postcranial pieces.

MATERIAL, Maxilla pieees: SKX 46394, 32708, 30858, 27817,26763, Mandible pieces: SKX 46899, 46312, 46208, Vertebrae: SKX 46400, 45792, 37172, 36296, 35857, 35270, 31818,31713,31638,31559,31558,31557,31554-31556, 20163. Scapuia pieees: SKX 22099, 22527, 34223, 35776, 35859, 37366, 38290, 39554 Humerus preces: SKX 22464, 28580. 33454, 31670, 22466, 29034, 32506, 38484, 39832, 40191, 48466, Radius pieees: SKX 19844, 28277, 30682, 32216, 36789, 47975, Ulna pieees: SKX 34471 e, 46146, 46198,40008,39862, Pelvis pieces: SKX 19441, 29135, 23051,32180,36853,36900,46732,24960a,24978,34762g, 46463, 46805, 39923, 38209, Femur pieees: SKX 19408, 34320,25466,26562, 26575, 26704, 26895, 29080, 31489. 33303, 33381, 38003, 34494, 19882, 22555, 22633, 24948, 26629,27756,28174,46573,46850,38489,39864,32611, 28910, 33568, 30525, Patellae: SKX 28571, 48537, Tibia pieces: SKX 22155, 32405, 20155, 34496e, 35773, 46145, 46632, 46691, 38629, 40068, 39649. Carpals/tarsals: SKX 27659 Astragalus: SKX 39857. Caleanea: SKX 22873, 33545,

58

COMPOSITION OF SONE ACCUMULATlONS

36329, 37749. Metapodial: SKX 22173. Phalanges: SKX 34624,25265,22171. Class AVES Ibis Two lower bilf fragments from a minimum 01 ene individual. MATERIAL. SKX 38064/5, 3891 6/2.

Francolin indet. Nine tarsometatarsus bones with spurs from a minimum 01 five Indlviduals. MATERIAL. SKX 39588, 37691, 37469,37271, 35641, 35405,26230,31938,25317.

35347, 35006, 33818, 33619, 33528, 33470, 33152, 33142, 31999, 31996, 31630, 31562, 30613, 30553, 29042. 28556, 26638,25277,25038,22363,22253,22217.21905.21509+ 21512,19703,19501. Tarsometatarsus pieces: SKX 48450, 46133,46099,39352,38533/1,37752,37037,36779.36632, 35620,33228,31935,31871,31396,27556,27478,27063, 26987, 26986, 26684, 26569. 26427, 26071, 25250, 25089, 25080, 25048b, 25037,24957,23061 ,22620,22222d, 21646, 20175,20106,37691. Phalanges: SKX 31666.30185,27910, 26831,26576,22015.21625. Limb-bone pieces: SKX 46203, 46435. 38308/8.

8ird síze class 111 One cranial and 48 postcranial pieces from a mínimum of

five incivicuals. Bird size class I Forty postcranial pieces froma mínimum of eightindividuals. MATERIAL. Coracoid pieces: SKX 33127, 32439. Sacra: SKX 34686, 31569a, 31564. Humerus pieces: SKX 39552, 39405,35896,35549,35502,34771,33129,33128,29114, 28599,27781, 27584, 26567, 25843, 25370, 25302, 25107, 22085. Ulna preces: SKX 46629,35655,35376,31039,27389, 26069,26050,22562,22517. Femur piece: SKX 35550. Tarsornetatarsus pieces: SKX 37598, 33002, 31958, 31854, 26645,26070. Phalanx: SKX 45965. Bird size class 11 Three craniat and 276 postcranlal pieces from a mínimum 0130 individuals. MATERIAL. Upper bilis: SKX 25014, 31636. Lower bill: SKX 34778. Keel pieces: SKX 38916/4, 35538, 35351, 35051, 31823,28595,27602,48230,48433,48486. Sacra and vertebrae: SKX 45958, 45925, 34685, 27231, 26693, 25147c. Coracoid pieces: SKX 48433, 48486, 46626, 38533/2, 37842, 37743,36902,35870,35179,34933,34738,34530,344961, 34338,33894,33891,33729,33311,33261,32024,31819, 31569,31409.30722. 30699, 28847, 28725, 28388. 28200, 28178,28180,27583,27311,27046,26370,25374,25261, 25137,24391.23373.22554,21976,21517.19869. Humerus pieces: SKX 48429, 46717, 45909, 39454, 37847, 37689, 37376,36739,36667,36495,36410,36227,35871,35869, 35774,35673,35371,35339,34934,34912,34671,34592, 34531,34217,34216,33662,33925,33502,33471,33413, 33365,33364.33313,33266,33244,32941,32828,32648, 32577,32508, 32280, 32074, 31913. 31895, 31748, 31563, 31531,29396,29346,34914,29254,29033,29008,28902, 28810, 28555. 28187, 27554. 27503, 27500, 27066, 27003, 26485, 26426, 25858, 25300, 24758, 23367, 22955, 22828, 22250, 21603, 19880, 19779, 19712, 19468, 19415. Ulna pieces: SKX 39560, 36789. 36682, 36668, 34529, 33705, 33501,32037,30346,30126.29936,28097,27325.27064, 26894,26832,26051,25015, 20040, 19968, 19404. Carpometacarpus pieces: SKX 46763, 35837, 33349. 29395, 26914. 26686, 26507, 25971, 25263. Pelvis pieces: SKX 35341, 35336, 35334, 26647. Femur pieces: SKX 46700, 46480,37739,37089.37088,37031,36633,36395,35875, 35397,35375,33902,33837.33301,33281.30299.30014, 29734, 39732, 28590, 28173, 24730, 20103, 20107, 20096, 19835, 19778, 19670. Tibiotarsus pieces: SKX 39044/5, 37511, 37171,36951, 36848, 36737, 36085, 35440,35382,

MATERIAL. Lower bill: SKX 3465ge. Vertebra: SKX 28471. Coracoid pieces: SKX 46857. 46686, 37182, 29829. 29516. 28062.26489,26471,26446,22128,24369. Scapula pieces: SKX 35747, 33491, 33431,32625. 31997. 22470. Humerus pieces: SKX 36785, 36387. 34913. 33259. 29785, 27712, 23149. Carpomelacarpus pieces: SKX 48053, 33618, 27526. 26792, 26312. Wing digit 111: SKX 27831, 35091, 33300, 32591, 29850, 28907, 26553. Femur pieces: SKX 35862, 26892. Fibula: SKX 28002. Tibiotarsus piece: SKX 35641a. Tarsometatarsus pieees: SKX 37208, 31565. Phalanges: SKX 36181,30011,23317.23062.22563,19416. Bird size class IV One cranial and 18 postcraníal pieces from a mínimum 01

three individuals. MATERIAL. Lower tnu: SKX 26685. Coracoid pieces: SKX 46129,37361,36122,32826.32480,30394,27582,21615. Carpometacarpus plece: SKX 46264. Tarsometatarsal pieces: SKX 37609, 31711,25772. Phalanges: SKX 26364, 37672, 33404,29242,26350,19899. Bird indet, (all slze classes). Limb-bone shafts: 108; rniscellaneous preces: 22.

Class REPTILlA Order CHELONIA Family TESTUDlNIDAE Chelonia indet,

Fifty seven postcranial pieces from a minimum 01 two individuals.

MATERIAL. Humerus pieces: SKX 46726, 46044, 22516. Limb-bone pieee SKX 20078. Carapace/plaslron pieces: SKX 46860.46807,46753,46667,46635.46445.46243,46229. 46040,45912,37872,37766,37751,36899.36270,35619, 35592, 34013/14, 32512, 31825, 31598. 31082, 30949, 30644.30643,30395,30285,38383,27663,27585,27295 27229,26796,26404.26308,26252,25329. 25019d,22199 22188,21823,21822,21500,19775,19738,19604,39035 38581,38575,40307,48258,48493.48501,48519 Reptile lndet. Four cranial ano 22 postcranial pieces from a minimum o two individuals. MATERIAL. Maxilla pieees: SKX 31786, 38061 Mandible

COMPOSITION OF SONE ACCUMULATIONS

pieee: SKX 34547. Head plate: SKX 31979. Vertebrae: SKX 46570,37828,36871,28179. Limb-bone pieees: SKX 39004, 38308/5,39590,46235,46222,37062,35008,34559,31792, 27757, 27725, 27616, 26957, 21572, 20506. Pelvis pieees: SKX 46045, 46033, 36580 Class AMPHIBIA Frog/toad indel. Nine postcranial pieces from a minimum 01 two individuals, MATERIAL. Urostyle pieees: SKX 37829. 36736, 26633, 29035. Humerus pieees: SKX 33327, 33346, 46066, 46577. Limb bone: SKX 27277. DISCUSSION Member 3, the 'tire member' beca use ot the occurrence 01 burnt bones that suggests the controlted use 01tire (Brain and Sillen, 1988), ís a very rieh deposit. The remains ot 54 different laxa were recovered, but the lack of Horno sp. banas is notable, although it does not necessarily mean that this genus was not present at the time. Meganterean sp. was still extant - a relatively recent record for this speeies (Brain el st., 1988). The presenee 01 Felis Iybica is a first record for the Sterkfontein valley, as is the pan gol in (Manis sp.) which was identified on the characteristic split third phatanx ot the fore umo, Also recorded ter the first time is Equus burchellii (Burchell's zebra), in coexisten ce with bath Hipparion /ybicum steytleriand Equus capensis. Member 2 material from the pre-1965 Swartkrans excavation referred to Equus quagga in Brain (1981) is now thought to belong to Equus burchellii (Chapter 5, this volume). Hipparion sp. appears to have died out befare Member 5 ot the present excavation, but in that Member E. capensis is still faund together with E. burchellii. The presence of htppopotarnus. alter (Aonyx capensis) and water mongoose (Ati/ax sp.) indicates the likelihood of fairly large expanses of water in the vicinity of the cave. The elephant (el. Elephas sp.) material recovered cornprised deciduous tooth plate pieces from a very young individual. The high minimum number of ?Gazeffa sp. individuals was based on horneares. The presenee 01 a large number of horneares and virtually no teeth ot any 01 the species is inexplicable. For th\s reason the identifieatian ls still in need of verification. The Prote/es sp. of Member 3 is of the large fossil variety. MEMBER 5 Figure 4 illustrates the macrovertebrate taxa found and the material reeovered is listed below. Class MAMMALlA Order PRIMATES Family HOMINIDAE Horno sepíene Five postcranial pieces from a mínimum of one individual. MATERIAL. Thoraeie vertebra and vertebral fragment: SKX 41692 + 41693. Astragalus pieee: SKX 42695. Distal metapodial: SKX 44173. Phalanx: SKX 41280.

59

Papio hamadrayas ursinus - ~ Ten cranial pieces from a mínimum 01three individuals, MATERIAL. Mandible pieee: SKX 44013. Isolated teeth: SKX 43980, 42257/3, 43454, 43668, 41259, 42207, 44206, 40807,44199. Cercopithecoid índet. Three cranial and five postcranial pieces. MATERIAL. Toolh pieees: SKX 42288, 43408, 43540a. Humerus pieee: SKX 45058. Radius piece: SKX 42462. Tibia pieee: SKX 43579/3 Metapodials: SKX 42036, 42495. Order CARNIVORA Family FELIDAE Panthera pardus Three cranial and nine postcrarnal pieces from a minimum of one individual. MATERIAL. Isolated teeth: SKX 44192, 42912, 41005. Humerus pieee: SKX 41619. Astragalus: SKX 43908. Tibia pieee: SKX 43909. Metatarsal pieees: SKX 42357, 43912, 42340,44878. Phalanges: SKX 41563, 42482. Family HYAENIDAE Subfamily Hyaeninae Hyaenid indet. Two pcstcranial pieces from a minimum ot two individuals. MATERIAL. Ulna pieees: SKX 42114, 43903.

Subfamily Protelinae Protetes cristatus Five cranial and three postcranial pieces from a minimum of three individuals. MATERIAL. Maxilla pieees: SKX 41413 + 41415 + 41417, 43038,44193 Mandible pieees: SKX 44108, 47328. Humerus pieee: SKX 42681. Phalanges: SKX 47256, 47392. Family CANIDAE Small canid gen. & sp. indet, Two postcranial pieces from a minimum of two individuals. MATERIAL. Tibia pieees: SKX 43812, 44189.

Canis ef. mesome/as Fifty cranial and 55 posteranial pieces from a minimum 01 one juvenile and tour adults. MATERIAL. Skull pieees: SKX 43480/5, 40863, 40852. Maxilla pieces: SKX 41238, 43013/10, 40845/1, 44205, 40855,40847. Mandible pieees: SKX 44187, 41496, 41493, 47242,40861,40856,40848,40849,40850,40846.lsolated teeth: SKX 44200, 43919, 43382, 43174, 43172, 43133, 43094,42524,41504,41498.44705,42523,43579/1,47362, 47274,45341,40864,40865,40866,40867,40869,40870, 40871,40872,40873,40874,40875,40876,40879,40880a, 40881. Vertebrae: SKX 43509, 43347, 40857, 40853. Seapula pieees: SKX 43879, 40781. Radius pieees: SKX 40833, 40837. Ulna pieees: SKX 40728, 40836, 41010.45378. Carpal bone: SKX 40831. Femur pieee: SKX 40629. Tibia pieees:

60

COMPOSITION OF BONE ACCUMULATIONS

SKX 4267B, 43432, 43418, 44103. Calcan ea: SKX 42800, 47241. Metapodials: SKX 40794, 47374, 41203, 44806: 43927, 43575, 42966, 42215, 42002, 41001, 43336/6 + 43336/17, 42967, 43569, 42591, 44126, 43423, 44818, 44692, 44198, 43336/11, 47335, 42936, 42791, 42778, 42474a, 42032. Phalanges: SKX 41723, 44124, 43566, 44925,44207,41008,44064/10,43579/9,47867,47805. Canid indel. Four eranial and twa postcranial pieces. MATERIAL. Mandible piece: SKX 43571. Isolated teeth: SKX 43387, 41202a, 42319. Fibula piece: SKX 40957. Tarsal bone: SKX 41068.

Family VIVERRIDAE Sublamily Herpestinae Suricata suricatta

Two cranial elements from a minimum 01 ene individual: MATERIAL. Two mandibles: SKX 45097/50, 45030. Mongoose indet. Three craniat and two postcranial pieces from a minimum 01 one juvenile and cne adult.

MATERIAL. Skull piece: SKX 43825. Maxilla piece: SKX 44064/20. Mandible piece: SKX 41448. Tibia pieces: SKX 42726, 44064c. Sublamily Vlverrinae Viverrid indet.

One cranial and ene postcranial piece trom a mínimum 01 one individual. MATERIAL. Mandible: SKX 47278. Scapula piece: SKX 41700. Carnivore indet. Fifteen cranial and 22 postcranial píeces.

MATERIAL. Maxilla pieces: SKX 42820, 42536. lsolated teeth: SKX 41211, 44208, 44203, 44194, 43925, 43067, 41527, 45205, 40997, 45097/39, 45097/6, 43013/9. Atlas vertebra piece: SKX 43994/4. Scapula pieces: SKX 42B38, 45352. Humerus piece: SKX 41949/1. Radius pieces: SKX 42460, 45258. Pelvis piece: SKX 47339. Tibia pieces: SKX 41263, 43909, 43420, 40720. Metapodials: SKX 4406417, 44988/1, 42797, 41998, 45385c. Phalanges: SKX 42004, 43462,40778/4,43391/2,45097/40,44487. Order HYRACOIDEA Family PROCAVIIDAE Procavia transvaalensis Twenty cranial and 11 postcranial pieces trom a mínimum ot cne juvenile and three adults.

MATERIAL. Maxilla pieces: SKX 43389, 43383, 41430. Mandible pieces: SKX 42919, 42894 + 42918, 42893/7, 43377. Isolated upper incisors: SKX 40830, 43336/16, 43831, 44314,43366. Isolated lower incisors: SKX 44632/1,47785, 41204. Isolated upper cheekteeth: SKX 43386, 43427, 43378, 43385, 43492. Humerus pieces: SKX 44975, 41236, 43580,

43561,43486,42488. Ulna pieces: SKX 43510, 43466. Calcanea: SKX 42779, 43593, 47812. Procavia capensis Three hundred and forty saven cranial and 115 postcranial pieces from a minimum ot three juveniles and thirty tour adults.

MATERIAL. Maxilia pieces: SKX 40615 + 40616, 41408, 41405,41394. 44343,44334,44261,44107,43830,43B21, 42965, 42277, 40772/1, 40802, 40825, 43099/1, 41023, 41030, 43245,43532,43092,44766,44755,45343,41226a, 41407, 44BOO, 44757, 44756, 44753, 44653, 44116, 43823, 43574,43531,43506, 42578,42361, 41557,40816,42773R, 43099/6,43438/12,43480/6,44312/3,45272,45074,41589, 41713,42366,42B12,43381, 45211,41227, 47767,42961/2, 44064/13, 43428, 44010, 47813. Mandible pieces: SKX 47327,41423,41245, 41230, 41226, 41218, 44763, 43533, 43477,42342,41702,45301,43050, 42321,43137A,41029, 41013,41002, 44767, 44312/4, 43994/3, 43579/5, 45097/42, 43213/11, 40814, 40779/3, 41495, 42276, 43136, 43246, 43558,43559, 44333, 44337, 44762, 44764, 44922, 41436, 47815,41244,41216,45351,45072,44765,44760,44338, 44123,43826,43649,45247,44331,42538,40985,43401/8, 41949/20, 43213/16, 42863/13, 42805/7, 41708, 42537, 42992, 43534, 43560, 44119, 44332, 44336, 44652, 44754, 45153,41400,41406,47866,41231,44335,43824,43244, 44330, 41591/9, 45097/48, 43556. 47855, 41212. Isolated upper incisors: SKX 47848. 41220, 41219, 41206, 45373, 44745,44741,44637,44325,44321,44320,44319,44318, 44316,44128,44127, 44125, 43999, 43832, 43835, 43540, 43434,42949.42937, 42513, 42389, 42320, 42275, 42242, 41920,41018,40996,44312/1,40976/2,40845/4,45319/9, 43099/9, 42893/3, 40839, 40819, 40537a, 40590, 47870, 47868,41225, 43710/1,44768, 44328, 43841,43542,43414, 42895, 42297, 42288, 42213, 41712, 43522/7, 43480/3, 43391/3, 43013/18, 40778/3. Isolated lower incisors: SKX 40577, 44324, 40808, 40817, 40818, 42203/3, 42203/4, 42203/5, 42798, 42893/2, 42893/4, 45318/8, 40976/1, 43336/16, 43438/8, 44064/13, 44064/16, 41011, 41012, 41022,41710,42231,42421,42802,42899,43398,43453, 43500, 43500a,43501,43541,43544,43545,43555,44315, 44317, 44322, 44323. 44326, 44327, 45340, 43710/2. 44989/1,44989/3,41207,41210,41213,41221,41223, 41239,47808,47851,47872,47873, 47783a. Isoiated cheekteeth: SKX 47285, 43410, 43399, 45387, 40974, 43837, 43548,43551,43550,43547,43536,43535,43497,43495, 43494,41717,41716,41699,41698.41697,41696,41446, 41445, 41445a. 41444, 43013/19, 43213/4. 43286/2, 45320/2, 4340113, 43438/1, 43522/2, 43996/2, 44064b, 44064/8, 44064/11.45220,45342,44985,41017,41019,43537,43543, 43549,42313,42314,42344,42903,42915,43135,43225, 43228,43245,43292.43375,43379,43380,43828,44311, 44736,44737.44740,43377,43384,43496,43838,45356, 41202,41215,47273,47365,47769,47861,47865,47816. Vertebrae: SKX 41217, 42572, 42017, 40573c. Scapula pieces: SKX 40701, 44405. Humerus pieces: SKX 47366, 472B4, 47258, 44816, 44345, 44344, 43819, 43663, 43464, 43406,42395,42392,42384,42018,41720,40998,40994, 40988. 44064/19, 45053/5, 43076/2, 40670, 40654, 40641. Radius pieces: SKX 41241, 45152, 44445, 44190, 44004, 43117,43033,42913,44946,44276,44268,43836,43339,

62

COMPOSITION OF SONE ACCUMULATIONS

Order PERISSODACTYLA Family EQUIDAE

Equus capensis Four cranial and two postcranial pieces from a minimum 01 cne juvenile and one adult. MATERIAL. lsolated deciduous teeth: SKX 43450 + 43451, 43552. Isolated permanenl teelh: SKX 42258, 41571. Carpal bone: SKX 42790. Patella: SKX 47755. Equus burchellii Nine cranial and ten postcrarual pieces from a mínimum 01 tour juveniles and two adults. MATERIAL. Isolated deciduous leelh: SKX 43735, 44234, 43759,43562,41569,44629,44125,42898. Deciduous looth in maxillary piece: SKX 41515. Radius pleces: SKX 41317 + 41332,41191. Ulna pieces: SKX 41194 + 41266. Melacarpal: SKX 44798. Tibia pieces: SKX 43484, 41201/1, 41192. Astragalus: SKX 44184. Phalanges: SKX 41027 + 41190, 43475. Order ARTIODACTYLA Family SUIDAE

Phacochoerus sp, Nine cranial and three postcraniat pieces from a mínimum 01 one juvenile and cne adult. MATERIAL. lsolated looth pieces: SKX 42799, 43522/6, 43221,43918,47847,45157/17 + 45384/22, 45053/2, 41688, 47829. Femur piece: SKX 41337. Metacarpal: SKX 44186. Phalanx: SKX 41831/10. el. Tepinocttoerus meadowsi Four cranial plecas from a mínimum 01 one individual. MATERIAL. lsotated looth pieces: SKX 41690, 41691, 43760,44704. Family BOVIDAE Sublamily Alcelaphini

Megalotragus sp. Two associated upper molars from a mínimum 01 one individual. MATERIAL. SKX 42786, 43452. Connochaetes sp. Sixteen cranial pieces from a mínimum 01 ene juvenile and two adults. MATERIAL. Mandible piece: SKX 42544. Isolated deciduous teeth: SKX 41817, 44221. Isolated permanenl teeth: SKX 42374,43208,43209,44640,44702,41588,43207,43451, 47326,44214,43657,43373,42327. Medium-sized alcelaphine: Beatragus sp./Alcelaphus sp. Forty two cranial pieces lrom a minimum 01 two juveniles and three adutts. MATERIAL. Isolaled deciduous leeth: SKX 44139, 45303, 42323. Isolated permanenl teeth: SKX 42527, 42889, 43128, 43288, 43290 + 43291, 44700 + 44710 + 44719 + 44732, 44703, 44712, 42867, 43388, 44110, 40935, 40787, 40800, 43235,42849, 43005, 44742, 42519, 42529, 42810, 40936,

42505,42809, 43069, 43734, 43748, 44115, 44696, 44715,

44749,45107,45122,4421217,44212/14,41257. DamaHscus sp. Thirty five cranial pieces from a mínimum 01 two juveniles and two adults. MATERIAL. Isolated deciduous teeth: SKX 44064/35, 45118,44636,45102,45134. Isolated permanenlleeth: SKX 42940/3,44746,44727,42061,42311,44748,40950,43129, 43289,44731,47311,41396,41580,42053,44245,45125, 44697,44983,44247,42562,40793 + 40844, 40803, 40842, 42807,44898/4,47259. Horneare pieces: SKX 40582,44990, 41377. Alcelaphine indel. Sixteen crarnal preces. MATERIAL. Isolated tooth pieces: SKX 42312, 43426, 43747,41884,42948,43125,43126,43374,43746,45140, 44241,43726,44223,44212/21,41566,42515. Sublamily Anlilopini

Antidorcas austraHs and/or marsupialis Twanty nine cranial preces from a minimum 01 six adults. MATERIAL. Mandible pieces: SKX 40952, 42341. Isolated permanenl teeth: SKX 44724, 44723, 44716, 42306, 43008, 43019,43036,41442,41551,41561,42064,41914,40955, 41497,41506,43020,43024,44695,41402,44706,40921, 41554,42507,42531,44720,45177,45257. Antidorcas bondi Three hundred and thirty seven cranial pieces trorn a minimum 01 níne juveniles and 37 adults. MATERIAL. Skull pieces: SKX 40579 + 40581 + 40705. Horneare pieces: SKX 42090, 42223, 42436, 42576. 42577, 44995,41372,41375,41376,41378,41393,40905,40972a, 40941a. Maxilla pieces: SKX 40786, 41409, 43049, 40591 + 40592. Mandible pieces: SKX 41424, 41491, 41505, 41428, 42542,43764,44752,43055,42293,40951,41427,47245, 44259,41425,42083,41492,44112,40962,42324,41421, 41426,42545,42081. Isolated deciduous leeth: SKX 44244, 40801,42535,42375,52046,42078,42871,44258,41403, 42512a, 45274, 45097/24, 42863, 40969, 42045/1, 42055, 44638, 45371, 45097/45, 40928, 43226, 43239, 4194917, 47346, 40943, 42511, 42502, 43237, 43226, 44212/9, 44212/11,44212/12,44212/15,42317 + 43233, 42071, 42530, 43223, 43236, 44118, 44973. Isolated permanenl teeth: SKX 42522, 42531, 44229, 44246, 44729, 40914, 40918,40587 + 40591,43222,45295, 45263, 45227, 44928, 44717,44222,44174,44141,44130,44114,43665,43241, 43230,43227,43217, 43215, 43091, 42876, 42532, 42309, 42303,42285,42069,42065,42048,42047,41553,41506, 41439,41412,41401, 42051, 45268, 43239a, 43205, 43046, 42977,42873,42518,42516, 42514, 42500, 42499, 42498, 42077,42075,42074,42068,42057,41570,41507,40922, 40922a,40924,40926,40927,40930,40931,40932,40944, 40948,40783. 40784,40790a, 40798,40841,40573,47265, 47266, 47267, 47280, 47286, 45320/10, 45285, 45265, 44701, 44242, 44233, 44215, 43750, 42961/3, 45319/3, 42988,42316,43996/3,45384/3,43758,47281,44212/8,

COMPOSITION OF BONE ACCUMULATIONS

42045/3,45232,45209,45105,43739,43218,42986,42806, 42520,42512, 42072, 44632/2,44212/18, 43216, 43131, 42869,42568,42508,42307,42067,41441,41435,40937, 40806, 47246, 47283, 47381, 41949/5, 44064/22, 45348, 45339,44980,44725,44721,44709,44699,43742,43740, 43618, 44726. 44243, 44235, 43231, 42506, 42050, 44212/19,45228,42579,42528,42315,42211,40947,40789, 40790,40804,47269,47356,43099/4,45346,45120,44924, 44251,44225,44220,44216,44213,45132,45114,44926, 44254,44131,44121,42808,42059,42052,45231,43130, 40821,47337,44708,44239,44227,41404,42509,42500, 42066,41437,40797,47336,43099/5,44248,44237,43736, 43650, 43287, 43132, 43127, 42866, 42850, 42840, 42525, 42358,42079.41397,42301,42305,40948,40799,44698, 44231,44007,43745,45262,45246,44713,44635,44228, 44224,43744,43028,43018,42218,42533.42076,42062, 40927a, 40939, 40828,47243, 37369,44728,44718,44217, 40780, 40965. 40942, 40925, 44226, 42054, 42056, 42504, 44929,47765,47781,47791,47814,47849,47763,47773, 47776,47807,47885,47864,40826. Antidorcas sp, Five orante! pieces. MATERIAL. Isolated lower premolar: SKX 45328. Horneare pleces: SKX 40809, 42571. 43017, 47380. Raphicerus cetnpestris seveo cranial pieces from a minimum 01 two individuals. MATERIAL. Mandible pieces: SKX 44751,42900. Isolated teeth: SKX 44064d, 42302, 42510,44212/13. Horneare pieee: SKX 40973.

Subfamily Bovinae

Syncerus sp. Two deciduous premolars possiblyfrom the sama individual.

MATERIAL. SKX 41500, 43485.

Taurotragus oryx MATERIAL. Isolated tooth: SKX 43737.

Sublamily Hippotragini Hippotragus et. niger Fifteen cranial pieces from a minimum 01 three juveniles. MATERIAL. Mandible pieee: SKX 40968. Isolated teeth: SKX 42070. 41438, 40945, 45217, 40576, 43738, 44250, 45276,44122,45119,44218,42058,43749,45121. Hippotragini/Reduncini indet. Two cranial pieces.

MATERIAL. Isolated toolh pieees: SKX 42985, 47264.

Subfamily Pelinae Pelea sp. Thirteen cranial pieces from a mínimum 01 three tndividuals. MATERIAL. Mandible pieces: SKX 44750.41906,41494. Isolaled teelh: SKX 41503, 44714, 43756, 44251. 40791. 40840.44176.44255. Horneare pieees: SKX 41946, 47370.

63

Subfamily Reduncinae

Redunca ct. arundínum MATERIAL. A deeiduous tooth: SKX 44219. Bovid slze c1ass I indet. Twenty six cranial and 113 postcranial pieces. Bovid slze c1ass 11 indet.

One hundred and twenty three eranial and 1439 posteranial pieces. Bovid size class 111 indet. Sixteen craníal and 122 posteranial pieees. Bovid size ciass IV indet. One cranial and twa posteranial pieces. Bovid indet. cranial (all size ctasses). Enamel tooth fragments: 692; isolated incisors: 101; isolated premolars: 2; horneare pieces: 78,

Order LAGOMORPHA Lagomorph gen. & sp. indel. Thirteen cranial and 27 postcranial pieces from a mínimum of four individuals.

MATERIAL. Maxilla pieees: SKX 47772. 41240, 41235. Mandible pieees: SKX 47857, 42283, 42214, 43391/6. lsolated teeth: SKX 41214, 41209, 41205, 44989/2, 45385b, 44770. Axis vertebra: SKX 41026. Seapula pieee: SKX 43429. Humerus pieees: SKX 43670, 43818. Radius pieee: SKX 44273. Ulna pieee: SKX 43029. Pelvis pieees: SKX 44129, 41996, 41093. Femur pieees: SKX 44817, 44191, 43601. 43430,40823,44266,41715. Tibiapieees: SKX 42013, 44267, 42318,45101. Astragalus: SKX 41229. Caleanea: SKX 42908, 41233,43662,43154/5. Metapodials: SKX 44811, 42033. Order RODENTIA Family PEDETIDAE

Pedetes sp. MATERIAL. One lemur pieee: SKX 42906.

Family HYSTRICIDAE Hystrix etriceeeustretts MATERIAL. An isolated moiar: SKX 43397. Mammal indet. Five cranial and 83 postcranial pieees from a minimum 01 six individua!s.

MATERIAL. Mandible pieees: SKX 47798, 44759, 43577. lsolated eanines: SKX 41689. 42969. Vertebrae: SKX 41050, 41463,41618,42265.43371,44263,43109,43111,45359. Seapula pieees: SKX 47378, 43344. 41357. 40706, 41058/1, 43480/1,43282,44446,40655,42012. Humerus pieees: SKX 47879, 47846. 47762. 47316, 40779. 41033. 44064/25, 41016.45115.45384/31,44277,44270,43260,43149,42999, 42902, 42227. 41694. Radius pieees: SKX 44667, 44280.

64

COMPOSITION OF SONE ACCUMULATIONS

Ulna preces: SKX 47862, 47778, 47354, 41672, 44347/1, 42805/16. Pelvis pieces: SKX 47373. 41346, 45244, 43920, 43504, 44264, 43433, 42923, 40690, 42003. Femur pieces: SKX 47758. 47842, 47353, 45194, 40829, 40776a. 45360, 45321, 4527S, 44659, 44178, 43409, 43032, 43007, 42944, 42844,42757,42655,41636/1,41621. Tibia pieces: SKX 43599, 42715, 42467, 44502, 45090, 44940, 43913. Calcaneus: SKX 43931. Metapodials: SKX 43468, 43431, 44654. 47874. Phalanges: SKX 44807, 42026. Miscellaneous: SKX 42008, 47332.

Reptile indet. Four postcranial elements possibly from the sama individual. MATERIAL. Vertebrae: SKX 40778/1 and 453851. Limb bone: SKX 43887.

ClassAVES

Class AMPHIBIA

Struthio sp, MATERIAL. One central phalanx piece: SKX 41349, and nine pieces 01 ostrich eggshell: SKX 47823, 47790, 44687, 44688, 44689, 44827, 44828, 44829,43996/16.

FrogIToad índet. Fourteen postcrarnal pieces Irom a minirnurn 01stx individuals, MATERIAL Urostyle pieces: SKX 43971, 43974, 43976, 43977, 43979, 45097/51. Scapula piece: SKX 43438/7. Humerus pieces: SKX 43970, 43401f7. Pelvis pieces: SKX 43972, 43973, 43978. Miscellaneous pteces: SKX 43975, 43367.

Ibis MATERIAL. A lower bill Iragment: SKX 43336/3.

Bird size class I Six postcranial pieces from a mínimum of two individuals. MATERIAL. Sacra: SKX 41701, 42792. Humerus pieces: SKX 45385e, 47775, 40778/2. Phalanx: SKX 43401/2. Bird size class 11 One cranial and 28 postcranial pieces from a minimum 01 three individuals. MATERIAL. Bill piece: SKX 43467. Vertebra and sacrum: SKX 40990, 45318/13. Coracoid pieces: SKX 44671, 41709, 41705, 41831/6, 45216/8. Scapula piece: SKX 42929. Humerus pieces: SKX 42463, 44269, 43419, 42930, 41722, 41711. Ulna pieces: SKX 43479, 40991, 43200. Carpometacarpus pieces: SKX 43584, 41703. Femur piece: SKX 41923. Tibiotarsus pieces: SKX 45085, 42198. Tarsometatarsus pieces: SKX 41706, 41704. Phalanx: SKX 44064e. Lirnb-bone shañs: SKX 44278, 43011,42433.

Bird size class 111 seven postcranial pieces from a mínimum of one individual. MATERIAL. Vertebra and sacrum: SKX 44272, 43438/2. Coracoid: SKX 44064/5. Scapula: SKX 43154/6. Ulna piece: SKX 40707/11. Wing digit: SKX 42203/6. Femur piece: SKX 45156.

Bird size c1ass IV Two cranta! and six postcranial píeces from a mínimum of two individuals. MATERIAL. Lower bill plecas: SKX 42805i12, 45319/15. Coracoid piece: SKX 41039. Scapula piece: SKX 41591/6. Ulna piece: SKX 40707f7 Carpometacarpus: SKX 43587. Phalanx SKX 47350. Limb-bone shall: SKX 47379. Class REPTILlA Order CHELONIA Family TESTUDINIDAE Chelonia indet. Twenty eight carapace pieces from a mínimum of ene individual.

MATERIAL. SKX 43391/5, 43401/10, 44028/2, 43600/5,

45384/35, 42271, 41431, 41627, 41719, 41908, 42261, 42267, 43966, 43967, 43968, 43969, 44058, 44339, 44340, 44341,44342,44808,45063,41195,41196,41197,41199, 47777.

D1SCUSSION The age 01 this relatively recent deposit ls estimated 10 be about 11 000 years B.P. according to radiocarbon dating rnethods (J. Vogel, personal cornmunication). 1t5 fauna is dominated by remains 01 the tossü springbuck, Antidorcas bondi, and the hyracoids Procavia antiqualcapensis and P transveetensis. During this period in its history, the cave was probably a leopard lalr with leopards preying upon what must have been abundant populations of A bondi and Procavia species, Protetes cristatus is present in this deposit whereas only the larger fossil Prote/es sp. is represented in Members 1 and 3. The large hyrax Procavia transvaalensis is, however, still extant, as is the robust Equus cepensis and a suid that is comparable in síze with Tapinochoerus sp. Ostrich eggshell is also present in the deposit, sorne 01which appears to have been worked. AGENTS RESPONSIBLE FOR SONE ACCUMULATIONS IN THE SWARTKRANS CAVE Table 5 shows the percent representation ot various higher categories of macrovertebrates in Members 1-3 and 5. The pattern is largely consistent throughout the depositional period of the cave, and suggests a predator/prey scenarío wilh the large cars. leopards and sabre-toothed cats almost certainly being responsible for much of the accurnulation 01 the faunal rernains. This is suggested by the combined interpretation of the size 01 prey species, anirnals represented ano their age classes, types 01 skeletal remains and carnivore damage cbaerved on bones. Evidence in support 01 this scenario is discussed separately below in respect 01the bovid remains. In Member 3 where the controlled use 01 flre (Brain and Sillen, 1988) and evidence 01 butchering 01 bones becomes apparent, hominids could also have been responsible lar bane accurnurations. These subjects are discussed in greater detail in Chapters 9-13 01this volume. The lossils tound in Member 5 were almost certainly largely the result 01 leopard predatian.

COMPOSITION OF BONE ACCUMULATIONS

Otherbone-accumulating agents undoubtedly also played a role in what has been found in the cave. For example, scavengers could haya intermittently used the cave as a feeding sita and breeding lair. Support ter this notion comes from the proportion of carnivora remains to other animal remains (Tabla 5), which is sufficiently high to suggest that hyaenas, in particular brown hyaenas, brought sorne ofthe banas to the cave. II is well known (Brain, 1981) that these animals hunt other carnlvores, aften accumulating their remains in larga quantitieso In addition. the type 01 damage found on many or the banas are suggestive 01 the cone-crunchinq abñitiee of hyaenas. The presence of carnivora coprolites in Members 1 and 2 also suggests that the cave was usad as more than a feeding sita, perhaps as a breeding lair. Porcupines were also responsible far sorne of the bone accumulations. Their characteristic gnawing patterns on bone are seen in Members 1-3. Small-rodent gnawing marks commonly seen on the bones in all Members were very )ikely made after the bones had found their way into the cave. Owls probably contributed largely to the microfaunal remains in the deposits, 8wartkrans appears to have served as an owl roost for the last two million years (Brain, 1981) and is still utilized by barn owls (Tylo alba). The role of black eagles in part of the hyrax bone accumulatían cannot be discounted. Although they do not enter caves, it is possible that they roosted on cliffs that may have existed in the vicinity 01 the cave. Unfortunately, the hyrax skull rernalns, although plentiful, are too fragmentary for the characteristic bone damage from black eagles, or even leopards. whose favoured prey are hyraxes, to be detected.

Bovids and the predator/prey scenario Evidence for the important role that predators played in the accumulation of macrovertebrates comes especlally from an analysis of the bovid assemblages. The size classes used in the analysis follows Brain (1981): Bovid i: 0-23 kg llve weight (upper limil: large female common duiker), Bovid 11: 23-84 kg live weight (upper limil: large maie blesbok). Bovid 111: 84-296 kg live weight (upper Hmit: large wildebeest or roan antelope). Bovid IV: more than 296 kg live weight (very large anlrnals such as eland and buffalo). The ratios ot number ot individuals in each size class to the total number of individuats for each Member are given in Table 6. This shows that most of the remains come from animals in the Bovid " and Bovid 111 size classes. (The minimum numbers used are based on postcranial rernains that have not been identified to species leve!.) Species identifications were basad on cranial remains exclusively, except tor klipspringer terminal phalanges which are distinct. tt is most Hkely that the postcranial remains recovered carne from the same individuals as the specifically identified cranial remains. Detailed skeletal part analysis is set out in Table 4a-c and shows that almost all of the bovid skeleton is represented, but to varying degrees, and that the percentage survival ot parts appears to be largely due to natural attritional processes (Binford and Bertram, 1973: Brain, 1981), the denser, more robust bones surviving better than the more delicate, tlimsy

65

Table 4

a: Bovid 11 slze creas skeletat part analysis. The values given are percentages of the total ournber of specimens per Member. Member 2

Number of specimens: MNI: SKELETAL PART Calvaría pieces Horneare pieces Maxilla plecas Mandible pieces Tooth pieces Hyoid bone pieces Atlas vertebrae Axis vertebrae Other vertebrae Scapula blade Scapula glenoid Proximal numerus Distal hume rus Humerus shalt Proximal radius Distal radius Proximal ulna Distal ulna Radio-ulnar shaft Proximal rnetacarpal Distal metacarpal Metacarpat shaft Carpals Pelvic pieees Proximallemur Distal femur Femur shatt Patella Proximal tibia Distal tibia Tibia shaft Proximal metatarsal Distal metatarsal Metatarsal shaft Astragalus Calcaneus Tarsal bones Proximal metapodial Distal metapodlal Metapodial shaft 1st phalanges 2nd phalanges 3rd phalanqes Phalanx indet. Sesamolds

3

5

372 6

330 6(2)=8

731 14

1562 33

0,8

0,6

0,3 0,1

0,2

1,9 8,1 0,3 0,5

3,0 1,2 0,3 0,3 2,1 1,5 0,6 1,2

5,1 5,1 0,8

0,5 0,3 1,6 1,1 3,0 1,3 1,1 1,6 1,3 0,3 3,2 1,6 0,3 3,5 3,5 0,8 0,8 0,8 1,9 0,3 1,1 4,3 1,6 0,5 3,5 1,3 2,2 1,9 5,9 0,3 15,3 7,0 9,7 0,8 4,0

0,7 1,9 0,5 1,9 0,8 4,4 0,7 0,5 2,7 2,9 1,2 0,5 1,4 0,3

4,8 0,6 3,9 0,9 3,0 1,2 1,2 0,3 7,3 3,9 1,8 0,6 2,1 4,8 0,3 1,5 0,3 1,2 0,6 3,3 3,9 1,5 0,9 7,9

4,5 7,3 1,5 2.1 1,1 2,6 1,2 1,2 0,3 4,0 0,8 0,5 2,3 3,8 3,6 0,7 4,4

9,4 9,1 9,1 1,2 1,8

9,6 8,2 7,3 0,4 0,8

0,2 4,9 2,2 0,4 0,2 2,5 1,7 3,3 0,5 4,1 1,7 1,6 1,2 2,1 0,2 1,9 1,6 0,7 0,3 3,7 4,4 1,4 2,2 2,0 0,2 1,4 4,3 1,8 2,3 1,1 1,9 2,8 2,8 2,4 1,3 6,5 1,5 11,5 7,2 4,4 0,1 1,2

b. Bovid 111 size class skeletal part analysis. The vatues given are percentaqes of the total number of specimens per Member. Member 2

Number of specimens: MNI: SKELETAL PART Calvaría pieces Horneare pieces Maxilla pie ces Mandible pieces

3

5

128 3(3)=6

520 13

138 5

1,3

0,8

0,7

1,0

3,9

0,4 0,2 0.2 5,2

300 9

2,9

COMPOSITION OF BONE ACCUMULATIONS

66

Table 4 - continued

Tooth pieces Hyoid bone pieces

6,0

Atlas vertebras

1,0 1,3 1,6 1,0 0,7 0,7 4,7 0,7 1,0 2,3 1,3

2,3 0,8 1,6 1,6 8,6 1,6 2,4 3,9 4,7

1,0 1,0 0,3

0,8 5,5 0,8

6,0 2,7 2,3 0,7

8,6 5,5 1,6 0,8

0,7 2,3 2,3

0,8 2,4 3,1

4,3 0,7 2,0 5,0 3,7 3,3 2,3 9,3 1,3 8,0 7,0 4,3 1,3 3,0

3,9 0,8

Axis verteorae Otner vertebras Seapula ceoe Scapufa qfeootd Proximal hume rus Distal humerus Humerus shatt Proximal radius Distal radius Proximal ulna

Distal ulna Aadio-ulnar shatt Proximal metacarpal Distal metacarpal Metacarpal shaft Carpals Felvic preces Proximal temur

Distal femur Femur snatt Patella Proximal tibia Distal tibia

Tibia shaft Proximal metatarsal Dista! rnetatarsal Metatarsal shaít Astraqatus Catcaneus Tarsal bones

Proximal metapodial Distal metapodial Melapodial snatt 1st phalanges 2nd phalanqes 3rd phalanges Phalanx mdet. Sesarnoids

4,3 4,3

5,8 1,2 0,7 1,2 2,1 0,6 1,2 1,0 7,7 0,6 4,6 1,3 4,0 0,4 3,1 4,6 0,4 0,6 3,5 5,0 1,9 0,6 0,2

0,8 1,7

5,5 4,7 1,6 0,8 2,4

2,9 3,6 4,3 2,9 0,7 2,2 0,7 2,9 0,7 3,6 2,2 2,9 2,9 2,2

2,9 2,2 0,7 0,7

8,1 0,2 0,2 3,3 1,9 2,5 1,3 5,4 0,2 4,4 4,6 2,1

4,3 5,1 4,3 1,4 5,3 0,7 5.1 5,4 3,6

1,9

2,9

c. Bovid size crasses I and IV skeletalpart analysfs. The valúes given are numbersof specimens.

8avid IV

Bovid I Member

MNlo

2

3

5

3

2

4

6

2

2

2 5 7

20 1 5

3

7

12 3 9

1 1 6 3 5 1 6

3 3 2 3 2 10 4 1 4 1

2

3

5

3

3

3

SKELET AL PART

Oavana preces Maxillary pieces Mandible plecas Horneore pieces Tooth pieces Hyoid bone Atlas Axis Other vertebrae Seapula blade Scapula glenoid Proximal humerus Distal humerus Humerus shatt

Proximal radius Distalradius Proximal ulna Distal ulna

2 3 2 2 1 3 1 3

2 1 1

3

2 1

Carpals 0,7 2,8 0,7 3,6

0,6 4,0

2,4 2,4 2,4 3,1 7,0

Radio-ulnar shaft Proximal metacarpals Distal metacarpals Metacarpal shatt

2 5 1

Pelvic peces Proximal ternur Distal femur Femur ahatt Patelía Proximaltibia Distal tibia Tibia shaft Proximal metatarsal Distal metatarsa! Metatarsal shatt Astragalus Calcaneus

Tarsal banes Proximal metapodial Distal rnetapodial Metapodial shatt 1st phalanges 2nd phalanges 3rd phatanqes Phalanx indet. Sesamoids

Total

8 1 4 3 1 2 2 2

3

1 4 2

2

1 1 2 2 15 2 3 4

5

6 6

27 1 6 1 2 1 2 3

3 5 4 2 3

1 3 5 3 7 1

9 9 22 1

91

56

146

2

1 2 1 9

6

2 5 5

3

1 5 7 9

2 1 11

139

30

3

2 1 3 6 3 10 2 3 1

1 13

57

3

bones. Although porcupines and hyaenas would have broughl

individual banes or groups 01 bones to the cave at intervals, whole carcasses were apparently brought to the vlcinlty of the cave, the feeding site, and the remains of meals found their way into the cave where preservation occurred. The síze classes of bovids represented in the deposits correspond with the preferred prey size of leopards. This is well illustrated in Tables 48 and 49 01 Srain's (1981) The

Hunters or the Hunted? Most of the Bovid 11 postcranial rernaíns frorn Member 5 are Antidorcas bondi, obviously a favoured prey species of the leopards at that time. Relatively complete limb bones, more extensively damaged vertebrae, scapulae and pelvises reflect the typical feeding pattern ot ieopards. Age-at-death categoriess of A. bondi specimens are shown in Table 7a. Teeth were allocated to these categories on the basis of eruption and wear entena, using a known-age collection lhal had been studied by Turner (1985). The ante-

lepe were taken at all ages but there was a distinct bias towards mature individuals. Antidorcas australis and/or A. mersupislis was present in Member 5 but numbers were relatively low and ageing of individuals was not undertaken. However, the species ís the most common bovid in Members 1-3, and age-at-death determinations could be undertaken using a known-age collection of A. marsupialis, with reference to Rautenbach (1971). The results are presented in Table 7b. As in the case of A. bondi, juveniles were taken to a certain axtent, but mature young adults seemed to be the preferred prey.

1 1 2

1 2

1 3 1

In the Bovid 111 and IV size classes a larger percentage of juveniles is represented in all species than in the Bovid I and II size cíasses (Table 8). This is particularly obvious in the Member 5 deposil. Leopards are known to prefer juveniles of larger prey species (Bram 1981). and such a preference may well have played a role in the accumulation of bovid rernarns at Swartkrans.

COMPOSITION OF SONE ACCUMULATIONS

67

Table 5 Occurrence (expressed as percentages) 01 various higher taxa of animals recovered trom Swartkrans. Member 1 % Primates Carnivores Hyracoids Periseodactvls Artiodactyls Aodents, tacomcrchs. tubulidents, elephants. pholidotes Birds Reptiles Amphibians

Member 2 %

Member 3 %

Member 5 %

14,2

21,4

11,6 12,5 2,6 22,1 9,2

12,3 9,9 3,1 34,5 5,5

10,2 13,9 6,8 2,9 35,5 5,5

2,3 9,6 22,9 4,4

8,0 1,2 18,0

7,6 2,3 3,8

22,1 1,8 0,9

5,7 1,2 3,3

47,1

3,4

Table 6 Bovid sfze ctasses based on postcranet material from Swartkrans.

Member 1 Member 2

n ratio n

ratio Member 3 Member 5

n ratio n ratio

Bovid 1

Bovid 11

Bovid 111

Bovid IV

Total

3 0,15 2 0,11 4 0,12 6 0,13

6 0,3 8 0,42 14 0,41 33 0,70

9 0,45 6 0,32 13 0,38 5 0.11

2 0,10 3 0,16 3 0,10 3 0,10

20

ENVIRONMENTAL RECONSTRUCTION Extant bovid spacies tall lnto three groups with respect to meir habitat requirements: sorne fraquent 1) open savanna, while others prefer 2) savanna woodlands with water or 3) rocky hillsides (Smithers, 1983). As regards the fossil bovid species, those referred to Antidorcaswere placad in the 'open savanna' group by assuming that, as the modern springbuck, they were primarily grazers. The habitat requirements or the fossil ?Gazella sp. are uncertain, but they could have been partiat browsers and partíal grazers as are their modern East African counterparts (Dorst and Dandelot, 1972). Owen-Smith (1985) implies thattheir niche could have differed to some extentfrom that of the Antidorcas species beca use they were of similar síze. However, I have placed them in the 'open savanna' group on the assurnption that they were primarily plains animals. Should they nave belonged to the 'woodland savanna' graup, the overall scenario would not be altered greatly in the sense that the open savanna species would sti/l be the most plentiful

19 34 37

(Table 9.) The MNl ratios in Table 9 show that there was a consrstent pattern over the time-span of the cave: most of the bovids represented graze on open savanna, while bovids that inhabit savanna woodland and the rocky hillsides are relatively poorty represented. At Swartkrans, the savanna woodland would most likely have been represented by a wooded area adjacent to the Blaaubank stream, while klipspringer and rhebuck wouid have lived on the rocky hillsides and krantzes aboye the cave. The rocky habitats would have been home also to the two hyrax species that coexisted throughout the history ot the cave. The results suggest that the water course at the base of the Swartkrans hill, the Blaaubankspruit, was in former times more extensive and permanent than at present. Not only is a fairly extensive rfpertan woodland envisaged, large enough to have supported buffalo, kudu, sable and elephant found in the deposits, but also a large and stable enough expanse of water to have supported hippopotarni, otters and watermongoose. Reedbeds and a vtei-Hke area would also have been requíred by reedbuck.

COMPOSITION OF SONE ACCUMULATIONS

68

Table 7 a. Age-at-death categories for Antidorcas bondi.

Table 8 Minimum numbers 01 adult and juvenile bovids in Members 1-3 and 5 ot Swartkrans.

Member 5

n

Age category

Bovid I and 11: minimum numbers; juveniles given in parentheses.

% 01 total

Member 0-3 3-12 1-2 2-3

months months years years 3-5,5 years 5,5 + years

3 6 8 5 13 11

Total

46

6,5 13,0 17,4 10,9 28,3 23,9

2 Steenbok

0-0,5 0,5-1 1-2 2-3

3 2 1 1

27,3 18,2 9,1 9,1

3-4 4-5 5+

2 1 1

18,2 9,1 9,1

Total

11

5 1(2)

2

14 14(4) 1

6

(Raphicerus campestris) Klipspringer

b. Age-at·death catecortes for Antidorcas australis and/or A. marsupialis. Member 1 Age category - - - - - (years) n %oftotal

3

Member 2

Member 3

n %

(Oreotragus oreotragus) ?Gazelfa sp. A. australislmarsupialis A. recki A. bondi Rhebuck (Pelea sp.) Reedbuck (Redunca ef. arundinum) Blesbok (Damaliscus sp.)

5 6(5)

5 7(3) 1(1)

2

37(9) 3 (1)

1(1)

5(1)

4(2)

14(6) 30

19(4) 17

37(8) 18

2(2)

n

% ot total

2 1 1

20,0 10,0 10,0

1 3 2

5,5 16,7 11,1

4

40,0

7

38,9

Bovid 111 and IV: minimum numbers; juveniles given in parentheses.

1 1

10,0 10,0

2 3

11,1 16,7

Medfurn-sized alcelaphine 2(1) (Alcelaphus sp. or Beatragus sp.) Wildebeest 5(2) (Connochaetes sp.) Kudu (Tragelaphus strepsiceros) Sable (Hippotragus cf. nigei) Megalotragus sp. Buffalo 1(1) (Syncerus sp.) Eland (Taurotragus oryx)

10

of total

18

Total % Juvenile

Total % .Juvenile

8(4) 50

3

5(2)

1 1(1)

12(3) 20

50(12) 19

5(1)

3(2)

13(4)

2(1)

1(2)

(3)

2(1) 1(2)

(1)

24(10) 29

6(7) 54

ACKNOWLEDGEMENTS

My thanks to C. K. 8rain for allowing me to work on the fossil collections and for all his help and guidance. I am indebted to R. A. Newman and G. Turner for numbering and cataloguing all the fossils, and to E. J. Watson for his help and encouragemen!. The staft 01 the Archaeozoology Department 01 the Transvaal Museum, especialiy 1. Plug and E, A, Voigt, have been 01 enormous assistance in allowing access to their comparative calleetion and in helping with identifications, and

far useful discussions. Access to the comparative bird skeleton collectíon of the Department of Birds atthe Transvaal Museum was also much appreciated. Thanks are also due to C. S. Churcher, E. Delson, F. E. Grine, R. Susman, A. Turner and E. S. Vrba far their help in specimen identification. I am also grateful to D, Panagos and R, Mathabathe lor their help with the project. Sincere thanks also to N. J. Dippenaar, A. Dreyer and E, M, Herholdt for editorial support,

COMPOSITION OF BONE ACCUMULATIONS

69

Table.9, Bovid spectes grouped according to habítat requirements and listed as a percentaqe of the total bovid MNls. % of total bovid MNI

Habitat & Species

Member

2

3

5

Open savanna Megalotragus sp. Connochaetes sp. Medium-sized alcelaphine Damaliscus sp. Antidorcas australislmarsupialis Antidorcas recki Antiaorces bond; Antidorcas sp. ?Gazella sp. Raphicerus campestris Total

20,0 8,6 5,7 31,4

2.5 17,5 7.5 15,0

25,0 5,0

3.8 21,S 7,6 7,6 22,8 1,3

1,2 3,7 6,2 4,9 7,4 56,8

5,8 14,3 2,9

12,5 2,5

17.7 3,8

2.5

88,7

87,5

86,1

82,7

Rocky hillside Oreotragus oreotragus Pelea sp.

2,9

1,3 2,5

3,7

Total

2,9

3,8

3,7

Savanna woodland Syncerus sp.

5,8

1,2 1,2

2,5

3,8 1,3 1,3 3,8

10,0

10,2

Taurotragus oryx Tragelaphus strepsiceros Hippotragus niger Redunca ef. arundinum

Total

5,8

5,0 2,5

3,7 1,2 7,3

REFERENCES

BINFORD, L. R. and BERTRAM, J. B., 1977. Sane trequencree-. an attritionat process. In: BINFORD, L. R., ed., Far theory buifding in archaeofogy, pp. 77-153. Academic Press, New York. BRAIN, C. K.. 1981. The hunters or tne hunted? The University ot Chicago Press. Chicago and Londcn. BRAIN, C. K. and SILLEN, A., 1988. Evidence from the Swartkrans cave for the earliest use of fire. Nature 336: 464-466. BRAIN, C. K., CHURCHER, C. C. S., CLARK, J. D., GRINE, F. E.. SHIPMAN, P., SUSMAN, R. L.. TURNER, A. and WATSON, V. 1988. New evidence ot early hominids, their culture and envtronment from the Swartkrans cave. South African Joumaf of Scíence 84: 828-835. OORST, J. and DANDELOr, P., 1972. A fie/d guide to the larger mammaJs of Africa. Collins, London. GILBERT, B. M., MARTIN, L. D. and SAVAGE, H. G., 1985. Avian oste%gy. Madern Printing Company, Laramfe. OWEN-SMtTH, N., 1985. Niche separatkm among African ungulates. In: VABA, E. S., ed., Species and speciation, pp. 176-171. Trans-

vaal Museum Monograph No. 4. Transvaal Museum, Pretoria. MAGLlO, V. J. and COOKE, H. B. S., 1978. Evolution ot African mamma/s. Harvard Universíty Press, Cambridge. AAUTENBACH, l. L., 1971. Agingcriteria in the Springbok, Antidorcas marsupialis (Zimmerman, 1780) (Artiodactyla: Bovidae). AnnaJs of the Transvaal Museum 27: 83-133. SMITHERS, R. H. N., 1983. The mamma/s of the $outhern African subregion. University of Pretoria, Pretoria. TARBOTON, W. R., KEMP, M.1. and KEMP, A. C., 1987. Birdsofthe Transvaal. Transvaal Museum, Pretoria. TURNER, A., 1985. Preliminary evidence for seasonal deposition patterns from Member 2 ot the Swartkrans hominid slte. Joumal ot ArchaeoJogical Science 12: 163-175. TURNER, A., 1987. New tosstl carnivcre remains from the Sterktonte in hominid site (Mammalia: Carnivora). Annals of the Transvaa/ Museum 34: 319-347. WATSON, V., 1991. Form, function and tibres: a preliminary study of the Swartkrans toss!l birds. Koedoe 34/1 : 23-29.

COMPOSITION OF BONE ACCUMULATIONS

70

Appendix I The SKR material

This sample ís derived lrom a large block 01 Member 1 Hanging Remnant breccia that became detached from the cave's west wall at the contact between the Outer and Inner Caves. lt contained a cranium of Australopithecus robustus, SKW 11, originally found by Elizabeth Voigt and later known as 'Liz's sku!l.' The whole block was processed in acetic acid so that all tossn banas and stanes were recovered for inspection. This provided the tirst complete Member 1 Hanging Remnant assemblage ever studled from Swartkrans. The previous sampíes, both frcm the Broom/Robinson excavatlon and mjning operatíons. were all selected to a greater or lesser extent. The breccia block,broken intopieces, was assignedto trays and weighed. Following acid-preparation,allbanes and stanes were weighed, while the banas were numbered and catalogued. From the 709 kg 01 matrlx that was prepared, 13,426 kg 01 stone (1,9 %) and 1,235 kg of bone (0,2 %) were recovered. The weight 01 the hominid skull referred lo above was 0,216 kg and that 01 a bovid scapula 0,056 kg, but in the analysls beJow these specírnens were not taken into account as they were in a separate block 01 matrix. Stone There wera 1340 stones in total which fall into the following categorías:

157 dolomite (11,7 %) 1158 chert (B6,4 %) -1057 stones; 92 water-worn pebbles; 7 tlakes. 10 quartz (0,7 %) 7 travertine (0,5 %) (0,7 %) 9 breccia other (0,1 %)

e) Bone ttekes: total number = 271. Size class n % of total

0-1 1-2 2-3 3-4

cm cm cm cm 4-5 cm 5-6 cm 6-7 cm 7-8 cm 8-9 cm Total

2 109 73 46 21 14 3

2 1 271

=

d) Skull fragments: total number 24. 1-2 cm, n= 7. SKR 174, 262, 274, 299, 438, 658, 819. 2-3cm, n= 13. SKR 83, 84, 99, 253, 258, 260 (three pieces), 278,416,443,759, no number. 3-4 cm, n = 3. SKR 309, 349, 364. 5-6 cm, n = 1. SKR 260.

=

e) Enamel tooth fragments (essentially bovid): total number 53. 0-1 cm, n = 1. SKR 824. 1-2cm,n=23. SKR 140, 155,229,243 (three pleces),261 , 264,330,425,445,447,484,511,597,643, 734,810, B23, B25 (tour pleces). 2-3cm,n=21. SKR6,48,62,63, 107, 112, 120, 126, 133, 135,248 (three pieces), 315, 502, 525, 779, 798, 899 (two pieces), 903. 3-4 cm, n = 4. SKR 92, 321,358,440. 4-5 cm, n = 4. SKR 420, 434 (two pieces), 862.

z

% MNI 9,5 52,4 4,8 9,5 4,8 4,B 4,8 9,5

g) Rib fragments: total number = 15 (12 shaft preces, three head pleces). 1-2 cm, n = 3. SKR 66, 94, 395. 2-3 cm, n 9. SKR 11, BB, 304, 335, 362, 363, 417, 497. 907. 3-4 cm, n 2. SKR 44, 101. 6-7 cm, n = 1. SKR 755.

In total, 1277 bonas were recovered. a) Microfauna: number of specimens = 137.

Murid

Cryptomys Rodent indet, Elephant shrew Shrew Sat Sird Total

NISP 4 17 1

MNI 2 11 1

3

2

1 2 1 4 33

1 1 1 2 21

=

1) Vertebral fragments: total number 21. 1-2 cm, n = 5. SKR 91,113,173,337,538. 2-3 cm, n = 6. SKR 14,389,400,452,849,909. 3-4 cm, n = 6. SKR 105, 117, 198,353,366,375. 4-5 cm, n = 4. SKR B7, 157,355,456.

Bane

Otomys sp.

0,7 40,2 26,9 17,0 7,7 5,2 1,1 0,7 0,4

= =

h)

b) Miscellaneous pieces: total number = 766. Size cJass n % of total 0-1 cm 12 1,6 1-2cm 617 BO,5 2-3 cm 103 13,4 3-4 cm 30 3,9 4-5 cm 4 0,5 Total 766

Macrovertebrate taxa: total number of bones = 94.

Cercopithecoid indet. Three specimens from a minimum of one individual.

MATERIAL. Molar fragment: SKR 900. Distal hume rus piece: SKR 571. Third phalanx: SKR 64. Panthera pardus MATERIAL. Radius piece: SKR 490.

COMPOSITION OF 80NE ACCUMULATlONS

Hyaenid indet, Two specimens from one [uvenlie and ene adult. MATERIAL. Phalanges: SKR 235.595.

Canis mesomelas MATERIAL. Fragment of a RP,: SKR 758.

Carnlvore lndet. MATERIAL. Two specirnens. Cervical vertebra 01 a small carnívora: SKR 652. Third phaJanx from a srnau canid.

Procevie antiqua Six cranial and tour pcstcranial pieces from a mínimum 01 cne juvenile and one adult. MATERIAL. Maxilia: SKR 902. Mandible pieces: SKR 414. 496. no number. Isolated teelh: SKR 544. 733. Thoracic vertebra: SKR 639. Seapula blade: SKR 517. Metacarpals: SKR

21,142. Equid inde!. Two cranial pieces from a minimum 01 one individual. MATERIAL. Molar pieees: SKR 271. no number. Connochaetes sp, Two craniat elements from a mínimum af two individuals. MATERIAL. Molars: SKR 684, 685. Raphicerus campestris MATERIAL. An erupling RM': SKR 510. Hippotraqini/Heduncln! Indet, MATERIAL. Damaged lowermolar: SKR 115.

Bovíd size crass I One cranial and three postcranial pieces. MATERIAL. Incisor pieee: SKR 822. Humerus pieee: SKR B07. Radius piece: SKR 835. Phalanx ptece: SKR 3. Bovid size class 11 Two cranial and 24 postcranial pieces from a mínimum of two individuals (one blesbok in slze, one smaller). MATERIAL. Mandible pieee: SKR 840. Isolated ineisor: SKR 916. Lumbar vertebra: SKR 232. Humerus pieees: SKR 311. 469,701. Radius pieee: SKR 692. Ulna pieces: SKR 638. 912. Carpal bones: SKR 514, 640, 641. 649, 653. Metaearpal pieee: SKR 421. Pelvis plece: SKR 751. Tibia pieces: 26.550. Tarsal bones: SKR 82, 715. Sesamoids: SKR 650. 661. Phaianges: SKR 4, 352, 368, no number. Bovid size class '" One cranial and 11 postcranial pieces . MATERIAL. Molar pieee: SKR 361. Atlas vertebra píece: SKR 9. Radius pieees: SKR 234. 625. Metaearpal piece: SKR 626. Pelvis piece: SKR 199. Tibia niece: SKR 776. Tarsal bono: SKR 741. Metapodial pieees: SKR 700. no number. Phalanges: SKR 325, no number. Bovid size class IV One cranial and ene postcranial piece trom a minimurn ot one individual.

ti

71

, MATERIAL. Molar pieee: SKR 236. Metatarsal pieee: SKR 489. Bovid indet. cranial (all size classes) MATERIAL. Two mandible pieees: SKR 306. 596. Three horneare pleces: SKR 123, 576, 771.

Lagomorph gen. & sp. indel. Two postcraniaJ pieces possibly from the sama individual. MATERIAL. Distal metapodials: SKR 197. 832. Mammal indet. Four cranial and eight postcranlaí pieces. MATERIAL. Mandible pieees: SKR 153. 154. Tooth pieees: SKR 735, 881. Humerus pieces: SKR 413, 607, 836. Carpal bone: SKR 323, Femur shaft: SKR 701. Phalanx: SKR 866. Unlused epiphysis: SKR 402. Shaft: SKR 384. Bird size class 11 (francolin to guineafowl in size) Fíve postcranial pieces from a minimum of one individual. MATERIAL. Ulna piece: SKR 486. Carpometacarpus piece: SKR 530. Tibiotarsus pieee: SKR 487. Shafts: SKR 268, 904.

Reptile indet, MATERIAL. Vertebra lrom a relatively large reptile: SKR

753. Damaged bonas: total number:::: 21. a) Apparent bone tools: two bones appeared rounded and smoothed al one ando Both were Ilakes. SKR 108. 703. b) Grey discolouration: SKR 488 and 492 were two bone flakes with a grey discolouration. On closer visual exami· nation they díd not appear bumt. SKR 492 al so has carnivare tooth rnarks on lt. c) Carnivore tooth and gouge marks: 11 bones fall into this category. Identifiable fragments: bovid 11I atlas vertebra piece: SKR 9. Procavia antiqua mandible piece: SKR 496. Bovid 111 [uvenlle 2nd phalanx: SKR no number. Shaft pieee: SKR 686. Miseellaneous piece: SKR 449. Flakes: SKR 167. 214, 233. 492, 604, 671. d) Digested bones: two bone pleces. both unidentifiable, appeared digesled: SKR 269. 693. e) Rodent gnawing: three bone flakes had the fine scratches associated with small rodent gnawing on them: SKR 444, 664, 724. SKR 444 also showed signs 01poreupine gnawing. f) lnsect damage: one miscellaneous piece appeared to have been bored by an inseel: SKR 696 + 711. These results are elucidating: from an initial 709 kg of matríx, 13,4 kg 01 stone and a mere 1,2 kg 01 bone (0,2 % 01 total weight) were recovered. The stones recave red were mainly chert (86,4 %). which was subdivided into stones (most Iterns), water-worn pebbles (about 7 % ot total) and flakes. The waterworn pebbles indicate the role of water in the original deposition of the rnatnx. Few flakes were faund, and these could have been formed naturally. A total number ot 1277 banes was recovered, of which 60 % fell into the miseellaneous category (Table 10). Mast 01 the

72

COMPOSITION OF BONE ACCUMULATIONS

misceUaneous fragments were srnau (1-2 cm). Bane ñakes, were the next mast numerous ltems, mast 01 them baing less than 4 cm in length. This is in accordance with the findings dlscussed elsewhere in this Chapter . Microfaunal remains were relatively seant, which indicates that the deposit was not in the vicinity 01an owl roost . Murines were the most plentilul, with otomyines (Otomys sp.) also presento Various insectivores and birds were tound. These results are in accordance with previous findings al 8wartkrans (Brain, 1981).

The macrovertebrates that could be assigned to taxa comprised 94 bones (7,4 % 01 total bone sample) (Table 11). Besides the robust apernan skull that was separately removed and not includad in the presant analysis, twelve taxa are represented in the small and fragmentary sample. The discovery 01two bone lools (both lashioned on Ilakes) in the material, was excJting. Mast 01 the damaged banes were carnivore-damaged and two appeared lo have been digested. Marks made by rodent gnawing were found on three banas and one had insect boring damage. The results presentad here are typical ot the Swartkrans cave Member 1 Hanging Remnant (Brain, 1981; Chapter 13, this volume). 01 interest ís the low density 01 bone in solid breccia, and even though Swartkrans could be considerad a rieh faunal site, the tlndings indicate just how much sediment had to be removed to obtain the fossil material.

Table 10 Number 01 banas in the vanous categorías.

Microfauna Miscellaneous pieces

Bone tlakea Skull fragments Enamel tooth fragments Vertebral fragments Rib fragments Macrovertebrate laxa Total

n

%oltotal

33

2,6 60,0 21,2 1,9 4,2 1,6 1,2 7,4

766 271 24

53 21 15 94 1277

Table 11 Macrovertebrate taxa identified in the SKR block 01 breccia from Member 1 Hanging Remnant. NISP = number of spectmens: MNI = minimum number ot individuals.

Cercopitheeoid fndet. Panthera pardus Hyaenid indet. Canis ef. mesomelas Carnivore indet. Procavía antiqua Equid indet. Raphícerus campestrís Connochaetes sp. HippotraginilReduncini indet.

Bovid 1 Bovid 11 Bovid 111

NISP

MNI

%

3 1 2 1 2

1 1

2 1

5,0 5,0 10,0 5,0 10,0 10,0 5,0 5,0 10,0 5,0

2

10,0

10 2 1 2 1 4 26 12

2 1

2 2 1 1

Bovid IV Bovid indet. crarual Lagomorph gen. & sp. indet. Mammal indet. Bird size class 11 RepUle lndet.

2 5

5,0

2 12

5,0

1

1 1

Total

94

20

5

5,0 5,0

Appendix 11 Remarks on Member 3 fossil bird remains Numerous bird remains have been found in variaus deposits (Table 3), and results 01 a pilot study on Member 3 are summarized below (Watson, 1991). For comparison the bird remains were placed in broad size categories on the basis of bone síze and robustness: Bird 1: small birds up to francolin in síze. Bird 11: francolin to guineafawl in stze. Bird 111: guineafowl síze to ibis in size. Bird IV: larger birds such as big raptors and ostrich. Sorne bones were characteristic and were assigned to familíes: for example, the spur on the francolin tarsometatarsus, the bill 01 the ibis, and ostrich banes. A total o/ 237 bones was assigned to 11 bird families (Fig. 5),

all of which are still represented in the area at present (Tarbotan, Kemp and Kemp, 1987). The bones were first identilied to skeletal part and then, by using the extensive comparative skeletal collection in the Transvaal Museum's Department al Birds, they were assigned to families, a largely practicable task (Gilbert el al., 1985). Identilication beyond family level is aften difficult. During initial analysis of bird banes from Member 3 it becarne apparent that certain skeletal elements were better represented than others (Fig. 6). More than a quarter (25,7 %) of all the bones assigned to families were humerus pieces. The coracotds also survived well and the ulna, tibiatarsus and tarsometatarsus were also well represented. The fact that

COMPOSITION OF SONE ACCUMULATIONS

73

Fig.6 Fig.5

Percentage skeletal part representation of the fossil bird collecticn identified to family level.

Btrd families identified in the Swartkrans Member 3 deposit, represented as percent mínimum number ot individuals. n '" 50. a. b. c. d. e. f.

g. h.

i. j. k.

Phasianidae (guineafowl and francolins) Sturnidae (starllnqs) Accipitridae (eagles, hawks, vultures. osprey) Corvidae (crows) Tytonidae (barn and grass owts) Threskiornithidae (ibises and spoonbills) Falccnidae (falcons and kestrels) Columbidae (pigeons and doves) Anatidae (typical ducks and geese) Ardeidae (herons, bitterns and egrets) Podicepididae (grebes).

tnese banas are better representad than other skeletal parts is probabfy because they are the more robust skeletal elements. In the case of the stemurn, the keel of which is generally very delicate, the more robust proximal end is the portian that tends to be preserved in the palaeontoloqical record. The pereent minimum number of individuals represented in the collection and in each family is given in Fig. 5. Of the total of 50 individuals, 12 belonged to the family Phasianidae (guineafowl and francolins). AH bones assigned ro this family were too small to be guineafowl and were thus assumed to be

francolins, with which they comparad favourably. At least four specimens had characteristic spurs on the tarsometatarsi. Swartkrans cave is believed to have been a lair tor carntvares and a shelter for fossil man during its deposttlonal phases (Brain, 1981). It is uncertain how the remains of the particular bird families identified found their way into the cave. Members ot most of the families may well have been either preyed upon or scavenged and brought back to the cave ter consumption. The remains 01 eave-dwelling species sueh as owls are not diffieult to account tor, but the presence 01 the raptor is enigmatic.

Chapter 3

Description and Preliminary Analysis of New Hominid Craniodental Fossils from the Swartkrans Formation Frederick E. Grine Oepartments of Anthropology and Anatomical Sciences. State University ot New York, Stony Brook, NY 11794. USA

Near!y 100 hominid craniodental tossfls have been recovered from tbe Swartkrans Formation through recen! excavations by C. K. Brain. Specimens derive from lhe Member 1 Hanging Remnant breccia. as well as from

in situ deposits that constttute Member 1 Lower Bank, Member 2 and Member 3. The remains of sorne six individuals are attributable te Horno, while the remainder ot the fossits that can be identified taxonomical1y are ascribable to Paranthropus. The Paranthropus scecrmens from the vartous Iithostratigraphic units within the Swartkrans cave are morphologically and metrically ver¡ similar to one another. tt te proposed that all are attributable to a single species. Comparisons between the Swartkrans ano Kromdraai Paranlhropus samples

reveal difterences in tooth size and shape that may atteet te tñeir taxonomic distinctiveness.

INTRODUCTION The site of Swartkrans has long been recognized as the single richest hominid fossil deposit in Africa. Well over 200 eraniodental íossils attributable to the 'robust' australopithecine, Paranthropus, have been recovered from tha Member 1 Hanging Remnant breccia through work by R. Broorn and J. T. Robinson between 1948 and 1953, and by C. K. Brain between 1965 and 1979. Mann (1975) has estimaled that at leasl113 individual s are represented by this sample, while more recent estimates place this number between 85 and 87 (Brain, 1981). Several specimens recovered by Broom and Robioson (i.e., SK 15, SK 18a, SK 18b and SK 43) that probably represe nt a single individual derived from a small pocket of Member 2 breccia (Robinson, 1953). Excavations by C. K. Brain from 1979 to 1986 have brought to a clase the third phase of palaeontological activlty lnitiated by him at the site. Not only has this third phase of research added significantly to the clarification the stratigraphy and geological history 01 Swartkrans (Brain, 1981, 1982a,b, 1985, 1988), but lt has also resulted in the discovery of well over 100 additienal hominid tossils from this rich repository. Mest 01 this recent collection comes from the excavation ot in situ deposlts (Braln, 1988; Srain el al., 1988). In addition to these 72 individually numbered jaws and teeth and 36 pcstcranial bones, sorne 21 craniodental tossus and a partía! cervical vertebra that have yet to be described have been recovered by Brain from Member 1 Hanging Remnant rubble. Eanier work at the site distinguished two separate lithostratigraphic units, Members 1 and 2 (Brain. 1976; Butzer, 1976). The recent excavations of in situ decalcified breccia and/or unconsolidated sediments have resulted in the recognition of an additional three Members (Brain, 1988; Chapter 1, this volume). This work has yielded sizeable fossll samples from the Member 1 Lower Bank and Member 2 deposits, and a

o,

somewhat smaller sample frorn Member 3 sediments. Several teeth and pcstcraníal bones were recovered from coordinates alonq the interface of Members 1 and 2, where the separation between them is not readlly apparent (Brain et et., 1988). These lossils have been described in detail by Grine (1989) and Susman (1989). The 21 individually numbered craniodental specimens recovered in tne 19705 by Brain from blocks of Member 1 Hanging Remnant breccia have yetto be described. Comparisons between the fossils comprising the Member 1 Hanging Remnant sample and those from Member 1 Lower Bank deposits revealed no instance of definite or even probable individual assoclanon, and none of the fossils from Member 2, Member 1-2 interface or Member 3 sediments appears to be associated with any ot the tossüs that constitute the Member 1 Hanging Remnanl sample (Grine, 1989). Similarly, comparisons between the large, previously available Member 1 Hanging Remnant hominid sample and the new specimens from that stratigraphic unit indicated no instance ot definite assoclatíon, The purpose of this paper is to provide a brief descriptiva account and an analysis of the recently recovered craniodental tossils, These new specimens previde evidence of Paranthropus from time-successive Member 1, 2 and 3 sediments, and th!s affords the opportunity to assess the reratlonshtps between chronologically successive hominid samples at a single South African locality. The descriptive terminology used here follows that ot Weidenreich (1936), as modilied by White and Johanson (1982) tor the mandibles, Tobias (1967), as modilied by Kimbel, Johanson and Coppens (1982) for the maxillae, and Rabinson (1956), as modified by Tobias (1967) and Grine (1981, 1984) for the dentition. Al! specimens were examined with a low-power binocular microscope; all measurements were taken using a dial vernier calliper with tape red ends and recorded to the nearest 0,1 mm. Measurements 01 the

NEW CRANIOOENTAL HOMIN10 REMAINS

76

mesiodistal (MD) diameters of worn teeth are provided, and in those instances where it was possible to estjmate the pristíne value with reasonable accuracy, this is recorded as 'MD est.' The maximum buccolingual (Bl) diameter was measured perpendicular to the MD axis of the tooth. Values provided in parentheses indicate lower confidence for the estimated MD and 8l vaJues for worn and/or damaged specimens. The descriptions that follow are grouped by stratigraphic unit, and in numerical order within each unit except where isolated teeth are regarded as having belonged to a single individual.

MEMBER 1 LOWER BANK SPECIMENS A total of 20 numbered specimens comprises this sample, which is considered to represen! between 13 and 16 individuo als. Of these 20 specimens, SKX 5007 and SKX 5014 possibly come from a single individual. At the same time, specimens SKX 3354, 3355 and 3601 definitely derive from a single individual, and while the me sial portion of the SKX 3356 crown is missing, il is likely to be associated with the previous three. As noted by Grine (1989), while SKX 3601 was provisionaJly catalogued as being from the Member 1-2 interface, it was found to be associated with SKX 3354 and 3355 through interproximal contact. SKX 3601 is, therefore. regarded as having derived from Member 1 (Grine, 1989). The SKX 5023 lower molar possibly comes from the same individual as the aforementioned maxillary teeth, and SKX 5002 and SKX 5004b may also come from this specimen. The Member 1 lower 8ank fossils are listed in Table 1. SKX 3300 RI 1 (Fig. 1). This is a complete, well-preserved crown and nearly complete root. Wear is slight. The mesial crown corner is comparatively sharp; the distal corner is more rounded. The sljght lingual cervical eminence is symmetrically disposed; there is no tubercle. The mesial and distal marginal ridges are moderate; the slight median ridge is short. The lingual surface is slightly shovelied. 1

Table 1 Hominid fossils excavated from Swartkrans Member 1 Lower Bank deposils.

Specimen

Excavalion coordinates

SKX 3300 SKX 3354 SKX 3355 SKX 3356 SKX 3601· SKX 3890 SKX 5002 SKX 5004b SKX 5007 SKX 5014 SKX 5015 SKX 5023 SKX 5024 SKX 6013 SKX 6277 SKX 7781 SKX 16060

RI' Lp· LM z LM 3 LM' Rp3 LM 3 LC RM3 LM 3 RM , LM? Re LM' Ap3 Rdj'

E3-N4 E3-N4 E3-N4 E3-N4 E3-N4 E3-N4 E3-N4 E4-N5 E3-N5 E4-N4 E5-N4 E4-N4 E3-N4 E3-N4 E2-N4 E4-N3 E3-N5

SKX 19031 SKX 21204 SKW 12

Al' mandible maxilla

S4-W1 E5-S4 W6-N1

L1,

'originally catalogued as having derived Irom (he Member 1-2 interlace; placed in Member 1 through individual association.

is moderate; the cusps are reduced to a nearly flat plane. A large dentine island is exposed on the protocone; smaller exposures are on the paracone and melacone, and the hypo-. con e shows the smallest exposure. The occlusal crown oulline

3

SKX 3354/3601/3355/3356 lp4, lM , lM 2 , lM (Fig. 2a, b). These four teeth are considered to derive from a single indio vidual.

Lp4 (SKX 3354). This is a complete, well-preserved crown with mostof the lingual and mesiobuccal roots and a short distobuccal root stump. Occlusal wear is moderate; the surface is reduced to a nearly flat plane, and large, concave dentine exposures mark the mesiobuccal and mesiolingual crown corners at the paracone and protocone positions. The occlusal crown outline is an irregular ovoid, with a reduced distobuccal corner; the lingual edge is slightly skewed mesially. The protocone appears to have been sítuated slightly mesial of the level of the paracone. 8uccally, a vertical depression demar· cates the distal edge of the paracone, and the dislal third of this face is f1attened behind it. There are three separate roots, and the neck IS 5,4 mm high buccaliy. Fig.1

LM (SKX 3601). This is a complete, weli-preserved crown with most of Ihe lingual rool and short buccal rool segments. Wear

.S

480-490 cm 490-500 cm 490-500 cm 490-500 cm 500-510 cm 530-540 cm 610-620 cm 610-620 cm 640-650 cm 600-610 cm 630-640 cm 600-610 cm 640-650 cm 610-620 cm 540-550 cm 540-550 cm surface lo 550 cm exactly 75 cm 220-230 cm 275 cm

SKX 3300, lingual view; scale = 1 cm.

NEW CRAN!OOENTAL HOMINID REMAINS

77

..'•

a

Fíg.3 SKX 5002, occlusal view; scale = 1 cm.

mesial face of Ihe distal marginal ridge. A moderate distobuccal cuspule is present. Lingually, there is a short accessory furrow that runs obliquely just dis1al to the principal groove. b

Fig.2 SKX 3354/3601/3355/3356 composile in occlusal (al and bucea! (b) views; scale = 1 cm.

is nearly square, and lhe tour principal cusps are well developed. The protocone is the largest; the paracone, metacone and hypocone appear lo have been nearly equal in size. The distal trigon cres! appears to have been well developed. The buccal groove ends in a pit enclosed by a low, lhick enamel wall. LM Z (SKX 3355). This is a complete, well-preserved crown with much of the lingual root and short segments of the buccal roots. Wear is moderate; all tour cusps are reduced to a nearly planar surface. Moderate dentine exposures are present on the protocone and paracone; the metacone and hypocone have small exposures. The occlusal crown outlíne is nearly square, wíth a slight mesiobuccal projection of the paracone. The four principal cusps appear to have been well developed. The protocone was evidently the largest; the other three appear to have been roughly equivalent in size. Most details ot occlusal morphology have been lost to wear, but it appears that the distal trígon crest was well developed. The root neck is 5,8 mm high buccally. LM 3 (SKX 3356). This is an incomplete crown with much ofthe lingual half of the lingual root. The crown is missing the mesial and buccal portíons and the cervical part of the distal surface. Wear on the preserved part of the crown is siight; it is evident that dentine was not exposed on the protocone. The hypocone is well developed. The distal trigon crest is incised al the base of the metacone by a deep, narrow físsure. A short accessory fissure runs distally from the talon basin to partially incise the

====..

; me

SKX 3890 Rp3 (No illustratíon). This is a very small triangular crown wedge wi1h much 01 1he lingual root. The crown preserves a band 01 lingual enameJ and part of the protocone dentine exposure. SKX 5002 LMs (Fig. 3). This is a complete, well-preserved crown with much of the mesíal root and a segment of the distal root plate. Wear is slight; all cusps show enamel tacets ánd the protoconid has a very smaH dentine exposure. The crown has a MD e/ongate, distally tapered occlusal outline. The f¡ve principal cusps are well developed, and a large tuberculum sextum is presen!. The metaconíd and protoconíd are nearly equal in size and are the largest cusps. The hypoconid ís slightly larger lhan the enloconid; the hypoconulid is the smalles!. The mesíal marginal ridge is complete and comparatively thick. There is no distincl fovea anterior or fovea posteríor. The metaconid contacts the hypoconid, and the mesiobuccal and lingual fissures occupy the same transverse plane. The lingual fissure is interrupted at the occlusolingual margin by an enamel wall. The mesiobuccal groove terminates at the end of a deep lissure that courses obliquely across the protoconid and is bound by an enamel wall that forms a cingular welt. The distobuccal groove is short; an accessory vertical fissure is located mesial to it. The distobuccal and distal crown faces display a number of short vertical furrows. The distal fa ce has a protuberance just below the rever ot the occlusal surface between the hypoconuJid and the tuberculum sextum; it expands cervically to lorm a prominent bulge at the level of the enamel margino The root neck is 2,4 mm high buccally. The MD compressed mesíal root plate has separate buccal and lingual canals; lhe large distal root has a single canal. SKX 5004(b) Lh (Fig. 4). This is a nearly complete crown with an intact root. EnameJ ís missing from the disto-incisal angle

NEW CRANIODENTAL HOMINID REMAINS

78

Fig.4

Fig.S

SKX S004b. lingual view; scale = 1 cm.

SKX 5007. lingual view; scale 1 cm.

and from the distal aspect 01 the lingual distal marginal ridge. Wear is slighí. The lingual cervical eminence is very slight; there is no tubercle. Weak mesial and distal marginal ridges bound a shallow central hollow and a barely perceptible me· dian ridge. The root is 17,7 mm long as preserved, and its originallength was likely 20,0 mm.

as the lingual groove are interrupted at the occlusobuccal and occlusolingual margins by enamel walls. The deep mesiobuccal groove terminates abruptly in a pit Irom which a deep, oblique lissure courses towards the protoconid tipo A thick enamel wall is developed buccal to this furrow. The distobuccal groove is slight. The broad mesial root plate has separate buccal and lingual root canals, and the distal root is robust. The root neck is 3,7 mm high buccally.

Le

SKX 5007 (Fig. 5). This is a complete crown with an intact root. Wear is moderate. There is a barely perceptible distobuccal depression bounded distally by a slight enamel ridge; there ís no mesial depression or ridge. The slight lingual cervical eminence is skewed distally; there is no tubercle. Weak mesial marginal and median ridges bound a shallow, V-shaped depression. The moderate distal marginal ridge is separated írom the median ridge by a deep cleft. There is no lingual shovelling. The robust root is 25,9 mm long.

=

SKX 5014 RM 3 (Fig. 6). This is a complete crown with much 01 the root system. Wear is very slight, with enamel íacets on

all cusp tips. The crown has a MD elongate. distally tapered, triangular outline. The five principal cusps are present, as is a large tuberculum sextum. There is no tuberculum intermedium. The metaconid is the largest cusp, followed by the protoconid and entoconid, which are nearly equal in size. The hypoconid is slightly smaller. and the hypoconulid is considero ably reduced, being smaller than the tuberculum sextum. The moderate me sial marginal ridge is complete. The lovea anterior is a thin, Iingually directed furrow írom the mesial end 01 the longitudinal lissure. The crown has a marginally crenulale appearance. The metaconid contacts the hypoconid, and the lingual groove is situated just distal to the level oí the mesiobuccal groove, providing a Y occlusal pattern. The lovea posterior is a deep. narrow íissure that curves around lhe tuberculum sextum and onto the distal surface on eilher side 01 this cusp. The mesiobuccal and distobuccal grooves as well

Fig.6 SKX 5014, occlusal view; scale

=1 cm.

NEW CRANIODENTAL HOMINID REMA/NS

SKX 5015 LM 3 (No illustration). This is a distal crown portion with a segm ent of the di stal rooL The crown fragme nt is broken mesially through the mesíal aspect of the hypoconid and entoconid tipo Wear is slight, with enamel facets on all preserved cusps. The hypoconulid and large tuberculum sextum appear to have been bifid.

.~jg.

79

7

SKX 5023, oeelusal

view (a), mesial view (b), distal view (e); seale ;;; 1 cm.

SKX 5023 RM, (Fig. 7a-c). This is a complete crown with an incomplete mesial and nearly complete distal root. Wear is moderate, with the cusps reduced to a nearly f1at plane. The crown has a rectangular occlusal outline. The five principal cusps are well developed. There is no evidence of the presence of a tuberculum intermedium, and the presence of a tuberculum sextum cannot be ascertained because of wear. The metaconid is the largest cusp, followed by the protoconid and hypoconid, which appear to be of nearly equivalent size. The entoconid is margínally smaller than the hypoconid, and the hypoconulid is the smallest. The metaconid contacts the hypoconid, and the lingual groove is situated distal to lhe level of the mesíobuccal groove, resulting in a Y occlusal pattern. The fovea posterior is a small, relalively deep pito The mesiobuccal groove is interrupted at the occlusobuccal margin, and the buccal grooves termínate abruptly. The end of the mesiobuccal is enclosed by a low, comparatively thick enamel wall. The mesial and disIal crown surfaces display carious lesions (Grine, Gwinnett and Oaks, 1990). The root neck is 2,5 mm high buccally. The mesial rool plate has separate buccal and lingual canals; the distal root has a single canal.

a

b

SKX 5024 LM? (No i1lustration). This is a small piece of a heavily worn crown with a narrow root fragment. The crown consists of a C-shaped enamel band around a deeply concave occlusal dentine exposure. The mesia! face preserved the lingual end of an interproximal contact facet. SKX 6013 Re (No iJlustration). This is an incomplete crown and incomplete rool. The distolingual portion of the crown and root are missing. Wear is slight, with the crown apex reduced to a flattened plane; a small dentine exposure is presenl. A faint, vertical buccal furrow is present near the distal margin, and this ís bounded distally by a slight enamel ridge. The mesial half of the face shows a small, barely perceptible depression. The slight lingual cervical prominence is skewed distally. There is no tubercle. A weak mesial marginal ridge is separated from a slight median ridge by a very shallow furrow.

c

1

SKX 6277 LM (No illustration). This is the mesiolingual corner of a maxillary permanent molar crown (considered to represent an M') with the mesial portio n of the lingual root and a ver¡ short mesiobuccal root stump. Much of the moderately worn protocone is preserved. Three small, faint depressions near the occlusal margin of the protocone represent the Carabelli trail. SKX 7761 Rp3 (No illustration). This is the buccal half 01 the crown with a small fragment of the distal face 01 the root. Wear is very slight. The paracone is well developed. The moderale mesial and distal marginal ridges each support two small cuspules; the mesial accessory cuspules are slightly larger than the distal. A faint distobuccal furrow is bounded distally

by a slight enamel ridge. Two buccal roots were present. 1

SKX 16060 Rdi (Fig. 8). This is a nearly complete crown and rool. Wear is moderate, and the mesial contact facet exposes dentine. The moderate lingual cervical prominence is skewed distally; there is no tubercle. A moderately broad, low median

NEW CRANIOOENTAL HOMINIO REMAINS

80

Fig.9

Fig.8 SKX 16060, lingual

SKX 19031, lingual

view; sca/e

view; scale

= 1 cm.

ridge rises Irom lhe apex of the cervical prominence. The mesial marginal ridge appears to have been weak; the distal margina! ridge is slight, and a very shallow V-shaped hollow separates it from the median ridge. There is no lingual shovelling.

SKX 19031 RI' (Fig. 9). This is a nearly complete crown with most 01 the root. The cervical enamel margin is damaged buccally. Wear is slight. The slight lingual cervical prominence is symmetrically posi1ioned; there is no tubercle. The mesial and distal marginal ridges are slight; the median ridge is very weak. Lingual shovelling is slight. SKX 21204 Mandibular corpus (Fig. 10a-b). This is a right mandibular corpus of a juvenile individual with the erupting 12 and developing P4 crown exposed. The mandibular symphysis is broken at the midline, and a short segment of the 12 root is exposed on this face. Posterior/y, the lateral surface of the body is preserved to the mesial root alveolus of the M1, while the internar surface is preserved to an oblique line that runs from the dm2 mesial root to below the M1 mesial root. The alveolar margin is damaged, especially along the outer borders of the incisor and canine sockets. Developmentally, the M, and h clearly had erupted, and the tip 01 the unworn 12 was emergent just beyond the level 01 the intact lingual alveolar margin of the 1, socket. X-rays reveal a very short root on the P 4, and a slighlly longer root appears to have been developed on the P3. The crown appears to have been incompletely formed at time of death.

e

Mandible. There is a very weak mental protuberance. A single mental foramen opens superiorly and slightly posteriorly. The shallow digastric furrow extends to below the level 01 the de from a point some 3,0 mm from the symphyseal midljne. Internally, the postincisive planum runs at 45° to the horizontal

Eat4-'--~ ,_~~_.z

=1 cm.

for some 10,0 mm before turning over a slight superior transverse torus. The genial depression is faint. An inferior transverse torus is not present. The symphysis measures 23,4 mm in height and 16,3 mm in thickness Irom a line tangent to the postincisive alveolar planum. Dimensions of the corpus (mm) are as follows:

di,/I, de dm, dm2

Breadth

Height

13,4 14,0 14,6 (14,4)

23,2 23,2 (23,1) (21,5)

R12 . The intact, unworn crown is partially exposed. The incisal margin has six small mammeHons. The mesial corner is strongly angled. The distal hall of the incisal margin has a pronounced cervical slope, and the distal corner forms a broadly obtuse angle. The /ingual cervical prominence appears to be symmetrically disposed; lhere is no lubercle. The marginal ridges are lacking, and there is no shovelling.

RP4. The occlusal aspect and portions ofthe buccal and lingual crown surfaces are visible. The crown has an ovotrapezoid occlusal outline; the mesiolingual corner is reduced and the crown is BL broader across the distal moiety. The protoconid ¡s well developed, while the metaconid is notably reduced and sítuated mesial to the level of the protoconidal apex. The strong mesia! marginal ridge is complete; its lingual end arises directly from the metaconid tip, while its buccal end terminates in a

NEW CRANIODENTAL HOMINID REMAINS

81

a

a

b Fig.10 SKX 21204, lateral view (a), occlusal view (b); scale "" 1 cm.

small accessory cuspulid. The large lovea anterior is enclosed distally by a thin, short crest from the metaconid and a larger crest from the protoconid. The fovea posterior is a broad, deep lransverse furrow that is enclosed by a comparatively low distal marginal ridge, the lingual end of which is partially separaled from the metaconid by a broad, shallow cleft. It supports five small accessory cuspulids along its Jength. Moderate mesiobuccal and distobuccal furrows are bounded by marginal ridges that terminate as the accessory cuspulids.

b Fig.11 SKW 12, lateral view (a), occlusal view (b); scale = 1 cm.

SKW 12 Left maxilla (Fig. 11a, b). This is an incomplete left maxilla with the poorly preserved crowns of p3_ M3, and part of the root socket. Laterally, the alveolar process is preserved from the level of the M 2 distobuccal root to the Q_ p 3 inter-radicular septum; lhe inferior face of the rool ot the zygomatic arch is preserved, as is a narrow segmenl 01 the lateral part of the floor of Ihe nasal cavity and a small section socket. of the naso-alveolar clivus superomedial to the 1 Medially, the alveolar wall is preserved opposíle the M and 2 M , and a very small segment of the horizontal palatal process of the maxilla is evident opposite the M 2 . A large triangle of bone is cemented against the alveolar process. The tooth crowns are very poorly preserved, being both heavilyworn and damaged. The buccal enamel face is missing from each of the crowns, and enamel is missing also from the mesial and/or distal aspects of lhe leeth; only the broken 3 mesial moiety of the M is present.

º

º

Maxi/la. The lateral portion of the nasal floor is smoolhly conlinuous wilh Ihe naso-alveolar clivus. lmmedialely inferior 10 Ihe lateral nasal margin and medial to Ihe socket, the naso-alveolar clivus faces anteromedially; it comes lo socket. The maxillary face turns face anleriorly over lhe posteriorly around the distal aspect of the socket to face late rally, where a broad and very shallow hollow is developed belween the canine eminence and the antero-inferior face ot the zygomatic root. This ascends al the level of the distal face 01 the p4, and il takes origin 14,0 mm above lhe level of the alveolar margin, which shows considerable resorption.

º

º

º

Oentition. The tooth crowns are heavily worn and damaged; morphological delail has been obliteraled, and meaningful crown dimensions cannol be recorded. It is evident from the

82

NEW CRANIODENTAL HOMINID REMAINS

exposed roots, as weu as from X-rays, that the p3 and p4~ possess two buccaí roots.

TAXONOMIC ATTRIBUTIONS Speeimen SKX 21204 has been attributed to Horno on the basís 01 its corpus size and morphology, the morphology ot the P, and the size and morphology olthe 1, (Grine, 1989). 01 the remaining specimens that can be diagnosed. al! are attributable to Paranthropus.

Table 2 Hominid tossils excavated

Specimen

SKX 2003 SKX 7992

trom Swartkrans Member 1-2 interface Excavation coordinates

E5-N2 E4-NS

250-270 cm 660-670 cm

MEMBER 1-2 INTERFACE SPECIMENS

This collection originally comprised nine isolated teeth, 01 which six are representad by incomplete fragments ar developing germs (Grine, 1989: table 2). Exeept lar SKX 3601, all were identified by me from storeroom boxes containing unidentified mammalian fossils. As noted above, SKX 3601 is undoubtedly assoeiated with 2 specimens from Member 1. SKX 308, an isolated R1 , is very likely associated with Member 2 specimen SKX 162, a maxnlary Iragment with the erowns 01 the Re-RP'. Not only do they derive from the sama grid coordinate and spit depth, but SKX 308 lits eomfortably within the open l' soeket 01 SKX 162 (Grine, 1989). Three Member 1-2 interface speeimens - SKX 2354 (L1,) , SKX 2355 (L1,) and SKX 2356 (Re) - probably denved Irom a single juvenile individual to judge trorn their deve'opmental status, as wen as their preservatian and provenance (Grine, 1989). SKX 1756 (t.om') may be Irom this same speeimen. It is possible that these tour isolated teeth are assaciated with the SKX 257/258 antimeres Irom Member 2. AII derive trom adjacent grid coordinates and depths, and they are compatible as the dental remains of a juvenile individual whose early develaping permanent teeth had nat yet attained complete crown farmation at the time ot death. They are, moreover, individually attributable to Homo (Grine, 1989), whieh is relatively uncomman in the Swartkrans cave deposíts. Finally, SKX 2671 lrom the Member 1-2 interface was identified by me as comprising the mesiolingual crown comer and short segment of the mestal face of the lingual root of a Rdm' (Grine, 1989: 424), and it was argued thatthis speeimen probably belonged to the same individual as representad by the Member 2 composite juvenile specimen SKX 267, 268 and 269 beca use all ofthe teeth derived from the same prcvenance (Grine, 1989: 444). Following the eompletion 01 that paper, specimens SKX 2671 and 267 were eleaned further and found to join perfeetiy. Thus, of the nine specimens originally catalogued as having derived trom the Member 1-2 interface, only two (SKX 2003 and SKX 7992) cannot be attributed with reasonable assurity to either Member 1 or Member 2 on the basis of individual assoeiation (Table 2).

SKX 2003 Rdi' (No iIIustration). This is a nearly complete crown and root. Wear is heavy. The slight lingual cervical eminence is skewed distally; there is no tubercle. Cervical remnants of the marginal ridges and a bread median ridge are preserved. The root is estimated to have been 20,3 mm long. SKX 7992 RI, (No illustration). This is an ineomplete erown

with a nearly complete root. Viewed buccally, the mesial and distal crown edges are strongly tapered cervically. A very shallow V-shapec depression is bounded mesially anddistally by bread, slfght ridges. The slight lingually cervical prominence is skewed slightly distally. There is no tubercle. There is a weak rnesial marginal ridge and a short, very faint median ridge. TAXONOMIC ATTRIBUTIONS Morphologically and metrically these twe incisors can be aceommodated eomlortably within the Paranfhropus hypodlgm.

MEMBER 2 SPECIMENS The 32 numbered specimens originally cata/agued as cornprising this sampte are considered to represent between 21 and 24 individuals (Brain el al., 1988). Five speeimens Ihat were catalegued provlslcnalty as having come from the Member 1-2 interface are likely associated with two of these individuals. and one tooth Iragment (SKX 2671) deseribed from the Member 1-2 tntertace ñts onto SKX 267 from Member 2. Thus, at least ñve, and possibly six camposite individual s can be reeognized: 1) SKX 162, 163 and 308 Iikely represent a single individual - SKX 1788 was noted by Grine (1989) as possibly being associated with these specimens, but the presence of a faint distal contact facet on the SKX 1788 U 2 makes this extremely unlikely in view 01the laet that the C. 01SKX 162 is still enelosed withln i1s erypt; 2) SKX 240 ano 242 are antimeres; 3) SKX 257 and 258 are antimeres, and they are possibiy assocíated with SKX 1765,2354,2355 and 2356; 4) SKX 267 (= 267 + 2671) and 268 are delinitely assoeiated, and they are very probably Iinked to SKX 269; and 5) SKX 31 O and 312 are antimeres. Finally, SKX 311 and 1016 are antimeres; they may be assoeiated with SKX 313 and SKX 1017, and all lour may be assoeiated with SKX 310/312. The Member 2 fossi/s are Iisted in Table 3. SKX 162 right maxilla with Re_p' (Fig. 12a-b). This consísts of the anterior alveolar and palatal aspects of a right juvenile maxilla with the complete, unerupted crowns of the RQ, Rp3 and RP4.The maxilla preserves the midline and patatal surface to the level ot the MI, and the bane over the sockets of the 1', l' and oc. The soekets tor theseteeth, the lingual root 01 the 1 dm 1, parts of the lingual roat eockete of the dm 2 and M , and portions of the crypts of the developing Q, p:J and p4 are preserved. Developmentally, the 1', l' and M' had erupted (at least Ihe crowns had emerged beyond the level of the alveolarmargin); the incisor root sockets are e/ongate and fully or nearly fully

NEW CRANIOOENTAL HOMINIO REMAINS

,Q.. This is a complete crown with part of the developing root.

Table 3 Hominid fossils excavated from Swartkrans Member .2

Excavaticn coordinates

Specimen

SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX

162 163 240 241 242 257 258 265 267 268 269 271 272 308' 310 311 312 313 333 334 339 341 610 1016 1017 1312 1313 1437b 1756' 1788 2003' 2354' 2355' 2356' 35S9 4446 5013

~originalJy

maxilla 2 Rdm

150-160 cm 150-160 cm 100-150 cm

E5-S2 E5-S2 E7-S2 E7-S2 E7-S2 ES-N4 E5-N4 E6-N1 E5-N4 E5-N4 E5-N4 E5-N3 E5-N3 E6-S2 E6-S2 E6-S2 E6-S2 E6-S2 E5-N2 ES-N3 E5-N2 ES-N2 E2-N3 E6-S3 E6-S3 E6-N2 E6-N2 E4-N3 E4-N3 E7-S2 E5-N2 ES-N3 E5-N3 E5-N3 E4-N4 E5-S8 E2-N9

L1'

Re RI' RM, LM, maxilla Rdm 2 RM' RQ

L1' frapment RI RQ

LP, LQ

L1' RM, RM' RI' RP1 RI'

RP,

L1, LM'

L1, LP' Ldm'

L1' Rd;l

L1, L1, RC RI,

mandible mandible

100-150 cm 100-150 cm 375-400 cm 375-400 cm 100-125 cm 300-325 cm 300-325 cm 300-325 cm 350-375 cm 350-375 cm 150-20D cm 150-200 cm 150-200 cm 150-200 cm 150-200 cm 140 cm 300-325 cm 150-175 cm 150-175 cm 400-450 cm 150-175 cm 150-175 cm 175-200 cm 175-200 cm 500-550 cm 400-450 cm 100-150 cm 250-270 cm 350-375 cm 350-375 cm 350-375 cm 510-520 cm 15 cm aboye b 1120 cm

catalogued as having deriven trom the Member 1-2 inter-

face; placed in Member2 through individual assocíation.

tormed. The de, dm' and dm 2 alveoli indicate that these deciduous teeth were probably still retained at the time of death. The crowns of the Q. and p4 are situated wen aoove the leve! of the preserved alveolar margin, whereas the cusps of the p3 are within 1,5 mm of ttiis margino 1

Maxi1la. The anterior surlace is slightly convex overthe 1 root, 2

slightly concave over the 1 root, and convex over the ce root and the anterior aspect of the underlying p3 crypt. The midline interalveolar septum is 18,7 mm from a line tangent to the anterior alveolar surtace to the anterior margin of the incisive canal. The palate is smooth and ñattened anteriorly; there Is no shelving_ To the level of the dm 2 there is a slight, narrow ridge along the midline, indicative of incipient development of gubernacular a palatal tarus. Anteriorly, the opening of the canal is presenl. The palatal and medial surlaces preserve remnants of the premaxillary-maxillary suture sorne 3,0 mm behind the incisar sockets.

e.

83

Viewed buccally, the crown has a symmetrical, ovoid outline. The tip is centrally positioned and the mesial and distal edges course cervically at about the same angle to the vertical. The mestal edge ts somewhat shorter than the distal; tne mestal comer is slightly more pronounced. There are two slight mesiobuccal furrows. The slight lingual cervical prominence is symmetrically positioned; there ts no tubercle. The moderate mestal marginal ridge is bounded by a curved furrow that extends from the mesial comer to the cervical eminence. The median ridge is barely perceptible. The weak distal marginal ridge is bounded by a shallow depression. The mesial and distal marqinal ridges are continuous with the cervical eminence.

p3. This is a nearly complete crown wíth a short segment of the developing root. The occlusal crown outline ls ovoid. The principal cusps are well developed and nearly equal in size. The paracone is symmetrically positioned. The protocone is sttuateo mesial to the level of the paracone apex; the protocone tip appears mesially tllted. The complete, strong mestal marginal ridge runs from the protocone to the mestal side of the paracone: its buceal end supports a small accessory cuspule. The fovea anterior ís a short buccal fissure that artses from the rnesiat end ot the longitudinal fissure. The distal marginal ridge ls complete, and longer than the rnesial. lts lingual end is comparatively thin, there is a srnall accessory cuspule in its middle, and the buccal end has a smaller "cuspule. The fovea posterior is a transversa fissure that is continuous with the longitudinal fissure. p4. This is a complete crown with a short segment of the developing root. The crown has an irregular ovoid occlusal outline. Bcth principal cusps are well developed; the protocone is larger than the paracone, and their apices occupy the same midcrown transverse planeo The strong mesial marginal ridge is complete. The fovea anterior is a short buccal furrow that ís continuous with the longitudinal fissure. The distal marginal ridge is complete; its buceal half is well developed, while the lingual half is thinner and lower. A moderate accessory cuspule is supported at its bucea/ end, a smaller euspule is present in the centre, and the thin lingual portion of the ridge supports two small cuspules. The fovea posterior is a narrow lingual fissure that is continuous wlth the large central basin. SKX 163 and SKX 308 are considered to represent the sama individual as SKX 162. While SKX 308 fits comfortably into the 2 open 1 socket of SKX 162, there is no contact between SKX 163 and the other specimens. They are described here separately. SKX 163 Rdm' (Fig. 12b). This is the mesial hall 01 a crown lacking the root. The crown is broken through the mestal maieties ot the metacone and hypocane. Wear is moderate to heavy; the pratocone and masiar marginal ridge are worn to a flat horizontal plane, with only slightly higher paracone relief. The protocone has a moderate, concave dentine exposure; a smaller expasure is present on the paracone. The protacone is largerthan the paracone. The mesial marginal ridge appears to have been moderately developed and complete; it ls

NEW CAANIODENTAL HOMINID AEMAINS

84

b

a Fig.12

SKX 162 wiCh premolars in situ (a), SKX 162.Q., p3 and p' with SKX 163 dm 2 and SKX 308 12 (b); scales = 1 cm.

restrieted to the bueeal half of the erown, where it arises from the mesiobueeal aspeet of the paraeone to beeome eonlluent with the mesial aspeet of the protoeone. The fovea anterior is restricted to the buccal half of the crown; it is enclosed distally by an epicrista that courses Irom the paraeone to the protocone. The buccal limb of the trigon basin is interrupted at the oeclusobuccal margin by an enamel wall. The distal trígon crest is complete. The buceal groove terminates abruptly on an enamel shelf. The Carabelli trait is represented by an L-shaped fissure. SKX 308 RI 2 (Fig. 12b). This is a complete, unworn crown wíth a short segment 01 the root. The buccal crown outline is somewhat trapezoidal, and a single small mammellon is situated just mesial 01 the midcrown axis. The mesial comer is strongly angled, while distal to the mammellon the edge follows a slightly curved course such that the distal comer is loeated cervical to the level 01 the mesial comer. The mesial half of the buccal tace has a low ridge. The slight lingual cervical eminence is symmetrically positioned; there is no tubercle or median ridge. The slight mesial and moderate distal marginal ridges are eonfluent cervieally; they bound a broad, shallow fossa. Shovelling is slight. 1

SKX 240/242 l/1 and RI (Fig. 13). These are the slightly damaged erowns of antimeres. SKX 240 preserves a segment

Fíg.13 SKX 240/242, lingual view; scale

= 1 cm.

"

NEW CRANIODENTAL HOMINID REMAINS

85

Fig.14 SKX 257/258 in occlusal, 1756 in mesial, 2354, 2355 and 2356 in lingual views; scale = 1 cm.

of !he root; !he SKX 242 root is broken at the cervix. Wear is slight. The slight lingual cervical prominence is symmetrically disposed; there is no tubercle. The slight mesial marginal ridge and somewhat stronger distal marginal ridge are separa!ed by shallow depressions from a weak median ridge.

Re

(No illustration). This is a nearly complete crown SKX 241 with mostol the root. Wear is moderate; the cusp tip is reduced to a large, ftat face! with a marked distal bevel. Buccally, the mesio-occlusal corner preserves the remnant 01 a shallow vertical furrow that appears to ha ve been bounded mesially by a slight enamel ridge. The moderate lingual cervical prominence is skewed distally; there is no tubercle. A narrow but prominent median ridge is present. The slight mesial marginal ridge ís separated from it by a broad, shaJlow depression; the moderate distal marginal ridge is separated from the median by a smaller, but deeper V-shaped hollow. The lingual surface is no! shovelled. SKX 257/258 RM 1 and LM1 (Fig. 14). These are the poorly preserved. developing crowns of antimeres. Root formation had not yet been initiated, although crqwn formation appears to have been nearly complete at the time of death. The crown has an ovorectangular outline with a rather prominent distobuccai bulge below the hypoconid. AII live principal cusps are well developed. The metaconid is the largest cusp, 101lowed closely by !he protoconid; the smaller hypoconid, entoconid and hypoconulid are nearly equivalent in size. The distal

.9.5S.Q

marginal ridge is well developed. It supports a barely perceptible tuberculum sextum on the left; this is not present onCthe right. The well-developed distal ridge of the metaconid forms a distinct postmetaconulid (= tuberculum intermedium). There is no mesial marginal ridge. The lovea anterior is a broad, triangular basin that opens mesially and is continuous distally with the broad. deep longitudinal trigonid lissure. The Iloor 01 the lovea anterior is slightly elevated on the left by a low transverse ridge that supports three small tubercles, and on the right by a small cuspulid. The end 01 the well-developed principal metaconid ridge shows slight distal deflection and incipient cuspulid development. The metaconid contacts the hypoconid; the lingual groove is situated well distal to the level 01 the mesiobuccal, resulting in a symmetrical Y occlusal patterns. The fovea posterior is a deep groove that is enclosed mesially by a high. narrow postentocristid. the apex of which is shallowly incised. Secondary ridges on the entoconid, hypoconulid and metaconid impart a somewhat crenulate appearance. The mesiobuccal groove ends in a small pit that is enclosed by a prominent enamel wall. On the left a faint, oblique furrow crosses the protoconid to end in a small pit 3,0 mm me sial to the mesiobuccal groove. A faint enamel ridge delineates the mesial aspect of the short oblique furrow. (This regíon is mjssing Irom the right tooth.) The distobuccal groove ends gradually. SKX 2354, SKX 2355, SKX 2356 and SKX 1756 are considered lo represent a single juvenil e individual, and it is possible

NEW CRANIODENTAL HOMINID REMAINS

86

'.

b

a

c

d

Fig.15 SKX 265 'In superior (a), inferior (b), facial (e), and medial (d) víews; seale = 1 cm.

that they derive from the same individual as is represented by the SKX 257/258 antimeres. Because this association cannot be demonstrated for certain, they are described separately. SKX 1756 Ldm' (Fig. 14). This is a small fragment o( the mesiobuccal crown corner with a nearly complete mesiobuccal root. Part of the paracone, the mesial marginal ridge and a small portion of the protocone are preserved. Wear has reduced the paracone and mesial marginal ridge to a nearly flat plane. The thick mesial marginal ridge is complete; it courses mesially tor a short distance trom the paracone before turning sharply Iingually toward the protocone. The tovea anterior is restricted to the buccal side of the crown; its lingual end appears to ha ve been conlinuous with the longitudinal trigon tissure. The tuberculum molar is slight, and the mesiolingual aspect of the protocone shows a number of small pits that probably represent the Carabelli trait.

SKX 2354 L12 (Fig. 14). This is an incompletely developed crown. Three mammellons are present. The distal third of the incisal edge slopes cervically so that the obtuse distal corner is situated cervical lo the level of the mesial cornero The lingual surface is slightly concave; there is no dislínct marginal or median ridge development. SKX 2355 L1, (Fig. 14). This is an incompletely developed crown. Six small mammellons are present. The incisal edge is nearly horizontal, and the me sial and distal crown corners are strongly angled. The lingual suriace is faintly concave, but there is no distinct marginal or median ridge development.

Re

SKX 2356 (Fig. 14). This is a poorly preserved, incompletely developed croWn. The apex has a sharp tip; the mesial crown COrner is obtuse, and the distal edge has a strong cervical slope. A median ridge and moderate mesial marginal

NEW CRANIODENTAL HOMINID REMAINS

87

ridge are preserved; the distal marginal ridge is incomplete. SKX 265 Left maxilla (Fig. 15a-d). Th;s ;s an edenlulous, left maxilla that preserves the palate and floor of the nasal cavity 1 to the level ofthe M and the sockets of l' to M'. The specimen ís broken along the intermaxillary suture. The lateral wall of the pyriform aperture projects just above the nasal lloor. The medíal part of (he nasal cavity floor curves gently ¡nto the incisive fossa, anteríor to which it extends horizontally to a small, superiorly projecting anterior nasal spine. The anlerior portion of the lateral part of the nasal cavity 1I00r forms a wide, flat platform to which the naso-alveolar c1;vus is rather sharply angled. The anterior margin of the nasal floor is delineated by a low, broad rídge that runs from the anterior nasal spine to lhe lateral nasal margino The spine is situated in approximately lhe same coronal plane as the anterior suriace of the lateral nasal margino The naso-alveolar c1ivus is virtually flal Irom the midline lo the ridge over the anterolateral aspect of the Q. root. The clivus is inclined at about 40 0 to the horizontal; il has a slighl posterolateral slope such that the anterior Q. socket 1 margin lies 3,5 mm behind the level 01 the anterior , socket rimo A sharp anterolaterally projecting ridge of bone is present over the mot. The alveolar process maxilla turns sharply posteriorly at the ridge. The palate is tlat anteriorly, and there is no palatine torus. The 1', 12 and sockets are aligned with a slight distal curve. The p3 alveolus reveals two roots. The p4 also has two roots; the socket reveals an incipient division of the buccal root into larger me sial and smaller distal parts.

e.

e.

SKX 267/268/269 Rdm 2 , RM', Re. (Fig. 16). This composite specimen consists of three isolated teeth that derive from the same grid square and depth. The interproximal contact Facets on the dm 2 and M' match in size, shape and configuration. Q. root formation had just been inítiated at the time of death; thus, the Q. is developmentally compatible with the slightly worn molars. 2

Rdm (SKX 267). This is a nearly complete crown with much of the root. Wear is slight, with enamel facets on all cusps, and a small dentine exposure on the protocone. The occlusal crown outline is nearly square. The four principal cusps are present; the protocone is the largest, and the paracone is slightly larger than the metacone and hypocone, which are nearly of equal size. The mesial marginal ridge is moderate, and the fovea anterior is a narrow transverse groove walled distally by a moderate epicrista. The buccal limb oF the trigon basin is interrupted at the occJusobuccal margin, and the thick distal trigon crest is complete. The well-deveJoped dístal marginal ridge ís complete; it supports two incipient accessory cuspules. The buccal end of the distal marginal ridge is thin, and situated well below the level oF the metacone apex. The fovea posterior is a deep, narrow transverse fissure. The slight buccal groove ends gradually. The Carabelli trait is represented by a narrow but deep transverse fissure that is bounded mesially by an enamel ridge. The root neck is 1,2 mm high buccally. I

RM (SKX 268). This ís a complete crown with segments of the roots. Wear is very slight. The crown has a rectangular, MD elongate occlusal outline. Al! four principal cusps are well

==..:.",."""',

Fig.16 SKX 267, 268 and 269 composite in occlusal view; scale

~

1 cm.

developed. The protocone is the largest, followed by the metacone; the slightly smaller paracone and hypocone are nearly equal in síze. The well-developed mesial marginal rídge is complete; its lingual end is 'waísted' between the Fovea anterior and the Carabelli fissure. The fovea anterior ís a long, broad transverse Furrow lhat is enclosed by confluenllransverse accessory ridges From the paracone and protocone. The trigon basin is deep and broad, The distal trigon cresl is broad, low and complete; a shalJow físsure courses over its apex to partially delineate an incipient cuspule on the protoconal halF ot the ridge. The distal marginal ridge ís well developed and complete; it supports a small cuspule near its buccal end, whích is located well below lhe level of the metacone apex. The fovea posterior is a broad transverse furrow. The buccal groove is barely perceptible. A prominent, vertical Cara.belli furrow is bounded mesially by a moderate enamel ridge. The root neck is 2,6 mm high buccally. R{2 (SKX 269). This is the mesíal halt of an unworn crown. Buccally, a very shallow triangular depression is present near the mesial edge. This depression is bounded mesially by a barely perceptible ridge that terminates al the mesial crown cornero The mesial edge is strongly inclined to the horizontal. There is no evidence of a lingual tubercle. A prominent median ridge courses vertically from the remnant of the cervical eminence to the cuspal tipo A slight mesial marginal ridge ;s separated from the median by a shallow triangular depression.

SKX 271 L1' (Fig. 17). This ís a complete crown with much of the rool. Wear is moderate. The moderate lingual cervical eminence is symmetrically disposed; there is no tubercle. A short, moderately elevated mesial marginal ridge is separated from the prominent median ridge by a small V-shaped hollow. The distal marginal ridge ís stronger than the mesíal; it is separated from the median ridge by a large, shallow depression. Lingual shovellíng is slight. SKX 272 Tooth fragment (No illustration). This is a narrow enamel wedge with a short segment of root of a permanenl tooth. Jt most likely represents part of the buccal Face 01 a worn premolar.

---------------------------------~_.-

. _._ -

aa

NEW CRANIODENTAL HOMINID REMAINS

Fig.17 SKX 271, lingual view; scale 1 cm.

=

SKX 310/312, 311/1016, 313 R + Le., L + RP3. L( These five specimens possibly derive from a single individual on the basis 01 provenance and wear. Because this association cannot be established with certainty the anlimeres and the incisor crown are treated separately. SKX 310/312 R and Le. (No illustration). These are the nearly complete crown and root of a RQ (SKX 310), and lhe nearly complete crown and incomplete root 01 a Le. (SKX 312). Wear is heavy, and the crown is reduced to a Ilat, distally bevelled suriace. The slight lingual cervical prominence appears to be slighUy skewed distally. The cervical end 01 the distal marginal ridge is preserved. SKX 311/1016 L and RP3 (No illustration). These are the incomplete crowns and complete roots of a LP3 (SKX 311) and RP3 (SKX 1016). Wear is heavy, and the crown is reduced to a (Iat plane with a slight distal bevel. A large dentine exposure on lhe proloconid is confluent with one in the centre 01 the mesial moiety. The crown has an ovorectangular occlusal outline. It is evident thal the protoconid was larger than the metaconid. There is a slight depression in the mesio-occlusal comer 01 the buccal face and a barely perceptible depression in the disto-occlusal comer. SKX 313 L1' (No illustration). This is a slíghtly damaged crown with much of the root. Wear is very heavy. with considerable reduction in crown height. The slight lingual cervical prominence is skewed distally. SKX 333 RM 1 (No illustration). Thís is a triangular crown fragment preserving much of the hypoconid and a short segment 01 the buccal root suriace. Wear on the hypoconid is slight. The mesiobuccal groove is interrupted at the occlusobuccal margin by a thick enamel wall between the protoconid and hypoconid. The deep, narrow mesiobuccal groove ends abruptly. The weak distobuccal groove terminates gradually. SKX 334 RM 1 (No illustration). This is an incomplete crown with smal1 segments 01 the lingual and mesial root suriaces.

Fig.18 SKX 339. lingual view; scale = 1 cm.

Wear is slight with fainl enamel facets on all cusp tips; dentine is not exposed. The lour principal cusps appear to have be en well developed. The moderate mesial marginal ridge is com-. plete. The fovea anterior is a narrow, centrally positioned transverse fissure; ít is enclosed distally by contiguous principal paracone and protocone crests. The well-developed distal trigon crest is complete. The lovea posterior is a large pit. The buccal end 01 the distal marginal ridge is moderately thick, but low; it supports a small accessory cuspule. and has a low conlluence with the metacone. A small but distinct Carabelli pit is presento and a slight, oblique enamel ridge runs cervical to il. SKX 339 R!' (Fig. 18). This is a nearly complete crown with a short root segment. Wear is slight. The weak lingual cervical eminence is symmetrically dísposed; there is no tubercle or median ridge. The short mesial marginal ridge is very weak; the longer distal marginal ridge is better developed. Shovelling is slight. SKX 341 Tooth fragment (No illustration). This is a small crown Iragment and part 01 a rool. It is interpreted as the mesiolingual comer of a permanent right mandibular premolar. SKX 61 ORI 2 (Fig. 19). This is a complete crown and rool. Wear is slight. Viewed buccally, the mesial and distal edges are stronglytapered some 4,0 mm from the incisal edge. The weak lingual cervical eminence is symmetrical1y disposed; there is no tubercle and no median ridge. The slight mesial and distal marginal ridges enclose a weakly concave lossa. The rool is 20,8 mm long.

------------_.---

-

NEW CRANIODENTAL HOMINID REMAINS

89

'~ot

exposed. Lingually, a short vertical furrow runs meslal to the principal groove. The Carabelli trait comprises two very laint, short vertical depressions near the occlusal margino

Fig.19 SKX 610, lingual view; scale

= 1 cm.

SKX 1313 U 2 (No ¡lIustration). This is an incomplete crown with a long segmentof rooL Wear is heavy, wilh a large dentine exposure. The slight lingual cervical eminence is symmetrieally disposed; there is no tubercle or median ridge. The mesial and distal marginal ridges appear to have been very weak. The lingual surface is ralher flaL SKX 1437a Lp3 (No illustration). This is a buccal erown fragment with a short segment 01 root. Most of the paracone is preserved; it is well developed, and symmetrieally posilioned between the mesial and distal edges. Three prominent ridges course lingually from its apex; the me sial is the strongest and follows a directly lingual course, while the two distal ridges pursue a dislolingual course. Small mesiobuccal and larger distobuccal accessory cuspules are present. The mesiooeelusal and disto-occlusal buceal corners have shallow furrows that course vertieally from the oeclusal ineisions between the paracone and accessory cuspules.

Fig.20 SKX 1017, lingual view; scale 1

=

cm.

SKX 1437b Lp 4 (No illustration). This is a small wedge from the distal side of the crown with a small segment of root. The fragment preserves a small portion of the distobueeal protocone ridge, and a short segment of the distal marginal ridge. The lingual end of the dista! marginal ridge supports a small aeeessory cuspule, and the lingual portion of a second, apparently larger cuspule is preserved. SKX 1788 L1 2 (Fig. 21). This is a nearly complete crown with a short segment of the rooL Wear is slight. Viewed bucea,IIY, the mesial corner is sharply angled; the distal corner is gently rounded. A faint median buceal eminenee is bounded by two shallow furrows lhat converge eervically from the mesial and distal corners; the furrows are bounded by low ridges. Língually, a faint median ridge is present; it becomes stronger incisally al the remnant of a central mammellon. A slight distal

SKX 1017 U2 (Fig. 20). This is a nearly complete crown with most of the rooL Wear is moderate, with a large dentine exposure. The slight lingual cervical prominence is skewed distally; there is no tubercle. The median and marginal ridges are lacking, and the lingual lace is virtually flat. Fig.21 1

SKX 1312 LM (No illustration). This is the lingual side of a erown with a short root slump. Wear is slight, and dentine is

±

SKX 1788, lingual view; scale = 1 cm.

NEW CRANIODENTAL HOMINID REMAINS

90

marginal ridge and very weak mesial marginal ridge are pre-,. sen1. There is no shovelling. SKX 3559 RI 1 (No illuslration). This is a complete crown and roo1. Wear is heavy, wilh considerable reduction of crown height. The slight lingual cervical eminence is skewed slíghtly distally; there is no tubercle or median ridge. The cervical portions of weak mesial and distal marginal ridges are preserved. The root is 17,2 mm long. SKX 4446 Right mandibular corpus with PCM2 (Fig. 22a-c). This is a nearly complete, subadult mandibular corpus preserving P4-M2, the P3 roots, incomplete 12 and e sockets, and the mesiaJ wall of the M3 crypl. Anteriorly the corpus is broken through the 12 alveolus but it approximates lhe internal midline in the region 01 the genial lossa. The corpus is fractured distally through the M 3 crypt, with only its mesial wall preserved. The ramus is broken off along a level wilh the alveolar margino

Mandible. The anlerior aspect forms a slraighl, nearly vertical profite in lateral view. A laint eminence just above the base in root represents a weak mental tuberde. The line with the base 01 the symphyseal region presents a moderate basal incisura and the remnanl ofwhat appears to have been a weak midline basal eminence. Laterally, the alveolar and basal margins are nearly paralleJ from P3 to M2. A single mental foramen opens posterosuperiorly al the level 01 the distal P3 rool. The mental foramen is situated jusi below midcorpus heigh1. The corpus inflates laterally atthe P4 , and the anterior margin 01 the ramus takes root at the level 01 the mesíal edge of M2. The extramolar sulcus is broad and moderately deep, extending back from the dislal half of the M1. The posterior aspect 01 the symphysis presents a postincisive planum that is slighlly hollowed in its transverse and vertical axes; it slopes posteroinferiorly for 16 mm before curving over the broad superior transverse torus, which is sítuated above midcorpus height. Viewed occlusally, the superior torus reaches the level 01 the mesial margin of the P 4. Below the torus the lateral wall 01 a deep genial pit is preserved. The inferior torus is tightly curved, extending posteriorly less than 3,0 mm beyond the superior lorus. Inferiorly, a broad digastric hollow tapers posteriorly to the level 01 the M1 . The digastric fossa is bounded anteriorly by a low ridge. The medial aspect of the corpus presents a broad swelling that extends from lhe superior transverse torus to the level 01 the M2. Below this, lhe medial lace slopes ínferolaterally without noticeable excavation. The symphysis is some 37,0 mm high from the estimated position of infradentale lo gnathion, and it is 26,7 mm thick from a line langent to Ihe postíncisive planum to the symphyseal base. The principal dimensions (mm) of the mandibular corpus are as follows:

a

e

8readth

Height

23,0 22,9 24,0 27,2

(37,5) est. 34,0 32,8 31,2

b

e Fig.22

SKX 4446 in lateral (a), medíal (b) and occlusal (e) views; scale = 1 cm.

P3· Only the damaged mesial root and part of lhe rool neck is exposed. X-rays reveal lhe presence of two well-separated roots. P4 . Wear ís slighl. The crown has an ovolrapezoid occlusai outline that is broader mesially lhan distally. Both principal cusps are well developed. The protoconid is somewhallarger

NEW CRANIODENTAL HOMINID REMAINS

91

lhan the metaconid, which is situated mesial 011he midcrown BL axis that bisects the protoconid. The well-developed mesial marginal ridge is complete; its buccal end forms a low, bulbous cuspulid. The fovea anterior is a deep but small basin that is completely enclosed dislally by a moderate crest between lhe distal aspect of lhe metaconid and the principal lingual ndge of the protoconid. The long distal marginal ridge is well developed and complete; ils buccal end is incised by a narrow, shallow fissure. The fovea posterior is long and broad. Buccally, the mesio-occlusal corner has a shallow, broad furrow that runs from between the protoconid and mesiobuccal cuspulid. A shorter furrow is present from between the protoconid and distal marginal ridge. MI. Wear is slight. The crown has a rectangular occlusal oulline. AII five principal cusps are well developed. The metaconid is the largest cusp; the protoconid, hypoconid and entoconid are nearly equivalent in size, and the hypoconulid is only slighlly smaller. The mesial marginal ridge is moderately welldeveloped and complete. The fovea anterior is a deep, narrow transverse fissure; il is continuous with the longitudinallrigonid fissure. There is no accessory trigonid lovea. The metaconid contacts lhe hypoconid; the lingual groove is siluated distal to lhe mesiobuccal, forming a symmetrical Y occJusal paltern. The mesiobuccal and dislobuccal grooves are partially interrupled by enamel walls between the buccal cusps, and the lingual groove is interrupled allhe occlusolingual margin by a comparalively ¡hin wall. There is no tuberculum intermedium. The distal marginal ridge is complete, but comparatively low; ils mesial tace is incised bylwo slight fissures but there is no clear development of a luberculum sextum. The tovea poste· rior is a narrow, short transverse fissure. The short, deep mesiobuccal groove lerminates in a pit lhat is enclosed by a low enamel wall. The short distobuccal groove ends abruptly on a narrow enamel shelt.

M2 . Wear is very slight. The crown has an ovorectangular occlusal outline. The five principal cusps are well developed. The metaconid is Ihe largest, tollowed by 1he proloconid, hypoconid, entoconid and hypoconulid in decreasing size order. The moderate mesial marginal ridge is complete. The lovea anterior comprises short buccal and longer lingual Jimbs thal radiate from the mesial end of a broad longiludinal tissure. The principal crest ofthe metaconid is 'waísled;' il contacts the hypoconid, and the mesiobuccal and lingual grooves lie in nearly the same transverse plane. An incipient postmetaconulid is present; lhere is a moderate tuberculum sextum. The fovea posterior has three distally direcled fissures, two of which delineate the tuberculum sextum. The principal ridges of the hypoconid and entoconid are 'waisled,' and the ends 01 these contiguous crests appear as incipient cuspulids. The lingual groove is interrupted at the occlusolingual margin by a moderate enamel wall, and the mesiobuccal and distobuccal grooves are partially interrupted in midcourse by thin, low enamel ridges belween the three buccal cusps. The crown has a slightly crenulate appearance due to ,he 'waisting' ot principal cuspal ridges and the presence of secondary ridges. The strong mesiobuccal groove ends in a pi! thal is partially enc10sed by a thin enamel wall. A protostylid is delineated (rom the hypoconid by an oblique fissure. The short distobuccal

.' ss

a

b

e Fig.23 SKX 5013 in lateral (a), medial (b) and oee/usal (e) views; sea le

=1 cm.

groove terminates abruptly on a narrow enamel shelf. SKX 5013 Left mandibular corpus with M, (Fig. 23a-e). This is a partialleft mandibular corpus with the distal root of the P4 • the intacl M" the Mz root neck and incomplete P 4 and M 3 mesial root sockets. The corpus is broken anteriorly at the mesial rool ot the P 4: the internal surface is missing at midcorpus height lo the M,. Posteriorly, the corpus is fractured through the rameal root and behind the M3 mesial root socket. Bone is missing Irom the lower half of the lateral surface at the level of the M3. The M3 mesial rool socket is fully lormed, indicating that thís tooth had erupted fully by time 01 death .

NEW CRANIODENTAL HOMINID REMAINS

92

Mandible. Laterally, the alveolar and basal margins are parallet below M , and M 2 . The lateral surtace is divided al midcorpus height by a slighl, triangular bulge that broadens posteriorly from below P4 to the anterior margin 01 M 2 • where the ramus takes root. The root 01 the ramus forms a distinct eminence: its anterior free margin arises at M3 level, and the broad, shallow extramolar sulcus runs from this point to the M2 rnesial rnarqín. The medial aspect 01 lhe corpus presents a broad swelling over the upper hall at the M, and M2; below lhis the medial face slopes inferolaterally. The principal dimensions (mm) 01 lhe corpus are as follows:

Breadth

Height

(22,7) 24,1 26,3

31,7 32,3 31,2

Table 4 Hominid tcssns excavated trom Swartkrans Member 3

Specímen

SKX 21841 SKX 19892 SKX 25296 SKX 26626 SKX 26967 SKX 27151 SKX 27524 SKX 26724 SKX 32162 SKX 32832 SKX 35025 SKX 35416

Excavation coordinates

RM' LM, RC LP'

L1, Ldm2

LI' RC RP, Rdm 2 RM~

RI,

W3-S3 W2-S5 W3-S6 W3-S3 W2-S5 W2-S4 W2-S5 W3-S3 W3-S2 W4-S2 W4-S2 W4-SJ

170-180 180-190 210-220 270-280

cm cm cm cm

330-340 cm 270-280 cm

280-290 cm 270-280 cm 280-290 cm 280-290 cm 450-460 cm 490-550 cm

specimens that are recognizable taxonorntcauy, are attributable lo Paranthropus.

MEMBER 3 SPECIMENS

P4. The distal wall 01the masial root socket and the distal root are preservad. x-rays reveal that the apical fourth 01the distal root has a very slight distal curve. M,. Wear is rather heavy with the cusps reducad to a flat horizontal plane, and a moderate dentine island is exposed on each cusp. The crown has a rectangular occlusal outline. The 1ive principal cusps are wen developed; relative size relationships cannot be ascertained due to wear. The metaconid contacts the hypoconid, and Ihe lingual groove ís disial lo lhe level 01the mesiobuccal with a symmetrical Y occlusal pattern. The mesiobuccal groove terminates abruptly on a short enamel shelf.

Twelve numbered specimens comprise thia sample. These are considered ro represent between 9 and 11 individuals. Specimens SKX 26625 and 28724 are Iikely associated on the basis 01provenance and developmental status, and these two incomplelely developed teelh may be associated with lhe SKX 32832 and 27151 deciduous molars (Grine, 1989), Because these assoclanons cannot be demonstrated with certainty, the specimens are treated individually. The Member 3 craniodental remains are listed in Table 4.

M3. The preserved mesial root socket indicates a robust píate with separata buccal and lingual apices. The mestal socket has a rather marked distal inclination.

SKX 19892 LM, (No illustration). This is an incomplele crown with part 01the mestal root and a nearly complete distal root, The mestal portions of the protoconid and metaconid are missing. Wear is heavy; dentine ls exposed on all cusps, whích are reduced to a nearly flat planeo lt IS evident that the five principal cusps were well developed. The metaconid contacts lhe hypoconid, and the lingual groove is situated disial lo lhe mesiobuccal groove resulting in a Y occlusal pattern. The root neck is 4,0 mm high.

TAXONOMIC ATIRIBUTIONS The SKX 257/258 M, anlimeres have been attribuled to Horno by Grine (1989) primarily on the basis 01 their síze, the presence 01 a postmetaconulid and the absence 01 a tuberculum sextum. As noted aboye, tnese antimeres are possibly associated wilh lhe proposed SKX 2354, 2355, 2356 and 1765 composite, and the latter group has been attributed to Horno on independenl entena (Grine, 1989), That is, SKX 2354 and 2355 are comparatively large with mammellate edges, and the 2356 crown displays a strong slope to its distal edqe. The SKX 267, 268 and 269 composite has been attributed also to Horno on the basis 01 occlusal morphology and the comparative shape oflhe molars (Grine, 1989), SKX 339 and SKX 61 Ohave been re1erred to Horno on the basis 01 their large sizes and weak developmenl 01 lingual reliel, and the SKX 334 M' has also been attributed to this taxon (Grine, 1989). The remaining

SKX 21841 RM 3 (Fig. 24), This is an unworn crown wilh 7,0 mm ot the developing root, which had nol formed beyond the level of bi1urcation. The crown has a trapezoid occlusal outline, with the buccal cusps placed rnesial to the lingual cusps. The paracone projects mesiobuccaHy, and the distobuceal comer is reduced. The tour principal cusps are well developed. The protocone is the largest, followed by the paracone, rnetacone and hypocone in decreasing size arder. Larga central and smaller buceal accessory cuspules form the mesial marginal ridge. The fovea anterior is a transverse fissure that is continuous with the longitudinal groove. The trigo n has a somewhat crenulate appearance. The buccal groove is uninterrupted at the occlusobuceal margino The moderate distal trigon crest is complete, although it is slightly 'waisted' at the metacone. The low but complete distal marginal ridge has a large accessory cuspule in its middle. The

M2. The exposed root neck ís preserved. X-rays reveal straight mesial and distal roots; the former ls slightly the longer.

NEW CRANIODENTAL HOMINID REMAINS

93

t~e midcrown axis. The mesia! and distal apical edges diverge al about the same angle to the horizontal; the mesíal edge is somewhat longer and the mesial crown corner is situated slightly cervical to the level 01 the more sharply angled distal corner. Cuspules are not presento The moderate lingual cervical eminence is symmetrically disposed; there is no tubercle. Moderate mesial and distal marginal ridges are present, and a broad, low median ridge courses to the cuspal apex. These ridges are separated by moderately deep c1efts. A prominenl accessory ridge runs obliquely Irom the distal corner into the hoflow between the distal marginal and median ridges. The root is 27,4 mm long.

SKX 26625/26724 Lp4 and Re.. These are complete, unworn crowns with incompletely developed roots. 80th are similarly preserved, and the amount of TOotformation suggests thal they derived from the same índividual. However, because this cannot be demonstrated with certainty, Ihey are treated individ ually.

Fig.24 SKX 21841, occlusal view; scale:: 1 cm.

fovea posterior has equal buccal and linguallimbs. The narrow buccal groove terminates gradually. The Carabelli trail is represented by a shallow, narrow furrow. SKX 25296 L.G. (Fig. 25). This is a complete crown and rooL Slíght wear has rounded the cusp tipo The crown has a pentagonoid buccal outline, with the apex set to the distal side 01

SKX 26625 Lp4 (Fig. 26). This is a complete, unworn crown wílh some 6,0 mm of roo1. Root development had not proceeded much beyond the prese rved margin; bifurcation 01 the buccal and lingual roots had been attained. The crown has an ovorectangular occlusal outline. 80th principal cusps are equally well-developed. The protocone is situated mesial 01 the midcrown transverse axis tha! bisects the paracone. The low, thick mesíal marginal ridge is complete; its buccal and lingual ends support incipient accessory cuspules. The fovea anterior has broad buccal and narrow lingual arms that radiate from Ihe broad longitudinal fissure. The principal protocone

Fi9·25 SKX 25296, lingual view;

Fig.26

scale:= 1 cm.

SKX 26625, occlusal view; scale :: 1 cm.

94

NEW CRANIODENTAL HOMINID REMAINS

ridge is 'waísted.' The distal marginal ridge is well developed~ it is thicker buccally than lingually, and incised at the base 01 the protocone. A large accessory cuspule is present at the buccal end 01 the distal marginal ridge; two smaller cuspules are present next to the buccal cuspule, and a small cuspule is present at the lingual end ot the ridge. The fovea posterior has shorter buccal and longer lingual Iimbs that radiate from the distal end of the longitudinal tissure. SKX 28724 Re. (Fig. 27). This is a complete, unworn crown with 14,5 mm ot root developed. The root margin ís slightly damaged. but very little of its original length has been lost. Viewed buccally, the crown has an irregular pentagonoid outline, with the típ situated just me sial of the midcrown transverse axis. The mesíal apical edge is longer and has a steeper slope than the distal edge. The mesial crown comer is a slight swelling close to the cervix; the distal comer is rounded. There is no mesial or distal cuspule. The lingual cervical eminence is slight; there is no tubercle. A broad median ridge and moderate mesial and distal marginal ridges are present. The mesial marginal and medían ridges are separated by a deep, narrow furrow; the distal marginal and median ridges are separated by a larger, shallow hollow.

Fig.27 SKX 28724, lingual view; scale := 1 cm.

Fíg.28 SKX 26967, lingual view; scale 1 cm.

:=

protoconid and hypoconid apices and mesial part of the hypoconulid are preserved. Wear is slighl. The mesiobuccal groove is interrupted at the occlusobuccal margin by a low enamel wall. The strong mesiobuccal groove terminates in a pit that is walled by a small protostylid lhat projects Irom lhe hypoconid tace. The deep and broad distobuccal groove ends abruptly. A barely perceptible, oblique lurrow is present on the mesial part 01 the buccal protoconid tace. The root neck is 2,0 mm high. SKX 27524 U' (No illustration). This is a complete crown with a substantial portion 01 the rool. Wear i$ moderate. The moderate lingual cervical eminence is symmetrically disposed; there is no tubercle. The me sial marginal ridge is virtually non-existenl; the distal marginal ridge is weak. Two barely perceptible 'median' ridges arise trom the cervical eminence. The lingual surface presents a broad, shallow tossa so that some shovelling is apparent despite the weak marginal ridges. Some 13,0 mm of root is preserved, and il appears that only about 0,5 mm has been losl.

SKX 26967 L/ 1 (Fig. 28). This is a complete crown and nearly complete root. Wear is moderate. The slight lingual cervical eminence is symmetrícally disposed; there is no tubercle. The mesial and distal marginal ridges are very slight and there ís a faint median ridge; lingual shovelling is not present. SKX 27151 Ldm2 (No illustration). This is the buccal side ot the crown with portions ot the mesial and distal roots. The

SKX 32162 RP4 (No illustratíon). This is an incomplete crown without the rool. The crown has a trapezoidal occlusal outline; the buccal tace of the protoconid projects beyond the level 01 the distobuccal cusp. Four distinct cusps are presento The protoconid and metaconid are well developed; the protoconid is the largest. The distal cusps are noticeably smaller, and of equivalent size. The metaconid tip ís set slightly mesial to the level of the protoconid tipo The mesial marginal rídge is high and complete; its lingual half is relatively thin, and its buccal end lorms a small accessory cuspulid. The middle 01 the mesial marginal ridge supports a second, small accessory cuspulid. The tovea anterior is a broad, centrally placed transverse fissure that is continuous with the talonid basin vía the broad longitudinal lurrow. The principal protoconíd ridge is bitíd; the stronger spur courses distolingually. The principal ridge of the metaconid has a strong distal dellection. The distal marginal ridge is thick and complete but low. The talonid basin is deep and capacious with short buccal and lingual limbs

NEW CRANIODENTAL HOMINID REMAINS

95

Buccally, a prominent V-shaped depression separates the protoconid and distobuccal cusp. A shallow vertical furrow runs from the occlusal incísion between the mesiobuccal cuspulid and protoconid.

. lingual cervical eminence is sym metricaJly disposed; there is no tubercle or median ridge. The distal marginal ridge is very slight. The root is 16,1 mm long, and a prisline length of 16,5 mm is Iikely.

SKX 32832 Rdm 2 (Fig. 29). This is an incompJete crown with the distobuccal root and most of the lingual rool. The mesial and buccal crown margins are missing. Wear is moderate; the cusps have been reduced to a nearly flat plane, and dentine is exposed on all four cusps. The crown likely had a square occlusal outline. AII four principal cusps are well developed. The protocone is the largest, followed by the paracone and metacone, which are 01 equal size; the hypocone is the smallesl. The buccal limb of the trigon basin is interrupted at the occlusobuccal margin by a thick enamel wall. The remnant of the distal trigon crest is moderately thick and complete, and the well-developed distal marginal ridge is complete. The fovea posterior has long buccal and short lingual limbs. The deep buccal groove terminales abruptly on a thick enamel shelf.

TAXONOMIC ATTRIBUTIONS Except for specimen SKX 35025, which cannot be assigned to a specific hominid taxon with any assurity because of its fragmentary nature, the entire Member 3 sample is attributable to Paranthropus.

MEMBER 1 HANGING REMNANT SPECIMENS

A total of 21 craniodental specimens comprises lhe recentJy recovered hominid sample Irom the Member 1 Hanging Remnant breccia. These specimens are listed in Table 5. There do not appear to be any individual associations among these specimens, nor can any be attributed with any certainty to fossils recovered earlier from the Member 1 Hanging Remnant breccia.

Table 5 Hominid fossils recently recovered from lhe Member 1 Hanging Remnant breccia

Specimen

Fig.29 SKX 32832, occlusal view; scale

= 1 cm.

Notes on Recovery

SK 14132 RM3 SK 14133 LP'-M' SKW 5 mandible Symphysis and corpora recovered separalely on 30 Oetober 1970 and 1January 1971 SKW6 LM 3 Recovered 6 November 1970. SKW 8 maxilla M' and M2 recovered in pie ce of breccia on 17 February 1971; maxilla found 12 Feb 1971 SKW 10 RM, Recovered 2 September 1971 SKW 11 cranium Recovered 22 March 1972 SKW 14 LM 2 Recovered 14 July 1970 SKW15 LP' SKW 29 cranium SKW 4767 RM, SKW 4768 LM 2 SKW 4769 LM2 SKW 4770 maxilla SKW 4771 ?RP, Found adjacent lo Ihe RP,-M, 01 SKW 5 SKW 4772 RP' Found aboul 4 cm above SKW 5 left corpus SKW 2581 temporal SKW 2637 enamel Found nexl lo lingual side al SKW 5 LM2 fra~ment

SKX 35025 RM? (No illustration). This is a fragment 01 the talonid of what may be a RM,. Most 01 the hypoconid, the distobuccal aspect of the metaconid and the mesiobuccal aspect of the entoconid are preserved. A short segmenl ot the buccal root neck is present below the hypoconid. Wear is slight to moderate; al! th ree cusps display well-developed enamel facets, but denline is nol exposed. The metaconid contacts the hypoconid and the lingual groove is set opposíte the hypoconid tip. The shallow distobuccal groove terminates gradually. SKX 35416 R 11 (No ¡lIustration). This is an incomplele crown with a nearly complete root. The mesial portions 01 the buccal and lingual faces are missing. Wear is moderate. The weak

SKW 2702 RM SKW 3033 Rp 3 SKW 3068 LM 2

3

SK 14132 RM (Fig. 30). This is the buccal side ot the crown wilh part of the rool. It preserves part of the metacone and Ihe buccal wall of the distal part 01 the paracone. The apex of the metacone is irregular, with a lurrow between the prominence and the buceal margin of the crown. The enamel along the oeelusobuccal margin is furrowed and pitted, and lhere is no discernible buccal groove. The preserved part of the crown is

NEW CRANIODENTAL HOMINID REMAINS

96

",

Fig.30 SK 14132, dislobuccal view; scale '" 1 cm.

comparable to the SK 13/14 homologue. The root neck is about 4,0 mm high. The mesiobuccal rool is unusual in torm; it has a slight distal tilt and a very strong buccal curvature, so that Ihe tip opens buccally. A short (5,0 mm long) accessory root emerges trom the mesial aspect ot the mesiobuccal root, and it also has a strong buccal curvature so that its patent apex opens buccally. SK 14133lp4 and M 1 (No illustration). This specimen consists of the damaged crowns of a lp4 and lM' in a small fragment of maxilla. The dístobuccal quadrant 01 the p4 crown is missing, as is the mesiobuccal rool and much of the distobuccal roo!. The M 1 crown is cracked, and it is missing a MD long wedge from the centre of the protocone and hypocone; small wedges of enamel also are missing from the mesiobuccal and distobuccal corners. Wear on the p4 is slight, with enamel facets 1 on both cusps. Wear on the M is moderate, with a nearly tlat occlusal tace and a large dentine exposure on the protocone. Smaller dentine exposures are present on the paracone and hypocone.

p4. The crown appears to have had an ovoid outline. The protocone is skewed mesially; Iingually it is more promínent mesially than distally. The protocone ¡s well developed, and the paracone appears to have been well developed also. The protocone is situated slightly mesial of the paracone apex. The moderately thick mesial marginal ridge is complete. The fovea anterior has short lingual and long buccal arms thal radiate trom the mesial end of the prominent longitudinal fissure. The distal marginal ridge is better developed than the mesial; a short, narrow físsure partially íncises the middle of its mesial face. The fovea posterior is represented by short tissures that radiate trom the distal end of the longitudinal turrow.

M r. The crown has a square occlusal outline wilh slighl mesial . displacement of the buccal cusps so that the mesiobuccal comer is more acutely angled than the distobuccal comer. Al four principal cusps are well developed. The protocone appears to have been slighlly the largest; the others are nearly equal in size. Occlusal morphological detail consists ot a remnant ot the triradiate tissure of the trigon basin and a pit representing the fovea posterior. The distal marginal ridge appears to have been moderately thick and complete. The buccal groove is represented by a V-shaped depression that terminates abruptly in a cleft.

SKW 5 Mandible wíth l+R P3-M3 (Fig. 31 a-c). This is a nearly complete mandibular corpus wilh much of the left ramus, the damaged crowns otthe LP3 and LM 1 , and the complete crowns of the RP3, L+RP4, RM" l+RM 2 and l+RM3. The incisor and canine alveoli are preserved; the right ramus is broken away below the level of the lateral rameal eminence, and the right gonial regíon is missing. The left condyle and coronoid process are míssing. The specimen was recovered in two pieces which were separated along an oblique plane that runs from the level of the lP4 and M 1 through the posterior part of the symphysis and to the level of the RM 1 and M 2 . A substantial amount of bone is missing from the left corpus along this break, and lhe symphyseal region contacts the corpora only at two points. The M3S were in the process of erupting at the time of death. The distal portion of the lM3 is situated below the rim of the alveolus, and only very faint enamel tacets are developed on the me sial cusps of these molars. Occlusal wear is generally slight, with small dentíne exposures onty on the M1s.

Mandibfe. Viewed from its occlusal aspect, the incisor and canine alveoli are arranged in an anteriorly convex arc, and the postcanine tooth rows are nearly parallel. The lateral edges of the corpora diverge posteriorly to a greater degree than the medial edges owing principally to the lateral swelling at the rameal root. Viewed late rally, the symphysis has a ver.¡ slight posterior inclination from the aveolar margin to the symphyseal tubercle, below which it curves posteriorly lo the base. The anterior surface is marked by prominent juga over the 1, rools. The postincisive planum is transversely concave; it slopes posteriorly and slightly inferiorly to the superior transverse torus. The torus ís situated above midcorpus height, and extends posteriorly to the level of the P4. The deep genioglossal fossa is bounded inferiorly by a broad inferior transverse torus. The superior and inferior corporeal borders are nearly parallel from the P4 to the M 2 ; on the right two mental foramina are located 14,0 mm and 17,5 mm below the alveolar margino A single foramen is discernible on the left, but the surface bone is eroded in this region. The ramus takes root at the level of lhe M 1 , with lhe oblique line originating some 12,0 mm below the level ot the alveolar margin. The extramolar sulcus ís shallow and ver.¡ broad. The ramus is broad (c. 62 mm), and the gonial angle is gently rounded. The base of the condyloid process (Le., lhe neck of the condyle) is preserved, suggesting that the ramus was rather short compared to its breadth. Medially, a massive triangular torus arises from poorly preserved endocoronoid and endocondyloid buftresses.

NEW CRANIOOENTAL HOMINID REMAINS

97

The principal dimensions (mm) of the corpus are as follows:

LPJ RP 3 LP. RP, LM, RM, LM 2 RM 2 LMJ RM J

8readth

Height

20,0 21,0

36,8 35,8

22,6 25,4 25,6 28,5 30,8 28,5 31,5

33,9 34,0 32,6 31,5 32,1 33,0 34,3

P3. The crown has a nearly square occlusal outline. The protoconid is well developed, and the smaller metaconid is situated mesial of the midcrown axis that bisects the protoconid. The moderate me sial marginal ridge is complete. The fovea anterior is represented by a small pit. The moderately thick transverse ridge is complete, as is the thick distal marginal ridge. The fovea posterior is better developed than the lovea anterior. Buccally there are barely pe rceptible furrows in the mesio-occlusal and disto-occlusal corners.

a

P4 . The crown has a somewhat trapezoidal occlusal outline; the protoconid projects mesiobuccally to a slight degree, and it is larger than the metaconid. The protoconid is situated slightly mesial to the midcrown transverse axis, and the metaconid is located slightly mesial to the' Jevel of the protoconid. The mesial and distal marginal ridges are thick and complete, and the moderate transverse ridge is located distal to the level of the cuspal apices. The distal marginal ridge is partially delineated from the cusps by shallow transverse lissures; the buccal fissure continues onto the buccal face as a distinct vertical furrow, and the lingual fissure continues onto the lingual surface as a shallow furrow.

b

MI. The crown has an ovorectangularocclusal outline. The five principal cusps are well developed. The protoconid is the largest, and the others are nearJy equivalent in size. The metaconid contacts the hypoconid, and the lingual groove is localed distal lo the level of the mesiobuccal groove, resulting in a symmetrical Y pattern. A distinct tuberculum sextum is present; there is no tuberculum intermedium. The fovea anterior is a centrally positioned transverse fissure that is enclosed distally by a narrow enamel ridge. The mesiobuccal groove terminales in a pit, and the slight dislobuccal groove ends gradually.

M2. The crown has an ovorectangular outline wíth a genlly rounded distobuccal comer. The ti ve principal cusps are well developed; the protoconid, metaconid, entoconid and hypoconid are approximately 01 equal size, while the hypoconulid is somewhat smaller. A moderate tuberculum sextum ís present; there is no tuberculum intermedium. The metaconid and

'E

e Fig.31

SKW 5 in oeelusal (a),left lateral (b) and right lateral (e) views; seales= 1 cm.

98

NEW CRANIODENTAl HOMINID REMAINS

hypoconid are in contact by virtue 01 a Iingually elongate, hypoconid; the lingual and mesiobuccal grooves are in ap- . proximately the same plane. The fovea anterior is represented by short buccal and longer lingual arms that radiate from the mesial end ofthe longitudinal lissure. The mesiobuccal groove terminates in a deep pit with a buccal wall that extends onto the distal aspect 01 the protoconid as an incipient singular shelf. The slight distobuccal groove ends gradually.

M3 . The crown has an MD elongate, distally tapered outline. The live principal cusps are present; the mesial two are dominant, and the hypoconulid is subdivided by a tissure. A well-developed tuberculum sextum is present, and there is a distinct postmetaconulid, which is more strongly expressed on the right crown. The fovea anterior is a lingual limb from the mesial end 01 the longitudinal lissure, and the tovea posterior is represented by a small pi!. The mesiobuccal groove ends abruptly in a pit that is surrounded buccally by a moderate enam el wall that extends onto the distal aspect of the protoconíd.

a

SKW 6 LM 3 (Fig. 32). This is a smalltriangle 01 enamel Irom the mesial moiety of the crown. It preserves the mesial marginal ridge, the base 01 the protocone and part of the apex of the paracone. There;s no wear apparent on this Iragment. The mesia! marginal ridge is comparativelythín and low; it is incised by a very narrow fissure. The paracone apex has an irregular appearance, with a curved ridge of enamel on the mesiolingual aspecl of the cusp.

b Fig.33 ~.

---'

Fig.32

SKW 6, occlusal view; scale '" 1 cm.

,"

'~.~ .... ' " ,... .' • "e'li .

~,

SKW 8 in occlusal (a) and lateral (b) views; scale = 1 cm.

.,.. . ...

SKW 8 Partial maxilla with u 2 , L + R.c., RP 4 _RM 2 (Fig. 33a, b). This is the incomplete anterior alveolar portion of a right maxilla with a small bit 01 bone around the root 01 the U 2. It preserves the naso-alveolar c1ivus and the anterolateral surface around the roots of the p3, and the anterior aspect of the patatal surface. The U 2 and L.c. are separate, but the f root joins the medial wall of the alveolus, and lhe Q. is joined by its interproxi1 mal contact facet and a small sliver of bone. The open U 2 alveolus is preserved, the root stumps of the Rl' and R1 are broken f1ush with the alveolar margin, and the Rp 3 roots are 1 exposed. The RM crown is rather Iragmentary, wilh pieces missing along cracks radiating from the centre 01 the protocone.

This specimen represents an adult individual. Occlusal wear is heavy, with large dentine islands exposed on the M' and p4. 2 The M is worn flat with protocone and paracone dentine exposures, and the incisor and canine crowns have been considerably reduced in heighl. The M 2 possesses a distal interproximal contact lacet. Maxilla. The tia t naso-alveola r clivus recedes from th e occl usal plane al about 45°, the anterior aspect 01 the maxilla turns 3 around the p roots, and the canine and incisor roots are situated in a nearly straight transverse line. The incisive foramen is large, and the palate is anteriorly shallow. The l' roots diverge apically al about 20° from one another.

,2.

The crown is heavily worn to a tlat surface wilh a slight lingual slope; crown height is reduced to a level coincidentwith the mesíal and distal cervices, so that only short cervical portions 01 the buccal and lingual surfaces remain. The mesial aspect 01 the crown is damaged.

NEW CRANIODENTAl HOMINfO REMAINS

99

k. The left tooth has an intact crown and root; the right crown is cracked mesially. Wear is heavy, with considerable reduc3 tion in crown height; on the left tooth the p contact facet is situated on the distolingual aspect ofthe crown, while this facet is restricted to the distal aspect of lhe righl crown. The crown has an ovoid occlusal outline, and Ihe buce al margin is slightly skewed mesially. The buccal cervical margin has a slight prominence; the lingual cervical eminence is symmetrically disposed

p3. This tooth is three-rooted, as índicated by roo! remnants in the alveolus. p4. Wear is heavy; the crown has been reduced to a flat surface with confluent concave dentine exposures on the prolocone and paracone. The crown has a Bl elongate, ovorectangular occlusal outline. The tooth has three separale rools. M t Wear is heavy; the crown has been reduced to a nearly tlat plan e, and occlusal enamel remains only on the hypocone. The crown has a square occlusal outline, and the contours suggest that the four principal cusps were well developed. M 2 . Wear is heavy; the crown is worn Ilat with concave denline

exposures on the paracone and prolocone. The dentine exposure on the paracone is larger than that on the protocone. The crown has a somewhat trapezoidal occlusal outline; the distobuccal corner is reduced and more obtuse Ihan the mesiobuccal comer. The crown is slightly BL broader across the mesial cusps. SKW 10 RM1 (Fig. 34). This is Ihe buccal and distal portions of the crown, and the buccal sides 01 the mesial and distal rool plales. The crown is broken along a longiludinalline thal cuts through the centre 01 the protoconid and hypoconid, and a transverse line lhal runs through the centre 01 the entoconid. The enamel is cracked, wilh slight displacement of the pieces. Wear is slight; the buccal cusps are reduced to a nearly flat plane with small dentine exposures on the proloconid and hypoconid. It is evident that Ihe five principal cusps were present; the hypoconid and hypoconulid appear equivalenl in size. The presence of a small tuberculum sexlum is indicated by a slight mesial swelling into the lovea posterior, which is represented by a small pi!. The moderale mesiobuccal groove terminates abruptly in a pit with a low buccal enamel wall; the modera te distobuccal groove ends gradually. The buccal apex of the distal root plate extends for a slight distance beyond the plate; ;t is 14,7 mm long. SKW 11 Partial cranium (Fig. 35a-c). This is the reasonably preserved maxilla and the fragmentary, crushed left side 01 the basicranium. The maxillary piece preserves the naso-alveolar clivus, the right side 01 lhe pyriform aperture, the paJate to the 2 1 level of Ihe M , the left alveolar process to the level 01 the M , and part of the right palatine. The basicranium preserves the mastoid process, the tympanic, part of the grealer wing of the sphenoid, the fragmented pterygoid plate, and part of the left palatine. The nasal bones are connecled lo Ihe basicranial piece by malrix. The basicranial piece connects to the maxil-

~-=

Fig.34 SKW 10, occlusal view; scale = 1 cm.

lary piece by the posterior wall of the socket of the distobuccal 3 rool of the lM , and the nasal bones make incomplete conlact with the fronlal process 01 the maxilla on the right side. The Rp3_RM3 and LM 3 are preserved intacI, Ihe lingual half of the 2 LM is preserved, as are the fragmentary roots of lhe Lp3_LM 1 . 1 The complele alveoli of the RI _RQ are preserved. Toothwear is slight; only the protocone of the M1 has a small 3 dentine island. The M s are fully erupled wíth very slighl occlusal wear; Ihe apices of the rools, however, are still palent. Maxilla. The naso-alveolar c1ivus is f1at, and recedes from the alveolar margin al aboul45° to the occlusal plane. The inferior margin of the pyriform aperture, in so far as it is preserved. indicates a gentle transition from the floor 01 the nose to 'the naso-alveolar c!ivus. The anterior nasal spine is situated posterior lo the level of Ihe inferior nasal margino The lacial portion is gentJy con ve x lateral to the pyriform aperture, which possesses a rounded margin, and it turns laterally at Ihe rool of the zygoma, which arises at the level 01 the p4. The nasoalveolar clivus extends laterally over the C. root, and the alveolar process turns posteriorly around the rools 01 the p3. The canine and incisor alveoli occupy the same transverse plane. Some 17 mm 01 the length 01 the nasal bones is preserved (nas ion and the inferior margins are missing). The nasals are quite narrow superiorly and diverge inleriorly; superiorly they project anteriorly in the midline while they are tlat inleriorly. Basicranium. The lateral plerygoid plate was evidently rather broad. The tympanic piate is deep supero-inleriorly, and its anterior surface is slightly concave in this direction. The tympanic plate projecls inferiorly into a well-developed, sharp crisla petrosa, the inferior margin of which is damaged except immediately lateral to the vaginal process. Here it altains a height of 6,5 mm Irom the pelrolympanic fissure al the stylomastoid foramen. The tympanic is thickened inferoposteriorly; its laleral margin is missing. There is no indication 01 an ossilied styloid process. The crista petrosa is separated from the mastoid process by a broad (c. 5,5 mm) furrow, and the mastoid process has a steep medial surlace.

NEW CRANIODENTAL HOMINID REMAINS

100

b

a

c Fig.35

SKW ~ 1 righ! lateral view of maxilla (a). occlusal view 01 maxilla (b), and left lateral view 01 maxilla with basicranium in posilion (e); scales = 1 cm.

p3. The crown has an ovorectangular occlusal outline. The two principal cusps are well developed. The protocone is slightly larger than lhe paracone and is situated slightly mesial 01 the midcrown BL axis lhat bisects the paracone. The lingual surtace is slightly more protuberanl mesially lhan distally. The moderate mesial marginal ridge is complete, and the lovea

posterior is restrlcted buccally. The distal marginal ridge thick and complete, and lhe fovea posterior has longer bucc: and shorler lingual arms.

p4. The crown has an ovoreclangular occlusal outline. The ''''' principal cusps are well developed. The protocone is slight

NEW CRANIODENTAL HOMINID REMAINS

larger than the paracone and is situated slightly mesial ot the midcrown BL axis that bisects the paracone. The strong mesial marginal ridge is complete, and the fovea anterior is represented by a buccallimb from the mesial end of the longitudinal furrow. The thick distal marginal ridge is complete; its buccal end is separated from the apex of the paracone by a narrow transverse fissure that extends from the fovea posterior. Buccally, the fissure between the distal marginal ridge and paracone extends as a moderate, vertical furrow. MI. The crown has a nearly square occlusal outline, wilh a slight curve to the distal margino The four principal cusps are well developed. The protocone is the largest; the paracone and metacone are equal in size, and the worn hypocone appears only marginally smaller. A small mesiobuccal pit ís the remnant of the trígon basin. The damaged distal trigon crest is well developed, and it appears to have been com plete. The moderate distal marginal ridge thins buccally, and it is incised at the base of the hypocone by a thin, distally directed fissure from the fovea posterior. The buccal groove terminates abruptly in a small pit that is surrounded by a low, thin enamel wall. M 2 . The crown has a nearly square outline wilh a slightiy reduced distobuccal corner that is more rounded than the other corners. The four principal cusps are well developed. The protocone is the largest, followed by the paracone; the slightly smaller metacone and hypocone are equal in size. The fovea ante rior is restricted to a small pit in the mesiobuccal corner of the crown. The moderately thick distal trigon crest is strongly 'waisted' in its middle by mesial and distal fissures. The moderate distal marginal ridge supports several incipient cuspules. The narrow buccal groove ends rather abruptly. The distal end of the lingual protoconal surface has a shallow vertical fissure, and several furrows incise the lingual surface of the hypocone. 3

M . The crown is somewhat trapezoidal in occlusal outiine; it is BL broader across the mesial than across the distal cusps. The four principal cusps are present. The protocone is the largest cusp, followed by the paracone, which is only marginally larger than the hypocone; the metacone is reduced in height and area. The strong mesial marginal ridge supports three incipient cuspules. The distal trígon crest is strongly 'waisted' on the right crown; on the left it is deeply incised. The thick distal marginal ridge is multicuspidate. The Carabelli trait is expressed as a V-shaped furrow. Vertical fissures incise the distal aspect of the lingual protoconal surface; these are better developed on the left crown. The left crown has a moderate, vertical buccal groove, while on the right side there are severa! vertical fissures and slight surface irregularities in this region. SKW 14 LM2 (Fig. 36). This is a nearly complete crown lacking the root. The cervical margin is missing from the entire periphery of the crown, and a wedge of enamel is missing from the distobuccal corner. Wear is very slight, with small enamel facets on the protocone and paracone. The crown has a trapezoidal occlusal outiine, with a reduced, obtuse distobuccal comer. The crown is BL broader across the mesial than across the distal cusps. The tour principal cusps are present.

.~ig.

101

36

SKW 14, occlusal

view; scale

=1 cm.

The protocone is the largest, followed by the paracone; this is larger than the hypocone, which appears slightly larger than, or at least equal in size to the metacone. The thick mesial marginal ridge is complete, and the fovea anterior is a buccally directed fissure from the mesial limb of the trigon basin. The moderately thick distal trigon ridge is high; it is incised in the middle by a deep, narrow lissure. The preserved portion of the distal marginal ridge is moderately developed. The fovea posterior is a moderately large basin. The buccal groove is a barely perceptible depression. The Carabelli trait is represented by a short, oblique fissure, and there is a shallow oblique furrow on the distal aspect of the protoconal lace. SKW 15 Lp4 (Fig. 37). This is the distobuccal comer of the crown with a short segment of the distal (or distobuccal) rooL The fragment preserves the distal half of the protoconid, and part of the distal marginal ridge. The protoconid has a slight enamel wear facet, and there is some rounding to the thick distal marginal ridge, the buccal end of which is delineated as a low, triangular cuspule from the protoconid by narrow furrows. The fissure between the cuspule and protoconid courses onto the buccal face of the crown as a deep furrow. It is evident from the curvature of the preserved portion of the root that separate buccal and distal roots were present. SKW 29 Badly crushed cranium (Fig. 38a, b). This is an incomplete, badly crushed and distorted cranium with the Lp3_LM3 and the RP4_ RM 3. The specimen preserves portions of the parietals and frontal, left orbit and zygoma, part of the maxilla, anterior aspect of the nasal cavity and part 01 the palate. The left occipital preserves part of the nuchal planum, the superior nuchal line and a distorted superior tableo The cranium is crushed flat, so that the orbit appears as a narrow slit, and the frontal and anterior parietal region forms a cracked, concave depression. Despite the crushing, the anterior part of lhe nasal cavity, including the vomerine insertion is preserved intact. Only a small segment of the mesiolingual corner of the p3 remains, the Lp4 is severely crushed and only the lingual half of the Rp4 is preserved. The 1 LM is crushed with displacement of the crown pieces, and enamel is missing from the mesiobuccal and distobuccal corners 01 the RM'. The LM 2 is nearly complete, while the

NEW CRANIODENTAL HOMINID REMAINS

102

Fig.37 SKW 15, occlusal view~ scale '" 1 cm.

3

mesial moiety is missing trom the RM 2 • The LM is complete, and lhe lingual surface ot the protocone is missing trom the RM 3 . Toothwear is gene rally moderate, with dentine expo· sures on the M 1s and enamel facets on all other crowns.

Cranium. Remnants ot lhe sagittal crest are preserved. The parietal is thin (e. 4,7 mm), and the supraorbital margin on the left side preserves details ot a vermieulate surface. The base 01 the vomer, which extends for some 3,0 mm in height, is present; its anterior end expands as a plate at lhe base of the anterior nasal spine, which is located slightly posterior to the level 01 the anterior margin ot the pyrilorm aperture.

a

MI. The crown appears lo have had a square occlusal outline, with the tour principal cusps well developed. The protocone is the largest; the other three appear to have been equivalent in size. The worn mesial and distal marginal ridges are moderately thick and complete. The worn distal trigon crest is 'waisted' by a mesíal tissure. The buccal groove terminates gradually.

M 2 . The crown has a square outline, and the tour principal cusps are well developed. The protocone is the largest cusp, lollowed by the paracone; this is larger than the metacone and hypocone, which are equivalent in size. The thick mesial marginal ridge is complete, and the moderate distal marginal ridge is also complete. The moderate distal trigon crest is damaged. The buccal groove terminates gradually. Two short, vertical turrows incise the distalmost aspect of the lingual surface of the protocone.

M 3 . The crown has a somewhat lrapezoidal' outline, with a reduced distobuccal comer. The four principal cusps are well developed. The protocone is notably the largest, tollowed by the paracone, metacone and hypocone in decreasing order 01 size. The thick mesial marginal ridge is complete, and the lovea anterior is a buccally directed transverse fissure. The moderate distal trigon crest is complete, but it is partíally incised by two mesíal furrows. The moderate distal marginal rídge is complete; it supports a small cuspule on the left crown and two incipient cuspules on the right tooth. The fovea

b Fig.38 SKW 29 in superior (a) and inferior (b) views; scale == 1 cm. . posterior is a narrow turrow. The buccal groove is very weakly developed. An oblique fissure incises the distal aspect ot the

NEW CRANIODENTAL HOMINID REMAINS

103

F19.. 40 SKW 4768, oeclusal view; scale '" 1 cm.

Fig.39

SKW 4767, oeclusal view; seale = 1 cm.

lingual protoconal surface, and on the right crown a seeond, mesially plaeed oblique furrow delineates a triangular Carabelli ·welt.' SKW 4767 RMl (Fig. 39). This is a complete, albeit rather poorly preserved crown with damaged roots. The crown has a numbe r of craeks, the mesial and lingual cervical margins are damaged, and the distobuccal moiety is displaced slightly. Wear is slight; enamel facets are present on all eusps, but dentine is not exposed. The crown has an ovorectangular occlusal outline. The five principal eusps are well developed, and a small tuberculum sextum is present. There is no development of a tuberculum intermedium. The metaeonid is slightly the largest cusp, followed by the protoconid, hypoconid and entoconid, which are nearly equal in size. The hypoconulid is the smallest cusp, and the tuberculum sextum is less than half the size of the hypoeonulid. The thick mesial marginal ridge is complete, and the fovea anterior is represented by short buceal and Iinguallimbs from the mesial end of the longitudinal fissure. No accessory fovea development is manifest. The metaconid contacts the hypoconid by a strong principal cres!, the distal end 01 whieh is formed into a small euspulid. The lingual groove is situated well distal of the level of the mesiobuccal groove, resulting in a rather symmetrical Y occlusal pattern. The mesiobuccal groove terminates abruptly on an enamel shelf that bridges the protoconid and hypoconid. The distobuecal groove ends gradually. Separate mesial and distal root plates measure some 16,2 mm and 17,5 mm in length respectively, and the root neck is 2,5 mm high Iingually. SKW 4768 LM 2 (Fig. 40). This is the distolingual comer of the crown and the distal half of the lingual rool. The fragment preserves the hypocone, distal marginal ridge, distal trigon erest and part of the metacone. Wear is slight on the hypocone and distal marginal ridge. The hypocone appears compara-

tively small (cf. SK 48), and the moderate distal marginal ridge is complete. It supports a small euspule adjaeent to the metaeone that is delineated by slight furrows. The fovea posterior comprises a short buccal and longer Iinguallimb lhat radiate from the lingual groove. The distal trigon crest is 'waisted' by a deep fissure on its mesial aspect that nearly incises the ridge. The apieal end of the lingual root, which measures nearly 14,0 mm in length, is ;ntact. SKW 4769 LM2 (Fig. 41). This ;5 the distal half of a eracked crown with the intact distal root plate and part of the mesial root plate. Broad, matrix-filled cracks radiate distally from the middle of the preserved mesial surfaee, whieh represents a transverse section through the hypoconid and the mesial part of the entoconid. Wear is moderate. The cusps are redueed lo a nearly tlat plan e, and the distal interproximal contact facet is large. The hypoconulid appears to have been comparatilJ'ely reduced, and it is evident that a large tuberculum sextum was present. The root neck is only about 1,0 mm high.

Fig.41

SKW 4769, ocelusal view; scale'" 1 cm.

104

NEW CRANIODENTAL HOMINID REMAINS

Fig.43 SKW 4772, occlusal view; scaJe = 1 cm.

Fig.42 SKW 4770, interior view; scale = 1 cm.

SKW 4770 Crushed maxilla (Fig. 42). This is a fragmentary, badly crushed and distorted edentulous maxilla. It preserves 1 the medial wall of the RI socket, lhe nearly complete socket 2 01 (he U ' , the severely damaged and crushed U socket. and the nearly complete socket ofthe lQ. The alveoli for the lingual and mesiobuccal roots of the lp3 are preserved, as is part of what appears to be the mes;al wall of the distobuccal root of the l p3, and the lingual root of the lP4. The specimen appears to have been water-Iogged prior to its preservation, especially to judge by lhe appearance of the facial aspect. The palate is crushed and open along the median suture. The palate is some 8,0 mm thick posteriorly. SKW 4771 ?RP4 (No illustration). This is a small fragment 01 the corner of a worn crown with a small piece of the root. Jt probably represents the distobuccal comer of a right mandibular fourth premolar. The enamel is very thick. SKW 4772 Rp4 (Fig. 43). This is a nearly complete crown lacking the cervical margin and rool. The mesial face is damaged cervical to the occlusal margin, and the unworn crown appears to have been incompletely developed. The crown has an ovoid occlusal outline. The two principal cusps are well developed. The protocone is slightly larger than the paracone, and both occupy the midcrown 8l axis. The moderately thick mesiaf marginal ridge is complete, but relatively low. The fovea anterior is a buccally directed fissure from the mesial end of the deep, broad longitudinal furrow. The distal marginal ndge is thick, and two moderate and one small cuspule are delineated along its length. The buccal end of the distal marginal ridge is separated from the paracone by a moderate fissure that extends onto lhe buce al face to form a rather deep,

-. _ _

~,~"07"'.-"_

'V-shaped depression. A barely perceptible mesiobuccal lissure is evident, and the lingual crown face shows a slight furrow between the disto lingual accessory cuspule and the protocone. SKW 2581 Right temporal (Fig. 44a,b). This is the isolated, rather poorly preserved, incomplele right temporal of an adult individual. The specimen preserves much of the mastoid process, about 30 mm of the lengths of the tympanic and petrous pyramid, and a portion of the postglenoid process. The vertical tympanic plate projects inferíorly some 19,0 mm· from the petrotympanic fissure, and it is siightly concave mediolaterally. It possesses a well-developed, sharp crista petrosa, and its anterolateral aspect is thickened. The plate extends some 17,0 mm below the postglenoid process, whích is fused to it. An ossified styloid process is present 27,1 mm medial to the inferolateral edge of the tympanic. The suprameatal roof ove rhangs (he lateral margin of the tympanic to only a very slight degree. The mastoid process is well developed and it appears to be rather extensively pneumaticized. The preserved tip extends approximately 10,0 mm below the inferior margin of the tympanic (when orientated approximately in Frankfurt Horizontal), and it curves slightly anteriorly. The anterior surface of the mastoid process is situated in close juxtaposition to the tympanic plate. SKW 2637 Enamel fragment (No illustration). Thís is a small piece of a worn molar or premolar. The occlusal remnant shows a flat enamel tacet, and a portion of a moderate, flat interproximal contact facet is preserved. The enamel is very thick. SKW 2702 Tooth fragment (No illustration). This is the corner of a crown with a portion of the rool. The disposition of the root indicates bifurcation, and the presence of an interproximal lacet aboye it suggests that the 'specimen represents the mesiobuccal aspect of a right maxillary molar. It preserves a small part of the corner of a worn crown. Occlusally the dentine island shows a thick enamel border.

NEW CRANIODENTAL HOMINID REMAINS

105

.{ig.45 SKW 3033, occlusal view; scale = 1 cm.

a Fig.46 SKW 3068, occlusal view; scale= 1 cm.

b

Fig.44 SKW 2581 in lateral (a) and inferior (b) views; scale = 1cm.

SKW 3033 R p 3 (Fig 45). This is the lingual half 01 the crown with a short segment ot the lingual root. The lingual aspect 01 the protocone has a concave dentine exposure that is connected mesially to the large paracone exposure by a transverse dentine isthmus. The mesial interproximal contact tacet is concave; the distal lacet is tlat. 2

SKW 3068 LM (Fig. 46). This is the buccal hall ot a crown with short root segments. The crown is broken along an irregular longitudinal line that cuts through the buce al aspects of the protocone and hypocone. Wear is heavy; the buccal cusps are worn to a nearly flat surtace, and a large, deeply concave dentine exposure extends trom the protocone to the paracone. The metacone is worn tlat, but dentine is not exposed on this cusp. Occlusal detail is limited to shallow remnants of the trigon basin and fovea posterior. The distobuccal crown comer ís reduced, being more obtuse than the mesiobuccal comer, which suggests that the crown had a trapezoidal occlusal outline. TAXONOMIC ATIAIBUTIONS Except for several 01 the small, fragmentary dental specimens that cannot be assigned to a specitic hominid taxon with

assurity, the entire Member 1 Hanging Remnant sample described here is attributable to Paranthropus. COMPAAISONS AMONG SWARTKAANS HOMINID SAMPLES The diameters 01 the recently recovered hominid teeth Irom Swartkrans are recorded in Tables 6-8. 01 the specimens that comprise this collection, the craniodental remaíns 01 perhaps six individuals can be ascribed to Homo, while the remainder are attributable to Paranthropus. Only one ot the Homo specimens (SKX 21204) derives from Member 1; the other Homo teeth for which diameters could be determined (i.e., SKX 257, 258,267,268,334, 339, 610,2354, and 2355) come Irom Member 2 sediments. Four 01 lhe new Paranthropus specimens trom Member 2 preserve gnathic parts. Specimen SKX 162, an incomplete right maxilla 01 a juvenile, is morphologically similar to Member 1 Hanging Remnant specimens in the slope of the nasoalveolar clivus, the slighl convexity over the " and concavity over the roots, and in the anteriorly tlat palate. Specimen SKX 265, an edentulous left maxilla of a subadult individual, is similar to Member 1 Hanging Aemnant specimens except in the relatively forward position of the anterior nasal spine, which is situated in the same coronal plane as the anterior margin ot lhe pyritorm aperture. The delineation 01 this margin in SKX

,2

NEW CRANIODENTAL HOMINID REMAINS

106

265 tends to be somewhat more marked trian in other Swartkrans maxillae, and the naso-alveolar clivus is bounded late rally by a sharp, anterolateral projecting bony ridge over rool. lhe The SKX 4446 and SKX 5013 mandibular corpora lrom Member 2 are similar to Paranthropus specimens from Member 1 Hanging Remnant in both size and morphology. The SKW 5 mandible lrom lhe Member 1 Hanging Remnanl possesses a reJatively low corpus and a relatively broad ramus by comparison with other Swartkrans Paranthropus homologues. The SKW 11 maxilla and partial basicranium are morphologically similar to other Paranfhropus specimens, as are the preserved parts al lhe SKW 8 maxilla and erushed SKW 29 cranium. The SKW 2581 Paranthropus temporal differs from other Paranthropus specimens in the presence of an ossified styloid process, Most of the new collectíon cornprises isol atad and/or associated teeth. The recently recovered dental remains from the Member 1 Hanging Remnant breccia are morphologically and metrically similar to other Member 1 Hanging Remnant specimens. At the same time, the teeth from Member 1 lower Bank, Member 2 and Member 3 sediments find morphological equivalents among the Member 1 Hanging Remnant specimens. An element-by-element comparison reveals that the morpholoqical variability within the Member 1 Hanging Remnant sample is as great as, ¡f not greater than the vanability evinced among the Member 1 lower Bank, Member 2 and Member 3 sarnples. The MD and Bl diameters of the Paranthropus teeth from Member 1 lower Bank, Member 2 and Member 3 deposits are compared with those recorded far Member 1 Hanging Remnant homologues in Figs 47-51. lt ts apparent that severat of the teeth from Members 2 and 3 tall slightly beyond lhe observed ranges for trie Member 1 Hanging Remnant sample. but in most of these instances the latter sample consists of comparatively tew specimens. In view of their morphometrical símllarlttes. the recently excavated Paranthropus fossHs undoubtedly represent the same specles as the previously available Member 1 Hanging Remnant specimens. The permanent and deciduous tooth dlameters for the combined Swartkrans Perentnropus sample are recorded in Tables 9-11. It is evldent that when sample size is taken into consideration, tooth diameter variation (as expressed by the coefficient of variation) within the combinad Paranthropus sample from Swartkrans is certainly no greater than that within any other early hominid species sample (Tables 12-18).

e

COMPARISONS BETWEEN SWARTKRANS AND

KROMDRAAIPARANTHROPUSSAMPLES The Swartkrans australopithecine sample is attributed to Paranthropus robustus (or Australopithecus robustus) by most workers. Howell (1978) and Grine (1982, 1985), however, haya argued that the Swartkrans sample should be accorded separate status as P. crassidens, the specific designation originally applied by Broom (1949), and thal lhe name P. robustus Broom, 1938, should be usad in reference to the Kromdraai hominid sample. In view 01 the increased number 01 Paranthropus teeth that are currently available from Swartkrans, it is perhaps timely to compare this sample with

other early hominid samples in an attempt to analyse the purported distinctiveness of the Swartkrans and Kromdraai 'robust' austraJopithecines. Because maxillary permanent incísors and canines, mandibular permanent caninas, maxnlary deciduous teeth, and mandibular deciduous central incisors are not known from Kromdraai, these teeth will not feature in the comparisons. The MD diameters 01 the mandibular permanent incisors from Swartkrans (P. crassidens) are comparad to those for olher auslralopilheeine samples in Table 12. The I,s al P. crassidens and P. boise! tend to be smaller than those 01 A. africanus and A. atarensis, whíle the P. crassidens, P. boisei and A. afarensis means for the 12 are similar to one another and smaller than the corresponding A. africanus sampíe mean. The central and lateral incisors from the single P. robustus individual (KB 5223) are slightly smaller lhan the smallest Swartkrans homologues. The M O and Bl diameters of the maxillary and mandibular permanent cheekteeth are compared in Tables 13-16 and Figs 52-55. In each instance the A. afarensis mean is the smallest and the P. boisei mean is the largest. The A. africanus sample means are larger than those for A. afarensis except with regard to Ihe MD diameler 01the P, and the Bl diameler of the M,. In these two ínstances the sample averages are nearly identical. At the same time, the P. crassidens sample means faH between the larger P. boisei and srnaller Australopithecus values except tor the MD and Bl diameters of the P3 and the BL diameter of the M3. In these cases the A. afarensis and/or A. africanus means are nearly the same as those for the P. crassidens sample. Metncal values for the Kromdraai P. robustusteeth are most similar to the P. crassidens sample means with regard to the MD and Bl diameters of the upper and lower prernolars, the Bl, diameler olthe M', and the Bl diameter 01 the M,. It should be noted. however, that the P. robustus, P. crassidens and A. africanus values are all similar to one anather in the MO and al diameters of the P3 and the Bl diameter 01the M3. On the other hand, the P. robustus values tend to be srnaller than those for P. crassidens in permanent molar diameters, except for the Bl diameters 01the M 2 and M 3 , where the means are nearly identical. The P. robustus permanent molar diameters are intermediate between the smaller A. africanus and larger P. crassidens means (MD and Blof M3) , mast similar to the A. africanus means (MD and Blof M' and MD of M3), or even smaller than the A. africanus means (MO and Blof M 1 and M,). Only with regard lo the MD and Bl diamelers 01 the M, and the MD diameter of the M,. do all Kromdraai specimens fall below the observed P. crassidens sample ranges. Nevertheless, it is clearthat in a numberof instances the A. africanus and P. crassidens sample ranges overlap to a considerable exlenl (e.g., MD and Bl al M'_M', MD and Blof P" and MD of M, and Blof M3), such that the P. robustus values fall beyond the limits of the P. crassidens sample ranges about as often as they fall beyond the limits of the A. africanus sample ranges. The MD and Bl diameters of mandibular deciduous teeth for which Kromdraaj specimens are available are recorded in Tables 17 and 18, and lhese values are depieted graphieally in Figs 56 and 57. As was noted fer the mandibular permanent incisors, the 8wartkrans and Kromdraai di2S tend to be smaller

NEW CRANIODENTAL HOMINID REMAINS

than A. africanus and/or A. etereneis homologues, and the single Kromdraai di2 is slightly smaller than Swartkrans crowns. Paranthropus crassidens and P. boisei des tend to be smaller than those 01 A. africanus and A. afarensis in theír M O diameters, and the single P. robustus crown ;5 notably the smallest in this dlameter. The observad Swartkrans sample range far the Bl marneter 01 the de encompasses all other australopithecine values. The slze patterns 01 the ceciduous rnolars are similar to those displayed by the permanent cheekteeth, where P. boisei values tend to be larger than those 01 P. crassidens. While the permanent tooth rneans for A. atarensis are almost unitorrnly smaller than those for A. africanus, the deciduous mandibular molars of A. etricsnus tend to be smaller than those of A. afarensís. This, however, is not the case for the decíduous maxillary rnolars (Grine, 1984). The mandibular deciduous rnclars of P. robustus are, on average, smaller than those comprising the P. crassídens sample. The Krorndraai means far the MD and BL diameters ofthe drn- are intermediate between the A. eterensis and P. crassidens sarnple averages, while the Kromdraai values far the dm2 are mare similar to those ter A. eterensis than for P. cressidens. Only with regard to the MD diameter of the drn, do all P. robustus specimens fatl outside the nmtts ot the observed Swartkrans sample range. In the other tnstances the Kromdraai values faJl within the lower limits of the Swartkrans sample ranges as well as within the A. eterensis sample ranges. While P. robustus premolars tend to be large by comparison to the Swartkrans sample, the deciduous molars from Kromdraai tend to be smaUer than Swartkrans homologues. Thus, compared to the size relationships of the permanent molars of P. robustus and P. cressidens, the premolars of the former tend to be large, but the comparatively large sizes of the P. robustus premolars are not reflected by their deciduous precursors. In addition to the size differences noted aboye, the Krorndraai and Swartkrans mandibular molars display proportional differences that do not appear to be allometrically related (Jungers and Grine, 1986). The Kromdraai and Swartkrans teeth also differ in a numberof marphologícal features (Howell,

107

,,1978; Grine, 1982, 1984, 1985), In sorne ínstances they differ in terms of 'central tendency,' and in most of these cases the Kromdraai teeth resemble those of A. etricenus andlor A. afarensis more closely than those trom Swartkrans (Grine, 1982,1984). Forsevera! morphological features, the character states presented by the Kromdraai specimens are shared with A. eiricsnus and A. eterensis specirnens in contrast to the presumably derived conditions displayed by the Swartkrans and P. boisei crowns. In other features, the Kromdraai teeth appear to be intermediate along a morphocline between the A. afarensis and A. eiricenus specimens on the cne hand, and the Swartkrans and P. boise! specimens, on the other. The marphological and metrical differences between the Kromdraai and Swartkrans dental samples suggest that a species level distinction should be retained tor P. robustus and P. crassídens. This is especially evident when the rnorphometrical similarities among the various Swartkrans lithostratigraphic sarnples are considerad, Te subsume P. cressidens inta P. robustus serves only to mask the possible evolutionary significance ot the Kromdraai hominid sample.

SUMMAAY AND CONClUSIONS Tne craniodental remains of perhaps six individuals of Horno, and some 50 to 60 individual s of Paranthropus have been recovered by C. K. Brain fram the Member 1 Lower Bank, Member 1 Hanging Remnant, Member 2 and Member 3 ssdlments ot the Swartkrans Formation. The 'robust' australcpithecines specimens are morphologically and metrically very similar to Paranthropus Iossils recovered earlier from the Member 1 Hangíng Remnant breccia, and all are considered to represent a single specles. These newly discovered, diachronous Swartkrans samples serve to highlight the differences between the Swartkrans and Kromdraai specimens of Paranthropus. Although the Kromdraai hominid sample is very small, there appear to be a number 01consistent differences between it and the larger Swartkrans hypodigm. These suggest that species level distinction be retained far P. robustus and P. cressitiens.

NEW CRANIODENTAL HOMINID REMAINS

108

Table 6 Permanent maxillary tootn diarnetera of new Swartkrans hominid soeomens

Tooth Member

l'

l'

Q

p'

p'

M'

M'

M'

'LB lLB lLB lLB 2 2 2 3 lHR 2 2 2 lHR lHR 2 2 2 3 3 lHR lHR lLB 2 lHR lHR lHR lHR lLB 2 3 lHR lHR lLB 2 2 lHR lHR lHR lHR lLB lHR lHR lHR lHR 3

Specimen

MD meas.

MO esto

SKX 240 SKX 242 SKX 3300 SKX 19031 SKX 271 SKX 313 SKX 339 SKX 27524 SKW8 SKX 308 SKX 610 SKX 1788 SKW 8 IL) SKW 8 IR) SKX 162 SKX 310 SKX 312 SKX 25296 SKX 28724 SKW 11 SKW 3033 SKX 7781 SKX 162 SK 14133 SKW 8 SKW 11 SKW 4772 SKX 3354 SKX 162 SKX 26625 SKW 14133 SKW 11 SKX 3801 SKX 268 SKX 334 SKW 8 SKW 11 SKW 14 SKW 29 SKX 3335 SKW 11 (L) SKW 11 IR) SKW 29 (L) SKW 29 IR) SKX 21841

8,6 8,5 8,1 9,6 9,5 8,5 10,3 8,0

8,8 8,8 8,3 9,8 9,8

6,5 7,3 6,3 7,8 7,7 8,8 7,8 7,4 9,7 8,9 9,7 9,1 9,6 10,0 9,5 9,8 10,3 10,7 9,9 10,8

10,3 8,2 6,5 7,4 6,3

8,8

9,7 8,9 9,9 9,4 9,6 10,0 9,8 10,1

10,5 10,8 10,5 10,8

12,1

12,1

12,6 12,4

12,9 12,5 (12,4) 13,2 (12,9) 13,2 14,0 14,0 13,1 (12,8) 14,2 14,8 12,7

11,5

13,2 12,7 12,6 13,9 14,0 12,8 12,3 14,2 14,8 12,7 13,1 15,5

Bl meas.

Bl est.

7,0 7,2 7,0 7,4 8,1 7,4 6,6 6,3 6,2 6,4 7,2 5,7 8,8 8,8 9,4 9,1 9,4 9,1 8,8

7,0 7,2 7,0 7,4 8,1 7,4 6,6 6,3 6,2 6,4 7,2 5,8 8,8 8,8 9,4 9,1 9,4 9,1 8,8

14,3

14.3

14,5

14,5

15,1 15,7 13,8 14,2 15,8 16,1 14,6 14,5 14,4 12,9

15,1 15,7 14,2 15,8 16,1 14,6 14,5 14,4 12,9

15,1 15,8 15,7 16,8 15,0 16,7 17,0 16,2

15,1 15,8 15,8 16,8 15,0 16,7

16,6

16,6

17,0

16,2

13,1

15,5

109

NEW CRANIOOENTAL HOMINIO REMAINS

Table 7 Permanent mandibular tooth diameters of rrewSwartkrans horninid specimens.

room

" 1,

e p, p,

M,

M,

M,

Member

lLB 2 2 3 3 lLB 1-2 2 2 2 lLB lLB 2 lHR 2 1HR lHR 2 3 'HR lHR 1HR lHR 1LB 2 2 2 2 1HR lHR 1HR 2 3 lHR lHR lLB lLB

Soecírnen

SKX 5004(b) SKX 2355 SKX 3559 SKX 26967 SKX35416 SKX 21204 SKX 7992 SKX 2354 SKX1017 SKX 1313 SKX 5007 SKX 6013 SKX 241 SKW 5 SKX 311 SKW 5 (L) SKW 5 (R) SKX 4446 SKX 32162 SKW 5 (L) SKW 5 (R) SKW 10 SKW 4767 SKX 5023 SKX 257 SKX 258 SKX 4446 SKX 5013 SKW 5 (L) SKW 5 (R) SKW 4769 SKX 4446

MD meas.

MD est.

Bl meas.

al est.

5.2 5.0 5.2 5,7

5.4 6.0

6.2

6.2

6.7 6.1 6.1

6.7 6.1 6.1

6.5

6.5

6.5 7.2 8.2 7,7 7,4

6.5 7.2 8.2 7.7 7.4

7.1 5.8 6.8 5.7 5.8 7.3 6.8 6,7 9.0 8.7 11.2 10,6

11,7 10.8 13,3 15.3 16,1 13,2 13.3 13.6 15,1 13.2 15.9 16,0

(5.9) 7.1 (6.9) 6.5 7.5 6.9 7,1 mino 9.3 9.0 11,4

11,2

11,2

11.9 12.7

11,1 11,9

12.5

11,9 12.7 12,5 12.5

13.3 13.2

13.3 13.2

14,4

14,4 12,8 11,7 11,9

12,5

11,3 mino 13.6 15,4 16.2 (14.1 ) 13.3 13.6 15.5 (13.61 16,1

12.8 11.6 11,8 14,3

14,3

12,0

12,0 14.4

16,2

14,4 14,4

17.1

16.6 15,8

14,4

SKX 19892

14,8

SKW 5 (L) SKW5 (R) SKX 5002 SKX 5014

16.5 15,8 14,8 (14,3) 14.0 13,9

15.0

17,1

17,6

17,8

(14,3) 14,0 13,9

17.2

17,2

15,0

16,7

16,7

Table 8 Maxillary deciduous tooth diameters of new Swartkrans hcminfd speclmens.

Tooth Member

di

1

dm'

1LB 1-2 2 2 3

Specimen

SKX SKX SKX SKX SKX

16060 2003 163 267 32832

MD meas.

MO

esto

6,6 5,4

6,8

107 10,1

(10,9) 10,7

Bl meas.

Bl esto

4.3 4,2

4,3 4,2

12,7 11,0

11.8

12,9

(11,2) 11,8

NEW CRANIODENTAL HOMINID REMAINS

110

Table 9 MD diameters of Swartkrans Paranthropus permanent teeth.

Tooth

n

Maxillary teeth 1 1 11 l' 5

ep' p'

M' M' M'

13 23 23 20 21 21

Mandibular teeth 1, 8 1, 5 C 11 p, 14 p. 18 M, 19 M, 16 M, 20

X

5.0.

S.E.

C.V. %

Observad range

0,61 0,19 0,40 0,64

«rr

0,18 0,09 0,11 0,13 0,12 0,12 0.17

0,98

0,21

6,75 2,89 4,55 6,36 5,27 4,08 5,46 6,58

8,2-9,8 6,3-6,8 8.3-9,7 9,3-12,2 9,8-12,1 12,2-14,1 12,8-15.6 12,9-16.7

0,27

0,09 0,21

4,86 7,38

0,21 0,11

8,50 4,08

0,13 0,17 0,21 0,30

5,00 4,93 4,97 8,01

5,1-5,9 5,7-6,9 6,9-9,1 9,0-10,5 10,5-12.5 13,6-16,2

Table 10 Bl, diameters ot Swartkrans Paranthropuspermanent teeth.

Tooth

n

5,56 6,37 8,12 9,81 11 ,26 14,81 16,50 16,98

x

Mesiodistal diameters di' 2 6.78 di2 d' dm' 1 9,10 10,90 dm' 6 di, 4,00 1 di;! 4,90 2 o, 6,03 4 dm, 6 10,38 12,77 dm, 10

°°

BuceoJingual dfarneters 4,60 3 O d' O dm' 10,05 2 12,10 dm' 6 di, 3,90 1 4,25 2 di:! o, 4 5.13 dm, 8,43 6 cm, 10.83 9 di' df2

5.0.

SE

15,11 14,70 16,07

0.55 0.33 0.74 0.62 0,86 0.76 0.80

16,81

0,71

0.16 0.11 0.12 0,16 0,19 0,16 0,18 0,15

6,21 6,89 8.38 11,89 12.94 13,66 14,96 14.61

0,44 0,40 0,89 0.88 0,89 0,83 0.91 1.09

X

C.V.%

Observen range

Maxillaryteeth 9,04 6.58 8,80 10,06 10,81 13,23 14,14 14,89

0,57

0,54

0,47 0,69 0,40 0,56 0,73 0.82 1,36

15,2-17,7 14,8-19,6

JI

12

l'

9 16 16 21 23 19 23

ep' p' M' M' M'

Mandibular teerh 10 1, 6 C 13 p, 15 p. 17 M, 18 M, 19 M, 20 1,

Table 11 MD and BL diameters of Swartkrans eeremtvoous deciduous teeth.

Tooth

n

5.0.

SE

C.V. %

Observed range

3,94

10,40-11.3

0,13 0,43 0,45

0,06 0,18 0,14

2,16

4,8-5,0 5,9-6,2

4,14

10,0-11,1

3,52

12,0-13,5

0.53

0.31

11,52

0,41

0,17

3,39

0.50 0,65 0,70

0,25

9,75 7,71 6,46

0,27

0,23

0,21 0,24

4,41

5,69 5,17

4,98 4,22 7,09 5,80 10,62 7,40 6,88 6,08 6,08 7.46

6,3-8.1 5,8-6,9 8,5-11,1 13,2-15,2 13,5-16,3 13.0-16,8 14,4-17,2 15,8-18,2 5,5-6.8 6,5-7,5 7.1-10,5 10,4-13,0 '1,3-14,5

12.0-14.7 13.0-16.5 12,6-16.2

Table 12

Tooth

1,

0,18

0,14 0,16 0,25 0,23 0,22 0,20

7,60 5,09 7,86

MD olarneters 01 mandibular permanent fncisors."

Sarnple

n

x

5.0.

SE

C.V. % Observed range

A. afarensis

3 3 1 8 5 5 3 1 5 4

6,63 6,00 4,85 5,56 5,36 6,46

1.23 0,35

0,71 0,20

18.55 5.83

5,6-8,0 5,6-6,2

0,09 0,29 0,30 0,03

4,86

12,69 10.22 0,83

5.1-5.9 4.2-5.9

7,27

0,27 0.68 0,66 0,06

5.65 6.37 6.30

0,47 0,26

0,21

7,38

0,13

4,13

A. atn'canus

6.75-6,8

0,43

7,24 6,49 9,41 14,06

1,

P. rooustus P. crassidens P. boisei A. afarensis A. africanus P. robustus P. crassidens P. totsei

5,7-7,2

7,2-7.3 5,7-6,9 6,0-6,6

'Data in this ano succeecñoq rabies ter A. stsrensts and P. botsei atter White et al. (1981); data ter A. africanusincludes Stw specünens 01 Sterktonteín Member

4 denvanon.

10,0-10,1 11,5-12,7 3.8-4,7 4,6-5,8 7,9-9.6 9,7-12,0

111

NEW CRANIODENTAL HOMINID REMAINS

Table 14

Table 13

MD oramerers

Tooth

p'

Sample

A. afarensis A. africanus P. robustus P. cressidens

p'

M'

M'

P. boisei A. afarensis A. africanus P. robustus P. crassidens P txnset A. afarensis A. africanus P. robustus P. crassidens P. boisei A. afarensis A. africanus P. robustus P. cressíaens

M'

P. boisei A. afarensis A. africanus P. robustus P.

omssiaens

P. boisei

o. maxlhary perrnanent cheekteeth.

n

7 15 1 23 6 10 11 1 23 4 8 18 4 20 5 5 14 1 20 3 7 16 3 21 4

ji

S.O.

8,73 9.26 10,50

0,57 0,32

SE

0,22 O,OB

e.v.%

6,53 3,46

10,06

0,64

0,13

6,36

10,62

O,SB

B,29

8.9B

0,61

0,36 0.19

9,61

0,48

0,15

6,79 5,00

Bl tñameters 01 maxillary permanent cheekteeth.

Observed range

Tooth

7,5-9,3

p'

0,57

0,12

5,27

12,08

0,30

0,15

2.48 11,7-12,4

12,21 12.54

1,06 0,73

0,38 0,17

8,68 10,8-13,8 5,82 11,2-13,8

12,55

0,98

0,49

7,81

13,23 14,96 12,78 13,89 13,80

0,54 0,89 0,50 0,77

0.12 0,40

4,08 12.2-14,1 5,95 13,4-15,6

0,22

3,91

0,21

5,54 12,6-15,2

14,14

0,77 0,82 1,35 1,61 0,15

0,17 0,47 0,51 0,40 0,09 0.21 0,22

16,50 12,04 13,60 14,53 14,89

0,9B

16.93

0,44

5,45 4,79 11,21 11,84 1,03 6,58 2.60

p'

9,0-10,6

10,75 10,81

9,8-12,1 M'

11,1-13,2

12,1-13,5

M'

12,8-15,6 15,6-17,2

10,9-14,3 10,8-16,8

M'

14,4-14,7

12,9-16,7 16,3-17,3

p,

p.

M,

M,

M,

Sample

n

afarensis 18 africanus 7 robustus 3 crassidens 14 boisei 5 eterenste 14 africanus 8 robustus 1 crassidens 18 boisei 9 A afarensis 18 A africanus 12 P. robustus 3 P. crassidens 19 10 P. boisei A. afarensis 19 A. etnaeruse 13 P. robustus 1 P. crassidens 16 P. ooiset 8 A. afarensis 11 A. africanus 14 P. robustus 2 P. crassidens 20 P. boisei 12 A. A P. P. P. A. P. P. P. P.

S.O.

SE

C.V. % Observed range

Toorn

9,61

1,00 0,71

0,24 0,27

10,41 7,29

p,

9,74 10,25

0,76

9,4-10,9

0,40 1,66

4,08 14,96 10,31

9,0-10,5 8,9-13,0

1,00

0,44 0,11 0,74 0,27

7,42

9,81 11,10 9,68 10,06 11,90 11.26 13,52

0,60

0,21

5,96

9,8-11,3

0,56

0,13

4,97

10,5-12,5

1,76

12.90 14,14

13,02 10,1-15,6 7,67 10,1-14,6

16,53 14,06

0.99 1,02 1,07 0.73 0,97 1,05

15,70

1,00

0,59 0,23 0,29 0,62 0,17 0,31 0,24 0,28

14,70 16,50

0,82

0,21

4,97 15,2-17,7

18,16

1,33

0,47

7.32

16,4-20,0

14,62 15.90

0,76

0,23 0,34

5.20 8,05

14,0-16,3 13,7-17,4

0,30 0,49

8,01 8,56

13,93 14,81

1,28

7,21

7,7-11,1

12,1-15,4

6.37

14,4-17,0

15,5-16,4

1,36

19,98

1,71

M,

7.68 12,7-14,6 4,93 13,6-16,2 5,87 15,4-18,6

15,95

16,9B

p.

12,4-15,2

7,47

14,8-19,6 17,6-22,4

S.O.

SE

e.v.% Observed

A. afarensis A. etttcenus P. robustus P. crassidens P. beisei A. afarensis A. africanus P. robustus P. crassidens

6 13 1 16 5 6 12 1 21 P. boise; 4 A. afarensis 8 A. stttcsnus 15 P. robuetús 4 P. crassidens 23 P. boisei 5 A. afarensis 6 A. africanus 16 P. robustus 1 P. cressídene 19 P. boisei 3 A alarensis 7 A. etncenus 17 P. ronustus 3 P. crassidens 23 P. boiset 3

12,40

0,74

12,71

0,95

13,80 14,06

0,62

15,18 12,OB

13,05 15,30 15,11 16,23 13,20 13,79 13,78 14,70

1,35

0,58 0,59 0,86 1,56 0,87 0,69 0,84 0,76 1.03

0,30 0,26

5,97 11,3-13,4 7,47

10,7-14,4

0,16 0,60 0,24 0,17

4,41 8,89

13,2-15,2

4,80 4,52

11,1-12,6 12,4-14,2

5,69 9.61 6,59 5,00 6,10 5,17 6,37 4,37 5,53

13.5--16,3 14,2-17,6

0,19 0,7B

0,31 0, lB

14,65

0,64

15,3B

0,85

0,42 0,16 0,20 0,26 0,21

16,IB

13,8-17,0

12,0-14,6

12,8-15,2 12,8-14,8 13,0-16,8 14,9-17,7 13,4-15,1 13,8-17,1

15,95 16,07

0,80

0,18

4,99

14,4-17,2

18,77 13,93

2,10 1,00

1,16 0,3B

10,71 7,15

17,1-21,0 13,0-15,5

15,34

1,49

0,36

9,71

13,2-18,1

16,33 16,81

0,42 0,71

0,24 0,15

2,57 16,0-16,8 4,22 15,8-18,2

18,57

1,69

0,97

9,10

17,4-20,5

BL dtameters of mandibular permanent cheekteeth.

ji

8,2-12,6 9.0-11 ,3

ji

Table 16

Table 15 MO dtametere ot mandibular perrnanent cheekteeth.

Tooth

n

range

8,9-9,B

9,3-12.2 9,5-11,8 7,&--9,7

Sample

M,

M,

Sample

A. afarensis A ettícenus

n

18 10 P. rooustus 3 P. crassidens 15 P. ootset 5 A. etoreneie 13 A. africanus 10 P. robustus 1 P. cressiaeos 17 P. boisei 9 A afarensis 15 A. africanus 11 P. robusfus 3 P. crassidens 18 P. bOise; 6 A. afarensis 18 A. africanus 15 P. robustus 2 P. crassidens 19 P. boisei 6 A. afarensis 11 A. africanus 12 P. robustus 2 P. cressiaens 20 P. boisei 9

ji

S.O.

S.E

C.V.% Observed range

10,62 11,82 11 ,B7 11,89 13,04

0.89 1,00 0,75 0,88 0.93

0,21 0,32 0,43 0,39

10,95

0.89

11,78 13,15

0,50

0.25 0,16

9,5-12,6 8,38 8,46 10,2-13,6 6,32 11,1-12,6 7,40 10,4-13,0 7,13 11 ,4-13,7 8,13 9,8-12,8 4,24 10,8-12,4

12,94

0,89 1,28

0,22

6,89

8,96 12,3-16,5 6,78 11,0-13,9

14,28 12,69

0,42

11,3-14,5

0,86

0,43 0,22

12,65

0,92

0,28

7,27

12,43 13,66

0,65 0,83

0,38 0,20

5,23 11,8-13,1 6,08 12,0-14,7

15,48

1,17

0,48

7,56

13,51 14,52

1,01 0.69

0,24 0,18

7,48 12,1-15,2 4,75 13,2-15,5

0.91 1,11

0,21

6,08 6,56

14,35 14,61

16,43

14,4-17,6

13,8-16,2

14,25

14,96 16,92 13,38 14,40

11,3-14,2

0,93 0,91

1,09 1,56

0,45 0,28

13,0-16,5 15,8-18,6

6.95

12,1-14,9

0,26

6,32

12,9-15,9

0,24 0,52

7,46

14,0-14,7 12,6-16,2

9.49

14,7-19,2

NEW CRANIODENTAL HOMINID REMAINS

112

Table 17 MD diameters o, mandibular decicuous teeth.

n

Sample

Tooth

Ji

S.O.

S.E.

Table 18 Bl diameters of mandibular deciduous teeth.

C.V.% Observed

Tooth

n

Sample

Ji

SO.

C.V. % Observad

S.E.

ranga

di,

A afarensis A. africanus P. robustus P. crassidens A. afarensis A. africanus P. robuetus P. crassidens P. baiseí A. afarensis A. africanus P. robustus P. crassidens P. botseí

o,

cm,

cm,

A. sterensie A. africanus P. robustus P. crassidens P. boisei

2 1 1 2 3 2 1 4 1 4 3 3 6 2 2 7 2 10 3

5,25 5,50 4,70 4,90 8,43 6,50 5,00 6,03 6,00 9,18 8,83 9,70 10,38

4,8-5,7

range

di 2

A. etsrensis

2 1 2 2 2 1 4 1 4 3 3 6 2 2 7 3 9 3

P. robustus P. crassidens 0,21

0,21

4,8-5,0 6,2-6,6 6,4-6,6

3,27

0,13

0,06

2,16

5,9-6,2

0,49 0,38 0,30 0,43

0,24 0,22

5,34 4,30 3,09 4,14

8,5-9,6 8,5-9,2 9,4-10,0 10,0-11,1 11,0-12,3 11,7-12,5 10,9-12,7 11,9-12,5 12,0-13,5 13,6-13,8

d,

dm, 0,17 0,18

11,65 12,08 11,88 12,20 12,77 13,72

Swartkrans

0,73

0,28

6,14

0,45 0,10

0,14 0,06

3.52 0,73

dm,

A eterensis A. africanus P. robustus P. crassidens P. boisei A. afarensis A. africanus P. robustus P. crassidens P. boísei A. afarensis A. africanus P. robustus P. crsssidens P. boisei

Paranthropus

• ••

3

2 1 (LB) 1 (HA)

2 Member 2

1

Member 1 (HR)

C

Member 3 Member 2 Member 1 (HA)

• •

" .¡.

• • •

1,

Member 3 Member 1 (LB) Member t (HR)

12

Member 2 Member t (HR)

e

Member 2 Member 1 (LB) Member 1 (HA)

.-+-

"

,2 C

•

3 2 1 (LB) 1 (HR)

• • ••

I

+

I

6

0,50

0,25

9,75

4,6-5,8

0,34 0,38 0,31 0,65

0,17 0,22 0,18 0,27

4,28 4,99 3,80

0,64 0,31 0,70 0,23

0,24 0,18 0,23 0,13

6,29 3,09 6,46 1,90

7,6-8,4 7,3-8,0 7,9-8,5 7,9-9,6 9,1-9,5 9,6-10,6 9,0-10,9 9,7-10,3 9,7-12,0 12,0-12,4

7,71

•

•

-+•

-

• •

Member Member Member Member

• • •

3 2 1 (LB) 1 (HA)

•

Member 1 (HA)

-

•• I 5

3,8-4,7 5,8-5,8 5,3-5,7

Member 2 12 Member 1·2

Member 3 Member 2 Member 1 (HA)

4

1,

• •

Member 2 Member 1 (HA)

4,2-5,0

6,64

BUCCOllNGUAL DIAMETER

BUCCOUNGUAl DIAMETER Member Member Member Member

0,23

MESIOOISTAL DIAMETEA

MESIOD1STAL DrAMETER Member Member Member Member

0,40

Swartkrans Paranthropus Mandibular Anterior Permanent Tooth Size

Maxillary Anterior Permanent Tooth Size

l'

4,63 3,90 4,25 5,80 5,48 S,DO 5,13 5,80 7,95 7,62 8,15 8,43 9,28 10,08 10,17 10,03 10,63 12,13

I 7

I 8

e

I

I

9

10

,

I 11

I

12

,

•

Member 2 Member 1 (L8) Member 1 (HR)

••

I

I

I

¡

¡

I

i

I

13

4

5

6

7

8

9

10

j ¡ji

11

12

j

i

13

Fig.47

Fig.48

Comparison ot maxüíary anterior permanent tooth diameters recorded lor new Swartkrans Paranthropus specimens and the Member 1 Hanging Remnant (HA) sample. Vertical \íne = sample mean, norizontallme mean ± 2 S. D., horizontal bar observed sample range. Scale in mm.

Comparison ot mandibular anterior permanent tooth diameters recorded ter new Swartkrans Paranthropus specimens and the Member 1 Hanging Remnan! (HR) sample. Legend same as ter Fig. 47.

=

=

NEW CRANIODENTAL HOMINID REMAINS

113

Swartkrans Paranthropus Maxillary ano Mandibular Premolar Teelh MES!ODISTAl DIAMETER Member 2 pa Member 1 (LB) Member 1 (HA)

-..

Member 3

p4~~g:~ ~

(LB)

Member 1 (HA)

P

Member 2

3 Member 1 (HR)

Swartkrans Perenmroous Maxil1ary Permanent Molar Slze MESIODISTAL DIAMETER

.

Membvr 3 P4 Memh&r 2 Membltr 1 (HR)

• M2 Member 1 (LB) Member 1 (HA)

BUCCOLlNGUAl OIAMETER

p3Member 2

M

Member 1 (HA)

3 Member 3 Member 1 (HA)

Memb9r 3

BUCCOLlNGUAL DIAMETEA

p4Member 2 Memb9r 1 (LB) Memller 1 (HR)

M'

Member 1 (LB) Member 1 (HA)

p

M<'

Memb&f 1 (LB) Member 1 (HA)

Member 2

3 M&mber 1 (HR)

P

Member 2

M3 Member 3 Member 1 (HA) i 11

4 Member 1 (HB)

i

i 9

a

I

I

'1

10

I

11

l'

12

13

1 '1

14

15

i

i

16

17

• f

i

I

i

I

12

13

14

15

16

I

[

I

i

17

18

19

20

Fig.49

Fig.50

comoartson of maxillary ano mandibular premolardiameters reccrded

Ocmoartscn ot maxillary permanent molar diameters recoroeo ror new Swartkrans Paranthropus specimens and the Member 1 Hanging Aemnant (HR) sample. Legend same as tor Fig. 47.

ter new Swartkrans Paranthropus specimens and the Member 1 Hanging Remnant {HA) sample. Legend same as for Fig. 47.

Mesiodistal Diameters: Maxillary Permanent Cheekteeth

p3

Swartkrans Paranthropus

A. afarensis A africanus

Mandibular Permanent Molar Size MESIOOISTAL OIAMETER

P'

Member 2 M 1 M&mber 1 (LB) Member 1 (HR)

A.2{¡mmsis A. etrcenus

P. robustus P e-assidens P. boi!,eí

M'

--. +

P robustus

P. crassidens P. bo,sa!

--:I

.

~

+

~

A. 1Jfsrensís A. afncanus

~

1

P robustus p. cresstaeos P bOisei

BLCCOllNGUAL OIAMETEA

M'

Member 2 1 (LB)

M 1 Mombar

Member 1 (HR)

Momber 3 MZ Membe' 2 Member 1 (HA)

M3

A. A P. P. P

~

a!arensis afncanus robustus

.

-+-

---

CfiJSSi{fBrlS

-+-

boise;

A. a!arensis

M3

A ilfric.:lnus P. robustus

-+-

P bOlse!

I

i

i

i

i

I

11

12

13

14

15

16

i

i 17

i

i

i

I

la

19

20

+

P. crassidens

Member 1 (LB) Membllr 1 (HA)

---

1

~Iililijilii

6

8

10

12

14

16

18

20

22

24

Fig.51

Fig.52

Comparison of mandibular permanent molar diameters recorded tor' new Swartkrans Paranthropus specimens and the Member 1 Hanging Remnant (HA) sample. Legend same as ter Fig. 47.

Ccmparisons ot MD diameters of maxillary permanent cheekteeth recorded ter difterent australopithecine sampfes. Verticalline :::: sampie mean, horizontal line :::: mean ± 28.0., horizontal bar e obaerved sample range. Data ter A. afarensis and P. bolee¡ from White, Johanson and Kirnbel (1981).

NEW CRANIODENTAL HOMINID REMAINS

114

Meslodistal Diameters: Mandibular Permanent Cheekteeth

Buccolingual Dtameters: Maxillary Permanent Cheekteeth

p'

.

A. afarensis A. lImcanus P. robustus P. crassidens P. txJissi

M'

A. afsf8fIsis A. africanus p. robustus P. crassidens p. I>oisei

M'

A. afsrensis A. africanus p. robUstUs P. etaSSidens

-+--

I

s

p. robu$llJS

A. afartlnsis

.

M,

P. boIs'"

+

I

•

I

10

, , ;

12

; 14

,; , te

I

I

A afric/lt'lus p. roblJSIUS

P. crassidens

A. afsr6nslS A africarlus M 3 P. robus!us P cra:s$Jdens P boisei

;

M

boiseí

P. crassidens P. boiseí

1

aftJref1SlS

afric6nU5

roOUltus crsS$idens

A. afarensis A. afridnus

M,

II

A. africanus 3 P. robustus P. crassidens p. boisei

I

A. afa,.nSis A. P. P. P.

p4

I

p. boisei A.

p3

~

-+-1

~

afsfflnsis africanos robustus crrJS3Idens p. bois4i

I

p. boísBi

p4

------------,.

A. A. P. P.

--+--

A. alarensis A. aln'canu$ P robustus p. crassKi~

, 20

---I

I

I

I

I

22

2'

6

8

i 10

f

i 12

i

14

i

I

'

16

i

i

18

i j j 2G 22

i

I

24

Flg.53

Fig.54

Oomparisons of al dlameters of maxillary permanent cheekteeth recoroeo for dltterent australoplthecme samples. Legend same as Ior Fig.52.

Comparison ct MO diametera of mandibular permanent cneekteetn reccrced for different australopithecine samples. Legend sarwe as far

Fig.52.

Buccolingual Oiameters: Mandibular Permanent Cheekteeth

-------

A. sfarensis A.

P3

~fricanllS

.:::1=-+-

P. robustus

P

C;f8SSídfmS

P. boisei

P,

M,

A. afarensis A. africanus P. robustus P crassidl!lns P. t;Joisei

Mesiodistal Diameters: Mandibular Deciduous Teeth A. afarensis A. afriCanus P robustus

--+I

A. afarensis A. afric.anlls P. rObustus p, CTSssid9r1s P. !>Oisai

-+-

de

--1

A afarensis A. africanus 2 P robustus P. crassidMs P. r>oisei

+

A, A. P P

afargnsis africanus robustus crassidens p, boisei

•

A afarensis A. africanus

l.,il

M

•

P. cmeeaere

drn, P. robustus

P C/7JSSldens

P

A. alarensiS A. africanus M 3 P robustus

beisei

A. afa",nsis A. afric~nus

dm2 p. robustus

P. crassidens P. boisei

P. crassidBns P boísei ~I

6

8

10

11

12

14

i

16

1

i

¡i¡~

18 20

22

24

3

4

\

1

1

1

5

6

7

8

(i

9

¡

¡

¡

i

(

i

t

10 11 12 13 14

Fig.55

Fig.56

Comparison ot Bl diameters o, mandibular permanent cheekteem recorded ter ditterent australopithecine sarnples. Legend same as ter

Comparison ot MO dtarneters of mandibular deciduous teelh recorded ter diHerent austratopithecine samples. Legend same as tor Fig. 52.

Fig, 52.

115

NEW CRANIOOENTAL HOMINIO REMAINS

BuccolingualDiameters: Mandibular Deciduous Teeth A. afarensis

di 2

Fe robustus Fe crassidens

--+• -+-

A. afarensis A. africanus

dc

•

P robustus P. crassidens P. boisei

.,.• •

A. afarensis A. africanus

dm,

dm2

4'

~ .¡.

Fe robustus Fe crassidens Fe boise;

"t

A. afar9nsis A. africanus P robustus

-+-1

P. crassidens P. boisei i 3

I

I

I

I

I

I

4

5

6

7

8

9

-+-

, I , I , I , I ,1 10 11

12

13 14

Fig.57 Comparison of BL diarneters ct mandibular deciduous teeth recorded

ter ditterent australopithecine samples. Legend same as for Fig. 52.

ACKNOWLEDGMENTS I am very grateful to C. K. Brain for his invitation to study the new hominid tossits that have been brought to light through his painstaking work al the site 01 Swartkrans, and for his generDUS hospitality during my numerous visits to the Transvaal Museum. R.l. Susman provided many hours offruitful discus-

sion and debate over the new Swartkrans fossils, and I thank l. Betti and L. Jungers for the artwork and J. voketson far photographic assistance. This work was supported by grants from the L.S.B. Leakey Foundation.

REFERENCES

SRAlN, C. K, 1976. A re-interpretañcn of the Swartkrans site and lts remains. South African Journal ot Science 72: 141-146. SRAlN, C. K., 1981. The hunters or Ihe hunted? An introduction to Alrican cave taphonomy. University of Chicaga Press, Chicago. BRAIN. C. K., 1982a. Cycles of oeposttton and erosion in the Swartkrans cave deposit. Paleoecology 01 Africa 15: 27-29. BRAIN, C. K., 1982b. The Swartkrans site: stratigraphy o, the fassi/ hominids and a reconstruction of the environment ot early Homo. Prétirage ter Congrés tntemetionet de Paféontology Humaine, Nice. Vol. 2, pp. 676-706. SRA/N, C. K., 1985. Cultural and taphoncmtc comparisons of hominids from Swartkrans and Sterkfontein. In: DELSON, E., ed.. Ancestors: the hard evidence, pp. 72-75. Alan R. use. New York. BRAIN, C. K., 1988. New information about the Swartkrans cave of retevance to 'robust' australopithecines. In: GAINE, F. E., ed., Evolutionary history 01the 'robust' australopithecines, pp. 311-316. Aldine de Gruyter, New York. BRAIN, C. K., CHURCHER, C. S., CLARK, J. O., GRINE, F. E., SHIPMAN, P., SUSMAN, R. L., TURNER, A. and WATSON, V.. 1988. New evidence of early ncmlnids. their culture ano environment from the Swartkrans cave, South Afriea. South Alrican Joumet 01 Science 84: 828-835 BROOM, R., 1949. Anolher new type of fossil ape-rnan. Nature 163:

57.

BUTZER, K. W., 1976. Lithostratigraphy of the Swartkrans Formatian. South Alrican Journaf 01 Science 72: 136-141. GAINE, F. E., 1981. A new composite iuveniíe specimenof Australopi· thecus africanus (Mammalia, Primates) from Member 4, Sterkfontein Formation, Transvaal. Annals ofthe Soutn African Museum 84:

169-201. GRlNE, F. E., 1982. A new juvemle hominid (Mammalia: Primates) from Member 3, Krcmdraal Formation, Transvaal, South Afriea. Annals of the Transvaal Museum 33: 165-239. GR1NE, F. E., 1984. The dec'duous dentition ot the Kalahari San, the South African Negro and the South African Pño-Pleistocene hcminlds. Ph.D. thesfs, Untversity of the Witwatersrand, Johannesburg. GRINE, F. E.. 1985. Australopithecine evclution: the deciduous dental evidence. In: DELSON, E., ed., Ancestors: the hard evidence, pp. 153-167. Alan R. Liss, New York. GRINE, F. E., 1988. Analysis ot new craruodentat soectmens of Paranthropustrom theSwartkrans Forrnaticn. In:GRINE, F. E., ed.. Evolutionary history 01the 'robust' australopithecines, pp. 223-245. Aldine de Gruyter, New York. GAINE, F. E., 1ge9. New hominid tcssus from the Swartkrans formation (1979-1986 excavations): craruodentat specimens. America.n Journal 01 Physical Anthropology79: 409-449. GA1NE, F. E., GWINNETT, A J. and OAKS, J. H., 1990. Early hominid dental pathology: interproximaf caries in 1,5 Myr old Paranthropus

116

NEW CRANIODENTAL HOMINID REMAINS

robustus from Swartkrans. Archives 01 Oral 8iology35: 381~386. HOWELL, F. C., 1978. Hominidae. In:MAGLlO, V. J. and COOKE,.H. B. S., eds. Evofution ot Alrican memmets, pp. 154-248. Harvard University Press. Cambridge, Mass. JUNGERS, W. L. and GRINE, F. E., 1986. Dental trenos in the australopithecines: the allometry 01 mandibular molar dimensions. In: WOOD. B. A., MARTIN, L. B. and ANOREWS, P., eds, Major tapies in pdmete and human evolution, pp. 203-219. Cambridge University Press, Cambridge. KIMBEL. W. H. JOHANSON. D. C. and COPPENS. Y.• 1982. Pllocene hominid cranlal remains from the Hadar Formation, Ethiopia. American Journa/ oí Physical AnthropoJogy 57: 453-499, MANN. A. E., 1975. Paleodemographic aspects ot the South African australopithecines. University of Pennsylvania Publications in An· thropology, No. 1, University ot Pennsylvania, Philadelphia. ROBINSON. J. T., 1953. Tefanthropus and its phylcqenetic significanee. American Joumaf of Physica/ Anthrop%gy 11: 445-501.

AOSINSON, J. T., 1956. The dentition ot the Australcpitnecinae. Transvaal Museum Memoir9: 1-179. SUSMAN, R. L., 1989. New hominid toeeue Irorn the Swartkrans Forrnation (1979-1986 excavations): postcrarual specimens. American Joumst ot Physical Anthrop%gy79: 451-474. TOSIAS. P. V., 1967. The cranium and maxillary dentition al Australopithecus (Zinjanthropus) boisei. Olduvai Gorge, Volume 2. Cambridge Uruversity Press, Cambridge. WEJDENREICH, F., 1936. The mandibles ot Sinanthropus pekinensis. a comparative study. Palaeontofogia sinica {D)7: 1-162. WHITE. T. D. and JOHANSON. D. C., 1982. Pñocene horninid mano dibJes from the Hadar formation, Ethiopia: 1974-1977 colíections. American Jouma/ oí Physicaf Anthrop%gy 57: 501-544. WHITE, T. D.• JOHANSON. D. C. and KIMBEL. W. H.. 1981 Australopithecus africanus: its phyletic position reconsiderad. Souft African Jouma/ of Science 445-470.

n:

Chapter 4

Hominid PostcraniaJ Remains from Swartkrans Randall L. Susman Department ot Anatomical Sciences, School ot Medicine. State University of New York, Stony Broo«. NY 11794~8081, U.S.A.

Thirty seven fossils representing the postcranial skeleton ot Paranfhropus robustue and Horno ct, erectus were excavated from Swartkrans from 1979 to 1986. These remains come from the earfjest three ot the ñve Members at Swartkrans. This contnbution addresses the descriptiva and comparativa morphology ot the new pcstcranial tossns from Swartkrans. In addition to the new tcssils. an important proximal radnrs, SK 18b, discovered by Broom and Robinson in 1949 and never fully described, 15 discussed. The behavioural implications 01 the selected new Swartkrans postcranial fossils attributed to Paranthropus are discussed. Detaüed Iuncticnal morphology of fassils assigned te Horno erectus e presented elsewhere.

INTRODUCTION Evar sínce the first week of operatíons at Swartkrans in 1948, the site has yielded a wealth of hominid fossils. Today, Swartkrans represents the principal source of intormation on the palaeobiology and behaviour of the 'robust' austratopíthecines. While the early years at Swartkrans were marked by the recovery of mostly craniodental remains, [he test decade of excavation al Swartkrans was punctuated not only by the continued discovery skulls and teeth, but also by the recovery ot postcranial remains. The present collection from Members 1-3 provides many new insiqhts into the mass and stature, the positional behaviour (including gait and terrestriality), and the capacity for too: behaviour in the Swartkrans hominids. Following ís a descriptive account of the 'new' Swartkrans postcranial remains (as well as sorne comparisons with apes, other horninids, and a few less well-known, earlter Swartkrans postcranial remains). Detalled tunctronal accounts and behavioural inferences are provided elsewhere (Susman, 1988; Susman and Brain, 1988; Susman and Grine, 1989; Grine and Susman, 1991). MEMBER 1 REMAINS

METATARSAL 1 (SKX 5017; Fig. 2). A complete and undistorted left first metatarsal attributed to Paranthropus. The tosen is 43,5 mm long and both articular surfaces are intact, The basal surface has a prominent tubercle tor the insertion of m. peroneus Jongus and a well-developed lateral margin, both of which contribute to the mildly concave appearance of the articular surface. The shaft is thick on Its basoplantar surface, indicating the presence ot a well-developed plantar aponeurosis. In profile, the distal articular surtace (head) of SKX 5017, reveals a bread extension anta the dorsurn of the bone, whiJe in distal (end on) view lt is narrow dorsally and widens markedly Dn its plantar extent (Fig. 2).

Morphological afflnities. SKX 5017 is essentially like O.H. 8-H in size and shape and both are essentially like modern humans. The only exception is seen in the morphology of the head of SKX 5017 whlch resembles, in distal view, that of African apes and AL 333-115, rather than later hominids (Fig. 1). For a detalled dtscussron ofthe functional morphology of SKX 5017 the reader is referred to Susman and Brain (1988).

POLLlCAL DISTAL PHALANX (SKX 5016; Fig. 1). This specimen is an undistorted pollical distal phalanx, missing only a smau portian of the dorsobasal marqín of the proximal articular surface. SKX 5016 has an extensíve fossa far the insertion of m. flexor polJicis /ongus and a massive apical tuft on its distal end (Fig. 1).

HALLUCAL PROXIMAL PHALANX (SK 45690; Fig. 3). This is a complete, undistorted left proximal phalanx that measures 23,2 mm in length. The basal articular surtace opens dorsally to accommodate an extensive range of dorsiflexion. There is a prominent dorsal tuberc!e on the basodorsal surface tor the ínsertíon of m. extensor digítorum communis. The plantar tubercles tor the attachment af the coüateral ligaments are prominent.

Morpho/ogica/ affinities. The closest morphological counterparts 01 SKX 5016 are Olduvai hominid (OH) 7 (FLKNN-A) and, to a lesser extent, modern humans. The functional morphology 01SKX 5016 is discussed in detail by Susman (1988).

MorphoJogica/ affinities. SK 45690 is human-like in both size and shape. When combinad wi1h SKX 5017, it yields a length ratio identical to that of modern humans (Asían lndian sample, n = 25).

HOMINIO POSTCRANIAL REMAINS

118

Table 1 Hominid postcranrar rossus from Swartkrans. h = hand; f = '001; I-V = individual manual and pedal digits. Taxon: P = Paranlhropus; H = Horno cf. erectus.

Specimen

Member 1

üescríptron'

Excavation coordinates

Distal phalanx-thumb Metatarsall Proximal phalanx (h,? IV) Middle phalanx (h) Metacarpal I

E4/N4, E4/NS, E3/N3, E3/N4, E3/N3, E3/N3,

NW quadrant. 5S0-560 cm SE quadrant, 6S0-660 cm 650-660 cm 640-6S0 cm NE quadrant, 630-640 cm NE quadrant, 630-640 cm

E3/N4, E4/NS, E2JN4, E4/N3,

SE quadrant, S00-S10 cm SE quadrant, 730-740 cm 500-510 cm NE quadrant. S70-S80 cm

Taxon

U

SKX 5016 SKX 5017 SKX 5018 SKX 5019 SKX 5020 SKX 5021 SKX 5022 SKX 3602 SKX 8761 SKX 8963 SKX 9449 SKX 13476 SKX 34805 SKW 3774 SKW 4776 SK 45690

Middle phalanx (h) Middle phalanx (h)

Distal radius Proximal ulna Distal phalanx (h) Middle phalanx (h) Middle phalanx

Distal humerus Distal humerus Cervical vertebra C3(C)4 Middle phalanx-baüux

W5/S4, 240-360 cm

E7/52, W nau, 148 cm

P P P P P P P P P P P P P P P P

Member 2·... SKX 247 SKX 344 SKX 1084 SKX 1261 SKX 3062 SKX 3342 SKX 3498 SKX 3699 SKW 2954 SKW 3646

Metatarsal. 11I Middle phalanx (f) PateUa Middle phalanx (f) Middle phalanx (h) Cervical vertebra Trlquetral Proximal radius

Metacarpallll

E7/52, E5/N3, E7/S1, E7/S1, E5/N3, E3/N4, E4/N4, E3IN4, E8/S3, E3/N4,

W Half, 100-150 cm 175 cm 150-175 cm 75-100 cm W half, 390-400 cm SE quaoraru. 480-490 cm NE quadrant, 510-520 cm NE quadrant, 510-520 cm 100-150 cm SW quadrant, 500-510 cm

Proximal phalanx (h) Proximal pnalanx (h) Proximal phalanx (h) Proximal phalanx (h, 1II) Distal phalanx (h) Medial cuneiform Middle phalanx (h) Metatarsal V Middre phalanx (h) Proximal phalanx (h) Middle phalanx (h)

WlIS4, W2/S4, W3/S2, W2/S4, W2/S5, W3/S3, W2/S4, W2/S4, S31W4, W3/S3, S3/W3,

NW quadrant, 170-180 cm SW quadrant, 190-200 cm SW quadrant, 210-220 cm NW quadrant, 300-310 cm NW quadrant, 380-390 cm NW quadrant, 350-360 cm SW quadrant, 370-380 cm SW quadrant, 390-400 cm 490-550 cm E Hall, 450-500 cm 600-610 cm

Metacarpal IV

P

?H P

Member 3 SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX SKX

19576 22511 22741 27431 27504 31117 33355 33380 35439 35822 36712

P

?H P

P

'Aadiographs of Ihese scecímensare providedin Figs 15-18. ··Specimens SK 50, 82, 84, 85, 97, 853. 854, 14147 and 3155b, recovereeearner, are from Member 1; SK 84, 85, 853 and 3155b are attributed lo Homo. '''Specimen SK 18b also recov ered earner. ls from Mernber2.

MANUAL DISTAL PHALANX (SKX 8963; Fig. 4). The distal two-thirds of a terminal phalanx. The base is missing. The well-developed apical tuft displays slight erasion an its margins. Morph%gica/ ettinities. What little there is 01 SKX 8963 is clearly human-Hke. AJthough the apical tuft is not 'spade-like' as in most modern humana, its development relative to radioulnar midbody diameter is clearly non-pongid and recatls the candition in humans (Susman and Creel, 1979; Stern and Susman, 1983).

METACARPAL I (SKX 5020; Fig. 5). A right metacarpal, 42,4 mm in length, missing the dorsomedial part ot the basal articular surface and bady as well as the distal articular surface. There are a number ot longitudinal cracks in the cortex 01the body which extend into the medullary cavity. The specimen passesses a broad, concavo-convex basal articular surface and a prominent crest which marks the insertion of the m. opponens pollicis on the lateral margin of the body. Morphological affinities. SKX 5020 resembles the first metacarpal ot modarn hurnans. The base is relatively broad

HOMINID POSTCRANIAL REMAINS

119

and saddle-shaped, the body is stout wíth a broad, flat dorsal surface. The distal articular surface is expanded anteroposteriorly. MIDDLE PHALANGES II-V (Fig. 6). SKX S021, 9449, 36712, 3S439 and 13476 are essentially complete, undistorted manual phalanges, representing at least two individuals. SKX S021 measures 26,4 mm, 9449 measures 19,2 mm, 36712 measures 21 ,8 mm, 3S439 measures 20,7 mm, and 13476 measures 19,8 mm. AII of these phalanges ha ve well-developed markings for the insert;on of m. flexor digitorum superficialis and pronounced basal coltars for the attachment of the proximal interphalangeal joint capsules.

Morphological affinities. These bones do not appear 'bottle shaped' as do the middle phalanges of O.H.7 (although it is important to note that the O.H.7 individual was a subadult). The closest morphological counterparts to the Member 1 middle phalanges are seen in the homologues 01 modern humans. PROXIMAL PHALANX (SKX S018; Fig. 7). A complete, undístorted proximal phalanx (11 or IV) with only minor erosion of the dorsal rim of the basal articular surface and of the dorsal margíns of the trochlea. The length of SKX S018 is 33,4 mm. The basal tubercles for the attachment of the interossei and the collateral Iigaments are well developed. The scars for the attachment of the flexor retinaculum are only moderately developed. The phalanx ;s essentially straight. The included angle (= longitudinal curvature) of SKX S018 is 27" (modern humans = 26°; chimpanzees = 42°). The cortex is extremely thick anteroposteriorly (AP) in proportion to overall AP midshaft diameter.

Morphological affinities. SKX S018 is unique among the extanl Hominoidea in its mosaic of primitive and derived features. In terms of size, curvature and muscle markings lhis phalanx is human-like (Susman, 1988). The thickened cortex is a more ape-like than human-Iíke trait (Susman and Creet, 1979). RADIUS (SKX 3602; Fig. 8). This specimen represents the distal end of a right radius. The dorsum of the distal end displays a prominent tubercle of Lister and a prominent (but smaller) tubercle for the (partial) attachment of the extensor retinacuJum (which also serves to separale the tendons of mm. extensor indicis proprius and extensor digitorum communis on one side from m. extensor polJícis Jongus on the other). SKX 3602 has a bony crest for the insertion of m. brachioradialis on its lateral side. This crest extends proximally from the ante rolateral aspect of the radial styloid process for a distance of 19,5 mm. The radial styloid process is well developed. The 'sel' of the radiocarpal joint is neutral (rather than flexed).

Morphological affinities. SKX 3602 is more human-like than ape-like. The distal end has a neutral set to the radiocarpal joint (unlike the flexed set of the radiocarpal joint seen in Pan and Pongo and to a lesser extent in gorillas). The distal radius bears a close morphological similarity to Stw 27, a distal radius from Member 4, Sterkfontein. In both SKX 3602 and Stw 27, joint morphology (articulations with the di~tal ulna, scaphoid,

Fig.1 Pollical distal phalanx. Modern human (Ieft), SKX 5016 (right). Dorsal view (top), palmar view (middle), mediolateral view (bottom). AII scales =1 cm.

and lunate) is essentially human-like. The proximal radius (SKX 3699) is discussed below. ULNA (SKX 8761; Fig. 9). Proximal third of a left ulna. SKX 8761 has suffered numerous small, longitudinal weathering cracks on its surface. The trochlear notch measures 14,1 mm AP. The olecranon process mea sures 10,4 mm. There is a very deep fossa tor lhe insertíon of the m. brachialis.

Morphological affinities. SKX 8761 ;s the same size as the

HOMINID POSTCRANIAL REMAINS

120

.....

Fig.2 Ha!lucai metatarsai. Distal view of SKX 5017 (top,left), proximal view (toP. right). Plantar view, dorsal view, medial view, lateral view (middle row, left to right). Distal view 01 metatarsal I in pygmy chimpanzee, SKX 5017, and modern human (bottom row, letl lo right).

.

Fig.3 Hallucal proximal phalanx. SKX 45690. Dorsal view (Ieft), plantar view (centre), medial view (right).

Fig.4 Manual distal phalanges. Palmar view of pygmy chimpanzee, SKX 8963. SKX 27504, and modern human (Ieft lo right)

___

'0

_

••

_

HOMINID POSTCRANIAL REMAINS

121

Fig.5 Metacarpall. Palmar view 01 SKX 5020 (Ieft) and SK 84 (righl), upper left, Lateral view 01 SKX 5020 (right) and SK 84 (Ieft), llpper righL Distal view of SKX 5020 (right) and SK 84 (Ieft), lower left, Proximal view 01 SKX 5020 (Iel!) and SK 84 (right), lower right.

Transvaal Museum (TM) 1517 ulnar fragment. The olecranon process is similar in both. 80th are roughly the size of a small modern human, Among extant hominoids the closest morphological counterpart 01 SKX 8761 is lound in modern humans. HUMERUS (SKX 34805; Fig, 10), Distal portion of a right hume rus, lacking part of each condyle, the trochlea, and the capitulum. SKX 34805 has a biepicondylar diameter of 47,2 mm and a shaft diameter 0118,2 (ML) x 17,4 (AP). There is a well-defined supinator ridge which extends for 39,7 mm along the lateral si de ofthe shaft to a pointjust above the lateral epicondyle, Morphological affinities. This specimen is a reasonably good match for TM 1517 in both size and shape. On these bases it is attributed to Paranthropus, Overall, this distal humerus is like that seen in small modern humans, HUME RUS (SKW 3774; Fig, 10). Small Iragment of a righl humerus preserving the trochlea, olecranon lossa, and part 01

the medial epicondyle, The trochlea measures 17,4 mm from the medial edge to the centre of the keel while the AP diameter 01 the medial trochlea is 21,3 mm, There is a prominent zona conoidea positioned medial to the capitulum. A sharp crest separates lhe zona conoidea Irom the trochlea, The inferior margín of [he medial epicondyle forms a right angle with the medial edge 01 the trochlea. Morphological affinities, Although this specimen is very fragmentary, the trochlea and the zona conoidea are ape-Iike while the inferior margin of the medial epicondyle resembles more its human counterpart rather than that seen in African apes, CERVICAL VERTEBRA, C 3 (or C 4 ) (SKW 4776; Fig. 11), Right hall of a hominid cervical vertebra with aft-intact right pedicle, articular process and lamina. The body is lackíng the left third and the left neural arch is also missing. Anteriorly, the body has a narrow, raised ligamentous area and a larger depressed muscular surface lateral to i1. The anterior and posterior tubercles on the transverse process are

122

HOMINID POSTCRANIAL REMA/NS

1 II • • Fig.6 Manual middle phalanges. Palmar view of SKX 5021,36712,9449 and 35439 (top row, left). Palmar view 01 SKX 33355,13476,5019 and 5022 (2nd row, left). Dorsal view of same (3rd row and boHom, left). Mediolateral view of same (upper right). Proximal view 01 SKX 5021, 36712 and 35439 (3rd row right, lelt to right). Proximal view of SKX 9449,13476 and SKX 5019 (bottom right, left to right).

broken in front of the foramen transversaríum. The costotransverse bar and lhe sulcus between the anterior and posterior lubercles are narrow. The articular processes are relatively well preserved save for the inferiormost part of the inferior zygopophysis. The spinous process is Iruncated by a poslmortem break at its distal extremity. The heighl of lhe body on its anterior surface is 11,1 mm, height of the superior zygopophysis is 9,7 mm, breadth (ML) 01 ¡he superior zygopophysis is 10,0 mm, lenglh of Ihe lamina (from posterior edge of zygopophysis lo lhe spine) is 12,5 mm, heighl of ¡he pedid e is 6,6 mm, the wídlh of Ihe sulcus on the costotransverse bar (at the midpoint of the foramen transver-

saríum) is 6,3 mm and the maximum diameter of the foramen· transversarium is 3,9 mm. Morphological affínities. SKW 4776 is considerably srl\,aller than its counterpart in modern humans, afthough it is larger than C 3-C 4 in bolh male and female Pan paniscus (n = 4). The costolransverse bar appears lo be oriented more anteriorly than in Pan paniscus or a sample ofsix modern humans (Asian Indians) indicatíng a more anteriorly direcled palhway 01 spinal nerve 3 (or 4). The variable orientation of the articular processes and costal process in both human and apes confounds funclional interprelation of this fealure.

HOMINID POSTCRANIAL REMAINS

123

'.

r

{

Fig.7 Manual proximal phalanges. Palmar view 01 SKX 501 B, 19576, 27431, 22511 and 22741 (top, left to right). Dorsal view of sama (middle); mediolateral view of same (bottom).

124

HOMINIO POSTCRANIAL REMAINS

/ Fig.8 Radius. Dorsal view of pygmy chimpanzee, SKX 3602 and modern human (upper left, top to bottom). Comparison 01 radial heads: SKX 3699 and SK 18b (Ieft lo right. boHom lett). Distal radius SKX 3602 (dorsal, upper right; distal, lower righl).

Fig.9 Ulna. Anterior view 01 pygmy chimpanzee, SKX 8761 and modern human (leH to right).

HOMINID POSTCRANIAL REMAINS

Fig. 10 Humerus. Anterior view of TM 1517 (casI), SKW 3774 and SKX 34805 (len to right, lop). Posterior view of same (boltom).

125

126

HOMINID POSTCRANIAL REMAINS

...... Fig. 11 Cervical vertebra. Superior view 01 SKW 4776 (top). Inferior view 01 same (bottom).

MEMBER 2 REMAINS

1

RADIAL HEAD (SKX 3699; Fig. 8). SKX 3699 is a right radial head which retains a virtually íntact articular surface measuring 19,5 mm in diameter. The neck is long; the radialluberosily is missing. There is a strong 'bevel' which surrounds the articular surface between the caput fovea and the articulation for the annular ligament and radial notch on the urna. This bevel creales, in effect, a three-part articular surface: these surfaces contact 1) the capitulum of the humerus, 2) the radial notch and annular ligamenl 01 the ulna, and 3) the zona conoidea 01 the humerus.

Morphological affinities. SKX 3699 is essentially chimpanzeelike with its 'third' articulation with lhe zona conoidea and its long neck. It is interesling to note thal although SKX 3699 was recovered Irom Member 2, it was found in the same grid square, and al roughly the same 10 cm level, as SKX 3602. The ratio 01 SKX 3602 and SKX 3699 (SKX 3699/SKX 3602) is 0,64 and falls within lhe 95 % conlidence limils 01 the mean 01 modern human s (0,67, n = 1.5) and outside the range 01 pygmy chimpanzees (0,77, n == 15). It thus appears that SKX 3602 and 3699 may represent ends 01 a single hominid right radius. 1I so, this radius displays a unique combination 01 primitive and derived leatures. The proximal radius is very chimpanzee-Iike, while the distal end (see ábove Member 1, SKX 3602) is more human-like. PATELLA (SKX 1084; Fig. 12). Upper hall 01 a left patella. Most 01 the superior surface is intact. The mediolateral (ML) breadth 01 SKX 1084 is 30,2 mm. The posterior surface is well represented, particularly the articular portion which contacts the lateral trochlea 01 the lemur. There is considerable erosion 01 what remains 01 the anterior surface.

Morphological affinities. Clearly, the c10sest morphological counterpart 01 SKX 1084 is found in modern humans. VERTEBRA (SKX 3342; Fig. 12). Righl half 01 the body 01 a thoracic vertebra (T6-9) with only a remnant 01 the right pedicle. Vesliges of the epiphyseal plates are seen on the superior and

'SK 18b is a proximal radius fragment lrom Member 2 (Fig. 8) lound with an isolated premolar and within a few ínches 01 (he SK 15 (type) mandible 01 'Telanthropus' (Broom and Robinson, 1949)_ RObinson (1953) provided the most extensive descriplion 01 Ihe 47-mm-Iong proximal radius. Robinson (1953) noted that SK 18b, SK 15, and the P3 might belong lo the same individual. Robinson, while providing detailed descriplions 01 Ihe jaws and teeth, slaled only that '{SK 18b] wlll not be described in detail as the features are Ihose 01 modern euhominids' (1953: 477). Given its associalion with 'Telanthropus' (= H. erectus) and the numerous olher hominid proximal radii (including SKX 3699), the morphology 01 SK 18b takes on added significance over tha! accorded earlier by Broom, Robinson and others. In racl, given lhe strong similarity of SK 18b, Homo ereclus and later hominids, aM [he differences between SK 18b and earlier hominids, apes, and SKX 3699, we have evidence 01 a morphological difference between the two hominids, H. ereclus and Paranlhropus, al Swartkrans. In lhe case 01 Ihe proximal radius, Paranthropus c10sely resembles apes and Ihe australopilhecines, while H. ereclus c10sely resembles later hominids. The SK 18b radial head displays a derived articulalion for conlact wlth Ihe humeral capltulum, Ihe ulna and annular Iigament. Lacking on SK 18b is lhe bevelled, three-part artlculalion lor conlacl with 1) Ihe capitulum 01 Ihe distal humerus, 2) the zona conoidea 01 lhe humerus, and 3) lhe radial notch and annular ligamenl 01 Ihe ulna. SK 18b lacks the articulation with ¡he zona conOidea. In proximal Vlew, Ihe radial head has a thin margin and lacks Ihe ,doughnul' shaped circumference 01 Ihe more primitlve SKX 3699. While SK 18b IS larger Ihan lis counterpart in Ihe average Bushman and SKX 3699, il is smaller than Ihe average Bantu (Robinson, 1953). For a detalled comparative and functional accounl 01 SK 18b and the new Swartkrans radlus see Grine and Susman (1991).

HOMINIO POSTCRANIAL REMAINS

127

Fig.12 Palella. SKX 1084: anterior, superior, medial and lateral views (lop, left to right). Thoracic vertebra. SKX 3342: anterior, superior, laleral and medial views (2nd row, left lo right). Tri.9uetral, SKX 3498: palmar. lateral, medial, proximal and distal views (3rd row, left to right). Medial cuneiform, SKX 31117: distal, superior, medial and lateral views (bottom, left to righl).

inferior surfaces. The height of the body measures 13,7 m m al the anterior margino There is a large demi-facet above and a faint remnant 01 one below. Morphological affinities. SKX 3342 is similar in size lo its counterpart in STS 14, Australopithecus africanus. STS 14-m measures 13,2 mm in height al the anterior margín olthe body.

SKX 3342 is considerably smaller lhan íts counterpart in modern humans. TRIQUETRAL (SKX 3498; Fig. 12). A complete, undistorted righl triquetral. SKX measures 13,3 mm (proximodistally), by 8,0 mm (ML) al lIs base. The pisiform face\ is relatively small compared to the size of lhis facet in olher hominoíds, and it is

128

HOMINID POSTCRANIAL REMAINS

elliptical (long axis = 6,9 mm, short axis = 4,3 mm) and convexo The lunate surface is slightly concave and the hamate surface is faintly saddle-shaped. There is a large posterosuperior surface for contact with the triangular disco Morphological affinitíes. SKX 3498 is essentially human-like with regard to Its overall shape; the lunate and hamate surfaces match those of modern humanso The size of SKX 3498 places il in the range of small (c. 1,54 m) modern humans. The pisiform facet, however, is oval in shape, relatively small, and is neither like that of humans (Asían Indians, n = 25) nor that of chimpanzees (Pan troglodytes, n = 15).

METATARSAL 111 (SKX 247). Left base and proximal third of a left third metatarsal. One demi-facet on the medial aspecl ot the base serves as Ihe articulation tor melatarsal 11 and a single demi-facet for metatarsal IV is found on the lateral side. Morphological affínilies. SKX 247 is simrlar in morphology to modern humans. lIs size ís within Ihe range of modern humans.

METACARPAL 111 (SKW 3646; Fig. 13). Proximal half of a right third metacarpal. The shaft is curved with a ML midshaft diameler of 8,1 mm and an AP midshaft diameter of 8,9 mm. The base has a fainl styloid process, a single tace! for conlact wilh metacarpal IV, and a bipartite facet tor articulation with the capitate. Morphological affinities. SKW 3646 is in lhe size range of small modern human females (n =50). The shaft curvature exceeds that of modern humans and recalls chimpanzees. The Iines on the dorsal surface ot the shafl marking the origins of the second and third dorsal ínterosseous muscles resemble their counlerparts in modern humans. SKW 3646 ís very simio lar in morphology to Slw 68, a complete righl thírd metacarpal from Member 4 at Sterkfontein.

METACARPAL IV (SKW 2954; Fig. 13). A complete right fourth metacarpal wilh erosion of the head on Ihe medíal and lateral sides. Four large weatheríng cracks are tound on the proximal part of the shaft. There appears to be a 'crook' in Ihe specimen which imparts an abnormally strong curvature to the shaft.

fíg.13 Metacarpals. Palmarvíew al SKX 14147, SKW 2954. SK 85, SKW 3646, SK 84 and SKX 502 (top, lelt lo righl). Dorsal view 01 same (2nd row). Mediolateral views (rows 3 and 4).

HOMINID POSTCRANIAL REMAINS

Radiographs 01 SKW 2954 do not reveal evidence 01 a healed fracture but this seems the mast likely cause 01 distortion. The interosseous markings on the dorsolateral aspects 01 the shaft are marked. The basal and distal morphology are unremarkable. Morphological sttimties. SKW 2954 is smaller than any lourth melacarpal in the present human sample (n = 50). II is also smaller than 8tw 65, a right fourth metacarpal from Member 4 at Sterklontein. Otherwise, SKW 2954 is essentially humanlike in morphology. SKW 2954 and SKW 3646 are similar in srze and shape and thay articúlate well. MANUAL MIDDLE PHALANX (SKX 3062). A complete undistorted (?fifth) manual middle phalanx measuring 15,3 mm in length. The specimen is short and stout, with a thick basal collar for attachment 01 the joint capsule. The m. flexor digitorum superficialis insertions are only faintly representad. Morphologica/ affínities. SKX 3062 is short and stout as are the rniddle phalanges 01 modern humans. This specimen is considerably smaller (shorter) Ihan even the shortest (fifth) middle phalanges in the present sarnple of modern humans (n= 50). PEDAL MIDDLE PHALANGES (SKX 344; SKW 1261). SKX 344 is a complete, undistorted pedal middle phalanx. The length ot SKX 344 measures 13,2 mm. The specimen ls short and stout with a weJl-defined rjdge for the tnsernon ot the extensor assembly. The basal articular surface slopes distalwards at its dorsal edge, reflecting the capacity for a bread range of dorsiflexion at the proximal lnterpnatanqeal jolnt in this individual. SKW 1261 measures 12,3 mm in length and Is less robust than SKX 344. The base is thick (doraoplantarty} and the basal articular surface, like SKX 344, opens dorsally to accornmodate fu!l extension at the proximal interphalangeal joint. Morphological affinities. Both SKX 344 and SKW 1261 are unique morphologically and in terms of their size and proportions. SKX 344 most closely approximates pedal middle phalanx 11 in modern humans, while SKW 1261 is more attenuated than its counterparts in humans. Neither specimen resernotes pedal middle phalanges of chimpanzees or any of the other great apes.

MEMBER 3 REMAIN5

129

like attribute as is the dorsally extended articular surface of tne m~tatarsal head. MEDIAL CUNEIFORM (SKX 31117; Fig. 12). A fragmenlary left medial cuneiform which lacks the infenormost portion of the metatarsal articulation. The proximal and dorsaJmost parts of the specimen are eroded. The bone corresponds in size to SKX 5017. The distal articular surface is tlat and measures 12,7 mm (ML) by approximately 23 mm (DP). Morphological affinities. SKX 31117 is badly damaged. Notwithstanding this fact, the specimen is unique among extant hominoids (and monkeys). The distai articular surface, the fJatness of which indicates an adducted, rather than an opposable hallux, is distinctly human-liks. MANUAL PROXIMAL PHALANGES (SKX 19576, 22511, 22741,27431; Fig. 7; 35822 not utustrateo). These middle phalanges are smaller than those of modern humans. Cornputation of phalangeal curvatura ís possible far two specímens, SKX 22741 and 27431. SKX 22741 is 30,1 mm long with an inctuded angla 01 2]0 (similar to that of SKX 501 8 from Member 1). This value piaces SKX 22741 elose to the human mean (26°, n = 146). SKX 27431 is Ihe same length (approximately 30,1 mm) but has an included angle of 34°, placing it outside the 95 % confidence limits of the human mean and within the 95 % confidence limits or garillas and chimpanzees (but outside the range of pygmy chimpanzees). SKX 22511 ls a complete but broken proximal phaJanx, similar morphologically to SKX 27431. Its length measures 30,3 mm. SKX 19576 and 35822 are proximal and distal fragments respectively. While there is considerable weathering and surface erosion of fhese specimens they, none the less, appear to have only weak to moderare f1exor sheath ridges, dorsal interosseous muscle insertions, and [oint capsules. Morphological affinities. The curvature of SKX 22741 is like that of modern humans. The phalangeal curvature of SKX 27431 is differenlfrom SKX 22741 and more like that ot Atrican apes and Stw 28 from Member 4 at Sterkfontein. One expíanation far this important difference in SKX 22741 and 27431 is that these specimens sample two different taxa in Member 3, viz. Paranthropus and Homo cf. erectus respectively. This is suggested by the very species-specific nature of phalangeal curvature. Morphologically, however, the phalanges from Member 3 are human-like, notwithstanding the ape-Iike curvalure of SKX 27431.

~METATARSALV

(SKX 33380; Fig. 14). Distal two-thirds 01 a metatarsal. The sha~ is broad (ML) and Ilal dorsally. There is a pronounced curvatura (concave laterally) and a erest on the plantar surface which separated the third plantar from the flexor disiti minimi brevis. On the dorsal surface there is a faint ridge forthe insertion of m. peroneus tertius. The distal articular surface of the head extends well onto the dorsum of the bone and is flanked by prominent epicondyles, especially that on the lateral side. le~ li~h

MorphoJogicaJ affinities. SKX 33380 js essentiaJly tluman-like in morphology. The m, peroneus tertius insertion is a human-

MANUAL MIDDLE PHALANGES (SKX 36712, 35439, 33355; Fig. 6). SKX 36712 and SKX 35439 are complete and undistorted middle phalanges; SKX 33355 represents the proximal half 01 a middle phalanx. AIl three speeimens possess very prominent bases, with thick basal collars for attachment of the proximal interphalangeal joint capsules. AII three have prominent dorsobasal beaks for insertion of the central tandon of the extensor assembly. SKX 36712 and 35439 have very pronouneed impressions which mark the insertion of the m. flexor digitortJm superficialis. These specimens lack any hint of longitudinal curvature.

HOMINID POSTCRANJAL REMAINS

130

Fig.14 Metatarsal V. Dorsal view 01 pygmy chimpanzee, SKX 33380 and modern human (Ieft lo righl).

Morphological affinities. The manual middle phalanges Irom Member 3 are unique morphologically. While they are more human- than ape-like, they ha ve stouter bases (and basal collars) and better developed m. flexor digitorum superiicialis impressions than do the middle phalanges 01 modern humans.

MANUAL DISTAL PHALANX (SKX 27504; Fig. 4). This specimen lacks its lateral basal tubercles and has suffered some erosion of the apical tuf!o It measures 14,0 mm in length; 3,8 mm at midbody (ML) and has an apícal tuft that is 5,5 mm (ML). The apícal tuft is well developed. Morphological affinities. This bone is decidedly human-like in síze and shape. The pronounced apical tuf! and its developmen! relative to the body, place it well within the modern human range.

DISCUSSION Taxonomy 01 the Swartkrans hominids Taxonomic attribution 01 individual specimens is made diffícult by the presence 01 two hominid taxa in Members 1 and 2. The diffieulty is lessened, however, with regard to Member 1 by the laet that over 97 % 01 the individua! hominids from that member represent Paranthropus and only 3 % are attributable lo Horno on lhe basis 01 craniodental remains (n = 130 lor

Member 1). In bolh Members 1 and 2 where the dental evidence indicates two hominid specjes, morphological criteria can also be used for taxonomic assessment. One such exampie regards the assignment 01 SK 18b and SKX 3699. In this case there are two morphs: one (SK 18b) is associated clearly wilh the type mandible 01 'Telanthropus' (= H. erectus); the olher (SKX 3699) is more primilive (ape-Iike) and is thus assjgned to Paranthropus (Susman and Grine, 1989; see also Grine and Susman, 1991). In another example in Member 1, the pollical metacarpal is also represented by two morphs. The larger 01 the lwo, SKX 5020, is assigned to Paranlhropus, lhe smaller, SK 84, is attributed to Horno erectus (Susman, 1988). 1I the attribution 01 SKX 5020 to Paranthropus is correct, then it lollows that SKX 5016 also belongs to Paranlhropus, based on the size compatibility 01 SKX 5020 and SKX 5016, and the incompatibility 01 SKX 5016 with SK 84 (Ior details see Susman, 1988). Taxonomic assessment 01 Member 2 hominids is somewhat more problematic than that 01 Member 1 fossils. Since1981, there were but a lew hominids recovered Irom Member 2. Earlier Brain (1982) noted the relative dearth 01 Paranthropus remains Irom Member 2 and suggested the possibility that the Paranthropus (= A. robuslus) remains in Member 2 eould have been recyeled into this later Member. The hominids deseribed earlier from Member 2 inelude the type 01 Telanthropus, SK 15, parts 01 two lower premolars, SK 43 and 18a, a proximal

HOMINID PDSTCRANIAL REMAINS

radius, SK 18b, and lWo melacarpals, SK 84 and 85 (Brain, 1981, 1982). On Ihe basis 01recenlly discovered craniodenlal remains (Braln el al., 1988; Grine, 1988), lhere are now roughly ten individuals representing Paranthropus and two individual Horno identified in Member 2. Thus the statistícal grounds for allocating Member 2 fossils to either taxon are weaker than they are with respect to Member 1. On the basis 01 cornparísons with specimens that are Jikely to be those 01 Paranlhropus (see above), I attribule Ihe palella (SKX 1084) and lhe radial head (SKX 3699) lo Paranthropus. The olher material from Member 2, including rnetacarpals 111 and IV (which are similar lo SK 85 and SKW 14147), metalarsallll (SKX 247), and pedal mlddle phaJanges (SKX 344 and SKW 1261) are possibly Ihose 01 Horno. AII 01 lhe lossils from Member 3 couio balong lo ParanIhropus. Grine (1988) attrlbules all 01 Ihe new eraniodenlal remains from Member 3 te Paranthropus. There 15 direct morphological justlfication to so assign the medial euneiform (SKX 31117) sinee it ñts well Ihe lirsl metalarsal Irom Member 1 (although one musl realize the circularíty of such a rationale). On lhe olher hand, rnetatarsal V (SKX 33360) and Ihe manual distal phalanx (SKX 27504) are essenlially modern humanlike. notwithstanding their small SiZ8. An ape-Iike phalangeal curvature in manual proximal phalanx SKX 27431 indicates a primitiva retention that 15 perhaps more likely to have been manifested in Homothan in the somewhat larger Paranthropus (Susman, Stern and Jungers, 1984; Johanson el al., 1987). There rs a hint (basad on its difference from SKX 27431 and lts similarity lo SKX 5016 from Member 1) Ihal one proximal phalanx trom Member 3 (SKX 22741) might be attributable to Paranthropus.

Morphology and behaviour 01 Paranthropus in Member 1 times The prevailing notion or the habitus of Paranthropus has been that they coexisted with Horno, that tney occupied a vegetarian niche, and did not possess tool-making (or perhaps even lool-using) capabililies (Robinson, 1972). (A similar eonception of the East African australopithecines dates to the early 19605; Leakey, 1961). The larga teeth and jaws, the small relative brain size 01 Paranthropus (Tobias, 1971), ca upled with the presence of a more advanced hominid such as Horno at sites in both South and East Africa were the principal reasans that tool behaviour was not imputed to Paranthropus (9.g., see Tolh. 1967). The recovery of fossif hand bones naw attributable to Paranthropus, makes it possible to address the issue of which hominid _al Swartkrans made the tools. With the recovery of fossil hand bones sueh as SKX 5016 and 5020 (Ihumb), SKX 8963 and others which represent the fingers, we can now comment on the marphafogicaf evidence for tool behaviour in

131

Paranthropus. As such, the fossil hand banes of Paranthropus

trom Member 1 reveal a pracision grip as refinad as that seen

in hominids such as Horno habjJjs and later modero humans. The lhumb 01 Paranlhropus was stout wilh a broad saddleshaped carpo-metacarpal [omt, a well-developed m. opponens pollicis, an extensive area of insertion for m. flexor pollícis longuson Ihe distal phalanx, and a prominenl apieallutt on Ihe tip 01 the lhumb. The fingers were straighl (nol eurved as in apes), and had broad apiees, lo maten that 01 the lhumb (Susman, 1988). The metacarpus was also Iike that of later hominids, including modern humans. These trails in Ihe hand 01 Paranlhropus Indieale Ihat Ihey, like Horno habilis (see Napier, 1962a,b) were tooJ-makers. AH of the salient features indicative of refined precisión grip that are in evidence in Horno habilis (Napier, 1959, 1962a,b) are also present in the new hand bones from Member 1 at Swartkrans. II is thus plausible, given Ihe tact that Paranthropus is regarded largely as a vegetarian, that tool behavlour In the 'robust' australopithecines was an adaptation tor the procurement and pernaps processing of plant toods. lndeed, the microwear on the bone tools from Swartkrans suggests that Ihey were used lar digging (Brain, 1965), and Ihe stone tools as well, could have been usad to exploit plant resources rather than for scavenging or meat-eating, A study of the edge wear and the function ofthe Swartkrans stone tocl assemblage should help c1arlty Ihis issue. In sum, a number of different lines of evidence are suqqestlve of too! behaviour in Paranthropus. These include 1) the hand and 1001 anatorny al Paranthropus in Member 1,2) Ihe diet reconstruction based on the microscopic anatomy of the occlusal surfaces ot dental remains (Grine, 1981), and 3) the presence of bone and stone artefacts. These all point to the likelihood that Paranthropus robustus was predominantly a tocl-using (and Iikelylool-making), essenlially terrestrial biped, with a relatively small component of climbing in its locomotor repertolre compared to olher Plio-Pleistocene hominids such as Australopithecus afarensis and A africanus. At the same time. Paranthropus was largely a vegetarian, while its hominid counterparts may have shared a more omnivorous diet and used their tool behaviour as much or more for scavenging and meat eating than for plant procurement. If Parsnthropus at 5wartkrans and other 'robust' hominids in South and East Africa used tool behaviour in conjunction with their craniodental specializations for a diet composed in large part of hard, fibrous planl toods (Grine, 1981), then 100/ behaviour may have constituted an important adaptation in both the gracile and robust hominid lineages. The presence or absence of tool behaviour may not have distinguished the robust from other hominid contemporarles. Instead it may have been the uses to which the tools were put that spelled the difference between Paranthropus and early Horno in Soulh (and perhaps also East) Africa, 1,5-2,0 million years a90.1

1Geochronological evidence from East Africa indicates fhat stone tools preceed the appearence of Horno by at least 400 000 years. Al this time Paranthropus appears to be the mast common hominid at the East African archaeologicai sites.

HOMINID POSTCRANIAL REMAINS

136

ACKNOWLEDGEMENTS " I Ihank Dr C. K. Srain for Ihe opportunity lo sludy the Swartkrans fossils and for his considerable efforts to explain the geology 01the Swartkrans cave. Dr Tim Brain is responsible for Ihe inilial description of SKX 5017 which constttuted his S.sc. in Anatomy al the University of the Witwatersrand.

Dr Fred Grine conceived this project and provided valuable insiqhts on Ptlo-Pleistocene hominid evolution. This work was supported by the National Science Foundation and a Basic Research Support Grant from the NationaJ Institutes ot

Health

REFERENCES BRA\N. C. K, 1981. The hunters or the hunted? An introduction lo African cave taphonomy. University 01 Chicago Press, Chicago. BAAIN. C. K., 1982. The Swartkrans stte: stratigraphy 01 the rcssü homlnlds and a reconstruclion 01 the environment of eany Homo. In: Congrés /ntemational de Paléontologie Humaine. ter Congres, Nice, pp. 676-706. BRAIN. C. K., 1985. Cultural and teobonomic comparisons 01 hominids from Swartkrans and Sterkfontein. In: DELSON. E., ec., Ancestors: the hard evidence, pp. 72-75. Alan R. Liss lnc.. New York. BRAIN, C. K., 1988. New lnformation from the Swartkrans Cave ct reJevance to 'robust' Australopithecines. In: GAINE, F. E., ed., Evolutionary history (he 'rooust' austra/opithecines. pp. 311-316. Aldine de Gruyter, New York. BRAIN, C. K" CHURCHER, C. S" CLARK, J. D" GRINE, F. E., SHIPMAN, P., SUSMAN, R. L" TURNER, A and WATSON, V" 1988. New evidence of early hominids, thelr culture and envirenrnent from the Swartkrans cave, South Africa. SouthAfrican Joumal of Science 84: 828-835. BROOM, A. and ROBINSON, J. T., 1949. A new type of tcesn mano Nature 164: 322-323. GRINE, F. E., 1981. Trophic differences between 'gracile' and 'robust' australcpithecines: a scanning electron microscope analysis of occfusal events. South African Joumal of setene». 77: 203-230. GAINE. F. E., 1988. New craniodental tcssüs 01 Paranthropus from the Swartkrans Formation and their slqnitlcance in 'robust' austraIopunecine evolution. In: GRINE, F. E., ed., Evolutionary history ot the robust australopithecines, pp. 223-243. Aldina de Gruyter, New York. GRINE, F. E. and SUSMAN. R. L., 1991, New ParanH1ropusrobustas radius trom Member 1, Swartkrans Formation: comparativa and tunctional morphology. American Joumal o, Physicaf Anthropology

o,

84: 229-248. JOHANSON, D. C" MASAO, F. T., ECK, G. G., WHITE, T. D., WAL TER. R. C., KIMBLE, W. H., ASFAW, B" MANGEPA, P" NDESSOKIA, P. and SUWA, G., 1987. New partial skeleton of

Horno habilis from Olduvai Gorga, Tanzania. Nature 327: 205-209. LEAKEY, L. S. B., 1961. Juvenile mandible from Otduval. Nature191: 417-418. NAP1ER, J. R.• 1959. Fossil metacarpals from Swartkrans. Fossif Mammals of Africa No. 17, pp. 1-18. British Museum (Natural History). NA?! EA. J. R., 1962a. Fossil hand bones from Olduva! Gorge. Nature 196: 409-411. NAPIER, J. R., 1962b. Evolution of the human hand. Scientific Ametícan 205: 2-8. ROBINSON, J. T., 1953. Telanthropus and its significance. American Jouma/ al PhysicaJ Anthrapology 11: 445-501. ROBINSON, J. T., 1972. Ear/y hominid posture and locomotion. University of Chicago Press, Chicago. STERN. J. T. and SUSMAN, R. L., 1983. Locomotor anatomy 01 Australopithecus alarensis. American Journal ot Physical Anthrop%gy60: 279-317. SUSMAN, R. L., 1988. Hand of Paranthropus robustus from Member 1, Swartkrans: toss!l evidence for too¡ behavior. Science 240'

781-784. SUSMAN, R. L. and CREEL, N. C., 1979. Funetional and morphological atñnities of the sub-aduft nano (O.H. 7) from Olduvai Gorga. American Journaf 01 Physical Anthropo/ogy51: 311-332. SUSMAN. R. L., STERN, J. T. and JUNGERS, W. L., 1984. Aboreality and bipedality in tne Hadar hominids. Folia Primatalogica 43: 113~ 156. SUSMAN. R. L. and BRAIN, T. M., 1988. New first metatarsal (SKX 5017) from Swartkrans and the gait ot Paranthropus robustus. American Journal of Physical Anthrop%gyTI: 7-15. SUSMAN. A. L. and GRINE, F. E., 1989. New Paranthropus robustus radius Irom Member 1, Swartkrans Fcrmation. American Journal ot Physical Anthropo/ogy 78: 311-312. TaBlAS, P. V. T., 1971. The brain in hominid evolution. Columbia University Press. New York. TOTH, N., 1987. The first technology. Scientific American 2-121.

Chapter 5

Additional Fossil Equidae from Swartkrans C. S. Churcher' & Virginia Watson

2

of Zoofogy, University of toranto. Toronto, Ganada M5S 1A1 & Department ot Venebrare PaJeantology, Royal Ontario Museum, Tocante. Ganada M5S 2C6 2Transvaal Museum. P, Box 413, Preroria, 0001 South Africa

1 Oepartment

o.

Further excavations at the Swartkrans cave have yielded additicnat dental and postcranial remains of Equus capensis aro Hipparion libycum steytler¡ trcm the Member 1 Lower Bank, Member 2 and Member 3. Equus capensis was also found in Member 5. Equus burchelfii was found both in Member 3 and Member 5. Ages ot Members are estimated as follows: Member 1 at about 1,7 mitñon years ago, Member 2 at about 1,5 m.y.a., Member 3 at about 1,0 m.y.a. and a more recent age ter Member 5.

INTRODUCTION Excavations at Swartkrans by C. K. Brain during the years 1979-1986 have resulted in the recovery of sorne 48000 cata/ogued ítems that inc/ude 158 equid specimens. Brain et al. (1988) investigated the stratigraphy 01the cave and divided the breccias ínto six units, fossils from four of which are considered here: Member 1 Lower Bank, Member 2, Member 3 and Member 5 (also referred lo as the 'Antidorcas bondi channel.') The extinct Cape zebra, Equus capensis Broom, 1909, and the last 01 the Southern African hipparicns or three-toed horses, Hipperion Iybicum steytleri Van Hoepen, 1930, are found in Members 1-3. Equus capensis is also found in Member 5, and Surchell's zebra. E. burchelfii Gray. 1824. is recorded from Members 3 and 5 and possibly also from Members 1 and 2. Churcher (1970) described the equid remains recovered from the 'Swartkrans Australopithecine Site,' í.e., the Hanging Remnant of Member 1 and mixed later deposita in a report on the equids of the Transvaal caves. Churcher and Richardson (1978) reviewed the history 01Equidae in Africa and followed Hooijer (1975) in referring all the African hipparions to Hipparion. However, they considered that all the advanced African hipparions belonged within the species H. Iybicum Pomel, 1897, unless further cranial material recave red indicates olherwise, and that the Southern African torrn might be referred to as H. Iybicum steytleri. We follow the classification 01Churcher and Richardson (1978) in preference to that 01 Churcher (1970). The new material now avaiíable from the Swartkrans cave augments the previous sample of 258 teelh and 13 limb bones by an additional51 dental and 107 postcranial elements (Table 1). The postcranial specimens largely comprise ends of metapodlals and tarsals, which are important elements in species ldentiñcañon: Ior instance, metapodial characters are irnportant in distinguishing between Hipparion and Equus, and

between E. capensis and E. burchellii. A mandibular Py--M3 toothrow and an isolated M 3 of Hipparion (Fig. 2) are Ihe first mandibular elements of the genus recorded from Swartkrans, and eight lower permanent cheekteelh of E. cepensis (Fig. 4) increase the sarnple 01 mandibular cheekteeth of the species to 15. No postcraníal elements of Hipparion and no tarsals and metapodials of E. capensis were previously recorded from the Swartkrans cave. Measurements of dental and postcranial elements are given in Table 2. MATERIAL The 1979-1986 material was catalogued with an 'SKX' pre!ix to distinquish it from earlier excavated specimens catalogued with an 'SK' pref¡x, and deposited in Ihe Transvaal Museum. Cross-sections of metapodial shafts jusi proximal to the distal or phalangeal articulations were used to identify taxa ano distinguish manual or pedal rnetapcdiats (Fig. 1). The cross-sections were obtained by moulding the outer curves in stift modelling clay and measuring the wall thickness at four positions (anteriorly, posteriorly., rnedíally and laterally).

DESCRIPTIONS Al! Equus specimens recovered are mottled with red or orange haematite or black manganese stains on a creamy bone surface. The staining usually occupies mosl ot the surface and the black stains predominate. In this the new material differs Irom Ihat described by Churcher (1970) which is less stained and more yellow. Encrustanon on the occJusal surfaces impedad observation of ename¡ patterns on sorne teeth, but generally the matr¡x cleared away to reveal the details without having to resort to potentially harmful procedures. Measurements of the better preserved teeth and postcranial pieces are given in Table 2.

FOSSIL EOUIDAE

138

Hipparion Iybicum

Equus capensis

Equus burchellii

Swartkrans

Elandsfontein

Swartkrans

~

Me 11

C..._~

Me IV

SKX 3206

ELF 428

SKX 25704 reversed

Metacarpal 111

Mt 11

Mt IV

SKX 2086

ELF 510

SKX 28495

reversed antenor

Metatarsal 111

medial _ . __ , -,

~

lateral

i

posterior

Fig.1 Cross-section ot distal ends 01 metaoocnats 01Hipparion Iybicum steytJ&rifrom Swartkrans Members 2 and 3, Equus capensisfrom Elandsfontein, southwestern Cape, and Equus burchelliifrom Swartkrans Member 3. Al! sectícns are oriented with anterior to the top and medial to tne left, and taken at tne leve! 01 the proximal border 01 the distal phalangeal articulanon on the posterior lace.

Member 1 Lower Bank Hipparion Iybicum steytleri Member 1 yielded eight specimens, including a wenpreserved left M, (SKX 5832), a possible left l' (SKX 4195), lhe proximal end of a right metatarsallll (SKX 8972) and other fragmentary banas. The left M, (SKX 5832, Fig. 2e,l,g) preserves the roots and aoout 40 mm 01 the crown. The metaconid is well separated from the metastylid by a wide, shallow valley, the metaconid IS in the outline 01 an angular club, the metastylid ís triangular, and the entoconid is squarely rounded. The floors 01the flexids are smoothly convex linguaJly and the buccal faces of the protoconid and hypoconid are buccally convexo No protostylid or entostylid is presant but a th¡n, elonqate ectostylid exlsts in the distobuccal area 01 the buceal valley. Elongate areas 01 secondary dentina are visible in the centres 01 the protoconid and hypocanid, and central dats only in the centres 01 the lingual cusps. Speeimen SKX 4195 (Fig. 2h) is probably a left 13 sínce advanced Hipparion has the parallel eurygnathohippus conformation 01 the lateral inclsors and thus its shallow curvatura, sma\! stze and weü-wom crown appear to fit appropriately in an upper jaw. The specimen is 43,7 mm high, and 13,3 by 9,2 mm in mesiodistal and buccolingual diameters. The occlusal face is 20,5 mm long. The proximal end 01the right metatarsallll (SKX 8972, Fig. 7a-d), ls well preservad and broken about 65 mm into the shatt. It is anteroposteriorly deeper in the shaft than it is wide

but wider transversely than lt is deep in the articulation. Interosseous areas tor metatarsats 11 and IV extend alonq raised medial and lateral areas separated by a con cave posterior face. The interosseous area for Mt IV is narrower than that tor Mt I1 and raised more above the posterior surface cr the shatt. The lateral groove ter trie great metatarsal artery runs obliquely distally and posteriorly above a buttress that supports the facets for the faurth tarsal and metatarsal IV. No such structures exlst medially for the great metatarsal vein. The ectocuneiform facet is compased of a large, broad, crescentic anterior portion and a smaller posterior and median facet conjoined through a median isthrnus. The broken shaft shows that the lumen is small, with diameters of less than one third ofthe shaft at that point.In these characters this specimen agrees with unpublished observations made by Churcher on Hipparion Iybicum from Olduvai Gorge, Tanzania. Partial specimens of the distal end of a left metacarpal III (SKX 13642), proximal end of a right metatarsalll (SKX 6778), a right ectocuneiform (SKX 8967) lacking its lateral area. and the right slde of a ?manual ungual phalanx are also represented. The metaearpallll distal end (SKX 13642) is split and water-warn, and exhibits the prominent medial epicondylar eminence typical 01 Hipperlon. The metatarsal 11 (SKJ"E778) proximal end shows the triangular anterior and oval posterior intermetatarsa! facets, a well-developed tace: for the mesocuneítorrn and an isoiated oval facet far the ectocuneiform. The partial ectocuneiform (SKX 8967) is short anteroposteriorly (35,2 mm) and proportionately deeper proximodistally

139

FOSSIL EQUIDAE

Table 1 List cf the SKX equid specimens recovered.

a: Hipparion Jyblcum specimens Member

SKX No.

Skeletal part

4195 5967 5832 6778 8967 8972 9166 13642 453 1059 1706+1708 2265 2626a,b 3206 20024+20154

R astragalus fragment LM, A proximal metatarsaltl A ectocuneiform traqment R. proximal metatarsallll Ungual phalanx fragmento digit 11I,?manual L distal metacarpal ltt L distal metatarsalHl Ungual phalanx traqment R~, damaged L proximal metatarsallV Unqual phalanx fragments, digit 111, ?manual R distal metacarpat 111 Metapodial llar IV Iraqment

22086+22126 22747-22750+ 29171+29388

L distal metatarsallll A dentary with alveofus for Pa, P3-M 3 present: end 01 tooth row towards ramus used as tool

3 3

22936 26812 28440 26975+28976

R distal metacarpal: porcupine gnawing Ungual phalanx fragment, digit 111, ?manual L distal metapodiallll, lateral condyle L proximal rnetatarsalltl, medial section L proximal metatarsallll Ungual phalanx fragment, digit 111 L ungual phalanx fragment, digit 111 L proximal metacarpal IV

3

L1'

29023+29027+29044 30160 39183 39618

Specirnena tentatlvely identttted to taxon 1058 L? ungual phalanx fragment, digit 111 3149 Ungual phatanx fragment 15822 Navicular or ectocuneiform tarsal fragment 21971 L proximal radius, proximomedial comer 22320c R distal radius fragment

Provenance

3 3 3 3 3

E4N5 NW 0-560 E3N4 NW 600-610 E3N5 SE 570-580 E5N4 SW 590-600 E4N5 NW 650-660 E4N5 NW 650-660 E3N5 NE 620-630 E5N4 NW 490·500 E5N4 375-400 E5N2 200-225 E6N440-100 E4S2 150-200 E5N2 175-200 E5N4 SW 450-460 W2S5 SW 200-210 W3S2 NE 90-200 W3S2 NE 210-220 W3S2 NW 220-230 W3S2 SE 220-230 W3S2 SE 200-210 W2S6 NW 250-260 W3S3 NW 290-300 W3S2 NW 240-250 W3S2 SW 220-230 W3S3 SE 330-340 W5S3660-700 W5S3 NW 71 0-720

2 2 2 3 3

E5N2 200-225 E5N4 W 1/2 400-410 E3N4 SW 500-510 W3S1 SW 180-190 W2S3 SW 160-190

1

E3N5 NE 640-650 E2N4 NE 560-570 E2N4 SE 560-570 E4N4 NW 590-600 E4N5 SW 640-650 E4N7 SE 770-780 Lower Bank clase to East waü E2N2 200-300 E5N3 275-300 E1S450-100 E1S450-100 E4N4 NE 500-510 E4N4 SW 500-510 E3N4 NE 530-540 E7S3 150-175 Exploratory trench against N wall Exploratory trencn against N wall Exptoratory trench against N wall Exploratory trench against N waü N wal! ESA gulley, 0~7m Collapsed interface W2S5 NW 170-180 W2S4 SW 170-180 W2S4 SE 160-170 W2S4 SE 180-190 W2S5 NW 180-190 W2S5 SW 210-220 W3S2 SW 200-210 W3S2 SE 200-210

1 1 1 1 1

1 1 1 2 2 2 2 2 2

3

3 3

b: Equus capensis specimens 4791 6288 6314 9596 12760 15351 45561 394 892 2390 2392 3415 3466+3467+15650 3837+15840+15988 15784 16815 16832 16917 17126 38763 39545 19563 19616 19658 19802 19606 20099 22729 22935

A navicular tarsal R astragalus A astragalus, two fragments A distal tibia, free epiphysis, juvenile L astragalus fragment R proximal metatarsal IV R distal radius Distal metapodialltl, digit III

RP' L distal tibia fragmenl Distal metapodiallll condyle L calcaneum, cuboid process L navicular tarsat L astragalus fragment A trapezoid carpal L navicular tarsal A astragalus, damaged mesiodistally A fused meso- and cubo-cuneiform tarsat Ab or U<: LM1, lacking anterior half, rootless

RM' L proximal metatarsal IV

RM, RM,

?

?

? ?

2/3 2/3 3

3 3 3

RP, L metatarsal

1 1 1 1 1 1 2 2 2 2 2 2 2 2

tJl, distotateral comer

L proximal metatarsalll

3 3

A aslragalus fragment L calcaneum

3 3

FOSSIL EQUIDAE

140

Tebte 1 - continued

26048

RM'

27230 27290+27285 28313 28881

LP, LP,

29344+29356 30256 30447 31932 33202 34477 35156+35157 35429 35761 36273

36416+36417+36418 36437 37356

37479+37480 37548 37551 38058

39179+39180 39181 39182 39815 48406 41571 42258 42790 43450+43451 43552 47755

Phalanx fragment Rp3 0' 4, juvenile L proximal metacarpallV R distal tibia R calcaneum fragment, calcar lacking LP4, rootless R proximal rnetatarsal IV L metatarsallv traqment RPJ Or 4, juvenile Lp3, juvenile R caput femoris L cuboid tarsal RM 3 , rootless and damaged LP4, rootless Upper molar fragment L astragalus fragments L phalanx 1, digit III L astragalu s fragment L astragalus fragment R proximal metatarsalll and l metatarsal 111, proximolateral fragment R proximal metacarpallll L phalanx 11, digjt 111 R phalanx 1, distal condyte. digit 111 L ulna shaft

3 3 3

3 3 3

3 3 3 3 3 3 3 3

3 3 3 3

3 3 3 3

3 3 3 3

3

RI'

5

LP4. unworn A uncitorm carpal

5 5

Li2 0' 3

5

Ril o • 3 R parella

5 5

Specimens tentatively identified to taxon 15075 L navicular tarsal fragment, [uvenüe 666+669 Upper deciduous molar Iragments, juvenile. 667+668 Lp , fragmentary and worn, juvenile 19808 L proximal metatarsat I1I fragment, posteno-medfal aspect

1 2

2 3

W2S5 NW 260-270 W254 NW 260-270 W2S4 NW 280-290 W3S3 NE 240-250 W3S2 SW 240-250 W3S2 SW 250-260 W3S2 SW 340-350 W3S2 SW 320-330 W2S3 SW 410-420 W3S4 NE 360-370 W4S2 SE 400-410 W6S3 SE+SW 190-200 W3S5 490-550 W2S3 NW 450-460 W5S2310-350 W5S2 450-500 W5S3 550-600 W4S2 600-650 W3S3 620-630 W3/4S3 550-600 W3/4S3 550-600 W5S4660-700 W5S3660-700 W5S3660-700 W5S3660-700 W5S3 NE 710-720 S3W3 NE 210-220 W5N6 W 1/2 650-675 Disturbed ground next te Channel A. bondi Channel N6W6675-700 N6W6725-750 A. bondi Channel

E3N6 NW 630-640 E2N4 425-450 E2N4 425-450 W2S5180-190

e: E. burchellil specimens 16862 16865 22815+22824 25704 28495 39617

41027+41190 41191 41192

41194+41266 41201/1

41317+41332 41515 41569 42898 43475

43484 43562

43735 43759 44125 44184 44234 44629

44798

L. phalanx 11, digit III L phalanx l, digit 111 L astragalus fragment, lateral condy!e L distal metacarpal 111 R distal rnetatarsal II1 L? ungual phalanx. ? manual L ñrst phalanx, juventle L radius shatt, juvenile R distal tibia epiphysis, juvenüe L proximal ulna fragment, juvenue R distal tibia R distal radius shaft, juvenile Lp3 o' 4 in maxillary fragment Deciduous premolar fragment, juvenile Rp2, juvenile R phalanx 11, digit I1I L distal tibia epiphysis, [uvenlle R lower deciduous premolar fragment. juvenile Lp3 or4 fragment, juvenile Rp3 or 4, [uvenile Lp 2 , juvenile R astraqalus Lp30r4, iuvenile Lp30r4, juvenile L distal metacarpallll

? ?

3 3 3

3 5 5 5

5 5 5 5

5 5

5 5 5 5 5 5 5 5 5

5

Exploratory trench against N waf Exploratory trench against N wall W3S2 NW 210-220 W2S5 SW 240-250 W3S3 N E 290-300 W5S3 NW 710-720 A. bondi Channel N6W5 Upper part N6W5 Upper part N6W5 Upper part N6W5 Upper part N6W5 Upper part A. bondi Channe! A. bondi Channel W6N6 E 1/2 650-675 W5N6675-700 N6W5675-700 A. bond; Channel A. bondi Channet A. bond; Channel A. bond; Channel A. bonai Cnannel A. bondi Channel A. bondi Channel A. bondi Channel

141

FOSSIL EQUIOAE

Teote 1 - contínued

d: Equ¡d Indet. specimens

5985 6001

6407 8501

8559a 9662 9989 12757b 13494

14222 14522 15370 18072a

21313 995+996

2549/2 19456 24951 24952 25475 26353 26745 28238 29524 29940+29960 30852 32976 33872 34326 34676 35007 37213 37368 46150 46528 46722

46908

Astragalus fragment R tersar fragmenl Molar fragment Ungual phalanx fragment Upper molar fragment Femur head fragment ? Molar fragment Astragatus fragment Molar fragment Molar fragment

1 1 1 1

1 1

1 1 1 1 1

Molar fragment

Astragalus fragment Molar fragment tnclsor fragment A astragalus, damaged R proximal metacarpallll fragmenl Incisor fragment Ungual phalanx 111 fragment Distal metapodtall!l tncisor fragmenl

1 1

1 2 2

3 3 3 3

3 3

Molar fragmenl L proximal metatarsalltl fragment Distal sptint Distal metapodiallll L radius fragment, proxlrnolaterar piece L proximal uína fragment Deciduous lnc'sor Incisor

3 3 3 3

3 3

L1'

3

Distal metapodial fragment l tibia fragment, cnemlat crest Incisorfragment Lower molar root fragment Deciduous incisor Incisorfragment Ungual phalanx 111 fragment Distal splint

over both the anterior plate (11,4 mm) and the posterior peg (18,7 mm). The manual ungual phalanx is so damaged tnat the extent of the angles posteriorly are not preserved. However, the volar area beneath the phalangeal tacet is deep, and possesses two nutrient foramina that are clearly visible in plantar aspect, as is typical of Hipparion but not of Equus. Equus capensis Member 1 yielded eight specimens, including a detached distal epiphysis 01 a right tibia (SKX 9596), a damaged righ1 astragalus (SKX 6288), a damaged adult (SKX 4791) righl and a juvenile (SKX 15075) left navicular, the major portion of a ri9ht metatarsat IV (SKX 15351), and olher partial specimens. The detached, fragmentary righ! distal tibial epiphysis (SKX 9596) has been glued logelher. The right astragalus (SKX 6288, Fig. 6c-f) is massive. has lost the lateral and ventral areas 01the navicular tacet, and is assigned to E. capensis on its general morphology and large slze. Two associated fra9ments of a second large right astragalus (SKX 6314) are also assigned to E. capensis purely on robustness. The left navicular (SKX 15075) is considered to derive from a [uvenile because of its thinness, light weight and lack 01 massiveness, and because the articular facets are less cornpact than is

3

3 3 3

3

3 3 3

E3N4 NW 590-600 E3N5 SW 630-640 E3N4 NW 550-560 E4N6 SE 620-630 E4N7 SE 740-750 E4N4 NW 550-560 E4N10 SW 1230-1240 E4N5 SW 640-650 E2N4 SE 520-530 E4N7 NW 820-830 E2N5 SW 500-510 E4N7 SE 770-780 E4N3 NW 610-620 E5S4 NE 230-240 E5N4 325-350 E5N2 225-250 W2S5 NE 170-180 W2S6 NW 210-220 W2S6 NW 210-220 W5S4 SE 200-210 W2S5 NW 290-300 S6W1 NW 260-270 W3S3 SE 250-260 W3S2 SW 270-280 W3S3 NE 310-320 W3S3 SE 340-350 W4S3 NE 350-360 W3S4 SE 460-470 W4S2 NE 320-330 W4S2 SE 390-400 W4S2 SE 440-450 W5S4 500-550 W4S2 600-650 W4S2 SE 430-440 W3S4 NE 380-390 W3S3 NE 270-280 W4S2 SE 460-470

usual tor acutt equids. It is assigned to E. capensis because tlle posterolateral navicular tacets are separate and the posteromedial confluent, as in Equus but not Hipperion. The adult right navicular tarsal (SKX 4791) is representad by the medial hal1only, in which the ectocuneiform and mesentocuneiform tacets meet in a fine ridge. The right metatarsallV (SKX 15351) is preserved lar about 85 mm from the proximal end, is massive, and shows a well-developed cuboid facet shaped Iike a dumb-bell. Member 2 Hipparion líbycum steytlerí Member 2 yielded six specimens assigned te Hipparion and an additional three specimens provisionally assigned to that taxon due lo their incompleteness or lack of ideal preservatíon. These inelude distal ends 01a righl metacarpallll (SKX 3206) and left metatarsallll (SKX 453), a left proximal metatarsallV (SKX 2265), fragmenlary ungual phalanges (SKX 1058, 2626a,b) and a damaged right p' (SKX 1706+1708). The metacarpal end (SKX 3206, Fig. 5a-c) differs from the metatarsal end (SKX 453) in being flatter anteroposteriorly, with a concave area just proximal to the phalangeal articulation

'42

fOSSIL EQUIDAE

as opposed to a slightly concave area wtth an eminence running proximally from the trochlear keel. som metapodial" ends show the strong epicondylar eminences typical 01 Hipparían, and a strong proximal spur on the trochlear keel that projects beyond the posterior condylar margins, a condition infrequently found in Equus. The Iragmentary ungual phalanges (SKX 1058, 2626a,b) are assigned to Hipparion because 01 their smal1 síze. larga volar areas with larga twin faramina and, in SKX 2626, an angle that projects posteriorly beyond the phalangeal facet, An indeterminate Iragment (SKX 3149) 01 an ungual phalanx could derive from either Equus or Hipparion. The right pa Iragmenl (SKX 1076 + 1078) preserves the two tossettes and the isolated protocone.The protoconeislaterally compressed and etonqate, there is a pl¡ caballine at the junction 01the protoloph and the hypoloph, two plis protoloph, a larga pli protoconule, no pli prefossette, one pli posttossette and one pli hypostyle, Although the tooth is badly damaged, the erown measures 27,9 mm mesiodistally over the fossettes, 23,5 mm buecolingually over the metastyle and protocone, and the protocona is about 10 mm long mesiodistally. Sorne 50 mm of crown height is preserved, whieh confirms the observation that advanced Hipparion may have crowns as hypsodont as 70 mm. A laterally compressed and elongate protocone is also charaeteristic of advanced Hipparion such as H. tybicum. (Hoo/jer. 1975; Churcher and Richardson, 1978). Fragment SKX 15822 represents part 01 the margin 01 the navicular or ectocuneiform 01 a small equid, and is previsiónally assigned to Hipparion.

Equus cepensis Member 2 yielded ten specimens, many of which are fraq-

mentary. A right upper premolar (SKX 892) and a left navicular tarsal (SKX 3466 + 3467 + 15650) are relatively complete. A partiai distal articulation 01 a left tibia (SKX 2390), a well-worn and Iragmentary left p' (SKX 667 + 668), fragments 01 a left aslragalus (SKX 3837 + 15840 + 15998), a left calcaneum cuboid process (SKX 3415), isolated metapodial condyles (SKX 394, 2392) and a damaged trapezoid carpal (SKX 15784), are presenl. None 01 these specimens, excepl the premolar (SKX 892), Ihe navicular (SKX 3466 + 3467 + 15650) and dental Iragments (SKX 667 + 668), warrant descriptlon. The premolar (SKX 892) was heavily encrusted with dark red matrix, is rootless and has test part of the distobuccal area. Development in dilute acetic acid to expose the enamel pattern resulted in exposure of a damaged occlusal faee in which sueh details were poorly preserved. The crown measures 31 ,3 mm buceoIinguaHy and 31,0 mm rnesiodistally and the protocone is 11,9 mm long. The left navicular tarsal (SKX 3466 + 3467 + 15650, Fig. 6a,b) was reglued but preserves ¡ts outline and shows the separated posterolateral cuboid and closely positioned ectocuneilorm and mesocuneiform facets. The deciduous second premolar Iragment Lp' (SKX 667 + 668) preserves a rounded protocone, but is so worn and fragmentary that only lhe lack 01 the pli caballlne and the plis within the fossettes may be deduced. Molar fragments (SKX 666 + 669) represent remnants of a well-worn upper deeiduous

molar in which only the mesial margin of the protocone and protoconufe are preserved.

Member 3 Hipparion iybicum steytieri Member 3 yielded 13 specimens assigned to Hipparion. The most interesting ís a right dentary with part of the ascending

ramus and P,-M, (SKX 22747 + 22748 + 22749 + 22750 + 29171 + 29388). The other specimens included the proximal end 01 a left melatarsallll (SKX 29023 + 29027 + 29044), the distal ends 01 a righl metacarpal 111 (SKX 22936) and 01 a left metatarsallll (SKX 22086 + 22126), Iragments olthe proximal end 01 a left metalarsal 111 (SKX 28975 + 28976) and 01 the distal end 01 a left metapodial (SKX 28440), a shaft 01 a rnetapodial llar iV (SKX 20024 + 20154), the proximomedial comer 01 a right radius (SKX 21971), a righl distal radius Iragment (SKX 22320c), a damaged pedal ungual phalanx (SKX 39183), a left rnetacarpallV (SKX 39618) and two ungual Iragments (SKX 26812, 30180), The righl dentary Iragment wilh P,-M, (SKX 22747 and associateo pieces, Fig. 2a-d) preserves only fragmentary portions 01 the bone, but all live teeth are undamaged. The premolars show the rounded metaconid, elongate and compressed metastylid and squared entoconid typical of the genus. The metaeonid-metastylid link is attenuated and the lingual valley is broad and semicircular. The floors of the flexids are slightly wavy and the f1exids are transversely narrow. The buccal valieys reaeh the base of the metaconid-metastylid isthmus and elongate separated ectostylids are present. No pli caballinid is present. The molars resemble the premolars in all characters except that the buceal valleys penetrate the metaconid-metastylid isthmus to separate the flexids. Ectostylids are present on P3-M2 but not on M3. Plis caballinid are absent but slight kinks in the distal walls ot the buccal valleys probably represent their locí. The maximum height of P 4 and M 3 crowns is about 45 mm. The broken end of the aseending ramus distal to M3 has been artifieially smoothed and rounded, especially on the medial and distolateral surfaces (Fig. 2a-c, arrows) (see Chapter 8, this volume), The proximal end 01 a left metacarpal IV (SKX 39618) exhibits lateral muscular insertions that are not strongly rugose but extend distoposteriorly more than in E. caballus, E. esinus or E. Z. zebra. The metacarpal 111 lacets are proportionately slightly larger than in Equus and the shaft appears to have been more robust in section. It is therefore provisionally assigned to H. Iybicum. The proximal ends 01 the rnetatarsals (SKX 29023 + 29027 + 29044; 28975 + 28976) are worn and spalled so that details 01 the laeets are obscured. The distal end 01the right metacarpal 111 (SKX 22936) is well preserved but anciently broken on the posterodistal surface ot the lateral condyle, as well as reglued. lt has been gnawed, possibly by a poreupine, on the posterior margin of the medial condyle. The distal end of the left metatarsal 111 (SKX 22086 + 22126) (Fig. 5d-l) is chipped on the posterior surface of the troehlear and reglued. Sorne tooth marks, probably from a carnivore, are present on the anterior surface over the margin 01 the phalangeal facel. 80th specimens exhibít the strong troehlear proxlmad projeetion typical 01 Hipparion, the strong epicondylar eminenees and

143

FOSSIL EQUIDAE

TabJe 2

Measurements of a) crarual and b) postcranlat SKX equid specimens. 8. Cranlal material

Abbreviations: M = Member; 10 = tooth: MOL= mesiodistallenqth: BLW= buccolingual wldth: BLWE = bucco!ingua! w!dth at enamel; PL = protocone lenqth: McMsL :: metaconid - metastylid length; EnFL = entoñexid length; MOl = mesiodlstal diameter ct infundibulum; BLI = buccolingual dlameter of infundibulum; e = estlmated. speces H.lyb;cum

SKX Number 5832

1 2 22747 3 22747 3 22747 3 22747 3 22747 3 22747 3 2 892 28881 3 26048 3 35156 3 19802 3 27290 3 27230 3 31932 3 36437 3 19658 3 19616 3 36416 3 42258 5 41571 5 43450+43451 5 43552 5 41515 5 41515 5 44125 5 44629 5

1706+1708

E. capensis

E. biJrche1fii

M

10

MOL

BLW

BLWE

PL

LM, RP' nP, RP, RM, RM, RM, Length Pe-Pe RP"

27,2 28,3+ 24,0 22,1 19,8 21,2 24,5 = 78e 30,5 39,1 26+ 36,7

10,8+ 23,5

10,8 22,5

11,4

R p 3 0. 4

RM' RP30r4

RP, LP, LP, LP, LP. RM, RM, RM, LP, RI'

15,8 13,2 15.1 12,1 13,3 11,2 12,9 11.8 10,6 11,4 Length M,-M3 = 65,9 31e 28.2 23,0 23,0 23,6 9,2 14.5 20,5 17,5 17,3 19,5+ 21,5 18.8 17,9+ 17,9 19,7 18,0 17,0 21,1

38,2

38,9 36e 33,0

38,6 29,8

?U 3

Ri3 Lp3 o' 4

Lp" Lp' Lp30.4

McMsL

18,3 15,3+

15,3

20,9

17,9

17,4

11,1

13,7

11,8

15,7 14.7

13,2 11,7 9,7 10,4 10.5

13,0

13,0 12.8

MOl

BLI

11,7

5 4 5

11,5 10,5 12,6 19.5 15,7 17,0

20,2 18,8 22,8 15.6 15,6 14.5 19,6

15,2

35e 34,9

EnFL

10,9 17,0 17,5

18,1 15,6 15e 7,4 7,8 8.5 15,6 15,8 13,5

21,0 17,5

8.2 8,5

31.4

23,7

11,4e

32,2 37,4 30,5

22,1 21,9

20.0

12,2e 7,0

13,7

11,4

15,1

11,3

b. Pcstcranlal material Abbreviations: TvD

POQ = proxrnodfstal diameter: APD = antercpostertor dlameter; AP = anterooostertcr: Tvs = = sustentacular: diam. = diameter; ht = height; max. = maximum; mino ::: rninlmum: ame. = articulation: prox = proximal;

= transversa diameter;

transverse: Sustento

dist. = distal. AIl measurements are in mm. SKX soecímen numbers are given in the ñrst columns following identifications L = lett: R =. right.

42790 R

POO 25,9

TvO 24.1

APO 27.6

15784 R

POO 19,4

TvO 14,8

APO 20.0

Unciform carpe!

E. capensis Trapezoid

E. ceoenee

Femur - caput temorts

E. capensís

35761

R

Max. dJam. Min. diam. 64,5 60,1

Patena E. capensis

47755

R

POO 59,1

TvD 52+

TvO 54,1

APD 49,7 40,6

67,7

48,0

Tibia

E. capensis

2390 L 30256 R R 4596

Ca/canea

E capensis

L 22935 L 30447 R

APO 42,0

? E. burcheJlií

Max.

Coroncld

Sustent.

length

depth

width

Coronoid calcar

Max.

Min.

Max.

diam. calcar

otam. calcar

diam. shalt

Min. diam. shaft

49,0

35.8

48,2

21,9

38.8 34.2

47,1

22.6

36,2

3415

1 t8,3

54,3

50.5

53,1

50,0

Astraqalus

Max.

Max.

AP diam.

E. capensis

Tvs diam. 60,3

6288 R 37551 L

63,9

72.0

Cuboid

process heighl

FOSSIL EQUIDAE

145

a~ .. . "'

"

- ..-

a

_...-...J_ _..L--~

l....'

d

b '--_~--'------'

d

~'_-~-~_---'

e

L '_ _L - _ ~ _ - - '

Fig.3

Fig.2 Hipparion Iybicum el. steytleri. Aspects 01 íncisor, lower eheekteelh

and enamel patterns 01 eheekteeth. a-d: right dentary wilh alveolus for P2 , P3-M 3 present, the end 01 the preserved portion olthe ramus has been used as abone tool (arrows), SKX 22747 (including 22748 + 22749 + 22750, 29171 + 29388). Member 3, speeimen reinforced wílh plaster on the ventral margino a: enamel patterns 01 eheekteeth; b: oeelusal aspeet; e: lingual aspect; d: bueealaspeet.

Equus capensis. Ocelusal aspeets 01 cheekteeth and enamel patterns. Speeimens and dlagrams onented with lingual margins lowards bottom.

a:. SKX 28881, Rp3 or p' (deciduous). Member 3; b: SKX 35156 + 35157, Rp3 or P4 (deeiduous), Member 3; e: SKX 31932, LP4 , Member 3; d: SKX 27290+27285, LP 2 , Member 3; e: SKX 19802, RPl , Member 3. Seales

= 30

mm; b and e to same seale.

e-g: left third lower molar, LM3, SKX 5832, Member 1. e: bueeal aspeet; f: enamel pa1tern; g: oeelusal aspeet. h: left third upper ineisor, L1 3 bueeal aspeet, SKX 4195, Member 1. Scales: a-e

~

30 mm; f-h = 20 mm.

condylar pits. The metacarpal has a more concave posterior surface proximal to the distal articulation than does the metatarsal and lacks the central ridge. The ungual phalanx (SKX 39163) lacks both lateral and medial margins and both angles. It shows the clearly visible and massive volar nutrient foramina within a broad depression. the hool appears to end at the level 01 the articulation and the angles, the dorsal margin of the articulation is sharp, and no median ridge is present within the lrog, as are characteristic 01 Hípparion.

Equus capensis Member 3 yielded 35 dental and postcranial specimens and one postcranial specimen intermediate in size between E.

capensís and E. burchellii. The dental specimens include right whole (SKX 28881) and left partíal (SKX 35429) deciduous 3 third or fourth premolars, a damaged right M (SKX 26048), an upper molar Iragment (SKX 37356), a decíduous lower P3 or P4 (SKX 35156 + 35157), a left and a right P2 (SKX 27290 + 27265,19802), a rootless and newly worn left P 4 (SKX 36437), a partialleft P3 (SKX 27230), a damaged left lourth premolar (SKX 31932), a damaged and well-worn right M 2 (SKX 19658) and damaged worn, and rootless, two newly worn right M 3 (SKX 19616, 36416 + 36417 + 36418). The postcranial materials comprise an ulna shaft (SKX 48406), a detached right caput lemoris (SKX 35761), the distal end 01 a right tibia (SKX 30256), complete left (SKX 22935) and partial right (SKX 30447) calcanea, the distal hall 01 a left astragalus (SKX 37551), fragments 01 three astragali (ríght: SKX 22729, 38058; left: SKX 37479 + 37480), a damaged left proximal phalanx (SKX 37548), a left cuboid (SKX 36273), the proximal ends 01 a left metatarsal IV (SKX 19563), a left

FOSSIL EOUIDAE

146

,

.-

&; .

'.

a

b

;

~

~

-,

." . ," , . ". ..'.'" .i:l.' ,.".

.... b

e

. :.~. '~i.:~1IJ

'--_...l..-_-'-_~

•

e

d

d

e

e

~,

_ _'--_-'-_-----'

Fig.4 Equus capensis and E. burchellii. Occlusal aspects of eheekteeth and

enamel pattems. Specimens and diagrams oriented with lingual margins towards boltom. a-e: E. capensis. a: SKX 19658, RM 2 , Member3; b: SKX 42258, LPA, Member 5; e: SKX 19616, RM3, Member 3.

Fig.5 Hipparion Iybicum steytieri.

a-e: righl distal metaearpal 111, SKX 3206, Member 2. a: anterior. aspeet; b: medial aspeet; e: posterior aspeet. ., d-f: left distal metalarsal 111, SKX 22086 + 22126, Member 3. d: anterior aspect; e: medial aspect; f: posterior aspeet. Seales

=30 mm.

d-e: E. burchel/ii. d: SKX 44125, Lp2 (deeiduous), Member 5; e: SKX 44629, Lp3 or p. (deciduous), Member 5). Seales = 30 mm.

metalarsal 11 (SKX 20099), fragments of right metatarsals 1I and III (SKX 39179 + 39180), a left metaearpallV (SKX 29344 + 29356), a right metatarsal IV (SKX 33202), a left metatarsal IV (SKX 34477), a posterolateral fragment of the proximal end of a left metatarsallll (SKX 19806), the proximal end of a right metacarpallll (SKX 39181), a left phalanx IIlacking ils proximolateral angle (SKX 39182), two phalanx fragments (SKX 28313,39815), a proximomedial left metatarsal 111 fragment (SKX 19808) of small size and possibly from E. burchellii. The deciduous upper premolar (SKX 28881, Fig. 3a) is sufficiently worn to show an occlusal pattern in which plis prefossette are developing, indications 01 a pli protoloph, and of a protocone that may broaden with wear. The right M 3 (SKX 26048) is unworn and rootless, with no occlusal pattern. However, il is wíthin the size range of upper molars 01 E. capensis, but the protocone is short, as in E. burchellii. The right P3 or P4 (SKX 35156 + 35157, Fig. 3b) also shows signs of wear but the cusps are not worn enough to show patterns distinc!ive 01 any equid; it is assigned to E. capensis on the basis 01 size

and deciduous premolars that are more late rally compressed than permanent premolars. The permanentpremolars P 2 (SKX . 27290 + 27285, Fig. 3d; SKX 19802, Fig. 3e), P3 SKX 27230, and P. (SKX 31932, Fig. 3c; SKX 36437) all show shallow buccal valleys; all except SKX 36437 show pointed metastylids and enloconids while SKX 27230 and 31932 (Fig. 3c) show rounded metaconids. The t100rs of the flexids are wavy, crested or smooth, exeept in SKX 36437 where, because of its newly-worn condition, there is a strong bifurcated pli lrom the entoflexid floor. No plis caballinid are present although kinks in the posterior walls 01 the buccal valleys in SKX 31932 and 27230 represent vestiges of these plis. The lower molars (SKX 19658, Fig. 4a; SKX 19616, Fig. 4c) possess the broad buccal valleys that penetrate to between lhe f100rs of the flexids, a characteristic of E. capensis molars. The floors of the f1exíds are genlly crested, the metaconids round, melastylids pointed and entoeonids squared (or reduced in M 3 ). The poslcranial elements are robust and well preserved. The caput femoris (SKX 35761) is massive, somewhat damaged, and shows a deep and strong triangular fovea, the distal end of the right tibia (SKX 30256) was reglued, a left

FOSSlL EQUIDAE

a

a---..

147

b

e

e

e

Fig.6 Equus capensis. a-b: left navicular tarsal bone, SKX 3466 + 3467 + 15650, Member 2. a: proximal aspeel; b: distal aspeet.

e-f: right astragalus, SKX 6288 Member 1. e: medial aspeet. d: lateral aspeet; e: tibial aspect; f: ealeaneal aspeet (damaged). Seales

=30 mm.

calcaneum (SKX 22935, Fig. 7e-h) is enlire, Ihe calcar af the right ealcaneum (SKX 30447) has been chewed off, the tibial and calcaneal faeets af the left aslragalus (SKX 37551) have been broken away, the lateral half of the distal artieulation al the proximal phalanx (SKX 37548) is missing, and the distal ends 01 the splint metapodials (SKX 19563, 20099, 33202, 29344 + 29356) are missing. The phalanx 11 (SKX 39182) is massive, robust and possesses foramina arranged as in Equus ralher than in Hipparion. The proximal end of right metacarpal lit (SKX 39181) has the remains 01 metaearpalll lused to the shaft and a roughened interosseus surface far melacarpal IV. It is interpreted as deriving from an old individual. Similarly the proximolateral areas of left metatarsals 111 and IV show fusion 01 the diaphysis of the splint metatarsal to that of the eannon bone. A fragmenlary right distal comer of a proximal phalanx (SKX 39815) is also present.

Equus burchellii Member 3 yielded four (possibly five) speeimens attributable to E. burchellii. Fragments ot a left aslragalus (SKX 22815 + 22824), distal ends of a left metacarpal 111 (SKX 28495) and

Fíg.7 o a-d: Hipparion Iybicum sleytlerí, right proXimal metatarsal 111, SKX 8972, Member 1. a: anterior aspect; b: lateral aspeet; e: posterior aspeel (a-c to same scale); d: proximal artieulation. e-h: E. capensis, lel1 caleaneum, SKX 22935, Member 3. e: anterior aspeet; f: posterior aspeet; g: lateral aspeel; h: medial aspect (e-h to same scale). Seales = 30 mm.

right metatarsal 111 (SKX 25704), a left ungual phalanx (SKX 39617), and the proximomedial angle of a left me1atarsal 111 (SKX 19808), which is intermediate in size between E. capen· sis and E. burchellíi but may derive from a small E. capensis. The metacarpal end (SKX 25704, Fig. 8a-e) shows a Ilattened surface proximal to the phalangeal arliculation, and the metatarsal end (SKX 28495, Fig. 8d-f) shows a flaltened area in the centre of which is a rounded eminence (not linear as in E. capensis). The trochlears of both metapodials are not prolonged proximally on the posterior surface as in E. capensiso A single ?Ieft ungual phalanx 111 (SKX 39617) is probably from E. burchellii: the volar insertion is deep and the volar foramina are just hidden by the semilunar crest; the anterior aspeet ot the hoot is finely rugose. Interface between Members 2 and 3 Two speeimens were reeovered fram deposits that eollapsed from the wall between Members 2 and 3. These

FOSSIL EQUIDAE

148

a

b

e

e

d

9

h

Fig.8 Equus burcheJlii.

a-e: left distal metacarpallll, SKX 25704, Member 3. a: anterior aspect; medial aspect; e: posterior aspect (a-c to same scale).

b:

d-f: right distal metatarsallll, SKX 28495. Member 3. d: anterior aspect; e: medial aspect; f: posterior aspect (d-Ito same scale). g-i: right phalanx 11, digitlll, SKX 43475, Member 5. g: anterior aspeet; h: proximal aspeet; 1: posterior aspect (g-i to same scale). Scales

=30 mm.

specimens cannot be assigned to eilher 01 the adjoining Mem· bers with certainty and are thus noted separately.

Equus capensis A right M 3 (SKX 39545) that lacks its ectoloph and protoloph, and a rootless left M, (SKX 38763) that lacks its protoconid and metaconid-metastylid, were recovered. Member 5

Equus capensis Member 5 yielded six specimens including the ename! crowns 01 a left upper or right lower second or third incisor (SKX 43450 + 43451). an enamel crown of a right i3 (SKX 43552), a rootless right 12 (SKX 41571), a rootless left P~ (SKX 42258), a right uncilorm carpal (SKX 42790) and a damaged right patella (SKX 47755). The caps 01 the incisor (SKX 43450 + 43451,43552) show

well-developed inlundibula without internal plis. but plis are 2 rare in deciduous incisors. The right 1 (SKX 41571) is maturel) worn and shows an infundibulum. The worn but rootless lef p~ (SKX 42258, Fig. 4b) has not developed a mature ename pattern, but the rounded metaconid, pointed metastylid, anc squarish entoconid may be envisioned as developing later. ¡. well-developed pli caballínid ís present. the floor 01 thE metallexid is wavy and that of the entollexid convex lingually as occurs in some E. capensis. The right unciform (SKX 42790) is entire and typical. ThE right patella (SKX 47755) lacks its medial third, is massiveanc shows deep ligamental insertions.

Equus burchellii Nineteen specimens of this taxon are recognized Irom Mem ber 5. Elements 01 possibJy four juveniles, including both teet~ and postcraníal elements, derive mainly from one individual Association 01 the dental and postcranial elements is insecure Dental elements which could be associated are newly-worr p3_ p• in a maxillary fragment (SKX 41515), a worn right p: (SKX 42898), an unworn left p3 or p. (SKX 43735), unworr right P3 or p. (SKX 43759), worn left p2 (SKX 44125), worn let P3 or p. (SKX 44629), a worn right P2, P3 or p. (SKX 43562), e worn left P3 or p~ (SKX 44234) and a lower deciduous premolal fragment (SKX 41569). Postcrania! elements are a left dista metacarpallll (SKX 44798), a right phalanx 11 (SKX 43475), é distal tibia fragment (SKX 41201/1) and a damaged astragalu~ (SKX 44184). Juvenile elements include a left radial diaphysi~ lacking the proximal end (SKX 41191). a right radial epiphysi~ and diaphyseal fragment (SKX 41317 + 41332), a left ulnal coronoid and shaft Iragments (SKX 41194 + 41266), a lel phalanx I with an unfused proximal epiphysís (SKX 41027 .¡ 41190), a right phalanx 11 (SKX 43475), and damaged left anc right tibial epiphyses (SKX 41192,43484). The deciduous teeth p3_p• (SKX 41515) and p2 (SKX 44125 Fig. 4d) are little worn but derive from different individuals. Th€ p2 (SKX 44125) has no pli caballinid, but a single pli protolo phid, and no other plis. The ectoloph has a flat paracona valle y, strong and broad metastyle, arcuate metaconal valley and a small distobuccal style. 80th the protocone and hypo cone are short and bulbous. The associated p3 and p' (SK) 41515) are so líttle worn that no plis are detectable, and onl: the relative elongation 01 the protocone and hypocone may b. seen. The ectoloph styles are heavy and the valley is open but theirconformation ís unclear. The P3 or p~ (SKX 44629, Fi¡; 4e) is worn to show a pointed bucea! valley penetrating be tween the f1exid Iloors, a compressed and elongate melaconic a metastylid and entoconid, a paraconid reaching the lingUe surface, and a right-angled bend indicating the ptychostylic Specimens SKX 41191, 41317 + 41332. 41194 + 41266, 4102 + 41190 and 41192 probably derive from a single individue because of situation, preservation and provenance. Thus tw juvenil e individuals are represented by the postcranial ev dence and four on the dental evidence. The left distal mele carpal 111, SKX 44798, is well ossilied, typical and shows preserved dry break at midshaft. The phalanx 11 (SKX 43475, Fig. 8g-i) is small and lightl built and possesses markedly raised eminences for th collateralligaments, suggesting an old individual, and reserr bIes that 01 modern E. burchellii in size.

149

FOSSIL EOUIDAE

Table 3 The Swartkrans cave Equidae: oistnbution ot me ldentftlable specimens endtaxa by unlt, within subdlvtsion of minimum numbers ot lndividuals :MN1) into adult and juvenile categorías. Figures in parentheses below Member 1 indtcate the number 01 teeth ano Individuals reporteo from tne AustraJopithecine Site' by Churcher, 1970.

Specimene

MNI

Member 5

Equus burcnetñt

Equus capensis

Hipparion fybicum steyíteti

Umt ano age

Specimens

MNl

Specimens

MNI

6

2 tadult

19

6 or 7 2 or 3 adults 4 juveniles

4 or 5

1

1 juvenlle Member 3

13

1 1 adult

35 or 36

2

Member 213

5 3 adults 2 juveniles

1 adult

1 1 adutt

Member 2

Member 1

(Cnurcner. 1970)

9

a

(1 )

TOTAL Specimens: 1140r116 MNJ: 23 or 24 Adults: 130r14 Juveniles: lD

1 1 adult

10

1 1 adult

a

(1)

(15)

3 1 adutt 2 juveniles

3 2 adults 1 juvenlte

(13)

(13)

(12)

30 3 3

61 or 62 14

a

7 or 8

O

6

3 or 4 4

Other material Spedmens were atso conectec from test trenches and it is uncertain to which Memberthey beleng. They include the cap al a deciduous Ri, cr Li', damaged buccally (SKX 17126), Irom an exploratian trench against the back wall about 11 m below the Member 2, a damaged right astraqalus (SKX 16832), a left navicular (SKX 16815), a right meso- and entocuneiform (SKX 16917), all from E. capensis, and aasociated left phalanges I and 11 (SKX 16865, 16862, respeclively) Irom E. burchelliifrom an exploratory trench along the north wall of the cave. Measurements ot the more complete specírnens are included in Table 2b. DISCUSSION The identification ofthe sarne equid taxa ín the stratigraphicalfy controlled additional sarnple from the Swartkrans cave reaffirms the conclusions from an ear1ier analysis (Churcher, 1970). Table 3 gives the distribution of specimens by taxon. level, and adult and [uvenüe cateqories, but note that Hipparion libycum steytleri íe not recordad from Member 5. Churcher (1970: tabíe 7) recorded onty one mature individual 01 Hipparion steytleri, no E. burchellií, 13 E. capensis and 12 E. quagga individuals. in the last two specles ranging in age from colts to very old. No temporal sequence for these taxa was suggested for individual caves, although a sequen ce far the fauna fram the caves was suggested (Churcher, 1970: table

23 or 24

8) following Well's (1962) temporal arder ter the caves. Ctturctter (1970: 151) concluded tha! the 'specíes of Equidae from the Krugersdorp cave deposits do not form a sequence in which faunal succession can be observed. The only suggestions are that Equus burcneltti and Equus quagga may not have been present at the beginning and that Hipparion steytleri and E. cspensis appear to have existed in the area and throughout the time interval during which fossils were being deposited.' Churcher's (1970) report al E. quagga from the Swartkrans and Kromdraai sites ls conaidered erroneous and thase specímens are now assigned to E. burctiettii. The single tooth from Kromdraai (KA 1908) was not securely identified as E. quagga as the only characteristic of E quagga tnat it exhibíts is the slightly convexly rounded walls to the ectoloph valleys. The other E. burcheflií (ex E. quagga) dental e/ements reported may derive from mixed later stratigraphic units but information on their provenance is not considered relíable. Churcher and Richardson (1978: table 20.2) collated the characters used by earlier authors to separata the teeth ot zebra s ¡nto taxonomic groups. While)t is difficult to distinguish E. quagga from E burchellií en the basis ot cheekteeth, as variation in au characters generally overlap, it now seems that E. quagga material earuer identified from Swartkrans should be reassigned ta E burchelJii primarily because there is no evidence that E. quagga occurred in the area between the Transvaal and the southwestern Cape.

150

FOSSIL EOUIOAE

It has been suggesled thal E. burchel/ii (Gray, 1824) (the., camman, plains or Burchell's zebra) is a synonym 01 E. quagga . Boddaert, 1785 (the quagga or half-striped Cape zebra). Aau (1974) first suggesled lhis on the basis 01 morphology and coat colour, and Lowenstein (1985) and Lowenstein and Rydar (1985) on Ihe basis 01 serology. Aau (1986) further advanced the argumenl and Thaekeray (1988) addueed evidenee from the entoflexid lengths o( the fourth lower permanent premolars that supports their synonymy. The status 01 these twa taxa is therefore unclear, 11, as seerns likely, E. quagga represents the southernmost variant 01the 'E. burcnetlii, 'cñne, the new name cornbination for the northern Southern African cammon zebra populalion will be E. quagga antiquorum (H. Smith, 1841), with two extmct raees or subspecies, E. quagga burchellHin central Southern Africa and E. quagga quagga in the Cape coastal mountains, being recognized. Tl1e Member 1 Hanging Remnant equids have not been Included in the sample of equids reported here because they cannot be reliably identified. While the absence of identifiable tossils o/ a taxon does not prave its absence, the absence of Hipparion libycum steytJeri from Member 5 and the possible absence of c. burchelliifrom Members 1 and 2, suggest that a tim e-transgressive faunal exchange did take place in which Hipparion libycum steytleri was replaced by, or gave way to, E. burchellii. The lack of E. burchellii specimens from Member 2 ls problematic in that two of 36 specimens were identified as E. burchelfiifrom Member 3, and the 14 speclmens from Member 2 might include at least one E. burchellii specimen. Equus ofdowayensis is known from aoout 1,8 m.y.a. (Bed 1,

Olduvai Gorge) and 1,9 m.y.a. (Member G, Shungura Formation, Omo deposits) until Recent in East Africa (Churcher, 1981). Equus capensis is conaldered to be the southern form of E. otaoweyensis and is known from the Late Pliocene or Early Pleistocene at Langebaanweg (Baard's Quarry) until the Early Hotocene at Equus Cave in the Buxton Limeworks at Taungs, Cape Province, and at Wonderwerk Cave, giving effectively the same temporal distribution as for E. oJdowayensls in reference to the Transvaal caves. Hipperion Iibycum, which became extinct during the Upper Pleistocene, is recordad from all tour beds in Olduvai Gorga, Makapansgat Umeworks Cave, and the later Omo Beds and Kaiso later fauna. lt ls not known from Elandsfontein (± 400000 years B.P.) or later dates. Thus the presence of an advanced Hippericn in all but the Member 5 deposits corresponds with the known time of extinction of the Afriean hipparions al aboul600 000 - 500 000 years B.P. Equus burchellii is known from Laetoli and Otduvai Gorge at about 2,5 and 1,8 m.y.a. respectively, but not from the Omo Beds. lt is also known from Makapansgar Cave of Hearths at about 17000 years B.P. (Mason, 1962). Thus E burchefliimay weu have been present in the Swartkrans cave area during the period 2,0-1,0 m.y.a., but the sample size is small, the elements not reliably diaqncstic, and the number 01 indlvlduals of both Hipparion Iibycum and E. capensis only seven. which could allow tor chance exclusion of this zebra. 11 this were so, then the E. burchelJii specimens reported from the Swartkrans Hanging Remnant breeeia (Churcher, 1970) may in faet derive from a time coeval with the Lower Bank ot Member 1.

REFERENces BRAIN, C. K., CHURCHER, C. S., CLARK, J. O., GRINE, F. E., SHIPMAN, P., SUSMAN, A. L., TURNER, A. and WATSON, V, 1988. New evidence 01 early hominids, thefr culture and envirenment from tne Swartkrans cave, South Alrica. South Atrican JournaJ ot Sciencs 84: 828-835. CHURCHEA, C. S., 1970. The fossil Equidae of the Krugersdorp caves. Annals 01 rhe Transvaal Museum 26(6): 145-168. CHUACHER, C. S.. 1981. Zebras (genus Equus) from níne Quater· nary sites in Kenya, East Africa. Canadian Journal 01EarthSciences 18:(2) 330-341. CHUACHEA, C. S. and RICHAADSON, M. L., 1978. Equidae. In: MAGUO, V. J. and COOKE, H. B. S., eds, Evofution o( Alrican mammals, pp. 379-422. Harvard University Press, Cambridge. HOOIJEA, D. A, 1975. Miocene to Pleistoeene hipparions 01 Kenya, Tanzania. and Ethiopia. Zoologische Verhandelingen 142: 1-80. LOWENSTEIN, J. M., 1985. Half·striped quagga was a plains zebra.

New Scientist 1465: 27. LOWENSTEIN, J. M. and AYDER, O. A., 1985. lmmunoloqical systematics of the extinct quagga (Equidae). Experiantia 41: 1192-1193. MASON, A., 1962. Preh;story ot tne Transvaal. Witwatersrand Unlversity Press, Johannesburg. RAU, A. E., 1974. Aevised list of the preserved material of the extinct Cape Colony quagga (Gmelin). Annals o( the South Afr;can Museum 65: 41-48. AAU, R. E., 1986. The quagga and its kin. Sagittarius 1(2): 8-10. THACKERAY, J. F.. 1988. Zebras Irom Wonderwerk Cave, northern Cape Provinee, South Afriea, with a potential distinction between E. burcheJlii and E. zebra. South Alrican Journa/ o( $cience 84(2):

99-101. WELLS, L. H., 1962. Pleistocene faunas and the distribution of mamma!s in Southern Alríea. AnnaJs 01the Cape Provincial Museums 2:

37-40.

Chapter 6

New Fossil Carnivore Remains from Swartkrans A. Turner Transvaal Museum,

P. O. Box 413, Pretoria, 0001 South Africa ~

New finds of tossu camfvore remains from the three oldest Membersofthe Swartkrans hominid stte are described and díscussed. The assemblage from Member 3, the youngest. contains the latest known African récords 01the dirk-toothed cat, Megantereon cultridens, and the hunting hyaena, Chasmaporthetes nitidu/a. The most common 01 the large preoators throughout the sequence Is the leopard, Panthera pardus, which appeers to nave been ot greater average size than módem !eopards in the same area. Members 1 and 3 contain remains 01 a large form of Prote/es, the aardwolf. AlI three Memberscontain remame of the ctawless otter, Aonyx cepensis. the

first occurrences ot the spectes in the hominid sites in the Sterkfontein valtey.

INTROOUCTION Previous detailed o'scussfons of the carnivore remains from Swartkrans have been given by Ewer (1955a-e, 195Ga-c, 1967), Hendey (1974a) and Turner (1984a,b, 1986a, 1987a). Those discussions were confined to material recovered from the site during earlier phases of work there, as summarized by Srain (1981). Continued excavations at the Site undertaken by C. K. Brain have produced further quantities ot rnarnrnalian fossils. These most recent investigations have led to significant refinements in our understanding of the site stratigraphy, and the new fossils are now seen to derive from a larger number of discrete tevels than (hose proposed in the o/der, more simple sequence (Brain, 1981). Previous efforts to date the deposits and their contents have relied on fauna! comparisons with dated East African assernblages (Vrba, 1985), and despite attempts to apply thermoluminescence dating (Vogel, 1985), lt seems that such relative approaches will continue to provide the best indication of age for sorne time to come. The purpose of this present paper is to record and describe the new carnivore remains from the earliest deposits at the site, Members 1, 2 and 3. In most cases, identifications are pcssible to species level, but sorne have been made only to the genus while a number of specimens could onty be identified ro family. MATERIAL ANO METHOOS !dentifications were based on direct comparisons with tossu collections and modern material held in the Transvaal Museum, Pretoria. Metrical comparisons were made with this body of material and with fossil and modern spectmens from a number of institutions in South Africa and Europe. Details are given where appropriate. Unless stated to the contrary, all measurements were taken by myself, using standard vernier ·Present address: Department 01 Anatomy and Cell Biology, University of Liverpool, Liverpool L69 3X, England.

callipers and a stee! tape. Al! quoted measurements of specimens are in milJimetres. Each new fossil specimen is referred to by its unique cataloque number assigned by Brain in the course of lnitial processing, and the prefix SKX ls applied to distinguish between the new material and that from the older warkings at the site. The older material is prefixed by the letters SK, or in sorne cases SKW, a designation made clear in any discussion ot old and new material. The specimens are cescribed systematically within each of the new Members delined by Brain (Chapter 1, this volurne). The carnivore remains from the recent excavations will be housed, together wlth other non-primate remains, at the Bernard Price lnstltute for Palaeontological Research (BPI), University of the Witwatersrand, Johannesburg.

MEMBER 3 SPECIMENS Family HYAENIDAE Chasmaporthetes nitidula (Ewer, 1955)

MATERIAL. SKX 29205, a right p3; SKX 22992/22972, a right p'; SKX 25775, a left p3 anterior fragment; 21615, a right P, anterior fragment. DISCUSSION. The nomenclatural history of African material now referred to the genus Chasmaporthetes was summarized e!sewhere (Turner, 1987b). The new specimens compare well with the sample of C. nitidula from the older workings at Swartkrans, measurements ofwhich are given in Tables 1 and 2 together with those of the new Member 3 specimens complete enough for measurement. The upper carnassial. 22992/22972 (Fig. 1A), ls relatively small, with a small but well~developed, forward~sloping proto~ cone and a marked absence of any internal cingulum on the metastyle. Specimen 29205 has a well-developed lingual bulge at the centre of the crown base and a prominent anterior cusp (Fig. 1S). Both 01 the tragmentary premolars show the

152

NEW FOSSlL CARNIVORE REMAINS

Fig.1 ChasmapOr1hetes teeth. A == SKX 22992172, right p4; 8 = SKX 29205, right p3; e = SKX 637, right P4; 0= SKX 635, righl P2. A and B lingual views. e, o and E bucea! views. AII natural slze.

E

characteristic development 01 the anterior cusp and the marked double curvatura 01the anterioredge 01the maincusp. Three 01 the teelh, 29205, 22992/22972 and 21615, are of sufficiently similar slze and wear to be from one animal. The left p3 fragmant is probably from a younger animal, giving a minimum 01 two individuals.

The p4 is from a large and robust hyaena. The shape and angle or the cusps do not agree with H. brunnea, ano there is a considerably better match with C. crocuta. The sale diHerence from the available comparative sarnple of Crocuta lies in the presence ot a well-defined cingulum on the protocone.

Hyaena brunnea Thunberg, 1820

Prote/es Geoffroy, 1824

MATERIAL. SKX 30824/30822/30820, a righl maxillary frag· rnent with pe and p'; SKX 30401, a lelt P,; SKX 20180/20176/20181, a right P,; SKX 35418, a Iragmenl of Ihe crown 01 a right p'; SKX 31535, a portian 01 lhe metastyle of a left p'; SKX 37303, a posterior fragmenl 01 a left p'; SKX 37433, an anlerior fragmenl of a left p3; SKX 38256c, a posterior fraqrnent 01 a left p3; SKX 31969, an anterior fraqmenl 01a rlght P3: SKX 34159, a left M,; SKX 20182, the anterior portian 01 a right M,: SKX 31858, a fragmenlola left uppercanine; SKX 30831, a righf M'; SKX 26268, a proximal portian 01 a righl lhird metatarsal; SKX 36942, a second phalanx. DISCUSSJON. The isolated and Iragmentary teeth can be identifiad as brown hyaena by their general robustness and relalively non-hypsodont crowns.Measurements are given inTables 3 and 4. Most 01 the teeth are worn, but two lower carnassials differ in the extent 01 advanced wear. Four fragmentary teeth show Ilttle or no wear but could all come lrom the same individual, suggesting a minimum of three individuals.

Crocuta crocuta Erxleben, 1777 MATERIAL. SKX 38256a, an anteriortragmentol a left p', with portions of the parastyle, paracone and protocone; SKX

MATERIAL. SKX 39109, a large, anterior righl mandible with the alveoli of the incísors and canine and broken fragments of three cheekteeth.

DISCUSSION. The specimen ís lIIustrated in Fig. 3. A second large specimen of Proteles has been found in the new Member 1 deposit, and both items wü! be discussed together in the section below dealing with the Member 1 material. Hyaenid gen. & sp. indet.

MATERIAL. SKX 38335, a left 1'; SKX 40043, a right 1'; SKX 38093, a right 1'; SKX 38334, a righl13; SKX 37994, a left 1,: SKX 38400, a left 1,: SKX 38399, a left p'; SKX 40029. a broken upper right canine; SKX 35325, a worn upper left canine; SKX 30575/30434, a worn upper righl canine; SKX 26190, a right p' Iragment: SKX 33644, a p' fragment; SKX 3111, a right p' fragment; SKX 39974, a posterior fragmenl 01 a p'; SKX 21547, 29750, 37314, three premolar fragments; SKX 37148, 35848, 37376, three proximal righl ulnae; SKX 37591, a proximal right radius; SKX 36835, a distal left libia: SKX 39116, a left calcaneum; SKX 27170, 29031, two first phalanges; SKX 37067, 35977, lwo second phalanges; SKX 29163, 33350, 28711,27291,34887, live third phalanges.

37235, a proxrrnal left second metacarpal; SKX 35528, a proximal right fifth metatarsal. DISCUSSION. The species is moderately well represented in the older samples from the síte (Ewer, 1955a; Turner, 1984a).

DISCUSSION. Since two or more species may be included, number esttrnanon is difficult, but at least three individuals are represented by rrght ulrtae. None of these ulnae offers a particularty good match with H. brunnea or C. crocuta.

NEW FOSSIL CAANIVOAE REMAINS

153

Table 1 Measurements ot Swartkrans Chasmaporthetes nitidula lower dentitlons. Land 8 are length and breadlh; Lp is length ot the protcconid: Lt ts length ot talonid; M/c récords the crear presence or absence of a metaconid.

p,

C,

p,

p.

M,

--~--

L

Spectmen

S

L

Oíd Member 1 spectmens (Hanging Remnant) 15,3 13,8 10,4 SK 302 15,4 alv. 8,5 SK 14005 17,5 SK 304 SK 301 SK 336

S

L

8

8,1 19,1 9,5

19,2 10,2 20,3

10,2 22,8

21,1

12,0

L

S

Lp

10,4

26,0

12,3

11,3

S

L

lt

M/c

broken 11,0

11,1

3,9

a26

11 ,3

11,3

3,7

New Member 1 specimens (Lower Bank) 14,3 10,0 13481

New Member 2 specimens 635 638 637

17,1

9,5 20,5

10,5 23,7

10,8

10,5

a - estimated Table 2 Measurements of Swartkrans Chasmaporthetesnitidula upperdentitions. Abbreviations L and B as in Table 1. Ba ts anterior breadth across the protocone; Bbl is breadth of blade: Lp ano Lm are Jengths of paracone and metastyle respectively.

p'

C· L

S

L

Qld Member 1 specimens (Hanging Remnant) SK 305 a16 SK 306 SK 312 SK 1853 13,9 10,0 SK 310 16,7 SK 379 19,5 SK 311 18,4 SK 309 SK 313 SK 307 SK 308

p'

p'

-----

Specimen

S

10,6

---~

S

L

Sa

Sbl

Lp

Lm

21,0 20,0 22,1

13,0 13,7 14,0

32,7

15,8 15,5 15,6

10,3 10,9 10,6

11,4 11,3

14,7

21,0 20,6

13,8 13,9 15,8

10,0

11,4

10,6

11,8

14,4 14,7

9,0

11,1

14,0

L

a12

11,7 12,0 12,3

New Member 3 specirnena 29205

22992/72

32,4

22,4

a13 a31

15,8

Family FELIDAE

Sublamily FELINAE

Sublamily MACHAIRDDDNTlNAE

Panthera pardus (Linnaeus, 1758)

Megantereon cuttridens (Cuvier, 1824)

MATERIAL. SKX 32467, a broken lower left canine: SKX 34193, a right p'; SKX 32203, a broken right p'; SKX 28098, a broken left p'; SKX 26759/26760/26761, a left maxillary fragmenl with p' and p4; SKX 35425/35426, a left P,; SKX 31654/31655, a left P,; SKX 27574, a broken left M,; SKX 21979, a rlght M, fregment; SKX 35238, a left second metatarsal, dístally unfused; SKX 36974, a left third metatarsal, distally unfused; SKX 35345, a left lourth metatarsal, distally unfused; SKX 33693, a complete right third metatarsat: SKX 33893, a proximal right second rnetatarsal: SKX 36407, a right fífth metatarsal, brcken distally; SKX 26483/26484, a complete right fifth metacarpal; SKX 37477, a proximalleft third metacerpal; SKX 38145, a fírst metapodial; SKX 39854, 39038, 39112, three distal metapodia; SKX 28067, a proximal left

MATERIAL. SKX 45391, the tlp 01 an upper canine. DISCUSSION. Although the specimen is only some 24 mm in

length. the identification is clear. The extreme flattening ofthe crown rules out the feline cats. The absence of beaded crenulations on the edges of the tooth rules out identification as the sabre-toothed cat Homotherium crenatidens (Fabrini, 1890), while the degree of pointing and flattening shows that it does not belong to Ihe genus Dinofelis Zdansky, 1924. Following a recent revision of the African Megantereon material. all the specimens are now consldered to belong to the single Eurasian species Megantereoncultridens (Tumer; 1987 a).

154

NEW FOSSIL CARNIVORE REMAINS

Table 3

Measurementsof Swartkrans Hyaena brunnea mandiblesand towerdentñíons. Abbreviations as in Table 1, plus PrM, is distancefrom anterior of Pi? to posterior 01 M,; Oepths A and Pare dorsoventral depths 01 mandibleanterior to Pi? and posteriorto M,; BP3 is mandible breadth below

P,.

p,

C, Specimen

L

S

L

p, S

Old Member 1 specimens (Hanging Remnant) SK 328 16,1 11,2 19,2 SK 329 New Member 2 specimens 591 a15 11.2 shed 590 alv. 15,5 9,0 365 15,2 11,5 3236/3237 1892

p,

L

S

L

S

13,6

23,1

13.1

broken 20,0

12,3

20,3

11,5

New Member 3 specimens 30401 20176/80/81 34159 20182

22,1

12,4 12,8

21.3

12,t

22,7 23,1

13,2

M,

Lp

L

L1

M/c

10.2

4.2

+

11,4

4,2

+

12,1

4,5

+

recte

12,0

22,5

S

Depths

S

Pz-M,

A

p, 18,4

a77

a81

P

37,3

18,4

11,5

23,2

12,0

Table 4 Measurements of Swartkrans Hyaena brunnea upper denüticns. Abbreviations as in Tables 1 and 2.

p'

C' Specimen

L

S

B

p' L

B

Old Member 1 specimens (Hanging Remnant) SK 326 alv. alv. alv. SK 335 15,8 10,8 SK 332 brcken SK 331 SK 327 Old Member 2 specimens SK 334

p'

p' L

t9,8

S

15,1

Sa

Sbl

33,0

20,2

11,1

32,9 31,8

broken 11,2 19,6 10,6

L

22,2

20,2

15,3

23,0

14,5

M'

Lp

Lm

S

12,4

11,1

13,2 12,2

10,9

11,6

f 1,2

20,3

New Member 3 specimens 30820/22/24 30831

ulna; SKX 30054, a proxlmalleft ulna: SKX 36825, a proximal right ulna; SKX 35801, a proximal lett radius; SKX 36644, a proximalleft radius: SKX 37364, a proximal right radius; SKX 39114, a proximal right femur; SKX 37552, a proxirnalleft tlbla: SKX 29652, a semi-complete axis vertebra; SKX 37679, a Iragment 01 axis vertebra; SKX 28870, 32973, two lumbar vertebrae; SKX 36955, 35313, two left ectocuneiforms; SKX 35234, a left cuboid; SKX 38424, a left astragalus; SKX 37594, a left fragmentary calcaneum; SKX 37029, a second phalanx; SKX 38578, a third phalanx. DISCUSSION. Measurements of the teeth are given in Tables 5 and 6, and those of metapodia in Table 7. Ewer (1956a) described leopard remains from the older workings at Swartkrans, and referred the material to a new subspecies, P. pardus incurva Ewer, 1956. The major character used in her diagnosis was the form of the angular process, said to be much more blunt and rounded in tne living animal based on her compara ti ve sample of nine modern specimens. My comparative sample of nineteen mandibles

32,3

19,7

11,9

14,2

shows distinction based on that character to be misleading. Four of the specimens have an angular process as fully developed as in the fossil material, and in the rest the extent of development ís variable and appears to be related to age. The distributions of sizes and wear stages in the teeth suggest the presence 01at least ñve animals, while the degree of size variation points to the presence 01 both sexes. The postcranial material comes from at least two individuals of different sizes. The immature metatarsals 35288, 36947 and 35345 lit together well and may come from one animal, while 33693 comes trom a srnauer individual in which the distal epiphysis is fully fused.

Acinonyx juba tus (Schreber, 1776) MATERIAL. SKX 36273, the crown 01a lower right canine. DISCUSSION. The tooth, which is clearly feline, is too towcrowned to be of leopard, and falls within the size range 01

NEW FOSSIL CARNIVORE REMAINS

extant Southern African cheetahs. The length (L) 01the crown is 10,1, the breadth (B) 7,8.

Family CANIDAE

Cenls cf. Canis mesomelas Schreber, 1778 MATERIAL. SKX 34254/34255, a left mandible with canine and premolaralveoli and ñrst and second molars in place; SKX 34988, a nghl M,; SKX 37056, a left M,; SKX 28800, a righl M,; SKX 23060, a broken right M,; SKX 27371, a broken right M,; SKX 21513, a broken M, in a Iragmenl 01 mandible; SKX 37811, a broken left M,; SKX 32890, a left M'; SKX 31653, a righl M'; SKX 21515, a right M'; SKX 36080, a left M'; SKX 37642, a broken right M'; SKX 22453, a broken left M'; SKX 30825, a left M'; SKX 30829, a righl M'; SKX 29038, a righl M'; SKX 30091, a righl M'; SKX 32874, a fragmentary left M'; SKX 30060, a right M'; SKX 36349, a righl maxillary fragment with M' ano M'; SKX 37434, a right p'; SKX 26989, a right p'; SKX 21514, a broken righl p4; SKX 30727, a broken left p4; SKX 32892, a broken left p4; SKX 29986/29987/29988, a right maxiilary Iragmenl wlth p' and p4; SKX 33445/33446, a left upper canine; SKX 38135, a right upper canine; SKX 38092, a canine fragment; SKX 19503, a left lower canine; SKX 37815, a left [a; SKX 28814, a right la: SKX 31919, a right la: 3; SKX 32602, a left 1,; SKX 33662, a right 1 SKX 35770, a lower rlqht deciduous molar; SKX 36352, 34822, 38392, 31839, 28329, 26991, 22514, 29965, 38261, 30138, ten broken oc very worn premolars; SKX 35792, 36124, 31851, three axis vertebrae: SKX 22975, a sacrum; SKX 37168, a lumbar vertebra; SKX 22448, 29535, two proximal nght ulnae; SKX 19531, a proximalleft ulna; SKX 38147, a proximal right ulna; SKX 33412, a distal left humerus; SKX 40151, a distalleft humerus; SKX 21540, a proximalleft radius; SKX 38152, the shaft al a left tibia; SKX 20128, a left calcaneum; SKX 40266, a left calcaneum; SKX 27354/27447, a right fifth metacarpal; SKX 36429, a right second metacarpal; SKX 38567/38571, a right second metacarpal; SKX 29490, a left fifth metacarpal; SKX 33689, a left second metacarpal; SKX 30300, 35164, 38506, three distal rnetapodia: SKX 39216, a first phalanx. DISCUSS10N. Measurements of the dentition are given in Tables 8 and 9, and measurements af metapodia in Table 10. Separatían 01 C. mesome/as from the side-striped [acka! (C, adustus Sundevaií, 1846) Is difficuil when deailng with fragmented material. The gene rally larger size of C. mesome/as carnassial and post-camassle! teeth suggesfs that it ls the species representad here, at least by the dental sample. Ewer (1956b) considered most, if not aií, 01 the material from the South African hornlnid sttes then available to her to be 01the black-backed jackal. On the basis of size and wear distributions in the dental sampte. at least five antmats are present.

Vu/pes chama (Smilh, 1833) MATERIAL. SKX 19697, a right p4; SKX 36620, an edenlulous righl mandible: SKX 23006, 36211, two righl Iragmentary scapulae; SKX 36127/36123, a right humerus missing the

155

distal end; SKX 36093, a proximalleft humerus; SKX 31820, '" di stal righl radius: SKX 36335, a proximal left ulna; SKX 31749 a left calcaneum; SKX 20113 a nght caícaneurn: SKX 29670, a left astragalus; SKX 33968, a metapodial fragmenl; SKX 25520, a Ihird phalanx. DISCUSSION. The upper carnassial and the edentulous mandible belong lo V. chama, while the fragmentary smaií canid postcranial remains are much more similar to ttiis species than to the bat-eared fax, Otocyon mega/olis (Oesmarest, 1821), Two right scapular specimens point to the presence of at least two individuals. The dimensions of the upper carnassial are iength (L) 10,8 mm, anterior breadth (Ba) 5,0 mm, breadth al blade (Bbl) 4,1 mm, and length of melastyle (Lm) 4,7 mm. Largo can id gen. & sp. indet, MATERiAL, SKX 35314, a fragmen! 01a left mandible wilh M3 in place: SKX 19502, 27480, 23059, three thlrd phalanges. OiSCUSS/ON. The mandlble 35314 Is too large to belong lo a species 01 jackal, The thr ee third phalanges show a dog-like morphology but are also too large lo be jackal. lt is tempting to refer these items to Lycaon pictus (Temminck, 1820), the hunting dog, but thelr identification as Lycaon cannot be posttively confirmad en the basis of this material. A larga spectes 'ot Caniswas clearly presentat Krorndraai (Turner, 1986a), and the material could equal1y belong to that taxon.

Famify MUSTELlDAE Subfamily LUTRINAE

Aonyx capensis (Schlnz. 1821) MATERIAL. SKX 30361, an edenlulous left mandible; SKX 35954, a broken right p4; SKX 35790, 35508, two righl upper canines. DISCUSSION. Measurements are given in Tabte 11 and the mandible is iIIustrated in Fig. 2. The material is from two or more large otters similar in size and morphology to specimens ot the extant species and quite unlike the smaller spottednecked otter, Lutre maculicolJis Uchtenste¡n, 1835

Family VIVERRIDAE Subfamily HERPESTINAE

Suricata eurcette {Erxleben, 1777) MATERIAL. SKX 39282, a righ! horizontal ramus ola mandible wtth broken and worn P2 and M, in place; SKX 35135, an edentulous right horizontal mandibular ramus: SKX 26913, an anterior portion of a right mandible with P2 and P3 in situ and the roots of P4; SKX 37483, most ef a right horizontal mandibular ramus with P2, P4 and M 1 in situ;8KX 36961, an edentulous, posterior left horizontal mandibular ramus; SKX 27001, an edentulous right horizontal mandibular ramus.

NEW FOSSIL CARNIVORE REMAINS

156

Table 5 Measurements of Swartkrans Panthera pardus rnancnbresánd lower dentitions. Abbreviations as in Febles 1 ano 3, plus e-co ls distance from anterior of canine to middle 01 condyle; HPC ls height 01 coronoid procesa.

C,

P3

L

Specimen

B

P.

L

B

M,

L

B

Oepths

B

e-ce

17.9 17.9

9.0 7.6

130

17,7

8.8 9.2

B

L

HPC

PJ-MI

A

45.9 44.1

26.4 21.6

a49 50,0

28.8

a45

23.6

40,2

21,0 21.0

P

P.

Old Member 1 soecrmans (Hanging Remnant)

SK 349 SK 18661 343a+b SK341 SK342 SK 344 SK340 SK 345 SK 339 SK 347

12,7 11.0

8.5 8.4

12.0 11.0

5.4 6.0

17.5 16.1

7.5 7.7

13.5 a13

a10 al0

12.2 a13 11,4

6.1

18.5 18.9 16.8 17,1 16,4

8.3 8.5 7.9 8,0 7,5

broken

5,7

alv.

5,8

19.3

28.0

13.9 12,0 a16 14.5

roer 17.5 17.6 a18 17,5

8.2 8,3 8,6 8,8

15,6

7,3

66,4

27,7 26.8

13.7

Old Member 2 specimens (mixed provenance]

SK 346 SK 348

13.7 10.5

8,7

11.9 10.2

6.3 5,5

15.3

7.4

16,0

7,6

12,6

New Member 1 specimens 8113 12,6 9.3

New Member 3 specimens 3165415 12,3 3542516 27574

6.0 8.2

Table 6 Measurements of Swartkrans Panthera pardus skulls and upper dentitions. Abbreviations as in Table 2, plus Bl ls basallength 01 skull: ZB is breadth aeross zygomatic arenes: AB is breadth across the externa! faces 01 the can ines; MB is breadrtt across the external faces ot the camassiaíe: POP is breadth across the postorbital processes: POC is breadth ot the postorbltal constncton: C8 ls the breadth of the condyles.

Specimen

L

p3

p'

C· B

L

B

L

p' B

L

M'

Ba

Bbl

Lp

Lm

B

BL

ZB

RB

MB

POP POC

11.9 9.5 14,6 a12 10.1

8.2 9.8 9.2

10,7

10,7

8,4

a170

a136

a54 a74 a60

a80 a54

a74

a54

aSO

44,7

CB

Old Member 1 specímens (Hanging Remnant)

SK SK SK SK SK SK SK SK SK

349 13.6 353 alv. 354 16.6 14010 351 alv. 352 alv. 355 356 358

10.0

alv. 12.7 ?alv.

a6

5.4 15.7 5,8

3.9 7.0 3,9

a17 24,4 19,4

a9 13.3 9,6

17.2 a4

7.8 15,0

25.4

14,3

7.7

16.0

8.0 8,2 7,3

15.1

7.0

17,5

25.3 8,4 27,7 24.2 9.1 a13 25.6

14.3

10.5 11.0 8,9

11,2

10.6

alv. 11.0

9.6

New Member 1 specimens 8528 15,6 12.8

9389 9145

813 593 812

11,6

9.5

New Member 2 specimens 6281629 a/v. alv. 16,0 14,1

7,6

24,9

12.3

alv.

7.5 25.1

9.8 12.4

10,0 7.5

9,8

10,1

24,1

12,0

7.2

10,2

9,6

29,4

13,8 11,5 12.1

9.6 8.0 8.3

12,0 9.1 10.0

11,8

10,3 7.5

New Member 3 specimens

267591 60161 34193 32203 28098

15.6

7,4

alv.

8.0

39

37.0

157

NEW FOSSIL CARNIVORE REMAINS

Fig.2 SKX 30361, left mandible 01 Aonyx cepenste. bucea! view. Natural size.

DISCUSSION. In all 01 these broken and frequently edentulous specimens, the size and morphology ot the mandible, and the number, size and spacing 01 the aíveo!'. suggest their identification as surícates.

Table 7 Measurements ot Swartkrans Panthera pardus metapodia. TL =: total length; PML =: medio lateral width o, proximal articutation: SW =: minimum meciolateral shaft width; DE =: medtotateral width ot distal epiconcíyle.

Cynielis penieillata (Cuvier, 1829) TL

Speclmen

PML

DE

SW

MATERIAl. SKX 36332, a posterior fragment 01 a right horizontal mandibular ramus with M1 in situ.

Metacarpals

Subfamily VIVERRINAE

New Member 1 soectmen (Lower Bank) 62,3 9715 Me 3

7,8

11.9

New Member 2 specimens 779/554 Me 2 65.3 769 Me 4 62,5 1554/1552 Me 4

8,5 8,2 6.9

13.7

1 t,9 10,4 10,1

7,0

11,6

13,8

9,0

12,6

cf. Genetta tigrina (Schreber, 1778) MATERIAL. SKX 35715, a distal righl radius. DISCUSSlüN. The specimen ls somewhat eroded, making a positive identifica tian difficult. MEMBER 2 SPECIMENS

12.1

New Member 3 soecunen 53,6 26483/26484 Me 5

11,0

Metatarsals New Member 3 speclmen 33693 Mt 3 79,1

Family HYAENIDAE Chasmaporthetes nitidula (Ewer, 1955)

Table 8 Measurements ot Swartkrans Canis cf. mesomelas lower dentitions. Aboreviations as in Table 1.

MATERIAL. SKX 635, a right P,; SKX 638, a right P 3 with a brcken main cusp: SKX 637, a right P4.

C Spectmen

DISCUSSlÜN. Measurements are given in Table 1, together with measurements 01specimens from Member 1 01 the older workings at the sita, and the specimens are illustrated in Fig. 1. AII three teeth are of similar appearance and exhibit compatible wear, strongly suggesting tnat they come from a single individual. Viewed buccally or Iingually, all exhiblt the trilobed appearance charactertstlc o, the Iower premolars in C. nitidula, and they match extremely well with other specirnens trom the older workinqs at the site.

L

M,

P. B

L

B

L

8

L

B

New Member 1 specimens (Lower Bank)

11,5

4,8

New Member 2 specimens 3016 8,2 5,4 1085 11,9 52 12,2 New Member 3 spectmens 34254/34255

16228/16231

-

5,4 5,6

34988 37056

28800 Hyaena brunnea Thunberg, '820

MATERIAL. SKX 590 and 591, respectively right and left horizontal mandibular rami from one individual; SKX 424, a premolar root that appears to belong to the P3 of 591; SKX 3236/3237, a left P,; SKX 365, a left P,; SKX 1892, a right P,; SKX 50/53, a Iragment 01right p 3; SKX 51. a Iragment 01right p'; SKX 2394, a distalleft tibia; SKX 577, a distalleft tibia.

21513

D!SCUSSION. Measurements of the dentition are given in Tables 3 and 4. The left and right mandibles belong to one old individual, whñe the three isolated lower teeth may betong to a single younger animal. The teeth are readily identifiable as brown hyaena by virtue

158

NEW FOSSIL CARNIVORE REMAINS

01 their small size, relative robustness and low crown heights in the case 01the prernolars, and by the length 01 tne talonid and shortness and relative proportions 01 the

Table 9 Measurementsot Swartkrans Canis d. mesomefas upper oentiticns. Abbrevlatfcns as in Table 2.

p'

p'

trigonidin the case 01the lower carnassials. Specimen

L

8

L

8a

M' 8bl

Lm

M'

L

----

8

L

8

7,5

9,2

7,3

9,1

7,8 7,5

9,7 8,8

Crocuta crocute Erxleben, 1777 New Member 1 specímens (Lower Bank)

MATERIAL. SKX 1146, a posterior Iragment 01 a right M,; SKX 15771, a posterior Iragment 01 a left M,; SKX 957, the broken metastyle 01 a right P'. DISCUSSION. Botn the M, specimens preserve the rearmast portion 01the protoconid and the talonid, a complete talonid and rear root in the case 01 1146. Bom also exhibit a small metaconid. The shortness 01 the talonid identifies the two specimens as C. crocuts: in the more complete specimen 1146 the length 01 the taionid is 3,0 mm. The metastyle 01specimen 957, aithough broken, is elongated beyond the condition seen in either Hyaena or Chasmaporthetes. The Interna! cingulum is apparent, but markedly less developed than in Hyaena.

8833 7997

17.9 18,2

8,4

6,8 7,3

7,9 7,8

19,9

10,1

7,2

8,6

New Member 2 specimens

1130/36/36a -

worn

1115 3990

a12

a16

New Member 3 specsnens

29986/7/8 12,0

4,1

37434

20,0 17,0

26989 36349 32890 31653 21515 36080 37642

8,6 8,0

18.4

6,9 6,1 6,8

8,8 7,2 7,0

30825

MATERIAL. SKX 303013028, 1262, posterior Iragments 01ieft Pes: SKX 3008, a posterior Iragment 01 a left p,; SKX 592, a right lower canine; SKX 1032, a right lower canine crown Iragment; SKX 1034, a broken right p'; SKX 3055/3056, an anterior Iragment 01 a right M,; SKX 3027, 3004, 959, 1201, tooth Iragments; SKX 640, a left 1'; SKX 642. a right 1,; SKX 545,621, 621a, three fragments ot axis vertebra, possibly a singie specimen; SKX 2541, a distal right humerus shaft; SKX 1921, a distallef1 radius; SKX 1920, a left pisiform; SKX 1185, 497, 191, three third phalanges. DISCUSSION. As in the case 01 the Member 3 sample, the likelfhood of more than one species being represented in the material renders number estimation 01 little use. Family FELlDAE Subfamily FELINA E Panthera perdus (Linnaeus, 1758)

MATERiAl. SKX 628/629. a right maxilla with aiveoli ter the canine, p2 and M\ and with p3 and p4 in place; SKX 813, an isolated left P';.oKX 812, a Iragment 01 an isolated P'; SKX 3544, a Iragment 01 a left upper canine; SKX 593, a left upper canine; SKX 1574, a Iragment 01 a ieft p'; SKX 600, a left 1'; SKX 630, a right 1'; SKX 639, a right 1'; SKX 632, a left glenoid tcsaa 01 the zygomatic process; SKX 431, the glenoid cavity of a left scapula; SKX 431 a, a fragmentof scapula that appears to go with 431; SKX 1337, a distal righthumerus; SKX 604/605, a distal left humerus; SKX 674, a proximalleft humerus; SKX

13,8 15,1 14,6

13,1

15,9 14,5

12,6 14,2 12,8 12,8

30829 29038 30060

Hyaenid gen. & sp. indet.

12,1 12,5 12,5

15,4

15,3

Table 10 Swartkrans Canis ct. mesomeJas rnetapodta. Abbreviations as in Table 7. Metaearpals

TL

Spedmen

PML

SW

DE

5,9 6,4 6,1 8,3

4,8 5,0 5,1 5,6

7,6 6,7 6,7 8,2

7,7

4,5 5,3

7,1 7,6

New Member 2 speomens 2870 Me 2 46,9

2691192 Me 3 2710/2690 Me 4

51,3 51,0

2693 Me 5

43,5

New Member 3 specimens

38567/71 Me 2 27354/27447 Me 5

47,2 42,3

Table 11 Swartkrans Aonyxcapensis isolated teeth. Abbreviations as in Tables 1 and 2.

p'

C' Specimen

L

8

L

M' 8

L

8

New Member 1 speeimens (Lower Bank)

13123 8719 8323

9,6

8,7 13,0 13,6

New Member 3 speeimens 35790 8,4 7,5

35508

9,0

8,6

14,0

NEW FOSSIL CARNIVORE REMAINS

586. a proximalleft radius: SKX 770. a distal left radius; SKX 580. a proximal left ulna; SKX 2880. a distal left ulna; SKX 779/554. a left second metacarpal; SKX 587. a left third melacarpal; SKX 769. a left fourth metacarpai; SKX 1554/1552. a right fourth metacarpal; SKX 782. 3808. 578. 558. four distal metapodia; SKX 2890. a left scapholunare; SKX 2885, a lett magnum; SKX 599, a left accessory carpal: SKX 2887 a lefttrapezoid; SKX 3307. 772. second phalanges; SKX 555. 556. 557. 559. first phalanges.

159

_PISCUSSION. As in the case of the Member 3 canid sample, these specimens clearly belong to a can id larger in size than a jackal, but of uncertain identity. The second phalanx 1386 is morphologically similar to Lycaon pictus, but somewhat larger than the comparative material available. Family MUSTELIDAE Subfamily LUTRINAE Aonyx cepensis (Schinz. 1821)

DISCUSSION. Measurements of teeth are given in Tables 5 and 6, and measurements 01metapodia in Tabla 7. Mast ottne post-canine dental material seerns to form a set belonging to one animal, but there are two left upper canines ano two left upper third prernolars, including one of the íatter in the seto The more worn af the two caninas may go with the Isorated third premolar, but the second canina does not seem to 90 with the set (because of the degree of wear). The dental evidence therefore points to three individuals. The more complete dentítion is af similar síze to that of the smaller animals in the Member 3 sarnple. The postcranial material contains parts of most elements ot the forellmb, and at most represents one individual. However, a second, smaller individual is al so presento The distal left humerus specimen 604/605 has a maxímum epicondylar width ot 51,6 mm and a minimum (estimated) mediolateral shaft width of 18,2 mm. The width of the distal articulation (trochlea ano capitulum) is 37.0 mm. Family CANIDAE

Cenis ci. mesometes {Schreber, 1778) MATERIAL. SKX 1130/1136/1136a. a left maxillary fragment with p' and M'; SKX 1115. a right M'; SKX 749 a left M'; SKX 3990. a left M'; SKX 1052. a left p'; SKX 3016. a leftlower canine; SKX 2772. a broken M,; SKX 1085. a right P,; SKX 52. a left p,; SKX 1436. a broken left P,; SKX 925. 2465. fragments of premcíars: SKX 62. an anterior portlon ot an axis vertebra; SKX 2695. a proximal right radlus; SKX 855. a distal left radíus; SKX 753, a proximal right radius; SKX 1652, a distal right radius; SKX 2693. a right filth metacarpat: SKX 2710/2690. a rlght fourth rnetecarpal: SKX 2692/2691. a right thlrd metacarpal; SKX 2792. a proximal right fourth merecerpal; SKX 2870. a left second metacarpal; SKX 962. a metapodial shaft and distal artlculation; SKX 2716. a right scaphclunare: SKX 18794. 2711. 2708. 15926.2712.2709. 2707. seven first phalanges; SKX 2714, 1045, two second phalanges; SKX 15897. 2346. 3989. three third phalanges. D1SCUSS10N. Measurements of the dentition are given in Tables 8 and 9, and measurements of metapcdla in Tabla 10. As in the case ot the Member3 samote. the teeth suggestthat Canis mesomelas is most likely to be the species representad. In terms of tooth size and wear, at least four individuals are present. Large canid gen. & sp. ¡ndel. MATERIAL. SKX 1386. a second phalanx; SKX 61. 15785. 2271. three third phalanges.

MATERIAL. SKX 3222. a worn upper left canine. DISCUSSION. Although loo worn lo be measurable. the specimen comes from an animal slightly smaller than those in the Member 3 sample. Family VIVERRIDAE Sublamily HERPESTINAE

Suricata suricatta (Erxleben, 1777) MATERIAL SKX 662. a right horizontal mandibular ramus with P3 and MI in situ; SKX 286, a tragment of a right horizontal mandibular ramus with P 4 in situ. DISCUSSION. Specimen 662 compares well with S. surtcette in terms or ovaran morphology and tooth numbers. Specimen 286 is too incomplete for a firm identitication, but the rnandible apparently had no P" ruling out Cynictis penicillata in which the P4 ts in any event somewhat different. The latter tooth is also different in Mungos mungo (Gmelin, 1788), the banded mongoose. ef. Cynictis penicillata (Cuvier, 1829) MATERIAL. SKX 15531. a proximalleft ulna. DISCUSSION. The specimen shows a clase resemblance to C. peniciIJata.

MEMBER 1 LOWER BANK SPECIMENS Family HYAENIDAE

Chasmaporthetes nitidula (Ewer. 1955) MATERIAL. SKX 13481. a left iower canine; SKX 5414. 9720. two lawer right third premolar tragments. DISCUSSION. Measurements 01 13481 are given in Table 1. The twa P3 fragments do nat [om, but they may ha ve come from the same toolh. Although only small fragments, both clearly exhibit characters of C. nitidula teeth: buccolingual narrowness and aceessory cusps that are well developed. The lower canine comes from a younger animal. Although clearly hyaenid, it is considerably smaller and more gracile in appearance than the canines af either H. brunnea or C. crocuta, and the crown is much lower.

160

NEW FOSSIL CARNIVORE REMAINS

Hyaena brunnea Thunberg, 1820 MATERIAL. SKX 5845, a fragment of the metastyle of a right

P'. DISCUSSION. The fragmen! is from a tooth with a short metastyle, as witnessed by the sharp curvatura of the enamel base towards the occlusal surface at the rear 01the tooth, and wttn a fairly well-developed lingual cingulum. Protetes Geoffroy, 1824 MATERIAL. SKX 13280, a rightanterior maxilla and premaxilla with the canina in place and the alveoli 01the incisors and 01 two post-canina teeth. DISCUSSION. Both SKX 13280 and the mandible SKX 391 09 trorn the new Member 3 deposit, together with the mandibte 01 a modern aardwolf, Proteles criststus (Sparrman, 1783), are illusttatsd in Fig. 3. The fossil mandibJe is similar in overall appearance to the equivalent portion of the modern species, but ccnslderably largar. The disposltion 01 the teeth is also different. The maxtña is broadly similar to the modern aardwolf bu! is considerably largar, although the disposition of the teeth in thís case shows no readily apparent differences. Hendey (1973, 1974a) referred material from Kromdraai 8 and the older workings at Swartkrans to Prote/es. He noted (Hendey, 1973) that the Kromdraai speeimen KB 2945, an anterior mandible fragmentsimilarto SKX 391 09, ts largerthan the modern material available for comparison, and that the number and/or the positioning of the post-canin e teeth appear to be different. He nevertheless decided against the erection of a new species on the basis of a single incomplete specimen. In a later work, Hendey referred two specimens from Swartkrans, SK 3173, a right mandible fragment, and SK 1851, a left maxillary Iragment (Hendey, 1974a: plate 3), to a new species, Prote/es transvaalensís Hendey, 1974, while referring two other specimens to the extant species. Brain (1981) listed the new species as eoming from the old Member 1, while the extant species was given as part of the old Member 2 assemblage. Hendey (1974a) argued that the Kromdraai specimen might also belong to his new species, although he also took the view that it might be a somewhat more advanced form largely on the basis 01inferred tooth dispositions. The two specimens from the old Member 2, SKW 122, a right maxillary fragment, and SK 11400, a left anterior mandible fragment, are very similar to the modern material in size, tooth numbers and posítioning, and overall morphology, and the;r referral to the extant species is evidently correet. The new material from Swartkrans is different in size from the extant species: should it be referred to P. transvaalensís, and how does the Kromdraai B specimen relate to that new speeies? The post~can¡ne mandibular dentition of the living aardwolf consists of a P3 and a P 4 set closeJy together sorne 18 mm from the canine alveolus, with an M2 set sorne 10 mm behind them (Fig. 3D). The P, tends to have a bifureated rool. Hendey (1974a) eoneluded that SK 3173 preserved a P, and a P 4, both with bifurcated roots. However, he also coneluded that the Kromdraai B specimen, KB 2945, differed from both the modern species and from SK 3173 in having a small P3 that

was set further forward ano in having an M, set very closely behind the P4, although he also aceepted the presence of root bifurcation in the P4 . When SK 3173 and KB2945 are now considered in conjunction with the new mandible SKX 39109 and with a modam specimen. it is apparent that the three fossils share distinetive characters. The second and third post-canine teeth of SKX 39109 falJ al the correet distance from the canine to be considered as P, and P, (Fig. 3B,C). It is nol entirely elear whether the inferred P, 01 SKX 39109 (2 in Fig. 3C) has a bifureated root although the shape of the alveolus suggests that it may do so, but the inferred P4 clearly does, smce the anterior root ts preserved together with the posterior alveolus (3 and 4 in Fig. 3C). The first post-eanine tooth in SKX 39109, of which the single root is preserved (1 in Fig. 3C), is therefore rnost probably a P,. Comparison with KB 2945 tnen suggests a similar presence af P2 in that specimen, and that the lncomplete alveolus at the rear 01 the fragment is ter the anterior root 01 the P4 and not for an M 1 as inferred by Hendey. Further examinatíon af SK 3173 serves to underline the similarities in al! three fossils.ln addition to the bifurcated roots of P3 and P4 inferred by Hendey, examination with a hand tens reveals the trace of an aíveoíus in the broken surface at the front ot SK 3173, in precisely the place where a P2 should be, basad on my interpretation of the tooth dispositions in SKX 39109 and KB 2945. The question then arises whether considerablv greater size than the extant species, in conjunction with the presence 01 a P 2 , is sufficient to warrant referral of the fossils to a new species. Of the fifteen modern mandibJe specimens available to me in the Transvaal Museum col1ection, tour have srnall, partly erupted teeth [ust in front of the PJ , each of which could be interpreted as a P 2 . These extra teeth vary somewhat in size, but al! appear to differ from the deciduous teeth in being more hypsodont and uniradial: in other words, they are unlikely to be retained deciduous teeth. The chlef distinguishing characters of the tossll mandibles would therefore appear to be size and the tendency for P3 to be biradial, while the presence of a P2 appears to be of less significance. However, Hendey's new species P. transvaalensís was in fact based upon the maxillary specimen SK 1851, with the mandible SK 3173 referred to it. Not only is the maxilla considerably bigger than any modern specimen of Proteles, but the teeth are also considerably less degenerate (Hendey, 1974a: plate 3). In addition, as Hendey (1974a) pointed out. the wear facets on the teeth indicate occlusion with lower teeth that must themselves have been more substantial (and probably more numerous) than those of the extant species. If the specimen SK 1851 is indeed an aardwolf, and I accept Hendey's argument that it appears to be so, then the mandible that went with it must indeed have differed from the modern formo The new maxilla SKX 13280 does not permit comparison with SK 1851, since it is a more forwardly-placed portion of the bone (Fig. 3A). it compares weli with lhe equivalent part of !he modern species, size apart, in the following features. First, in overal! shape and appearance. Second, in having smalJ incisors set well forward in the premaxilla, with a marked canine-incisor diastama; the alveoli of all three incisors, while showing a reduction in size from 13 to I 2, exhibit nothing like the size reduction between their counterparts in other

161

NEW FOSSIL CAANIVOAE AEMAINS

carnlvores. Third, in having a maxillary-premaxil· lary suture that passes sorne 3,5 mm in front ot the anterior edge 01the canina alveolus and which then runs almost straight across the palate at right anglas to the intermaxillary suture 01the patate: in canids, viverrids and telida the maxillary-premaxillary suture runs into the alveolus sucn that the front edge 01 the canina socket is formad by the premaxilla, while the entíre socket is formad in the maxiña in Prote/es and in the specimen. Fourth, in having two post-canina teem with alveoli separsted by a similar distance (sorne 5,0 mm) to that separating the first post-canina tooth from the canina alveolus. Fifth, in having a second post-canina tooth with partly blfurcated roots, Sixth, in having a buccolingually flattened canine. The argument for specific separation of the Swartkrans and Kromdraai B specimens from the extant species therefore rests on the characters 01 the maxillary Iragment SK 1851, which certainly warrant taxonomic distinction. Whether the three large mandibular specimens and the rnaxílla belong to that species, or simply represent a larger variery of the extant species in which the P2 is more frequently retained, is difficult to say on the evidence presently available.

1 2

~

e

."**~~.

Hyaenid gen. & sp. indet. MATERIAL. SKX 9479, a deciduous lower left canlne; SKX 14481, a premolar Iragment; SKX 7854, an anterior fragment of an axis vertebra; SKX 9571, the glenold cavity 01 a left scapula; SKX 8673, 8675, fragments 01 a left ulna that probably go together; SKX 8674, the olee rano n process of a right ulna, with untused epiphysis; SKX 7822, a proximal right radius: SKX 6988, a distai left tibia; SKX 13664, a distal portian 01 a right lifth metatarsal; SKX 4673, a distal metapodial; SKX 6990, a right calcaneum; SKX 5795, a right ectocuneiform; SKX 13722, a second phalanx.

__4

~~\ ,,:'.:.,;

o Flg.3 Proteles specirnens. A

=

=

SKX 13280, right maxilla and premaxtña. bucea! view, B SKX 39109, right mandible, bucear view; = SKX 39109 occlusal view; D = AZ 330, right mandible of modero Protelee cristatus, Transvaal Museum coltecttcn, bucea! vtew. AII natura! size.

DISCUSSlüN. The hyaenid materiai represents at least ene adult and ene juvenile, but the minimum number of adults is impossible to estímate.

Family FELIDAE 5ublamily FELfNAE

Panthera pardus (Linnaeus, 1758) MATERIAL. SKX 8528, a right upper canine; SKX 5099, a crown fragment of a right uppercanine; SKX 8113, a right lower canine: SKX 9145, a fragmentary righl p'; SKX 9389, a ieft p3; SKX 13053, a righ1 P, fragment; SKX 9402, a left upper deciduous fourth molar; SKX 6345, a proximal fragment ot a right ulna; SKX 6346, the distal end al a right ulna; SKX 13232, a proximal fragment of a right ulna; SKX 11925, a proximal

e

right tibia; SKX 8164, a distalleft libia; SKX 14332, a righl ectocuneiform; SKX 7122, a rightthird metatarsal: SKX 13227, a left lourth rnetatarsal: SKX 5471, a left fifth metatarsat: SKX 21337, a right fifth metatarsal; SKX 4922, a left third metacarpal; SKX 9715, a righl third metacarpal; SKX 14701, a left lourth metacarpal; SKX 4168, a left second metacarpal; SKX 5736, a left third metacarpal; SKX 4203, a left patella: SKX 8110,6746, two first phalanges; SKX 81 09, 8896, 6745, 7225, tour second phalanges; SKX 6747, a third phalanx.

DISCUSSION. Measurements of the dentition are given in Tables 5 and 6, and measurements ot the metapodia in Table 7. Two adult cocer right canines and a left upper deciduous carnasstat point to at least three individuals in the dental sample. Among the postcranial elements, at teast two

162

NEW FOSSIL CARNIVORE REMAINS

differently sized anirnals are present. One of the smaller metapcota Is of an adult animal. Although tbe largar metapodia". are generally too incomplete to snow whether tuaion of the distal epiphyses has occurred, their size suggests that they come from eouns. The presence of a deciduous tooth together with these metapodia therefore again points te at least three individuals. Panthera leo (Linnaeus. 1758)

two and poasibly three ages are present in the dental sample, and lt would be difficult to 90 beyond a mínimum figure of three. Small canid gen. & sp. indet. MATERIAL. SKX 13538, 7135, 8605, 13557, four distal metapodía of a small, tox-slzed canid. family MUSTELIDAE Subfamily LUTRINAE

MATERIAL. SKX 8464A, a righl second phalanx, Aonyx DISCUSSION. The speeimen ís unqueslionably lelid, and while it is too larga to belong to a leopard it also appears too small to belong to Megantereon or Homotherium. The match with a modern lion specimen is 900d. Fefis servalSchreber, 1776 MATERIAL. SKX4116, a righlcalcaneum. DISCUSSION. The specimen, allhough somewhat smaller, matches very well with comparenve material of tms species.

cepensts (Schinz, 1821)

MATERIAL. SKX 14686, 13508, two broken nghl mandibles; SKX 8323, a right M'; SKX 9717, a left M'; SKX 8719, a ieft p4; SKX 13123. a right upper canine. DISCUSSION. Measuremenls are given in Table 11. The material is from anirnals similar in size to those represented in Members 2 and 3. Family VIVERRIDAE

Herpestes ícrmeumon (Unnaeus, 1758) Felis caracal Schreber, 1776 MATERiAL. SKX 14520, a right p'; SKX 6939, a left third MATERIAL. SKX 5792, a distal metapodial.

rnetatatsal.

DISCUSSION. The specimen exhibits the curvatura of the

DISCUSSION. Identlflcalion 01the p' is seeure, the identification of the metate-ea! stightly lesa so. The dimensions of the p' are lenglh (L) 10,1, anlerior breadth (Sa) 7.3, breadth 01 btade (Sbl) 3,7, length 01 paraeone (Lp) 5,0 and length 01 metastyle (Lm) 4,3.

upoer edge ot the articular surface sean in felid metapodia. It Is somewhat too Jarge to be F. servet. Family CANIDAE

Canis ef. mesomelas (Schreber. 1778) MATERIAL. SKX 16228116231, tne horizontal ramus 01 a left mandible with vetv worn P 2 -4 and alveoli forthe canina, PI and the molars; SKX 16256/16230, the anterior portion 01 a right mandible with a very worn P2 (the specimen may palr with 16228/16231); SKX 6175, a broken right M,; SKX 13668, a broken left M,; SKX 8833, a left maxillary tragment with a p'; SKX 7997, a broken right p'; SKX 4937, a left p'; SKX 4939, a lett p'; SKX 4940, a righl p'; SKX 14772, 5314.4933. three righl 1'5; SKX 4934, a left 1'; SKX 8844, 5047, 14069, lhree '3S; SKX 7875, a worn inctscr: SKX 4930, a rert lower carune. SKX 4929, a righl lower rightcanine; SKX 8084, 10021,4843,4566, 4938, five broken premolars; SKX 9542, a proximalleft radius; SKX 4585, a distalleft radius; SKX 9149, a right femur head; SKX 13422, a distal right libia; SKX 12111a, a left aceessory carpal; SKX 5864, a left calcaneum; SKX 13421, a right astragalus: SKX 4793, a prox.imalleft fourth metatarsal; SKX 13384, a proximal righl lourth melalarsal; SKX 8355, 8356, 8931, 15044, four distal melapodia; SKX 8358, 11866, 7994, 18988,8424,5447.8357,8289, eight tirsl phaianges; SKX 18177,12070,6602,8016,7030,13323, six lhird phalanges. DISCUSS/ON. Measurements of the dentitions are given in Tables a and 9, and measurements of the metapodia in Table 10. Number estimation is difficult: on the basis of wear, at least

GENERAL DISCUSSION

"he presence of Megantereon cultridens in the Member 3 sample may be me tatest record of the animal in Atrica. Previously known late occurrences on the continent 1all in tne 1,0-2,0 m.y. S.P. braeket (Turner, 1985,1 98Gb), and although the date at the specimens from Elandsfontein in the Cape Province of South Africa (Hendey, 1974b) is uncertain, lt is likely tnat they are 01 Lower Pleistocene age (Hendey, personal communication).ln Europe the species appears to ha ve become extmct towards the end of the Villafranchian faunal stage (mic-Pliocene to Lower Pleistocene), although jn Asia it ts recordad in the Middle Pleistocene deposits of Zoukoudian (Teilhard de Chardin, 1939; Turner, 1987 a). lt therefore seems that the Old World extinetion of the species was broadly contemporaneous. What is now abundantly clear is that the machairodont cats coexisted in Africa with the modern feline cats far a very long period of time, considerably reducing the force of the argument that theirdemise may have resultad from competition with the modern species (Turner, 1985; 1986b, 1989). The presence of Chasmaporthetes nitidula in Member 3 is yet another latest occurrence, with the previous youngest record being that in the old Member 1 Hanging Remnant sample from Swartkrans (Turner, 1986b). Two species ot Chasmaporthetes are known from both Swartkrans and Sterk·

NEW FOSSIL CARNIVORE REMAINS

fontein (Turner, 1987c), but only C. nitidula is present in the new material. The relationship otthe African Chasmaporthetes to those in Eurasia and the New World, and that between the Eurasian and New World forms, has been the subject 01sorne debate over lhe past two deeades (Kurtén, 1968; Hendey, 1974a, 1975, 1978; Galiano and Frailey, 1977; Kurtén and Anderson, 1980; Berta, 1981). A study ot the Eurasian and New World material now in progress may help to clarity a number 01 the taxonornic problems (Kurtén, in Jitt.), and until that ls completad there seems little point in attempting to relate the African material to Eurasian laxa currently given in the literatura. One point that remains worth stressing al this stage, however, is the length 01 time over whích these now extinct hyaenas coexisted with the modern species in Africa. The size range of the lower first rnolars of Canis ct. mesame/as from Member 3 places a number ot the animals in the upper size ctasses for that species given by Klein (1986) in a survey of the assoclation between size and latitude in extant and fossil Sauthern African carnivores. Aecording to Klein's figures, three of the Swartkrans teeth tal! at the top of the size ranga fortha substantial fossil samples of large [ackels from the cool-cllmate deposits at Equus Cave and other sites. The Equus Cave deposits contain sorne of the largest jackals in Klein's sample, althcugh the assemblages trom the site do not have the highest average values. Klein argues that the correlation between latitude and size observed in modern samples of this and other carnivora species permits gross climatic interences to be made based on sizes in fossil sampies. However, cautlous interpretation is warranted in the case of smaíl sample sizes, sinee, as Klein's figures show, the variation in tooth síze for a speeies at any given latitude may be considerable. This is particularly true in the case of his C. mesamelas sample where despite a sampJe size of 308 the sígnificance of the linear ccrrelation coefficient (r) ls only 0,397. In the Member 1 and Member 2 sample of Can;s el. mesame/as, breakage and wear make it diffieult to assess the size variatlon in relation to either the Member 3 sample or to Klein's fossil and modern samples. The sample o( leopard remains from the new excavations adds substantially to the already moderately large sample from the old Member 1 Hanging Remnant assemblage, and makes the animal the best represented of the larger carnívores and by far the best represented ofthe cats. Brain (1981) discussed the Iikely importance of leopards in the accumulation ot the bone assemblage at the site, although he al so stressed that many of the bovíds in the old Member 1 sample may have been too large to have been taken by the species, based on the work of Vrba (1975,1976). It is elear, given the range 01 predators represented in the deposíts and therefore by implicat~on active in the area during the period of deposit formation, that no one species can be singled out as the sale agent of accumulation. However, in view of the relative abundance of the leopards at the site, it may be worthwhile to consider the question of theír role in the assemblage formation a little further. As shown by studies of African carnivoras, such as those carried out by Kruuk (1972) and Schaller (1972), and as reeognized by Vrba (1976) and Brain (1981) in lheirdiseussion of fossil assemblage formation, the size of prey taken is related to the body size 01the predator. Sinca the leopard is a solitary hunter, it does not act cooperatively as Hans may do, to attack

163

prey much larger than itself. But it is a powerful animal, ano in generallhe bigger fhe leopard, the larger the prey that it may take. How big are the Swartkrans leopards? Ewer (1956a), in the eourse of her description of the older sample of leopard remains from the sites, concluded that her fossil subspecies was slightly smaller than the living species. Her comparative sample eonsisted of 'nine modern leopard skulls Irom various loealities in South Afriea' (Ewer, 19568: 85), but she gave no details of the sexual or geographic consñtution of the sample. Such an ornísston makes it very difficult to evaluate the relative size or her specimens, for reasons given below. The sample sizes for the toss¡l specimens are small, and there is litile point in employing sophisticated statistlcal te ehniques to examine the size distributions. particularly since sex-based size dlmorphlsm tends to eonfuse the issue in unsexed fossil samples of eats (Turner, 1984c). tnstead, Ishall consider the problem graphically. In Fig. 4, size dlstributions ot upper carnasslals are shown for a series of modern, sexed leopards from Southern Africa. Also shown in Fig. 4 are the size distributions ot carnassials fram the Swartkrans samples, grouped by Member using the figures given in Table 6. Considerable vartation ano overtap may be seen within and between the sexes, but it is evident that leopards from the Cape Province of South Africa tend to be smaller than tncse trom more northerly areas and that males from a given area tend to be eonsiderably larger than females. Far these reasons, it 15 pointless to compare a fossil sample with an unsexed and unprovenanced modern couecñon. It may be seen that the bulk ot the modern speeimens have a greatest lengfh falling between 22 and 27 mm, and thatthose from more northerly areas alone also vary by about 5 mm in totallength.lgnoring lhe very large lossil speeimen SKX 34193 from new Member 3, ene eould argue thatthe remainder ofthe fossil carnassials are entirely males, since the range falls within that of the larger modern males. However, inclusion of SKX 34193 impnes that the two specimens from Member 3, separated by just over 5 mm in length, are most likely to represent different sexes. In that case, the Member 1 sample is likely to be mostly females with one male, and the two antmals trom Member 2 are prabably af the same sex and probably female. Of course, if the environmental conditions, and hence the aceessibility of prey, differed significantly between Members, then we could not assume that aH the smaH fossjl leopards were females and all the larger ones males. Some of the smaller speeimens might be small males fram a period of general size reduction. Sut the fact that none of the speeimens faUsbelow the 24 mm greatest length, and that the total range of variation above that point closely approximates that seen in the northern leopards. implies that the range of variatíon for both sexes is represented in the foss1l sample. If that conclusion is correet, then there is sorne suggestion in Fig. 4 that the Swartkrans fossil leopards tended to be larger than those occurring in the area today, and the maximum possible size of prey taken may also have been somewhat greater than previously assumed. The new speeimens of Pro/eles from Member 1 augment the material previously described by Hendey. It is elear from the specimens reeovered trom the old Member 1 Hanging Remnant and old Member 2 samples that the extant species,

NEW FOSSIL CARNIVORE REMAINS

164

12

Lm

* •

•

•«•

•

10

o

• • •

• 20

o

• o

•

22

2.

2.

L

30

Fig.4 Total Jength (L) and Jength 01metastyle (Lm) 01 p4 in modern and fossutecpards. Closed cfrctes, modern males from Transvaat and Matawi. Open circles. modern tamales from Transvaal ano Malawt. Clcsed trtanqles. modern males from Cape Province. Open triangles, modern tamales from Cape Province. Clcsed stars, Swartkrans Member 3 specimens. cceec diamonds. Swartkrans Member 2 specimens. Ooen stars, Swartkrans Member 1 speclmens. Modern speclmens trom conections in the Transvaat Museum, Bernard Price lnstitute ter PalaeontologicaJ Aesearch (University 01the Witwatersrand) ano the South Atrican Museum. AII measurements in miñtmetres.

P cristetus, was present in post-Member 1 times, and that a dífferent species was present in Member 1. But the timing of the speclatlon event tnat led to the appearance ot the extant species remains unresolved. The new Member 1 material, like that from Kromdraai B, appears to be from a form that differs in size from the modern species and in which the frequency of retention in P2 may have be en greater, but lt may simply nave been a larger variety of the same sp ecies. Nor is it olear how long the older species, P. trensveeiensis, may have continued to exist following the speciaticn event. The new finds of spotted and brown hyaenas add to the

existing samples ot these species wímout shedding greater Iight on the palaeoeocology of either, or their particular rotes in the formation of the deposits. The finds ot Aonyx cepensis, the clawless otter, are the first from the hominid sites in the Sterkfontein valley, a\though Brain

(1981: 165) quates an unpublished identilieatian 01 A. el. capensisby H. B. S. Cooke in the material from the nearby site 01 Bolt's Farm. The species occurs in areas with standing and running water, implying a greater quantity ot surface water than that to be found in the valley taday, although it is also known to wander far trorn water at times (Haltenorth and Diller, 1980).

REFERENCES

BERTA, A., 1981. The Plio-Pleistocene hyaena chesmeoorttetes ossifragus from Florida. Jaurnal ot Vertebra te Paleontology 1(3-4):

341-356. BAArN, C. K., 1981. Thehunters or the humad? Unlversity of Chicago Press. Chicago. EWEA, R. F., 1955a. The tossu carnfvores of the Transvaat caves. The Hyaentdae, other than Lyeyaena, of Swartkrans and Sterktontein. Proceedings ot the Zootoqicei Soeiety 124: 815-837. EWER, R. F., 1955b. The tossü carntvcres ot the Transvaat caves. The lycyaenas ot Sterkfontein ano Swartkrans, together with sorne general considerations of the Transvaat fossit hyaenids. Proceedings ol the Zoological Society 124: 839-857. EWER, A. F., 1955c. The fossil carnivores of the Transvaal caves: Machairodontinae. Proceedings ofthe ZoologicalSociety125: 587-

615. EWEA, R. F., 1956a. The fossjl carnivores 01 the Transvaal caves: Fellnae. Proceedings of the Zoo/ogieal Society 126: 83--95. EWEA, A. F.. 1956b. The fossit carnivores of the Transvaal caves: Canjdae. Proceedings of the Zo%gical Society 126: 97-119. EWER. R. F.o 1956c. The lossil carnivoresof the Trans\laalcaves: two new viverrids logether with sorne general considerations. Proceed· ings of the Zoologicaf Society 126: 259-274. EWER, R. F.. 1967. The 10ssit hyaenids of Atrica - a reappraisat. fn' BtSHOP, W. W. and ClARK, J. O., eds, Background to evolution in Africa. pp. 109-123. University 01 Chicago Press, Chicago.

GAlIANO, H. and FRAILEY, D.. 1977. Chasmaporlhetes kani, new specles from China, with remarks on phylogenetio retationships of genera within the Hyaenidae (Mammaua, Carnivora). American Museum Novitates 2632: 1-16. HALTENORTH, T. ano DILLER, H., 1980. A field gu;de ro rhe rnammals of Africa. CoUins, London. HENDEY, Q. B., 1973. Carnivore remains from the Kromdraai australopithecine site (Mammatia: Camfvora). Annals ot the Transvaal Museum 28: 99-112. HENDEY, Q. B .. 1974a. New Iosai' camivcres from the Swartkrans australopithecine site (Mammalia: Carnívora). Annafs of the Transvaal Museum 29: 27-48. HENDEY, a. B., 1974b. The Late Cenozoic Carnivoraotthe south-western Cape Province. Anna/softhe SouthAfriean Museum 63: 1-369. HENDEY. Q. S., 1975. Relationships al North American hyaenas. $outh African Journa/ ol Science 71: 187. HENDEY, Q. B., 1978. Late Tertiary Hyaenidae from Langebaanweg, South Africa, and their relevance to the phytogeny 01 ¡he lamily. Annals 01 the Sourh African Museum 76(7): 265-297. KLE\N, R. O., 1986. Carnivore size and Quaternary climatic change in southern Africa. Quatemary Research 26: 153-170. KURTÉN, B.. 1968. Pfeistocene mammals of Europe. Weidenteld and Nicolson, london. KUATÉN, B. and ANDERSON, E., 1980. Pleistocene mammals 01 Norrh America. Columbia Universit"y Press, New York.

NEW FOSSIL CARNIVORE REMAINS

KRUUK, H., 1972. The soonec hyaena. Chicago Universily Preas, Chicago SCHALLER, G. B., 1972. The Serengeti /ion. University ot Chicago Press, Chicago. TEILHAAD DE CHARDIN, P., 1939. On two skulls of Machairodus from the Lower Plelstocene beds of Choukoutien. Bulletin o, tne Geological Society al China 19: 235-256. TURNER, A., 1984a. The Interpretation ofvariation in fossü specimens ot spotted hyaena (Crocuta crocuta Erxleben, 1777) from Sterxtcntein valley sites (MammaJia: carrworaj. Annals ot me Transvaal Museum 33: 399-41 a. TURNER, A., 1984b. Panthera crassidens Broom, 1948: the cat tnat never was? South African Joumal of Science SO: 227-233. TURNEA. A., 1984c. Dental sex dtmorptusm in European lions (Paothera leo (L.)) ct the Upper Plelstocene: palaeoecoloqical and paJaeoethoJogical implicaticns. AnnaJes zoologici feooici21: 1-8. TURNER, A., 1985. Extinetion, speciaticn and dispersa! in African largar camfvores, from the Later Miocene to Recent. Souttv Africao Joumal of Seienee 81: 256-257. TURNER, A., 1986a. Miscellaneous carnlvore remains from PltoPtetstocene deposita in the Sterkfontein vauey (Mammalia: Carnivora). Aooa/s of the Transvaal Museum 34(8): 203-226. TURNER, A., 1986b. Sorne features of Arncan larger camlvcre bioqeoqraphv. Palaeoecology of Afriea 17: 237-244.

165

TURNER, A., 1987a. Megaotereon cútuídens Irom Püo-Ptefstocene ---'age deposita in Atrica and Eurasia, with comments on dlspersat and the possfbility of a New World origin (Mammalia, Feíidae. Machairodontinae). Journal ot Paleootology61(6): 1256-1268, TURNER, A., 1987b. New tossn cartuvcre remaíns from the Sterktcntein hominid sita (Mammalia: Carnívora). Annals ot the Transvaal

Museum 34(15): 319-347. TUANER, A., 1989. Late Neogene/Lower Pleistocene Felidae al Atrica: evolution and dispersat. OuartarpalaeontoJog;e 8: 247-256. VOGEL, J. C., 1985. Further attempts at dating the Taung tulas. ín: TOBIAS, P. V., ed.. Hominid evofution, pp. 189-194. Alan R. Uss lnc.. New York. VRBA, E. S., 1975. Sorne evidence of chronalogy and palaeoeccloqy 01 Sterkfontein, Swartkrans ano Krcmdraa! fram the tossn Bovidae. Nature 254: 301-304. VRBA, E. S., 1976. The significance 01bovtd remains as lmñcators ot enviranment and predation patterns. In: BEHRENSMEYER, A. K. and HILL, A. P., eds, Fossils in the making, pp. 247-271. University of Chicago Presa, Chicago. VRBA, E. S., 1985. Early hominids in southern Atnca: updated observations on chronological ano ecological background. In: TOBIAS, P. V., ed., Hominid evotuíion, pp. 195-200. AJan R. Liss Inc.. New York.

Chapter 7

Stone Artefact Assemblages from Members 1-3, Swartkrans Cave J. D. Clark

Oeoeameot of Anthropology, University ot California, CA 94720. USA

The artefact assemblages from each 01 the three Swartkrans Members show no significant difference and can all be assigned to the ccre/cnopper/nake traoftlon, although tnere Is a suggeslion that biface technology might alsc be assoclated in Members 2 and 3. As the full range of flaking üebrls is absent and no flaking tloor can be oenuñeo. it is apparent that the assemblages are lncornpfete and tna: they rnost probably reached their position in tne cavem by falling in from the sida ot the hill outsfce at dltterent times. or by being cerned in by homlnlds. This ts confirmad by the horizontal and vertical distribution pjots which show that artefacts are more usually scattered, in particular vertically. rather than concentrated. The Member 3 distribution, however, lmptiea that artetecte may neve been usad in the cave itseñ, wntcn by tnen was a more cpen ñeevre. The cultural record from Swartkrans is a significant ene among the oldest databte contexts from Southern Africa.

AII stone excavated by C. K. Brain lrom Swarlkrans Members 1-3 was preserved and torms the bas¡s tor the present study. The material was analysed and measured, with the exception of the angular waste from Member 1, which was counted but onJy 200 píeces measured. NATURE Of THE OCCURRENCE

The assemblage in the breccia 01 the Lower Bank 01 Member 1 forms part of a talus cone so that the strata dip at a steepish angle, and thin lenses of angular stone fragments show the s/opa of the bedding; a similar dip can be seen in Member 2. Since the stone tragments, consisting almost entlrely of quartz and chert, are not compactad together, it is less Iikely that they were fractured under pressure in situ; rather, they were broken as they entered the cave from the outside, as sorne weathered examples $u9ge5t, or, as in the case of the great majority, in fresh condition, by being derived from solution 01 the walls of the cavern system itself. In all three Members, artefacts and natural stone occur isoleted and at different levels. In the later stages ot the excavation (C. K. Brain, in liN.) it was found that the central area of the cavern, in front ot a very large dolomite block, was almost f1at during the accurrunanon 01 Members 1 and 2 and tha! the back parl 01the lloor 01!he Member 3 guliy was also relatively tlat. It would have been possible, therefore, lor hominids to have used those parts 01 the cave for shelter and sleeping, in relative comfort, and sometimes to have carried artefacts ¡nsida. There ¡s, however, no indication from the artefact assemblage5 from each of the three Members, that thay represent in situ flakíng areas and, in fact, the interior 01 the cave (particuJarly during the accumulation 01 Members 1 and 2) Is likely to have been dark and not conducive to tool

manufacture. An undisturbed flaking area will show a fuI! ranga of tlake sízes and types and, in this case, there is an almost total absence 01 the smali t1ake category in the 10-20 mm range. This and the overall paucity 01 debitage as shown by the small number of preces in each assemblage, make it unlikely that tool-rnakinq took place in the cave. Rather, il seems that the artefacts reached the positions in which they were found as a result of rolling down-slope through the cave entrance on the hillside and that the flaking areas must, therefore, have been outsrde the cavern itself, presumably associated with sorne immediately adjacent rock outcrop, or clase to the stream-bed which was the source of the cobbles used for some of the artefacts. The nature of the occurrences at Swartkrans when Members 1, 2 and 3 were accumulating is therefore one in which the artefacts reached their positions in the cavern by having fallen in, having been thrown in or sometlmes. perhaps, having been cartied in at times of hornlnid visitation. RAW MATERIAL

Quartzite occurs in the form of mostly sub-angular but wellrounded cobbles and was therefore available in the immediate vicinity in the Blaaubank stream. The cobbJes were generally of fist size, which possibly shows intentional selection as Fig. 3 suggests. Quartz occurs rarely as cobbles or pebbles but most commonly as fractured blocks, chunks and fragments, mostly in fresh condjtion though sometimes weathered. Both quartz and cher! (see below) break up readily along jolnt planes. In one instance, however, a weathered block of quartz trom Member 3 had broken ¡nto seven pieces that could be refitted. None of these pieces was unquestionably a flake, and two only

168

STONE ARTEFACT ASSEM8LAGES

doubtfully so, and normal pressure from the weight o. the., show no signs ot having been used as manuports since it is overburden is more likely to haya been the agency involved in . not possible to show that they were intentionally brought lnto the breaking. Whereas much of the quartz and chert from the the cave by hominids. They could have fallen in naturally with country rack appears to haya besn naturally introduced and other components 01 the assemblage. fracturad, this is not the case wtth all pieces ofthese rocks and Artefacts have been divided into three major categories: there are sorne that possess the usual hallmarks of artificial retouchedlmodified, modifiedledge-damaged and eores and fracture and rnodlñcatíon. unmodlfied waste (Fig. 1; Table 1). The first of these is further Owing to the small number 01 artefacts from Member 3 it is subdivided into heavy-duty and ilght-dutycomponents. It musí not possible to show any undoubted rncrease in the use ot be emphasized that very few, if any, artefacts show the kind of either of these rocks through time, but they are the predomiregular and unítorrn retouch associated with scrapers and nant raw materials used during the Middle Stona Age (Memother unitacial 10015. Retouch, if such tt ls, takes the form of interrupted flaking 01 an edge, sornetimes steep, sometimes ber 4) and were. presumably, locally obtalned. Chert occurs as angular blocks and chunks and as flat plates shallow, and notching. Much 01this ts also marginal or, al mast, serni-invasive and it grades into examples where the sporadic that break up into thin, plate-like fragments. Experimental flaking shows that chert fractures very erratically and only one flake scars rasemble modification rather than retouch. At the or two flakes with bulbs of percussion result; the rest of the lower end of the modification/retouch category, artefacts pass block breaks along fracture planes ínto a number ot ñat, into the modified/edge~damaged class. Modified pieces are angular pieces. A search revealed sorne nodules that are more examples where the primary form has been modified by the homogeneous, producing conchoidal fractures and, clearly, removal of one, two or more flakes but does not show any the Middle Stone Age inhabitants of Swartkrans were selective regular retouch. Edge-damaged forms are those that show in the chert sources they used. srnall marginal scars, both interrupted and continuous, normal The only other raw material s found are rare examples ot and/or inverse, on part or all ot ene or more edges. This may be the result of use or, in view of the nature of the occurrence, diabase and shale, the tatter in flat, angular fragments that show no definite signs of artificial fracture, but the diabase has might equally wj311 be due to natural causes. Cares have been been used for trie production of f1akes, notably during described as casua/where one, perhaps two. flakes haya been removed. They have also been described on the basis ot the Member 3 times. number of striking platforms - single, two or three p/atformsPHYSICAL CONDITION and as bipolar, biconicet and disc cores. Cares on cobbles that have been Ilaked bifaeially along a lateral edge cr end ('pebbie By far the majcnty of artetacts, fragments and chunks are in cores' in Kleindienst, 1962) have been classed as corelchopfresh condition with sharp edges. A few, however, show sorne pers and placed in the retouchedlmodified category without degree of weathering and abrasion and are described as necessarily implying secondary use as choppers. Leakey's 'slightly abraded.' Very rarely there is a piece that ie more (1971) chopper Iypes are also given. heavily abraded or weathered. It ts obvious that when the Angular fragments (possible flake fragments) and chunks material originally entered the cave, it was already in the (angular blocks) have been mcluoed with the unmodified condition in which it was found and that it had not subsequently waste (Table 1; Fig. 2) but, sinee a hlgh proportion olthese are been subjected to further weathering. The sxceptions are the probably of natural origin, lhey have also been classified few diabase flakes that are all heavHy weathered as a result, separately (Fig. 2; Tabie 2). it seems, of incorporatíon in a wet matrix where in situ weathering had been able to take place. TECHNOLOGY TEAMINOLOGY AND CLASSIFICATION Leakey's (1971) c1assification of Oldowan, Oeveloped Oidowan and Acheulian artefacts is used here with the exception that 'choppers' are reclassified as core/choppers since, from the studies 01 assemblages at ülorgesailie (Isaac. 1977), at Koobi Fora (Bunn et al., 1980) and the experiments of Toth (1985) and olhers, It is clear that the primary use of these artefacts, as was earliar appreciated at Sterkfóntein by Masan (1962), was as cores for the production of flakes, even though sorne ofthem may subsequently have been used as choppers. Core classification follows that suggested in the Kalambo Falls monograph (Clark and Kleindienst, 1974). The six flake types (Types 1-3 with cortex platlorm, and 4-6 with plain platlorm and showing progressively later stages of flaking) are those established as meaningful by Toth (1985) following hls experimental studies and analysis 01 Oldowan and Developed Oldowan assemblages in East Turkana and elsewhere. It was considered preferable not to refer to the natural cobbles that

No difference could be discerned in the flaking methods used to produce the assemblages from each of the three Members. AIl flaking is by hard hammer but it is of ¡nterest that no certain hammer·stones are present - un[ess they are the subspheroids (e.g., Figs 12:1,2; 21). Quarlzite and quartz were worked by bipolar lechnique, the cobble or chunk being rasted on a rock anvil and struck at the upper end with a hammer5tone. This resultad in typical bipolar splintering where smaller quartz chunks are concerned and, in the case of quartzite cobbles, in the removal of a f1ake or f1akes from one end and evidence of crushing at the other. Several typical examples are found (e.g., Flgs 8: 1,2; 16: 1,2). Blfacial flaking, using the scar on one face as the platform for removal of a new flake from the opposite face and so alternately forming a wavy cutting edge, is well seen in the core/choppers (Figs 9:1; 14:3.4; 15:1,2; 19:7,8). There is also sorne evidence for unifacial retouch of an edge on light-duty artefacts; this edge may be convex, concave ordenticulate (e.g., Figs 10; 1,2; 17: 1.2). Splil cobbles are somallmes Ilaked as single-platlormed cores (Fig. 14:2).

STONE ARTEFACT ASSEMBLAGES

Sometimes the right-angled edge so formad shows several smaller sears on the platform (e.g., Fig. 10:2) that eould be the natural product 01 flaking or could result from the intentional preparation 01 the platform, or from use. In the latter case, the specimens might be classified as heavy-duty core-scrapers. None 01the examples from Swartkrans, however, appears to warrant such reclassification sinca the chipping is not sufficiently definite or regular. The only lndlcation 01any difference in the technique shown by the assemblages from the three Swartkrans Members is the presence 01what appears to be a very weathered Proto-Levallois flake and a fine spheroid from the surfaee of Member 3 (Fig. 22) (lhese may be younger than Member 3 itself - Editor). ARTEFACT ASSEMBLAGES The artefact component af each ot the three assemblages Is shown in Table 1 and Fig. 1, by majar categories and classes, exclusive of fragments and chunks. Figure 2 shows the total stone assemblage broken down by raw material and with the fragments and chunks shown separately. There appears to be no significant difference between the three except that the assemblage trom Member 2 contains the greatest number 01 artefacts and there are appreciably fewer in Member 3. The same classes occur, however. The number ot quartzite cobbIes rematns much the same in each of the two earlier Mem· bers but there is a signíficant difference in the percentages ot chert and quartz, as between Member 1, where chert predorninates, and the two later Members where chert is replaced by quartz. Table 2 shows the maximum, minimum and mean measurements of fragments and chunks and raw material details. Fragments in the two later Members increase in size and possibly it ís the same with the chunks, though, again, numbers are too small to be siqnificant. These differences are essentially thase reflectad by the fragments and angular frag· ments, most of which could be natural, so that the significance of these changes is unclear, since the number of flaked artefacts is too small to show whether this could be a reflection of intentional selection on the part ot the tool-makers. In any case, we are not dealing with discrete assemblages but rather with pieces that haya found their way at different times into the cave. Figure 3 does suggest that there is an optimal size and shape for core/choppers when length and breadth measure· ments of these are compared with those of casual cares and cobbles frem each of the three Members. Casual cares have been included here as they are believed to represent the initial stage in splitting and reduction of a core/chopper for the production of flakes. It appears that bread and relatively thick cobbles were preferred aUhough this could have been the predominant form present in the stream-bed. There are, however, no very smaJl examples such as those from the DK1 site al Olduvai Gorge (Leakey, 1971). AII Ihree Members have generally símilar forms, the great majority of cobbles, casual cares and core/choppers bejng longer than broad but not markedly so. Figure 4 shows a similar distribution diagram for the flakes, which are predominantly short, irregular, end-struck examples mostly clustering between 60 x 60 mm with sorne larger flakes, usually primary cartex examples. Tabla 3 gives details

169

of flake measurements, raw material and flake type. Noticeabte is a considerable difference between the maximum ano minimum Ilake size, a difference that probably depends on the size of the original piece selected. Mostly the quartzite and chert flakes are longer, and sorne ot the quartz examples are quite srnall and derive from smatl angular chunks, sometírnes worked as bipolar cares. Comparing the mean measurements, however, these are remarkably consistent as also te the predominantly fresh condilion of the material. The flake lype shows the extent to which reduction of a core or cobble has been taken. There are too few f1akes from Member 3 to be diagnostic but, whereas flakes from Members 1 and 2 are predornlnanny of Type 1 (cortex flakes), lhere is a range ot other types which, if Types 5 and 6 are combined, outnumber the Type 1 exarnples. It is olear that this is a result of care reduction that 90es beyond any primary flaking ot a cobble or chunk, showing that cares were worked from two or more platforms befare abandonment. This is important as it elirninates the possibility that the cortex f1akes and casual cores might be the result of the cobbles' faUing into the cave and being fractured on irnpact with a hard surface, or from other rocks' falling onto them. In a normal flaking assemblage, however, cortex flakes form a srnall part only ot the total flake types. The only other mam class of artefacts represented rs that ot poiyhedrons (e.g., Figs 9:2: 16:4,5). These appear lo be the eutcome ot attempts to remove flakes from severa! dlñerent striking-platforms ano they were discarded when the platform edges became too obtuse for removing flakes. Sorne of these may have been subsequently used as harnrner-stones and, by degrees, became reduced to sub-spheroids. Polyhedrons are mostly ot quartz (12), one is of diabase. two of quartzite and two of chert. There is no significant difference in size or form between the polyhedrons in the three Members. The characteristic artefacts of each Member assemblage are discussed below: Member 1 The Member 1 Lower Bank breccia yielded 402 pieces ot fractured stone of which 340 are angular, small fragments and chunks, mostly of chert and quartz. Sorne of these are arte· factual, but (hey cannot be readily separated from the naturally fractured pieces. ot the 62 artefacts that the Member pro· duced, there was one core/chopper worked bifacially at the dislal end (Fig. 9: t); lhree polyhedrons (Fig. 9:2), one Iightduty, steeply retouched, single side~scraper on a weathered fragmenl 01 quartz (Fig. 10:2) and a similar flal chunk wilh steep scraper retouch on the distal end and part of one lateral edge (Fig. 10:4). There is also a small f1ake01quarfz modilied by allernate fiaklng to form a bee (Fig 10:8). The nine modified/edge-damaged pieces are mostly flake fragments of chert and quartz with marginal nibbled retouch, sometimes dentícu· late, on one or more edges or ends (Figs 10:3-7; 11:2,4.5). The casual ceres or cobbles and angular chunks (Fig. 8: 1,2) clearly show the method of splitting and fracture by bipolar technique and most examples show incipient flake scars and crushing on the anvil end that result trom indirect percussion from the hammer's striking the upper end to detach the flake. Flakes are mostly end·struck, in fresh condition and of quartz and cherf (Fig. 11: 1,3,6) and, in general, relalively broad. The flakes are mostly ot Type 1 with cortex or point platform and

170

STONE ARTEFACT ASSEMBLAGES

cortex on the dorsal face, but a proportion are 01 Types 4-,6, where the flake has a plaln platform showing evidence 01' bidirectional f1aking; platform anglas vary between 80° and 131°.01 lnterest al so are two small angular pieces 01 quartzite and chert that show con cava spalling such as is associated with tiest fracture; there ts, however, no discolouration as ís cften associated with heañnq in a tire. 11 these are tire-fracturad pieces, they probably result from natural bush tires across the hillside (Fig.11:7.8). Member 2 (Flgs 12-18) This Member contains the greatest number 01 retouchec/modilied and modified/edge-damaged artelacls bul a reduced number 01 angular fragments and chunks, which are now moslly in quartz, so lhat the total number 01 pieces olfractured stone (403) is virtually lhe same as thal Irom Member 1 (402). Among the heavy-duty retouched/modilied artefacls, the commonest are polyhedrons (10) (Fig. 16:4-5). 01 these, saven are 01 quartz and one each 01 quartzite, chert and weathered diabase. These artelacts are interpreted as multifaceted cores and as the end-product 01 lithic reduction through forms such as eore/choppers and bícomcat cores (Tolh, 1982, 1985). There are also two sub-spheroids in quartz with the ridges between the facets well-battered, perhaps from use as hammers (Fig. 12: 1,2). Core/choppers total seven, with four of quartzite, two 01 chert and one 01 quartz. Five are bifacially worked on the distal ene (Fig. 14:3,4) and two have one lateral edge bilacially worked (Fig. 15:1,2). Length/breadth measurements are recorded in Fig. 3. Numbers of flake scars vary between 2 and 5. In état physique three are fresh, three slightly abraded and cne more heavily so - conditions acquired before incorporation into the cave breccia. Light-duty artelacts (Figs 17, 18) comprise side- and enestruck flakes that show marginal micro-ehipping and are classified in the retouched/modified eategory when this is more regular (Figs 17: 1,2,8; 18: 1,2). Alllie within lhe mean range 01 unmodified flakes; the raw material ís quartz ano quartzite. There are also three artefacts, two in quartz and the third in quartzíte, that have been elassified as borers with what appears to be retoueh on two convergent lateral edges to form a blunl point (Fig. 18:3). The lhird example also appears lo show bilacial flaking (Fig. 18:5). One small chunk 01 quartz exhibils two burin-type scars opposed to a single sear and might be called an asymmetric, dihedral burín but is here classifled as 'lechnical' (Fig. 18:4). Among the modified/edge-damaged pieees are the flakes with discontinuous and irregular mlero-chipplng on an edge or edges (Figs 17: 1,2,8; 18: 1). Where this chipping is on bolh the dorsal and ventral faces and the artefact is in fresh condition, it could be due to its having been used for cutting but, if abraded, the chipping is more Iikely to be natural. There are 11 angular chunks that show variable modification and marginal steep scarring of an edge that, jf more regular and consistent, might be classified as core-scrapers, but these examples, especially those in quartz and chert, are more likely to be naturally worked (Fig. 18:6-8). One specimen (not iIlustrated) is of interest, however, and is an ellipsoid, split cobble 01 quartzite measuring 153 x 102 x 49 mm. The distal end ls snapped and missing but the right lateral edge has been

modified by the removal of two or more flakes from the ventral face and two from the dorsal. If flaking hact continued this would haya been classified as a side-chopper but it ls not ímpossiote that it could represent an initíal stage in the manufacture of a biface. The casual cores (Fig. 14: 1) are all 01 quartzite and similar lo lhose in Member 1. Bipolar cores (6) are atl 01 quartz (Fig. 16: 1,2). They snow charactenstic splintering and crushing at one or both ends and are what is to be expected where quartz is the material used. Lengtlls vary between 26 and 96 mm. 01 special interest are two cores that can be classified as biconical (Fig. 15:3.4). Both are en quartzite cobbles, one being Iresh and the other sligh~y abraded. They have been radially worked on both the dorsal and ventral faces for the rernoval ot ñakes, ano show a biconvex or pyriform cross-sectien. Perhaps the rnost significant artefacts are twe disc cores, also fresh, in quartz and quartzite (Fig. 16:3). These are typical, radially flaked disc cores such as are found wlth the Oldowan, Developed Oldowan and the Acheulian. The remaining cores are mosny single plalform examples (9) (Fig. 14:2). They are made on cobbles and chunks, two in chert, two in quartzite and five in quartz. Two of the quartz examples have two platforms at right angles and one of those in quartzite has three plattorrns. Flake details (Fig. 4; Tables 1, 3) show a slight increase in dimensions over those from Member 1. There are a tew large, bread, convex Ilakes (Types 1 and 2) struck Irom quartzite cobbles, that exhibit marginal chipping which is. however, probably natural (Fig. 13: 1-3). The majority (23) are irregular, end-flakes (Figs 17:1,2,5-7; 18:1) and only seven are sideflakes (Figs 17:3,4,8; 18:2). Cortex plalform Ilakes (Types 1-3) are the commonest forms but there are ten flakes with prepared plain platforms, six ot which belong to Type 5. With one exception, all are in fresh condition. Cobbles are similar to these found in Member 1 (Fig. 3). Member 3 (Figs 19--22). This Member fllls a deep fissure or gully eroded into Members 1 and 2 along the cave's west wau. There are appreciably fewer artefacts and there is less angular waste than in elther of the earlier Members (Table 1) and, owing to the small number 01 Ilaked artelaets (33) and the absence of any diagnostic forms, no reliable assessment of the affinities can be made. There are, however, certain pointers that suggest an Acheulian contexto This small assemblage is also a randem one that most probably reached lts position in the gully as a result of natural gravitation down the slope, or of having been carried in. With few exceptions, the artefacts cal1 for no special comment. The care/choppers (Figs 19:7,8; 20:2) are similar to those from Members 1 and 2 except that there is now somewhat more Ilaking down a laleral edge and end. Artefacts shown in Fig. 19:8 and Fig. 20:2 begin to resemble 'prote-bifaces' and there is one large example on an angular chunk ef chert in which randem bifacial flaking ef ene lateral edge recalls a similar specimen from Member 2 and could be an incipient roughout for a biface. There is one geed example of a biconical core (Fig. 20:3). a dlscoid core (Fig. 20:4) and a quartzite spheroid (Fig.21). The only other heavy-duty artefact has been classed as an asymmetric, double side-seraper (Fig. 20: 1). It ls made on a

171

STONE ARTEFACT ASSEMBLAGES

Table 1 largo side-flake 01 quartzite that has Swartkrans artetact ñst. been retouched along both lateral edges on the dorsal tace so that the piece has Artefacttype Member 1 Mernber 2 Member 3 Total a plano-convex cross-sectlon. The distal Lower 8ank end is snapped and rnissinq but it is possible that the specimen couid have Retouched/modifled been lntended as a cteever. Heavy-duty: There are only seven measurable 1 Double side-scraper/?cleaver 1 Ocre/chopper - end 5 2 complete flakes and again lhese do nol 8 2 1 3 side appear to be different from those in Memside & end 3 3 bers 1 and 2 (Fig. 4; Table 3). They are Polyhedron 10 1 14 3 mostly in fresh condition, are predorni1 Sub-spheroio 2 3 1 Spheroid 1 nantly 01quartz and show a similar ranga Light·duty: 01Ilake types lo those in the other MemScraper - side 4 5 bers. A selection of whole and snapped 2 3 end flakes (Fig. 19: 1-6) ís included lor comaorer 3 3 Bec 1 paríson. These are mostly thin, and the Burin (technical) 1 faet that they are also snapped across Total Retouched/modified 29 10 46 the short axis, exhibit plain ano simple7 faceted platforms and uní- and bidirecModl1iedledge-damaged 4 ti anal ffaking. suggests that sorne 01 Flake 1 3 6 6 these flakes could be waste products 01 Flake fragment Angular chunk 11 13 2 biface manufacture. Total Moditied/edge-damaged 14 23 9 There are two additional artefacts that could be important far assessing the afCeres: Unmodi1ied waste Ca res: Casual - cobble 4 11 7 finities 01 this assemblage. At the south 4 angular chunk 10 6 end 01 the ñssure, where it narrows, the 7 Bipolar 6 1 two artefacts illustrated in Fig. 22 were 2 1 Dlsc 3 recave red. They were found superñ2 1 3 Btccnical 2 Single plattcrm - cobble 2 cially, befare any stratigraphic control angular chunk 7 8 was introduced, ano there is uncertainty Two-plattcrm 2 2 as to whether they were derived from 2 Three-platform 1 Member 3. One (Fig. 22: 1) ís a very 48 26 Total cotes 13 9 weathered flake 01 diabase. The bulb is Flakes - end-struck 10 23 15 48 flat, the platform shows three faceting 6 7 4 17 sfde-struck scars and the dorsal face appears to fragments 61 12 299 226 Angular chunks 27 221 show the remains ot scars resulting from 362 114 Cobbles 10 22 32 radial flaking of a large coreo This, there2 2 Fire-traeture:.:d=--fore, is believed to be a flake from a 360 62 808 386 Proto-Levallois coreo The spheroid (Fig. Tota,,-I.:cco,,-r.:c0.:csl.:cu,,-n::m:.:o:.:d::ifl.:ce.:cd-,w.~a:.:sc::!e=---__ :.::..::.:..-~ 72 877 402 403 22:2) is of quartz and shows inclpient Grand total concholdal fracture over the whole sur-----------~------------~------face to produce the only well-made spheraid from the site. lt is important also to note here that in coordinates ter most of the archaeological stone assemblage. the collection 01artefacts from the dumps described by Leakey This makes it possible to show the position 01 artefacts in the (1970). there were three hand-axes, a paralleí-sided cleaver cave in relation to the 1-m grid squares 01the permanent metal with straight butt and a pick-like tool. The hand-axes are ovate grid and the depth within 10-cm sptts at which the artefacts to elongate-ovate in plan lorm, and were of quartzite, diabase were found. In the earlier stages ot the excavation, depths in and dolerite. A further pointed ovate and very weu-meoe excess of 10 cm were sometimes used tor the spits. In such hand-axe in ?horntels was picked up by C. K. Brain from the cases, to maintain uniformity, the artefact depth has besn dumps but its provenance in the sequence is unknown. It is plotted to the nearest 10 cm to the spit mean. symmetrical and made on a side-flake from whích the platform The earlier stages 01 the excavetlon showed a steeply and bulb have been removed by retouch. The dorsal tace is dipping con e 01detritus that had entered the cave through the finely worked by soft hammer technique and it has a regular opening in the roof (Brain, 1981). C. K. Braln informs me that biconvex cross-section.lt measures 156 x 96 x 41 mm. as the excavation continued, it was found tnat the central area in front 01 an enormous dolomite block had been almost ñat ASSEMBLAGE DISTRIBUTION PATTERNS: MEMBERS 1-3 during the accumulation 01 Members 1 and 2 and that it was Subsequent to completing the analysis of the artefacts in also fairly flat in the back part of the Member 3 gully. Since, in Pretoria. C. K. Brain provided me with three-dimensional

=-

....:=

....::=-

---'=-_

--==__

STONE ARTEFACT ASSEMBLAGES

172

Table 2 Length measurements and raw material breakdown Ior fragments and chunks from Swartkrans Members 1-3.

Max. M.mUr 1 Fragments Chunks Member 2 Fragments Chunks M.mb.r 3 Fragments Chunks

Ouartz

Quarrzite Min. Mean

Max.

Mean

Olher

Nambers '

Total

Min.

Max.

Chert Mean

Min.

Max.

Mean

Min.

Ole

al

Cl

o

80

50.8

45

36 96

14.77 35.32

6 17

41 131

15,42 48.9

6 15

184

104,3

52

5

32 41

168 61

7

200 114

72 65

45.9 59.5

16 40

53 87

23.9 30.8

12 9

60 86

38.8 47.0

14 15

109

58.7

16

8 8

43 159

10 47

60

61 220

48

44.3

38

56 72

31,4 24.2

19

52 93

39.8 76,0

24 59

3

5 23

4 2

'Ote e quartzite; Qz

= quartz;

11

12 25

el_ chert; o _ oíner.

these places homlnids and other animals could comfortably have lived, the relationship of the artefacts to these areas is cJearly important. Accordingly, the numbers af artefacts in the gr\d squares and the depth rar-qes over which they occurred below datum. have been plotted for 68Ch of the three Members and are shown in Figs 5,6 and 7 on ptans of the cave system. Artefact totals differ from those shown in the analysed total s in Rg. 2 as soma artefacts are without provenance and sornetimes new data nave become available. Artefact assemblages in each of the three Members can be seen to differ and are dlstinctive as Figs 5, 6 and 7 cJearly show. Each assernblaqe is discussed below.

Member 1 Lower Bank (Flg. 5) The plot shows that there are two areas where artefacts are concentrated - in the northeast and southeast quadrants ot the grid. These are north and south respectively of the two larqe dolomita blocks previously fallen from the cave root. AII the artefacts of the southeast concentration - with the exception of two (one each in ES/53 and W3i52) -lie in Ihe eastem ofthe two flat áreas. The artefacts in squares E6/51 and E6/52 (13) are presumed to abut ihe southern dolomite block and not te overlie i1. Artefacts in [he southern flat area (57) range between 60 and 200 cm below datum but lhe majority (55) occur between 150 and 260 cm. There is nowhere any indication of a horizontal concentration in this profile and the 1,1O-m ranga is consistent with sporadic entry of artafacts inta the cavern as Member 1 breccia accumulated. The concentration in the northeastern quadrant contains the greatest number of artefacts (162) and almost all occur at a greater depth below datum. 11 they do not Ile on a sloping !loor then they represent the earliest artefacts to haya accumulated in the cave. Most were found jn an Brea situated west of the northern dolomite block and extending almostto the north wall. Only in grid squares E5/N3 and E4/N3 three and eight artefacts respectively abut or may in part Qverlie this northern block. Oepths at which artefacts occur range between 350 and 820 cm except tor ane surface specimen in square E2/N2 and one at 150-175 cm in E6/N2. Eighty nine artefacts occur in tour grid squares at depths between 490 and 750 cm - a vertical range of 260 cm. Again, looking at the numbers of artefacts within eaeh 10-cm spit, there are never more than

three arteracts in any one. Between 580 and 600 cm in the tour grid squares w;th the highest number of artetacts, there are 13 specimens present, which again le inconststent with identityjng an activity honzon. The profile shows that mast artefacts come from the lower tevels and this is more indicative of sporadíc accumulation due to natural causes. Member 2 (Fig. 6) Artefacts are concentrated in thrae secnons: 1) the flat area in the southeast quadrant (14 grid squares), 2) in the northern area (14 grid squares) north 01 the lwo large doiomite blocks, and 3) in the Brea between and to the east of these (11 grid squares). There are twe artefacts en the east side of the western flat area. In three squares, artefaets appear to líe over the northern and southern dolomrte blocks in the eastern quadrants of the grid. There ts also a scatter 01 artetaets to the south and/or east of the eastern flat area, The two first concentrations complement those seen in Member 1 and suggest that the main access to the cave had not significantty changed, although the scatter between the two large blocks might indicate that the entrance had become enlarged by Member 2 times. There is a significant difference in the depths below datum at which Member 2 artefacts occur. There are 13 specimens that come from the surface of grid squares, but this could be an artificial surface created in the 1930s by limeminers. In the excavation, specimens come from depths belween 10 and 1230 cm. but mostly between 20 and 190 cm as is seen in the profile (Fig. 6), as is the smaller number al artefacts trom the deeper levels in the northern area. Again, there is no significant eoncentration in any one 10-cm spit. There are five artefacts from square ES/SS between 70 and 80 cm, four artefacts between 140 and 150 cm, and nine between 150 and 160 cm. In E3/N4 a further five artefacts are plotted between 51 O and 520 cm though these figures, where the spit range is greater than 10 cm, may retleet only the mean selected for plotting the prefile as mentioned earlier. The grid square plat and profile together seem to confirm a sporadic accumulation of artefacts in the Member 2 breccia, the greater numberofspecimens being closerto the access to the outslde and those from the deeper spits in the outer cave reached their position by falling down-slope, though one cannot exclude the possibility of their having been carried into the cave during temporary occupation by hominids.

173

STONE ARTEFACT ASSEMSLAGES

Table 3 Length measurements and other detalla ot whofeüakes from Swartkrans Members 1-3.

Measurements Length

Breadth

Thickness Max. Min. Mean

Physical Gondition'"

material

F S/A VlA ate

oz

Min.

Mean

97

13

40,5

81

9

36,1

41

3

12,9

18

2

Member 2 104 Member 3 169

18

43,2

106

16

37,8

53

5

13,6

29

1

lB

46,8

139

15

46,1

49

3

15,2

11

t

Max. Member 1

Max.

Min. Mean

Raw

et

Flake type

Di

2

3

4

S

6

Total

11

3

6

2

4

4

4

21

12 13

5

9

6

5

6

3

30

1

3

4

1

4

14

2 2

.

----

2

9

1

2

-F '" fresh; S/A '" slightly abraoeo: V/A _ very ebradeo: Qte - quertate: Qz '" quartz; Ct '" chert; Oi _ diabasa.

78

80

~

60

Total 121

Total 62

~

Total 33

46

ca

O 40

2E

10

:J

Z 20

14 19

3 4

Rlm

MEMBER 1 Swartkrans Artefacl Assemblage (t.ese Fragments and Angular cnunks)

Rlm Mled C/uw

Miad C/uw

MEMBER 2 _

Light duty

MEMBER 3

I!!IJ]]] Heavy duty

RetouchedlModified Modífied/Edge-damaged Cores and Unmodified Waste

Fig.1 8reakdown ot artetact assemblages from Swartkrans Members 1-3 by majar categories, but not including angular fragments and chunks.

Member 3 (Fig. 7) Although the number of plotted artefacts from Member 3 is small, they all üe within the Member 3 gully, as is to be expected. They occur in the tairly flat area towards the backof this gully and tend to cluster over the eastern part of thia area in the southwestern quadrant of the grid. The depth coordinates range between O and 550 cm but the majority of the artefacts (42) cluster between 160 and 220 cm. This 60 cm thickness of spits is an indicatia n of possible temporary but sporadíc occupatíon of the gully by hominids during this period in the tormatton of Member 3 breccia. The published profile tor cne grid square (W3/S3} (Brain ano SiJJen, 1988) shows bone fragments concentrated between c. 270 and 470 cm, burnt

bane between 300 and 360 cm, and a single cut-rnarxeo bone at 440 cm. This grid square is recorded as producing 207 preces or fractured stone, mast ot it occurring between 250 and 460 cm depth. Mest ef this is presumably naturally fractured quartz and chert, as 1 record only seven pieces that show evidence of artificial fracture. Of tnese, three come from the 180-190 sptt. two from the 350-370 spit and three from between 430 and 450 cm, which is in general agreement with Srain and Si/len's protile. However, the protile in Fig. 7 shows that tt-e artetacts in my analysts are more trcquent in tha upper part of Brain and 5illen's profile, and in the spits that lie aboye this concentration of bone, burnt bone and fractured stene that they record Ior W3/S3. This implies that mast of the artefacts

STGNE ARTEFACT ASSEMBLAGES

174

were íntroduceo after mast of the bone fragments and burnt bone had been incorporated into the breccia, Sane tools lnW3/S3 occurred at much the sama spit levels as the lower íímrt 01the 'concentration' shown by the presant analysis.

SWARTKRANS RAW MATERIAL

% 100

AII artefacts and fractured stcne MEMBER 1

MEMBER 2

MEMBER 3

Tolal 373

Total 403

Total 68

80

,.-'"

DISCUSSION

--.-,

80

The artefact distributions in each 01 the three Members at Swartkrans províde no cartain evidence as to the agencies that inlroduced Ihe flaked stone assemblages and cobbles into the cave. None 01the assemblages show the ranga of artefacts to be found in a flaking area, and the type ranges and flake types suggest a random accumulation rather than the full ranga to be expected in a workshop area.ln this raspeet itean be clearly stated that this analysis does not show evidence for patterned flaking 01 stone in the cave. This does not, however, excJude the possibility that artefacts may from time to time have been introduced into the cave by hominids using it for temporary shelter. Since so me of the concentrations ot artefacts coincide wtth the flat areas of the cave floor when Members 1 and 2 breccias were accumulating. opportunistic introduction by hominids would have been quite possible, especially if Iight from the entrance shaft(s) penelrated lo this part of the cavern. On the other hand, artefacts, bone and other material falling into the cave from an activity area on the hillside close by can also be expected to have been more numerous closer to the access. The profiles (Flgs 5, 6), whiie showing times when flaked stone accumulated more frequently than at others, also show a smaH but steady addition of artefacts throughout the times when Members 1 and 2 were forming. The range of weatheringlabrasion categories shown by the artefacts are atso suggestive of intermittent exposure on the hills before falling into the cave. Without more knowledge of the circumstances under which the different breccias accumulated and the time-span involved, it is unlikely that more definite conclusions could be drawn from this analysis. It is important to know to what extent artefacts could have gravitated down-slope after having fallen or having been washed into the cavern or introduced by hominids. lt ls equally important to determine how much of the artefact and bone material that accumulated during Member 1 times might. after resoíuticn. have been reincorporated into the Member 2 breccia. It would also be important to know how easy access and exit would have been for hominids and anirnals wishing to make use ot the cavern. The Member 3 concentration and profile is more suggestive of hominid activity in the gully, especially when read in conjunction with the evidence from bumt bone and bone tools as shown by the W3/S3 grid square profile. At this time access may also have been easier.

CONCLUSIONS The evidence presented shows that none of the assemblages from the three Members carne from an in situ activity area in the cave. The full range of flaking debris is absent and there is no refitting of pieces. There is, however, considerable natural selection and the numberofartefacts is relatively small. This is lo be expected al a site of this kind where soil, angular

.C.J :...:...-~ .~

.~

~~

40

'-:-1 .~

~ 20

r-:....:...

..

~

¡

.'--<

oi

QI QzCh O

oecn o

¡7-:QIQzChO

%

80 -

Fractured stone only as percentage o, total assemblage Total 314

Total 262

Total 37

60

40

20

¡- -.

r

~.

~'-

.

¡: ... Qt azCh O

QI QzCh O

Qt

=

Quartzite

Qz

=

Quartz

Ch = Chert

al oz c» o O = Other

FIg.2 Breakdown of artefact assemblages from Swartkrans Members 1-3 by raw material and with angular fragments and chunks shown separately.

fragments, blocks, faunal remains, including that of the normnids, and some artefacts found their way into the cavern system mostly, the assemblage suggests, by gravitation down the entrance shafts from the hlll-stde outside. Comparing the assemblages from the three Members, there appears to be no significant difference in the core/chopper, core and ñake components and the retouched and modified pieces. The core/choppers are, as Leakey (1971) suggested from earlier analysis ot the artetacts from the durnps, of Developed Oldowan, rather than Oldowan type, mainly on the larger dimensions of the specimens. AII the main stages in the process of core reduction are representad from the selected cobbles: the initial splitting by bipolar technique, primary bltacial flaking through the casual core stage to the core/chopper and so on to the biconical or multi-platformed core to the polyhedron discard. The flake types also demonstrate this reduction very clearly from the

175

STONE AATEFACT ASSEMBLAGES

140

cortex flakes and plattorrns through to the tater stagas 01 reduction represented by Types 5 and 6. There is a suggestion that rather largar flakes are presant in Members 2 and 3 than in Member 1, while the rare exam-

pies 01 split cobbles or chunks, which haya been bifacially worked along one edge, may possibly be an indieation 01 bifacial technique, and that hand-axes might haya been a component, if a complete assemblage had been presen!.

Member 2 may have more in camman with Member 3than it does with Member 1 but it ts not possible to be sure on such small assemblages. There is reason to belleve that the artefacts from Member 3 are part 01 an Acheulian assemblage on the basis 01 the two examples from the south end 01 the fissure, and the handaxes and cleaver from the dumps. These last are unlikely to have come from the Middle Stone Age Member 4, so thal they more probably are derived from Members 3 or 2, or both, though the former appears more likely. Such an interpretation would be in accord with the inferred age 01 the Members. Member 1, dating

Member 2 Member 31 Á 1

'--... Member Core/Choppers • Casual Cores () ,Cobbles O

,1

•

D,

O

O

I

I

•

o·

•

100

o

o

cE

'O

t»JJ

'"~

CD

~()

o

6

O '\.

60

00 00 O

o"

~O

O

o o

O"

•

ó> O • () O o CI"o°

o

o

O O ~8."

o 20

20

40

60

60 100 Length (mm)

120

140

160

Fig.3 Scatter diagram to show length/breadth measurements 01 corelchoppers, casual coree and cobbles from Swartkrans Members 1-3.

(on faunal evidenee) from 1,8-1,5 m.y. S.P., would be the equivalent of an Oldowan assemblage from the East African sites, Member 2, dating between 1,5 and 1,0 m.y., probably would be Developed Oldowan but contemporary also with an early Acheulian Complex. Member 3, dating to approximately 1 ,O rn.y., would be contemporary with a later Acheulian stage, artefacts of which became incorporated in the gully from time to time, together with faunal remains. That the Member 3 assemblage was made by a more advanced hominid is also likely from Brain's discovery 01 burnt bone and bane fragments with cut marks in Member 3 (Brain, 1985, and Chapter 9, this volume). While utilized bone fragments with distal poíish and striations are found from Member 1 times onwards, evidence of fire ano butchery is present only in Member 3 times (Brain el al., 1988; Brain and Sillen, 1988), suggesting, as may also the stone artefact dlstributlon. possible occupation 01 the Member 3 gully. The Swartkrans topography must have offered sorne specially favourable features - a shelter, fissure or cave for protection and a possible source of water, special foods or raw materials tor tools in the vicinity - for the hillside to have been utilized for so long a time, presumably when the large camivcres were absent but, since erosion has removed the whole of the upper part of the cavern system, including the root, we are probably never likely to know precisely why lt was such a favoured locality. Finally, since both Australopithecus robustus and Homo sp. were present in Members 1 and 2, it may be asked who was the tool·maker? Did both these hominids make stone tools? The robust australopithecine líne ls known to have been present 2,5 m.y. ago in East Africa and it seems to have beeome extinet about 1,0-0,9 m.y. ago (Walker el al., 1986);

Horno habilis is present in the Lower Memberofthe Koobi Fora Formation in East Turkana al about 2,0-2,2 m.y. ago (Walker and Leakey, 1978) and is also presant at Sterkfontein in Member 5 (STW 53), established en bovid evidence (Vrba, 1975,1985) to be about 1,6 m.y. old ano somewhat olderthan Swartkrans Member 1 (Hughes and Tobias, 1977; Clarke, 1985b). The earliest Horno erectus from East and West Turkana dates to about 1,6 m.y. (Brown el al., 1985); stone tools appear by about 2,5 m.y. at the Gona River localitíes at Hadar in lhe Alar Rift (Harris, 1983) and al about 2,4-2,3 m.y. in the Lower Omo Basin (Howell, Haesaerts and De Heinzelin, 1987). Anatomically it would appear from the postcranrat evidence (Chapter 4, this volume) that there was no reasan why A. robustuscould nothave made stone tools il it had been expedient to do so and lt can hardly be doubted that a bipedal horninid would not hava made and used simple tools in a more advantageous manner than that seen in chimpanzee behavlour today (Goodall, 1986). The evolutionary hislory 01 the Horno lirieaqe is usually constructed as developing from a gracile austratopltheclne to Horno habilis and H. erectus and there can be no doubt of the increasing technological skills shown by the artefact assemblages with which, from H. habilis times onwards, they are assoctated - Oldowan, Developed 01dowan and Acheulian. Especially telling is the speed with which H. habilis, associated with the first stone tools in Sed 1 at Olduvai Gorge, (Johanson et al., 1987) and in the Koobi Fora Formation, was replaced by H. erectus. Such rapid morphological changes must be the outcome of very different patterns of social organizatian and economics. Stone·tool

176

STONE ARTEFACT ASSEMBLAGES

technology seerns to have been cne important adjunct to, achieving increasing cultural and technological skllls , Auetreloplttvecus robustus does not appear to show any comparable morpholoqical change and presumably, if it was Jargely vegetarian, it is less likely that stene tools would haya been a necessity far continuing lo make use 01 a traditional vegetarian dieto Recent work al Sterkfontein on the stane artefacts fram Member 5 and Ihe associated horninid (Clarke, 1985a) has shown that tne artefacts are similar to those from Swartkrans, except ter the presence al Sterkfontein 01 'two definite and two possible cleavers made on side-struck flakes, ane handaxe on a tlaxe and ane bifacial handaxe.' The assemblage is constdereo to be early Acneulian comparable to that from EF-HR al Dlduvai (Leakey, 1970) and lhat Hamo habi/is (represented by STW 53 and other fossil remains from Member 5) was the tcol-rnaker, While this may be so, sorne caution is still needad since it is not clear whether these finds came from resolution cavities in the Member 5 breccia. albeit eroded to a iow level, or whether they were recovered from in Sltu in the breccia itself. If the latter ís the case, there can be little reason to doubt an Acheulian association but, if the former is true, then artefacts and bone might have dropped down from aoove on different occesíons and further clarification is needed on this írnportant poínt. In favour ot contemporaneity ts the fact that fragments 01 SlW53 lrom the solution cavity and the bracera were found to join. The Swartkrans artefacts in Member 1 were most probably mada by an early representative of the Homo grade (Clarke et al .. 1970: Clarke, 1985b) and those in Members 2 and 3 by anatomicañy more advanced (H. erectus) torrns that had also begun to make use 01lire and to process meat by Member 3 times (Brain el al., 1988). The cultural record from Swartkrans ts an invaluable one

120l-···~ Flakes Member 1 . , Member 2 .... 100 Member 3 DI

o

... ...

80

•

E

.s oC

'5

•

...

• • 60

al

~

o

tIl

40

ib ...

20

....

... ....... o

ro

~

... L

0

4

... ...~

1.0

.t:Je-,

•

¡.

•

·fiP.. 20

I

40

60

BO

100

Lenglh (mm) Fig.4 Scatter diagram of whole tlakes from Members 1-3.

and, although it iS unlortunate that the assemblages are incomplete. these are among the oldest in datable contexts known from Southern Africa and show the general affinities of the different Industrial Complexes to which they belong. We are all deeply indebted to C. K. Brain 10r this meticulously excavated record made over the last twenty years.

/

I

~

• '

......1

1 ~I ,

NORTH I

I I I I I 1 1 1, I >

'.'-. -:

t

e

,

4... - 1

5

/

HAY ARU

__

"T"-. I

...

HAT AR[A

.1.

1

2

/

.'

,Li,~.,

~:

"'1,

11

00

_

./1

I 6

-'\.110:"'':'·11 a

t

t

e

___'o

'Olllh·'

...•1,'

.

'ttbt ~

•

-'-

•~

__

...--......

'

,

EAST

-,

HANGING REMNANT

O

r

/

~

N

-,------,-----,-1

"",/

--j---~)

"'-... -------

----

I

.m

1"

¡•

,- ,

--tL---,

I I

--+-r;---J

I

1

10

1

11

s

- ---..1'

-,.-J ,

,.

",:J'

r

.... ....

(1)

m

» o

s:

•

CIl CIl

» o .... »

-n

.... m

m

"

m

»

l)

ro r-

" "re

" to

"¡ O z

=-.J "

e

1"

o, Member 1 artetacts

•

---d' ,

l'

•

•

1,

H.

." -}

'"

'"

..,

'"

.'"

'"

~,

..,-p'

B. DEPTH PROFILE - TOTAL N=231

'"

350

'"

250

200

'"

roe

"'"

cms.

A. Plan 01 Swartkrans cave showing the Hanging Remnant 01 Member 1, the tlat areas when Members 1 and 2 accumulated, Ihe permanent grid, lhe number in each grid square and the depth ranges over which these occurred. B. Profi1e lo show the depth betow datum over which al! plotted Member 1 artetacte were found.

Fig.5

A. GRID SQUARE DISTRIBUTlON

SWARTKRANS MEMBER 1 ARTEFACT DISTRIBUTION

- - ... -

/ "

I L......

I

\

I

I I \

I I

(

,/ /'"

1

NORTH

O

,

1

~

...

.. 5/1

__

-,

1/

;¡UT ARfA

_

,

,

A

,

•

,

;);.1

1

t\

_,

I

._.

H~n

I

j

F(JlT ARfJI

.,

L .rI

I.

'"'\: I . !i

I

j

,

,

,

'''¡'

-,

T

ro

A

1: ,.. !

!

)

,a_..,

I

.ll--:lt.~

\,1...1ooot.......j.~_"

I

Al<

g

, O

'--9-'

1""1'"01.2.101 1 \1 -._

I~ "

1 -----'-------.J

I 1

---

lJiIJDTI iXrll!!il if(1

-

• •• -~i...."'1¡:-',:-,.,::..,-~''~ 101 I I 1 _,

',-------,

I

'"

~

I I,

"

,

"..

HANGING REMNANT . "¡

"

5m

~I"""II"""I

o

N

1

l'

--/

118

121

,,,

B. DEPTH PROFILE - TOTAL N=318

""

"

"

112

-----j--'1 .~". ,'i

00

•

,

"oo~ 'U~

, a

~r

550--+'

00'

"O

..

A. Plan 01 Swartkrans cave showing the Hanging Remnant 01 Member 1, the uat areas when Members 1 and 2 accumulated, the permanent grid. the number o, Member 2 artelacts in each grid square and the depth ranges over which these occurred. B. Profile to show the depth below datum cver which aü plotted Member 2 artefacts were tcund.

Fig.6

A. GRID SQUARE DISTRIBUTION

SWARTKRANS MEMBER 2 ARTEFACT DISTRIBUTION

1/ Il . . . . . ---- __

)/

\.

I\

I I I I

I

1

to

cms.

I

I V>

V>

m

o>

r-

'"

s:

m

V> V>

>

--j

O

>

-n

m

--j

Jl

>

m

z

o

--j

ce

-J

I \

)

/

NORTH

I

I

I

i I

I

MEMBER 3

I

'c

\.

°,-1..,1"'1 I I 1/ + I I I Ji< I I I 1;0

'1 1 1 I 1 1 I 1 1 1 1 1 1 I I

I

654321~123

EAST

O

I

¡-¡

l' ¡-¡

N

¡

.m

¡,

! 12

B. DEPTH PROFILE - TOTAL N= 77

83);

."

oo,

'"

'''E:

I

,

A. Plan of Swartkrans cave showing the Member 3 gully, the permanent grid, the number 01 artefacts in each grid square and the depth ranges over which these occurred. B. Profite te show oeprbs below datum at wrucn all moneo Member 3 artefacts were found.

Flg.7

A. GRID SQUARE DISTRIBUTlON

SWARTKRANS MEMBER 3 ARTEFACT DISTRIBUTlON

I

(

I

f f I

/

z

.., '"

m

Gl

>-

lE

tn m ;;:

i\;

>o --<

-n

:D

>--< m

m

'"b

180

STONE ARTEFACT ASSEMBLAGES

o

¡ 6CM

Fig.8 Artefacts from Member 1 Lower Bank, 1. Casual ccre on a sub-anqular, elongate quartzite cobble showing bipolar flaking. one large cortex ñake (Type 1) and two subslciary tlakes removed from tne distal end. opposed to severa! small scars and crushing al the proximal end. Condilion: fresh. Dlmensicns: 121 x SS x 40 mm. Catalogue no.: SKX 12098, Grid reterence: E2/N4. Oepth: 550-560 cm. 2. Casual care on a sub-anquiar quartzite cobble shcwinq oipotar tlakínq. A single tarqe cortex flake (Type 1) has been removed from the distal end. Opposed at the proximal end on the same tace are severa! smaJl üake scars and crushing. ccndmon. fresh. Dímensions: 78 x 66 x 52 mm. Catalogue no.: SKX 9905. Grid reterence: E3/N4. Depth; 610-620 cm. In both speclmens the opposed scarring at the proximal end of the cobble rs due te lndirect percussion resulting trom the blow that detached the large cortex tlake.

181

STONE ARTEFACT ASSEMBLAGES

i

O

6 CM

Fig.9 Artefacts trom Member 1 Lower Bank. 1. Core/end-chopper on a ñatttsn, angular cobble of chert. bltaciauy ñaked al the distal eno lo remcve severa! small flakes trom both faces, and two parallel flakes and crushing on the right proximal edge. Condition: fluvially abraded. Dimensions: 73 x 63 x 37 mm. Catalogue no.: SKX 13408. Grid reterence: E2/N5. Oepth: 490-500 cm. 2. Polyhedron on a weathered nodule ot quartz, multi-taceted with tndícaticns 01bilateral flaking. Condition: tresh toslightlyabraded. Dimensions: 76 x 73 x 69 mm. Catalogue no.: SKX 7111. Grid referenee: E5JN4. Oepth: 530-540 cm.

182

STONE ARTEFACT ASSEMBLAGES

2

I O

i

4CM

Fig.10 Artefacts Irom Member 1 Lower Bank. 1. Sice-ñake fragment of chert with concave, scraper-Hke retouch. Ccndition: fresh. Dimensicns: 33 x 59}( 13 mm. Catalogue no.: SKX 14822. Grid reterence: E4/N4. Oepth: 690-700 cm. 2. Flat chunk ot quartz with two steeply retouched scraper edges. Condition: fresh. Dimensions: 55 x 44 x 15 mm. Catalogue no.: SKX 21334. Grid reterence: ES/54. Oepth; 220-230 cm. 3. Modified flake fragment of chert with ñat, lnverse llaking on the lett lateral edge. Condition: tresn. Dimensione: 62 x 67 x 24. Catalogue no .. SKX 4103. Grid reterence: E4/N6. Dapth: 660-670 cm. 4. Cortex flake fragment 01 quartz with retouched/modified right lateral edge. Condilion: fresh. Dimensions: 37 x 35 x 10 mm. Catalogue no.: SKX 14427. Grid reterence: E4/N3. Depth: 590-600 cm. 5. Flake of quartzüe with point platform, distally snapped and with inverse marginal modification of the right lateral edge. Condition: fresh. Dimensions: 40 x 34 x 10 mm. Catalogue no.: SKX 16241. Grid reference E4/55. Oepth: 190-200 cm. 6. Small flake fragment of quartz with dorsal modification ot the right lateral edqe. Condition: lresh. Oimensions: 18 x 16 x 10 mm. Catalogue no.: SKX 21306. Grid reterence: ES/54. 230-240 cm. 7. Flake fragmentof quartz with distal modification on the dorsal face. Condition: fresh. Dimensione: 22 x 12 x 10 mm. Catalogue no.: SKX 13244. Grid reterence: E4/N4. Depth 590-600 cm. 8. Small chunkof quartz aüemately flaked to form a beco Condition: tresh. Dimensions: 19 x 17 x 9 mm. Catalogue no.: SKX 13283. Grid reference: E4/N4. Depth: 580-·590 cm.

STONE ARTEFACT ASSEMBLAGES

183

o

i

4CM

Fig.11 Artefacts Irom Member 1 Lower Bank. 1. End-struek. eortex (Type 1) flake 01 chert. Condition: fresh. Dimensicns: 97 x 55 x 21 mm. Catalogue no.: SKX 8692. Grid reterence: E4/N5. Depth: 680-690 cm. 2. Small, side-struck cortex ñake (Type 1) of quartz with ínverse distal modíñcatton. Candition: fresh. Dimensions: 20 x 23 x S mm. Catalogue no.: SKX 8165. Grid reterence: E5/N4. Oepth: 650-680 cm. 3. End-struck flake (Type 6) 01 chert with plain platform and unfcñrecttonat dorsal flaking. Ccndüion: tresn. Dimensicns: 38 x 24 x 9 mm. Catalogue no.: SKX 6199. Grid reterence: E2/N4. Depth: 550-560 cm. 4. End-struck Ilake (Type 6) of quartz with a dihedral platform, semi-corte of perccss.cn aeo right, in verse modification. Condition: fresh. Dlmensions: 58 x 34 x 18 mm. Catalogue nc.: SKX 5601. Grid reterence: E4fN5. Depth: 620-630 cm. 5. Hinged, end-struck cortex flake (Fype 4) ot quartzite, with a plaln striking p'atform and distar. marginal mocññcanon. Condition: slightly aoradec. Oimenslons: 31 x 29 x 10 mm. Catalogue no.: SKX 14030. Grid reference: E4/NS. Depth: 520-530 cm. 6. Hinged, side-struck flake (Type 5) ot quartzite with a point prattorm. Condition: fresh Oimenslcns: 42 x 51 x 14 mm. Catalogue no.: SKX 9535. Grid reference: E2/N2. Depth: 350-360 cm. 7-8. Angular chunks ot chert with coneave spalling suggestive ot thermal fracture. Condítion: tresn. Dlmenslons: 41 x 37 x 20 mm and 37 x 26 x 12 mm Catalogue nos: SKX 21039 and SKX 7018. Grid reterences: E5/S5 and E4/N4. Oeptns: 180-190 cm and 730-740 cm.

184

STONE ARTEFACT ASSEMBLAGES

1 O

1

6~

Fig.12 Artefacts from Member 2. 1. Sub-spheroid of quartz. CondiUon: tresh. Dimensions: 74 x 64 x 59 mm. Catalogue no.: SKX 2101. No provenance. 2. Sub-spheroid of quart: Oondition: lresh to slighUy abraded. Dlmensions: 103 x 102 x 89 mm. Catalogue no.: SKX 19621. Grid reference: W2/S2. Oepth: 170-180 crr

STONE AATEFACT ASSEMSLAGES

.'

185

,.

,':··{:':\W\V\ .:~

¡

o

Fig.13 Artefacts from Member 2. 1. Large cobble end-tlake (Type 3) 01quartzfte with inclined cortex plattorm. Condition: fresh. Dimensions: 91 x 87 x 23 mm. Catalogue no.: SKX 1650. Grid referenee: E6/N4. Depth: 160-170 cm. 2. Large cobble slde-tlake (Type 4) 01 quartzite with plam striking pJatform and sorne right lateral edge mcdfication on the dorsal face. Ccno'ñon: fresh. Dimensions: 95 x 105 x 42 mm. Catalogue no.: SKX 2367. Grid referenee: Ea/Nl. Depth: 90-100 cm. 3. Larqe, thick, end-struck flake (Type 1) of quartzite with cortex prattorrn and semi-cone 01 percussicn. Circumlerence en ventral face shows four radial flake scars suggesting modificatían lar use as a naturally hqh-backed. radially struck coreo Condüion: tresh. Dimensions: 130 x 117 x 65 mm. Catalogue no.: SKX 2159. Grid reference: E4/N1. Depth: 70-80 cm.

STONE ARTEFACT ASSEM8LAGES

186

i

o

6 CM

".

......

~ :".~

.

~.u·

\- ,

-,

I

\

\ \

\

~

-,

,<

/ /

----

Fig.14

Artefacts 110m Member 2.

1. Casual core on a small, angular quartz cobble. One cortex tlake has been removed trom the dorsal tace. Condition: fresh. Dimens! 51 x 33 x 27 mm. Catalogue no.: SKX 12363. Grid reterence: E4/SG. Depth: 10-20 cm. 2. Casual core/unitacial ene-chopper? on an an! quartz cobble. One f1ake has been removed trom the dorsal tace. Condition: fresh. Dlmensions: 77)( 61)( 49 mm. Catalogue no.: SKX 19 Grid reterence: E6/S5. Oepth: 80-90 cm. 3. Corezenc-cbooper cn a cuartzüe cobbte. bnaterauv flaked atthe distal end. Condition: slightly abra Dimenslons: 86x 73 x 57 mm. Catalogue no.: SKX 2944. Grid raferenee: E7/N4. Depth 120-130 cm. 4. Core/end-chopper. bilaterally ñake a quartzíte cobble. Condüion: tresn. Dimensions: 84 x 72 x 59 mm. Catalogue no.: SKX 2039. Grid reterence: E7/S6. Depth: 80-90 cm.

..

STONE ARTEFACT ASSEMBLAGES

187

Fig.15

Artefacts trorn Member 2. l. Core/slde-chopper on a quartz cobble. bilaterally flaked down one lateral edge. Ccndition: fresh. Dimensions: 74 x 47 x 38 mm. Catalogue no.: SKX 2038. Grid reterence: Ea/56. Deptn: 90-100 cm. 2. Core/side-chopper on a split cobble ñake (Type 1) struck trom an angular cobble 01 chert. Condltlon: very slightly abraded. Oimensions: 65 x 74 x 47 mm. Catalogue no.: SKX 2193. Grid reterence: ES/52. Oepth: surtace. 3. Hiqh-backad biconical core with radial flaking, or a circular chopper, made on a quartzite cobble. Condñion: slightly abraded. Dimensions: 72 x 60 x 57 mm. Catalogue no .. SKX 2430. Grid reference: E7/N2. Depth: 120-130 cm. 4. Hiqh-backed biconical ccre/stde- and end-chopper, made on a quartzite cobble. Condition: very slightly aoraoec. Dimensions: 67 x 59 x 60 mm. Catalogue no.: SKX 2040. Grid reterence: E7/S7. Depth: 90-100 cm.

188

STONE AATEFACT ASSEMBLAGES

o

i

4 CM

Fig.16 Artetacts from Member 2. 1. Bipolar core on a chunk 01quartz. Condition: fresh. Dimensions: 32 x 22 x 20 mm. Catalogue no.: SKX 1175. Grid reference: E4/N2. Depth: 150-160 cm. 2. Bipolar core on a quartz cobbJe. Condition: fresh. Dimenstons: 48 x 56 x 47 mm. Catalogue no. SKX 19084. Grid reterence: E6/S7. Depth: 50-60 cm. 3. Disc core on a quartzite cobbte fragment with six main radial scars on the dorsal tace. Condition: fresh. Dimensians: 67 x 62 x 32 mm. Catalogue no.: SKX 263. Grid reference: ES/N4. Depth: 280-290 cm. 4. Polyhedron 01quartz. Condüion: fresh. Dimensicns: 66 x 65 x 57 mm. Catalogue no.: SKX 12364. Grid reference: E4/S6. Oepth: aO-90 cm. 5. Polyhedron of quartz. Condition: very slightly abraded. Dimensions: 72 x 64 x 64 mm. Catalogue no.: SKX 2027. Grid reterence: ES/56. Oepth: 50-60 cm.

STONE ARTEFACT ASSEMBLAGES

189

5

o

i

4 CM

\ \

,( ,

10~ Fig.17 Artefaets trorn Member 2.

1. End-flake (Type 5) 01 quartzite with a plain platform and marginal chipping or edqe-cíamaqe on the right lateral edge 01 the ventral tace. Condition: fresh. Dlrnensfons: 44 x 32 x 9 mm. Catalogue no.: SKX 19168. Grid reference: ES/52. Oepth: 70-80 cm. 2. Obñque, end-struck cortex flake (Type 1) 01quartzite with marginal ehipping en the dorsal tace of the left lateral edge. Ccnditicn: fresh. Dimensions: 48 x 32 x 14 mm. Catalogue no.: SKX 19097. Grid reference: E5/S6. Oepth: 120-130 cm. 3. Side-ñake with a plain platlorm (Type 3) in quartz. Ccncltion: fresh. Dimensions: 22 x 32 x 10 mm. Catalogue no.: SKX 1865. No provenance. 4. Side-flake (Type 6) of quartzite. Condition: fresh. Dimensions: 34 x 49 x 14 mm. Catalogue no.: SKX 19050. Grid reference: Ea/57. Depth: 40-50 cm. 5. End-tlake with cortex ptattorm (Type 3) of quartzfte. Conoüion: fresh. Dimensions: 36 x 31 x 11 mm. Catalogue no.: SKX 260. Grid reference: E5/N4. Depth: 280-290 cm. 6. End-ñake (?Type 6) ot quartz with poinl platform. Condition: fresh. Dimensione: 31 x 25 x 9 mm. Catalogue no.: SKX 336. Grid reference: E5/N2. Depth: 160-170 cm. 7. End-Ilake (Type 6) of chert with a plain plattorrn. Condition: fresh. Dlmensions: 59 x 37 x 11 mm. Catalogue no.: SKX 1109. Grid referenee: E6/N2. Oepth: 130-140 cm. 8. Cortex sfde-ttake (Type 1) of quartzite with marginal cbippinq at the distal end and right lateral edge. Condition: very slightly abraded. Dlrnensions: 35 x 54 x 11 mm. Catalogue no.: SKX 12454. Grid reterence: E5/S6. Depth: 30-40 cm. 9. End-ñake with cortex platform (Type 2) 01 quartzite. Condition: fresh. Dimensions: 64 x 39 x 13 mm. Catalogue no.: SKX 2581. Grid reference: E7/N1. Depth: 60-70 cm. 10. End·flake with platn plattorm (Type 4) ot quartzite Ccndition: fresh. Dimensions: 64 x 38 x 12 mm. Catalogue no.: SKX 2369. Grid reterence: W2/S4. Depth: 200-210 cm.

190

STONE ARTEFACT A5SEMBLAGES

I

4CM

Flg.1a Artefacts trom Member 2. 1. End·flake with plain platform (Type 5) of quartzlte, with marginal chipping on the distal end and part of the left lateral edqe. Condition: tres Dimenslons: 46 x 37 x 13 mm. Catalogue no.: SKX 19053. Grid reterence: EBlS7. Deoth: 40-50 cm. 2. Side-ñake with plain plattorm (Type 01 quartz, with marginal chipping on the distal eoqe. Ccnditicn: tresh. Dlmensions: 27 x 36 x 8 mm. Catalogue no.: SKX 19148. ano reterenc E6JS7. Oepth: 60-70 cm. 3. Proximal flake tragment 01 ouartz on whlch a blunt borer pcnu has been worked by steep convergent dorsal flakir alo09 Ihe distal ends of two converging lateral eones. cononon. fresh. Dimensions: 26 x 23 x 9 mm. Catalogue no.: SKX 337. Grid reterenc E5fN2. Depth: 160-170 cm. 4. Thick fJake tragment of quartz with opposed flakjng on the tnickness ot the piece to form a dihedral burin; probat a technical bunn. Condition: very slightly abraded. Dímenslons: 43 x 47 x 25 mm. Catalogue no.: SKX 19193. Grid referenee: EB/S7. Dept 20-30 cm. 5. Smal! chunk 01 quartz wJth what appears to be bifacia! retouch on both faces converging to form a blunt, borer-like point. Conditic fresh. Dlmensions: 26 x 21 )(8 mm, Catalogue no.: SKX 262. Grid reterence: E6/N1. Depth: 110-120 crn. 6. Quartzitecobble ehunkwith rnodltir distal end and sorne marginal edge damage. Condition: fresh. Dimensions:49 x 39 x 28 mm. Catalogue no.: SKX 17044. Grid reference: E3//\ Deptb: 1100-1110 cm. 7. Smalt, lhick ehunk of quartz with steep convex retouched edge. Condition: fresh. Dimensions: 42 x 31 x 24 m Catalogue no.: SKX 331 Grid reterence: E5/N2, Deptn: 180-190 cm. 8. Small, thiek ehunk quartz with two steeply retouehed scraping edg in two planes al right angles. Condltion: tresh. Dimensions: 49 x 39 x 28 mm. Catalogue no.: SKX 335. Grid reterence: E5/N2. Deo 160~170 cm.

o,

STONE ARTEFACT ASSEMBLAGES

191

o

•

•

I .CM

Fig.19

Artetacts from Member 3. 1. Sub-triangular ñake (Type 6) 01 quartz with point platform and unldlrecncnal flaking on me dorsal tace. Condition: fresh. Dimensions: 18 x 16 x 3 mm. Catalogue no.: SKX 33990. Grid reterence: W4/S3. Depth: 400-416 cm. 2. Short quadrüateral snapped flake (Type 2) of quartz with cobble cortex plattcrm and uníclrecttonal flakjng from the proximal end en the dorsal face. Condition: fresh. Dimensions: 34 x 23 x 9 mm. Catalogue no.: SKX 32041. Grid reterence: W3/S3. Depth: 440-450 cm. 3. Short quadrilateral, snapped, end-struck ttake (Type 6) ot quartz with faceted striking platform and unidirectional flaking lrom the proximal end. Condition: fresh. Oimensions: 29 x 27 x 6 mm. Catalogue no.: SKX 34942. Grid reference: W4/S4. Depth: 360-370 cm. 4. Irregular, snapped tlake (Type 6) 01 quartzfte. oblique and end-struck with a plain striking platform and unidirectional flaking at right angles to the platform. Cond'tion: fresh. Dimensions: 41 x 38 x 16 mm. Catalogue no.: SKX 33859. Grid reterence: W3/S4. Depth: 440-450 cm. 5. Irregular, snapped, end-struck flake (Type 6) 01 quartz, with a plaln striking platform and bidirectionaJ flaking on the dorsal tace. Condition: fresh. Dtmensfons: 44 x 45 x 12 mm. Catalogue no.: SKX 25606. Grid reterence: W4/S4. Depth: 200--210 cm. 6. Irregular, longitudinally split flake (Type 2) from a quartztte eobble with a cortex plattorm and unidirectional ñaktnq. cooeaon: fresh. Dimensions: 64 x 41 x 14 mm. Catalogue no.: SKX 31789. Grid referenee: W3/S3. Deptn: 420-430 em. 7. Bitacially ñaked pebble core/side- and end-chopper on an angular quartzite ccbble. Condinon: moderately abraoeo. Dimensions: 80 x 71 x 42 mm. Catalogue no.: SKX 31743. Grid reterence: W3/S3. Oepth: 430-440 cm. 8. Bifacially ñakec pebble core/stce-chopper on an abraded ehunk ot chert. Condition: moderately abraded. Dlmensions: 82 x 64 x 45 mm. Catalogue no.: SKX 27636. Grid reterence: W2/S4. Depth: 310-320 em.

192

STONE ARTEFACT ASSEMBLAGES

i

O

1 • CM

Flg.20 Artetacts from Member 3. 1. Umtace wlth snapped distal end made on a larga side-struck ñake from a quartzlte cobble with a number 01 smaü flakes removed by stee retouch atong both lateral edges. The butt is pointed; possibfy a pick or c1eaver. cononcn: fresh. Dimensions: 107 x 60 x 36 mm. CatalogL no.: SKX 26168. Grid referenee: W3/S3. Depth: 180-190 cm. 2. Proto-bitace or pebble core on an angular quartzlte cobble, with partl-bñaci tlaking down the right lateral edge, distal end and distal part 01the !eft lateral edge, lorming a steep scraping edqe. Ccnditicn. tresh. Dimension 91 x 76 x 51 mm. Catalogue no.: SKX 31060. Grid reterence: W3/S3. Depth: 350-360 cm. 3. Biccnical core on a qcartzite cobble, ractanv ñake on the dorsal and ventral faces. Condition: very sHghtly abraded. Dlmensicns: 62 x 58 x 54 mm. Catalogue no.: SKX 34866. Grid referenc W4/S4. Oepth: 510-520 cm. 4. Discoid core on an abraded, angular chunk of quartzite, with radial flaking on both faces. Condition: ver¡ slighl abradeo. Dimensions: 91 x 86 x 48 mm. Catalogue no.: SKX 32784. Grid relerence: W3/S4. Depth: 280-290 cm.

STONE ARTEFACT ASSEMBLAGES

193

Fig.21 An artefact from Member 3. Spheroid on a quartzite cobbíe, characteristicaJly tlaked and battered over mast ot the surtace. Condition: slightly

abraded.

Oimensions:

103 x 101 x 87 mm. Catalogue no.: SKX 32792. Grid reterenea: W5/S4. Depth: 220-

230 cm.

I

i

o

CM

8 CM

i

• "

,é ".':. -¡-:--.

..

~ t}:-:~-~. .:-;.;."

{6':~:'

~-

..

"

.-,.,..-1

Fig.22 Artefaets from the Member 3 gully area. No provenance was recordad as they were collected superñclally, befare the construetion ot the metal grid. 1. Larga, bread f1ake ot diabasa with a radial tlake scar pattern on the dorsal tace and a taceted striking platform with three scara. Probably from a Proto-Levallois coreo Conditlon: very weathered. Dimensions: 169 x 139 x 28 mm. 2. Ouartz sphercld. manqanese-stained. Dimensicns: 70 x 70 x 69 mm.

i

o

¡ 7 CM

194

ACKNOWLEDGEMENTS

STONE ARTEFACT ASSEMBLAGES

..~

Mest grateful thanks are recorded here to Dr C. K. Brain, the excavator and Director of the Swartkrans investigation for his

are also due to Dr C. K. Brain, then Director, and to his staff at the Transvaal Museum, Pretoria, for the laboratory facilities to

kind invitation to examine and describe the artefacts from Members 1-3, tor providing the coordinates and for most helpful visits to the site itself to examine and discuss the stratigraphy and other features. My thanks and appreciation

undertake the analysis; lo my wife. Betty C. Clark, for help wilh lhe lables and lo Judilh Ogden who drew the artetacts, texl figures and distribulion plots thal accompany Ihis chapter,

REFEAENCES

BAA1N, C. K., 1981. The hunters or the huntecJ? An introduction lo cave taphonomy. University ot Chicago Presa. Chicago. BRAIN, C. K., 1985. Cultural and taphonomic comparisons ot hominids trcm Swartkrans and Sterkfontein. In: DELSON, E., ed., Ancesto-s. tre nera evidence, pp. 72-75. Alan Uss Inc, New York. BRAIN. C. K, CHURCHER, C. S., CLARK, J. D., GRINE, F. E., SHIPMAN. P.• SUSMAN. R. L. TURNER. A. and WATSON. V.. t 988. New evidence or early hominids, their culture and envirenment from the Swartkrans cave. South Atnca. South African Joumaf

o, $cienee 84: 828-835. BAAIN, C. K. and SILLEN, A., 1988. Evidence from the Swartkrans cave for the eartiest use 01 ñre. Nature 336: 464-466. BROWN. F.. HARRIS. J., LEAKEY. R. E. and WALKER. A., 1985. Early Horno erectas skeleton from west Lake Turkana, Kenya. Nature 316: 788-792. BUNN, H.• HARRIS. J. W. K.• ISAAC. G.. KAUFULU, Z.• KROLL. E.• SCHICK, K. TOTH. N. and BEHRENSMEYER, A. K.. 1980. FxJ j5D: an early Pletstocene site in northern Kenya. Worfd Archaeology 12: 109-136. CLARK, J. D. ano KLE1NDtENST, M. R., 1974. The stcne age cultural sequence: terminology, typology and r3W materials. In: CLAAK, J. D., ed., Ka/ambo Fafls prehistoric site 1/: the later prehistoric cultures, pp. 71-106. Cambridge University Press, Cambridge. CLAAKE, R. J., 1985a. Early Acheulian with Horno habilis at sterkfontein. {n:TOBIAS, P. V., ed., Hominidevolution: past, presenrand tuture, pp. 287-298. Alan Lisa, New York. CLARKE, R. J., 1985b. Australopithecus and early Horno in southem Africa. In: DELSON, E., ed., Ancestors: the hard evidence, pp. 171-177. Alan Liss, New York. CLARKE, R. J.• HOWELL. F. C. and BRAIN. C. K.. 1970. More evtdence of an advanced hominid at Swartkrans. Nature 225: 1219-1222. GOODALL, J., 1986. The chimpanzees ot Gombe: patterns of behaviour. Bellknap Press, Cambridge. HARRIS, J. W. K., 1963. Cultural beginnings: Ptio-Pletstocene archaeological occurrences trcm the Afar, Ethiopia. The Alrican Archae· ologieal Review 1: 3--31. HOWELL. F. C" HAESAERTS. P. and DE HEINZELlN, J., 1987. Depositional environments, archaeological occurrences and hominids Irom Members E and F 01the Shunguru Formation (Omo Basin,

Ethiopia). Joumal ol Human Evolution 16: 665-700. HUGHES, A. R. and TOBIAS, P. V., 1977. A tesen skull, probably of the genus Horno from Sterkfontein, Transvaal. Nature 265: 310312. ISAAC, G. U., 1977. Olorgesailie: archaeologiea/ studies 01a Middle Pfeistocene tske basin in Kenya. University of Chicago Press, Chicago. JOHANSON, D. C., MASAO. F. T" ECK. G. G.. WHITE. T. O.. WALTER, R. C.. KIMBEL. W. H.• ASFAW, B.• MANEGA, P., NDESSOKIA, P. and SUWA, G" 1987. New partial skeleton of Horno habiJis from Olduvai Gorge, Tanzania. Nature 327: 205-209. KLEINDIENST, M. R., 1962. Components of the East African Acheulian assemblage: an anatytica! approach. Jn: MORTELMANS, G. and NENQUIN, J., eoe. Actes du IVeCongres panafricam de prehistoire et de rétude du Ouaternatre. pp. 81-112. Musée Royal de I'Afrique Centra/e, Tervuren. Sér. to-e. Scienees Humaines 40. LEAKEY, M. D., 1970. srooe artetects from Swartkrans. Nature225: 1222-1225. LEAKEY, M. D., 1971. Olduvai Gorga: excevetons in Beds 1 and 2 (1960-1963). Cambridge University Presa, Cambridge. MASON, R. J., 1962. The Sterkfontein stone artetacts and their maker. South Afriean Archaeological Bulfetin 17: 109-125. TOTH, N., 1982. The stone technologies 01 early hominids at Koobi Fora, Kenya: an experimental approach. Ph.D. diesertation, University 01California, Berkeley. TOTH, N., 1985. The Oldowan reassessed: a close look at early stone artetacts. Journal of ArehaeologicaJ Science 12: 101-120. VRBA, E. S., 1975. Some evidence of chronology and palaeoecology of Sterkfontein, Swartkrans and Kromdraai from the fossil Bovidae. Nature 254: 301-304. VRBA, E. S., 1985. Early hominids in southern Africa: updated cbservatlcns on chronological ano ecological background. In: TOBIAS, P. V., ed., Hominid evolution: past, present and tutúre, pp. 195-200. Alan uss. New York. WALKER, A. and LEAKEY, R. E. F., 1978. The hominids of East Turkana. Scientific American 238: 44-56. WALKER. A.. LEAKEY, A. E. F., HARRIS, J. M. and BROWN. F. H.• 1986.2.5 m.yr. Australopithecus boiseifrom west 01Lake Turkana, Kenya. Nature 322: 517-522.

Chapter 8

The Swartkrans Bone Tools C. K. Brain' & Pat Shiprnan"

p o. Box413. Pretoria. 0001 Soutn Atrice 220 11 Wilson Road. White Hall, Maryland 21161, USA

1Transvaal Museum,

Recent excavations at Swartkrans have produced 68 fossil borres that appear to have been used as digging tools, although sorne show subseouent modification through presumed use in tbe preparation of animal skins

by earty hominids.

INTRODUCTION

In the course of the excavations al Swartkrans between 1979 and 1986, a total 0168 fossil bones were found that appear to haya been usad as tools. 01 these, 17 came from the Lower Bank of Member 1, 11 from Member.2 and 40 from Member 3. As mentioned in Chapter 1 ofthis volume, the Middle Stone Age unít designated Member 4 has not yet been excavated, while the excavated portions of Member 5 have not yielded any bane artefacts. The typical feature of the Swartkrans bone tools is that they taper to smoothly worn polnts, with sorne ot them showing a superimposed poJísh. The majority 01 the speclrnens consist ot bone flakes, which have been defined (Brain, 1981) as coming fmm the shafts of long-oones, such as the temur, radius or metapodial, but lacking complete articular ends ano not preserving more than half the circumference of the longbone shaft; a smaller number are parts of ante lo pe horncores, while other skeletat parts are only marginally represented. When pondering about the possible use to which these bones had been pul, it occurred lo one ot us (C.K.S.) that their worn appearance was very reminiscent of that of our excavation tools, In the softer parts of the Swartkrans deposit, as ter instanee the Lower Bank ot Member 1, we used long screwdrivers to loasen the partially calclfied sediment from quarter-rnetre grid squares, in 10 cm-deep splts. The loosened sedirnent was then lcaded into buckets and carried to the sieve. When new, the screwdrivers had normal square ends, but after sorne days of use, these were modified into smoothly tapering points, with characteristic scratching, as shown in Fig.l. It seems Jikely that the onJy raason that early hominids would have had to dig in the ground, would be the procurement of tood. Indeed, in the course of making a living on the open grassland habitat that appears to have existed around Swartkrans during the rimes of accumulation ot Members 1,2 and 3, jt is very probable that hominids would have been obliged to aU9ment their vegetable toad with edible bulbs and other underground storage organs. An undisturbed dolomite environment, with sorne of its indigenous fauna, still exists

about 15 km northeast of Swartkrans in the John Nash Natura Reserve on (he farm Uitkomst. Observations made by one of us (C.K.B.) over many years on this reserve have shown that chacma babocns (Papío hamadryas ursinus) dig edible bulbs from dolomitic soil during the dry winter months when other vegetable food ís less abundant than in summer. Two plants in particular were involved in these observations, SciJIa (or Ledebouria) marginata and Hypoxis costata, the bulbs of which were dug out by baboons using anly their hands, in situations where the ground was sufficiently soft ter them to do this. Normally, such digging took place in the ailuvium of valley bottoms and, although the plants in question are more abundant on the rocky dolomitic hillsides, baboons are unable to excavate them there. Lilies 01 the genus Scitle, or Ledebourie as lt is usually now called, belong to the family Liliaceae and have decorative undulating Ieaves blotchcd with purple, up to 25 cm in length, and short inflmescences bearing purple and green flowers. The underground slorage organ is a moisture-Iaden bulb, up lo 20 cm in length and weighlng up lo 250 9 (Fig. 2). Sorne specíes of Scitte are said to be poisonous fa humans (Lucas, 1987), but they do not seem to cause harm to the Uitkomst baboons during the highveld winters. 'Grass stars' ot the genus Hypoxis, belonging to the family Hypoxidaceae, have golden ñowers, 2 cm or more in diameter, borne in inflorescences that are covered with silver or golden hairs. The leaves of H. costata are bread and sheath-like, fringed with long. rough hairs, and up to 20 cm long. A second species, H. rigidula, common on the Swartkrans hill, has narrow fibrous leaves. up to SO cm in length, which used to be plaited into rope by indigenous people (Letty, 1962). The storage organs of Hypoxis plants are tunicated corms (Fig. 3), weighing up to 5009, extremely fibrous and difficult te chew. Uitkomst baboons have been observed to eat them (C.K.B. personal observation), so it is possible that they were also used by early hominids. It is c1ear that, if hominids had access to sorne sort of digging tools. they would have been able to use the toad resouree of Seif/a, Hypoxis and other underground storage organs, unobtainable by baboons on lhe rocky dolomite hillsides. Their

196

SONE TOOlS

Fig.1

Fig.2

The elfecl of digging on the appearance 01 metal tools - in lhis case. screwdrivers used as digging tools in lhe course 01 the Swartkrans excavation. The tool on lhe lelt is, as yel, unused. while lhe other two show characterislic wear and scratching reminiscent of that seen on the worn bones from Swartkrans described in this chapter.

A plant and bulb of the Iily Scil/a margina/a, dug 'rom the Swartkrans. hillside with an experimental bone tool shown in lhe photograph and also in Fig. 6.

advanlage over baboons in this regard, particularly during lhe dry winter months, m ust have been significant. The majority of bone tools found at Swartkrans, as mentioned above, are made on bone f1akes and il may seem strange that these thin and sometimes delicate pieces should have been selected as digging tools, but the reason becomes obvious when one personally digs an edible bulb from a dolomite hillside, such as that at Swartkrans. The bulbs 01 Scilla and Hypoxis are typically 10-20 cm below the surface, but are wedged between and beneath a mosaic of angular chert blocks, which have to be removed one by one if the bulb is to be extracted whole. In arder to remove a chert block, it is fjrst necessary to scratch out the soil between it and adjacent blocks, after which it can usually be lifted out (Fig. 4). Experience has shown that the intervening soil can be best scratched out with a narrow. pointed instrument, such as is provided by abone Ilake. In less rocky situations, a more robust tool such as an antelope horn is very effectlve. Bone f1akes can be generated from long-bone shafts either by the impact 01 a hammerstone on an anvil, or by the feeding 01 large carnivores, such as hyaenas. For the purpose 01 digging experiments, some bone lIakes were collected at a blue wildebeest kili in the Kruger National Park that had been

worked over by spotted hyaenas. These flakes, some of which are shown in Fig. 5, were between 8 and 12 cm long and were fresh at the time of collection. Two flakes were selecled for a series 01 digging experiments at Swartkrans: Tool A was 105 mm long, with a maxim um width 0125 mm, while Tool B was 82 mm long and 18 mm wide. Tool A had been used lar a total of 8 hours digging atthe time the photographs (Fig. 6a,b) were taken, while Tool B (Fig. 7a,b) had seen 4 hours' service. It was lound that noliceable smoothing 01 the sharp edges 01 the flakes was apparent after one hour of digging use, while rounding of lhe digging tip was well-developed afterfour hours 01 use. This rounding did not appear to be much accentuated in the subsequenl four hours of digging to which Tool A was subjected, suggesting that the greatest visible modification of the original sharp edges had taken place wilhin the first four hours 01 use. The length 01 time taken to dig a Scilla or Hypoxis bulb Irom the Swartkrans hillside varied from 14 lo 30 minutes (n = 12), according to the sloniness 01 the ground. The two experimental bone lools, togetherwith the collection 01 fossil equivalents lrom Members 1, 2 and 3 al Swartkrans, were taken lo the Oeparlment 01 Cell Biology and Anatomy of lhe Johns Hopkins University Medical School in Baltimore, during April 1987. Here selected parts 01 many of the speci-

BONETOOLS

Fig.3 A plant and tunicated corm 01 Hypoxis costala, dug from the Swartkrans hillside with an experimental bone tool shown in the pholograph and also In Fig. 6.

Fig.4 Experimental use 01 abone tool: Conrad Brain digs a Hypoxis plant out 01 the rocky Swartkrans hillside.

mens were replicated, using silicone-based dental impression material to make negative impressions, in conjunction wilh epoxy resin used lo make posilive casts, according to the method described by Rose (1983). Scanning electron microscope photographs of the characteristic patterns 01 wear and

scratching on experimental tools A and B are shown in Figs 6b and 7b, while those of selected Swartkrans lossil bone tools are provided with the descriptions of each specimen. The microscopic patterns of wear on the possible bone tools from Swartkrans were compared with those resulting from use

SONETOOLS

198

a "---~........~;~ .-_ .. .

-

• 1

Fig.5 Sane Ilakes, broken (rom the limb bones 01 a blue wildebeest by spotted hyaenas. Naturally pointed flakes ot this kind appear to have been selected by early hominíds and used as digging tools in the vicinity of the Swar1krans cave.

on the experimental bone tools, to testthe hypothesis that such use had caused the modilications lhal originally drew attention to the lossil specimens. Some specimens showed modifications that appeared to be functionally identical to those on the experimental digging tools; these are described and figured below. The use-wear on these specimens fulfilled expectations based on previous studies (Shipman, 1989; Shipman and Rose, 1988}, in that the wear was confined to specific regions, immediately adjacenl lo the working edge or surface. Unaltered areas of bone could be observed a few millimetres away; lhese showed no traces of wear or modification. Other specimens were more ambiguous, showing modifications that could not be confidently distinguished from those produced by trampling. abrasion, wealhering, carnívore dígestion or other natural taphonomic processes. These were eliminated from the sample 01 bone tools, since it could not be demonstrated with confidence lhat the observed modífications were caused through use by hominids.

DESCRJPTlONS OF SWARTKRANS BONE TOOLS MEMBER 1 LOWER BANK Horncore pieces used as tools SKX 4228b, Fig. 8b. Grid square. E5/N5, SW quarter; depth: 560-570 cm. Date fDund. 9 July 1980.

Fig.6 An experimental bone tool (A), based on a flake trom a wildebeest limb bone. a: the tool showing wear resulting from eight hours 01 digging in the ground at Swar1krans. b: SEM image 01 the surtace of the bone tool showing detail al scratching on the smoothly worn surtace. Scale bar = 1 mm.

DescriptiDn. A smaJl tapering piece of bovid horneo re, 43 x 15 mm, with smooth wear along two sides. Interpretation. It is presumed to have been part 01 a much larger digging tool. SKX 7068, Fig. 8a. Grid square. E5/N4, SW quarter; depth 470-480 cm. Date fDund. 19 March 1980. Descrip/iDn. Parts 01 a fragmented bovid horncore with a smoothly worn tipo A distinct facet on one side of the tip suggests that the horn had an incomplete sheath on it when used. The horncore was at least 10 cm long before it was broken. Interpreta/ion. The smooth wear and scratching suggests that the horn had been used for digging and that the sheath-tip 'Nore through to expose the end of the horncore. SKX 5011> Fig. 9a,b. Grid square. E3/N10; depth uncertain as the Member 1

BONE TOOLS

199

a SKX 7068

b ~

SKX 4228B

Fig.8 Bone lools Irom Swartkrans Member 1 Lower Bank. These are pieees Irom the tips of antelope horneores showing wear consisten! with digging. a: SKX 7068, b: SKX 4228b.

SKX 5011

Fig.7 An experimental bone tool (B), based on a flake from a wildebeest limb bone. a: the tool showing wear resulting from lour hours 01 digging in the ground at Swartkrans. b: SEM image 01 the surlaee of the bone tool showing detail of seratehing on lhe smoothly worn surfaee. Seale bar = f mm.

sediment here had been disturbed by mining activities. Date found. 2S March 1981. Descríptíon. An almost complete horneore of a blesbuek, cf. Damalíscus dorcas, 145 x 30 mm, the distal 3S mm of which has been smoothly worn and polished. Interpretation. SEM examination (Fig. 9b) eonfirms the visual impression that all sides of this horncore's tip have been evenly worn, and that fine longitudinal scratches with oecasional transverse ónes, are visible under a smooth and even poJish. It is inferred that the horneore was used as a digging tool, presumably without íts sheath, and then used for rubbing soft material, perhaps animal skin. Sone flakes apparently used as digging tools

Fig.9

SKX 794, Fig. 10a Grid square. E2/N3; depth: 350-400 em. Date found. 10 May 1979. Description. The broken-oH end of abone flake, 28 x 20 mm,

SKX 5011: specimen from Lower 8ank 01 Member 1. a: a horneore 01 a blesbuek, el. Damaliscus dorcas, wilh a smoolhly worn and polished tipo This is interpreted as a digging 1001 thal has also been used for the rubbíng of sofl materíal, sueh as animal skin. b: SEM ¡mage 01 the worn surlaee showing polish superimposed on a smoolh surtaee, with linear seratehing. Sea le bar = 100 11m.

80NETOOLS

200

~~.\. b

a SKX 794

SKX 1429

a

'<:;;~~ . e

SKX 5005

SKX 3477/8

~~~." ~ d

SKX 5003

SKX 5001

Fig.10 Sane tools made on f1akes from (he Lower 8ank of Member 1: a: SKX 794; b: SKX 1429; e: SKX 3477/8; d: SKX 5003; e: SKX 500t.

smoothly worn and shows some scratching, although detail is partly obscured by adherent matrix. Interpretation. Apparently the end 01 a digging tool that was originally much longer. SKX 1429, Fig. 10b. Grid square. E4/N3; depth: SOO-SSO cm. Date {ound. 17 May 1979. Description. The tip of a symmetrical worn point 18 x 8 mm, made on abone flake. Longitudinal and transverse scratching is visible. In terpretation. This appears to be the tip 01 a bone-flake digging tool which was originally much longer. Fig.11 SKX 3477/3478, Fig. 10c. Grid square. E4/N4, NE quarter; depth: Si 0-520 cm. Date {ound. 28 February 1980. Descriptíon. Two parts 01 abone f1ake without a clear join, 41 x 16 and 18 x 27 mm respectively, tapering to a smooth and well-worn tipo Interpretation The smooth wear and scratching suggest that this bone flake, much longer when complete, was used as a digging tool.

SKX 5005: abone tool, assumed to have been used lor digging, made on abone f1ake trom the Lower Bank of Member 1. a: the whole specimen; b: SEM delail 01 the worn tip showing cross slriations superimposed on linear scralching. Scale bar = 1 mm.

SKX 5003, Fig. 10d. Grid square. E4/NS, NW quarter; depth 660-670 cm. Date {ound. 25 March 1981. Description. Abone llake in lour pieces, without clear joins, which had been at least 170 mm long when whole. The tip has been worn to a smooth conical point, showing line longitudinal and transverse scratching. /nterpretatíon. An unquestionable digging tool made on a long-bone Ilake. Prolonged use, in excess of lour hours, is indicated.

SKX 500S, Fig. 11 a,b. Grid square. E3/NS, SW quarter; depth 610-620 cm. Date {ound. 11 February 1981. Description. A ralher tlat bone f1ake, 100 x 20 mm, tapering to a symmetrical conical point. SEM detail 01 thís point is shown in Fig. 11 b, with prominent transverse scratches superimposed on smooth wear c1early apparent. /nterpretation. An exceptionally good example 01 a digging tool. The considerable wear suggests several days 01 use, implying that lhe tool was carried around lar some time by its hominid owner.

SKX 5001, Fig. 10e. Grid square. E3/N4, NW quarter; depth: 610-620 cm. Date {ound. 18 February 1981. Description. Abone Ilake, 51 x 18 mm with a rounded and

smoothly worn tipo Fine longitudinal and transverse scratching is visible. Interpretation. A digging tool that was probabJy longer when used.

SKX 5006, Fig. 12a,b. Gríd square. E3/NS, SE quarter; depth: 640-6S0 cm. Date {ound. 18 March 1981.

SONE TOOLS

a

201

SKX 5006

. SKX 500a t.

Fig.12

Fig.13

SKX 5006: the tip 01 abone 1001, assumed lo have been used lor digging, made qn abone f1ake trom the Lower Sank of Member 1. a: the specimen: b: SEM detail ollheworntip showing coarse scratching. Scale bar = 1 mm.

SKX 5008: pieces 01 abone 1001, assumed lo have been used tor digging, made on abone flake from Ihe Lower Sank 01 Member 1. a: the specimen; b: SEM detail 01 the worn lip showing prominent crossslriations. Scale bar = 100 .um.

Description. The tip of abone flake, 35 x 13 mm, worn lo a smooth, conical point. As shown in the SEM photograph (Fig. 12b), prominenl longitudinal and transverse scratching is superimposed on the smoothly worn suriace. Interpretation. This specimen appears lo be the típ of a bonef1ake digging tool thal was oríginally much longer.

sedimenl disturbed by lime-mining operations. Date found. 19 May 1981. Description. Abone flake, 112 x 22 mm, wilh smooth wear, scratching and polish on one end, as detailed in the SEM photograph (Fíg. 14b). Interpretation. Regarded as a digging tool.

SKX 5008, Fig. 13a,b. Grid square. E3fN5, SE quarter; deplh: 580-590 cm. Date found. 5 February 1981. Description. Abone flake in three pieces, wilh imperiect joins, that would have been at least 150 mm long. The terminal 50 mm lapers lo a smooth and symmetrical point, showing longitudinal and cross-scratches, some of which are visible in the SE M image in Fig. 13b. Interpretalion. Appears to be a typical example of a bone-fJake digging 1001.

SKX 501 Ob, Fig. 15a,b. Grid square. E4/N9; deplh about 1000 cm, in Member 1 sediment disturbed by mining operations. Dale found. 19 May 1981. Description. Part of abone flake, 85 x 15 mm, with a worn tip that is incomplete as a result of a longitudinal split. As shown in the SEM photograph (Fig. 15b), longitudinal and transverse scralchíng is superimposed on the smooth wear. Interpretation. This is regarded as part of a larger bone-f1ake digging tool.

SKX 501 Da, Fig. 14a.b. Grid square. E4fN9; depth about 1000 cm, in Member 1

SKX 7062. Grid square. E5fN4, SW quarter; deplh: 460-470 cm.

BONE TOOLS

202

a

a

SKX S010A

SKX 50108

Fig.14

Fig.15

SKX 5010a: abone tool, assumed to have been used tor digging, made on abone llake from the Lower Bank of Member 1. a: the specimen; b: SEM detail 01 the worn lip showing fine lin ear scratching. Scale bar =1 mm.

SKX S010b: abone tool, assumed to have been used tor digging, made on abone f1ake Irom the Lower Bank 01 Member 1. a: the ¡¡pecimen; b: SEM detail 01 the worn tip showing prominent cross· scratching. Scale bar = 180 ).lm.

Date found. 19 March 1980. Description. Part of abone flake, 81 x 15 mm, tapering to a smoothly worn point. Longitudinal and transverse scratches are visible. Interpretation. Thís appears to have been a digging tool on a bone flake that was originally longer and wider. Sone flakes showing a high polish s uperímposed on wear and scratch es SKX 5000, Fig. 16a,b. Grid square. E3/N5. SW quarter; depth 620-630 cm. Date found. 11 February 1981. Description. Three pieces of abone flake. without good joins, indicating that the original piece of bone was alleast 130 mm long. The largesl piece, 62 x 25 mm, tapers to a very smooth and symmetrical point, showing scratching and pitting, over which is superimposed a surprisingly high polish (Fig. 16b). Interpretation. It is suggested that this bone f1ake, which c(early tape red to a rough natural point, was used as a digging tool for a considerable period, and was then rubbed on some soft

substance, such as animal skin. SKX 5009, Fig. 17a,b. Grid square. E3/N5. NE quarter; depth: 750-760 cm. Date found. 10 June 1981. Description. A piece of the tip 01 abone flake. split longitudi- . nally, 38 x 8 mm. The complete tip clearly tape red to a point, and what remains shows smooth wear with longitudinal and transverse scratches. on which a fairly high polish has been superimposed, as shown in the SEM photograph (Fig. 17b). Interpretation. The specimen seems lo ha ve been part of a larger bone Ilake that had been used firsl for digging and then for rubbing some smooth substance, such as animal skin. SKX 5012. Fig. 18a,b. Grid square. E4/N9; depth: 1230 cm. Dale found. 3 June 1981. Description. A Ihick bone flake, 113 x 25 mm, tapering to a chisel-like end, which shows smooth rounding and wear. Unidirectíonal scratching is visible beneath a well-developed

BONE TOOLS

a

203

a

~:

SKX 5009

SKX 5000

Fig.16

Fig.17

SKX 5000: abone tool made on a f1ake Irom the Lower Bank of Member 1. It is assumed that this tool was originally used over a long period lar digging, but has subsequently been rubbed on soft material such as animal skin. a: the specimen; b: SEM detail 01 the worn point showing polish superimposed on the scratched surface. Scale bar = 100 !-1m.

SKX 5009: part 01 abone tool on a f1ake Irom Ihe Lower Bank of Member 1. When complete, it is assumed that this was a digging loo! tha! has subsequently acquired a polish, probably as a resull 01 rubbing on animal skins. a: the specimen; b: SEM detail showing polish superimposed on the scratched surface. Scale bar = 100 !-1m.

polish, as shown in the SEM photograph (Fig. 18b). Interpretatíon. It is possible that the bone f1ake was first used for digging, but Ihereafter íl appears lo have been used as a rubbing tool, perhaps for use on animal skin.

mueh longer, was used as a digging tool and that, in the course of use, the tip was broken off. Further use Ihen rounded the broken surface. It appears that the eore was used without the original horn sheath in place.

MEMBER 2

SKX 15536, Fig. 19b. Gríd square. E6/S1 ; depth: 150-200 cm. Date found. 6 June 1979. Descríption. Two pieces 01 a broken horncore. 120 x 22 mm, the tip 01 whieh lorms a smoothly tapering conical point. The natural point 01 the horneore has been blunted and worn with use, with wear visible up to 45 mm from the tipo Interpretation. This was clearly part of a much longer horneore that appears to have been used as a digging tool. The smooth wear extending backwards Irom the tip suggests Ihat the eore was used without the eovering sheath in place.

Horneare pieces used as tools SKX 12383, Fig. 19a. Gríd square. E4/S6, SE quarter; depth 80-90 cm. Date found. 26 May 1983. Descríption. The distal end 01 a bovid horneore, 50 mm long and aboul 24 mm in diameter. The tip 01 the horneore had clearly broken off obliquely and the Iractured surface has since been worn smooth with use. AII surfaees 01 the pieee show smooth wear, with some scratching visible on the roughly textured bone. In te rpretation. It is surmised that this horneore, when very

SKX 17211, Fig. 20a,b. Gríd square. E4/S3. NE quarter; depth: 160-170 em, at the

SONE TOOLS

204

a b

SKX 15536

SKX 5012 '\

•

a

-=~-

.

SKX 12383 Fig.19 The tips of two antelope horneares Irom Member 2 that appear lo hay been used as díggíng too/s: a: SKX 12383; b: SKX 15536.

a

Fig.18 SKX 5012: abone tool made on abone Ilake Irom the Lower Sank 01 Member 1, assumed to have been Llsed both lor digging and rubbíng 01 anímal skins. a: the specimen showíng íts blunt, chisel-Iike end; b: SEM detaiJ of Ihe worn end showing fine linear scratchíng on whích polísh ís superimposed. Scale bar = 1 mm.

interface belween Member 2 and the underlying Lower Bank 01 Member 1. Date found. 22 June 1983. Description. The complete right horncore of an alcelaphine antelope 01 the genus Beatragus, 285 mm long and 55 mm in diameter at the base, with the tip worn off at right angles to the axís of the eore. Smooth wear exlends baekwards from the worn tip for 15 mm on two lalerally opposed surfaees. The surface of the rest of the horneare is no! worn or polished. Interpretation. 11 appears that this horneare, with its sheath in place, was used as a digging tool and thal the sheath was worn baek, exposing (he truneated coreo Furthermore, the shealh seems to have worn through on two opposing surfaees for a dístance of about 15 mm from the end, exposing the core in these areas to smooth wear. Sane flakes apparently used as digging tools SKX 105, Fig. 21 a,b. Grid square. E5/S2; depth: 180-190 cm.

b SKX 17211

Fig.20 SKX 17211: a complete Beatragus horneore from Member 2 th appears to have been used as a digging tool. a: the whole speeíme b: detail of the worn típ.

Date found. 26 July 1979. Descriptíon. Tip of abone flake, 36 x 13 mm, tapering to smoothly worn and almost symmetrical poin!. Longitudinal ar transverse scratches are visible, as in the SEM photogra¡ (Fig.21b). Interpreta/ion. The worn end 01 a once longer bone-flal digging tool.

BONETOOLS

a

S'KX 105

·-a

205

SKX 1142

rl==:=:i__

Fig.21

Fig.22

SKX 105: the tip 01 a bone-fJake digging tool from Member 2. a: the specimen showing the well-worn point; b: SEM detail showing wear and scratches. Scale bar", 1 mm.

SKX 1142: part 01 a bone-rlake digging tool from Member 2. a: the specimen showing its worn tip; b: SEM detail showing longitudinal scratching on the worn point. Scale bar = 1 mm.

SKX 1142, Fig. 22a,b. Grid square. ES/Ni; depth: 175-200 cm. Date found. 13 June 1979. Description. The broken end ot a fairly tlat bone flake, 60 x 20 mm, with a symmetrically worn point. Prominent longitudinal scratching is visible, as shown in the SEM photograph (Fig. 22b). Interpretation. The poínt 01 a once longer bone f1ake that had been used as a digging tool.

Date found. 2 March 1983. Descriptíon. Two pieces 01 a thick bone flake, without a clear join, that would have been at least 130 mm long. The tip piece, SKX 10158, 61 x 20 mm, tapers to an almost symmetrical buliet-shaped point, with smooth, even wear and some scratching. Interpretation. Almost certainly a digging tool that has seen prolonged use.

SKX 3227, Fig. 23a,b. Grid square. E4/N4, SE quarter; depth: 4S0-470' cm. Date found. 26 February 1980. Description. Piece of abone tlake, 42 x 15 mm, with a smoothly worn tapering end. Coarse longitudinal and transverse scratching is visible, as shown in the SEM photograph (Fig. 23b). In te rpretation. The worn end 01 a once longer bone-1lake digging tool. SKX 10158/10159, Fig. 24a Grid square. E1/S7, SE quarter; depth: 100-110 cm.

SKX 16976, Fig. 24b. Grid square. E4/N9; depth: 1200-1230 cm. Date found. 3 June 1981. Description. Piece of a transversely broken bone Ilake, 67 x 21 mm, that has also been split longitudinally. The shaft shows rodent gnawing, while the chisel·like tip has been smoothly worn and shows longitudinal and transverse scratches. Interpretation. Appears to be part 01 a bone-tlake digging tool that was subsequently gnawed by rodents, presumably after it had been discarded in the cave.

SONE TOOLS

206

a

SKX 3227 a SKX10158/9

b SKX 16976 Fig.24. Parts of two bone-Ilake digging tools Irom Member 2. a: SKX 10158/9; b: SKX 16976.

a

SKX 352

Fig.23 SKX 3227: the end of a bone-lIake digging 1001 from Member 2. a: the speeimen showing its worn tapering tip; b: SEM detail 01 longitudinal and transverse seralehes. Seale bar = 1 mm.

Sone flakes showing a hígh polish superimposed on wear and serateh es SKX 352, Fig 25a,b. Grid square. E5/N2; depth: 250-275 cm. Date found. 21 June 1979. Description. Part of the tip of abone flake, 41 x 19 mm, lhat has been broken lransversely and longitudinally. The tip shows smooth, rounded wear with scralching, superimposed on which is a high polish, as shown in the SEM image in Fig. 25b. lnterpretation. Initially a bone-flake digging tool, which has subsequently acquired a polish through being ·rubbed on soft material su eh as animal skin. SKX 1141/1143, Fig. 26a,b. Gridsquare. E6/N1; depth: 175-200 cm. Date {ound. 13 June 1979. Description. Two pieees of abone flake that would have been at Jeast 100 mm long. The tip piece, SKX 1143,52 x 7 mm, tapers to a smoothly worn point, showing scratching on whíeh is superimposed a high polish, as shown in the SEM image in Fig.26b.

- - - - - - - - - - - _ . - - _.... -

Fig.25 SKX 352: part 01 the lip 01 abone f1ake from Member 2 that had apparently been used lor digging and rubbing. a: the speeimen; b: SEM delaíl showing polish superimposed on wear and seralehes. Seale bar 100 lim.

=

BONE TOOLS

a

207

-·,a

SKX1141/3

SKX 3287

_1It::::::::_.~

Fig.26

Fig.27

SKX 1141/3: The tip 01 a bone-flake digging tool from Member 2 that appears to have served both lordigging and rubbing. a: the specimen; b: SEM detail o/ the polished suriace. Scale bar = 50 ).1m.

SKX 3287: A bone-Ilake 1001 Irom Member 2 that appears to have served both lor digging and rubbing. a: the specimen; b: SEM detail of Ihe polished lip. Scale bar = 50 ).1m.

Interpretation. A bone-f1ake digging too! that has subsequently been used for rubbing, perhaps on animal skin.

Date found. 15 March 1984. Description. A complete left horneore of an unidentified antelope, including the frontal bone, 155 mm long. Smooth wear and rounding is visible on the distal 15 mm of the eore, but not lower down. In terpretation. It appears that the horneo re, at that time still proteeted by its sheath, was used as a digging tool and that the tip of the sheath had worn away to expose the core, which suffered typieal abrasion.

SKX 3287, Fig. 27a,b. Grid square. E4/N4, NE quarter; depth: 490-500 cm. Dale found. 26 February 1980. Description. A complete bone flake, 130 x 25 mm, which tapered naturally to a chisel-like end. This shows even wear and rounding, as well as prominent scratching, on which is superimposed a high polish, as shown in the SEM image in Fig.27b. Inlerpretation. A bone-flake diggíng tool which has subsequently been used for rubbing soft materíal, such as animal skin.

MEMBER 3 Horneore pieees used as tools

S KX 21790, Fig. 28a. Grid square. W3/S 1, SW quarter; depth: 190-200 cm.

SKX 288768, Fig. 28b. Grid square. W3/S2, NE quarter; depth: 280-290 cm. Date found. 14 February 1985. Description. Four pieces of a somewhat curved horncore, probably from an alcelaphíne antelope, that had an original length 01 at least 140 mm. The típ has been blunted by use and shows smooth wear and seratehing. Interpretation. The specimen appears to have been used lor digging, but owing to the fragmentary nature 01 íts distal end, it is uncertain whether it was with or without lhe sheath.

SONE TOOlS

208

SKX 30246/9, Fig. 28c. Grid square. W3/S3, SE quarter; depth: 320-330 cm. Date found. 7 March 1985. Description. An almostcomplete right horneo re, 120 mm long, probably of a springbuck, Antídorcas sp., with smooth wear and scratching round [he tip, which has been blunted with use. In terpretation. The horncore appears to have been used as a digging tool, while a certain amount of wear and polish on the shaft of the core suggests that it was used without ils shealh. SKX 23567, Fig. 29a. Grid square. W2/S4, NW quarter; depth: 200-210 cm. Date found. 1 August 1984. Description. A small piece 01 horneore tip, 32 x 15 mm, with smooth wear and seratching. lnterpretation. Appears to be the tip of a horneare, which when more complete, had been used as a diggíng tool. SKX 26234, Fig. 29b. Grid square. W2/S5, NE quarter; depth: 270-280 cm. Date found. 11 January 1985. Description. The distal end, 45 x 12 mm, of a small, straight horneare trom a steenbuek-like antelope, with smoolh wear and scratching round the tipo The specimen has also, almost eertainly, been burnt. In terpretation. It appea rs that the horncore was used as a digging tool befare being discarded clase to a camp-fire in the cave, where it was subjeeted to a temperature of 300400 oC.

SKX 28437, Fig. 29c. Grid square. W3/S3, NE quart~r; depth: 280-290 cm. Date found. 14 February 1985. Description. The distal end 01 a large horneore, 51 x 25 mm, which has been considerably modilied by wear to a rounded point, showing polish and scratching. Interpretation. The indication is that, when this horneare was complete, it saw prolonged use as a digging tool, apparently with its shealh on, which

SKX 21790

a

SKX 288768

e

SKX 30246/9

Fig.28

Three horneare pieees from Member 3 showing wear to their lips, apparently the result of digg use. a: SKX 21790; b: SKX 28876B: e: SKX 30246/9.

a

b SKX 23567

SKX 26234

..-"

d

e SKX 28437

~~

e

SKX 30215

SKX 34570

f SKX 36485

9 SKX 36861 Fig.29

Seven horneore pieees from Member 3 showing wear lo their tips assumed to have resulted Ir( digging use; two speeimens designated b and I also appear to have been burnt. a: SKX 23567; SKX 26234; e: SKX 28437; d: SKX 30214/5; e: SKX 34570: f: SKX 36485; g: SKX 36861.

SONE TOOLS

209

lad worn away on one side more (han he other. 3KX 30214/5, Fig. 29d. Grid square. W3/S3, NE quarter; jepth: 300-310 cm. Date found. 27 February 1985. Description. Two pieces from the distal end of a small horncore, but without a clear join. The lerminal piece, 20 x 11 mm, has been worn to a bJunl point, showing polísh and scralching. Interpretation. It appears that the horncore: when more complete, was used as a digging tool. SKX 34570, Fig. 2ge. Grid square. W4/S2, SE quarter; depth: 370-380 cm. Date found. 24 July 1985. Description. A small piece, 27 x 12 mm, of whal appears to be horneore, worn lo a smooth and symmetrical tapering point. Polish and .scratchin9 is visible. Interpretation. Thís is clearly Ihe tip 01 a much longer and well-used digging tool.

..

:

FIg.30 a: SKX 29388 + 22747-22750 + 29171: the left mandible 01 a three-Ioed horse. Hipparion Iybicum sfeytieri, from Member 3. The ascending ramus has been broken off ¡nto a point (arrow) and appears to have been extensively used as a digging 1001. b: SKX 26324/8. A shaft piece from an anlelope limb bone from Member 3, the poinl 01 which appears to have been used as a digging 1001.

SKX 36485, Fig. 29f. Grid square. W5/S3; depth: 550600 cm. Date found. 24 May 1985. Description. A small pieee, 30 x 9.mm, of horncore that has been worn lo a smooth, bullet-like tipo Polish and seratehes are visible and Ihe pie ce looks as if it has been burnt. Interpretation. It appears Ihat Ihis horneore, when much longer, was used for digging and was subsequently heated in a fire. SKX 36861, Fig. 29g. Grid square. W5/S2; depth: 600-650 cm. Date found. 27 September 1985. Description. A small piece, 25 x 14 mm, Ihat has also been split longitudinally, from a horncore tipo It has been worn lo a smooth, rounded end, with scrateh marks. interpretation. Appears to be part 01 the tip 01 a horneore Ihat had been used extensively lor digging.

used as the point of a digging tool, with the toothrow having been grasped in the hand. The amount 01 wear on Ihe rounded point suggests that the 1001 saw extensive use, probably over several days. Long-bone shaft piece apparently used as a digging tool SKX 26324/8, Fig. 30b. Grid square. W2/S5, SE quarter; depth: 300-310 cm. Date found. 11 January 1985. Description. A piece of antelope Iimb-bone shaft, 95 x 30 mm, one end of which had been fraetured obliquely and then worn smooth. The worn surfaces show coarse scratehing. Interpretation. It seems that Ihe naturally pointed end 01 this shaft pieee had been used as a digging 1001. The other end was very probably longer at the time of use. Sone Ilakes apparently used as dígging tools

Horse mandíble that had been used as a tool SKX 29388 + 22747-22750 + 29171, Fig. 30a. Grid square. W3/S2, NW & SE quarters, depth: 220-230 cm. Date found. 20 February 1985. Description. Left mandibular ramus 01 a mature three-toed horse, Hipparion Iybicum steyUeri, with an almost complete looth-row. The ascending ramus had been broken off into a natural poinl and this has been smoothly rounded by wear, with transverse scratehing visible. Interpretation. It appears that the broken ascending ram us was

SKX 10859, Fig. 31a. Grid square. W6/S4; depth: 0-100 cm, found in a hole in the southwest wall 01 the cave. Date found. 17 November 1982. Description. The end 01 a large bone flake, 45 x 20 mm, that has also been split longitudinally. The tip shows considerable wear, resulting in smooth rounding with longitudinal and transverse scratehing. Interpretation. This appears to be part of Ihe tip of a mueh longer digging tool that had seen fairly extensive use.

BONETOOLS

210

SKX 10978, Fig. 31 b. Grid square. W6/S4, depth: 0-100 cm. Date found. 7 November 1982. Description. The end 01 abone Ilake, 63 x 14, that has been worn ¡nto a tapering, bullet-like point. The smooth surface shows scratching and a certain amount 01 potish. Interpretation. Almost certainly a digging tool that has seen prolonged use, extending over several days. The bone Ilake is likely to have been longer at the tíme 01 use, whíle the polish may indicate that i1 was al so used lor rubbing skins.

'~~ ~.~..~ ; .. ='f .(.~ ..

_. -

..

a

b SKX 10859

SKX 10978

e SKX 20046

.¡a..,

.~~~" ~~~ d

SKX 20081

e SKX 21617

f

SKX 22885

SKX 20046, Fig. 31 c. Grid square. W2/S5, NE quarter; SKX 22933A SKX 25678 depth: 200-210 cm. SKX 26112 Date found. 17 November 1983. Description. Abone Ilake, 87 x 29 mm, Fig.31 with a rounded, chisel-like tip, showA series 01 bone flake pieces Irom Member 3. assumed to have been used as digging tools. , ing smooth wear, line linear and occaSKX 10859; b: SKX 10978; c: SKX 20046; d: SKX 20081; e: SKX 21617; f: SKX 22885; g: SK sional transverse scratching. Polish is 22933a; h: SKX 25678; j: SKX 26112. likely to have been present but this ís obscured by a superfícial deposit 01 manganese dioxide. Interpretation. Almost certainly a digSKX 22933a, Fig. 31g. ging tool that may have been longer when used. Grid square. W3/S2, SE quarter; depth: 200-210 cm. Date found. 21 March 1984. SKX 20081, Fig. 31d. Grid square. W2S5, SE quarter; depth: 210-220 cm. Description. Part 01 lhe tip 01 abone Ilake, 25 ;0('13 mm, tha Date found. 16 November 1983. has been broken transversely and Jongitudinally. The chisel like tip shows smooth wear and abundant scratching. Description. The end 01 abone Ilake, 62 x 19 mm, tapering to a symmetrical and rounded tipo The smoothly worn surlace In terpretation. This is almost certainly part olthe tip 01 a digginr tool, originally much larger, lhat has seen prolonged use. shows linear and transverse scratching. /nterpretation. Almost certaínly a digging tool, showing ev;SKX 25678, Fig. 31h. dence 01 prolonged use. The bone Ilake may have been longer Grid square. W2/S5, SW quarter; depth: 240-250 cm. originally. Date found. 11 January 1985. SKX 21617, Fig. 31e. Description. Parl 01 the tip 01 abone flake, 31 x 11 mm, tha Grid square. W2S5, NW quarter; depth: 220-230 cm. has been broken transversely and longitudinally. The rounde, Date found. 2 February 1984. chisel-like lip shows smooth wear and scratching, benealh al encrustation 01 manganese dioxide. Description. Part 01 a thick bone Ilake, 52 x 19 mm, broken Interpretation. This is almosl cerlainly part of a digging tool tha transversely and longitudinally. Whal remains 01 the tip is had been much larger when in use. chisef-like, showing smooth wear on three surfaces, with so me polish. SKX 26112, Fig. 31i. Interpretation. The piece appears to have been part 01 a larger Grid square. W2/S3, SW quarter; depth: 170-180 cm. bone Ilake that had been used exteñsively lor digging or Date found. 16 March 1984. rubbing skins. Description. Part 01 a somewhat shattered bone flake SKX 22885, Fig. 311. 93 x 32 mm, tapering to a smoothly worn bullet-like poín° Grid square. W3/S1, SW quarter; depth: 210-220 cm. Scratching is clearly visible under an encrustalion 01 manga Date found. 21 March 1984. nese dioxide. Description. The end 01 a spirally-Iractured bone flake, Interpretation. This is almosl cerlainly a digging tool thal ha 37 x 17 mm, tapering to a smoothly worn point, showing seen protracted use, extending over many days. scratching. SKX 26113, Fig. 32a. Interpretation. This is almost certainly the workíng end 01 a digging too! made on abone flake which, in use, would have Grid square. W2/S3, SW quarter; depth: 170-180 cm. beenlonger. Dale found. 16 MaTch 1984.

9

h

SONE TOOLS

Description. Part of abone flake, 49 x 17 mm, with an oblique chisellike point that has been smoothed on all sides by wear and also shows scratching and polish. Interpretation. This is the end 01 a bone Ilake that appears to have been used as a digging and/or a rubbing

211

b

SKX 26113

SKX 26138

SKX 26139

SKX 26149

1001.

SKX 26138, Fig. 32b. Grid square. W3/S2, NE quarter; depth: 210-220 cm. Date found. 21 March 1984. Description. The end of a fairly delicate bone f1ake, 47 x 11 mm, which tapers to a finger-like point that has been worn smooth on all sides. The smooth surface shows linear and transverse scratching. In terpretation. The wear and scratching suggests strongly thal this piece, though small, had been used for digging over an exlended periodo

9 SKX 26624

•

SKX 28076

SKX 28487

h SKX 28828

; SKX 30052

Fig.32 A seríes 01 bone flake pie ces from Member 3. assumed to have been used as digging lools. a: SKX 26113; b: SKX 26138; e: SKX 26139; d: SKX 26149; e: SKX 26624; f: SKX 28076: g: SKX 28487; h: SKX 28828: i: SKX 30052.

SKX 26139, Fig. 32c. Grid square. W3/S2, NE quarter; depth: 210-220 cm. Date found. 21 March 1984. Description. The end of abone flake, 28 x 9 mm, lhat has been broken lransversely and longitudinally. What remains 01 lhe lip shows lhal it lapered lo a smoolh, butlel-Iike poi nI, showing even wear and scratching. Interpretation. This is almost certainly part 01 a well-used digging tool, which originally consisted 01 a much longer bone f1ake. SKX 26149, Fig. 32d. Grid square. W3/S2, NE quarler; deplh: 200-210 cm. Date {ound. 21 March 1984. Description. The end of a delicale bone f1ake, flat on one side but rounded on the olher, which tapers to a smooth point. It shows even wear with fine linear, and occasional transverse, scralching. Interpretation. The smoolh wear and scralching suggests strongly lhat this was part of a longer digging tool, which must have been used over several days. SKX 26624, Fig. 32e. Grid square. W2/S5, NE quarter; deplh: 280-290 cm. Date {ound. 11 January 1985. Description. The end 01 abone flake, 51 x 27 mm, hemispherical in cross-seclion and tapering to a point in the form 01 a ha/f-cone. The point shows smoolh wear and fine scratching. Interpretation. This is the end of a once longer bone f1ake that had, almost certainly, been used as a digging tool over an extended periodo SKX 28076, Fig. 321. Grid square. W3/S4, NE quarter; deplh: 230-240 cm.

Dale found. 23 January 1985. Description. A pencil-like pie ce of bone flake, 51 x 9 mm, tapering lo a rounded, sym metrical point. Interpretation. The smoolh wear and scratching suggests tha! thís rather delicate tool was used for digging. . SKX 28487, Fig. 32g Grid square. W3/S3, NE quarter; depth: 290-300 cm. Date found. 14 February 1985. Description. The end of abone flake, 35 x 11 mm, tapering 10 a symmetrical and rounded poinl, with smooth even wear and fine scralching. Interpretation. When complete and much longer, this bone flake appears to have been subjected to prolonged dígging. SKX 28828, Fig. 32h. Grid square. W3/S3, NE quarter; depth: 270-280 cm. Date {ound. 14 February 1985. Description. A small, flal bone flake, 43 x 14 mm, with a worn scoop-like end. The worn area shows scralching and polish. Interpretation. The bone flake, when more complete, was almosl certainly used lor digging.

SKX 30052, Fig. 321. Grid square. W3/S3, NW quarler; depth: 320-330 cm. Date {ound. 7 March 1985. Descrip/ion. The tip 01 abone flake, 25 x 19 mm, with a wellworn symmetricai end, showing scratching and polish. Interpreta/ion. When much longer, this bone Ilake was almost certainly used as a digging tool.

BONETOOLS

212

SKX 30141, Fig. 33a. Grid square. W3/S3, NW quarter; depth: 300-310 cm. Date found. 7 March 1985. Description. The end of abone flake, 40 x 10 mm, tha! has been broken transversely and longitudinally. 1I shows smoolh wear and scratching. Interpretation. When more complete, this bone f1ake almost certainly saw prolonged use as a digging tool.

1."'_"'II8IHP!.•

b

a SKX 30141

SKX 32582

1.. . .,lI!J~

e

SKX 32897

~~ ~~'':'"'''''' d

SK)( 33654

SKX 35196

e

~ii

SKX 32582, Fig. 33b. Grid square. W4/S2, NE quarter; deplh: 230-240 cm. Date found. 15 May 1985. Description. A thin sliver of bone, 25 x 5 mm, which has split longitudinally from the tip 01 abone flake, and which tape red lo a smoothly worn point. What remains of the worn surlace shows fine scratching. Interpretation. When complete, this bone flake had almost certainly been used lor digging.

SKX 37703

l

SKX 45806

SKX 38041

SKX 38830 Fig.33 A series 01 bone flake pieces from Member 3 assumed lo have been used as digging tools. a: SKX 30141; b: SKX 32582; e: SKX 32897; d: SKX 33654: e: SKX35196; f: SKX 37703; g: SKX 38041; h: SKX 38830; i: SKX 45806.

SKX 32897, Fig. 33c. Grid square. W4/S2, SE quarter; depth: 270-280 cm. Date found. 21 May 1985. Description. The end 01 abone f1ake. 49 x 9 mm, which could be a piece 01 rib split longitudínally, and whích has been worn to a symmetrical point. The smoothly worn surface shows fine scratching. Interpreta/ion. Although delicate, this specimen appears to have been extensívely used as a digging tool. SKX 33654, Fig. 33d. Grid square. W3/S2, SW quarter; deplh; 420-430 cm. Date found. 29 May 1985. Description. A thick and rather flat bone flake, 71 x 14 mm, tapering to a smoothly worn and bullet-like point. Despite a coating 01 manganese dioxide, longitudinal and transverse scratching is visible. Interpretation. This is almost certainly a digging tool that has seen prolonged use, extending over many days.

Description. A long-bone flake, 131 x 22 mm, one end of which has been worn lo a symmetrical and rounded point that shows linear scratching and a cer!ain amount 01 polish ..' Interpretalion. Almos! certainly a digging 1001 that has seen prolonged use over a number of days. SKX 38041 , Fig. 33g. Grid square. W5/S4; depth: 660-700 cm. Date found. 11 October 1985. Description. The end 01 a slightly weathered bone flake, 44 x 16 mm, that has been worn to a smoothly rounded point, showing scratching. In/erpretalion. When more complete, this bone piece appears lo have been extensively used for digging.

SKX 35196, Fig. 33e. Grid square. W4/S2; deplh: 470-550 cm. Date found. 8 August 1985. Description. A thin, flat bone flake, 57 x 15 mm, one end of which is worn lo a smoothly rounded spatulate tip showing linear and occasional transverse scratching. Interpretation. Almost certainly a digging tool that has been extensively used.

SKX 38830, Fig. 33h. Grid square. W4/S1; depth: 70 cm, on north wall 01 the Member 3 gulley. Date found. 23 October 1985. Description. Abone flake, 86 x 19 mm, one end of which is pointed as a result 01 a diagonal break. Thís point has been worn smoolh and shows scratching. Interpreta/ion. This naturally pointed bone Ilake has almost cer!ainly been used lor digging.

SKX 37703, Fíg. 331. Grid square. W3/S3; depth: 550-600 cm. Date found. 28 September 1985.

SKX 45806, Fig. 33i. Grid square. W3/S2, SW quarter; depth: 260-270 cm. Date found. 27 February 1985.

213

SONE TOOLS

a

SE 612

-_±==::::jp

l

Fig.34

Fig.35

SKX 37052: a delicateawl·like tooi made on abone f1akefrom Member 3. a: the specimen; b: SEM detail 01 the worn tip showing scratching and polish. Scale bar = 100 11m.

SE 612: abone tool from Sterkfontein Member 5 descríbed by Robinson (1959) as a 1001 for lhe working 01 animal skins on accounl 01 its smooth wear, linear scratching and polish. a: the whole specimen; b: SEM detaíl 01 the polished surface on the tace! indicated wilh the arrow. Scale bar = 100 11m.

Descriplían. A smali piece. 24 x 10 mm, from the tip of abone flake that has had its natural point rounded by wear. Scratching is visible despite an encrustation 01 manganese dioxide. Interpretatian. When much longer. this bone flake was almost certainly used for digging.

DISCUSSION

A delicate bone flake fashioned ¡nto an awl-like tool SKX 37052, Fig. 34a,b. Grid square. S3f\N3; depth: 610-620 cm. Date found. 26 September 1985. Descriptian. A delicate bone flake, 58 x 6 mm, tapering at one end to a sharp point that has been consíderably worn. lis smooth surface shows fine linear scratching and a fairly high polish, as shown in the SEM ímage in Fig. 34b. Interpretatían. This delicate bone flake is unlikely to have been used as a digging tool. Its fine linear scratching and polish suggests rather lhat it was used for piercing softer material, such as animal skin.

Although the number of bone tools found at Swartkrans is smali in relation to lhe tolal number of fossil bone pie ces from the relevant Members, there appears to be a fairly even scatter 01 these tools throughout the excavated areas and depths. In the Lower Bank of Member 1, 17 bone tools have been described out of a total of 52496 bone pieces from animals larger than the 'microlauna' assumed to have been collected largely by owls. Thís gives a figure of 0,032 % for the incidence 01 bone tools in Member 1 Lower Bank. It is interesting to note that only two bone tools have so far been found in the sediment of the Hanging Remnanl of Member 1, the probable reason lor this being discussed in Chapter 13. In Member 2, 11 bone tools ha ve be en described out of a total 01 34 312 bone pieces, which gives the same percenlage incidence as for Member 1 Lower Bank, namely 0,032. The situation in M ember 3 is interestingly dirferent, how· ever, wilh 40 bone tools having be en found among a total of 63 125 bone pieces, giving a percentage incidence of

SONE TOOLS

214

Fig.36 A reconstruction by Imogen Serry showing a group of Swartkrans hominids digging edible bulbs from (he ground. One ot them has a carrying bag made on an animal skin for the transport of tools and gathered foods. The use 01 such conlainers is suggested by the bone tools found in Swartkrans Members 1,2 and 3.

0,063, or approximately double that for the earlier two Members. It is of ¡nteresl to compare Ihe abundance of bone artefacls in the three Swartkrans Members with that 01 stone artefacts:

Stratigraphic unit Member 1 (Lower Bank) Member 2 Member 3

Number of bone tools

Number 01 stone artefacts

17

402 403 72

11 40

A curious facl emerges from thesé figures - that Member 3 contained far fewer stone artefacts Ihat the earlier two unils, but appreciably more bone tools.lt should be remembered that Member 3 is the deposít containing evidence far early firemaking, in the form of burnt bones distributed throughout the vertical profíle (see Chapter 10, this volume). The Member 3 bone tools occur in close association with burnl bones and Iwo of the tools show evidence of having been in a camp-fire (SKX 26234, 26485, both burnt horneo re pieces). Within the excavated area of Ihe Member 3 gulley, burnt bone pieces have been found in 17 grid squares, while ten ot these have also

provided bone tools, suggesting that the fire-lending hominids were also (he users 01 the bone tools. As is apparent from the descriptions and interpretatjons of individual bone lools earlíer in Ihis chapler, al! bu! one 01 !he 68 tools appear to have been used tor digging in the ground, although some have been modified by subsequenl rubbing. 01 the 67 presumed digging tools, 49 show evidence of prolonged use, almost certainly in excess of one day, and probably exlending over many days, if not weeks. The question arises as lo how the Swartkrans early hominids managed to keep bone, and possibly stone artefacts, in their possession for days or weeks at a time, wilhout losing them in the course 01 their daily food-seeking rounds. An obvious answer would be that these hominids had access to carrying bags in which their tools, and possibly their gathered loods, were transported. As detailed in the descriptions 01 the individual bone tools, three specimens from each of the three Members show a polish superimposed on the wear and scratching which, originally, appears to have been caused by digging. The suggesled interpretation is that the digging tools had been used also tor rubbing a soft substance, presumably animal skin. Of interest in this regard is the description 01 abone 1001 by Robinson (1959) from his excavation 01 Sterkfontein Member 5. The tool, SE 612. is shown in Fig. 35a, while SEM detail 01 the polished surlace is given in Fig. 35b. Made on abone tlake with a natural

SONE TOOLS

point, the tool has a well-defined worn tacet, showing fine linear scratching and a high polish. Robinson's interpretation was that the bane had been repeatedly rubbed on a sott substance, presumably animal hide, a conclusion with which we cancur, although the onginal use of the tool, resulting in general wear on the tip, may have been digging. Ethnographic bone tools usad in hide-burnishing, or working of other sort substances. show a fine polish (Shipman and Rose, 1988; Olsen, 1984), similar to that sean on the Sterkfontein and Swartkrans polished bane tools, Another bane tool, trorn Swartkrans Member 3, is suggestive 01 use on animal hides. It ls the delicate awl-like artefact SKX 37052, shown in Fig. 34a,b, which ccnslets of a thin flake ot bone tapering to a worn point showing longitudinal and circumferential scratching, tagether with polish. This tool may well have been used far piercing hales in skins or other sott materials, as similar microscopic wear has been documentad on experimentally made and used awls (Olsen, 1984). The evidence discussed here suggests that the Swartkrans hornlnids of Member 1-3 times may well have mace simple carrying bags from animal skins, in which 1hey transported their tools, as weü as possibly their gathered food. This could explain the evidence for the apparent use of the same 100ls over successive days or weeks. The evroeoce also emphesizes the importance of food-procurement by digging during

215

the daíly Jife of these hominids on the open grassland envirenments of Southern África, an activity portrayed in the recenstruction by lmogen Berry shown in Fig. 36. A study has been made by one of us (P.S.), uslng the same techniques as those described in thls chapter, of bone tools lrom Olduvai Gorge (Shipman, 1989). Points 01 interes! are that rninlrnalty shaped bone toote also occur at Qlduvai between two and one miUion years ago. These comprise a small percentage of the total bone assembtage, ranging from 0,2 to 2,0 % at various sites. If the assemblages ot 32 373 bones, excluding unidentifiable fragments, from Beds I and 11 at Olduval are combined, then the 41 confirmed bone tools make up 0,13 % oflhe total. 1Ithe Sed 11 assemblage only is consídered, the bane tools comprise 0,5 % of the total, while the íncrease in the incidence 01 bone tools in Sed l! mirrors the increase seen at Swartkrans as one pass es from Member 2 to Member 3. Sorne of the Olduvai bone tools, íike their counterparts at Swartkrans, show wear suggestive 01 use on soft animal sklns , while others appear to haya been anvils, which might have been usec in the working 01 sktns. and which occur with etone awls. However, the type 01 bane from which the Olduvai tools are made is different, with a strang bias towards very large anlmals, while the apparent use to which the tools were put was also different, with very few having been used for digging.

REFERENCES BRAIN, C. K. 1981 The hunters or the hunted? An introduction to African cave taphonomy. Chicago University Press. Chicago LETTY, e., 1962. W¡fd ffowers ot the Transvaal. Trustees, Wild Flowers of tne Transvaaí Book Fund, Pretoria. LUCAS, A., 1987. Wild flowers 01 the Witwatersrand. C. Struik Pub~

lishers, Cape Town. QLSEN) S. L.• 1984. AnaJytjcal approaches fo the manufacture 01bone artefacts in prehistory. Ph.D. thesis, Univers¡tyof Londan. ROBINSON, J. T.. 1959. Abone implemenl from Sterkfonlein. Nature

184: 583-585.

ROSE, J. J., 1983. A replicaticn technique ter scanning electron mcrcscopy: applications for anthropologists. American Joumet 01 PhysicaJ Anthropology 62: 255-261, SHIPMAN, P., 1989. Altered bones fram Olduval Gorge, Tanzania: techniques, problems and implications of their recognition. In: BONNICHSEN, R. and SORG, M. H., eds, Bone modification, pp. 317-334. Centre for Ihe study of the first Americans, Orono, Maine. SHIPMAN, P. and ROSE, J. J., 1988. Sone tools: an experimental approach. In: OLSEN, S. L., ed., Scanning electron microscopy in archaeofogy. S.A.R. lnternational Series 442: 303-335. Oxford.

Chapter 9

The Incidence of Damage Marks on Swartkrans Fossil Sones from the 1979-1986 Excavations Rosemary Newman Transvaal Museum, P. O. Box 413. Pretoria, 0001 South Africa"

A sample 01147 311 tossll bone peces trom Members 1,2,3 and 5 at Swartkrans was examinad tor the presence of naturally oceurring damage marks. Such marks were classitied as toothscratches, punctures, chewed edges, rodent·gnawed banas, insect borings and digested banas. A listing of the specimens showing damage attributable to the above categoríes is províoed. Cut marks were also noted on sorne specsnens but these are not dealt with here. Epoxy replicas ot selected specímens were made and examined with a scanning electron microscope.

INTRODUCTION The usefulness 01damage marks on borres in the taphonomic reconstructian of fossil assemblages has been appreciated far a long lime (e.g., Shipman, 1981; Bunn, 1981; Potts and Shipman, 1981). In view ot this, it was decided ta examine a large sample of fossil bone pieces from the most recent Swartkrans excavations tor the presence of carmvore feedingmarks and other recognizable traces. It proveo to be easy to separate recent damage marks, such as those resuJting from excavation or fossil preparation, from pre-fossilization rnarks. Three categories of camlvore-darnaqed bone were recagnized: taoth scratehes, punctures and chewed edges; two of rodent gnawing; ane ot inseet boring and one of digested bone pieces. The sample used consisted of 147311 fossll bane pieces from faur Swartkrans Members: Member 1, Lower Bank: 52 496 specírnens: Member 2: 34 312 specimens; Member 3: 40 239 specimens and Member 5: 20264 specimens. Each specimen was individuañy examined far the presence ot damage marks attributable to one af the categories mentioned aboye and 1038 pieces were selected far further study. These specimens were examined in greater detail under a stereo microscope and, in many cases, replicas 01 the damaged surfaces were made tar study with a scanning electrón mícroscope. This process followed a well-established procedure described by Rose (1983), but adapted to locally available materials. The pracedure may be briefly described as follaws (Newman and Brain, 1989): each fossíl surfaee was first thoroughly cleaned with acetone, blown dry and then covered with a thick layer oí vinyl polysiloxane Express ímpresslon material supplied by 3M. When set, the impression material was carefully removed from the fossil and Plastomax Araldite epoxy M resín, treated with hardener HY956, at a ratio of 5:1, poured into it. Each replica was air-dried far three days, kept in a vacuum desiccator tor a similar pertod. rnounteo on a

"Present aooress: 38 King Street, Irene, 1675 $outh Africa

stub, gald-coated, and examined in a Hltachi S510 scanning electron mícroscope. Sorne examples or SEM images ot damage marks are provided in this chapter. DAMAGE MARKS Further informatian of the damage categories and the fossil specimens referred to them is now given. TOOTH SCRATCHES (Figs 1, 2) Shipman (1981: 365) provided the following definilion: 'A tooth scratch is produced by drawing a pointed tooth cusp across the surface ot a bone as the animal closes its mouth. This type of mark is probably mast often made by carnivores' canines. Tooth scratches are elongate grooves that may vary from V-shaped to u-shaped in cross-section, depending on the morphalogy of the taoth cusp. The bottom or nadir of the groove is smooth. Tooth scratches may occur on bones singly, as sets ot parallel and sub-parallel marks, or as ctusters ot marks differing widely in orientation.' Marks conforming to this description were observed on 311 specimens, as detailed befow: Member 1, Lower Bank A total of 66 specimens, rnade up as follaws: Bavid stze ciess I/(as defined in Braín, 1981). Rightastragalus, SKX 9809; left proximal metatarsal, SKX 15417. Miscellaneous. Rib fragments, SKX 8726, 9265, 11977, 47012; unfused proximal humeral head, SKX 21315; femoral epiphysis, SKX 28357. Bone flakes (as defined in Brain, 1981). Fifty two pieces ofthe following lengths: 1-2 cm,1; 2-3 cm, 9; 3-4 cm, 12; 4-5 cm, 9; 5-6 cm, 9; 6-7 cm, 5; 7-8 cm, 3; 8-9 cm, 3; 11-12 cm, 1. Indeterminate tragments. 6. Member 2 A total of 47 specimens, made up as rcnows: Bovid size ctess l. Mandibular ccranotd process, SKX 46924;

DAMAGE MARKS ON FOSSIL SONES

216

SKX 40214 SKX 47002

] I

'1 ¡

]

~

SKX 45882

~O SKX 45885

-O SKX 30379

SKX 29691

-~_._-~._---'

Fi9·1 Examples of Swartkrans fossils showing looth scratches. Member 1, Lower Sank: SKX 40214, SKX 47002; Member 2: SKX 45682, 45685; Member 3: SKX 30379, SKX 29691.

metapodial shaft, SKX 3809. Bovid size class fl. Left proximal metacarpals, SKX 696, 45852; right femur shaft, SKX 1153; right humeral piece, SKX 2264. Miscelfaneous. Femoral head, SKX 15309; epiphysis fragment, SKX 15541. Bone ffakes. Thirty three pieces of the following lengths: 1-2 cm, 1; 2-3 cm, 6; 3-4 cm, 10; 4-5 cm, 6; 5-6 cm, 4; 6-7 cm, 2; 7-8 cm, 2; 8-9 cm, 1; 10-11 cm, 1. Indeterminate fragments. 5. Member 3 A total of 103 specimens. made up as follows: Damaliscus sp. Left horneare base, SKX 36309. Canis mesome/as. Right proxim~.l ulna, SKX 30498. Procavia antiqua. Left distal humerus, SKX 36086. Cercopithecoid indet. Left humeral fragment, SKX 19495; ríght distal humeral fragment, SKX 37838. 80vid size cfass 11. Right and left pelvic pieces, SKX 27008, 35245; phalanx piece, SKX 29955; calcaneus fragment, SKX 31362. 80vid size c1ass 111. Calcaneus fragment, SKX 31606; right distal hume rus. SKX 32346; right distal radius, S KX 34804; left proximal radius, SKX 37135; proximal metatarsal, SKX 37586. 80 vid size class IV Immature distal metapodial, SKX 28671; left calcaneus, SKX 29880.

Aves, d. franco/in. Tarsometatarsal spur, SKX 37691. Miscelfaneous. Skull fragmenls, SKX 31600, 19684; horneare fragment, SKX 22042; rib fragments, SKX 21989, 28096; femoral pieces, SKX 30525, 33568, 33381, 26895; proximal humeral pieces, SKX 26562, 33451. 80ne ffakes. Fifty pieces of the following lengths: 2-3 cm, 7; 3-4 cm, 11; 4-5 cm, 15; 5-6 cm, 6; 6-7 cm, 5; 7-8 cm, 4; 8-9 cm, 1; 9-10 cm. 1. Indeterminate fragments. 25. Member 5 A lotal of 96 specimens, made up as follows: Equus burchef/ii. Left proximal ulna fragment, SKX 41194. Procavia capensis. Pelvic fragment, SKX 41591/5; right proximal femur, SKX 42038. Procavia transvaalensis. Left proximal ulna, SKX 43466. Bovid size class l. Left proximal metacarpal, SKX 41856; femoral shaft, SKX 47262. Bovid size class fl. Scapula piece, SKX 42128; left distal radius, SKX 40599; !eft proxim'al radius, SKX 41292; radial shaft piece, SKX 44188; distal humeral pieces, SKX 41286, 44428,44893; humeral shaft pieces, SKX 41769, 4161 O; distal femoral pieces, SKX 42337, 42687; femoral shaft pieces, SKX 42369, 42597; left distal tibia, immature. SKX 42647; calcaneus pieces, SKX 41526, 42412, 44385, 47237; astragalus piece, SKX 44448; proximal metatarsal pieces, SKX 41939,

OAMAGE MARKS ON FOSSIL SONES

219

F'ig.2

Scanning electron microscope images of looth scratches on Swartkrans fossil bones. a: SKX 21864 (rom Member 3; b; SKX 37691 from Member 3; e: SKX 40593 ¡rom Member 3; d: SKX 42897 from Member 5.

42644; dístal metalarsal piece, SKX 44797; proximal metacarpal pieces, SKX 42642, 42646, 44898; distal metacarpal piece, SKX 41943; metapodial shaft pieces, SKX 44821, 45240,47231; phalanx pieces, SKX 42398,45354,47799. Bavid size c/ass 111. Left dislal humerus, SKX 42336; pelvic fragment, SKX 44471. Miscellaneaus. Scapula piece, SKX 43594; rib fragmenls, SKX 42991,43797; femoral pieces, SKX 47343,47353. Sane lIakes. Thirty nine pieces of the following lengths: 12 cm, 3; 2-3 cm, 3; 3-4 cm, 10; 4-5 cm, 9; 5-6 cm, 6; 6-7 cm, 5; 7-8 cm, 1; 8-9 cm, 1; 10-11 cm, 1. /ndeterminate /ragments. 12. PUNCTURES (Figs 3, 4) Shipman (1981: 366) provided the following definition: 'Punctures are produced by the concenlration of a biling force through a single tooth cusp, oflen a canine, at an angle roughly perpendicular to the bone surface. The result is a

depressed fracture with rounded, roughly circular outline. In some cases, microscopic fragments of bone can be seen pushed inwards into the surtace. Punctures may have a stepped appearance because the area of depression is greatest al the bone surface and decreases as the distance from the surface increases.' Marks conforming to this description were observed on 74 specimens, as detailed below. Member 1, Lower 8ank A tolal of 12 specimens, made up as follows: Procavia antiqua. Righl proximal femoral piece, SKX 5803. Savid size class ffI. Cervical vertebral piece, SKX 8842. Miscellaneous. Vertebral fragment, SKX 5830; left astragalus, SKX 6703, and rib fragment, SKX 8091. Sane {fakes. Five pieces of lhe following lengths: 2-3 cm: 1; 3-4 cm, 1; 4-5 cm, 3. Indeterminate fragments. 2.

DAMAGE MARKS ON FOSSIL SONES

220

1 \l II

5KX 2262

J

SKX 35563

SKX 39984 - - - - - , SKX 42477 Fig.3 Examples of Swartkrans fossils showing puncture marks caused by carnivore teeth. Member 1, Lower Bank: SKX 47121, SKX 8091; Member 2: SKX 2262; Member 3: SKX 35563, SKX 39984; Member 5: SKX 42477.

Member 2 A lolal of five specimens, made up as follows: Bovid size class 11. Juvenile distal libia piece, SKX 2262. Bovid size class 111. Right distal radius, SKX 2042. Miscellaneous. Vertebral fragment, SKX 1128; humeral Iragments, SKX 3413, 3732. Member 3 A total of 20 specimens, made up as follows: Bovid size class 11. Right scapular piece, SKX 39840; juvenile distal femur shaft, SKX 37385; right proximal ulna, SKX 39984. Bovid size class 11/. Scapular fragment, SKX 19587; carpal, SKX 31910. Bovid size class IV, Radial shaft, SKX 28742. Suid indet. Femoral fragment, SKX 38618. Miscellaneous. Vertebral pieces, SKX 25433, 25880, 33509, 35261,35563,37848,39983; femoral pieces, SKX 28910, 33159. Indeterminate fragments. 4. Member 5 A total of 37 specimens, made up as follows: Antidorcas bondi. Left mandibular fragment, SKX 42542. Procavia capensis. Left distal tibia, SKX 43812. Bovid size class l. Humeral fragment, SKX 42837; proximal radial fragment, SKX 43354.

Bovid size class 11. Atlas vertebra, SKX 41456; thoracic vertebra, SKX 42564; lumbar vertebra, SKX 42269; pelvic pieces, SKX 40603, 43902, 44916; scapular pieces, SKX 41790, 42222,42351,42477,42956,42974,42975,47835,44437; proximal humerus, SKX 44081; distal radial pieces, SKX 41546,41551. Bovid size class 111. Proximal ulna piece, SKX 41194. Miscellaneous. Cranial pieces, SKX 42861, 43393; vertebral pieces, SKX 40858, 41471, 4164217, 42928, 43579; pelvic pieces, SKX 40692, 43213, 41346; rib fragments, SKX 42473, 41949b; proximal humeral piece, SKX 45384; femoral fragment, SKX 41636. CHEWED EDGES (Fig. 5) The diagnostic feature of this category 01 damage is that the specimen's fea tu res have been altered by the removal 01 bone, inwards from an edge or surface. An irregular or ragged edge . normally resu/ts, similar to that seen on bones chewed by carnivores. Marks conlorming to this description were observed on 267 specimens, as delailed below: Member 1, Lower Bank A total 01 53 specimens, made up as follows: Bovid size class l. Scapular piece, SKX 11976; melacarpal SKX 6877

DAMAGE MARKS ON FOSSIL SONES

221

Fig.4

Seanning eleelron mieroseope ¡mages of punelure marks on Swartkrans fossil bones. a: SKX 46995 from Member 1 Lower Sank; b: SKX 47121a from Member 1 Lower Sank; e: SKX 31910 trom Member 3; d: SKX 47835 from Member 5.

Bovid size class 11. Proximal humeral fragment, SKX 8464; distal humeral pieces, SKX 6276, 8055, 11939a; humeral shaft, SKX 17230; pelvic pieee, SKX 8720; distal metapodials, S KX 9016, 21084; phalanx, 9359. Bovid size class 111. Dislal humeri, SKX 4162,6397; proximal metapodial, SKX 47040; juvenile distal metapodial, SKX 14803. Aves indet. Distal tibio-tarsal piece, SKX 8485a. Misce/laneous. Cranial fragments, SKX 7949, 8457; rib fragmenls, SKX 5679b, 6456,6779; scapular piece, SKX 4312; pelvic piece, SKX 17248; proximal humeral pieces, SKX 4852, 8012; proximal femoral pie ce, SKX 9048. Bone f1akes. Twenty one pieces of the following lengths: 2-3 cm, 2; 3-4 cm. 6; 4-5 cm, 5; 6-7 cm, 4; 7-8 cm, 2. Indeterminate fragments. 6.

Member 2 A total of 21 specímens, made up as folJows: Bovid size class 111. Left proximal humerus, SKX 2043; tibial fragment, S KX 3435; calcaneus, S KX 15933. Miscellaneous. Skull condyle, SKX 3159; vertebral fragment, SKX 46923; humeral epiphysis, SKX 991. Bone flakes. Thirteen pieces 01 the folJowing lengths: 2-3 cm, 3; 3-4 cm, 4; 4-5 cm, 3; 5-6 cm, 2; 9-10 cm, 1. In de termina te fragments. 2. Member 3 A total of 74 specimens, made up as lollows: Equus capensis. Proximal phalanx, SKX 28313. Procavia antiqua. Left ulna, SKX 38048. Cercopilhecoid indet. Immature left mandible piece, SKX 27988. Lagomorph inde!. Left distal femur, SKX 38003.

DAMAGE MARKS ON FOSSIL SONES

222

SKX 6276

o. \:; SKX 44074 Fig.5 Examples 01 Swartkrans fossils showing chewed edges. Member 1 Lower Bank: SKX 8055, SKX 6276; Member 2: SKX 2043; Member 3: 22743; Member 5: SKX 44074. ". Bovid size class ,. Skull condyle, SKX 37749; left scapular píece, SKX 39080; petvic fragment, SKX 33616; humeral pieces, SKX 19555, 25841, 27648, 29840; radial shaft, S KX 39015; astragali, SKX 27608, 29701, 30721, 28740; calcaneus, SKX 32320; patella, SKX 22072; phalanx, SKX 30783. Bovid size cJass 111. Pelvic pieces, SKX 22743, 36310; distal humerus, SKX 27762; right proximal ulna, SKX 28479; left femur head, SKX 36610; let! astragalus, SKX 38559; metapodial fragment, SKX 26932; carpal, SKX 27831. MisceJlaneous. Vertebral fragments, SKX 22973, 32297; rib fragment, SKX 33255; femoral fragment, SKX 31489; calcaneus, SKX 37749; phalanx, SKX 25450. Bone flakes. Twenty eight pieces of the foHowing lengths: 2-3 cm, 1; 3-4 cm, 6; 4-5 cm, 7; 5~6 cm, 9; 6-7 cm, 2; 7-8 cm, 1; 8-9 cm, 1; 10-11 cm, 1. In de termina te fragments. 9.

Member 5 A total of 119 specimens, made up as tollows: Equus burcheJlii. Immature right distal tibia, SKX 41192. Bovidsize cJass 11. left mandibular pieces, SKX 40651,42541; atlas vertebral fragment, SKX 43306; axis vertebral fragment, SKX 41474; lumbar vertebrae, SKX 41335; sacrallragments, SKX 42268, 42951; scapular pieces, SKX 40608, 40613,

40617,40745,41058,41064,41279,41609,41638,416é 41819,42111,42143,42331,42379,42413, 42470, 428~ 42856, 43273, 43283,43876, 43892; pelvic pieces: Sb 40598,40601,40606,40658,41283,42273,42348,4247 42491, 43894; humeral shafts, SKX 41284, 41340; dis humeral pieces, SKX 41554, 42328, 42698, 44074, 473C 41550; left proximal ulna, SKX 45017; right femoral head, S~ 42391; distal femoral pieces, SKX 41626, 47329, 415~ 45178, 44430; right proximal tibia, SKX 43884; calcaneL SKX 45048. Bovid size class 111. Vertebral piece, SKX 47338; scapul plece, SKX 45075; pelvic piece, SKX 41850; femoral fragme SKX 43163. Bovid size cfass IV Mandibular fragment, SKX 41398. Misce/laneous. Skull fragments, SKX 40691, 41470, 418~ 42881, 42910; vertebral fragments, SKX 40688, 4161 41639, 42954,43572,43716,43917,44075,44681; rib~a. menls, SKX 41608, 43415; scapular fragments, SKX 406~ 41835,43253,43344,44446; pelvíc fragments, SKX 418~ 42446,40690,42923, 42931,43984,44512,44865;proxirr humeral fragments, SKX 45384, 47316; proximal femoral fra ment, SKX 41639; proximal tibial fragment, SKX 4391 phalangeal fragmenl, SKX 43914. Bone flakes. Fourteen pieces ofthe following lenglhs: 4-5 c 2; 5-6 cm, 5; 6-7 cm, 3: 7-8 cm, 1: 8-9 cm, 2: 9-10 cm, 1 Indeterminate fragments. 8.

DAMAGE MARKS ON FOSSIL SONES

-o

223

6-

SKX 35445

SKX 5445

SKX 12388 Fig.6 Examples of Swartkrans lossils showing porcupine gnawing marks. Member 1 Lower Bank: SKX 5445; Member 2: SKX 12388: Member 3: SKX 35445.

PORCUPINE-GNAWED SONES (Figs 6, 8a) Marks caused on bones by the gnawing 01 porcupines are highly characteristic, resulting in the reduction of skeletal elements to distinctively-shaped cores (Maguire, 1976; Brain, 1981). Porcupines gnaw bones to wear down lheir constantly growing incisors and to augment their phosphate intake. The gnaw marks result lrom the drawing 01 the upper and lower incisors overthe surface of the bone being gnawed. The marks are shallow, Ilal-bottomed scoops, each the width oi the porcupine's incisors. Such marks were observed on 46 specimens, as detailed below.

Member 1, Lower Bank A total of eight specimens, made up as follows: Miscellaneous. Vertebral fragment, SKX 38479. Bone flakes. Four pieces oi the following lenglhs: 2-3 cm, 1; 4-5 cm, 2; 5-6 cm, 1. Indeterminate fragments. 3. Member 2 A total 01 12 specimens, made up as lollows: Hipparion Iybicum steyt/eri. Left distal metatarsallll, SKX 453. Bovid size c1ass /11. Scapular fragment, S KX 12388. Bone flakes. Nine pieces 01 the foliowing lengths: 2-3 cm, 1;

3-4 cm, 1; 4-5 cm, 2; 5-6 cm, 2; 6-7 cm, 1; 7-8 cm, 1; 11-12 cm, 1. Indeterminate fragmento 1. Member 3 A total 0126 specimens, made up as follows: Hipparion Iybicum steytleri. Right distal metacarpal 111, SKX 22936. Bovid size class 111. Right astragalus fragment, SKX 38251. Bovid size c1ass IV Right calcaneus, SKX 22651. Bone flakes. Twelve pieces of the following lenglhs: 2-3 cm, 1; 3-4 cm, 5; 4-5 cm, 1; 5-6 cm, 2; 9-10 cm, 1; 12-13 cm, 1; 15-16cm, 1. Indeterminate fragments. 11 . Member 5 None found. SMALL-RODENT-GNAWED SONES (Figs 7, 8b) The marks in this category are similar to those on porcupinegnawed bones, but are made by the incisors 01 mouse- and rat-sized rodents. Individual marks resemble vertical-sided grooves, more than the shallow scoops of porcupine-gnawed bone. Such marks were observed on 141 specimens, as detailed below:

DAMAGE MARKS ON FOSSIL SONES

224

SKX 41537

SKX 42484

SKX 41424

t_j

Fig.7

Examples 01 Swartkrans fossils showing small-rodent gnawing marks. Member 3: SKX 41424, SKX 41537, SKX 42484, SKX 44961.

Member 1, Lower Bank A total 01 14 specimens, made up as lollows: Bovid size class 11. Left proximal metatarsal, SKX 14597; 1st phalanx, S KX 14670. Miscellaneous. Tooth fragment, SKX 4677; rib fragment, SKX 47062; carpa!, SKX 9416. Bone flakes. Seven pieces of the following lengths: 2-3 cm, 2; 3-4 cm, 2; 4-5 cm, 1; 5-6 cm, 2. Indeterminate fragments. 2. Member 2 A total of 12 specimens, made up as follows: Procavia antiqua. Right mandibular Iragment. SKX 665. Bovid size class 11. Phalanges, SKX 403, 3130. Bovid size class IV Right proximal radial fragment, SKX 294. Misceflaneous. Toolh fragment, SKX 45939. Bone flakes. Five pieces of the following lengths: 2-3 cm, 1; 3-4 cm, 2; 4-5 cm, 1; 6-7 cm, 1. Indeterminate fragments. 2. Member 3 A total 01 15 specimens, made up as follows: Bovid size class 1/. Pelvic piece, SKX 26921. Bovid size class 111. Left distal humerus, SKX 38509. Hyaenid indet. Right ulna shaft, SKX 36657; 1 st phalanx, SKX 32397.

Misceflaneous. Vertebral fragment, SKX 27076; rib fragmen SKX 36432. Member 5 A total of 102 specimens, made up as follows: Antidorcas bondi. Right mandibular fragment, SKX 41424. Procavía capensis. Ríght mandibular piece, SKX 41226. Bovid size class l. Right humeral shaft fragment. SKX 4381' Bovid size class 11. Atlas vertebral fragment, SKX 4184 1 scapular pieces, SKX 41838, 42484; humeral pieces, SK 41120, 41298, 42372. 42748; left proximal ulna, SKX 4265~ distal femoral pieces, SKX 42686, 42697; femoral sha pieces, SKX 41531, 43181; tibial pieces, SKX 40599, 4259; tarsal bone fragment, SKX 45367; proximal metacarpals, SK 41535,41537,45204; proximal metatarsal, SKX 41320; meté podial pieces, SKX 41762, 42609; astragali, SKX 4287; 44859; calcaneus pieces, SKX 43404, 47330, 47803; phi langes, SKX 40605. 40638, 42414, 42426, 42891, 44370. Bovid size class 111. Sternal fragment, SKX 42749. Aves indet. Tarsometatarsus, SKX 42865. Miscel/aneous. Mandible pieces, SKX 42704, 41380; rib Ira! ments, SKX 40976, 47453; scapular fragments, SKX 4186/ 41949, 43278, 47378; radial shaft, SKX 42176; metapodi, pieces, SKX 43122, 41535. Bone f1akes. Forty five pieces of the following lengths: 1-2 en 8; 2-3 cm, 10; 3-4 cm, 11; 4-5 cm, 8; 5-6 cm, 3; 6-7 cm, '

DAMAGE MARKS ON FOSSIL BONES

225

.1VI.ember 1, Lower Bank A tolal of 19 specimens, made up as follows: Bovid size class 11. Phalanx, SKX 8336. Bov;d size class 111. Humeral pieces, SKX 11899, 45410; libial pieces, SKX 45392, 45409; phalanges, SKX 45504,45659. Bav;d size c/ass IV Distal metapodial, SKX 14300. Miscel/aneous. Horncore fragment, SKX 5271 b. Bone flakes. Five pie ces of the following lengths: 1-2 cm, 1; 2-3 cm, 1; 4-5 cm, 1; 7-8 cm, 1: 9-10 cm, 1. Indeterminate fragments. 5. Member 2 A total of eighl specimens, made up as follows: Bovid s;ze class/l. Phalanges, SKX 3972, 1245, 3417. Bovid size class 111. Righl aslragalus, SKX 3988. Miscel/aneous. Ulna shaft, SKX 726. Bone flakes. Two pieces of the following lengths: 3-4 cm, 1; 4-5 cm, 1. Indeterminate fragmento 1. Member 3 A total of nine specimens, made up as follows: ?Gazella sp. Horncores, SKX 29281, 36311. Bavid size class l. Right distal femur shaft, SKX 36480. Bovid size class 11. Phalanges, SKX 26897, 27263. Bovid size class 111. Right distal femoral piece, SKX 36304. Miscel/aneous. Horncore fragments, SKX 29791, 39829; femoral fragment, SKX 32611. fndeterminate fragmento 1. Member 5 A total of lour specimens, made up as follows: Bovid size cfass/1. Right distal femur, SKX 42653; left proxil11al tibia, SKX 42664. Indeterm;nate fragments. 2.

Fig.8 Scanning electron microscope images of gnawing marks on Swartkrans fossil bones. a: porcupine gnawíng marks on SKX 47030 from Member 1 Lower Bank; b: small rodent gnawlng marks on SKX 42484 Irom Member 5.

7-8cm,1. Indeterminate fragments. 9. INSECT-BORED BONES (Fig, 9) As shown in Fig. 9, bones in this category show circular holes, penelraling inlo lhe specim en, either al right angles or obliquely to the surface. The holes vary in diameter from two to fíve millimetres and may penetrate for up to one centimetre. Walls of the tunnels are typically smooth and their bases are rounded. In some cases Ihere is a raísed lip around the periphery 01 the hole. It is surmised that the smaller of lhese holes and lunnels were made by lermites, while the larger exampJes were the work of beette or moth larvae. Marks conforming to this description were observed on 41 specimens, as detailed below.

D1GESTED BONES (Fig. 10) Digested bones (hardly ever exceeding six cenlimetres in length) are assumed to have been swallowed by a carnivore, affected by the digestive juices, and then eilher regurgitaled or voided with the droppings. They typically show an overall rounding 01 sharp edges, etching or polishing. In all, 156 bone pieces conforming to this description were found, as detailed below. Member 1, Lower Bank A total of 92 specimens, made up as lollows: Bovid size class /l. Astragalus fragment, SKX 6575. Miscel/aneous. Rib fragment, SKX 13880; humeral shaft fragment, SKX 6310; tooth fragments, SKX 4659a, 4890, 5785, 18056. Bane ffakes. Seventy one pieces of the following lengths: 1-2 cm, 27; 2-3 cm, 33; 3-4 cm, 8; 4-5 cm, 3. Indeterm;nate fragments. 14. Member 2 A tolal of 28 pieces, made up as follows: Bov;d size class /l. Left patella, SKX 2765. Miscellaneous. Pelvic fragment, SKX 231; distal femoral condyle, SKX 1097; phalanx, SKX 46938.

DAMAGE MARKS ON fOSSll BONES

226

'f[ SKX 45410 Fig.9 Examples 01 Swartkrans fossils showing insect boring. Member 1: SKX 45409, SKX 45410.

Sone flakes. Twenty two pieces of the following lengths: 1-2 cm, 3; 2-3 cm, 14; 3-4 cm, 3; 4-5 cm, 2. Indeterminate fragments. 2. Member 3 A lotal of 23 pieces, made up as follows: Bovid s/ze class /. Righl radial carpal, SKX 48437. Bone tlakes. Eleven pieces of Ihe following lengths: 1-2 cm, 1: 2-3 cm, 3; 3-4 cm, 5; 5-6 cm, 2. Indeterminate fragmenls. 11. Member 5 A total of 13 pieces, made up as follows: Miscellaneous. Skull fragment, SKX 42897. Bone flakes. Twelve pieces 01 the following lengths: 1-2 cm, 4; 2-3 cm, 4; 3-4 cm, 4.

DISCUSSION Carnivore-induced marks and their implications As detailed aboye, marks attributed to carnivore feeding activity have been described under the headings 'tooth scratches,' 'punctures' and 'chewed edges.' The incidence of these marks relative lo the enlire sample is shown in Table 1. The incidence 01 carnivore-induced damage marks shows

that carnivores, presumably both predators and scavengers, were involved in Ihe bone-accumulating process in Members 1,2,3 and 5. The assemblages from Members 1 and 2 show approximately the same percentage incidence of damage marks (0,25 and 0,21). The figure for Member 3 is appreciably higher (0,49), while it is higher still in Member 5 (1,24). While these incidence figures indicate unquestionable carnivore involvement in the history ofthe bone assemblages, the percentages are remarkably low. Blumenschine (1988) showed that, in an experimental assemblage worked over by spotted hyaenas, an average of 15 % of long-bone fragments showed distinctive tooth marks. However, most of the individual Swartkrans fossil specimens are small pieces, presumably a result of breakage caused by lrampling and weathering, while they were exposed on the cave floor before fossilization. To this must be added the damaging effects of excavation and removal from the matrix, which applies to assemblages from all the Members studied. It was noticeable, however, that the individual fossil specimens from Member 3 tended to be more complete than those from Members 1 and 2, while fossil bones from Member 5 were the most complete of all, although they were often noticeably weathered. On the basis of the observed damage marks, it is very difficult to be sure of the identity 01 the carnivores involved. Damage marks consistenl with those caused by both leopards and hyaenas were seen.

DAMAGE MARKS ON FOSSIL SONES

227

SKX 4876

SKX 9304

Fig.10 Examples 01 Swartkrans fossils showing lhe inlluence 01 carnívore digeslion. AII the specimens shown are Irom the Lower Bank 01 Member 1.

Implications of porcupine-gnawed bones II is well known Ihat porcupines (Hystrix africa eaustralis) collect large numbers of bones in Iheir retreats and breeding lairs and that they gnaw them there. The incidence of gnawed bones in a porcupine-collected assemblage will vary according to the availability of bones in the home range 01 the porcupines involved, while the incidence has been shown to vary from 22 to 100 % in different porcupine lairs (Brain, 1981). Porcupine-gnawed bones were lound in the fossil assemblages Irom Members 1, 2 and 3, though they were absent from the Member 5 sample. However, the numbers were extremely small: eight for Member 1; 12 tor Member 2; and 26 for Member 3. This indica les thal porcupines did make use 01 the cave during the accumulalion periods 01 Ihese Members, but their contribution to the building up 01 the assemblages was almost insignificant.

Implications of small-rodent-gnawed bones It is not unusual to find bones in Southern African caves that have been gnawed by rodents of mouse- or ral-size. This indicates that these rodents either live in the caves or visit them, while their reasons for gnawing bones are presumably the same as for porcupines, Le., Ihe need to wear theír incisors and lo augment their phosphate inlake. It is unlikely that such

Table 1 The incidence 01 carnivore-induced damage marks on fossil bones trom Ihe various Swartkrans Members. Pie ase note tha! Member 1 relers exclusively lo the Lower Sank 01 Member 1. Slraligraphic Members

2

3

Toolh scratches 103 66 46 Punctures 12 5 20 21 Chewed edges 53 74 Bones wilh carnivore marks 131 197 72 Total number ot bones 52496 34312 40239 0,25 % showing carnivore marks 0,21 0,49

4

5 96 37 11 9 252 20264 1,24

small rodents were instrumental in bringing bones lo the caves, bul they simply modilied those bones Ihey encountered on the cave floors. As detailed above, 143 small-rodent-gnawed bones were found in the assemblages examined, individual numbers being as follows: Member 1, 14; Member 2, 12; Member 3, 15, and Member 5, 102. These figures indicate that mice or rals, of unknown identity, occasionally gnawed bones on the cave floor during the accumulation períods 01 the four Members studied. They were clearly much more active in the case of

228

DAMAGE MARKS ON FOSSIL SONES

Member 5 than in the others, aJthough the absence of porcupine-gnawed banas in this assemblage suggests thatfhe Member 5 depositional space, or 'Bond; channel' as it is callad, was ínaccessibre or unattractive to porcupines, but that this was not trua far smaller rodents.

Implications of lnsect-bored banes Little can said about the fact that 41 fossil bones in the collections examinad had been borad by insects, either termites or the larvaa 01 moths or beetlas. Such insects were obviously occasionally active on the floor of the cave during the accurnutatíon periods 01all Members investigated. Implications ot digested banes The presence 01 digested banas on a cave floor suggests that the cave was frequented by carnivores that either regurgitated the banas or vaided them with their oroppings. The mast likely carnivores to be involved in a contem-

parary Southern Atncan context are leopards or brown hyaenas. Of the 156 bone fragments considered te have been digested, 92 occurred in Member 1, 28 in Member 2, 23 in Member 3, and 13 in Member 5. The carnivores responsible left these traces either inaide the cave, or within the catchmen1 area of its entrance.

CONCLUSION This study ot the occurrence and incidence of damage marks on Swartkrans fossil bones suggests that carnivores, probably hyaenas and larga cats, were involved in the boneaccumulating process of all the assemblages examinad. The role 01 porcupines appears to have been negligible. ACKNOWLEOGEMENT I thank Virginia Watson for her help with specimen identifications.

REFERENCES

BLUMENSCHINE, R. J., 1988. An experimental model 01 the timing of hominid and carnivore influence on archaeological bone assemblaqes. Journal of ArchaeoJogical $cience 15: 483-509. BRAIN, C. K., 1981. The hunters or the hunteen An introduction to African cave teononomv. University of Chicago Press, Chicago. BUNN, H. T., 1981. Archaeological evidence tor meat-eating by Plio-Pleistocene hcrmnids from Koobi Fora and Olduvai Gorge. Nature 291: 574-577. MAGUIRE, J., 1976. A taxonomic and ecoloqical study 01 the living and tossn Hystricidae, with particular reterence to southern Africa. Ph.O. dissertation, University of the Witwatersrand.

NEWMAN, A. A. and BRAIN, C. K., 1988. A scanning electror microscope study of damage marks on toss!l bcnes. Proceeding~ of the Electron Microscopy Society of Southern Africa 19: 137-138 POTTS. R. and SHIPMAN, P., 1981. Cutmarks made by stone toots on bones from Olduvai Gorge, Tanzania. Nature 291: 577-580. SHIPMAN, P., 1981. Applications of scanning eleetron mlcroscopy le laphonomic problema. AnnaJs of tne New York Academy 0, Sciences 376: 357-386. ROSE, J. J., 1983. A replication technlque far scanning electror microscapy: applications ter anthropoloqtsts. American Joumal o, Physical Anthropology 62: 255-263.

Chapter 10

The Occurrence of Burnt Sones at Swartkrans and Their Implications for the Control of Fire by Early Hominids c. K. Brain Transvaal Museum, P.

o. Box 413, Pretoria, 0001 $outh Africa

In the course of the excavatlcn 01Member3at Swartkrans, 270 preces ot tosen bone were recovered that snowed signs 01 having beeo burnt. 01 tnese, 45 were buff in colcur, suggestive 01 brial heating, 51 were dark brown as il carbonizad, and the remainder had apparently been calcinad to varying degrees. Histoloqical and chernlcal examination 01 the darkened banas confirmad that they had been aubjected to temperaturas similar to those

reached in camp-ñres. while their distnbutlcn in the sediment profile suggested that tires had been tended repeateuly in the Member 3 gulley during lts period 01infilling. Member 3 has yielded remains 01 nine individua!s 01 Austra/opithecus robustus, but it ls not known whether it was these homfnlds. or Homo erectus individuals, who were responslble for tbe control of the tires.

INTRODUCTION Calciñed sediment referred to Member 3 at Swartkrans was found to occupy a steep-sidad gulley, up to seven metres deep, that had been eroded into Members 1 and 2 along the southwest wall 01 the cave. The reader is referred to Chapter 1 01 this volume, and particularly to Fig. 14 in that chapter, far further detalla 01 the extent 01 Member 3 and its place in the overall history 01the cave. The gulley, at least 20 m long and up to 5 m wide, appears to haya been below a substantial dolomite root, with a steeply inclined opening somewhere aboye the southern margin of the cave. Sediment which washed into thís gulley trom the hillside aboye shows layering inclined upwards towards the south, although stratification ís almost horizontal towards the back 01 the gulley. A large sample 01 the Member 3 sediment was excavated between 1984 and the end 01 1986, as lhe last phase 01the Swartkrans project. On 21 March 1984, when we were excavating in grid square W31S2 at a depth 01200 em below the permanen! grid datum, clase to the east wall 01the Member 3 gulley, we carne across a piece ot fossil bone that appeared to have been burnt. Now designated SKX 26158, this is part 01 abone Ilake, generally brownish in colour, but with about ha!f 01 its thickness diseoloured to a mueh darker and characteristic brownish blaek (Fig. 1). Had this come lrom a tater archaeoJogical excavañon, I wouJd have had no hesitation in assuming that it had been partJy charred in a fire, but at an estimated age in excess of ene million years, the presence of burnt bone in a hominid living area required further verifieation.1 therefore took part 01the broken bone f1ake to the National Physical Labora· tory of the e.S.I.R. in Pretoria, where a test was carried outfor the presence of carbono The result confirmed the existenee 01 free carbon, suggesting that the blaekening wa5, in fact, the result of charring.

As the exeavation of the Member 3 sediment progressed, numerous other pieces of fossil bane carne to light that appeared to have been burnt, Allhough partially calcified, the sediment of Member 3 was soft enough to be excavated with hand-tools, so that every scrap of bane was recovered when the material was sfeved and washed. The pieces 01bone which had the appearance of having been heated eventually nurnbered 270, out 01 a total 01 59488 bone speeimens Irom Member 3 (excluding mierofaunal bones assumed to have been introduced by owls). Excavation proceeded in quartermetre squares, controlled by the overhead grid, and in spits 10 cm thick, so that all specirnens faund could be located spatially with reasonable aeeuraey (Fig. 11). It soon became elear that if the suspected bones had in fact been heated in ñres, the practice ot ftre-maklnq, or at least ñrs-tendtnq, must have been a regular activity within the cave, as apparently bumt bones oceurred in as many as 23 suecessíve excavation spits within the profile of a single one-metre grid square. Experimental contlrmatton 01 heating was obvlousty required in the case of these very old bones and various tests were eonsequently made. A considerable Iiterature aJready exists on the effects of lires on bones, (see, e.g., Shipman el et., 1984, lar a review), so that the unambiguous recognition of artificial heating is now possible. While the microscopio inspeetion af the entire specimen and histological exarnination 01 thin sections is very uselul, the surest confirmation 01 charred bone is the demonstration of free carbon by che mical means. This was carried out by Andrew SiIlen at the Archaeometry Laboratory 01the University 01Cape Town (Brain and Sillen, 1988), whose full descripticn 01the procedures and results forms Chapter 11 01the present voJume. Although ¡nfarmation is now available on the temperatures normally attained in wood fires (e.g., Gilchrist and Mytum, 1986; SeHorno and Harris, 1990), it seemed desirable, soon

OCCURRENCE OF SURNT SONES

230

Fig.1

The lirst piece of burnt bone, designated SKX 26158, found in Member 3 sedimenl al Swarlkrans on 21 March 1984. It is part 01 abone flake, generally brownish grey in colour, but as shown in lhis broken section, is discoloured a black-grey on one side, apparently through charring.

SKX 26158 t Fig.2

after the linding of apparently burnt bones at Swartkrans, to confirm what temperatures were attained in a camp-lire made trom wood lrom the site. The most common tree currently growing at the Swartkrans cave is the white stinkwood, Celtis kraussiana. which has a rather solt, pale-coloured wood. 1I is lound throughout the Transvaal dolomite areas and olten grows direct/y out 01 the enlrances to shalt-like cave openings, where its seedlings have been protected Irom trost and lire in the otherwise olten open grassland environment. Although the temperatures attained in many experimental lires were monitored, the lirst test will be described here. It was carried out at Swartkrans on the evening 01 25 September, 1985, which was clear and moonlit, with a light wind Irom the northwest and an air temperature 0126 oC. We selected one dry Ce/tis log, 2.5 m long and 30-40 cm in diameter, (rom a tree that had grown directly out 01 Ihe Member 3 gulley. The lire was started at 17:35, using several smaller branches Irom the same tree, 10-15 cm in diameter, and temperatures were monitored using a long thermocouple probe Iinked to a digital thermometer (Fig. 2). Seven minutes alter the lire was started, a temperature of 650 oC was reached in the llames. After one hour the log was burning strongly and a hole had developed midway along its !ength, surrounded by glowing coa/s. It was here that the maximum temperature attained in the fire was recorded, 860 oC, as shown in Fig. 3. At that time the temperature 01 the flames immediately aboye the log was 690 oC, which declined to 550 oC at their upper Iimit. Coals around the periphery 01 the lire reached a temperature 01 717 oC and these were used 90 minutes alter slarting the lire to grill the lorequarter 01 a sheep, which was placed directly upon a flat bed 01 coals. The maximum temperature attained within lhe intercostal muscles 01 this lorequarter, and lhose surrounding the scapula. was 153 oC: The Ce/lis log burnt away entirely in the course 01 [he night, leaving a bed of white ash that retained a temperature 01394 oC 16 hours alter the lire had started. This declined to 162 oC at 24 hours and 88 oC at 28 hours, when observations were discontinued. Subsequent experimental lires using wood from Acacia karoo trees did not attain temperatures as high as those of [he lirst lire, the maximum reached being 688 oC. In experiments to determine the effect 01 specific temperatures on bones, an electric kiln litted with precise temperature control was used. A large selection of Iresh and weathered

An experimental fire of Celtis wood, made at Swartkrans on ; September 1985 and described in the texl. Temperatures altained varíous parts 01 lhe tire were measured with a digital thermomel attached to a long thermocouple probe as shown here.

bones was used and specimens were typically heated to selected temperature and maintained there lor 30 minutl belore cooling. The most obvious effect was tha!, al 300 0, the bones generated a good deal 01 acrid smoke and changl colour to a dark brownish black. At higher temperatures n· discolouration laded and the bones became calcined, white colour and very light. In order lo lurther quantify lhese changes, a Iresh, dellesh, radius 01 a hartebeest. Alcelaphus buse/aphus, was cut in seven transverse segments. Each 01 these was then heatl to one of the lollowing temperatures. maíntained there lor ; minutes and then allowed to cool slowly: 200, 300, 400, 5e 600,700 and 800 oC, Each segmenl was lurther divided: 01 part was embedded in clear resin, mounted on a microscol slide, and ground into a thin section; one part was sent A. Sillen lar chemical determination olfree carbon conten!, wh the remaining portions were retained lor luture reference. An indication 01 colour changes with heating of the vario segments is given in Fig. 4. Appearance changes may , summarized as lollows: At 200 oC, the colour becomes slightly darker than thal the unheated bone, and a thin laye r 01 yellow·brown lat ex ud over the surface. In thin section (Fig. 5b), histological detail similar lo Ihat of unheated bone, except for the presence numerous small masses 01 carbon surrounding lacunae wilt Ihe bone struclure. These result in the slight buff discolourali 01 the whole bone. At 300 oC, the colour changes to a dark brown-black and t superficial lat layer is carbonized to a glislening bla· encrustation. In Ihin section (Fig. 5c), a striking change apparent: the lame llar struclu re 01 the bone is accentuated carbon deposjtion within its structure, while the brown-bla discolouration is intense. Cracks, spreading outwards fr( Haversian canafs, are often observed. At 400 oC, the dark brown-black colour starts lo fade, p ticularly close lo the outer surface 01 the bone. Some 01 1 encruslation loses its glistening black appearance. In ti section (Fig. 5d), the heavily carbonized parts 01 the bone é

OCCURRENCE OF BURNT BONES

231

Fig.4

The effect 01 controlled heating on lhe appearance 01 bone. As described in the text, a tresh bul delleshed radius 01 a hartebeest was cut transversely inta seven segments and each was heated to a selected temperalure in an electríc kiln and held there lor 30 minutes befo re beíng coaled. At 300 oC the bone turned black, as a resull of the carbonization of its collagen, bul this darkening gradually laded at higher temperalures, as the sample was calcined.

Fíg.3

Temperatures allained in the experimental Ce/lis wood fire shown in Fig. 2. The maximum temperature of 860 oC was reached in a hole burnt into the log, midway along its length.

restricted to areas c10se to Haversian canals, but the whole matrix 01 the bone shows numerous cracks. At 500 oC, further lading occurs so that the outer margins of the segment are a uniform grey-buff, while the interior remains a darker brown. The superficial encrustation fades to white. In thin seclion (Fig. 5e), only a lew pockets 01 unoxidized carbon remain, while the bone matrix in general is pale coloured, with little lamellar structure but numerous cracks. At 600 oC, the outer margins of the piece are an even, pale grey, but a reduced area within the bone's interior retains a darker grey coJour. In thin section (Fig. 5f), the bone structure is still visible, but cracking is more prominent. At 700 oC, the entire piece is a brilliant, chalky white, except for lhe outer surfaces which acquire a líght blue tinge. In thin section (Fig. 5g), shrinkage 01 the pale-coJoured matrix results in the widening of cracks. At 800 oC, the appearance 01 the piece is unchanged from that subjected to 700 oC. In thin section (Fig. 5h) however, the microstruclure of the bone has largely disappeared due lo

recrystallization 01 the hydroxyapatite, whjle widening 01 cracks has progressed further. On the basis 01 appearance, bolh normal and microscopic, il is possible lo classify the 270 pieces of fossil bone from Member 3, which are suspecled of having been burnt, into four categories, reflecting the temperalure to which it is assumed they were subjected. The reconslruction of the precise temperature to which a bone has been heated is complicated by the fact that prolonged heating al, say 300 oC, may have had lhe same result as a short interval of heating to a much higher temperature. This applies particularly to the bone-carbonization phase. It has been observed that at about 300 oC, a Iresh bone undergoes carbonization of its collagen and other organic components, with the result that a black colour is imparted to the specimen. Prolonged subsequent heating causes oxidation 01 this carbon, which escapes as carbon dioxide, leaving only the mineral component 01 the bone in the form of hydroxyapatite. In this form the bone is said to be calcined. It is difficult, however, to decide whether a white, calcined bone lost its carbon rapidly at a high temperature, or more slowly in the course 01 less intense, but prolonged, healing. The four categories of Member 3 burnt bones are defined by the typical colour of lhe specimens: 1. Buff: specimens are typically a uniform light buff colour, and some examples of bones in this calegory are shown in Fig. 7; in thin section (Fig. 6a), the appearance is little different from that of normal fossil bone from Member 3, except for an accentuation of the lamellar structure as a result of slight carbon deposition within il. lt is assumed that this discolouration was índuced by heating to a low temperature, below 300 oC. 2. Brown: specimens are typically a dark chocolate-brown colour, though this IS often surrounded by paler margins. Some examples 01 bones in this calegory are shown in Fig. 8. In thin section (Fig. 6b), the specimen typically shows an accentuation of the lamellar slructure of the bone, pronounced darkening due to the presence of diffused carbon

232

OCCURRENCE OF SURNT SONES

Fig.5 Thin seelions of the experimentally heated bone samples shown in Fig. 4. Sea le bar = 1 mm. a: normal, unheated bone; b: 200 oC; e: 300 oC; d: 400 oC: e: 500 oC; f; 600 oC; g: 700 oC and h: 800 oC. Further details are given in the 1exl.

OCCURRENCE OF BURNT BONES

233

Fig.6 Thin sections of fossil bones assumed to have been burnt, from Member 3 at Swartkrans. Sca)e bar = 1 mm. a: a 'buff' category specimen, SKX 29612 ; b: a 'brown' calegory specimen, SKX 33063; e: a 'brown-grey' category specimen, SKX 29681, and d: a 'grey' category specimen, SKX 33939. Further details are given in the tex!.

and, often, the presence of cracks running outwards from the Haversian canals. It is assumed that such specimens were subjected to a temperature of 300-400 oC. 3. Brown-grey: specimens lypically show a gradation from brown to grey, some examples being shown in Fig. 9. In lhin section (Fig. 6c), darkening due to the presence of carbon is still visible in isolated pockets, and cracking hasoccurred. S ubjection to a tem pe ratu re in ex ces s of 400 oC is suggested, or prolonged heating al a lower temperature. 4. Grey: specimens are typically a uniform pale grey, are frequently light in weight and have a talc-like feel. Some examples are shown in Fig. 10. In thin section (Fig. 6d), the bone matrix has lost much of its structure and wide cracks, often infiltraled by manganese dioxide, have for!TIed.

SKX 39562

SKX 28391

SKX 26128

SKX 30623

SKX 28973 SKX 37763 Fig.7 Examples of 'buff' category burnt bones from Member 3 al Swartkrans.

~,

OCCURRENCEOF BURNT BONES

234

SKX 33607/8

-

SKX27717

SI<X 29458/ SO

SKX 29121

SKX .23649/50

SKX 33067/8

~

SKX 39316

SKX 30106

SKX26174

Fig.8

SKX 37932

SKX 37830

Examples 01 'brown' category burnt bones from Member 3 at Swartkrans.

Fig.9 Examples of 'brown-grey' category burnt bones from Member 3 al Swartkrans.

THE SWARTKRANS BURNT BONE ASSEMBLAGE FROM MEMBER 3 The 270 bones regarded as having been artificially heated, grouped according to the four categories described above, are listed below, providing an identification of each piece, its catalogue number and grid square, and the depth at which it was found. Three further specimens, SKX 37763, 39466 and 39562, lack detailed stratigraphic data. Category 1. Buff-coloured bones (Fig. 7) Total number 01 pieces: 45 Bovid size class I Mandible piece W3/S2 NW SKX 28973 Bovid size class III R. distal humerus SKX 29876 W3/S3 SE Bovid horneo re pieces SKX 28391 W2/S3 NE W3/S2 SW SKX 30622/3 W3/S3 SE SKX 45801 Miscellaneous pieces W3/S2 NW SKX 28974 W3/S2 SE SKX 29079 SKX 29135 W3/S2 NE W3/S2 NE SKX29138 SKX 30112 W2/S3 NE W3/S2 NE SKX 45699 Sone flake, 1-2 cm long W3/S3 SW SKX 31051 Sone tlakes, 2-3 cm long W3/S2 NE SKX 23124 W2/S4 SE SKX 23652 SKX 23976 W2/S4 SE SKX 24417 W2/S4 NE

240-250 cm

310-320 cm 230-240 cm 300-310 cm 240-250 cm 240-250 cm 240-250 cm 230-240 cm 230-240 cm 300-310 cm 320-330 cm 350-360 cm 210-220 200-210 220-230 230-240

cm cm cm cm

SKX 28958 SKX 28992 SKX 29204 SKX 29475 SKX 29495 SKX 29946 SKX 29948 SKX 30196 SKX 33748 Sone f1akes, 3-4 cm long SKX 22057 SKX 23974 SKX26123 SKX 26137 SKX 28957 SKX 29472 SKX 29667 SKX 29712 SKX 29947 SKX 29950 SKX 29952 SKX 30767a SKX31905 Bone flakes, 4-5 cm long SKX 22061 SKX 30189 SKX 33087/90 Bone flakes, 5-6 cm long SKX 26128 SKX 29612 SKX 32934 Sone flake. 7-8 cm long SKX 27415

W2/S2 W3/S2 W3/S2 W3/S2 W3/S2 W3/S3 W3/S3 W3/S3 W4/S2

NW NE SE SE SW NE NE NE NE

240-250 cm 220-230 cm 230-240 cm 270;:;-280 cm 270-280 cm 310-320 cm 310-320 cm 300-310 cm 310-320 cm

W3/S2 NE W2/S4 SE W3/S2 NE W3/S2 NE W3/S2 NW W3/S2 SE W3/S2 SE W3/S2 NW W3/S3 NE W3/S3 NE W3/S3 NE W3/S3 NE W3/S3 NE

190-200 cm 220-230 cm 210-220 cm 210-220 cm 240-250 cm 270-280 cm 260-270 cm 260-270 cm 310-320 cm 310-320 cm 310-320 cm 340-350 cm 400-410 cm

W3/S2 NE W3/S3 NE W3/S4 NE

190-200 cm 300-310 cm 360-370 cm

W3/S2 NE W3/S2 SE W4/S2 NE

210-220 cm 280-290 cm 290-300 cm

W2/S4 NW

300-310 cm

OCCURRENCE OF BURNT BONES

SKX 30873

r SKX 31957

~

SKX 31456

SKX 31578

SKX 29660

SKX 39466

Fig.10 Examples ai 'grey' category burnl bones from Member 3 at Swartkrans.

Category 2. Brown-coloured bones (Fig. 8) Tolal number of pieces: 51 cf. Equus capensis R. astragalus fragment W2/S6 SW SKX 26524 cf. Procavia sp. Shaft piece W1/S5 NW SKX 21538 Aves: Phasianidae L. distal coracoid W2/S4 SW SKX 24369 Bovid size class 1I Astragalus W2/S4 SE SKX 23649/51 Metapodlal fragment SKX 33073/80 W3/S4 NE Bovid size c1ass I1I Metatarsal piece W3/S2 SE SKX 29458/80 Bovid horneo re pieees W3/S3 NW SKX 30678 W3/S4 NE SKX 35630 SKX 35926b W5/S3Miscellaneous pieces W3/S2 NE SKX 26727/9 W3/S2 NW SKX 28966 W3/S4 SW SKX 29784 W2/S3 SW SKX 29814/30 Bone flakes, 1-2 cm long W2/S4 SE SKX 23982 SKX 26129 W3/S2 NE W3/S2 SW SKX 26144 SKX 26157 W3/S2 SW SKX 32932 W4/S2 NE Bone flakes, 2-3 cm long SKX 23835 W2/S4 SE SKX 23836 W2/S4 SE SKX 26126 W3/S2 NE SKX 26131 W3/S2 NE SKX 26136 W3/S2 NE SKX 26151 W3/S2 NE

340-350 cm

200-210 cm

230-240 cm

200-210 cm 380-390 cm

270-280 cm 340-350 cm 560-570 cm 290-310 cm 210-220 240-250 290-300 300-310

cm cm cm cm

220-230 210-220 210-220 200-210 290-300

cm cm cm cm cm

210-220 cm 210-220 cm 210-220 cm 210-220 cm 210-220 cm 200-210 cm

W3/S2 NE SKX 26161 W2/S6 NW SKX26810 W2/S4 SW SKX 27949 W3/S2 NE SKX 29130 W3/S3 SE SKX 30002 W4/S2 NE SKX 33749 W4/S2 NE SKX34315 Bone flakes, 3-4 cm long W2/S4 SW SKX 22496 W2/S4 SW SKX 23702 W2/S4 SE SKX 23832 W2/S5 SE SKX 24614 W3/S2 NE SKX26152 SKX 26159/60 W3/S2 NE W2/S5 SW SKX 26174 SKX 29920 W3/S3 NE SKX 29921 W3/S3 NE SKX 29999 W3/S3 SE W2/S4 NW SKX 33063/6 W2/S4 NW SKX 33908 Bone flakes, 4-5 cm long W3/S2 NE SKX 26130/2/4 SKX 34629 W4/S2 SE SKX 37190/2 W3/S3 ~ Bone flakes, 5-6 cm long W2/S4 NW SKX 33058 W2/S4 NW SKX 33067/8 SKX 33070/2 W3/S4 NW W3/S4 SE SKX 35607/8 W5/S3SKX 38389

235

200-210 250-260 350-360 230-240 320-330 310-320 320-330

cm' cm cm cm cm cm cm

190-200 cm 210-220 cm 210-220 cm 230-240 cm 200-210 cm 200-210 cm 280-290 cm 320-330 cm 320-330 cm 320-330 cm 370-380 cm 460-470 cm 210-220 cm 370-380 cm 630-640 cm 410-420 cm 420-430 cm 350-360 cm 550-560 cm 600-650 cm

Category 3. Brown-grey-coloured bones (Fig. 9) Total number of pieces: 47 .::' Bovid size class 1/ Mandible piece SKX27717 W2/S4 SW 320-330 cm Mandible pieee W3/S3 NW SKX 30672 340-350 cm Bovid size c1ass 111 Cranial piece W4/S3600-610 cm SKX 37603 Bovid horneore pieces SKX 19771 W2/S4 NW 170-180 cm SKX 22451 W2/S3 SW 170-180 cm W3/S2 SE SKX 29633 280-290 cm SKX 29661 W3/S2 SE 260-270 cm Miscellaneous pieces W2/S4 SE SKX 23984 220-230 cm SKX 24446 W2/S4 SE 230-240 cm SKX 30012 W3/S3 SE 320-330 cm SKX 30106 W3/S3 NE 330-340 cm W3/S3~ SKX 37200 630-640 cm 230-240 cm SKX 45698 W3/S2 SE Bone flakes, 1-2 cm long W2/S4 SE SKX 23837 210-220 cm SKX 23838 W2/S4 SE 210-220 cm W3/S2 NE 210-220 cm SKX 26124 260-270 cm SKX 29724 W3/S2 NW W3/S3 NE 340-350 cm SKX 30767b 430-440 cm SKX 33637 W3/S2 SW

OCCURRENCE OF SURNT SONES

236

Sone flakes, 2-3 cm long W2/S4 SE SKX 19648 SKX 23693 W2/S4 SE W2IS4 SE SKX 23833 W2/S4 SE SKX 23834 W2/S4 SE SKX 23839 SKX 26135 W3/S2 NE SKX 27542 W2/S5 SE SKX 29101 W3/S2 NE SKX 30032 W3/S2 SE W2/S3SW SKX 30236 SKX31103 W3/S3 NW W3/S3 SE SKX 45693 Sone f1akes, 3-4 cm long W6/S4SKX 10862 SKX 29460 W3/S2 SE SKX 29618 W3/S2 SE W3/S3 SE SKX 30000 SKX 30561 W3/S2SW Sone f1akes, 4-5 cm long W2/S4 NW SKX 19686 W3/S2 NE SKX 26150 W3/S2 NE SKX 29121 SKX 34487 W4/S2 NE W4/S2 NE SKX 34489 SKX 39316 No data Sone f1akes, 5-6 cm long SKX 26158 W3/S2 NE W3/S2 SW SKX 29681 SKX 39511 No data Sone flakes, 6-7 cm long W5/S3SKX 37830 W4/S2 NE SKX 37932

--

160-170 cm 200-210 cm 210-220 cm 210-220 cm 210-220 cm 210-220 cm 380-390 cm 250-260 cm 290-300 cm 310-320 cm 350-360 cm 230-240 cm 0-100 270-280 280-290 320-330 310-320

cm cm cm cm cm

180-190 cm 200-210 cm 230-240 cm 360-370 cm 360-370 cm

200-210 cm 260-270 cm

600-650 cm 330-340 cm

Category 4. Grey-coloured bones (Fig. 10) Total number 01 pisces: 127

AustralopithecU5 robustus Distal 2nd phalanx SKX 35822 Papio sp. Proximal utna píece SKX 31456 Radius shaft SKX 31957 Radius shaft SKX 33939 ProcBv;a sp. lncisor fragment SKX 25425 ef. Cynictis sp. R. mandible plece SKX 23736 Bovid size class I

W3/S3-

450-500 cm

W3/S3SW

380-390 cm

W3/S3NW

440-450 cm

W2/S4 SW

470-480 cm

W3/S6 NE

150-160 em

W2/S4 SE

210-220 em

W3/S2 SE

200-210 cm

W2/S5 NE

330-340 em

W5/S3SW

80-90 em

Mandible pisce

SKX 22949 Mandible píece

SKX 27042 Tooth fragment

SKX 25545

R. proximal ulna

SKX 25426

W3/S6 NE

150-160 cm

W3/S3-

450-500 cm

W4/S3-

490-550 cm

W3/S6-

150-160 cm

W3/S3 SW

400-410 cm

W3/S6 NE

150-160 cm

W2/S4SW

160-170 cm

W3/S2 SE

280-290 cm

W3/S6 NE W3/S3NW W3/S2 SW W3/S3 NW

150-160 cm 290-300 cm 310-320 cm 330-340 cm 630-640 cm

Acetabular piece

SKX 35702 Proximal metacarpal SKX 35467 Terminal phalanx SKX 25427 Bovid stze cíass 11 lIium fragment SKX 31578 Carpal SKX 25431 Carpal SKX 19749 Bovid size class 111 Tooth fragment SKX 29631 Bovid horneare pieces SKX 25424a SKX 28564 SKX 30545 SKX 30872/3 SKX 37196 Miscellaneous preces SKX 23983 SKX 25391 SKX 25424b SKX 25430 SKX 25432 SKX 25464 SKX 27579 SKX 28208 SKX 29267 SKX 30663 SKX 31341 SKX 31884 SKX 31899 SKX 35497 SKX 35499 SKX 35598 SKX 35700 SKX 35704 SKX 35706 SKX 35709 SKX 35712 SKX 35713 SKX 37290 Sone ñakes, 1-2 em long SKX 23428 SKX 23433 SKX 23707 SKX 23708 SKX 23721 SKX 24236 SKX 25416 SKX 25417 SKX 37909 SKX 45802 Sane flakes, 2-3 cm long SKX 23668

W3/S3W2/S4 SE W3/S6 NE

W3/S3W6/S3-

220-230 cm 180-190 cm 150-160 cm 150-160 cm 150-160 cm 160-170cm 370-380 cm 270-280 cm 250-260 cm 340-350 cm 350-360 cm 420-430 cm 410-420 cm 470-550 cm 470-550 cm 590-600 cm 450-500 cm 450-500 cm 450-500 cm 450-500 cm 450-500 cm 450-500 err 550-600 err

W2/S4SW W2/S4SW W2/S4SW W2/S4SW W2/S4 SE W2/S4SW W3/S6 NE W3/S6 NE W3/S4 NE W3/S4 NE

200-210 err 200-210 crr 210-220 err 210-220 err 210-220 err 220-230 err 150-160 err 150-160 err 210-220 err 390-400 err

W2/S4 SE

200-210 err

W3/S6 NE NE NE NE NE NW W3/S2 NW W3/S3 SW W3/S3 SE W3/S3 SE W3/SE SW W4/S4W4/S4W3/S4SW

W3/S6 W3/S6 W3/S6 W2/S5 W3/S5

W3/S3W3/S3W3/S3W3/S3-

W3/S3-

OCCURRENCEOF BURNT BONES

SKX 23749 SKX 23750 SKX 23257 SKX 23959 SKX 24309 SKX 24420 SKX 254'8 SKX 25420 SKX 25422 SKX 25434 SKX 25438 SKX 26758 SKX 27048 SKX 27633 SKX 27653 SKX 28679 SKX29118 SKX 29125 SKX 29128 SKX 29182 SKX 29620 SKX 30030 SKX 31096 SKX 31132 SKX 31455 SKX 33706 SKX 35548 SKX 35665 SKX 35793 SKX 45697 SKX 45784 Bene ñakes. 3-4 cm long SKX 19462 SKX 22780/96 SKX 24452 SKX 25415 SKX 26644 SKX 27115 SKX 29063 SKX 29067 SKX 29120 SKX 29660 SKX 30037 SKX 30115 SKX 30145 SKX 30173 SKX 30191 SKX 30788 SKX 31010 SKX 31539 SKX 32461 SKX 32982 SKX 35545 SKX 35584 SKX 35926a SKX 45789 SKX 45790 Sane flakes, 4-5 cm long SKX 22773 SKX 29071

W2/S4 SW W2/S4 SW W2/S3SW W2/S4 SE W2/S4 SE W2/S4 NE W3/S6 NE W3/S6 NE W3/S6 NE W3/S6 NE W3/S6 SE W2/S6 NW W2/S5 NE W3/S3 SE W2/S4 NW W3/S3 NE W3/S2 NE W3/S2 NE W3/S2 NE W3/S2 SE W3/S2 SE W3/S2 SE W3/S3 NW W3/S3 SE W3/S3 SW W3/S2 NW W3/S4 SE W3/S4 NW

W4/S4W3/S2 SE

W5/S3W2/S5 W3/S2 W2/S4 W3/S6 Wl/S6

W4/S3 W3/S4 W3/S4 W5/S3W4/S2W2/S5 SW

170-180 cm 210-220 cm 230-240 cm 150-160 cm 320-330 cm 300-310 cm 240-250 cm 240-250 cm 230-240 cm 260-270 cm 290-300 cm 300-310 cm 300-310 cm 330-340 cm 300-310 cm 340-350 cm 220-230 cm 370-380 cm 210--220 cm 370-380 cm 570-580 cm 580-590 cm 290-310 cm 580-590 cm 200-210 cm

W3/S2 SE W3/S2 SE

210-220 cm 240-250 cm

W21S5

W3/S2 W3/S2 W3/S2 W3/S2 W3iS2

W2/S3 W31S3

W3/S3 W3/S3

W3/S3 W3!S6

W3/S3 W4/S2

SE SE NE NE NW NW SE SE NE SE SE NW NW SE NE NE NW SE SE NE SE SW

210-220 cm 210-220 cm 200-210 cm 220-230 cm 230-240 cm 230-240 cm 150-160 cm 150-160 cm 150-160 cm 150-160 cm 150-160 cm 220-230 cm 330-340 cm 220-230 cm 310-320 cm 280-290 cm 230-240 cm 230-240 cm 230-240 cm 220-230 cm 280-290 cm 290-300 cm 350-360 cm 350-360 cm 380-390 cm 440-450 cm 570-580 cm 590-600 cm 550-600 cm 230-240 cm 310-350 cm

SKX 29180 SKX 29622 SKX 29626 SKX 29849 SKX 32478 SKX328S5 SKX 35001 SKX35086 SKX35693 SKX 36640 Bone tlake, 5-6 cm long SKX 24676 Sane t1akes, 6-7 cm long SKX 32531 SKX 32984 SKX 33441

237

W3/S2 SE W3/S2 SE W3/S2 SE W3/S3 SE W4/S2 SE W4/S3 NE W4/S2 SE W4/S2 NW W3/S3W5/S2-

220-230 280-290 280-290 300-310 210-220 240-250 440-450 450-460 450-500 590-600

W2/S5 NW

230-240 cm

W4/S2 SE

250-260 cm 370-380 cm 410-420 cm

W4/S3 NE W2/S4 NW

cm cm cm cm cm cm cm cm cm cm

SIZES OF BURNT BONE PIECES FROM MEMBER 3 All 01the 270 fossil bone pieces from Member 3, assumed to haya been burnt, are between 1 and 9 cm in maximum length. As snown in Tabla 1, the great majority 01 the pteces are between 2 and 4 cm in qreatest length.

SPATIAL DISTRIBUTION DF BURNT BONE PIECES WITHIN THE MEMBER 3SEDIMENT As the excavation 01the Member 3 gulley proceeded (Fig. 11), burnt bone pieces were found in 17 one-metre grid squares. as shown in Fig. 12. Numbers 01burntbone pieces per square variad from ane, recordad in five squares. to 79, found in grid square W3/S2. The verncat dtstrlbutlon 01 bumt bone pieces, relative to total bone specimens in the 17 grid squares, is given in Table 2, while detall 01the incidence of burnt bone pieces in each 10 cm-thlck excavation spit per squars is provided in Table 3. In the 17 squaras, burnt banas occur in from one to 23 spits per square. with a mean 01 6,5 spits. While there was generally a fairly aven scatter 01burnt bone pieces throughout the excavated protiles, it was found that tour 10 cm-thick spits within single grid squares yielded more than 10 pieces 01 burnt bone each. These were W2/S4 al 210220 cm (14 pieces); W3/S2 al 210-220 cm (15 preces) and 230-240 cm (13 preces), and W3/S6 al 150-160 cm (15 plecas). 01 these, one distinct cluster 01 eight bumt bone preces was recordad and plotted in the SE quarter of grid square W2/S4, at a depth 01210-220 cm below dalum. The position in which these eight pieces lay, relativa to the abundant chert stones. as well as ta a quartz flake and normal fassil bone pieces, is shown in Fig, 13.

IDENTIFIED ANIMALS REPRESENTED BY THE BURNT SONES Mast ot burnt bone pieces from Member 3 are c1assified as bone ttakes from the Hmb banas of antelope and ather mammals, or as miscellaneaus fragments which cannot be

OCCURRENCE OF SURNT SONES

238

- ,Table1 The maximum lengths ot 270 burnt bone pieces from Swartkrans Merilber 3, in tour categorías, relatad to the degree 01 heating to which they are assumed lo have been subjected. Length in cms

1-2

2-3

3-4

4-5

5-6

Buff Brown Brown-grey Grey

2 6 6 28

16 14 19 45

16 16 8 29

4 6 7 14

5 9 5 5

Total

42

94

89

31

24

8-7

>7

Total

45 51 47 127

2 5

8

2

270

Tabla 2 The distribulion ot 270 burnt bane plecas from Member 3 al Swartkrans, in the profiles of individual cne-metre grid squares, relative to the numbers ot normal banas tound al each level.

Grid squares

Depth (cm)

Wl/S5

0-50 50-100 100-150 150-200 -/399 200-250 1123 250-300 -/9 300-350 350-400 400-450

WllS6

-/129

-/133 -/31 1/96

--151

500-550 550-600 60lH;50 650-850

1/431

11749 V647 3/553 -/376 -/285

--1357

4SG-SOO

Total

WVS3

1/440

W'2JS4

WVS5

4/1207 1/1306 34/2088 4/1654 -/511 1/1298 3/1300 31747 211011 21637 3/453 21181 -/86 -/69

6/3118 4816906 1115642

W2IS6

-/896 V431

W3/S2

V1352

W3/S3

-/233

48/2395 V660 1/106 24/1835 311425 11219 3/1717 27/2668 -1126 -/801 9/1664

V982

W3/S4

511235 91298 --1145 --1819 31381 -1455

119 --128

-1164 41362 --1374 --158 --194 7/329 -/261

W3/S6

W4/S2

181400

--1382

-{lB1

21644

--119

31723 31849 4/1429 1/901

11888

W41S3

-/11 1/186 -/224 --17

2/1039 --1186 --1292 l/-

W5/S2

W5IS3

--112

--1189

-/9

-{121

--16

-/259 -/251 -/103

11-115 --133 -1242 V56 1/226 -1106

W4IS4

-/23

-/26 -19

--196 --1632

-/525 1/687

V1fl9 --1203 --1110

4/1778 79110064 5819983 1'lI1805 181600 15fl61O 513157

3/448

-/126

1/660

-/499 -/635

--171

-fl -1208

W6IS3

1/--1337 -/457

-/129 -/19

-/37

--117 -/9 -/5 -/34

--1798

--1682 V1715

1/-/56

113035

613361

1/962

1/563

W6IS4

1/148

Overall totals: 270159488

assigned to species. The pieces that have allowed more precise identification come from the following animals: AustraJopithecus robustus, one individual represented by a calcined distal second phalanx, SKX 35822. Papio sp., one individual representad by a calcined proximal ulna piece, SKX 31456, and two calcined radius shaft pieces, SKX 31957,33939.

Procavia sp., two Individuats represented by a calcined incisor fragmen1, SKX 25425, and a charred shaft piece, SKX 21538.

Equus cf. capensis, one individual representad by a charred right astragalus fraqment, SKX 26524. cf. Cynictis sp., one individual representad by a calcined right mandible piece, SKX 23736. ct. Phasianidae (a guineafowl or francolin), one individual, represented by a charred left distal coracoid piece, SKX 24369.

In addition to these, ante\ope of Bovid size classes 1-111 clearly contributed significantly to the burnt bane assemblage, though the numbers of individuals mvotved cannot be established. A visual impression of anirnals invotved in this assemblage of 270 pieces 01 burnt bone is given in Fig. 14.

DISCUSSION The abrupt appearance of burnt bones in the Member 3 assemblage requires an explanation. Over 52000 pieees of fossil bone from the Lower Bank of Member 1 were examined critically for the effects ot heating, but none was found. The same is true far the assemblage from the Hanging Remnant 01 Member 1 described earlier (Brain, 1981). In Ihe case 01 1he slightly later Member 2 sediment, aver 34 000 pieces of bane were found to inelude only two specimens that appeared to haya been heated. They were a buff-coloured equid lateral metapodial, SKX 2372, found on 18 June 1979 in grid square ES/Sl al a depth of 50-100 cm, and a charred bone f1ake. SKX 34102, lound on 4 January 1984 in grid square E5/S5 at a depth 01 160-170 cm. It is to be expected that occasional lightning-induced fires would nave swept through the grassland at the cave's entrance and could nave ignited loqs and branches Iying there, which in turn may have heated pieces of bone on the cave floor. This could well have happened in the case 01 the two pieces mentioned from Member 2 and in the case of a few of the burnt banes from Member 3. Butthe difference in incidence

239

OCCURRENCE OF BURNT SONES

Table3

The cetañed distribution of 270 burnt bone peces Irom Member 3, in the profiles of 17 ene-mene grid squares. according lo their presence in 1O-cm-thick excavation spits.

Grid squares Oepth 01 sptts (cm) W1/S5 Wl/S6 W2JS3 W2JS4 W2JS5 W2JS6 W3IS2 W3IS3 W3IS4 W3/S6 W4/S2 W4/S3 W4/S4 W5/S2 W5/S3 W6IS3 WS6IS4

0-50 150-100 100-110 110-120 120-130 130-140 140-150 150-160 160-170 170-180 180-190 190-200 200-210 210-220 220-230 230-240 240-250 250-260 260-270 270-280 280-290 290-300 300-310 310-320 320-330 330-340 340-350 350-360 360-370 370-380 380-390 390--400 400--410 410--420 420--430 430--440 440--450 450-460 460--470 470--480 480--490 490-500 500-510 510-520 520-530 530-540 540-550 550-560 560-570

15 2

2 1

1 1 6 14 7 6

2 2 2

2 1 2

2 8 15 3 13 9 2 6 5 7 4 1 2 1 1

2

1 1 2

1

5 6 8 3 6 5

2 1 1

1 2

1

2

2 2 1 1 1 1 4 2 2 1

2 1

2

1 1 2 1 2

570-580

580-590 590--600 600-650

2

3

Total: burnt

6

48

11

4

79

58

12

18

15

5

3

6

Number of spits with burnt 1 bones: 1 5

17

8

3

15

23

10

3

11

4

2

5

bones:

1

270

111

240

OCCURRENCE OF BURNT SONES

+-- W 5

6

4

2

3 .1 .

~

'. 3113~

~ 7610

6 3361

5 3157

1

.'

X-

1,

( ('

1 962

II

1

148

I

3 448

~

.

?!

79 10064

' O

11"

1---o" .

m

3

) 12 48 1805 , 6906

4

11 : \ 1 5642 ~ 431

.'~

31I

2

1~1~

58 9983

~ 1-I

~ ~.

18 600

!

:\ \,

4 1778

1, 440

"",

7

8

\

Fig.12 A plan of the Member 3 gulley al Swartkrans, showing 17 grid squares

in whích burnt bones were found. In each square, Ihe upper figure indicates [he number 01 burnt bones, while the lower one reflecls Ihe total number 01 bone pieces recovered from the pro/ile of tha! square. Fig_ 11

Excavation 01 the Member 3 gulley in progress at Swartkrans during 1984. Positional control 01 Ihe excavation was made possible by hanging plumb-bobs and measuring tapes from Ihe corners 01 each grid square delined by the overhead slructure.

01 burnl bones between the earlier Member 2 and the Jater Member 3 deposits is striking. These two Members are separated by an interval during which lhe Member 3 gulley was eroded inlo older Member 1 and 2 sedimenls; such an inlerval may have lasted as long as 50 000 years. As described above, the burnt bones are confined laterally by lhe extent of the Member 3 gulley, while they are distributed vertically through a six-metre-deep profile. In one grid square, W3/S3, burnt bones occur in 23 excavation spils, each 10 cm deep, indicating ,hat the bones were heated in frequently recurríng fires during the depositíon period of this straligraphic unít, whjch may have spanned several thousand years. The spatial distribution of the búrnt bones suggests strongly that the fires occurred within the confines 01 the gulley, which was beneath a dolomite roof and had an inclined entrance towards the southeast. That Ihis gulley was used as a shelter by hominids is confirmed by the presence throughout the excavated prolile of artilicially fractured stones and bone tools, while the presence of the remains of nine indívíduals of Australopithecus robustus, details ot wh ich are given in Chapter 3 01 this volume, indicales that this hominid was still living in the area at thal time. Figure 15 presents the occurrence of burnt bones wilhin the

excavated protile of grid square W3/S3, relative to the total number ot bone pieces, fractured slones, bone tools, cutmarked bone and hominid remains in each 01 the 10 cm-deep excavation spits. It will be seen that hominid activity is reflected throughoutthe deposition period of lhis Member 3 prolile, while in thal partof the profile where bone pieces are most abundant, so too are burnt bone pieces. The ínterpretation suggested here is thal hominids were tending fires wilhin the Member 3 gulley as this space filled with sediment and that many pieces of bone, presumably hominid food remains, were discarded ínto the ash of the tires. 1\ is of interest that two bone tools, both made on horncores and numbered SKX 26235 and 36485, show signs 01 healing. These tools were presumably Iying on the cave floor close to one of the fires, and were incidental1y heated, The same may be true of the hominid phalanx SKX 35822, which appears lo have been calcined. The presence of artificial cut marks on Member 3 bones suggests that hominíds were feeding on meal while sitting around their Jires. Mention has already been made of the sudden appearance of burnt bones in Member 3, after their virtual absence in the earlier Members 1 and 2. The same is true of the appearance of cut marks on Member 3 bones; at present, 13 specimens with artificial cul marks on them have been recorded from Member 3, while two show chop marks, as further documented in Chapter 13 of this volume. By contrast, only one specimen wilh possible cut marks on it has been recorded from the earlier Members: a horncore piece

OCCURRENCE Of SURNT SONES

....

Fig.13

241

"

A cluster of burnt bone fragments found in che SE quarter of grid square W2/S4 at a depth ct 21 0-220 cm below datum. Also shown In lhis plot are the pcsitions of chert stones, normal bone pieces and a quartz ñake .

.¿ P.. . ...•

,

'<;,'>

e

'1

•

mJ

BURNT eONE

@J

CHEAT STONE

~OR"'Al

11II

Ol.JARTZ Fl..AKE

80NE

Fig.14

A chart 01the identifiable animals represented by the 270 pieces ot bumt bone from Mernber 3 at Swartkrane. Figures indicate the mínimum number of indiviouals present, although it ts not possible to estímate the numbers 01 anteiope involved.

A. robu$luS

d. GuirUiO' Fowl

1

Vjverrid ct, c.Yl!iii.!.!!.

1

~~ BQYid I

from the Lower Bank ot Member 1. It therefore appears that, with the management of fire, hominids started the habit of eating meat there as well. This tapie is further discussed

BOl/id

n

." Bovid III

in Chapter 13, which ateo explores the implications 01 lire management in hominid evolution.

ACKNOWLEDGEMENTS

1 am very grateful to Virginia Watson for help wíth the preparation 01 figures and tables in this Chapter. She and Tim Braln helped with the management 01 experimental tires; Rosemary Newman prepared and cataíoqueo me Member 3 bones and allowed her pottery kiln to be extensively modified

for the experimental heating 01 bones, while Tim and Laura Brain ground thin sections of fossil bones. The collaboration 01 Andrew Sillen on the chemical characterization 01 burnt bones is very much appreciated.

OCCURRENCE OF BURNT BONES

242

W3S3

BURNT BONE n

~

TOTAL BONE PIECES n = 9983

56 10

ISO

,O .

soo

B

=

200

2SO

3SO

J: f-

400

~

c..

O

-

20

'--'--'---'---'

BONE TOOLS n O

=9

n =

2

~

•

HOMINID REMAINS

n =

1 1

L-.-.o.

11

=

~

= =

E3

=

W

o

ClIT·MARKED BONE

'ª= =

300

E

FRACTURED •• ~ONE n ='207

4SO

=

soo sso 600

650

Fig.15

-

=

=

=

Protiles through the exca'/ated depth 01 grid square W3/S3 in Swartkrans Member 3 sediment. The occurrence 01 burnt bone pieces relative to total bone pieces, artificially Iractured stones, bone tools, cut-marked bone and Australopithecus robustus remains is indicated.

REFERENCES

BELLOMO, R. V. and HARRIS, J. W. K., 1990. Preliminary report 01 actualistic studies of tire within the Virunga National Park, Zaire: towards an understanding 01 archaeological occurrences. Virginia Museum of Naturaf History Memoir 1: 317-338. BRAIN, C. K., 1981. The hunters or the hunted? An introduction to African cave taphonomy. The University 01 Chicago Press, Chicago. BRAIN, C. K. and SILLEN, A., 1988. Evidence Irom the Swartkrans cave lor the earliest use 01 lire. Nature 336: 464-466.

GILCHRIST, R. and MYTUM, C., 1986. Experimental archaeology and burnt animal bone Irom archaeological sites. Circaea 4(1): 29-38. SHIPMAN, P., FOSTER, G. and SCHOENINGER, M., 1984, Burnt bones and teeth: an experimental study 01 color, morphology, crystal structure and shrinkage. Journal of Archaeological Science 11 (4): 307-325.

Chapter 11

Chemical Characterization of Burnt Bones from Swartkrans Andrew Sillen' and Thomas Hoering' 'Archaeometry Laboratory. Department ot Archaeofogy, Universily o/ Cape Town, Rondebosch, 7700 South Afrlca 2Geophysical Laboratary, Carnegie fnstitution ot Washington, 5251 Bmad Branch Road NW. Washington O.e. 20015-1305. U S.A.

Some tossu banes from Swartkrans Member 3 were otecseneo due ro the presence of carbcn-containlnq char, indicating that they were burnt. To determine the temperatures to which the Swartkrans bones were heated. modern bone was heated to a ranga of temperaturas; it was found tnat cnar extsts mainly in banas neated between 300 and 500 "C. Modern and Swartkrans bcne chars were further examinad by CHN elementalanalysls. analytical pyrolysis. GC~MS and HPLC. In additionto 'free carbón,'Swartxraos bcne charswere feundte centainmeasurabfe nitrogen (mainly in lhe torrn of arnmonium ion), and diagenetical1y altered animal tata probably derivad from the

animal whose borreswere burnt. lmplicaucnster pataeodietary and ñre-tapboncmy stud'es are discussed.

INTRODUCTIDN The identification of burning in modern bones is a simple matter, beeause the blackening caused by cna- is readñy observable with lhe naked eye. Using a visual approach ís unsatistactory for tossus, however, sinee they may be blackened for other reasons, chiefly the incorporalion of diagenetic metate. In the South African context for example, purportedly burnt bones from Makapansgat have been shown to be actually blackened due to the presence of oxides of manganese and iron (Oakley, 1954; 1955a,b; 1956). The demonstration that Makapansgal bones were blackened by manganese was usad to disprove Raymond Oart's assertion that austratopíthecines found at the cave were capable of making fire. Where fossil specirnens are suspected of being bumt, a demonstration using sorne analylical means is therefore necessary. A number of such techniques have been proposed and applied sinee al least 1925, when blackening in burnt bones was shown to be due to insoluble particles containing carbon (Dart, 1925; Oakley, 1955). Organic resldues do not always survtve the fossifization process, however; therefore most recent research has explored histological changes, and lhose changes that occur in the inorganic phase of bone due to heat. Histological study has shown that. when cortical bona is heated above 200 oC, the canalicuH beeome increasingly prominent, whiJe at higher temperatures the circumferential lamellae surrounding the Haversian canals become coarsely granular in structure and osteocytic lacunae begin to disap~ pear (Forbes, 1941). Increased prominence of circumferential lamellae, foHowed by structural disintegration at higher tem~ peratures, are characteristics of Swartkrans burnt fossils and are described in detad in Chapter 10, this volume. With regard to the mineralogy of the inorganic phase, at reasenably high temperatures (over 700 oC), apatite is trans~ formed into whitlockite, and this is readily observed using X·ray

diffractometry (XRD) (Posner, personal communication). It is also possible to identify crystallographic changas in apatite at temperatures below 800 "C using XRD (Shipman el al" 1984). Unfortunately, crystallographic changes in the 200-700 "C regíon amount to increases in crystallinity, which are very similar to those that occur in normal Iossllization. Therefore lt ls difficult to distinguish diagenetic crystal·growth trom that due to burning at temperatures likely to have been encountered in camp-ares. Since it was suspected that Swartkrans tossils were blackened due to the preaence of char, it was unnecessary to resort to crystallography of the inorganic phase in arder to prave that the bones were indeed burnt. Instead it was possible to concentrate en recovery and examlnation o: the chars themselves. The purpose of the present study is first to explore in an experimental context the temperatures at which char is encountered in burnt bones. With these data in hand, a second objective is to document the presence of char in Swartkrans tossils, and to demonstrate by cornparison with the experimental series. the temperatures to which Swartkrans bones had been heated in antiquity. During the course ot this Invesnqation, it became clear that the char of Swartkrans fossils contains organic moJecules, and not merely 'free carbón' as described by Oakley in early investigations (1954; 1955 a,b ). Therefore we examined the composition of char itself, using CHN microanalysis, analytical pyrolysis, gas chromatography/mass spectrometry (GC/MS), and high-performance liquid chromatography (HPLC); the implications 01 lhese analyses for future research are explored. The archaeological context of the finds, and tlle bellavioural implications tor early hominids have been discussed previously (Brain and Sillen, 1988; Sil/en and Brain, 1989). Carbon in fresh bone resides in a number of different phases which must be removed or olherwise accounted for when examining char carbon. In the inorganic phase, carbon is present in the form of the carbonate ion, C03 , at the surface

CHEMICAL CHARACTERIZATION OF 8URNT BONE

244

Table 1

Table 2

Replicate recovery data on 300 vO and 400 "C chars.

Carbon, hydrogen and nitrogen in experimental cnars.

Experimental series: Sampfe resídue % weigllt

C

H

1,4 0,8 2,8 2,4

72,0 74,0 57,4 58,8

8,9 9,3 4,1 3,5 2,9

400°C

2,2

(rpl. a) (rpt. b) 600 ~C

2,6 3,4 1,4 0,3 0,0

58,9 58,6 51,3 52,4

unheated 200 ce

300

~C

(rpt. a) (rpt. b)

700 'C

BOO°C

65,8

N

C:N

300 -c % yield; 2,8 °/0 400 -c % yield: 2,2 %

11,3

5,1

12,0

4,9 4,9 4,2

13,4

13,4

3,4 3,6 4,4

13,3

11,6 8,6

2,6 % 3,4 %

Aeplicate CHN anavsee on ene 300 -c charo

trace trace

4,1

Mean:= 2.67 % Mean := 2.7 % S.O.:= 0,61

2,4 %

4,2 4,4 6,0

1

e

H

N

C:N

57,4

4,1 3,5 2,5

11,3 12,0

5,1 4.9

13,4

4,9

2 58,8 65,8 3 Mean C:N = 5,0

lnsutñclent reecue for enejysrs InsuffiCient residue for analysis

ot apatite crystalütes. and as structural inorganic carbonate, which replaces pnosphate in the apaute crysta! Iattíce (Biltz and Pellegrino, 1977). In the organic phase, carbón ls present both ln protelns (malnly callagen), and in other organic molecuJes, such as lipids, Fosslls may have additional inor-

ganic carbonate, in the forms of adsorbed C03 on biogenic apatite. recrystallized camonate-apante, and calci1e. Fossils of the age of Swartkrans material (c. 1,0 m.y. a.p.) have generally lost mast 01 the organic phase. except wheTe the organic phase has been converted to insoluble molecules by

incomplete combustión. Char ítselt occurs when the organic components of bone are incompletely combuated. (Wl1en combustion is completed, all of the carbón in the organic molecules ts oxidized to COz gas.) Because chars consist chiefly of carbón. they are insoluble in both water and non-aaueous sotvents. This insoluble property is one reason that burnt bones survive so well at rnany archaeological sites.

PART 1: CHAR RECOVERY The objective of the initial study was to previda background informatian on the amaunt of char which forms in bones heatad to specific temperaturas under cantrolled conditions. In order to previde a series ot bone samples that had been heatad to known temperatures, a fresh hartebeest (Alcelaphus busefaphus) radius was cleaned of adherent tissue and cut trans· versely into saven segments. Each segment was individually heated in air in an electric tumaee to a specific temperature, maintained there for 30 minutes and then slowly cooled. Individual segments reached temperatures of 200, 300, 400, 500, 600, 700 and 800 oC respeclively. As reported by previous workers, changas in the gross appearance of the bone are very striking as Dne progresses through the temperature series. At 200 oC the bone appears unaffected but at 300 oC lts colour has changed to a glistening blackish brown. Sorne of this darkening has laded at 500 oC and lightens progressively through shades of grey up to 600 oC. After exposure to temperatures of 700 or eoo oC, the bone is chalky white and extremely light and brittle.

Melhods In tne taboratory, samples were mi/led using a Spex model 6700 freezer-mlñ. One gram samples of each powder were treated with 15 % EDTA, a chelating agent which decalcifíes the inorqanic phase. After decalciñcation, the preparations were centrifugad, and the supernatant removed. At this point (dependíng upon the temperatures reached), the residues may conté-n the insoluble organic pnase, lncludinq proteins and lipids, and charo"rn¡e residue was then trsateo with 6N Hel at 100 "C far two days, in arder to hydrolyze any remaining prota¡n. Tnis preparation was centrifuged and the supernatant, containing the solubilized protein, removed. The insoluble residue, now containinq only lipids and char, was washed, recovered, freeze-dried, weighed, and analysed tor C, H, and N using a Heraeus CHN-Rapid thermat-conducuvtty detector. Results The percent contribution of char to íhe total weight of experimentally heated bones is presented in Table 1. As expected, no cnar rs found in unheated bones, and bonas heatad up to 200 "C. Bones haated to between 300 and 400 "O contain the mast char (2,45-3,40 %); the char content decreases sharply al 500 'C. Replicate recovery data on lhe 300 and 400 oC chars are presentad in Table 2. Although char is unformed, lipids are present in the unheated and 200 specimens. They are clearly identified both by the absence of nitrogen and because they are soluble in chlorolorm melhanol (1 :1). The quantity of char is directly related to observable blackening, and it is of interest that relatively small contribu~ tions of char (1-3 %) cause considerable blackening. The C:N ratio 01chars is in lhe regíon 014 to 5 (Tabie 1). This is considerably higher than that seen in coJlagen (2,9-3,6) (DeNiro, 1985), and is due to 1) the inclusion in the charoflipid carbon, and 2) the destruction of N·containing molecules. However, what is most interesting about these values is that nitrogen is present in char at all. The data thus demonstrate that bone chars do not on\y contain 'free carbon' (ef. Oakley, 1954. 1955), but may also contain surviving organic molecules, and/oT other molecules that are the products of intermediate combustion. Swartkrans fossils W6re examined using the same techniques. While ccllagen does not survive in Swartkrans fossils, the hydrolysis step was still ¡ncluded because bacterial

ce

CHEMICAL CHARACTERIZATION OF BURNT SONE

Table 3 CHN analyses 01 Swartkrans foss¡l residues.

residue Sampre SK 26157 SK Pool

Table 4 Total elemental carbón. hydrogen and nitrogen in 11ght· and dark· coccreo Swartkrans tossss.

% weight

e

H

N

C:N

Llqht-ccloured bones

1,3 1,6

25,7 33,8

2,5 2,5

3,2 2,9

8,0 11,7

SK 292B1 SK 3007. SK 30205 SK 26952 SK 29756 Mean = 1,0·, n = 5

Light-coloured Swartkrans Iossüs SK 28952

SK 30074 SK 29756

245

No resldue No residue No residue

Sumt bones

proteins may be presant in fossils. No char Is presant in light-coloured specimens from Swartkrans {Table 3). That ts, no residue occurs when the inorganic phase is removed. On the other hand, btackened tossüs were found to contain 1,31,6 % char, which preves that they were burnt. Swartkrans chars are more depleted in nitroqen than banes from the experimental series (Table 3); this ls conslstent wrth elevated C:N reported for organic residues from other burnt archaeological banas (Debliro, 1985), and is al 50 due to the natural diagenetic toes of nitragen with time. Given the observation from the experimental series that char is present mainly in banas haated 300-500 "C. it is mostlikely that 'blackened' Swartkrans banas were heatad to this ternperature. Sorne Swartkrans specirnens have a grey ashy colour, indicannq that they were heatsd to relatively high temperatures; not surprisingly, one such specimen examinad in íhis study ccntained relatively low char content (Tabla 4). To gauge the total contribution 01 char carbon to the total weight of Swartkrans fossils, an alternatíve method was aiso used. Fossils from Member 111 were mllled using the Spex treezer mil/o Fossil bone mineral contams a certam amount cr diagenetic carbon in the form 01 highly soluble inorganic caro bonate, which resides at the surface of apatite crystallites. This inorganic carbonate was removed by treating the powdered bone samples overnight in 1 M acetre acid (Sullivan and Krueger, 1981; Lee Thorp and Van der Merwe, 1987), The insoluble fraclion that remains consists of apatite, char, and any remaining organic molecules. This fraction was washed in deionized water, treeze-drreo. weighed, and analysed ter carbon, hydrogen and nltroqen Results are presented in Table 4. (It should be cautioned that the weightpercentages do not pertain to whole speclmens, but whole specimens which have been epecially treated to remove all but structural apatite and char.) AH dark·coloured bones contajn over 2 % e, while the light-caloured bonas contain approximately 1 %. There is no overlap between the two distributions, and the difference is significant (P = < 0,0001). Dark·coloured banas also contain residual nitrogen (0,20,3 %): this element is not detectable in the light~coloured banes. The value af 1 % e in light~coloured banes isduetothatpresent as structural carbonate in bone and fossil apatite crystals (Biltz and Pellegrina, 1977), as mentíoned earlier. The additional 1-2 % carbon in dark-coloured bones is due to char, and matches the results obtained by ;solatin9 and weighing the char.

e

H

N

1,1

0,2 0,2 0,2 0,2 0,3

0,1

1.2 1,2 0,7 1,0

e

H

N

SK 26130 SK 26131 SK 26136

2,2 3,1 2,8

0,3 0,5 0,3

0,2 0,3 0,3

Chocolate-brown

SK 26151

2,3

0,3

0,2

Ohocolate-brown

SK 26126 Mean» 2,5·, n = 5

2,3

0,4

0,2

·significant difference using Stucent's r-test: P SK 35545

0,5

0,2

0,1

Oolcur descriptlon

Cnocolate-brown Cnocclate-brown

< 0,0001

Grey/ashen

SK burnt bone (control) SK 26126 2,3 0,4 0,2 Same specimen atter heating to 800 -c 1,3 0,2 0,1

PART 11: CHAR CHEMISTRY 8ecause C:N analysls of Swartkrans chars indícated that sorne organic rnolecuíes may remain, wa usad analytical pyrolysis, a technique whereby sol id organic matter is heatad rapidly in a helium cardar gas stream to a temperatura where the sample fragments into volatile molecules. These products are swept onto a gas chromatographic column where they are separated and then identified by mass spectrometry. Sorne of the chemical structure of the unknown salid material is preserved in tne fragments and valuable clues about the nature of the sample can be obtained (MeuzeJaar et aJ., 1982). Analytical pyrolysis has been successful in a wide ranga of applicatians for characterizing intractable, insoluble orqanic matter and rs particularly usefu! when pyrolytic 'fingerprints' from an unknown can be compared with thermal deccrnposítion products of known material (Irwin, 1982). A fundamental difficulty ls the interpretation of the amount of data generated by the experiment. A cornplex mixture of organic moJecules ís produced by pyrolysis and although the molecules can be sorted out by automation, the problem is one of determining which ofthe products gives significan! informa· tion on the basis of the ¡nitial material. Figure 1, for exampJe, is the pyrolysis chromatogram (pyrogram) of fresh cow cortical bone. Each of the numerous peaks represents a pyrolysls product of the numerous organic constituents of bone. In general, those peaks that appear early in the chromato~ gram (0-15 minutes) are the pyrolysis products of amino acids. The am;no acids proline and hydroxyproline are majar constltuents of callagen. Pyrrole, methyl pyrrole, and pyridine are major and easily recognized products of bone pyrolysis that occur from dehydrogenation, dehydration, decarboxyla·

CHEMICAL CHARACTERIZATION OF SURNT SONE

246

tíon and reduction during the pyrolyis experiment (Hoering and Si/len, 1990) (Fig. 2). Pyrrole, wilh a mass number of 67, and methyl pyrrole, with a mass numberof 81, are easily detectable using the selected ion-monitoring mode of the mass spectrometer. In this moda, selected masses are repeatedly and sequentially measured for 250 rnilliseconds each. The peaks lhat appear late in the chromatogram shown in Fig. 1 (25-40 minutes) are, in general. the pyrotysis products 01 latty actds. As their ammonium salts are heated, these molecules form alkyl nitriles, distinctive molecules in which a nitrogen atom is linked te a terminal carboo atom by a triple bond (Fig. 3). In mass spectroscopy, these molacules haya a dístinctive signatura-ion with a mass of 97 (Budzikiewicz et al.. 1967), which can al so be seen using the selective ion-monhortnq mode of the

.-.•

,

~ Fig.1

I~

••• s

.1

~~ JI¡JII

.-

17.3e

.,. 1.1

w

\¡,,¡,

-

"

1

-

.,......

~,-

Pyrogram ct fresh cow bone. Pyrclysls or 5 mHligrams 01the sample was canee out for 20 secenes at 750 "C in a CDS Model120 pyroprobe. The analysis of the pyrolysata was performed with a Finnigan 4500 GC-MS with an INCOS 2300 Data System. The GC coJumn was a 60 m by 0,025 mm tused suica cepnuery coturnn with bcnded SE·54. Column temperature was programmed from 60 to 300 "C at arate of 6 degrees per minute.

Ha

o

\;/

N

I!i

C, mass spectrometer. The <; H-C·H H-<¡:-H identification of these com, Proline Hydroxyprolin" H-C-H H-C-H pounds using the selected , H-C¡;.H H'9- H ion-monltoring mode is H-C-H H-C, H , shown in Fig. 4. H_C·H H-C·H , , Using the same experiH.<¡.H H·C·H , mental conditions, a pyroH-C-H H-e¡:-H , Iysis experiment was conH-C-H H-C·H , , ducted on a Swartkrans Pyrrole 2 Methyl Pyrrole H-<¡:.H H-?-H burnt bone; the chromatoH·C-H H-C·H , , gram is presentad in Fig. 5. H-C-H H-C·H , There ís no evidence for the H-C_H H·C-H , presence of arnlno actos. H-C_H H-C-H , Virtualiy ali 01 the peaks ocH-C-H H·C-H , cur in the region of 25 to 50 H-C·H H-C-H minutes, and selected ion Pyrldlne Toluene PlIlmltlc acld PentadecanitrUe monitoring confirmed the Fig.2 Fig.3 presence of nitriles. This reMolecular structures of sorne of the organic Molecular structure of palrnitic acid and tts pyrolysis sult, however, was not conproduct in bone, pentadecanltrñe. Analagous tcrmaticn molecules produced by the anaiytical pyrosistently reproducible, lysis of bone arnino acids. of octacecal nitnla from steanc acid also occurs. elther because the sample was inhomogeneous. beof fatty acids using gas chromatograph-mass spectrometry cause of the inherent complexity 01 pyrolysis experfments, or (GC-MS). Two hundred mg 01 Swartkrans burnl bone was because 01 chanca contamination. extracted wlth a chloroform:methanol mixture (1:1 per volTo determine if the nitriles were pre-existinq components of urne). The extract was concentrated, and divided into two the Swartkrans char, or jf they were tormed from fatty acids parts, The first part was injected directly mtc the GelMS during pyrolysis, we examined the burnt bane for the presence

247

CHEMICAL CHARACTERIZATION OF SURNT SONE

system. The second was treated with boron trifluoride-methanol to convert any free fatty acids into their methyl esters.

Figure 6 is a gas chrornato-

~

.

SI&<. gram al the total extract (Iower trace) and a rness chromatogram al mass 57 (upper trace). The ion al this mass (C 4H g" ) is characteristic 01 saturated hydrocarbans. Several pertinent features are evident. No alkyl nitriles were detectad. Thus, this class of compound observad in Fig. 5 has probably bes n generated by pyrolysís 01ammonium salts 01fatty '==1 acids. The large peak centred al sean number 2860 IS due to di-isoctyl phthalate, a ubiquitous contaminan! from plastícízers. The regular series 01 peaks, laId JJt.N.,.u.!. ~ , belled 1 lo 11 and identified in the l~ I..:B. 17.311 26.IS captíon 01 Fig. 6, are due to norFig.4 mal paraffinic hydrocarbons with Mass chromatogram and total chromatogram ot pyrolysis of tresh cow bone trom the same sample a predominance ot even-carbonshown in Fig. 1. A: mase chromatogram due to pyrrole (m/e 67); 8: mass chromatogram due to methyl numbered molecules. The peaks pyrrole (m/e 81); C: mase chromatogram due to nitrite siqnature-icns (m/e 97); D: total cnromatogram. preceding those of odc-caroonnumber hydrocarbons (2,4,6, and 8) are due to olefinic hydrocarbons. The source of this suite of compounds is unknown, but does not correspond ta comman cantarnínants such as paraffin wax. Figure 7 is a total cnromatoA gram (Iower lrace) al the methyl estar traction from the extraet of Swartkrans burnt borre, plus _. _. .. -mass chromatograms at masses e» 57 (upper trace) and 74 (middle trace). lons at mass 74 are specific for methyl esters of saturated fatty acids. No iso-octyl (al phthalate contamination was obB served. The smooth distribution of peaks in the mass 57 chromatagram corresponda to high moI I , 11I II ~. lecular weight, normal hydrocarl',,,. •• .os bons eharacteristic of paraffin wax contamination. The two Fig.5 large peaks in the mass 74 chroPyrogram ot Swartkrans burnt bone. Pyrolysis of 15 mg. The conditionswere identical te those described matagram are due to methyl for Fig. 1. A: Mass chromatogram at m/e 97 corresponding to signature ion 01 nitnles: B: total hexadecanoate and methyl oechromatogram. The mass epectra of peaks A and 8 correspond to that 01 pentadecyl nitrñe (C 16H 3 1N) tadecanaate that have been deand heptadecyl nitrile (C18H~sN) respectively. rived from palmitic and stearic aeid, common tatty acres ot living to heating. Common contamination due to soaps, oü from ski n organisms. Very few unsaturated fatty acids (e.g., oleic acid) during handling, or dandruff flakes have abundant oleie aeid. were identified in spite of the ract that they are major constituNo branched-chain fatty acids, a common constituent of bacents of animal fats. Unsaturated fatty acids are unstable in teria/ cel! waus, were detected, precluding any major contarnigeofogicaf environments and their aosence in the case ot the nation from mfcrobes. In summary, the chromatograms are Swartkrans bonos could be due to either natural diagenesis or

I

..

1

h

-

~.N

.

I

,

-

,- I

-- J ~

I

-".

CHEMICAL CHARACTERIZATION OF SURNT SONE

248

consistent with the presence of diagenetically altered animal fats, in alllikelihood ariginating from the animal whose bones were burned. Although pyrolysis did not indicate the presence of amino acids in Swartkrans chars, the C:N ratio and the formation of nitriles during pyrolysis indicate that nitrogen is present in some formo To further address this, Swartkrans burnt bone was examined using high·performance liquio chromatography (HPlC). The resulls confirm lhe indication from analytical pyrolysis that absolute amounts of amino acres are extremely low. Only trace amounts of amino actos were detectad. in lhe region 0120-150 picomoles/mg. In contrast, fresh bone contains over 1000 times this amount, approximately 2500 nanomoles/mg. Two observations suggest that the trace levels of amino acids present are due to contamination. First, tne relative levels of amino acids were not representative of bone collagen. Forexample, glycine accounted for only 10 % of the amino acids present; it accounts for 30 0/0 in collagen. Second, aspartic acid showed no racemization, indicating tnat it is probably modern in origino The highesl peak seen in the HPLC experiment was for the ammonium ion. NH/-. Previous studies nave shown that ammonium is elevated in burnt calcified tissues, including bone and ostrich eggshell (Brooks el al., 1990). Therefore il is probable that lhe arnmonium ion Is a product of the heating of collagen, and represents the majar source of nitrogen in these chars. DISCUSSION

7

s

11

9

3

e

•

10

-~,.

Fig.6

"""

Total ion chromatogram ano mase 57 chromatogram of total extract ot Swartkrans burnt bone. GC conditlons were the same as in Fig. 6. ldentification of the numbered peaks is as lollows: n-ccsane (1); C2oH.42 (2) n-henelcosane: C2, H•• (3) n-docosane; Cn H46 (4) n-tricosane: C23H.e (5) n-tetraccsane: Casl-lec (6) n-pentacosane; C:n H s6 (7) n-hexaccsane: C 26Hs4{S) n-heptacosane: C 21Hs6 (9) n-octacosane: C28H58 (1O)n-nonacosane: C 29H61l (11) n-tricosane: C30H 62 .

.',

A

.. .

J~L h -

.8._ "".." .1

.l

¡ l ...

,."".

Stearie Palmib..

B

en o.,; c" eso 02' Experimental work with the hartebeest el' c" I radius has shown that insoluble char can be recovered from banas neated lo temperaturas batwean 300 and 500 oC. Sinca char , can also be recovered from 'blackened' e Swartkrans fossils it follows that these tossüs were heatec to similar temperaturas in n l' antiquity, Histologícal investigations showing ."'~ TU. crear alterations in structure at high temperaturas complement these data; using the two Flg.7 techniques together it is possibJe to deterMass ehromatograms and total ehromatogram ot methyl ester traction-extract 01 mine whether banas haya been heated to the Swartkrans burntbone. A: mass chromatogram due te C4H7· ion, cnaracterfstic of saturated hydrocarbons; B: rnass ehromatogram at m/e 74 due to ion C 3H e0 2· , speciñc ter methyl ranges 0-200, 300-500, and over 500 -c, estere of saturated latty acids: C: total chromatogram. Apart from proving that cenain Swartkrans fossits were burnt, the data haya certaín impücatlons for the direction 01 future reThe composition of bone chars deserve more detailed attensearch. Foremost among them is the observation that char tion, since it might be possible to develop a more precise does not consist only of 'free carbón.' but rather contains fatty means 01 identitying the temperatures to which bones have acids, which are consistent with fats derived from the animal been healed. At present, the existence ot cnars indicates that whose banas were burnt.

=

..

~

.

-,

.•1 11. 1

1,1 I

--

CHEMICAL CHARACTERIZATION OF BURNT BONE

bones have been healed lo belween 300 and 500 "O: however, organic analyses 01the chars themselves could Improve our resolution considerably. Two chernical markers that might be further examinad in this regard are the thermal degradation 01amino acids and the production 01ammonlum ion, using HPLC (Brooks et et., 1990), and lhe lhermal degradation 01 fatty acíds. Efforts to reconstruct heating history using such data, however, are Ilkely to be complicated by diagenetic phenomena in fossils. The demonstration that organic residues survive in the form 01 char frcm Pleistocene banas also has implications for palaeodietary studies lnvolvinq isotopic analyses 01 carbono Although carbon isotoplc analyses may províoe evidence 01 dietary patterns, until recently an effective time limit has existed on these analyses due to diaqenetlc loss of the organic phase. To circumvent this problem ano make carbón isotopes workable for the hominid fossil record, recent research has focused on analyses of dental enamel structural inorqanlc carbonate (Lee-Thorp and Van der Merwe, 19a7; Chapler 12, this volume.). An alternative strategy could be to use char-

249

derived carbon, or even better, carbon obtained from charo derived fatty acids. While burnt specimens are scarce and shouJd not be casually destroyed, lsotopíc analysis 01 some faunal char carbon could provlds useful environmental inlormatíon (Van der Merwe, 19a9). Such sludies may become feasjble as new techniques make it possible to measure stable isotopes in microgram samples of carbono In summary, the recovery of char from fosslls is a simple and useful technique lar indicating burning, and can also provide mtormation as lo the bread lemperalure range (± 200 "C) lo which the fossHs have been heated. In addition, fossil chars contain organic molecules in sufficient quantity te be recovered and examlned, which do not appear to be extraneous in origino The stucty reported here is at best a beginning; clearly more work is necessary to characterize these molecules. Since they are probably derived from the living animal, however, chemical and isotopic study 01 such molecules, both at Swartkrans ano at other sites where burnt bones occur, may previde useful taphoncrruc, palaeobiological and palaeoenvironrnental information in the future.

ACKNOWLEDGEMENTS

We thank C.K. Brain of the Transvaal Museum for providing burnt bones from the Swartkrans excavations for this study. We are also grateful to P.E. Hare of the Carnegie Institutlon of

Washington Geophysical Laboratory, who conducted the HPLC anafyals 01amino acíds. The research was funded in part by the Foundation for Research Development (South Africa).

REFERENCES

BILTZ, R.M.andPELLAGRINO, E. 0.,1977. Thenatureofbonecarbonate. ClinicaJ Orthopaedics and Related Research 129: 279-292. BRAIN, C. K. and SILLEN, A., 1988. Evidence from the Swartkrans cave ter the earliest use of tire. Nature 336: 464-466. BROOKS, A. S, HARE, P. E.. KOKIS, J. E., MILLER. G. H., ERNST, R. D. ano WENDORF, F., 1990. Dating Pfetstocene archaeological eltes by proteín diagenesis in ostrich eggshell. 5cience 248: 60-64. BUDZ1KIEWICZ, H., DJERASSI,C. and WILLlAMS, D. H., 1967. Mass spectrometry ot organic compounds. Holden-Day, Inc., San Francisco. DART, R., 1925. A note on Makapansgat: a site ot early human occupation. South African JournaJ ot Science 22: 454. DeNIRO, M. J., 1985. Postmartem preservaticn and alteratlon ot in vivo bone coüaqen isotope ranos in retatton to paleodietary reconstruction. Nature317: 806-809. DeNIRO, M. J., 1987. Stable isotopy and archaeology. American Scientist 75: 182-191. FORBES, G., 1941. The effects ot heat on the histological structure 01 bone. Police Journal (London) 14(1): 50-60. HOEAING, T. and SILLEN, A., 1990. Assessing the state of preservation of fassil collagen by analytical pyrolysis. Annual Reporto! the Director, Geophysical Laboratory, Carnegie Institution of Washington number 2200, Washington D.C. IRWIN, W. J., 1982. Ana/ytical pyroIY$is. Marcel Dekker, Inc., New York.

LEE-THORP, J. A. and VAN DER MERWE, N J, 1988. Carbon isotope analyais ot tosen bone apatite. South African Journal ot Science 83: 712-715. MEUZELAAR, H. L. C.. HAVERKAMP, J. and HILEMAN, F. D., 1982. Pvrotvets-msss spectrometry of Recent and fossiJ uometenee.

EJsevier Scientlñc Publishing Co., Amsterdam. OAKLEY, K. P., 1954. Evidence ot fire in South African cave deposlts. Nature 174: 261-262. OAKLEY, K. P., 1955a. Fire as a paleoiithic tool and weapon. 8uJletin 01 the Prehistoric Society 21: 36-48. OAKLEY, K. P., 1955b. Earflestuse ot tire. In: CLARK, J. O., ed., Pan African Congress on Prehistory, 3rd edn., pp. 385-386. OAKLEY, K. P., 1956. The eartiestfire-makers. Antiquiry30: 102-107. SHIPMAN, P.. FOSTER, G. and SCHOENINGER, M., 1984. Burnt

bones and teeth: an experimental study ot color, morphology, crystal strueture and shrinkage. JournaJ of Archaeological Science

11;307-325 SULLlVAN, C. H. and KRUEGER, H. W., 1981. Carbon isotape analysis of separate chemical phases in modern and foss)) bone. Nature 292: 333-335. VAN DER MERWE, N. J., 1989. Natural variation in 13 C concentration and ils effect on enviranmental reconstruction using 13 C¡12C ratios in animal banes. fn: PRICE, T. D., ed., Chemistry 01archaeofogical bone, pp. 105-125. Cambridge University Press, Cambridge.

Chapter 12

Stable Carbon Isotope Studies of Swartkrans Fossils Julia A. Lee-Thorp and Nikolaas J. van der Merwe' Department al Archaeology, Univcrsity al Cape Town, Private Bag. Rondebosch, 7700 South Africa 1 etso Peabody Mueeúrr, Harvard University, Cambridge, U.SA

In this chapter we describe stable carbon isctope ratio ('3C¡12C) studíes ot the Swartkrens tossus. The primary aim was to determine the proportions ot toces ct C 3 (trees. shrubs, torbs and tubers) or C 4 (tropical grasses) origin in ttte tñets of varfcus epedee representad al the sita ano lo trace the Isotopic and trophic reíatlonsbfps between them. These mcuoec the hominid AustraJopilhecus rooustus, and various cercopithecids: Parapapio

ionesi. Papio nsmedrves robinsoni, Papio (Dinopithecus) ingens end Theropithecus darti. 13C¡lZC ratios were measured in tooth enamel, after appropriate cnernca¡ pretreatment to remove secondary carbonates. The gramioiv"orous diet of T, asnne isotopical/y clearlydistinguishable lrom tne cs-oasec ctets 01the ottter baboons.

Amongst the cartuvcres, leopard vafues most closely resemble the C 3 baboons and robust australopithecines. while ñon apparentfyconcentrated en grazers. Values for A. robustus indicate an overall reliance on Cs-based tocos. but surprisinglyalso a substantlal contribution from C4 grasses.Whether this 1$ derivad c1irectly or indirectly vta grass-eating animals. the Isctcpic evioence indicates Ihat thís hominid was a generalized rather than a specíalized feeder.

INTRODUCTION Stable carbon isotape ratio analysis ot archaeoloqicat bane ís now an established technique for reconstructing ancient diets and toad webs (tor a review see Van der Merwe, 1982). The baste prmciotes of the method are brtefly as fo/lows. At the base of the food web, plants discriminate against the heavier isotope o] carbón 3 C) during photosynthes!e. The degree of diserimination, or fraetionation, is markedly greater in plants following the Calvin-Benson or C3 photosynthetic pathway (named after the number of earbon atoms in the first metabolie procuct. phosphoglyceric acid) than in plants following the Hatch-Slack or C 4 pathway (again after its first product, aspartate or rnalate). As a result, the two groups are tsotoplcalty distinct (Smith and Epstein, 1971). In warm savanna environments, trees. shrubs, forbs, corms and tubers are C 3 plants, while the grasses are overwhelmingly C 4 (Vogel et al., 1978; Livingstone and Clayton, 1980). In temperate environments the grasses, too, are CJ. These distinct isotopic patterns are passed along the food chain to the tissues of the consumers (Fig. 1), with sorne further fractionation (OeNiro and Epstein, 1978). Gollagen, the main protein constituent of bane and dentine, is enriched (more positive) by about 5 %o (parts per thousand) relative to the diet (Vogel, 1978: Krueger and Sullivan, 1984; Lee·Thorp el al., 1989), while carbonate substituted in tha mineral phase is enriched by about 12 %o (Sullivan and Krueger, 1981; Krueger and Sullivan, 1984; Lee- Thorp el al.• 1989). These isotopic patterns (Fig. 1) find applications in palaeoenvironmental and animal palaeodietary studies. Until recently, only bone collagen was routinely used as sample material in these studies, because wel1·preserved collagen is isotopically stable and contaminants are relatively easy to eliminate (Hassan, 1975). Eventually, however,

e

collagen hydrolyses and dissolves away, especially in warm envlrcnrnents (Hare, 1980) such as the Southern African interior. Thus appílcations have been limited to lhe recent past and excluded interesting evolutionary periods in the more distant human past, The use ofthe lonqer-Iived mineral phase, a series of bicloqicat apatités containing substituted carbonate (LeGeros, 1981), was inhibited by disagreement over the effect of post-depositicnal changes an the original isotopic signature. The results of eariier work seemed contradictory. Values ter a 2-m.y.-old giraffid and an equid tooth correctJy reflected the expected diets (Ericson el al., 1981 l. In anolher study, it was eoncluded that suitably pretreated mineral provídeo reHable resulte. beca use a consistent difference between collagen and apatite lsotopíc values was rnaintained far archaeological specimens (Sullivan and Krueger, 1981 ). Their finding was disputad, however, bec8use lhis difference fluctuated for other archaeological samples (Schoeninger and OeNiro, 1982). As it turned out, the collagen-apalile test was not entirely suitabJe, partly beca use the difference between the two phases varies with trophic level (Krueger and Sullivan, 1984; Lee·Thorp el al.• 1989), and partly because it Is Iimited in time by the longevity ot collagen. Recent studies using fossB material from South African sites demonstrated that expected browsing or grazing isotopic signatures are maintained for at /eas1 3 m .y. (Lee- Thorp and Van der Merwe, 1987; Lee-Thorp, 1989). It was found that fossil herbivore values are altered slightly, by about3-4 %0, depend¡ng on age, from the values observad tor their modern eounterparts (Fig. 2). The effect is less marked for grazers. Sorne of the difference may be accounted for by the burning of fossil fuel over [he las[ 150 years, which has led ta anomalously depleted isotopic values for atmospheric CO 2 as well as an increase in current C02 concentrations (Friedli et al., 1986).

252

STABLE CARBON ISOTOPE STUOIES

Table 1 values of acetite for fauna from Members 1 and 2, Swartkrans. Al! analyses were performed on tooth enamel, except for Procavía sp. (mainly whole teeth) and Lagomorpha (bone). Specimens marked ~ are from Member 2, aflthe others are from Member 1.

ArMas CO2

l 3c

a

-7,5 %.

-, C4 photosynthesis

~

TREES &

TROPICAL

SHAUBS

GRASSES

-26.5 %.

-12,5 %.

t

BROWSER APATITE

-14,5 'Jo.

INTERMED FEEDERS

t

GRAZER

APAme -O.. ""

Fig.1

A simple model Icr stable carbón isotope pathways in a modern savanna environment, showing the fractionation occurring al various tevets o, the toad web. IJC¡12C ratícs are expressec in the notaticn as per mil (%o) according fa the following relationship: S13C (%o) = (RJR••r-1) x 1000, where R = 1~C¡12C.

o

This means that values for present-day plants and anirnals are depleted by at leasl 1 %. and thus provide a slightly shifted 'reference point.' On present knowledge, the remaining differenea 01 about 2 %o must be attributed to diagenesis - alteration due to exchange with matrix carbonates over time. Nevertheless, an isotoplc distinction 01 at least 10 %o remains between browsers and grazers. Tooth enamel has been found to be the most reliable mineral tissue far ísotoplc studies (Lee-Thorp and Van der Merwe, 1987, 1991; Lee-Thorp, 1989; Koeh et al.,

1990), Thus 13C¡12C·ratia analysis of fossil tooth enamel provides a useful approach for tracing diets and isotopic relationships in andent palaeoecosystems such as those represented at Swartkrans during the depcsitlcn 01 Members 1 to 5. At the outset we assumed that tne Swartkrans environment remained, in general, a summer-rainfall savanna region during the Pleistocene and that the isotopic food web model shown in Fig. 1 applies. We used a comparative ísotopic approach, using known browsers and grazers to establish isotopic 'endpolnts' in the system, for investigating the diets 01 extinct species. Known browsers includad Tragelaphus cf. strepsiceros (kudu), Tragelaphus el. scriptus (bushbuek) and Oreotragus ef. oreotragus (klipspringer). Grazers included Connoehaetes sp. (wildebeesl), Aleelaphus sp., Equus burchelfii (zebra), Equus eapensis (giant zebra) and Megalotragus cf. priscus (giant hartebeest). The latter two species are extinct, but their grazing habits are unmistakable from tooth morphology. Small herbivores with unknown diets ¡ncluded hare (Lagomorpha sp.) and two speeies 01 hyrax, differing in size (Procavia transvaaJens;s being larger than P antiqua). The Iimited sample of carnívores included Panthera pardus (Ieopard) and P. leo (lion). Extinct cercopithecid specimens included Papio hamadryas

UCT

Catalogue

number

number

2448 2449 2450 3399 3400 341034113523 3792 3793 3524 3794 3795 2451 2452 2453 3401 3402 3403 34133414· 34153796 3797 2454 2455

SK 3023 SK 2304 SK 2681 SK 2576 SK 2541 SK 2329 SK 1631 SKX 8876 SK 14631 SK 14643 SKX 9958 SK 8483A SK 9604 SK 2261 SK 2354 SK 5946 SKX 9644 SKX 5844 SK 2063 SK 2626 SK 3990 SK 1869 SK 11798 SK 8089 SK 411 SK 495 SK 507 SK 2158 SK 14162 SK 430 SK 431 SK 439 SK 497 SK 2319 SK 512 SK 502 SK 503 SK 586 SK 628B SK 352 SK 5960 SK 1869

2456 2457 2458 2459 2460 2461 2462 2463 3405 3406 3407 3408 3409 3517 3518 3420-

Species

Tragelaphus cf. strepeíceros Tragelaphus cf. strepeceroe Tragelaphus ef. strepsiceros Tragelaphus cf. streostceros Tragefaphus cf. strepsiceros Tragefaphus d. scriptus Oreotragus cf. oreotragus Procavia transvaalensis Procavía transvaalensis Procavia transvaalensis Procevie antiqua Procavía antíqua Procavía antiqua Connochaetes sp. Connochaetes sp. Connochaetes sp. A/celaphus sp. Alcelaphus sp. cr. Megalotragus Equus capensis Equus capensis équus burchellií

Lagomorpha sp. Lagomorpha sp. Theropithecus dartí Theropíthecus darti Theropithecus darti Theropithecus aerti Theropíthecus derti Papio h. robínsoni Papio h. robinsoni Papio h. robinsoni Papío h. robinsoni Pap;o h. robinsoni Parapapio jonesí ? Parapapio iones! Papío (D.) ingens Peoio (D.) ingens Papio (D.) ingens Panthera pardus Panthera pardus Panthera leo

-10,7 -10,9 -10,9 -10,8 -12,4 -1,4 -11,7 -9,9 -9,0 -7,1 -8,5 -8,8 -5,2 -1,2 -1,2 -0,9 -3,6 -2,6 +2,2 -2,1 0,0 -2,8 -4,0 -5,1 -4,0 -2,8 -0,3 +0,4 -3,7 -12,3 -10,2 -11,3 -10,8 -10,9 -9,5 -12,8 -10,1 -10,8 -10,9 -9,8

-10,7 -2,2

rooinsoni, P. (Oinopithecus) ingens, Theropitheeus darti (formerly known as Simopithecus darti daniefl) and Parapapio ionesi. The ecological relationships between these species are not clear, but based on the morphological similarities in denti· tion between P h. robinsoni and its modern descendant, the ehaema baboon (P. ursinus) (Freedman, 1957), and T darti and the modern T. gefada, it may be inferred that P.h. robinsoni had a broadly C3-based diet and T. darti a graminivorous, C.-based one (Lee-Thorp et al., 1989). Papio (O.) ingens appears to have resembled P h. robinsoniin al! aspects except síze: it was much larger (Freedman, 1957). Little is known about its dietary adaptations or the n¡che relationship betwean the two species. The ecological position of P. jonesi is even less clear. There have been differing taxonomic interpretations since Freedman (1957) first recognized the genus Parapapio at Swartkrans, and others have reclassified specimens into

253

STABLE CARBON ISOTOPE STUDIES

either Cercocebus (mangebeys) or Papio (Szalay and Delson, 1979) or all into Cereoeebus (Delson, 1984). It is possible to determine modem mean the validity of the dietary predietions and the • ± 1 S.O. o ecologicallisotopic relationships amongst these 0 0 primates using carbon lsotcpe anatys¡e 01 tooth J -12 enamel. o o Flnatlv, the diet 01 Australopithecus robustus. o whose remains are abundant throughout o Swartkrans Members 1 to 3, may be evaJuated using 13C¡12C_ratio analysis. The diet 01 this -6 hominid has long presentad an intriguing probMAK Caco,¡lj lem for palaeoanthropologists, primarily beKRM -3 deposil , cause a number 01physical features which disSIS breccia tinguish it from other hominids (A. africanus and I Horno), appear to suggest dietary specialization (Roblnson, 1954; 1963; reviewed in Grine, 100 120 140 16D 1 2 3 20 40 60 80 1981). The massive chewing apoaratus, large x 1000 years million years rnolars and reduced caninas (Wood ano Stack, Fig.2 1980), appear to be adaptations for heavy crush13 I) C valúes lor browsinq bovtdsin a chronoloqical sequence. Sitas of varying age are ¡ng and grinding. Dietary hypotheses propasad representeo. the youngest is about1000years. and the oloest.Makapansqat Member to accaunt far these features have included 3, about 3 rnillion years. Open symbols represent bone samples, cíosed symbols 'seed-eating' (Jolly, 1970), bone-crushing enamet Diagram reproduced trom Lee-Thorp and Van der Merwe (1987). (Szalay, 1975), consumption oftough vegeta tian (DeBrul, 1975), or 01 hard nuts, beans and fruit effervescence ceased (several days), washed until neutral, (Walker, 1981; Kay, 1985; Peters. 1987). Scanning electrón and freeze-dned. The same pretreatment procedure was 101microscape studies af occlusal enamel microwear suggested lowed for the bone and whole tooth samples. For the calcite anly that the robust australopitnecines ate small, hard objects and breccia samples, the acetic acid step was left out, since (Grine, 1981; Grlne and Kay, 1989). Current opinión seems to acetic acid dissolves away CaC0 3. CO2 was produced in the favour the hard fruitlnuts option, and few have questioned lhe standard way by reaction with 100 % phosphoric acid and idea of specialized herbivary, except for Szalay (1975) and collected by cryogenie distilJation, as described previously Mann (1981). Stable carbon isotopes are capable 01 distin(Lee-Thorp, 1989). Gas yields were determined manometriguishing between many of these options; for instance cally_ 13C¡12C ranos were measured on a VG 602E dual-inlet gramlnívory is isotopically distinct from frugivory, since the ratio mass spectrometer aqainst a secondary CO2 reference latter is C 3-based and the former Ca-based, gas calibrated againstfive N.8.S. stanoards. 13C/2C ratios are reportad in the 15 notation relative to the P.D.B. standard, as METHODS per mil (%0), according to the relationshíp given in Fig. 1. The teunat sample consisted 01 tooth fragments of the anirnals Precisión. as determined from duplicate analyses, is better than 0,1 %O. discussed above, with the excepticn of bone samples for hare and whole teeth (Le., enamel plus dentine) for hyrax, Species, RESULTS AND DISCUSSION numbers and provenance are given in Table 1. Most of the faunal specimens are from Member 1; their ages are believed The resuíts for the fauna are reported in Table 1. In general, to be between 1,5 and 1,8 m .y. on the baels of comparativa grazer (Connochaetes, Alcelaphus, Megalotragus and Equus biostratigraphy (Vrba, 1982; Brain el al.. 1988). The rew specisp.) Ó13C values are clase to 0%0 (1he modern mean ís -0,5 %0) mens from Member 2 are younger by an indeterrninate and ccnñrm that grasses in the Transvaat Highvefd were amount. Whale calcite crystals faund in the pulp cavlties ot predominantly C4 when the fassiliferous sediments accurnusome tooth specimens were extracted for separate analysis, together with an attached breccla sample (Table 2). The hornilated in Members 1 and 2, as they are today. Grazer values are gene rally similar for the two Members, indicating that no nid sample consisted of eight fragments of teeth from Memmajor vegetaban shifts occurred during theír formatian (e.g., bers 1, 2 and 3 (Table 3). The two fragments of SK 879 were analysed separately, since their colour and appearance sugfrom a warm C.¡ grassJand to a temperate C3 grassland). This gested that they might be derived from different teeth. is consisten1 with the suggestion that the various Members Enamel was manually separated from dentine using ;ewelwere formed during deposiUonal episodes with similar enviranmental conditions (Brain et al., 1988). Values far the extinct lers sidecutters and a scalpel, and the fragments ground in a giant zebra, E. capensis, resemble those far E. burchemi. The cooled Spex milI. An aliquot of the powder was reaeted oversingle Megalotragus specimen is more enriched (positive) night in a dilute solution of sodium hypochlorite (2 %). to eliminate bacterial proteins and humates, after which it was than Ihe other grazers, which might imply greater specializa· alternately centrifuged and rinsed several times, to neutrality. tion on C 4 grasses. Interestingly, the only ather published 013C values for Megalotragus, from Equus Cave, are also The anarganic powder was bathed in 1 M acetic acid until

"=8'1lo

"

t

"

"I

STABLE CARBON ISOTOPE STUDIES

254

Table 2

¿¡13C and 0180 values for matrix carbonates al Swartkrana.

Member 1 Matrix

UCT number

Carbonate

2449 2451 2452 2455 2459 2460

Calcite Calcite Calcite 8reccia Calcita Calcite

Catalogue number

SK SK SK SK SK SK

2304 2261 2354 495 430 431

13 Ó C (%o)

¿¡130 (%o)

-0,6 -1,6

-4,4 -4,7

-1,1

-4,6

-1,4 -1,6 -1,6

-4,8

-4,7

T. strepsiceros

• • od

Megalotragus

••

A/celaphus sp. Connochaetes sp

-4,7

._.

Lagomorpha sp.

somewhat enriched comparad to other grazers from the sama levels (Lee-Thorp and Beaumont, 1990). The hare (Lagomorpha sp.) values are relatively enriched, indicating a predominantly C, (grass-based) diet, 013C values for the browsers (T. el. strepsiceros and O. et. oreotragus) cluster near -11 to -12 %o, with the exception 01 the single T. et. scriptus specimen at -1 ,4 %o. We interpret this anomaly as misidentification 01 the specimen (Lee-Thorp, 1989). As in the earlier study (Lee-Thorp and Van der Merwe, 1987), browser values al Swartkrans are about 3 %o more positiva than tne modern mean. Part 01 the difference, about 1.5 %o. may be due to the fossil-fuel effect, as discussed aboye. The remaining diHerence probably represents post-rnortern alteration, resulting in a shift towards matrix carbonate values. The mean S13 C and S18 0 values obtained for the calcite and breccia specimens from Member 1, are -1 ,3 %o and -4,6 %o respectively (Table 2), which is consistent with the derivation of the breccia from a Proterozoic marine dolomite. Isotopic atteratíon is less obvlous for the grazers, because the original biogenic values cluster near O %0, similar to values ter the matrix. In other words, post-mortern isotopic alteration is unmistakabie for animals eating C, loods, but barely detectable for those eating C4 grasses. This factor must be borne in mind when interpreting results for extinct species which consumed unknown diets. Nevertheless, a wide separation between browsers and grazers remains at thls site. The cercopithecids may be compared against the background provlded by the browsers and grazers (Flg. 3), Values fer T. darti cluster at the positive end, confirming that thls species was predominantly dependent on C4 grasses, although the spread of values may imply dietary heterogeneity. The fossil Theropithecus seems to have had a diet similar to that described lor the modern T. gelada (Dunbar and Dunbar, 1974), provlding lurther support lor the Idea that a gramlnivorous dietary specialization developed early in thls genus (JoUy,1972). Values lor P robinsoni show that It was a conslstent C 3 consumer. Indeed, the values are slightly more negative than those of modern Papio, which has S13 C values of about -12 %o (Lee-Thorp et al., 1989), rellecting a diet which includes a little C 4 grass and/or grazing animals, The more negative values for P robinsoni indicate a complete lack of Ca-based items. The difference may be related to sympatry 01 Papio and Theropithecus species and niche separation. In the Ethioplan highlands, where sympatry 01 Theropithecus and Papio stlU exists (JoUy, 1970; Dunbar and Dunbar, 1974), Papio concentrates on tree products and seldom forages in grassland areas.

Procavia sp. P robinsoni

D. ingens

T. darti P jonesi

.-

• •

•

•• ••

••

P pardus

O. oreotragus

••

-• -

-• .

•

Member 2

Equus sp. P leo

-10

·5

o

Ol3e (%0) Fig.3 Dlstributlon of Ó'3C vatues for fossil fauna, including herbivores, caro nivores and cercopithecids. from Members 1 and 2 at Swartkrans.

The three values for P (D.) ingens are conslstent, between -10 and -11 %0, and indicate a concentratíon on C3 foods. The isotopic data thus support the idea that its diet resembled that of P. h. robinsoni, at least isotopically. The values obtained tor the two Parapapio specimens are difficult to ínterpret, because they differ by more than 3 %. On the lace 01 It, these values suggest that SK 502 had an exclusive C,-based diet, while SK 512 also Included sorne C,-based items in the die!. It ís possible that the diet of Parapapio varied widely, but we favour an alternative explanation - that in fact two species are represented, one 01 which (SK 502) may be Papio. Thls explanation ís supported by the observatlon that SK 502 was only tentatively assigned to the genus Perepeplo. and in any case, assigning these specimens has presented taxonomic problerns, as discussed aboye. Carnivores represent the highest trophic level. S13 C values for the two leopards, P. pardus, from Member 1 are both depleted and show that they ate prey with C, diets, From the suite of Member 1 launa analysed, the likely prey candidates in the smaller size range, with suitable isotopic values, include only Papio baboons, A. robustus and possibly hyrax (Procavia

255

5TASLE CARSON ISOTOPE 5TUDIES

Table 3

S'3C values of Australopithecus robustus and faunal tooth enamel specimens from Swartkrans. Two fragments marked SK 879 were analysed separately (as UCT numbers 2541 and 2542) because tney were considered to be from different teeth (on the baste of colour and appearance).

UCT

Speces

number 2541 2689

2540 2542 2543 2690 2691 2692

A. robustus A. robustus A. robustus A. robustus A. robustus A. robustus A. robustus A. robustus

Member

Catalogue number

Tooth

5K 879 5K 5015 5K 878 5K 879 5K 1512 SK 1312 5K 333 5KX 35025

M, LM3 M3(?) crown M? fragment M? fragment LM' RM, RM,

ólJC

T. strepsiceros

-8,5 -9,6 -6,8

1

-8,8

2 2 3

-10,0 -7,9

-8,1 -8,1

•

Connochaetes

T. datli

(%o)

1 1 1 1

•

P. robinsoni A. robustus

P. petdus

.-_.-

•

•••

•• -10

-s li 13C (%a)

o

Fig.4 sp.) (Figs 3, 4). The two Swartkrans hyrax species, P. trensveetensis and P antiqua, rather surprisingly do not seem to have had isotopically distinct diets; the average far al! six samples is -8,1 ± 1,7 %0, or -8,7 ± 1,0 %"if the outlier at-5,2 %o is left out, suggesting a predominantly C 3 diet with a contribution from C" grasses. These values must be interpreted cautiously beca use the samples consísted mostlyofdentine, whlch yields less reliable results trian ename!. The single non, P leo, from Member 2, has an enrtched S13C varue similar to that of the grazing anirnals, suggesting that grazers, which are very abundant in the site, were íts main prey. A similar isotopíc relationship between leoperds and lion has been observad in modern ecosystems, such as the Kasungu National Park in Malawi, where the larqer-bodied animals preyed upon by lions tend to be grazers. Ca-feedera are either very large (e.q.. elephant) or small, solitary animal s (e.g .. Raphicerus sberpet; which form the main prey of leopards (Van der Merwe, 1966). It is unfortunate that so few carnivore specimens were available for analysis, because their isotopic vatues may provide use fui clues about trophic relationships in a food web. The isotopic data support the proposal that at least sorne of the fossil accumulations at Swartkrans are the result of leepard activities (Brain, 1981). Finally, we turn to a discussion of the hominid results (Table 3). &13C values ter A. robustus range between -10 and -6,8 %0,. with an average of -8,5 ± 1,0 %o.. When compared with the other Swartkrans fauna, A. robustus S13 C values are more positíve than those 01 the fossil browsers and Papio baboons (Fíg. 4). The pattern 5uggests an overall dependence on C 3 foods, but at the sama time there is a substantial cQntribution from C" sources. The C 4 contribution can onJy be explained by the inclusion of C4 grasses directly in the diet, or índirectly by consumption of grass·eating animaJs, or both. Grass-eating by itself is most unlikely, beca use tooth structure is unsuitable and enameJ microwear studies discount it (Walker, 1981), but seasonal use of grass seeds and rhizomes are possibilities. We tend te favour the alternative explanation that A. robustus ate some animal foods such as grass-eating vertebrates or insects. We are unable, on the basis afcarban isotopes alone,

Distribution e¡ ¡:¡13 C values for Australopithecus robustus tooth enamel from Members 1 and 2 at Swartkrans. compared with sorne of the fossil fauna.

to distinguish between the two atternatíves. In any event, the isotopic data leads us to the somewhat surprising conclusión that A robustus had a generalized, not a speclallzed. diet. The data do not support proposals for specialized graminivory, or specialized frugivory, but also do not contradict proposals for hard frults and nuts as the principal dietary items. Although data are few, there appears to be no trend between Members 1 and 3. l.ikewise, there are no obvious differences for various teeth (Table 3), although they have different develcpment times. This suggests that infant, juvenile ano adult diets are lsotopically similar, and that the choice of a particular tooth, e.g., M3, does not signiñcantly affect the resulte. CONCLUSIONS We have shown that the diets and isotopic relationships between extinct animals may be determined by 13Cl2C_ratio analysis of tooth enamei. Amongst the baboon species at Swartkrans, T. darti is shewn to llave been graminivorous and isotopically c1early distinct from the other three baboons, P. n. roblnsoni, P (D.) ingens and Parapapio jonesi, which al/ depended on e3 foods to varying degrees. Niche separation between several sympatric babeon species may account for the observatíon that P. h, robinsoni lacked any C 4 contribution to it5 diet, unlike its modern descendant P. ursinus. Leopard 013C values most closely resemble those of the Papio babaons, and hominids, suggesting that these formed its principal prey. The lion clearly ate mainly larger·bodied grazers. 13C/12C.ratio analysis af A. robustus tooth ename fragments provides compelling evidence for an unexpectedl generalized, passibly even omnivorous diel. The isotopi approach offers a significant new tool for examining the diets of hominids and evaluating the varíoLJs dietary hypotheses.

256

STABLE CARBON ISOTOPE STUDIES

ACKNOWLEDGEMENTS

We are grateful to C.K. Brain tor his substantial contribution to this projeet. We thank our colleagues J. Sealy, A. Sillen, J. Lanham, F.Thackeray and J. Kovacs for tneír advice. Financial

support was provided by the University of Cape Town, the Harry Oppenheimer lnstitute for African Studies and the Foundation tor Research Development.

REFERENCES BRAIN, C. K.. 1981. The hunters or the hunted: an introducticn re Afric8n cave taphonomy. University ot Chicago Press, Chicago. BRAIN, C. K., CHURCHER, C. S, CLARK, J. D., GRINE, F. E.. SHIPMAN, P., SUSMAN, R. L.. TURNER, A. and WATSON, V.. 1988. New evidence of early hominids, their culture and envirenment from the Swartkrans cave. South African Joumal af Science 84: 828-835. DEBRUL. E L., 1975. EarJy honunid Ieedinq mecnanlsms. American Journalof Physical Anthropology 47: 305-320. DELSON, E., 1984. Cercopithecid biochronology 01 the Atncan PlioPleistocene: ccrrefaticn among eastern and southern hominldbeartnq Iocaünes. courier Forschungsinstitut Senckenberg 69: 199-218. DeNIAO, M. J. and EPSTEIN, S., 1978. lnfluence 01 diet on the dtstr'butlon ot carbon isotopes In animals. Geochimica et cosmoctumice Acta 42: 495-506. OUNBAR. A.I, M. and DUN8AA, E. P.. 1974. Ecological relationsand ruche separation between sympatnc tertestrtal primates in Etniopia. Folia primato/ogia 21: 36-60. EAICSON, J. H., SULLlVAN, C. H. and BOAZ, N. T., 1981. Diets 01 Pliocene mammals from Omo, Ethiopia, deduced from carbón isotope ranos in tooth apetite. Palaeogeography, Peteeocñmetotogy, Palaeoecology36: 69-73. FREEDMAN, L., 1957. The tcss!l Cercopithecoidea or South Atríea. Arma/s of lhe Transvaal Museum 23: 121-262. FRIEDLI, H., LOTSCHER, H, OESCHGER, H., SIENGENTHALER, U, and STAUFFER, B., 1986. Ice-core record ot the 13C¡J2C ratio of atmospheric C02 in trie past two centuries. Nature 324: 237-238. GRINE, F. E., 1981. Trophicdifferences between 'gracile' and 'robust' Australopltnecines: a scanning electron microscope analysis of occlusat events. South African Journa/ ot Science 77: 203-230. GAINE, F. E. and KAY, R. E.. 1989, Early hominid diete from quantitative image anafysis ot dental microwear. Nature 333: 765-768. HAAE, P. E., 1980. Organic geochemistry ot bone and íte relation to the survival of bone in the natural environment. In: BEHRENSMEYEA, A. K. and HILL, A. P., eds, Fossils in the making, pp, 208-219. University 01 Chicago Press, Chicago. HAS SAN, A. A., 1975. Geochemical and mineralogical studies of bone and their implications for radiocarbon dating. Ph.D, thesis, Southern Methodist University, Texas. JOLL Y, C. J., 1970. The seed-eaters: a new model of hominid differentialion based on a baboon analogy. Man 5: 5-26. JOLLY, C. J., 1972. The classification and natural history of Theropithecus (SimopithBeus) (Andrews, 1916), baboons of the Afriean Plio-Pleistocene. Bu/letin 01 the British Museum (Natural History) Geology22: 3--123. KAY, A. F., 1985. Dental evidence lor the diet of Australopithecus. Annual Review of Anthropology 14: 315-341. KOCH, P. L., BEHRENSMEYER, A. K., TUROSS, N. and FOGEL, M. L., 1990. Isotopic fidelity during bone weathering and burial. Annua/ Reportolthe Geophysical Laboratory, Carnegie Institution of Wash· ington, 1989-1990, pp. 105-110, Geophysical Laboratory, Wash· ington D,C. KRUEGEA, H. W. and SULLlVAN, C. H., 1984. Models for carbon isotope fractionation between diet and bone. In: TURNLAND, J. R. and JOHNSON, P. E., eds, Stabfe isotopes in nutrition, pp. 205220. American Chemical Soeiety Symposium series 258. LEE-THORP, J. A., 1989. Stable earbon isotopes in deep time: the diets of fossil fauna and hominids. Ph.D. thesis, University 01 Cape Town. LEE-THORP, J. A. and BEAUMONT, P., 1990. Environmental shifts in the last 20 000 years: isotopic evidence from Equus Cave. South African Journal 01 $cienee 86: 452-453.

LEE-THORP, J. A., SEALY, J. C. and VAN DER MERWE, N. J., 1989. Stable carbon tsctcpe ratio differences between bone collagen and bone apatite, and their retaüonshlo to diet. Joumal of Arehaeological Science16: 585-599. LEE-THORP, J. A. and VAN DER MERWE, N. J., 1987. Carbon isotope analysis ot tossü bane apatite. South African Journa/ ot Seience 83: 712-715. LEE-THORP, J. A. and VAN DER MERWE, N. J., 1991. Aspects of the chemistry of módem and rossn biological apetites. Journal 01 Archae%gical Scienee 18: 343-354. LEE-THORP, J. A., VAN DER MERWE, N. J. ano BRAIN, C. K., 1991. lsotoplc evidenee ter dietary differenees between two extinct baboon soeces from Swartkrans. Joumaf 01 Human Evotutkm 18: 183-190. LEGEROS, A. Z., 1981. Apetites in biological systerns. Proceedings ot Crystal Growth cnerecteasnoe 4: 1-45. UVINGSTONE. D. A. and CLAYTON, W. O., 1980. An artitudinal cline in tropical African grass floras and its paleoecoloqtcaí significance. Quaternary Research 13: 392-402. MANN, A. E., 1981. Diet and human evojutlon. In: HAADING, R. S. O. anñ TELEKI, G., eds, Omnivorous primates, pp. 10-37. Columbia University Presa, New York. PETERS, C., 1987. Nut-llke oil seeds: food for mcnkeys. chimpanzees. humana and probably ape-rnen. American Journal ot Physieal Anthropology 73: 333-363. AOBINSON, J. T., 1954. Prehominid dentition and hominid evolution. Evo/ution 8: 324-334. ROBINSON, J. T., 1963. Adaptive radiation in the australcpithecfnes and theorigin 01 mano In: HOWELL, F. C. ano BOUALlERE, F., eds. African ecotoqy and human evolution, pp. 385-416, Aldine, Chicago. SCHOENINGER, M. J. and DeNIAO, M. J., 1982. Carbón isotope rattos of apatlte from fossil borre cannot be used to reconstruct diets of animals. Nature 297: 557-558. SMITH, B. S. and EPSTEIN, S., 1971. Two categories 01 13C¡12C ratics for higher plants. Pfant Physiology 47: 380-384. SULLlVAN, C. H. and KRUEGER, H. W., 1981. Carbon isotope analysis 01 separate chemical phases in modern and fossif bone. Nature 292: 333-335. SZALAY, F. S., 1975. Hunting-scavenging protohominids: amodelfor hominid origins. Man 10: 420-429, SZALAY, F. S, and DELSON, E., 1979. Evo/utionary history 01 the primates. Aeademic Press, New York. VAN DEA MERWE, N. J., 1982. Carbon isotopes, photosynthesis and archaeology. American Scientisr10: 596-606. VAN DER MERWE, N. J., 1986. Isotopie ecology of herbivores and carnivores. Palaeoecology of Africa 17: 123-131. VOGEL, J. C., 1978. Isotapic assessment of the dietary habits of ungulates. Sourh African Journaf o, Scienee 74: 298-301. VOGEL, J. C., FULS, A. and ELUS, R. P., 1978. The geographical distribution of kranz grasses in South Africa. South African Journa/ of $cienee 74: 209-215. VRBA, E. S., 1982. Blostratigraphy and chronology, based particularly on Bovidae, of southem hominid-associaled assemblages : Maka· pansgat, Sterkfontein, Taung, Kromdraal, Swartkrans; also Elandsfontein (Saldanha), Braken HiII (now Kabwe) and Cave of Hearths, In: Actes 1 er Prétirage Congres /nternational de Paléontologie Humaine, pp. 707-752, Niee. WALKER, A., 1981. Diet and !eath. Dietary hypotheses and human evalution. Philosophical Transactions of the Royal Society 01 London (B)292: 57-64. WOOD, B. A. and STACK, C. G" 1980. Ooes allometry explain the differences between -graeile" and "robust- australopithecines? American Joumal Physical Anthrop%gy 52: 55-62.

o,

Chapter 13

A Taphonomic Overview of the Swartkrans Fossil Assemblages c. K. Brain Transvaal Museum, P,

o. Box 413. Pretoria, 0001 South Africa

Cornprehensiva tcssu assemblaqes are now avanabíe from uve different stratigraphic units al Swartkrans, namely Member 1, Lower Bank; Member 1, Hanging Remnant; Member 2; Member 3; ano Mernber 5, tne latter being ot reoent age in cornpartson with the othera. Sediments compris;ng each ot these units enterad through difterent entrances at intervals throughout the evolutlon o, the cavern system, while thereis evidence that the structure 01 each of the entrance areas influenced the composition ot fossil assemblages Ihat passed through

them. The rtcnest assemblage, in terme of hominid remains, rs that from the Hanging Remnant ot Member 1, which apparently accumulated under a vertical shatt above the cave'e north wall, with large cats being among the rnam bone-cojectínq agents.

Fossil assemblages to be considered here come from the following stratigraphic unus: Member 1, Lower Bank; Member 1, Hanging Remnant; Member 2. from the 1979-1986 excavatíons; Member 3; and Member 5. It should be notad that, befare the 1979-1986 excavations, a large collection ot foesüs, derivad from earlier mining activities at the cave, was collectively referred to as the Member 2 assemblage (Brain, 1981). It is now known tnat this is a very mixed collection. caming fram virtually an the sediments preserved in the Guter Cave that nad been removed by miners and dumped around and within the cave. As prevlcusty described, the sample consisted of 5894 specimens from aminimum of258 individual arurnate, but they clearly carne nat only from Member 2 as currently defined, but also from Member 5 as well as from a range of intermediate-age sediments that have been Iost through the mining process. They will not be considered further here.

AGE RELATlONSHIPS OF MEMBER 1 COMPONENTS As outtíned in Chapter 1 ofthis volume, Member 1, as currently defined, consists of two separate sediment masses that entered the cave through different openings in the roof of the original cavern. The Lower Bank of Member 1 entered via a steeply-inclined opening above the southeast wall of the cavern and consisted of a massive deposit of earth and chert blocks covering much of the floor of the cave, inclined downwards towards the north at angles of up to 30~. In the central area of the Quter Cave, two large dolomite blocks that had fallen from the roof of the cave, actad as a retaining wall and resulted in approximately horjzontal sedimentation immediately to the south of them. In general, Ihe deposit is pink in colour and lightly calcified, suggesting that it was a fairly rapid accumulation. Excavated parts of this unit provided 3570 water-worn chert pebbles derived, in all probability, from alluvium along the bank of the Blaaubank stream, a

short distance to the south of the cave's opening. The Hanging Remnant unit of Member 1 is confined to Ihe irnmediate vicinity of the cave's north wall, as described in Chapter 1, and ís undercut along its entire width by an erosionar surface. Sediment comprising this unit appears to nave entered via a near-vertical shaft directly above the north wall and to have built up a talus cone radiating southwards. As indicated in Chapter 1 01 this volume, it is assumed that the base 01this unit rested originally on the upper surface of the Lower Bank of Member 1, but that eroslon within the cave created a wide space. on either side of this junction, that was later filled by Member 2 and subsequent deposits. On this assurnption, the Hanging Remnant is younger than the lower Bank, but it must be stressed that this ls an assumption. An alternative interpretation ís that the Hanging Remnant entered the cave first and fanned out over the ñoor in, at least, the northern parts of the Outer Cave. The existence of an additlona! erosrone! cycle would then have to be invoked, which removed the entire base of the Hanging Remnant deposit before entry of Lower Bank sediment beneath it. On stratigraphic grounds it is not easy to determine which of the two units of Member 1 entered the cavern first, but there is sorne palaeontological evidence to suggest that the Hanging Remnant is the older of the two units. lt revolves around the presence of species in the Hangjng Aemnant assemblage that are not known from other Swartkrans deposits: Parapapio jonesi is a small baboon representad by 29 cranial pieces from a minimum of eight individuals in the Hanging Remnant assemblage, but has not been recorded from the Lower Bank of Member 1 or from subsequent depos~ íts at Swartkrans. Elsewhere in the Sterkfontein val1ey it is known from Sterkfontein Member 4 and Kromdraai A. It seems likely that Parapapio jonesi disappeared from the vicinity of Swartkrans after Member 1 times, but it may have done so in the glacial episode represented by the erosion separating the Hanging Remnant unit from the Lower Bank, jf the former is older than the Jatter.

TAPHONOMIC OVERVIEW OF FOSSIL ASSEMBLAGES

258

Table 1 The representation ot various categories of mammals, with adult ñve weights in excees of 5 kg, in five Swartkrans stratigraphic units: M 1LB = Member 1, Lower 8ank; Ml HA = Member 1, Hanging Remnant; M2 = Member 2; M3 = Member 3 ano M5 = Member 5. The flgures represent minimum numbers 01 individuals present, while those in parentheses indicate the percentage contributicn ot these individuals to the total number of animals whcse remalns were found in the assemblage. Unit

Hominids

M1LB M1HR M2 M3 M5

14(11.9) 90 (20,3) 19 (16,7) 9 (5,1) 1 (0,6)

Baboons 11 117 10 15 3

(9,3) (26,4) (8,8) (8,5) (1,9)

Carnívoras 19 54 16 33 17

(16,1) (12,2) (14,0) (18,8) (10,6)

Dinofelis sp., the false sabre-loolhed cal, is known from lhe Hanging Remnant by two cranial and twe postcranial fossils, probably from a single individual. It has not been recorded Irom any other Swartkrans assemblage, but D. barlowi is known from lhe Makapansgal Limeworks, Slerkfonlein Member 4 and Bolts Farm. It is possible that this cat, which has a long earuer fossil record. disappeared from the Swartkrans vicinity immediately after the Hanging Remnant infilling. cf. Makapania sp. ís a remarkable ovibovine antelope representad by ten cranial fossils Irom a minimum of three adult individuals in the Hanging Remnant of Member 1. It is not known from the Lower Bank or subsequent Swartkrans deposits. Originally described as Makapania broomi from Makapansgat Limeworks, lt was reeognized as Makapania ef. broomi by Vrba (1976) from Member 4 al Slerkfontein. The dentíttons from the Hanging Remnant at Swartkrans are appreciably smaller but in other respects reminiscent 01Makapania. It is possible that these specimens represent the last 01 the Makapania lineage, which may have become extinct soon aftar the Hanging Remnant depositian.

TAPHONOMIC FEATURES OF THE MEMBER 1 HANGING REMNANT FOSSIL ASSEMBLAGE The excavations described in this volume did not penetrate into the Hanging Remnant unit 01 Member 1, except lar the analysís 01 the fossils contained in a large fallen block and designated lhe SKR assemblage. This ls described by Virginia Watson in Chapter 2 af this volume. The main palaeonlological sample from lhe Hanging Remnant was derived lrom the Broom/Robinson excavations between 1948 and 1953, as well as from lime-mining operations af 1949-1951. The macrovertebrate assemblage has been described in detall (Brain, 1981), and was faund ta consist of 2381 skeletal parts from animals belonging lo 41 Identified taxa. As detailed in Table 1, primates, consisting of homlnids and cercoprthecoids, are better represented than any other category of mammals, making up 46,7 % of the total number of individuals. They are followed in abundan ce by ungulates, (34,5 %) and carnivores (12,2 %). 11 is unfortunale thal lhe Hanglng Remnant fossil sample cannot be compared directly, in taphonomic terms, with the other assemblages, more recently excavated and described in this volume. The Hanging Remnant matrix is heavily calcified, with the result that all the fossils have required laborious preparation, either w;th mechanical tools or acetic acid. The

-------------

Ungulates 39 153 49 90 92

(33,1) (34,5) (43,0) (51,1) (57,5)

Hyracoids 20 26 13 16 41

(16,9) (5,9) (11,4) (9,1) (25,6)

Others 15 3 7 13 6

(12,7) (0,6) (6,1) (7,4) (3,8)

Total 118 443 114 176 160

earlier excavation technique involved the use af dynamite to loasen blocks of breccia, after which these were broken up with hammers in search of fossils. Insignilicant-Iaoking fossil banas were discarded and anly those of reasonable cornpíeteness were retained tor preparation and study. The same is true 01 fossils recovered during the lime-mining episode. The resultant tossu sample is Iikely to give a reasonably true reflection of animal taxa represented, since the recognition of these is aften based on cranial fossils that would have been retained, but would not provide reuabte infarmation on skeletal part representation or the occurrence 01 bone Ilakes or indeterminate bone fragments. The nature of the damage observed on many 01the Hanging Remnanl fossils led me (Braln, 1981) lo conclude tnat lhe assemblage had resulted largely from carnivore leeding activitieso It was necessary to account for the very large number of primate individuals whose bones had contributed to the assemblage. Of particular mterest was the factthat while the tour species 01baboons and two of hominids showed a wide range 01 adult body weights, lhe percenlage of subadult individuals presant in the sample increased as body weight increased. This suggesled lhal lhe predalors responsible for hunling these primates had preferentially selected prey-animals within a particular weight bracket. It ís probably not a coincidence that this weight bracket coincided with that preferred by leopards, which were by far the most cammon predators whose remains form par! of lhe Hanging Remnant assemblage (Fig. 1). The proposed scenario (Brain, 1981) was that hominids and baboons came to sleep within the entrance area of the cave on cold winter nights, as baboons still do in other dolomite caves in the vicinity today, and that they were preyed on there by leopards. They were then consumed, either in trees growing out of the cave's entrance, or within the dark recesses of the cave itself, where the remains 01 their meals contributed ultimately to the fossii assemblage. While this scenario might provide a reasonable explanation for the abundant babaon and hominid remains in the Hanging Remnant assemblage. it does nat account for the abundant non-primate remains from mammals 01 heavier body weight. Among the 82 fossil antelope individuals, for instance, whase remains are found in the Hanging Remnant assemblage, 47 came from animals too large ta have fallen prey to leopards. It was therefore suggested (Srain, 1981) lhal these had been hunted by predalors with a preferenee for larger prey, Ihe mosl likely candidatas having been the sabre-toothed cat, Meganterean, the false sabre-toothed eat, Dinafelis, and a variety of

TAPHONOMIC OVERVIEW OF FOSSIL ASSEM8LAGES

SKR 703

259

SKR 108

Fig.2 Two bone lools in the SKR assemblage Irom lhe Hanging Remnant 01 Member 1_ Each is based on abone flake that shows wear and scratching assumed lo have been caused by digging use. These are the first bone tools to have come from the Hanging Remnanl, but indicate that many others were probably present in this stratigraphic unit, though they did not find their way into the Transvaal Museum collection.

Fig.1 A reconstruction in the Transvaal Museum of a leopard canying an Australopithecus robustus child that it kílled and ate in the Swartkrans cave. The reconslruclion is based on the juveni/e cranium SK 54, Irom the Hanging Remnant of Member 1, tha! has a puncture mark in each of its parietal bones, assumed to have been made by Ihe lowercanines 01 a leopard.

hyaenas. It was assumed thal these carnívores also used lhe cave, or the catchment area of its entrance, as a feeding lair, with the remains 01 their meals contributing to the fossil assemblage. It is 01 interest lo nolice thal only one stone artefact is known with certainty lo have come from the Hanging Remnant breccia (Brain, 1981). It is c!assified as a 'heavy-duty scraper: struck from a quartzite cobble and showing _a certain amount 01 abrasion. It is quite possible that other slone artefacts were present in lhe Hanging Remnant breccia excavated in lhe Broom/Robinson and lime-mining operalions, but that they were not recognized, and were lost in the process. My impression is, however, that lhey were less common in this stratigraphic unit than in the Lower Bank of Member 1, Member 2 or 3. Similarly, no bone tools were recognized in the Hanging Remnant assemblage. It is likely that this was due to lhe fact that bone flakes and non-diagnostic skeletal parts were discarded onto the breccia dump, where many of them were carried away over the intervening years by casual visitors to the site.

- - _ ... _._--

-.

As mentioned above and described in detail by Virginia Watson in Chapter 2 of this volume, the first complete bone assemblage from the Hanging Remnanl. designated SKR, is now available from the large breccia block tha! was processed in its entirety with acetic acid_ Here, 709 kg of breccia matrix yielded 1277 pieces of fossil bone. among which were two bone tools shown in Fig. 2. Designated SKR 108 and 703, each consis!s 01 a short pie ce of bone flake, one end of which shows the wear and scratching so characteristic of lhe digging tools from the Lower Bank of Member 1, as well as Members 2 and 3, described in Chapler 8 of this volume. It .therelore seems very likely thal il all the bone pieces that had originally been present in the Hanging Remnant assemblage were available lor study, they would include a fair number 01 bone tools. TAPHONOMIC FEATURES OF THE MEMBER 1 LOWER BANK FOSSIL ASSEMBLAGE

As described in Chapter 1 of this volume, sedimentcomprising the Lower Bank 01 Member 1 entered the Swartkrans cavern through a steeply inclined entrance above the cave's southeast walL This sediment formed a talus slope, dipping lo the north al about 30°. It is Iikely that the cave was much more accessible to hominids and other animals during the accumulation period of the Lower Bank Ihan it was while the Hanging Remnant deposit was building up. The cave ilselt, or the immediate catchment area of its entrance, appears to have been regularly used by hominids, as suggested by the presence, in the excavated part of the Lower Bank, of 402 pieces of fractured stone, otwhich 62 are regarded by Desmond Clark (Chapter 7, this volume) as being artefacts. In addition to these, the excavation produced 17 bone tools, described in Chapter 7 of this volume. The cave was also, apparently, frequently used by carnivores as a feeding or breeding ¡airo Apartfrom abundan! carnivore remains themselves, detailed by Alan Turner in Chapter 6 of this vo)ume, a total of 50 coprolites, assumed to

• 260

TAPHONOMIC OVERVIEW OF FOSSIL ASSEMBLAGES

"'. their victims in the cavern itself. A contemporary baboon sleeping-cave no! far from Swartkrans was documented in detail in an earlier work (Brain, 1981). The entrance area 01 this cave is shown in Fig. 3, with the inclined talus slope leading down the entrance clearly visible. Fossil bones in the Lower Bank assemblage tend to be much more Iragmentary than Ihose in the Hanging Remnant colleclion and this can, I think, be attribuled to the tact that the floor 01 the cave was clearly much more accessible lhan had been the case in Hanging Remnanl limes, when a vertical shaft linked the cavern with the hjllside. The presence of coprolites shows lhat hyaenas made regular use of the cave and theír activity would have added lo the fragmentation 01 the bones, an influence superimposed on the lrampling effecl of hominids and other animals. There is no doubt that fragmentalion has had the etfect of lowering the incidence of carnivore-induced damage marks on the individual specimens. The figure of 0,25 % incidence of carnivore damage marks would doubtless have been much higher had the bone pieces remained more intact, before fossilizalion.

TAPHONOMIC FEATURES OF THE MEMBER 2 FOSSll ASSEMBlAGE

Fig.3

The entrance and talus cone in a cave on the larm Uitkomst, about 20 km northeasl 01 Swartkrans, used regularly in the past by baboons as a sleeping sile. A similar cave lorm is visualized for the Swartkrans cave's entrance area during the accumulation 01 the Lower Bank 01 Member 1, and Member 2, when hominids and baboons slept there.

have come Irom hyaenas, was recovered in the lower Bank excavation. Furthermore, as described by Rosemary Newman in Chapter 9, 131 bones showed recognizable carnívore damage in the lorm 01 tooth scralches, punctures and chewed edges. There can be little doubt that carnívores not only brought many bones to the cave during lower Bank accumulation times, but also fed upon them there. Porcupines, on the other hand, appear to have had little inlluence in the accumulation of this assemblage, as only eight bones wilh porcupine gnaw marks were lound. Identifiable bones in lhe Lower Ban/< assemblage ca me from a minimum 01 118 individual mammals, with adult body weíghts in excess of 5 kg, as detailed in Table 1. Primates, both hominids and baboons, accounted for 21,2 % of all individuals, as compared to 33,1 % for ungulates. The prominence 01 primates in this assemblage, though still remarkable, is not as great as it was in the Member 1 Hanging Remnant sample, where they consliluted 46,7 % of all individual animals present. Once again, it seems reasonable to suggest !hat hominids and baboons used the entrance area of the cave as a sleeping site and that they were preyed upon there occasionally by leopards and, possibly, sabre-toothed cals, which ate

As described in Chapter 1 ofthis volume, the form ofthe cave's entrance area and talus slope during Member 2 accumulation times was very similar to that which existed in Lower Bank times, although the entrance itself was now further to the north. The fossil assemblage of 70 524 bone pieces was just under half lhe size of that recovered from the lower Bank. Hominid activity is once again indicated by the presence of 403 pieces of fractured stone (see Table 1 in Chapter 7 of this volume), and 11 bone lools.
TAPHONOMIC FEATURES OF THE MEMBER 3 FOSSll ASSEMBlAGE As indicaled in Chapter 1 of this volume, the Member 3 sedimenl accumulated in a roofed gulley, eroded between the west wall of the cave and a vertical bank of calcified lower Bank and Member 2 sedimento The Iloor of the gully was inclined upwards towards the entrance, positioned over the

TAPHONOMIC OVERVIEW Of FOSSIL ASSEMBLAGES

;ave's southern wall, although sedimentation in the back part )f the gully was more or less horizontal. The preserved part of :he gully is over 20 m in length and up to 5 m wide. Excavated sediment from the Member 3 deposit provided 3n assemblage of 59 488 fossil bone pieces, excluding microfaunal bones. Considering mammals with an adult live weight in excess of 5 kg, the fossils come from a minimum of 176 individuals, as detailed in Table 1. Although hominids are proportionately less well represented than in the earlier units (see Fig. 8), baboons still make up 8,5 % of the mammal total. Most abundant were ungulates at 51,1 %, followed by carní'lores, making up 18,8 %. Carnivore involvement in the boneaccumulation process is suggested by the presence of carnivore tooth marks on 197 bones, or 0,49 % of the sample examined. Although stilllow, this percent incidence is twice as high as that in the two earlier units, a possible reason being lhat the bones in this assemblage have been less fragmented than were the earlier ones, perhaps due to more rapid accumulation and less lrampling. Carnivore coprolites were not found in the Member 3 collection. Porcupine gnawing was seen on 26 specimens, indicating that these rodents had a small role in the bone-collecting process. Hominid activity is reflected throughout the six-metre excavated profile of Member 3, although stone artefacts, 72 in . number, are less abundant than was the case in the Lower Bank and Member 2. Bone lools, on Ihe other hand, were more abundant, wilh 40 having been described in Chapter 8 of this volume. Of special importance is the abrupl appearance of burnt bones throughout the excavated depth and in 17 grid squares of Member 3, totalling 270 pieces (Brain and Sillen, 1988) and further described in Chapter 10. On the basis of the analysis and interpretation provided there, we concluded that fires were tended within lhe Member 3 gully, or its entrance area, on repeated occasions as the sediment gradually accumulated. In addition lo the sudden appearance of burnt bones, however, evidence of hominid meat-eating also makes an unprecedented entry. This takes the form of cut marks observed on 14 fossil bone pieces, and chop marks on a further two. Details of these specimens are as follows: Miscellaneous cut-marked skeletal parts: SKX 23222: proximal radial piece Irom an antelope of size class 11; grid square: W3/S3 NE; depth: 200-210 cm; date found: 20 June 1984. SKX 24414: rib fragmenl; grid square: W2/S4 NE; depth 230-240 cm; date found: 19 September 1984. SKX 28966: vertebral process; grid square: W3S2 NW; depth: 240-250 cm; date found: 20 Janua.ry 1985. The specimen has been heated in a fire and is placed in the 'brown' category. Cut-marked bone 11akes: SKX 23189: 1-2 cm in length; grid square: W3/S3 NE; depth: 200-210 cm; date found: 20 June 1984. SKX 24144: 3-4 cm in length; grid square: W2/S4 SW; depth: 220-230 cm; dale found: 12 September 1984. SKX 24351: 3-4 cm in length; grid square: W2/S4 SW; deplh 230-240 cm; date found: 19 September 1984. SKX 24400: 1-2 cm in length; grid square: W2/S4 NW; depth

'Q

261

l_IIi:::=::::;:j'_ _

S'l(}( 26164a

Fig.4 a: abone flake trom Member 3 showing cut marks (arrows), apparently • caused by a sharp-edged stone artefact, during hominid meat-eating. b: SEM detail of one 01 the cut marks. Scale bar = 1 mm. Pholograph by Pal Shipman.

230-240 cm; date found: 19 September 1984. SKX 24494: 4-5 cm in length; grid square: W2/S5 SW; depth 230-240 cm; date found: 28 September 1984. SKX 24497: 3-4 cm in length; grid square: W2/S5 SW; depth 230-240 cm; date found: 28 September 1984. SKX 26164a (Fig. 4): 6-7 cm in length; grid square: W3/S2 NE; depth: 200-210 cm; date found: 21 March 1984. SKX 30767a: 1-2 cm in length; grid square: W3/S3 NE; depth: 340-350 cm; date found: 13 March 1985. The specimen has been healed in a fire and is placed in the 'buff' category. SKX 33441: 6-7 cm in length; grid square: W2/S4 NW; depth 410-420 cm; dale found: 12 June 1985. The specimen has been heated in a lire and is placed in the 'grey' category. SKX 34315: 2-3 cm in lenglh; grid square: W4/S2 NE; depth: 320-330 cm; date found: 17 July 1985. The specimen has been heated in a tire and is placed in the 'brown' category. SKX 35444 (Fig. 5): 10-11 cm in lenglh; grid square: W4/S3; depth: 490-550 cm; date found: 1 August 1985. Chop-marked bones SKX 22023 (Fig. 6): abone flake, 4 cm in length, with several dístinctive marks apparently caused by a sharp-edged stone

..

-

-

-

~-".--_._--------------

TAPHONOMIC OVERVIEW OF FOSSIL ASSEMBLAGES

262

_t===-...\

a

.J

SKX 22023 Fig.6 Abone IJake Irom Member 3 showing several chop marks (arre apparently caused by a sharp-edged slone artefact, during hon meat-ealing in the Swartkrans cave.

Fíg.5 a: abone flake from Member 3 showing cut marks (arrows), apparent\y caused by a sharp-edged slone artefact, during hominid meat-ealing. b: SEM detail of ane al lhe cut marks. Scale bar = 1 mm. Pholograph by Pat Shlpman.

artefact; grid square: W3/S2 SW; depth: 170-180 cm; date found: 14 March 1984. SKX 23583 (Fig. 7): abone Ilake, 4 cm across, with a straightsided faceto apparently the result of having been chopped with a stone artefact; grid square: W2/S4 NW; dep1h: 200210 cm; date lound: 1 August 1984.

BEHAVIOURAL IMPLlCATIONS OF BURNT AND CUT-MARKED SON E IN MEMBER 3 As discussed in Chapter 10 01 this volume, the distribution of burnt bones throughout the Member 3 profile suggests that fires were tended. either withín the gully itself, or in the immediate calchment area of its entrance. The tactthat burnt bones have been found in up to 23 consecutive excavation spits in the vertical sediment profile, indicates that fire-tending was a repetitive activity during Member 3 accumulation times. This evidence forms a contrasl to the virtual lack of burnt bone in the two earlier stratigraphic units: the Lower Bank of Member 1, and Member 2, where abundant evidence 01 hominid occupation, in the form 01 stone and bone artetacls, has been found.

The interval separating the accumulation times 01 Membe and 3 is assumed to have been a glacial episode, during w the Member 3 gully was eroded. tt seems likely that the al to manage fire in the vicinity 01 Swarlkrans was achievel hominids duríng lhis non-depositional interval, althOugh nature of the slone lechnology, according to Desmond CI. analysis in Chapter 7, is not appreciably differenl bew,' Members 2 and 3. Likewise, no striking faunal difference~ apparent in the fossil assemblages available from these uníts. On the basis of faunal composition and stone cul level, il is likely that the age of Member 3 is in the vicini one million years. Although it has been possible to estímate temperatur( which Member 3 tossil bones had been heated, no evidt is forthcoming on how the camp-fires had been started . unsubstantiated suspicion is that the early stages of management would have relied on galhered tire from natl occurring conflagrations. At present, the onset of the sun rainy season on the Transvaal highveld in October or No' ber is usually characterized by violent late-afternoon stc accompanied by lightning, which Irequently igníles the ~ that has dried out during the rainless winter months. : grass lires could well have been the source of lire taken to the Swartkrans cave by hominids who sheltered ti Although the gathering of fire is not as impressive an int( tual achievement as deliberate tire-making. il must neve less have been an immensely important milestone ir manipulation 01 the natural envíronment by early homi Perhaps the most ímmediate benetit that a camp-lire sleepíng site would have ímparted was a measure of protE from nocturnal predators, conslantly on the lookout for prey. On lhe basis 01 the Member 3 fossil assemblage c1ear that leopards. sabre-toothed cats 01 lhe genus MI tereon, and a variety ot hyaenas frequented the cave a

TAPHONOMIC OVERVIEW OF FOSSIL ASSEMBlAGES

263

60

Ungulates

50

40

30 20

... ·0

..... "0

10

Carnivores Hominids

o+---.------,-------,----~'---

M1

M2

M3

M5

Fíg.8

SKX 23583 Fig.7

Abone Ilake from Member 3 showing a slraight-sided chop mark (arrows), apparenlly caused by a sharp-edged stone artefact, during hominid meat-ealing in the Swartkrans cave. time. The presence 01 a camp-fire in the cave's entrance could well have served to protect the hominids from the attention 01 such predators. There is some evidence that hominid mortality at Swartkrans did, in lac\, decline during the accumulatíon of Member 3. As shown in Fig. 8, hominids made up an average of 16,3 % of mammal individuals in the assemblages from Members 1 and 2, while this ligure declined to 5,1 % in Member 3. It had declined fúrther to almosl zero in the very much later Member 5 assemblage, although this, too, is assumed to have been a leopard lair accumulation. The presence 01 bones, including four burnt ones, bearing cut and chop marks, as documented aboye, sUggesls that hominids were using stone tools for the removal 01 meat from bones while sitting round theircamp-lires. Evidence 01 hominid meal-eating at lhe Swartkrans cave before Mp-mber 3 times has not been lound; it is possible that the bringing 01 meat to a cave, without the protection 01 fire, would have invited lhe attention of carnivores. It would be interesting to know whether Member 3 hominids were eating meal that they had scavenged, or whether they hunted it themselves. In this regard, Shipman (1986) has suggested thal the position 01 cut marks on long-bones can be instructive in deciding between scavenged or hunted meat: a concentration 01 cut marks near joints indicates systematic butchering following hunting, while the presence 01 cut marks in mid-shaft posítions suggests opportunistic utilization 01 scavenged carcasses. The great majority of the Member 3 cut marks are on mid-shaft bone Ilakes, suggesting that scavenging was probably the means 01 meat acquisition during Member 3 times at Swartkrans, although the sample size 01 cut-marked bones ;s very small.

Adiagram showing the percent representation 01 hominids, carnivores and ungulates in the lossil assemblages from Swartkrans Members 1, 2, 3 and 5. While the representation 01 carnivores remains rairly constant through the four Members, lhat 01 hominids declines mark· edly in Members 3 and 5, while that of ungulates increases. It is conceivable (hat (he decline in hominid mortality in Member 3 was linked to a reduction in carnivore predation, following the protective influence of fire-management.

With the presence of two ditterent hominids in the vicinity of Swartkrans during the accumulation 01 the earlier Members, the question remains as to which 01 these was responsible lor the tending 01 fires at the cave and the eating of meat there. Strangely enough, the only hominid taxon currently known from Member 3 is Australopithecus robustus, 01 which nine individual s are represented in lhe lossil assemblage. Horno erectus remains have been found in both the earlier stratigraphic units, Members 1 and 2, so it seems reasonable to assume that these people were present during Member 3 times, although no skeletal remains have been tound. The conventional view is that Horno erectus was intellectually more advanced than Australopithecus robustus and that early humans would have had the added insight to manage lire at their sleeping cave, in contrast to their auslra/opithecine relatives. The Swartkrans investigation has been characlerized by surprises, so an australopithecine fire-tender would nol be out of character tor lhis remarkable cave, particularly in the light of Randall Susman's conclusion (1988) thal robust austra/opithecínes had hands adapted lor precision grasping and the use 01 tools.

TAPHONOMIC FEATURES Of THE MEMBER 5 fOSSIL ASSEMBLAGE

In comparison to the earl;er fossil assemblages from Swartkrans, that Irom Member 5 is bolh small and very lale in time. Its 21 324 fossil pieces came from a mínimum 01 160 mammals with live body weights in excess 015 kg, and the assemblage ;s strongly dominated by antelopes and hyraco;ds. Wilhin the antelope component, remains of the small extinct springbuck, Antidorcas bondi, are especially common. I concur with the conclusion reached by Virginia Watson in Chapter 2 that this assemblage consists very largely of leopard food remains, lollowíng predation particularly on hyracoids and

TAPHONOMIC OVERVIEW OF FOSSIL ASSEMBLAGES

264

springbuck. Baboon remains are not common, suggesü¡;¡g that lhere was not a sleeping site al Swartkrans during the accumulation of Member 5 sediment, while remains of only one human are present. The Member 5 sediment accumulated in a series of

channels ramifying through the calcified cave filling CIOSl the north wall. Such channels would have been dark : unattractive as human sheltering places, lhough they app to have been used by leopards, and possibly hyaenas, feeding or breeding lairs.

Taphonomic reconstructions at Swartkrans have emphasized the importance ot feline predalion on early hominids. al a stage when human influence on the natural environment was very slíght. and when the balance 01 power lay with the cats. Butthe Swartkrans record also documents a technological innovation of immense importance: the management of fire, providing a measure of protection from nocturnal predators. In Member 3 at Swartkrans, a hint is discernible of a shift in the power balance towards the hominids - the firsl in a series of technologicallybased triumphs Iha! have established people as lhe dominant animals on earth.

A reconstruction in lhe Transvaal Museum by Imagen Berry, showing hominids al lhe enlrance to Ihe Swartkrans Member 3 gully, wílh a camp-fire Ihal they had learned lo manage.

REFERENCES BRAIN, C. K., 1981. The hunlers or the hunled? An introduclion lo Atrican cave taphonomy. Chicago University Press, Chicago. BRAIN, C. K. and SILLEN, A., 1988. Evidence from the Swartkrans cave tor Ihe earliest use o/ fire. NalUre 336: 464-466. SHIPMAN, P., 1986. Scavenging or hunting in early hominids: theoretical framework and lests. American Anthropologist 88: 27-43.

SUSMAN, R. L., 1988. Hand of Paranlhropus robustus from Me 1, Swartkrans: fossil evidence lor 1001 behaviour. Seienee 781-784. VRBA, E. S., 1976. The fossil Bovidae ofSterkfontein, Swartkran Kromdraai. Transvaal Museum Memoir No. 21. Transvaal Mu~ Pretoria.

Index A Acacia karoo, 230 Accipitndae, 73 Acheutean. xi. 168, 170, 175 Acinonyx jubatus. 36. 39, 52, 154 Addis Ababa, viii eoustus. Canis. 155 Afar Rift, 175

atarensis, Auslralopithecus, 106, 107, 110113,131,136 afer, Orycteropus, 36, 39, 42, 48, 54 atricaeaustrahs, Hystrix, 36. 40, 44, 50, 57,

63,227 African apes. 117, 121, 129 africanus, Australopithecus, 1, 106, 107,

110-112,115 alba, Tyto, 65 Alcelaphinae, 43, 45, 49, 55, 56. 62 Alcelaphini, 62

Alcelaphus, 36, 40, 43, 49, 55, 56, 62, 69, 252, 253

buselaphus. 230, 244 amino acids, therrnal degradation, 249 Amphibia, 41, 45, 51, 59, 64 Anatidae, 73 Anderson, E.. 163 Andrews, P., ix animal skin. working ot, 4, 215 Antarctica, 5 Antidorcas, x. 36, 40, 43, 45, 63, 67, 69, 208 Antidorcas australis, 36, 40, 43, 49, 56, 62,

66,68,69 aqe-at-death cateqories. 68 australislmarsuplalis, 36, 43, 49, 56, 62

bondi, 4, 33, 36, 40, 62, 64, 66, 68, 69, 140,220,224,263 aqe-at death cateqcnes. 68 marsuplalts. 36, 62, 66, 69 recki, 36, 40, 49, 56, 69 'Antidorcas bond¡ channel,' ix, 4, 137, 227 Antifopinae, 43 Antilopinl, 49, 56, 57, 62 AntilopinilNeolragini 44 antiqua, Procavia. 36, 40, 42, 48, 54, 64,

71,72.218,219,221,224,252,255 Aonyx capensls. 36, 39, 42,46,52,59,

151,155,157-159,162,164 cranial dimensions, 158 Ardeidae, 73 Ardrey, R" 2, 9 Arnct. A. A., vi artetacts, vi, ix, xi-xii, 32,175,180-193,259 lilhie, ix, xíi Artiodactyla, 43, 49, 55, 62 Asfaw, B., 131 astnus. Equus, 142 Atilax, 36. 39, 48, 54. 59 australis, Antidoreas. 36, 40, 43, 49, 56, 62,66,68,69 australopithectne. xi, 2, 13, 23, 33, 113, 114,115,126,243,263

graeile, 131, 175 Aus!ralopithecus, 1, 21, 106, 194,256 (Zinjanthropus) botsei, 116, 194 aíarensis, 106, 107, 110-113. 131, 136 atncanus. 1, 106, 107, 110-112, 115 robustus, 3, 10, 12, 19,33,36,37,39,

45,51. 70, 106, 130, 175, 176, 229, 236, 238, 240, 242, 251, 253-255, 259,263 Avery, M., 35 Aves, 44,50, 58, 64, 70, 71,221,222 Member 3,72 B

Baard's Quarry, 150 bartowl. Dmoteñs. 258 Beatraqus. 36, 40, 43, 49, 56, 62, 69, 204 Beaumont, P., 254 Behrensmeyer, K., 3 aetomc, R. V., 229 Berry, l., 19,214,215,264 Berta, A, 163 Bertram, J. B., 65 8eth, L., 115 Binford, L. R., 17,65 biostratigraphy, viii Blaaubankspruit, 34, 36, 67, 167,257 81itz, R. M., 244, 245 Btumenschine, R. J., 226 boisei, Australúpitheeus (Zinjanthropus),

116,194 Paranthopus, 106, 107, 110-112, 113 Bolt's Farm, 164,258 bondi. Antkíorcas. 4, 33, 36, 40, 62, 64, 66, 68,69,140,220,224,263 bone. accumulating agents, 2, 3, 64, 65,

257 artiñclaüy heated, 19 assemblages, macrovertebrates. 35 bird, 44, SO, 58. 64, 71, burnt, ix, xil. 4, 18,19,34,173-175,

190,214,208,209,229.231,233235.237-249,261-263 behavioural implicatlcns. 262 categories, 231 chemical charactertzanon, 243 dimensions of speclmens. 237, 238 dsitribution in Member 3, 238. 239 ust ot soecenens. 234 taxa represented, 237 chop marks, x.ii, 240, 261,263 controlíed heating, 231 cultural use ot. 16 cut marks, IX, xii, 173, 175, 240, 242.

261,262 behavioural impticatlcns. 262 damage marks, x', 71, 217, 228 chewed edges, 217, 220, 222, 226,

260 digested, 25, 217, 225, 227, 228 gna wed,3.65,223,225

tnsect damaqe, 71,72,225-228 Member 1 Lower Bank. 217 Member 2, 217 Member 3, 218 Member 5, 218 ocrcuotoe-cnaweo. 223, 227 punctures. 219 rOdent-gnawed, 70, 71, 72,205,217, 223-225, 227

tooth scratcnes. 217 digging tools, 195, 209 experimental heating, 232. 241 tlakes. awl-llke tools. 213 digging tools. 199, 204, 209 high polish, 202 goat, 3 modern, vi, 244 regurgitated.228 rubbing on animal skin, 199, 202-204, 206,207,210.211,214,215 stains. 137 toots. vi, xii, 3, 4, 14,35,131,145,174,

195-198,200-204,213-215,240, 242, 259-261 anvil, 196,215 awl-like, 213, 215 ethnographic, 215 experimental, 196 -199 horneare, 198, 203, 207 horse mandible, 209 Member 1 Lower Bank, t 98 Member 2, 203 Member 3, 207 Bovidae, 43, 49, 55, 62, 71 habitat requirement categories, 69 minimum numbers in Members 1·3 & 5.

68 predator/prey scenanc, 65 sue ciasses l-IV skeletal parte & MNI,

65-66 Bovinae. 44, 49, 57, 63 Brain, e., vi, 3,16,197 Braln. e. K., vii-Jx. xi, xli, 1, 7, 8, 10, 11,

21.23.24,26,27.30,32,35,45,59, 64-66,68,72,73.75,79,107,115, 130, 131,136,137,151,160,163,164,167, 171,173,175,176,194,195,217.223.

227,229,238,243,249,253,255-257, 260, 261 Brain. L. H., vi, xn, 10,241 Brain. T. M., vi, 14, 19, 117, 136,241 Bretz, J. H., vii Brock, A., 33 Brooks. A. S., 248, 249 8room, R, Ix. xi, 1.7.23,75,106,117,

126,137 Broom-Robinson excavaticns. 2, 7, 9, 24,25,70,259 broomi, Makapania, 258 brown breccla. 24, 26 Brown. F., viii, 176 browsers, 252, 254

ro.

INDEX

266

brunnea. Hyaena, 36, 39, 41, 46, 52, 152, 154,157, 159, 160 BuckJand, D., ix Budzikiewicz, H., 246

Bunn, H. T, 168.217 burcheüí. Equus, 36, 40, 55. 59, 62, 137,

138,140,143--150,218,222,252,253 burchellii, Equus, 150 Burckle, L.• 5 buselaphus, Alcelaphus, 230. 244 Bushman, 126 butchery practices, xii, 64, 175, 263 Butzer, K. W., vi, ix, 2, 13,25, 75 Buxton, vii Limeworks, 150

C cabanus, Equus, 142 Cainozoic Era. 4 Calvin-Benson pathway, 251 cerne-n-es. 2, 208, 214, 229, 243, 262,

263,264 campestre. Raphicerus, 36, 40, 44, 49, 56, 63,71,72 Canidae, 36, 41, 46, 52, 59. 60, 155, 159, 162

Canis. 155 adustus 155 mesometas. 36, 39. 41. 46, 52, 59, 71,

72,155,157-159,162,163,218 cranial dimensions, 157, 158 canniballsm. 2 capensts. Aonyx, 36, 39, 42, 46. 52, 59,

¡. ¡

r j

¡

¡

151,155,157-159,162,164 Equus, 36,40,43,48,55,59,62,64, 137-140,143--150,221,235,252,253 Precavía, 36, 40. 60, 64, 218, 220, 224 Telanthropus, 23, '25 caracal, Feñs. 36, 39, 41, 45, 162 carbon, 13C¡12C_ratios, 252, 253, 255 C3 planta. 251, 253-255 C4 ptants, 251,253-255 oecenese. 252 isotopes. 20, 251-253 carnívora, ix, 41, 42, 46, 48, 51, 54, 59, 60,71,151 Member 1 Lower Bank, 159 Member 2, 157 Member 3,151 carnivore damage, 64, 65, 71, 217, 226, 227,258,260,261 implications of camlvcre-lnduced marks, 226 lair, 73. 259 predaton. 263 carrying bags, 214, 215 Celtis, 19,230,231 kraussiana 230 Cercocebus, 253 Cerccpfthecoidea, ix, 37, 46, 51, 59,70, 218,221 Cercopithecoides, 36, 37, 39, 45 chama, vuoes. 36, 39, 52, 155 char, 229, 243--246, 246, 249 chemistry, 245 recovery, 244 Charlton, R.. vi

Chasmaporthetes. x. 151, 152, 158, 163 kani, 164 nitidula, 36, 39, 41, 46, 52,151,153, 157,159,162,163 cranial dimensions, 153 ossifragus, 164 Chatwin, B., 18 Chelonia, 41. 45, 50, 58, 64 Chesselet, L, sea 8erry chimpanzee. 126, 128, 128, 175 Chiroptera, 70, 69, 171, 82 Churcher, C. S., vi, 8, 68, 75,131,137, 138,142,149,150 Clark, B. C., 194 Clark, J. D.• vi, xi, 14, 75,131,167,262 Clarke, R. J., vi, vil, 11, 15, 21, 176 ctaytcn. W. D, 251 Columbidae, 73 Connochaetes, x. 35, 36, 40, 43, 49, 55, 62,69,71,72,252,253 Cooke, D., 18 Cooke, H. B. S., 18, 164 Coppens, Y., 75, coprolites, x, 42, 48, 65, 259- 261 Ccrvtcae. 73 ccstata, Hypoxis, 195-197 crassidens, Panthera. 165 Paranthropus, 106, 107,111,112 creer N. C., 118, 119 crenatidens, Homotherium, 153 crtstatus, Proteles. 36, 39, 59, 64, t 60, 161,164 Crocuta crocuta. 36, 39, 46, 52, 152, 158, 159,165 Cryptomys, 70 cultridens, Megantereon, 36. 39, 52, 151, 153,162,165 Cynictis, 326, 238 penicillata 36, 39, 48, 54, 157, 159

D

Damaliscus, x. 36, 40, 43, 49, 56, 62, 29, 218 dorcas. 199 Dandelot, P., 67 Dart, M., 16 Dart, R. A., 1-4, 16,243 darti ríaruell, Slmcpithecus. 252 Theropithicus, 252, 254, 255 dating, cyclic nick-point migratian, 33 Geomagnetic ReversalTlme Scale, xi macrovertebrate biostratigraphy, xi magnetic imprinls, xi poUen successions, xi radlocarbcn. 4, 64 raoromemcztsotoptc rnethods. xi thermofuminescence, 151 De Brul, E. L., 253 De G raaft, vi De Lumley, H., viii Deacon, H. J., 14 Deaeon, J., vi Delson, E., viii, 45. 68, 253 DeNiro, M. J., 244, 251 Dentan, G.• 5 Developed Oldowan, xi. 170, 174, 175, 168

dietary patterns, 20,131,249,253 ancient. 251 edible bulbs, 3. 195 meat-eatlnq, 131, 241, 261-263 stable carbón isotope inslghts, 255 Diller, H., 164 Dinotelis, x. 153, 258 barlowi, 258 Oippenaar, N. J., vi. 68 dorcas, Damañscus. 199 oorst. J., 67 Oreyer, A., vi, 68 Ou Toit. A. L.. 4 Dunbar, E. P., 254 Dunbar, R. 1. M .• 254

E East Turkana, 168, 175 Eck, G. G., 131 Efremov. ix Elandstonteln. 138. 150, 162 Eldrige. M.• 17 Eldriqe. N., 17 elepbant shrew. 70 Eíepnantidae, 42, 54 Elephas, xi, 36, 39, 42, 54, 59 environmental reconstruetion, 67 Epstein, S., 251 Equldae. 43, 48.55,62,71.137 measurements, cranial & postcranlal. 143 Member 1 Lower Bank, 138 Member 2, 141 Member 3. 142 Members 2-3 interface, 147 Member 5, 137, 148 table SKX specimens, 139 of taxa and MNI, 149 Equus, 137, 138, 141,142,147,150,25: asinus, 142 burcheltii, 36, 40, 55, 59, 62,137,138, 140, 143--150, 218, 222, 252, 253 burchetlii burcheüii. 150 cabañus. 142 capensis. x, 36, 40, 43, 48, 55, 59, 62, 64,137,138,139,140-150,221,2, 238,252,253 oldowayensis 150 quagga, 59, 149, 150 quagga antiquorum, 150 taxonomy & distribution, 149 zebra, 150 zebra zebra 142 Equus Cave, 150, 163, 253 Erasmus. M. C., vi erectus, Hamo. 4. 23, 25, 117, 126, 130, 175,176,194,229,263 Ericson, J. H., 251 Eurygnathohippus, 138 Ewer, A. F., 151, 152, 154, 155, 163 Extensfon pit, viii

o,

F Falconldae. 73 faunal succession, 149

267

INDEX

feeding experiments, 3 taír, 260

Felidae, 41, 46, 51, 59, 158, 161 Felinae, 51 Felis caracal, 36, 39, 41, 45, 162 lybica. 36, 39, 52, 59 serval, 36, 39. 41, 45, 162 tire, control of, 59, 64, 229 experimental, 2, 19, 22. 32, 230, 231, 241 grass, 262 lightning-jnduced, 4, 238 manaqernent, xñ, 4, 64,175,176,214,

229,240,241,243,261-264 recurring, 240

wood,229

v..

FitzSimons, 1 Fratley, D.• 163 trancoñn, 41, 44, 58, 71-73, 218 Freedman, L., 252 Frick. C" vi Friedli, H., 251 trugivory. 253, 255 G Gailiano, H., 163 Gazella, x. 36, 40, 43, 45, 49, 56, 59, 67, 69.225 getada, Theropñnecus, 252, 254 Genetta tigrina, 36, 39, 54, 157 geomorphological stucñes. vfií Gerrns, J. B., 5 Gilbert, 8. M" 72 Gilchrist. R_, 229 Giraffidae, 36, 49, 55 Gana River, 175 GoodaJJ, J.. 175 qoríuas. 29, 120 graminivory, 255 grazer, 252, 254 Gnne. F. E., vi, viii. ix, xi. 4, 12,20,68,75, 76,79,82,92,106,107,126,130,131, 136, 253 Grobbelaar, 8., 8 guineafowl, 71, 72, 73 Gwinnett, A. J., 79 H

habilis, Horno, 131, 136, 175, 176, 194 Hadar, 175 Haltenorth, T., 164 hamadryas robinsoni, Papio. 36, 37, 39, 45,46,51,251,254,255 ursinus, Papio, 36, 39, 59, 195 Hanging Remnant Member 1, vlli. 2-4, B, 12, 15, 25, 27, 28, 31-33, 75, 95, 106, 107,137,150,163,177,178,213,238, 259 Hare, P. E., 249, 251 Harris, J. W. K., 175,229 Harts valley, vii, vii( Hassan. A. A., 251 Heidelberg Man, 23 Hendey, Q. B., 151,160,162,163 Henneberq. M., vi Herholdt, E. M., vi, 68

Herpestes ichneumon, 36, 39, 42, 162 Herpestinae. 48, 54, 60, 155, 159 HiII, A. P., 3 hillside. significance 01, vii, 3, 16, 23, 33, 67,195,197 Hippanon, x. 59, 137, 138, 141, 142, 145, 147,150 lybicum, 137-139, 142-144, lS0 Iybicum steyt'ert, 36, 40, 43, 48, 55, 59, 137,138,141,142,145-147,149, 150,209,223 Hippopotamidae. 43, 55 Hippopotamus, x, 36, 40, 43, 55 Hippotraginae,SO Hippotragini, 63, 71 HippotraginilReduncini 50,57,63,71 Hippotragus niqer, xi, 36, 40, 50, 57, 63, 69, Hoering, T., vi, 243 Hoiocene, 33 homlnid. cranlodental tossils. 75 Member 1 Lower Bank, 76 Member 1 Hanging Remnant, 95 Members 1-2 interface, 82 Member 3, 92 tooth dimensions, 108 pcstcrantal rossus. 117 tabte ot tossits. 118 Member " 117 Member 2, 126 Member 3, 129 taxonomy, 130 Horninidae. 36, 45, 51,59 Horno, vili, xi, 1" 36, 37, 39,45,51,59, 82,92,105,107,115, 118, 130, 131, 175, 176, 194, 253, erectus. 4, 23, 25,117,118,126,129, 130, 175, 176, 194, 229, 263 erqaater, xi habilis, viil, 131, 136, 175, 176, 194 sapiens, 36, 39, 59 Homotherium, x crenaudens, 153 Hooijer, D. A., 137, 142 Howell, F. C., vi, 11,106,107, 175 Hughes, A. R., vi, vii, viii, 10, 11, 18,21, 175 Hyaena, 158 brunnea, 36, 39, 41, 46, 52, 152, 154, 157,159,160 crania! dlmensiQos, 154 Hyaenldae. 36, 41, 46, 52, 59, 71, 151, 152, 157-159, 161,224 Hyaerunae, 41 , 46, 52, 59 Hypoxis, 16, 195-197 costata, 195, 197 nqiduta. 195 Hyracofdea, 42-44, 48, 50, 54, 57,60, 63 Hystrlcidae, 44, 50, 57, 63 Hystrix atricaeaustraüe, 36, 40, 44,50, 57, 63, 227

Ibis, 41, 44, 58, 64, 72 Icnneurnoo. He-pestes, 36, 39, 42, 162 tncurva. Panthera panthera 154 Industrial Complexes, 176

ingens, Papio (Dinopithecus), 36, 37, 39, 251,252,254,255 Inner Cave, Ix. 24, 26-28, 70 Interface, Members 1-2, 82 Members 2-3, 147 Isaac, B., 14 Isaac, G. L/., 14, 168

J Johanson, D, C" vi,viií,21, 75,113,131,175 Johanson, L, Vi, 21 John Nash Nature Reserve, 195 Jolly, C. J., viii. 253, 254 jonesi, Parapapio. 251,252,255,257 jubatus. Acinonyx, 36, 39, 52, 154 Junaers. W. L., 107, 115,131 K Kaiso, 150 Kalambo Falls, 168 kani, Chasmaporthetes, 164 sarco. Acacia, 230 Kay, R. F., 253 Kemp, A. e., 72 Kemp, M. r., 72 Kimbel, W. A., 75, 113, 131 King, L. c., 1,23 Kitching. J., vi Klein, R. G., vlit, 163 Kleindienst. M. R., 168 Koch, P. L., 252 Koobi Fcra, 168, 175 Kovacs, J., 256 krausstena. Oeltla. 230 Kromdraai, vii, 3, 106, 107, 149, 155, 160, 161,164,257 Paranthropue, 106 Krueger, H. W., 245, 251 Kruuk, H., 163 Kurtén, B., 163

L Laetoli, 150 Lagomorpha, 36, 44, 50, 57, 63, 71, 221 Langebaanweg, 150 Lanham, F., 256 Le Geros. R. Z., 251 Leakey, L. S. 8., vii, 131 Leakey, M. D., xi, 1" 168, 175, 176 Ledebcuna. 195 Lee-Thorp. J. A., vi, 20, 33, 245, 249, 251, 252,254 leo, Panthera, 36, 39, 41, 45, 162, 165, 252, 255 leopard lafrs. 3, 260, 263 predation. ix, 66, 260, 263 remams, 163 Letty, e., 195 Lewin, R., 17 t.naceae. 195 Uvingstone, D. A, 251 Lobelo. A, vi, 10 Loock, J. c., vi louw, G. N., vi Lowenstein, J. M., 150

INDEX

268

Lower Bank Member 1, ix. 1-4, 14, 15, 16, 17,27,31,32,150,167,195,213,227, 238,241, 26G-262 Lower Cave, ix, 24, 26-28 Lucas, A., 196

Ndessckia.. p., 131 Newman, R. A., vi, 9,16,68,217 nftioula, Chasmaporthetes, 36, 39, 46, 52, 151,153,157,159,162,163

Lutra maculicollis, 155

o

Lutrinae, 46, 155, 159, 162

Masao, F. T., 131 Mason, A. J., 150 Mathabathe, R., 68

Oaldey, K. P., vn, 1,8,24,243,244 Oaks, J. H., 79 Ogden, J., 194 Oldowan, 168, 170, 174, 175 oldowayensis, Equus, 150 Olduvai, H7, 138, 150, 169, 175, 176, 215 Olorgesailie, 168 Oleen, S. L., 215 Omo, vil, viii, 150, 175 orange breccia, 27 Oreotragus, 69 oreotragus, 36, 40, 44, 56, 252, 253 Orycteropidae, 42, 48, 54 Orycteropus afer, 36, 39, 42, 48, 54 oryx, Taurotragus, 36, 40, 49, 57, 63, 69 Osmundsen, L.. vi, 2, 13 ossifragus, Chasmaporthetes, 164 Osteodontokeratic Culture, 1,3 Ostrich, 72, 248 oswaldi danieü. Theropithicus, 36, 37, 39, 46,51 Otocyon rneqatotis. 155 Otomys, 70, 72 Outer Cave, IX, 4, 24-26, 28, 29, 70, 257 Owen-Smith, N., 67 owls, ix. 72, 73 Ozarks, vii

McFadden, P. L., 33 meadowsi, Tapinochoerus, 36, 40, 43, 49,

P

Iybica, Felis, 59 Iybicum, Hipparion, 137-139. 142-144, 150 Lycaon, 155

pictus. 159 Lycyaena, 164 M Machairodontinae. 52, 153

Machairodus, 165 macrovertebrate bcne assemblages, 35 % occurrence 01 higher laxa, 67 in SKA breccia, 72

Maguire, J., 223 Makapania, xi, 258 broomi, 258 Makapansgat, 3, 150, 243, 258

Mammalia, 36, 45, 51, 59 Mangepa, P., 131

Manidae.54 Manis, 36, 39, 54, 59 Mann. A. E., 21,75, 253 Mann, M., 21

marsupialla. Antidorcas, 36, 62, 66. 69

55, 62 megalotis, Otocyon, 155 Megalotragus, x, 36, 40, 49, 51,55,62, 69, 252,253 priscus, 252 Megantereon, x, 59, 153, 258, 262 cultrtdens. 36, 39, 52, 151, 153, 162, 165 Member 1, Hanging Remnant, 95 Lower Bank, 36, 76, 159, 172,217,219, 220,223-225 Members 1-2 interface, 82 Membera 2-3, interface, 147 Menzelaar, H. L. C., 245 mesomelas, Canis, 72, 155, 159, 163,218 rnicrotauna. 35, 70, 72 Middle Stone Age, ix, 4, 33, 168, 175, 195 mining, vii, viii, 2, 7, 9,11,14,23-25,28, 32,33,70,172,201,257,258 Miocene,30 Moenda. G., vi, 2, 13,-15 moisture index, 33 Moshele, O., 8, 24 Mungos mungo, 159 Mustelidae, 36, 42, 48, 52, 54, 60,155, 159,162 mutiwa, Paracolobus, 45 Mytum, C., 229 N Namib Desert, 3 Napier, J. R., 131

palaeodemography, 21 palaeoecology, vi, viii, ix, 2, 117, 164,251, 252 palaeomagnetic studles. xi, 33 Pan, 119 pancos. 122 troglodytes, 128 Panagos, D., vi, 68 panicus. Pan, 122 Panthera crassidens, 165 100,36,39,41,45,162,165,252,255 pardus, 36, 39, 41, 45, 46, 51, 59. 70, 72,151,153,156-158,161,252,254 cranial dimensions, 156, 157 pardus incurva, 154 Papio, 236, 238, 253- 255 (Dinopithecus) ingens, 36, 37, 39, 251, 252,254,255 hamadrayas robinsoni, 36 hamadrayas ursinus, 36, 39, 59, 195 hamadryas robinsonl. 36, 37, 39, 45, 46, 51,251,254,255 robinsont, 254 ursinus, 252, 255 Papicnidae. 37, 46 Paracolobus mutiwa, 45 Paranthropus, xi, 75, 82, 92, 105, 106. 107,110,112,113,117,118,121,126, 129,130,131,136 aethiopicus, xi bcisei. 106, 107, 110-113

crassidens. xi, 106, 107, 111, 112 robustus, xi, 4, 106,107,111,112,11 117,131,136,264 morphology & behaviour (Member 1) 131 Parapapio. 252, 254 jonesi, 251, 252, 255, 257 pardus. Panthera, 36, 39. 41, 45, 46. 51 59,70,72,151,153,156,157,161, 252,254 Partridge, T. C., vi, 5, 14, 33 Peabody, F., vil Pedetes, 36. 40,44, 50,57,63 Pedetidae, 44. 50. 57, 63 peklnensis. Sfnanthrcpus, 116 Pelea, 36, 40, 57, 63, 69 Pelinae, 57, 63 Pellegrino, E. D., 244, 245 penicillata, Cynictis, 36, 39, 48, 54, 157 159 Perissodactyla, 43, 48, 55, 62 Peters. C., 253 Phacochoerus, 36,40, 43,49.55,62 Phasianidae. 73 Pholidota. 54 pictus. Lycaon. 159 plnk breccia. 23-·27 Plug, l., vi, 68 Podlcepkñcae. 73 Pongo, 119 pcrcupme. bone accumulations, 65, 22 gnawing, 223, 225, 260, 261 Port, A., 3 Potts, R., 217 Precambrian Era. 5 precision grasping, 4,131,263 Primates, 36, 45, 51, 59 Prinsloc, L., vi Proboscidea, 42, 54 Procavia, 64, 235, 236, 238, 255 entiqua, 36, 40, 42. 48, 54. 64, 71, 7 218,219,221,224,252,255 capensis, 36, 40, 60, 64, 218, 220, :: transvaalensis, 36, 40, 42, 48, 54, 61 64,164,218,252,255 Procavñdae, 54, 60 Proteles. 36, 39, 41, 52, 59, 152, 160 crtstatus. 36, 39, 59, 64, 160, 161, 1 taxonomy & identification, 160 transvaalensis, 160 Protefinae, 41, 52. 59 Proterozoic sediments. 5, 254 Proto-Levallois. 169. 171, 193 pygmy chímpanzee. 120, 124, 129, 13 pyrolysis, 245-248 Q

Quaternary, xii, 18 quagga antiquorum, Equus, 150 Equus, 59,149,150

R Haath, M. A., vi Haphlcerus, 69, 255 campestrts. 36. 40. 44, 49. 56, 63, "i Rau, R. E., 150

INDEX

Hautenbach, l. L., vi, 66 recki, Antidorcas. 40, 49, 56, 69

Redunca, 69 arundinum, 36, 40, 63 Reduncinae, 63 Heduncinl, 63, 71 Reptilia, 41, 45, 58, 64, 71 Richardson. M. L., 137, 142, 149 rigidula, Hypoxis, 195 Robinson, J. T., vi, ix, xl, 1,7,8,23,25.26, 75,117,215,253 robinsoni, Pacto. 254 robust auatraícplthecine. viii, 20, 75, 106, 107,117,131,175,176,253,263 hominid lineage, 131 robustus. Australopithecus, 3, 10, 12, 19, 33,36,37,39,45,51,70,106,130, 175,176,229,236,238,240,242,251, 253-255,259,263 Paranthropus. 4, 106, 107, 111, 112, 116,117,131,136,264 Hodentia, 44. 57, 63 Rose, J. J., 197, 198,215,217 Rotifera.5 Rubidge, B., vi Ryder, O. A., 150

s sapiens. Horno, 36, 39, 59 Schaller, G. B, 163 Schirge, G. U., vi Schoeniger, M. J., 229, 251 Schopf, W. J., 5

Scñla. 195, 196 marginata, 195, 196 scriptus, Tragelaphus, 252, 253 SeaJy, J., 256 Sellschop, J. P. F., vi, 21 Sepeng, O., vi serval, Felis, 36, 39, 41, 45, 162 Shipman, P., vi, 3, 17, 75, 131, 195, 198, 215,217,229 Shungura Forrnation. 150 Sillen, A., vi, xli, 4, 19, 32, 59, 173, 175, 229,230,241,243,256,261 Simopithecus. 256 darti danfela. 252 Sinanthropus peklnensis. 116 Singer, R., vii Sivatherium. x SKR material, 70 Smith, B. S., 251 Smithers, R. H. N., 67 Soricldae. 70 Stiv, C. G., 253 Sterkfontein, vñ, viii, 2, 3,10,12,23,34, 45,59,129,162/163,164,168,176, 214,215,257,258 Stern, J_ T., 118, 131 steytleri, Hipparion, 149 Iybicum 36, 40, 43, 48. 55, 59,137, 138,141,142,146,147,149,150, 209,223 stone artefacts. assemblages, 76, 167 Acheulian, 175 Oldowan, 174

bec, 169, 171, 182 borer, 170, 171,190 biface, xi. 167, 170 tecnnoioqy, 167 burin, 170, 171 cleaver, 171, 175, 176, 192 chopper, 167-171, 174. 175, 181, 186, 187,191 cores, 168. 169, 170, 171, 175, 180, 185-188,191-193 dimensícns, ot fragments & chunks, 172 ot uakes, 173 flake types 167, 169. 170, 175 hammerstone, 168-170, 196 handaxe. 171, 175, 176 heat fractured, 170, 183 Member 1, 169, 172 Member 2, 170, 172 Member 3,170,173 pick, 171, 192 polyhedron, 169-171, 174, 184 Proto-Levaílols. 169, 171 scraper, 168, 169, 170, 171, 182, 259 stone technology, 2, 4, 35, 131, 214, 261,262,263 table of artefacts, 171 terminolagy & ctassiticaticn. 168 thermal fracture, 170, 183 stratigraphy, ot Swartkrans, 23 current interpretation, 28 historical developrnent of concepts. 23 strepsiceras, Tragelaphus, 36, 40, 57, 69, 252,254 Stringer, C., 18 Struthio sp.. 40, 49, 64 Sturnidae, 73 Suidae. 43, 49, 55, 62, 220 Sullivan, C. H., 245 Suncata aurtcatta, 36, 39, 48, 54, 60, 155, 159 Susman, R. L., vi, xi. 4, 20, 68, 75, 115, 117-119,126,130,131,263 Suwa, G., 131 Swartkrans, age of deposits, 33 agents responsible for bone accurmrlations,64 artefact assemblages, 169 benavioural implications of burnt and cutmarked bone in Member 3. 262 birds in Member 3 , 72 bone, tabje ot numbers excavated, 36 flakes, total recovered. 37 tools, 195, 198 accumulations, 35 burnt bcne, aturnals represented, 237 assemblage from Mernber 3, 234 implications for control 01 fire by early hominids, 229 stzes. 237 spatial distnbuticn. 237 camivore remains, 151 cave structure & stratigraphy, 23 errar, chemfstry, 245 recovery, 244 chemical characterizaticn 01 burnt bones, 243 cñmatic lmplications of depositional sequence,33 companson, ot hominid samples. 105

269

01 Swartkrans and Kromdraai Paranthrcpus. 106 damage marks on bones. 217 environmental reconstruction, 67 Equidae, 137 Formalion, 2, 13, 25, 107 formation ot cave & oepcsus. 30 hlstoncal development of stratigraphic

concepts, 23 hominid cranlodental fossils, 75 postcranlet remains. 117 macrovertebrates, bone component, 35 Member 1 Lower Bank, 36 Member 2,45 Member 3, 51 Member 5,59 SKR, 70 taxa identltled. 39 mammals represented, x, 36 Member 1, age relationships at cornponents. 257 pictoríal overview 1948-1992,7 SKR material, 70 stable carbon isotope studies. 251 stages in formation ot cave & depcslts. 30 stone artefact assemblaqes, 167 structure & stratigraphy, 23 current mterpretation of stratigraphy,28 hlstcrical development ct stratiqraphic concepts, 23 taphcncmic overview of foss!l assemblaqes. 257 taphonomy, Member 1 Hanging Remnant fossil assemblage, 258 Member 1 Lower Bank foss!l assemblage, 259 Member 2 fossil assemblage, 260 Member 3 tcesñ assembfaqe. 260 Member 5 tossu assemblage, 263 Syncerus. xi, 36, 40, 44, 49, 57, 63, 69 Szalay, F. S., 253

T taphonomy, Member 1 Hanging Remnanl, 258 Member 1 Lower Bank, 259 Member 2, 260 Member 3, 260 Member 5, 263 Tapinochoerus, 64 meadowsi, 36, 40. 43, 49, 55, 62 Tarboton, W., 72 Tattersall. l.. viii Taung, vii, v;¡¡, 150 Taurotragus, xi oryx, 36,40,49,57, 63,69 Tei1hard de Ohardin. P., 162 Telanthropus, 25, 26, 126. 130 capensts, 23, 25 Testudirudae, 45, 50, 58, 64 Thabaseek travertlne. vii Thackeray, J. F.. 12, 33, 150. 256 Theropithicus, 254. 256 gelada, 252. 254

n

_ INDEX

270

darti, 252. 254, 255 oswaldi daníeu, 36. 37, 39,46,51 Threskiomithidae, 73 tigrina. Genetta, 36, 39, 54 Tobias, P. V., vi. vü. 2, 10, 12, 13, 75,131, 175 Toth, N., 131, 168 Tragelaphus scriptus, 252, 253 strepsíceroe. 36, 40, 57. 69, 252, 254 transvaalensls. Precavía, 36, 40, 42, 48, 54,60.64.164.218.252,255 Proteles, 160 troglodytes. Pan, 126 Tubuüdentata. 42, 48. 54 Tumer, A., vi, 20, 45, 66, 68, 75,131,151153, 155, 162, 163,259 Tumer, G., vi. 20, 68 Tyson, P., vi Tyto alba, 65 Tytcnldae. 73

u

v Van, Hoeoe». E., 137 Van Couvering, J., viü Van den Heever, J., 21 Van der Merwe, N. J ., vi, 20, 245, 249, 251.252,254,255 ViUatranchian, 162 Viverridae, 36, 42, 48, 54, 60, 155, 159, 162 Viverrinae, 48, 54, 60, 157 Vogel, J. C., vi, 33, 64, 151, 251 Voigl, E. A., vi, 12, 17, 68, 70 vroa, E. S., vi, vil, viil, xi, 2, 5,12,14,17, 24, 33, 68, 151, 163, 253, 258 Vulpes cherna, 36, 39, 52, 155

y Yellow Sands Mursi Formation, viii Yokelson, J., 115

w Walker, A., 20, 175, 253, 254 Waller, M. R., 5

z

waner,

zebra zebra, Equus 142 zebra, Equus. 150 Zihlman, A. L., 9 Zoukoudian deposlts. 162

R. C., 131

Watson, E. J., 68 Uitkomst tarm, 195, 260 ursinus, Paplo. 252, 255

weapons, 1 Weideman, W. J., vi Weidenreich, F., 75 weiqelt, Ix Wells, L. H.• 149 Werdelin, 163 West Turkana, 175 White, T. D., 75, 11D, 113, 131 Wolpoff. M. H.• 18 Wanderkrater peat locality. 33 Wonderwerk Cave, 33. 150 wooo, B. A., 19,253

wetson, v., vi, vii. xi. 9, 19, 30, 33, 35, 75, 131.137,228,241,258.259,263