Journal of the History of Biology, Vol. 1, No. 1

Spring 1968/ Volume1: Number I

Journal of the History of Biology Published by the Belknap Press of Harvard University Press Cambridge Massachusetts

Journalof the Histori of Biology SPRING 1968: VOLUME 1, NUMBER 1 Editor: Everett Mendelsohn, Harvard University Assistant Editor: Judith P. Swazey, Harvard University

CONTENTS Leeuwenhoek as a Founder of Animal ')emography

1

FRANK N. EGERTON

The Founding of Population Genetics: Contributions of the Chetverikov School, 1924-1934

23

MARK B. ADAMS

Trigonia and the Origin of Species

41

STEPHEN JAY GOULD

Sherrington's Concept of Integrative Action

57

JUDITH P. SWAZEY

August Weismann and a Break from Tradition

91

FREDERICK B. CHURCHILL

Thomas Hunt Morgan and the Problem of Natural Selection

113

GARLAND E. ALLEN

First Steps in Claude Bemard's Discovery of the Glycogenic Function of the Liver

141

M. D. GRMEK

Essay Review R. C. LEWONTIN

161

Editorial Board: Bentley Glass, State University of New York, Stony Brook; Hebbel E. Hoff, M.D., Baylor University; Ernst Mayr, Harvard University; Everett Mendelsohn, Harvard University; Jane Oppenheimer, Bryn Mawr College. Advisory Editorial Committee: Enrique Beltrin, Mexico; Georges Canguilhem, France; John T. Edsall, M.D., U.S.A.; A. E. Gaissinovitch, U.S.S.R.; Ralph W. Gerard, M.D., U.S.A.; John C. Greene, U.S.A.; Marc Klein, M.D., France; Vladislav Kruta, M.D., Czechoslovakia; Joseph Needham, England; Dickinson W. Richards, M.D., U.S.A.; K. E. Rothschuh, M.D., Germany; Conway Zirkle, U.S.A. JOURNAL OF THE HISTORY OF BIOLOGY is published semiannually in the spring and autumn by the Belknap Press of Harvard University Press, 79 Garden Street, Cambridge, Massachusetts, 02138. Editorial Correspondence and manuscripts should be sent to Professor Everett Mendelsohn, Editor, Journal of the History of Biology, Holyoke Center 838, Cambridge, Massachusetts, 02138. Subscription and advertising correspondence should be addressed to Christopher D. Reed, Harvard University Press, 79 Garden Street, Cambridge, Massachusetts, 02138. Subscriptions, which are payable in advance, will start with the first issue published after receipt of the order. Please make remittances payable to Harvard University Press. Subscription rates are $7.50 a year; $4.50 for a single copy. Fourth-class

postage paid at Cambridge, Massachusetts,

02138.

Journal Design by David Ford C Copyright 1968 by the President and Fellows of Harvard College

EditorialForeword

Interest within the scientific community in the historic development of the special fields of science probably has been sharpened by the extremely rapid rate of current scientific growth and by the striking nature of the conceptual and technical changes which one witnesses almost daily. A sense of time and history is often now sought by the working scientist and by students in the sciences. The past decade has also seen the history of science emerge as an active field of research and teaching in the universities. New standards of scholarship have been established and new areas of interest have been explored. While the physical sciences have long served as the paradigm for work in the history of science, and several specialized journals have published articles in this field, this imbalance is now being redressed. Many historians of science are now turing their attention to the complex and often challenging problems of the history of biology, and a new generation of scholars has taken biology as the focus for their historical analyses. Contemporary scholarship in the history of science makes changed demands upon the author; these are demands for methodological awareness and realization that other fields of historical study have brought new sophistication to the writing of history. While hard data will always serve as the basis of history, the simple narrative is no longer acceptable, particularly when dealing with the emergence of ideas. The best history will be characterized by penetrating and critical analysis of changing concepts and altered methods of experiment and observation. Biology, in particular, must be studied in terms of its relationships with the other sciences and with the intellectual currents of its day. It may be examined as well for its interaction with the institutions of the society which spawns it. The Journal of the History of Biology will attempt to serve as a forum both for the working biologist and the historian of the biological sciences. All periods of history will fit within its

Editorial Foreword scope, and special attention will be paid to the developmentsof the last half-century. Authors are invited from the laboratory as well as the library. In the first instance the Journal will be published semiannually. Suggestions, comments, and criticism are invited from potential authors and readers alike. Everett Mendelsohn

iv

Leeuwenhoekas a Founderof AnimalDemography FRANK N. EGERTON Hunt Botanical Library, Carnegie-Mellon University,

Pittsburgh, Pennsylvania

Antoni van Leeuwenhoek (1632-1723)1 is one of the most fascinating figures in the history of science, but many aspects of his work have not yet been closely studied. In this paper his important and original contributions to what we now know as animal demography will be described. Although observations on animal populations had been recorded since antiquity, at the time that Leeuwenhoek wrote there was no formal scientific discipline of ecology, let alone that branch of it now called animal demography.2 Nor did Leeuwenhoek formally organize animal demography as a scientific discipline. What he did do 1. Clifford Dobell, Antony van Leeuwenhoek and His 'Little Animals': Being Some Account of the Father of Protozoology & Bacteriology and His Multifarious Discoveries in These Disciplines (1932; 2nd ed., New York: Russell and Russell, 1958). Arthur William Meyer, "Leeuwenhoek as Experimental Biologist," Osiris, 3 (1937), 103-122. Maria Rooseboom, "Antoni van Leeuwenhoek vu dans le milieu scientifique de son epoque," Archives internationales d'histoire des sciences, 12 (1959), 2746. Abraham Schierbeek, Measuring the Invisible World: The Life and Works of Antoni van Leeuwenhoek F R S, with a Biographical Chapter by Maria Rooseboom (abridged trans. from Dutch ed., 2 vols., 1950-51; New York and London: Abelard-Schuman, 1959). W. H. van Seters, "Antoni van Leeuwenhoek in Amsterdam," Notes and Records of the Royal Society of London, 9 (1952), 3645. Francis Joseph Cole, A History of Comparative Anatomy from Aristotle to the Eighteenth Century (London: Macmillan, 1944), pp. 255-270. 2. The terms "ecology," "demography," and "microbiology" are anachronistic when referring to the knowledge of Leeuwenhoek's day, since these subjects had not become formal scientific disciplines. The use of these terms in this paper is always in the sense of what Leeuwenhoek contributed that was later incorporated into these subjects. The term "animalcule" has been used instead of "microorganism" since the former term has often been used by the English translators of Leeuwenhoek. For a general sketch of the history of animal ecology, see W. C. Allee, "Ecological Background and Growth before 1900," in W. C. Allee, A. E. Emerson, Orlando Park, Thomas Park, Karl P. Schmidt, Principles of Animal Ecology (Philadelphia and London: W. B. Saunders, 1949), pp. 1343.

1

FRANK N. EGERTON

was to make a series of contributionsthat in retrospect can be seen to have been an important foundation for the modem development of this science. There is a charm in Leeuwenhoek'slife story that has universal appeal. From humble origins he rose to achieve international acclaim from the scientific world. His scientific education was acquired informally; and it was limited by his inability to read any language except his native Dutch. Furthermore, his investigations were carried out with only slight assistance from, or even interaction with, other scientists. He did, however, achieve a position of such high esteem in the scientific world that the Royal Society of London and individual scientists took the trouble to communicate with him through translators. And since there was not a significant body of literatureon animal demography,his isolation was not a serious handicap to his research relating to this subject. For fifty years Leeuwenhoek wrote scientific letters which reported a wide assortment of observations.Written entirely in Dutch, they were first published in Dutch, English, or Latin. Many of them first appeared, usually abridged, in the Philosophical Transactions of the Royal Society, but some were

published as separate works by Leeuwenhoek or incorporated into works published by others.3 It is a nice question whether Leeuwenhoek'sobservationson population were much influenced by previous discussions. Vital statistics had been widely discussed since John Graunthad first published his Natural and Political observations Mentioned in a following Index, and made upon the Bills of Mortality (Lon3. Leeuwenhoek's bibliography is complicated, and several guides are needed. For the titles of his publications, see the bibliographies in Dobell, Leeuwenhoeh, and Schierbeek, Leeuwenhoek. A bibliography for each of his letters is included in The Collected Letters of Antoni van Leeuwenhoek, edited, illustrated and annotated by a Committee of Dutch Scientists (7 vols., Amsterdam: Swets and Zeitlinger, 1939-64). However, the seventh volume of this continuing production reaches only to August 24, 1688, and it will be many years before the set is complete. Meanwhile, an indispensable guide is by Francis Joseph Cole, "Leeuwenhoek's Zoological Researches," Annals of Science, 2 (1937), 1-46, 185-235. Cole indicated the source of publication of the letters in Dutch, English, and Latin, and he also provided a subject index. The Latin volumes of Leeuwenhoek's letters appeared under several titles in the first edition, but they will be referred to here by the title of the second edition, Opera Omnia (lst ed., 4 vols., 1685-1719; 2nd ed., Leiden, 1722). The only presently available English translation for many of the later letters is that of Samuel Hoole, The Select Works of Antony van Leeuwenhoek, containing his Microscopal Discoveries in Many of the Works of Nature (2 vols., London: Henry Fry, 1798, 1807). Hoole arranged passages topically, without indicating when they were written. However, Cole's article provides dates for these passages.

2

Leeuwenhoek as a Founder of Animal Demography don, 1662). Leeuwenhoek's scientific correspondents knew Graunt's book, and most likely he himself had heard of it, but he could not have read it. It seems reasonable to conclude that he had heard of some discussions of population which might have sharpened his interest in the subject, but that the observations he made concerning population were dictated by whatever biological materials came to hand. Some of his contributions to Animal demography were indirect. His investigations of spermatozoa and parthenogenesis helped eventually to clarify reproduction, with which animal demography has always been closely tied. His discovery of microorganisms was necessary before food and energy chains, with their important implications for animal demography, could be comprehended. He was one of the first to describe food chains, and this important contribution to what we consider ecology was enhanced by his studies of life cycles, age determination, and reproductive potentials. Before he began his sciendfic studies, he had studied mathematics in order to get a surveyor's license, and that background must have given him the confidence to attempt mathematical calculations relating to population.4 Leeuwenhoek recorded his first observations of populations as part of his description of animalcules. In April, May, and June of 1676, he observed the organisms that appeared in water containing pepper, and made some unexpeLted discoveries. "The 24th of May observing this water again, I found in it the oval little animals in a much greater abundance. And in the evening of the same day, I perceived so great a plenty of the same oval ones, that there were a thousand if there was one which I saw in one drop; and of the very small ones, several thousands in the same drop."5 These statements were naturally met with incredulity by other members of the Royal Society, who had never seen microorganisms. On March 23, 1677, he replied to Henry Oldenburg in a slightly defensive tone: "Nor do I wonder, they could not well apprehend, how I had been able to observe so vast a number of living Creatures in one drop of water, that being very hard to conceive without an ocular inspection. Mean time I never 4. E. J. Dijksterhuis has discussed "Mathematics in Leeuwenhoek's Letters," in Leeuwenhoek, Letters, III, 443453; see especially p. 451. 5. Letter dated 9 October 1676. An abridged translation first appeared in the Philosophical Transactions of the Royal Society, 12 (25 March 1677), 821-831. Letters, II, 61-161; quotation on p. 99. All quotations from the Leeuwenhoek Letters are used with the permission of the present editor, Dr. J. J. Swart. Cf. Leeuwenhoek to Constantine Huygens, 7 November 1676. First published in Dutch by J. F. Snelleman, Album der natuur (Haarlem, 1874), pp. 360-362. Letters, II, 179, 181.

3

FRANK N. EGERTON

affirned, that there are so many of these anmals in this water, but I generally said, that I imagined I saw so many."8 Although Leeuwenhoek'searlier statement had not been fanciful, it had lacked precision. After the challenge of the Royal Society, his ingenuity was equal to the task of devising a method for counting animalcules. He reasoned that a drop of water was the size of a pea and that a seed of millet was 1/91st as large. Then he took a fine glass tube and marked divisions on it that he presumed were 1/30th the volume of a millet seed. He estimated that in one of these units there were 1000 of a certain animalcule, and that in the drop of water there were therefore 1000 x 30 x 91 = 2,730,000.7 This technique was very good, but it would be fair to ask how accurate it was. He faced this question and warned his readers that it was only a rough approximation: For my computation is as uncertain as that of those who, seeing a large flock of sheep being driven, tell you, by sight alone, how many sheep there are in it. The most exact manner to do this is to imagine that the sheep walk togetherbroadwise in a certain conjecturednumber, and to multiply this number with the conjectured length of the flock, and thence to conclude the size of the flock. And just as the conjecturednumber of sheep may differ from the real number at the rate of 100, 150 or 200 in the case of a flock of 600 sheep, my computationof the very little animals may differ.8 However, he thought that his estimate would probably be less rather than greater than the real number. But this did not quite satisfy all the members of the Royal Society, because in a letter of October5, 1677, he had to explain how he estimated that the millet seed was to the size of a pea as 1:91, and he enclosed letters from eight men to whom he had demonstratedthe numbers of microorganismsin a drop of water.9 Presumably this last letter satisfied the skeptics, but if these results were widely read and accepted, they were also eventually forgotten. At least, Leeuwenhoek'sdiscussion was apparentlyunknown to William Scoresby,who published in 1823 his own technique for estimating numbers of aquatic microorganisms.10 6. Phil. Trans., 12 (23 April 1677), 844-846. Letters, II, 197-207; quotation on p. 199. 7. Letters, II, 199, 201. 8. Letters, II, 203-205. 9. In Lectures and Collections . . . , ed. Robert Hooke (London: J. Martyn, 1678), pp. 81-82. Letters, II, 253-271. 10. William Scoresby (1789-1857), Journal of a Voyage to the Northern Whale-Fishery; including Researches and Discoveries on the Eastern Coast

4

Leeuwenhoek as a Founder of Animal Demography Techniques of counting are important for certain types of studies of animal population, but beyond that it is interesting to see how Leeuwenhoek's attempt to solve the problems of numerical estimation led him to pay closer attention to population quesidons. He discovered spermatozoa in 1677 (see below), and in a letter written February 21, 1679, he estimated that in cod there were more than 10,000 sperm in the space of a sand grain, and that the males had several thousands of sperms for every egg of the females.11 He returned to this subject in a letter written two months later,,2 in order to demonstrate that the numbers of sperms in a single cod exceeded the population of the world. He estimated that 100 grains of sand placed side by side would measure an inch, and that the volume of the milt of a cod was 15 cubic inches. Therefore, there should be 150,000,000,000 sperms in the milt of a single cod. Then turning to human population, Leeuwenhoek did not attempt to estimate the real population of the world, but only the maximum possible population. He assumed that one third of the earth is land, that one third of the land is uninhabited, and that the inhabited portion was of an average population equal to, or less than, that of Holland. He estimated that Holland had an area of 154 square miles and had a million inhabitants, and that the habitable land of the earth was 13,385 times larger than Holland. Therefore, the maximum population of the earth should be less than 13,385,000,000,13 or about eleven times less than the numbers of sperns he estimated were in the milt of a cod. There were those who doubted the number of sperms which he estimated for the cod, and in 1688 he made a new calculation, of West Greenland, made in the Summer of 1822, in the Ship Baffin of Liverpool (Edinburgh: Archibald Constable, 1823), pp. 353-356. Earlier, Scoresby had reported his discovery of plankton and his less precise estimate of their numbers in An Account of the Arctic Regions, with a History and Description of the Northern Whale-Fishery (2 vols., Edinburgh: Archibald Constable, 1820), I, 176-180. On Scoresby's life, see John Knox Laughton, D.N.B. Robert Edmund Scoresby-Jackson, The Life of William Scoresby (London, Edinburgh, New York, 1861). 11. This letter was addressed to Nehemiah Grew, Secretary to the Royal Society, but it never reached him. It was first published in Dutch in 1933. Letters, II, 409-423; see p. 421. 12. Dated 25 April 1679. In Philosophical Collections . . ., ed., Robert Hooke (no. 1, London: J. Martyn, 1679), pp. 3-5. Letters, III, 3-35; see pp. 25-35. 13. Leeuwenhoek's calculation of the world's population was later discussed by Johann Peter Sissmilch, who then attempted to calculate the world's population by taking into account the unequal distribution of people in different countries. Die g6ttliche Ordnung in den Veranderungen des menschlichen Ceschlechts, aus der Geburt, Tod, und Fortpflanzung desselben erwissen (2nd ed., Berlin, 1742), ch. 3, sec. 23, pp. 76-77, 97.

5

FRANK N. EGERTON

and concluded that there were really thirty times more sperms in a cod than people on the earth.14He also calculated the number of eggs in a large crab-more than 2,000,000, he thought.15

Leeuwenhoek was impressed by these numbers, but he did not speculate on why anmals should produce so many sperms and eggs. His attention was often upon the reproductiveprocesses rather than upon the demographic consequences. His greatest discoveryim the field of reproductionwas the existence of spermatozoa, which he reported to the Royal Society in 1677.18 He believed after 1683 that the individual was already complete within the sperm.17The preformation theory of development had become generally accepted by then, and those concerned with reproductionwere tuming their attention to the question of whether the individualwas preformedin the father's sperm or the mother's ovum.'8 A private letter, dated June 13,

1679, revealed Leeuwenhoek'sdoubts about one of the prime sources which the ovists could cite-Aristotle's famous report of mouse reproduction.'9At the same time Leeuwenhoek revealed some of his own thinking about reproductivepotentials: Nor can I understand how any animalcule can originate without fertilization. I have been told that young unbom mice in their turn have mice in their bellies, but I have always rejected this, for if we state the fact that mice (like our tame rabbits) produce young every month, and that the young mice when a month old are fit for copulation, and that in each cast there are six young, to wit 3 males and 3 females, and that the first cast is in April while the last is in November,eight months altogether,we shall find-this being the case-that one pair of mice can produce about ten thousand mice.20 14. Dated 24 August 1688. Leeuwenhoek, Natuurs verborgentheden ontdeht . . . (Delft, 1689), pp. 237-260. Letters, VII, 343-393; see pp. 389-393. 15. Dated 10 June 1686. Leeuwenhoek, Ontledingen en ontdekkingen ... (Leiden, 1686), pp. 66-86. Letters, VI, 85-123; see p. 115. 16. Dated November, 1677. Latin trans., Phil. Trans., 12 (1678), 104043. Letters, II, 279-299. Arthur William Meyer, The Rise of Embryology (Stanford and London: Stanford University Press, 1939), ch. 9. 17. Francis Joseph Cole, Early Theories of Sexual Generation (Oxford: Clarendon Press, 1930), p. 13. 18. Ibid., pp. 53.57. Meyer, Embryology, chs. 4-5. R. C. Punnett, "'Ovists and Animalculists," American Naturalist, 62 (1928), 481-507. Jacques Roger, Les sciences de la vie dans la pensee frangaise du XVIII' sikcle. La g6n6ration des animaux de Descartes ta l'Encyclop6die (Paris: Armand Colin, 1963), pt. 2, chs. 2-3. 19. Aristotle, Historia Animalium, 580blO-581^5. 20. Addressed to Lambert van Velthuysen; first published in Dutch

6

Leeuwenhoek as a Founder of Animal Demography Leeuwenhoek, in this quotation, rejected parthenogenesis in mice (he later discovered it in aphids, see below) because it was an unnecessary way to account for the rapid reproductive capacity of mice. In his letter, he did not include computations with his statement, and it is not clear how he thought that one pair of mice could produce 10,000 mice in eight months. Obviously, he did not believe that they all came from the same mother, and it seems very likely that he did not mean to take infant mortality into consideration. Although it is not explicit in his statement, he apparently meant to indicate that the gestation period lasts a month (this fits the facts for the house mouse, Mus musculus). Assuming this, the offspring of any month could only produce their first offspring two months later (allowing a month for puberty and a month for gestation). Working on this assumption, and letting G with subscript represent the number of the generation, we find the following: Total Producers Existing Offspring 8 6G2 1G, 14 IG1 6G3 38 24G4 1+3G2=4 80 4+3G3=7 42G5 194 7+12G4=19 114G6 434 19+21Gs=40 240Gr 1016 582G8 40+57G6=97 2318 97+120G7=217 1302Gg However, the final total of 2318 differs considerably from Leeuwenhoek's "about ten thousand," from which we must conclude that he worked with other assumptions or that he only gave an estimate. Essential for calculating the reproductive potential of a species is a knowledge of the period of time needed to reach sexual maturity and the average longevity for the species. Leeuwenhoek studied ways of ascertaining the age of plants

Month April May June July Aug. Sept. Oct. Nov.

Adults 2 8 14 38 80 194 434 1016

in 1924. Letters, HI, 73-83; quotation on p. 77. "For the greater part L.'s observations are correct. Mice on an average are twenty-one days in young; they can become pregnant again during lactation. Young mice, however, cannot propagate when they are one month old; they have to be three months old. The average throw is six, as L. says. The ratio of males and females is approximately equal in the case of a great number of litters. During the cold season their fertility decreases." Note by J. Freud, ibid. The following data comes from Philip A. Altman and Dorothy S. Dittmer, eds., Biology Data Book (Washington: Federation of American Societies for Experimental Biology, 1964) p. 57. For Mus musculus, puberty at 35 days, breeds all year, gestation 19-31 days, litter 1-12 with average of 6.

7

FRANK N. EGERTON

and animals, though he did not always relate these studies explicitly to his consideration of populations. Previously, Anrstotle had reportedreproductiverates,2' ages of sexual maturity in mammals, length of breeding season, number of years in which individuals of various species remain fertile,22and ways to determine ages of several animals. Aristotle determinedages by an approximate correlation between gestation period and longevity,23 or by characteristics of skin, teeth, and antlers; by the supposed absence of gall bladders in certain animals;

and by the size of snail shells and fish scales.24However, few other attempts to discover ways of determining age seem to have been recorded before Leeuwenhoek. In fact, exaggerated stories of the long life of some animals had accumulatedin the literature.For example, Pliny had reportedthat a stag had lived more than one hundred years,25and the story of a pike which lived 267 years had been widely accepted.28It is true that the correlation between the age of a tree and the number of rings on its stump had been drawn before Leeuwenhoek,27but he was the first to study how these rings develop.28Having realized that the seasonal rate of growth is reflected in the structure of trees, when he came to examine fish scales under his lens he similarly concluded that the lamellae of each scale represented yearly growths. He was correct in surmising that scales show a seasonal growth pattern, but he was mistaken in believing that age could be as simply determinedin fish as he had found to be the case in trees. He first studied scales from an eel, which species was an unlucky choice; eels have few scales and these are often rudimentaryand atypical of teleost scales. Nevertheless, Leeuwenhoekwas able to determine that the scales were seven years old; but unknown to him, the eel itself would have 21. Aristotle, Historia Animalium,

542b-544b.

Aristotle contrasted the

fecundity of different species of birds in Generatione Animalium, 749b_ 751j24. 22. Aristotle, Historia Animalium, 544 ff. 23. Aristotle, Historia Animalium, 578b; idem, Generatione Animalium, 777a32. 576g. Antlers-ibid., Animalium, 578g. Teeth-ibid., 24. Skin-Historia 547"10. Gall bladder607b30. Snail shell-ibid., 611". Fish scales-ibid., De Partibus Animalium, 677a31. 25. Pliny, Historia Naturalis, bk. 8, ch. 119. 26. The long credence of this story has been described by A. Hauber, "Kaiser Friedrich II. der Staufer und der langlebige Fisch," Archiv fur die und der Technik, 3 (1912), 315-329. Geschichte der Naturwissenschaften 27. George Sarton, "When Was Tree-Ring Analysis Discovered?" Isis, 45 (1954), 383-384. 28. Letter dated 12 January 1680. Phil. Trans., 13 (1683), 197-208. Letters, III, 151, 185. Letter dated 10 July 1686. Opera omnia, 1st ed., I, 87-110. Letters, VI, 143-145.

8

Leeuwenhoek as a Founder of Animal Demography been ten or eleven, since the scales do not grow before the third year.29When he extended his study to other teleosts, he assumed, incorrectly, that only one lamella was laid down per year, and he therefore greatly overestimated their age.30 This error,which was rectified in 1898,31was probablyan important influence on his conclusion that fish do not undergo senescence. He postulated that since their environment was more constant than that of terrestrial animals, they probably lived until an accident befell them, and that they probably grew as long as they lived.32This was a reasonable conclusion; the degree to which-fish age is still a moot point.83Leeuwenhoekalso used the clue from tree rings to interpret the incremental lines which he detected in the cement of an elephant's tooth.34His guess that these lines were caused by periodic deposition was correct, and for seals, the correlation between incremental lines and age has been found to be valid.35Since elephants live in areas that lack cold seasons, however, it is not as certain that the correlation would be as reliable for elephants as Leeuwenhoek thought. In connection with the subjects of longevity and rates of growth and reproductionshould be mentioned the influence of temperature, of which Leeuwenhoek showed an awareness (see below). So far, we have seen how Leeuwenhoekstudied animal populations from the three different standpoints of what we would call microbiology, reproduction, and growth rate. Even more importantwas still another approach: his studies of animal life histories, of which he was one of the first great students.36The way in which these investigations involved him in studies of population can be seen if we start with an early life history 29. Letter dated 25 July 1684. Phil. Trans., 15 (1685), 883-895. Letters, IV, 293-297 and n. 48. 30. Letters dated 12 October 1685, 2 April 1686, 22 May 1716. Letters, V, 337; VI, 35. Works, I, 65-72. Opera Omnia, 1st ed., IV, 212-218. 31. C. Hoffbauer, "Die Alterbestimmung des Karpfen an seiner Schuppe," Allgemeine Fischerei-Zeitung, 23 (1898), 341-343. 32. Letter dated 22 May 1716. Works, I, 68-70. Opera Omnia, 1st ed., IV, 212-218. 33. Alex Comfort, "The Life Span of Animals," Scientific American, 205 (Aug., 1961), 108-119; see pp. 115-116. 34. Letter dated 4 April 1687. Leeuwenhoek, Vervolg der brieven . (Leiden, 1687), pp. 1-16. Letters, VI, 191-221; see p. 195. 35. Victor B. Scheffer, "Growth Layers on the Teeth of Pinnipedia as an Indication of Age," Science, 112 (1950), 309-311. R. M. Laws, "A New Method of Age Deterination for Mammals," Nature, 169 (1952), 972-973. 36. Most of these life histories were of insects. His contributions to entomology have been discussed by Schierbeek, Leeuwenhoeh, ch. 6, and by Frederick Simon Bodenheimer, Materialen zur Geschichte der Entomologie bis Linn6 (2 vols., Berlin: Junk, 1928-29), I, 367-379; II, 363-367.

9

FRANK N. EGERTON

study in which his speculations about population are not conspicuous and then see what led him into greater involvement with population studies. One of his most thorough life histories, written August 6, 1687, was of the grain weevil Calandriagranaria L. He stated that he undertookthis study to disprovethe common assumption that Calandriaarose spontaneously.87(He was always alert for an opportunity to disprove spontaneous generation.) He received some calanders on March 13, and, as he reported, I took three distinct small glasses, and into the same I put 6. 8. or 9. Calanders,and 6. 10 or 12. Wheat-grainsof which I made sure that they were Wheat that contained no Calander; and so these had been standing for a few months upon my Study in a box....

And as the weather was cold, and these

Animals were mostly lying without any movement, I carried the small glasses in leather cases in my pocket.88 He observed them mating on March 27 and dissected some females to remove their eggs, but he failed to find any. On June 10 he found two larvae, and inside the grains he discovered others. From a different batch of calanders, he dissected two females and removed five eggs from one and two from the other.39Later he observed that one female laid four eggs in twenty-four hours.40The scarcity of their eggs seems to have led him to an important generalizationconcerning the balance of nature: After this I opened several more Calanders, and discovered therein an entire Ovary,and I saw that some of the Eggs had their full-grown size, and that the other Eggs were smaller by various degrees. From this I concluded that, whereas the Moths of the Silk-wormlay many eggs within a day or two, after which they quickly die, the Calander, on the contrary, lays only one, or a few, Eggs, each day, and that for this reason the Calander keeps alive so long, in order to be able to multiply as amply as the Silk-worm or other creatures. For, the Calanders of which I am speaking here have been alive ever since last Summer.41 This conclusion was a restatement of Sir Thomas Browne's generalization that short-lived species produce more offspring 37. Leeuwenhoek, Vervoig der brieven, pp. 73-95. Letters, VII, 3-45; see pp. 5, 7. 38. Letters, VII, 9. 39. Ibid., p. 19. 40. Ibid., p. 27. 41. Ibid., p. 19.

10

Leeuwenhoek as a Founder of Animal Demography at one time than do long-lived species.42 However, Leeuwenhoek could not have read Browne, and it was obviously an independent discovery. Leeuwenhoek went on to explain why calanders lay so few eggs: they deposit one egg, and only one, in a grain. Observing this behavior, he realized that frequent stirring of the grain would prevent calanders from depositing their eggs and thus effectively prevent their multiplication.43 This was a modest, but nevertheless noteworthy, application of a knowledge of life history to the control of a species' population. In working out the life history of the calander, Leeuwenhoek had made observations that would be important for aniimal demography. Since he had discovered the obvious theoretical and practical advantages of knowing life histories, he would subsequently repeat the procedures for other species when the opportunity arose. Consequently, he was bound to make further related contributions when the species under observation was favorable for such observations. A favorable species would be one that was 'opposite" to the calander in having a short life cycle and many offspring, because this situation was most conducive to demographic calculations. For a species producing only a few offspring, one could easily discuss its reproduction in qualitative terms, but for a species having many offspring, there would be more of a tendency to make numerical statements about its reproduction. Such a species came under Leeuwenhoek's attention during that same summer of 1687. Although Browne, Graunt, and others had calculated the probable rate of increase in human populations, Leeuwenhoek was apparently the first to publish calculations predicting the rate at which an animal species might increase during more than one generation. Since his calculations were based upon his observations on the life cycle of the fly, carried out in 1687, these observations should be examined first. Although his style was discursive, his statements were explicit. He had attentive readers who might have made their own tabular summaries of his reports, and it is convenient and appropriate to tabulate here his report on the flies: July 27 Aug.

1

received larvae in tissue; fed them meat three times pupation

42. Sir Thomas Browne (1605-82), Pseudodoxia Epidemica: or, Enquiries into Very Many Received Tenents, and Commonly Presumed Truths (London; Edward Dod, 1646; 6th ed., 1672), bk. 6, ch. 6, 1st ed., p. 297. Geoffrey Keynes has reprinted the 6th ed. with notes in The Works of Sir Thomas Browne (6 vols., London and New York, 1928-31; 2nd ed., 4 vols., Chicago, Toronto, London, 1964), 2nd ed., II, 428. 43. Letters, V1I, 31, 33.

11

FRANK N. EGERTON

14-15 ffies emerged; fed them sugar 28 dissected three flies; removed eggs Sept. 7 all had escaped except two, thought to be male and female (one lived until beginning of October,other until Oct. 16) 9

145 eggs laid

put some in glass tube and carriedin pocket; some hatched 10 remaining eggs hatched and larvae were twice size of eggs

13-14 larvae reached their full size 17-18 pupation;carriedsome pupae in pocket, left others in box in study Oct. 1-2 flies emerged from pupae carried in pocket 12 flies emerged from pupae left in box. Based upon these observationswere the following calculations: 144. flies in the first month. 72. females 144. Eggs each female 288 288 72 10368. Flies in the second month 5184. females 144. Eggs each female. 20736 20736 5184 746496. Flies in the third month.44 This was an important step for animal demography.Not because it iMdicatedthe high reproductivecapacity of insectsthat was nothing new-but because he had set the example of trying to actually calculate theoretical rates of increase for a particular kind of animal. There were, of course, some limitations to his calculations. These should be made explicit as an indication of the kinds of knowledge that was needed before improvements could be made in this type of investigation. Some of these defects Leeuwenhoekhimself took steps to improve later in his work. Other defects were eliminated by his followers, who included Reaumur and Buffon. 44. Letter dated 17 October 1687. Vervolg der brieven, pp. 115-140. Letters, VII, 81-133; quotation on p. 115.

12

Leeuwenhoek as a Founder of Animal Demography The first difficultyis that, in spite of Leeuwenhoek'sdescriptions and illustrations, the species he discussed, probablyCalliphora erythrocephala Mg., cannot be positively identified.46 Demographic traits vary from species to species, and without positive species identification, demographic observations cannot be readily substantiated.This variabilityis one reason why further quantitative precision than he gave was desirable. Even among closely related species of flies, the period of developmentis not the same,46and the genetic control of growth rate can be significantly altered by temperature. Pliny and Redi had noted the differentrates of development,47but without correlating it with reproductivepotential. The brilliant Dutch microbiologist Jan Swammerdam (1637-80),48 with whom 45. He stated that it was the largest species in Holland, ibid., p. 111, and he also provided illustrations, ibid., pl. 6, figs. 20-22, but positive identification remains problematic. J. Meltzer, who thinks it is C. erythrocephala, provides the following supporting data: "The number of eggs which the female of C. erythrocephala lays varies between -1--150 and 300. The eggs hatch within 24 hours. Larval development, at a temperature of 25CC, takes about eight days; at lower temperatures this may be considerably longer. Pupation takes place I places that are not wet, but just slightly moist; in nature, often in the soil or more at the surface of drying carrions. At 25CC the duration of the pupal stage is about two weeks. In optimal conditions, therefore, the life-cycle from egg to fly is completed in three weeks. The pre-oviposition period of C. erythrocephala lasts about five days; during the first three days of this period, only sugar containing food is taken. Thus, the imagines therefore may often be seen on flowers, especially umbellifers, from which they suck honey. Only after the third day meat becomes attractive. It is very striking that L. already observed this. From the fifth day on the first oviposition may be expected. The eggs are preferably laid in crevices or cracks; on carrions chiefly at the edges of body-openings such as the corners of eyes or mouths, or the anus." Note in Leeuwenhoek, Letters, VII, 105. 46. G. Bakri, "Ueber die Vorzugstemperatur und Vorzugsfeuchtigkeit der drei Calliphora-Arten erythrocephala Mg., vomitoria Mg. und uralenis Villen (Diptera, Calliphoridae)," Zeitschrift fuir angewandte Zoologie, 46 (1959), 495-511. Adel S. Kamal, "Comparative Study of Thirteen Species of Sarcosaprophagous Calliphoridae and Sarcophagidae (Diptera) I. Bionomics," Annals of the Entomological Society of America, 51 (1958), 261-271; see p. 265 for a table giving the hours that eleven species spent in various stages of development when raised at 80? -e- 2'F and 50% -+- 2% relative humidity. 47. Pliny, Historia Naturalis, bk. 11, ch. 43. Francesco Redi (1626-97), Esperienze intorno alla generazione degli insetti, scritte in una lettera a Carlo Dati (Firenze, 1668; 4th ed., 1688; Latin trans., Amsterdam, 1671); trans., Mab Bigelow, as Experiments on the Generation of Insects (Chicago: Open Court Publishing Co., 1909), p. 29. 48. Abraham Schierbeek, Jan Swammerdam (12 Februari 163717 Februari 1680), Zijn Leven en Zijn Werken (Lochem: De Tijdstroom, 1947). Bodenheimer, Geschichte, 1, 342-366; II, 361-362 et passim. Cole, Anatomy, pp. 270-305.

13

FRANK N. EGERTON

Leeuwenhoek commumncated,had occasionally mentioned the length of time an insect spent in certain stages of its life cycle. In particular, he called attention to the dramatic cycle of the May fly, which lives three years as a larva and only about five hours as an adult.49However,Swammerdamrealized that insect development was highly dependent upon prevailing temperatures,50 and he did not make many records of the time needed

for maturation. Both Swammerdam and Leeuwenhoek were, of course, working before thermometerswere readily available or standardized.Leeuwenhoeknoticed that flies reached maturity faster if carried in his pocket than if left in a box, but since he did not state the temperatureof the box, this remained a qualitative observation. Furthermore,he needed more data on longevity, and also information about whether individuals might reproduce more than once, im order to predict accurately the ideal rate of increase. His calculation was based upon the implicit assumption that each female reproducedonce and only once. This may be something that he never thought about; but, on the other hand, he might have felt that, since every female might not live to reproduceeven once, once was probably a reasonable average. Nor did he investigate limiting factors. It was common knowledge that insects die in cold weather. Besides this, he apparently thought that the availability of food was the main factor that controlledfly populations,for he wrote: Many persons have been inordinately amazed at the large multitude of Flies by which the People (before a prominent besieged Town) were being seriously plagued. But we can understand this, when we realize that it is impossible for the commanders to have all the People buried that are shot dead. Moreover,the number of Flies greatly increases when no care is taken that the entrails of slaughtered Beasts are not left lying in the field, but are buried every day.5' Leeuwenhoekdevelopedfurther his ideas on the factors that 49. Swammerdam, Ephemeri Vita, of, afbeeldingh van's Menschen Leven . . . (Amsterdam, 1675); trans., Edward Tyson, as Ephemeri Vita, or the Natural History and Anatomy of the Ephemeron . . . (London, 1681). It was included in Swammerdam, Biblia naturae; sive historia insectorum . . . (Leiden, 1737-38); trans., Thomas Flloyd, as The Book of Nature, or the History of Insects Reduced to Distinct Classes, Confirmed by Particular Instances, Including the Generation of the Frog, with the Life of the Author by Herman Boerhaave (London, 1758), pt. 1, 2nd Order, ch. 4. Cf. Aelian, de Animalibus, bk. 2, ch. 4. 50. Swammerdam, Book of Nature, pt. 1, 3rd Order, ch. 6, p. 173. 51. Letters, VII, 113.

14

Leeuwenhoek as a Founder of Animal Demography check insect populations as a result of studying the crane fly (Tipula). In the month of May [1693] I was shown, by a countryman, a meadow which though good land, was very thinly covered with grass, and the reason he gave for it was, that a certain species of black, thick, and short maggots devoured the roots of the grass; and he added, that the grass would not grow, until there should be some hot weather, by which these maggots (called in our language den Hemelt) would be killed. . . . And the country people say, that after a few hot days they often see these maggots lying dead in the fields.52 It is now known that his countryman had given a reliable report.53 After keeping some of these maggots, Leeuwenhoek found that they did not become adults before the end of August. They were vulnerable to drought because they spent practically the whole summer in the soil. On becoming adults, they did not lay eggs until September, he reasoned, because there were not enough rains before then to keep the eggs moist. From one female he removed 200 eggs and concluded that in three years "they would so multiply, as to devour all the roots of our grass; but by droughts in the earth, great rains and storms, and severe frosts, many of them are destroyed, and we are not infested with them equally every year.54 Insect plagues naturally brought to his mind the locust plagues of the Mediterranean lands. He had never seen locusts, but once he had had the opportunity to count the eggs in a grasshopper and found eighty. He felt certain that locusts were similar enough to grasshoppers for him to conclude that their sudden increases were due to favorable weather conditions, "which, however, I think more constant to right reason, than the notions of those who dream that Locusts came out of the clouds for the punishment of mankind, as I have often heard in conversation."55 Thus far, Leeuwenhoek had concluded that the populations of calanders and of scavenger flies were checked mainly by food supply, and that certain vegetarian insects were checked mainly by weather conditions. On July 10, 1695, he recorded his observation on aphids, and although his letter on this occa52. Letter dated 20 December 1693. Opera Omnia, 1st ed., II, 345-363. Works, II, 174. 53. The observations are in accord with those made by Frederick Simon Bodenheimer, which he summarized in A Biologist in Israel: A Book of Reminiscences (Jerusalem: Biological Studies, 1959), pp. 14-15. 54. Works, II, 178. Opera Omnia, II, 351. 55. Works, II, 179. Opera Omnia, II, 351-352.

15

FRANK N. EGERTON

sion might be most notable for reportinghis discoveryof parthenogenesis, it also contained interesting statements about rates of reproductionand checks on the population. These observations were occasioned by his concern for his gooseberry,cherry, and peach trees, the leaves of which had been curled by these insects. At first, he had thought that it was the work of ants, but closer inspection revealed the aphids. "Uponsight of these creatures, I concluded that the ants resortedto the contractedleaves for no other purpose,than to devour these Animalculeson them; and I was confirmedin my opinion by seeing several, both of the smaller and larger sort, to be almost wholly consumed; so that This discovery nothing except their skins and feet remained."5B caused him to wonder whether it was better to tolerate the ants because they ate the aphids, or to destroy them because they ate the fruit later in the summer.57Of course we now know that circumstantial evidence is not sufficient to establish that ants eat aphids: the ants may have been only collecting the aphid honeydew. Later, Leeuwenhoekbelieved that ants ate some aphids he had in his house; but again, he did not see them do it. As had become his custom, Leeuwenhoek searched for the eggs of this new species, but without success. Therefore,he dissected what he presumed were females and found within, not eggs, but mniature aphids. The first one he dissected contained four young, but later he removed up to sixty from one female.58 Since he had found that they matured in two weeks, he became alarmed at the thought of how much damage they could do to his trees. He had the urge to destroy them as quickly as possible, but, after reflecting,he realized that the effectiveness of this would be limited by the fact that new aphids could easily fly in from neighboring gardens.59 Leeuwenhoek'sconclusion about the factors limiting numbers of aphids grew out of his observations,just as had his conclusions concering the checks on flies, calanders, and crane flies. He seems not to have asked himself what limits their populations in advance of his observations.His conclusion conceming aphids was reached after having searched in January for overwintering individuals. Those he found "werenot only dead, but the hind parts of their bodies were perforatedwith a round hole, and their entrails gone, whence I gathered, that provident Nature had assigned these creaturestheir enemies, to prevent their species increasing too fast, and also for the sustenance of other 56. 57. 58. 59.

16

Works, Works, Works, Works,

II, II, II, II,

193. 197. 194, 199.

Opera Omnia, II, 488. Opera Omnia, II, 492. 199. Opera Omnia, II, 495. Opera Omnia, II, 494.

Leeuwenhoek as a Founder of Animal Demography animals."60 He was correct in noticing that aphids have many enemies, but if he had considered the question more deeply, he would undoubtedly have concluded that climate was also a limiting factor. His experience had not indicated that food might be a serious limiting factor. (Bodenheimer recently concluded that, in spite of heavy predation, climate and food are the most important limiting factors for aphids.611)The above passage from Leeuwenhoek is significant for showing that, in spite of having slight knowledge of ancient literature, he knew and accepted the ancient concept of the balance of nature. By 1696 Leeuwenhoek had more knowledge and experience of reproductive rates than anyone else had ever achieved. And he was bound to be skeptical of assertions that did not agree with his experience. He undertook a very interesting investigation to test such an assertion: "The Louse is so prolific an animal, that it is a common vulgar saying, that it will be a grandfather in the space of twenty-four hours. This I could never believe to be the fact, but rather that it would require nearly a month for the offspring of a Louse to be capable of producing young of its kind."62To find out, he put two lice into his stocking, which he wore without removing for six days. At the end of that time, one louse had laid fifty eggs, and its body, which he then dissected, contained at least fifty more. The other louse had disappeared, after having laid about forty eggs. He persisted in the experiment a while longer, but then terminated it for esthetic reasons: Having worn the stocking ten days longer, I found in it at least twenty-five lice of three different sizes, some of which I judged were two days old, others a day old, and the rest newly come out of the egg, besides others ready to come forth, as I found upon opening some of the remaining eggs. But I was so disgusted at the sight of so many lice, that I threw the stocking containing them into the street."63 There were two conclusions that Leeuwenhoek drew from this experiment. One concerned the rate of propagation. He did not give his calculations, but he presumed the following: that two females could lay 200 eggs 12 days after mating, which would hatch into 100 males and 100 females in 6 days; and that this generation would be able to breed 16 days later. At this rate, he 60. Works, II, 195. Opera Omnia, II, 490. 61. Frederick Simon Bodenheimer, Animal Ecology To-day (The Hague: W. Junk, 1958), pp. 83-86. 62. Letter dated 20 February 1696. Works, II, 167. Opera Omnia, II ("Continuatio"), 56-82. 63. Works, II, 168.

17

FRANK N. EGERTON

concluded, there would be 10,000 offspring from the two females within eight weeks. He based this calculation on the premise that within eight weeks the second generation of offspring would have hatched (if he had included the first generation, the total would have been 10,200). He then asked the significant question, 'who can tell, whether in the heat of summer these creatures may not breed in half the time I have mentioned?"64 The second conclusion was on the biologicrole of lice and was derived from this calculation of reproductiverate. He realized that lice increased rapidly enough to become very numerous on the body of anyone who did not change his clothes or wash very often, and he suspectedthat such a person "mayin a few months (if I may use the expression) be devoured by these vermin."65 This was an ambiguous statement, and it is not certain that he meant the person might die; but at least he had begun to suspect that a large number of parasites could be dangerous, and that it is possible for parasites to multiply rapidly on a host. Apparently, he also felt that the population increase of an external parasite would depend, at least as a parasite on human beings, upon the grooming habits of the host. The way in which Leeuwenhoek'sstudies on population grew out of his investigations of microorganismshas been described above. In the last letter that he wrote about protozoans,on November5, 1716, he once again marveledat their rate of increase. Judging from his description, the species he observed was a phytoflagellate,most likely Polytoma.66Careful observation revealed to him that the time between generationswas very brief, and that the species he observedhad a peculiar form of reproduction: . . . in the end I found out that these animalcules lived for no longer than 30 or 36 hours, and that they then fixed themselves upon the glass, and stopped there without moving: while soon after, their body burst asunder, and lay divided into eight portions: and these were actually young animalcules, for in five or six seconds some of them swam off.67 In this case he did give calculations for rate of increase. There were the novel features of there being no sexual reproduction 64. Works, II, 169. 65. Works, II, 168. 66. Dobell, Leeuwenhoek, p. 298. 67. Opera Omnia, IV, 279-286; see p. 284; trans., Dobell, Leeuwenhoek, pp. 297-298. I wish to thank Russell and Russell, Inc. for permission to quote this and the following passage.

18

Leeuwenhoek as a Founder of Animal Demography and the disappearance of the parent in the production of the offspring: [At zero hour, In 36 hours, or 1 /2days, in 3 days, in 41/2 days, in 6 days, in 71/2 days, in 9 days,

1 8 8 64 8 512 8 4096 8 32768 8 262144

individual] animalcules animalcules

animalcules.68

Also important for developing an understanding of population dynamics were Leeuwenhoek's investigations of the feeding habits of fish. Some fishermen with whom he discussed the matter believed that herring did not eat. He opened the intestines of several herring in 1695 and concluded that they eat the plankton which he had previously studied under the microscope: . . .we must conclude, that there are more animalcules or minute fishes in the sea, than has ever yet been thought of: and hence we are not to wonder that Herrings are sometimes caught in one, and sometimes in another part of the sea; sometimes in the shallows, and sometimes in the deep water, according to the places where the small fishes on which they feed do from time to time resort; so that, here I think, the words of scripture may be applied "wheresoever the Carcase is, there will the Eagles be gathered together."89 About twenty years afterward he again investigated the food of fish, this time in order to explain the sudden increase of haddock and cod: In the months of April and May 1716, there were brought to our town of Delft, from the sea coasts at Schevling,70 Cat68. Opera Omnia, IV, 285. Dobell, Leeuwenhoek, p. 298. 69. Letter dated 28 December 1695. Works, II, 11. Opera Omnia, II, 45. Leeuwenhoek described oceanic plankton in a letter to the Royal Society dated 9 October 1686, which was abbreviated in Phil. Trans., 12 (1677), 821-831. It is fully translated by Dobell, Leeuwenhoek, pp. 112-166; see p. 129. Letters, I, 87-89. The biblical passage quoted by Leeuwenhoek here and in the following quotation is Matt. 24:28. well-known sea-side resort near The 70. "Now called Scheveningen-the Hague . . ." Dobell, Leeuwenhoek, p. 129.

19

FRANK N. EGERTON

wick, and Terheid, a great quantity of the fish called haddocks, which, though very fresh and good, were sold at a low price. The glut of this fish was so great, that though in general they are caught with hooks, they were on this occasion taken in nets. Seeing this, I considered, that there must be some particular reason, why these fish should at that time resort to our coasts in such multitudes, and I was afterwardsconfirmedin that opinion, for in a month or two afterwards, not one of these fish was to be taken: and the reason which I assigned to myself for the abundance I have mentioned was, that at that time, there was a greater quantity of food for them on the coast than usual, whereby they had been tempted thither. In order to investigate this matter, I opened the stomachs of many haddocks, and found them to be filled with a certain small species of shrimps, called by our fishermen meutjens, which are taken among the common shrimps, and are used for food by people living along the shore. About a fortnight afterwards, on examining the stomachs of the haddocks, I found some of them quite empty, and others not more than half filled with the before mentioned small fish; and so much was the glut then diminished, that few or no haddocks were taken. Upon enquiring the reason of this diminution from a fisherman, he answered only, that every sort of fish had its season, though I should rather have said, in the words of scripture, "that where the food is, there will the eagles be gathered together." At the time there was this glut of haddocks, there was a great quantity of cod fish caught on our coast, the reason of which I took to be, that these cod flocked to our shores in pursuit of the haddocks which are their food. About the beginning of October, in the same year, there were taken on our coasts, great quantities of the common shrimps, and these in better condition than they are generally found in the summer time. Hence I concluded, that the haddocks would again resort to our coasts, and that the shrimps, to avoid them, would crowd in greaterquantities to the shores and shallows. In the month of Novemberin the same year, there was another great draught of haddocks on our coast, whereupon I went to the fish marketto examine the intestines, when newly taken out of the fish: I found most of the stomachs to be empty of food, but some remains thereof in the intestines; and at the same time, great plenty of cod fish were caught, I

20

Leeuwenhoek as a Founder of Animal Demography judged that the haddocks, avoiding the pursuit of the cod, and these pursuing the haddocks, was the reason, that both were in such abundance.71 In these two quotations, taken from letters written two decades apart, Leeuwenhoek explained fluctuations in fish populations as the result of irregular migrations. He gave no reasons for his rejecting the fisherman's belief that these fluctuations were seasonal. Apparently it did not occur to Leeuwenhoek to apply to fish his conclusions on the rapid reproductive potentials of insects. Leeuwenhoek's observations on fish included early descriptions of food chains. Some food chains, such as fish to duck to hawk, might have been known earlier,72 but no conclusions of ecological significance had been derived from this knowledge. Leeuwenhoek did not dwell upon the subject more than is indicated in the above quotations, but his investigations of the food of fish and his observations on plankton provided an early clue for understanding the economy of oceanic communities. SUMMARY AND CONCLUSIONS Leeuwenhoek's observations relating to animal population, though scattered through many letters written during a period of over forty years, when seen in toto, were important contributions to the subject now known as animal demography. He maintained enough contact with other scientists to have received encouragement and some helpful suggestions, but the language barrier and the novelty of doing microscopic work forced him to be resourceful, inventive, and original. His multifarious investigations impinged upon population biology before he discovered a direct interest in it. He devised methods for estimating numbers of animalcules, and then he went on to estimate the 71. Letter dated 10 September 1717. Works, I, 283-285. Opera Omnia, IV, 396-399. 72. Several food chains had been described by Abu 'Uthman 'Amr ibn Bahr al-Jahiz in his Book of Animals, bk. 6, ch. 133. The passage has been translated from Arabic into Spanish by Miguel Asin Padacios, "El 'Libro de los Animales' de Jahiz," Isis, 14 (1930), 20-54; see pp. 38-39. Padacios also compiled a list of al-Jahiz's references to generation, pp. 2828, and to conflicts between species, p. 32. The passage on food chains has been translated from Spanish into English by Conway Zirkle in "Natural Selection before the 'Origin of Species,'" Proceedings of the American Philosophical Society, 84 (1941), 71-124; see p. 85. On alJahiz, see: George Sarton, Introduction to the History of Science (vol. 1, Baltimore: Williams and Wilkins, 1927), p. 597. L. Kopf, "The 'Book of Animals' (Kitab al-hayawan) of Al-Jahiz (ca. 767-868)," Actes 7' Congr0Ts International d'Histoire des Sciences (Paris, 1953), pp. 395401.

21

FRAN}K N. EGERTON

population of the world. His interest in reproduction was an important avenue by which he approached the subject of reproductive capacity. Other important approaches were his studies of growth, longevity, and life histories. He discovered relationships between aspects of the life history, longevity, and reproductive capacity of several species of insects, notably calanders, scavenger flies, crane flies, aphids, and lice. An important feature of these investigations were the arithmetical calculations which he made of reproductive potentials. In spite of several limitations, these calculations were an important innovation to the study of animal population. In his later years, his investigations came more and more within the sphere of ecology. He made the first significant observations on food chains. It is especially interesting that fish were the subject of these observations, because it was not until the latter half of the nineteenth century that scientists realized that fish ultimately depend upon phytoplankton. These accomplishments did not pass unnoticed. Although Leeuwenhoek never synthesized his scattered observations concernig population, his originality and perception were appreciated by outstanding biologists of the eighteenth century. The important discussions of population biology by Reaumur, Buffon, and Bonnet all derived inspiration and assistance from the writings of Leeuwenhoek.73 This ingenious Fellow of the Royal Society, "by detecting through diligent application and scrutiny the mysteries of Nature and the secrets of natural philosophy,"74 became one of the founders of animal demography. This paper is based upon part of a dissertation submitted to the University of Wisconsin in partial fulfillment of the requirements for the Ph.D. degree. It was written under the direction of Professor Robert Clinton Stauffer, to whom I wish to express my appreciation for his assistance. Undocumented statements in this paper concerning the history of demography are based upon this dissertation. 73. The influence of Leeuwenhoek upon these men has been shown in my dissertation, ch. 3. 74. From Leeuwenhoek's epitaph, trans., Dobell, Leeuwenhoek, p. 100.

22

The Founding of Population Genetics: Contributionsof the ChetverikovSchool 1924-1934 MARK B. ADAMS Department of the History of Science, Harvard University, Cambridge, Massachusetts

Of the eighteen founders of the synthetic theory of evolution listed by G. G. Simpson in his book, The Meaning of Evolution,' four are Russian in origin and training: S. S. Chetverikov, N. V. Timofeev-Resovsky, N. P. Dubinin, and Th. Dobzhansky. All are significant primarily for the same type of studies: analyses of the genetic variability of wild populations, and the development of sophisticated notions of the role of the genetic and environmental backgrounds in determining the expression and fitness of genes. Furthermore, Dobzhansky's work comes later than that of the first three, and he himself was among the first to credit their work with originating many of the concepts and experimental approaches which he has applied so fruitfully. Thus it is especially lamentable that an informed student of evolution today in the West, though he would probably be familiar with the work of the Western founders of the synthetic theory, might well have only a vague notion of the contributions of Chetverikov, Timofeev-Resovsky, and Dubinin. This lack is perhaps natural enough: it reflects the scant treatment given them in the biological literature, which generally has only brief mentions of their early works. And this lack in turn is to be largely explained by the unavailability of many important articles published in the 1920's and 1930's in Russian, and often unavailable in most Western libraries in any language. In order to put this study in the proper perspective, it is perhaps advisable to delineate what will not be discussed. First, the Russian School made important contributions to genetics which, however significant, do not bear directly on population genetics. Hence I will not discuss the contemporaneous work on position 1. George Gaylord Simpson, The Meaning of Evolution (New Haven: Yale University Press, 1949), p. 278. In later editions, Simpson adds I. I. Schmalhausen.

23

MARK B. ADAMS

effect (by Dobzhansky, I. B. Panshin, and others), on chemical mutagenesis (I. A. Rapoport and colleagues), or on the substructure of the gene, termed "step-allelomorphism" by Russian workers (I. I. Agol, A. S. Serebrovsky, and Dubinin.) Second, I wish to concentrate only on work completed and published before 1935, since our concern is with the founding of population genetics and not its later development. Finally, I wish to restrict myself to studies of the genus Drosophila. This is a natural enough restriction, since Drosophila was by far the best understood genus genetically, thanks to the Morgan School, whose work together with that of the Russian School has made Drosophila the mainstay of most experimental population genetics since then. I wish to focus on the work of Sergei S. Chetverikov and of the students who worked with him in the decade after the Bolshevik Revolution. Their scientific contributions are threefold. First, the experimental work under Chetverikov's direction by Timofeev-Resovsky on a naturally occurring Drosophila population led to the development of clear ideas conceming the influence of genetic and environmental backgrounds on the fitnesses and effects of genes. Second, it was Chetverikov's 1926 theoretical paper, "On Certain Features of the Evolutionary Process from the Viewpoint of Modem Genetics,"2 which initially bridged the gap between Mendelism and Darwinism, or, to be more precise, between the genetics of the Morgan School; biometrics and mathematical studies as developed by Karl Pearson, G. H. Hardy, H. T. J. Norton, and others; and studies of natural variation from natural history. Finally, in order to test experimentally certain theoretical conclusions, Chetverikov and his students undertook the first genetic analysis of free-living Drosophila species and founded experimental population genetics. This led almost immediately to a series of studies by Dubinin, of which the first is of special interest. Accordingly, we will consider in order; the formation of the Russian School, its scientific contributions, and its historical significance. The impact of the Morgan School on Russian genetics was 2. Sergei S. Chetverikov, "O nekotorykh momentakh evoliutsionnogo protsessa s tochki zreniia sovremennoi genetiki," Zhur. Eksper. Biologii, 2 (1926), pp. 3-54. The Russian original is reprinted in Biulleten' Moskovskogo Obshchestva Ispytatelei Prirody, Otdel Biologii, LXX (1965), 4:33-74. For an English translation, see that done by Malina Barker, edited by I. M. Lerner, which appeared under the title in the text, Proc. Amer. Phil. Soc., 105 (1961), pp. 167-95. In general, quotations in the text are taken from the Lerner translation.

24

The Founding of Population Genetics post-revolutionary, and this impact was heightened by efforts of the new Soviet regime to stimulate the development of genetics. L. C. Dunn relates the following episode which illustrates the light in which the development of biology was regarded by the new Soviet regime: "Koltsov . . . walked with Lenin in the 1920 Leningrad famine. Lenin said, 'The famine to prevent is the next one, and the time to begin is nowI' "3 As a result of this conversation, emergency funds were partly spent to build the Institute of Applied Botany, and biological work received priority support. The presence of a promising number of experienced biologists, coupled with government interest in the development of biology for practical reasons, no doubt contributed to the rapid development of the three major genetics schools which arose in Russia in the early twenties.4 One group developed around I. A. Philipchenko in Leningrad; also in Leningrad was a second group, headed by N. I. Vavilov, who had moved from Saratov to establish a department of applied botany and plant breeding that later developed into the USSR Institute of Plant Breeding. WVhile Leningrad had been developing as a center for research in plant genetics, Moscow was developing as a center for animal genetics, due largely to the efforts of N. K. Koltsov, S. S. Chetverikov, and A. S. Serebrovsky. Sergei S. Chetverikov (1880-1959) was a butterfly taxonomist by training, but his concem with entomology and evolutionary problems was complemented by an interest in genetics and biometrics.5 By the time he had graduated from Moscow University in 1906 he had already published an article6 in which he called attention to the evolutionary significance of what he termed "population waves": periodic and radical decimation of insect populations which in his view allowed the role of natural 3. L. C. Dunn, "Science in the USSR: Soviet Biology," Science, 99 (1944), pp. 65-67. 4. These three schools are discussed briefly by Theodosius Dobzhansky, 'The Crisis in Soviet Biology," Continuity and Change in Russian and Soviet Thought, E. J. Simmons, ed. (Cambridge: Harvard University Press, 1955) and also by Sos I. Alikhanian, "Soviet Genetics," Soviet Life, January 1966. 5. For material on Chetverikov's life and work, see Sergei S. Chetverikov, "Autobiographical Note," written in 1956, Nova Acta Leopoldina, N. S., 143 (1960), pp. 308-310. Some additional information is also available in I. M. Lerner's introduction to the Malina Barker translation of Chetverikov's "On Certain Features . . . ," Proc. Amer. Phil. Soc., 105 (1961), pp. 167-69; also B. L. Astaurov, "Two Landmarks in the Development of Ispytatelei Genetical Concepts," Biulleten' Moskovshogo Obshchestva Prirody, 70 (1965), pp. 25-32. 6. Chetverikov, "Volny zhizni" (Waves of Life), Dnevnik Zootd., Moscow Society of Naturalists, 3.

25

MARK B. ADAMS

selection to be periodically "swamped"by chance phenomena. It was thus one of the earliest papers to call attention to what Sewall Wright would twenty-five years later term "genetic drift,"

and according to Chetverikov'sown evaluation, the paper "produced a sensation in Russian readership circles."7

In the decade preceding the Revolution, Chetverikovtaught entomologyat the MoscowUniversity for Women and published papers on entomology.After the OctoberRevolution,the University for Women was merged with Moscow University; Chetverikov remained on the faculty where he taught entomology and "theoreticalsystematics."By 1924 he had developedtwo entirely new courses in biometry and genetics which he taught until 1929. Aleksandr Serebrovskyhad established a department of genetics at Moscow University, and it was to Professors Chetverikov and Serebrovskythat H. J. Muller in August 1922 brought laboratory Drosophila melanogaster strains from the United

States.8Theodosius Dobzhansky,then an instructor at the University of Kiev and subsequently of Leningrad, borrowedfrom these strains, and their introduction was a major stimulus to laboratorywork on Drosophilain Russia. Investigationshad begun as early as 1920 on free-living Drosophila from suburban Moscow, but the Muller strains were the first available with a known genetic history.9 According to the testimony of N. K. Koltsov, Muller's impact was also a personal one, in that he 7. Chetverikov, "Autobiographical Note." Translated by the author from German. Chetverikov borrowed certain features of the theory proposed by Rev. John T. Gulick which suggested that non-adaptive evolution could occur as a result of inbreeding of a few isolated individuals. This notion Chetverikov applied to the case of radically fluctuating insect population sizes. The author is presently engaged in a study of the intellectual currents of thought which led to the almost simultaneous exposition of a theory of "genetic drift" by Sewall Wright, and of a strikingly similar theory of "genetico-automatical processes," by D. D. Romashov and N. P. in 1931, and apparently independently. Dubinin-both 8. The significance of Muller's 1922 visit is repeatedly emphasized in Russian genetics literature of the period. For example, Th. Dobzhansky, 34 "Kleinere Mitteilungen," Z. Induktive Abstammungs-Vererbungslehre, (1924), p. 245, refers to a culture brought by Muller in August, 1922. (Dobzhansky's fuller communication [43, 1927, p. 330] mistakenly gives the date as August 1923 due to a misprint.) See also N. K. Koltsov, "On the Work of the Institute of Experimental Biology in Moscow," Uspekhi Eksperimental'noi Biologii, 8 (1929), p. 23; and A. S. Serebrovsky and V. V. Sakharov, "New Mutations in Drosophila melanogaster," Zhur. Ekaper. Biologii, 1 (1925). All these sources may be referred to for brief descriptions of Muller's visit. 9. Koltsov, "On the Work . . . " p. 23.

26

The Founding of Population Genetics "infected" young Russian workers with a sense of "enthusiasm for the study of Drosophila genetics."'0 Koltsov, who earlier had operated an experimental station in animal genetics near Moscow, had been chosen to direct the recently established Institute of Experimental Biology which had been established in 1916 and reorganized after the Revolution. In 1922 Koltsov entrusted to Chetverikov the organization and direction of the genetics section of the Institute, a post which he held until 1929 when, according to B. L. Astaurov, one of his students, he was "forced to break off his work on Drosophila population genetics";"' he left Moscow, for reasons which remain obscure.'2 He never retumed to his earlier Drosophila studies. For the next three years he worked as a zoo consultant in Sverdlovsk, and from 1932 to 1935 he taught mathematics at a tekhnikum in Vladimir, just east of Moscow. In 1935 he went to Gorkii University to teach genetics, and he soon became head of the biology faculty. He worked there until 1948 and lived in the city of Gorkii until his death on July 2, 1959.13 The period of Chetverikov's tenure at Moscow University (1919-1929), and especially at the Institute of Experimental Biology (1924-1929), was the formative period of the Russian School of population genetics. According to his own recollection, Chetverikov "collected a narrow circle of students and co-workers" about him, and over a number of years gave a seminar in "the relationships between evolutionary theory and the newest results in genetics."'4 This group included a num10. Ibid. (Translated by the author from Russian.) 11. B. L. Astaurov, "Two Landmarks ...," p. 27. (Translated by the author from Russian.) 12. The only published suggestion as to the reason for Chetverikov's departure comes from Th. Dobzhansky: "In 1929 [Chetverikov] was banished from Moscow, as were some of his collaborators. In their enthusiasm they forgot caution. They organized a closed genetics and evolution discussion group, the acceptance into which of new members was by unanimous secret ballot of the old members. This was too much for Stalin's secret police." Whether this was the sole reason, who was responsible for the banishment, how it was engineered, and how and by whom resisted (if at all): these and other matters remain unclear. Dobzhansky's article, "Sergei Sergeevich Tshetverikov: 1880-1959," (Genetics, 55 [1967], pp. 1-3) from which the above quotation is taken, contains useful biographical information on Chetverikov not previously published in English. 13. Chetverikov writes in his autobiographical note: "In 1948 I resigned from all positions." There can be little doubt that the official victory of Lysenko over his geneticist rivals in that year was the major cause of Chetverikov's resignation; health became a contributing factor, since in the following year Chetverikov had a series of heart attacks and became blind. 14. Chetverikov, "Autobiographical note."

27

MARK B. ADAMS

ber who were later to become prominent in world science, among them N. V. Timofeev-Resovsky,N. P. Dubinin, B. L. Astaurov, and D. D. Romashov, all of whom Chetverikovhad initiated into research. Dobzhanskyhas acknowledgedhis debt to the work of members of this group-indeed his earliest work on Drosophila (1923-1927) was done on flies obtained from the Chetverikov Laboratory.Hence, in terms of training and intellectual influence, we are justified in speaking of Dobzhansky as an offshoot of the Russian School, though he left for America in 1927; likewise, Timofeev-Resovsky, who studied with Chetverikov for several years, is clearly part of the Russian School, even though after 1925 he did most of his work in Germany, based at Buch, just north of Berlin. In the twenties Chetverikov'sgroup developed and clarified a number of concepts which were to lead to important work in later decades, and initiated the wide-scale genetic analyses of natural populations of Drosophila on which much modem population genetics is based. It is to one of the most important of these concepts-his idea of the "genotypic milieu"-that we shall now turn. As Chetverikovreadily admitted,15his school did not originate the notions of pleiotropic and epistatic gene action. It was William Bateson who first demonstrated the role of gene interaction in producing a phenotypic character in 1907.16 At roughly the same time, the studies of Nilsson-Ehle on the genetics of cereals were showing that many cases of continuous variation could be explained if it was assumed that certain major genes were interacting with other genes so as to increase, decrease, or alter their effects.17 Thus, NilssonEhle wrote that an inherited difference between individuals or strains may be due to "the joint actions of many genes, 15. Chetverikov, "On Certain features . .. ," (Lerner trans., p. 189).

16. William Bateson and R. C. Punnett, "Experimental Studies in the Physiology of Heredity" (1905-1909), in J. A. Peters, Classic Papers in Genetics (Prentice-Hall, 1959). One of these investigations concerned the genetic basis of the shape of poultry combs. Bateson showed that when a gene 'R' (which by itself produced a comb shape termed "rose") was present with another gene 'P' (which yields by itself a shape termed "pea,") the resulting combination 'RP' produced a comb of an entirely different shape, which he called "walnut." The two genes had thus interacted to produce a phenotypic character, the new comb shape. 17. H. Nilsson-Ehle, "Kreuzungsuntersuchungenan Hafer und Weizen," Lund# Univ. Aarsh. N.E. Afd. 2, 5, 2: 122. Cited in Th. Dobzhansky, Genetics and the Origin of the Species, 3rd ed., rev. (New York: Columbia University Press, 1951), p. 71.

28

The Founding of Population Genetics each having a small effect in relation to the total nonheritable fluctuation of the character in question."18 Effects that would now be termed pleiotropic and epistatic were also discovered in Drosophila. In work reported 19121914, J. S. Dexter did experiments on Drosophila involving "beaded" wing, a highly variable character which is often nearly normal in appearance. Dexter( showed that "the degree of abnormality and the proportion of abnormal offspring are both capable of being altered, within limits, by selection or by crossing to a normal stock."19After 1914, Morgan, Muller, Altenburg, and Dexter showed that many modifier genes existed in Drosophila, and that they were inherited in Mendelian fashion. To Chetverikov, however, belongs the credit for clarifying the importance of gene interaction for evolution. In his lengthy theoretical article (1926) which will be discussed later, his treatment of the evolutionary importance of gene interactions and of the genetic background, which he terms the "genotypic milieu," stands out for its clarity and insight. Chetverikov develops the earlier notion of pleiotropy, which was previously applied to one gene affecting a limited number of characters, into a more generalized concept of the "genotypic milieu": Each gene does not act isolately from the whole genotype, is not independent of it, but acts, manifests itself, within it, in relation to it. The very same gene will manifest itself differently, depending on the complex of the other genes in which it finds itself. For it, this complex, this genotype, will be the genotypic milieu, within the surroundings of which it will be externally manifested. And as phenotypically every character depends for its expression on the surrounding external environment, and is the reaction of the organism to the given external influences, so genotypically each character depends for its expression on the structure of the whole genotype, and is a reaction to definite internal influences.20 From here he moves to a discussion of the evolutionary implications of the "genotypic milieu." True, Chetverikov agrees, selection cannot alter the gene itself-a point made 18. Ibid. 19. T. H. Morgan, A. H. Sturtevant, H. J. Muller, and C. B. Bridges, The Mechanism of Mendelian Heredity (New York: Henry Holt, 1915), p. 195. 20. Chetverikov, "On Certain Features..." (Lerner trans., p. 190; Russian reprint, p. 66).

29

MARK B. ADAM S

by Morgan repeatedly-but it can and will alter the expression of the gene in subtle ways and hence is a 'creative process" in evolution. Any newly arising mutation may appear in connection with the selected feature either as an "enhancer" or a "weakener." In the case of an "enhancer," selection will pick it up and spread this gene in subsequent generations through the whole population, enhancing the selected trait. In this way selection does not cease with the passage of the selected character into the homozygous condition, but is extended further for an indefinitely long time, acting on the whole genotype. Exactly this process occurs also in nature under the influence of natural selection. It no longer merely selects a given mutation, nor only selects genes favored by it; its influence extends a great deal further over the total complex of genes, over the whole "genotypic milieu," on the background of which a given gene will manifest itself in various ways. In selecting one trait, one gene, selection indirectly also selects a definite genotype milieu, a genotype most favorable for the manifestation of the given character. By removing thus unfavorable combinations of genes, selection aids the realization of a more advantageous genotypic milieu. Selection results in the enhancement of the trait, and in this sense it actively participates in the evolutionary process.21 Hence Chetverikov put forth the first clear statement of the Importance of the "genotypic milieu." Its experimental demonstration and further clarification, however, was the work of N. V. Timofeev-Resovsky. It was he who, in the words of Fothergill,22 "stabilized" the concept of the interaction of genetic factors in a series of papers published 1925-1934, reporting work begun under Chetverikov's direction, 19231925. In the first of these papers, "On the Phenotypic Expression of the Genotype,"23 Timofeev-Resovsky used stocks of a 21. Ibid. 22. P. G. Fothergill, Historical Aspects of Organic Evolution (London, 1952), p. 237. 23. N. V. Timofeev-Resovsky, "O fenotipicheskom proiavlenii genotipa: 1. Genovariatsiia radius incompletus u Drosophila funebris," Zhur. Eksper. Biologii, 1 (1925), pp. 93-142. An English article covering much of the same material appeared under the title, "Studies of the Phenotypic Manifestation of Hereditary Factors: I. On the Phenotypic Manifestation of

30

The Founding of Population Genetics mutant in Drosophila funebris called "radius incompletus" (ri) in order to demonstrate that the phenotypic expression of ri varies according to the genetic environment in which it occurred. This work led him to distinguish three phenomena in the phenotypic manifestation of the gene which were shown to vary independently: In the intensity of the gene manifestation, the frequency of appearance, or penetrance, must be distinguished from the degree of expression of the character, or expressivity; the third phenomenon is specificity, or localization, extent, array of variants, and morphophysiological nature of the character.24 The ri character, however, proved unsuitable for the analysis of the third phenomenon, specificity, and hence work was done on another recessive autosomal gene of Drosophila funebris whose expression depends on the presence of ri: this mutation is called vti (venae transversae incompletae) and breaks or abolishes the crossveins of the wings. Although this work was briefly reported earlier and was "essentially completed by 1928,"25 it was most completely described in an article published in 1934-5, "On the Influence of the Genotypic Milieu and of the Environment on the Expression of the Genotype."26

Timofeev-Resovsky employed the following strategy: in order to evaluate the effect of the genotypic milieu on the expression of the trait vti, he created a series of uniform but different genotypic backgrounds into which he introduced vti and ri in the homozygous condition; whereupon he tested the penetrance, expressivity, and specificity of the vti trait. To get the most diverse array of genetic backgrounds possible, he crossed flies homozygous for the vti and ri traits with various laboratory cultures and with wild flies from geographically diverse populations (from Moscow, Leningrad, Kiev, Central Russia, Saratov, the Genovariation Radius incompletus in Drosophila funebris," Genetics, 12 (1927), pp. 128-165. 24. N. W. Timofeef-Ressovsky, "Uber den Einfluss des genotypischen Milieus und Aussenbedingungen auf die Realisation des Genotyps," Nachr. (Biologie) Ges. Wiss. Goettingen. Math.-Physik. Ki. N.F. Fachgruppe IV vol. I (1934-5). Dr. Roger Milkman kindly made available to me his unpublished translation of this article into English, under the title "On the influence of the genetic background and of the environment on the expression of the genotype: the mutation vti (venae transversae incompletae) in Drosophila funebris." Quotations used in the text are taken from his translation. 25. Ibid. (Milkman trans., p. 1). 26. Ibid.

31

MARK B. ADAMS

the Crimea, the Caucasus, and so on). Homozygousvti ri flies which appearedin the F2'swere inbred and selected for various expressions of the trait. These populations were inbred for 25-35 generations (until selection had no further effect on the expression of the trait), resulting in populations essentially homozygous for vti modifiers. The penetrance was then measured as the percentage of individuals from such lines showing the trait; expressivity was measured as the percentage of offspring exhibiting the trait which totally lacked the posterior crossvein; and specificity was tabulated using a simple classification system based on the amount and location of crossvein deletion. When data on penetrance in the thirty cultures were collected, a variation in penetrance was found ranging between 41% and 100%. Since these data were gathered simultaneously and under identical environmental conditions, and since all cultures are homozygous for vti ri, these differences are inherited, and are caused by the different array of modifying genes present in each culture. Expressivity also varied, ranging from 12% to 100%. In general, high penetrance was accompanied by high expressivity; but when we consider only those cultures with 100% penetrance, expressivity ranged between 29.3% and 100%, and hence expressivity was shown to be in large part independent of penetrance. The cultures also varied in the fields of influence or specificity, and this variation failed to correlate with either penetrance or expressivity. When Timofeev-Resovskywent on to test how environmental factors can influence penetrance, expressivity, and specificity, he found that while changes in food and humidity did not noticeably affect the vti phenotype, temperature affected penetrance and expressivity at two key points in development: the first larval stage and the pupal stage. The specificity, however, was not affected by temperature, but only by the "genotypicmilieu." The influence of Timofeev-Resovsky'sconceptual innovation did not await the 1934 publication of his most complete treatment of the subject. Rather the impact of his 1925 article was immediate among Russian workers: as early as 1926, Russian work on Drosophila mutants began distinguishing between penetrance, expressivity, and specificity.27Later in 1925, Timo27. E.g., E. I. Balkashina, '"Vlianie genotipa na mnozhestvennoe vyrazhenie genovariatsii Alae curvatae u Drosophila funebris Meig.," (The influence of the genotype on the multiple expression of the genovariation [mutation] Alae curvatae in Drosophila funebris Meig.) Zhur. Eksper. Biologii, 2 (1926), no. 2-3.

32

The Founding of Population Genetics feev-Resovsky left Moscow and moved to Buch, just north of Berlin, where he continued his work, keeping in close contact with his Russian colleagues. Within a year of Timofeev-Resovsky's departure for Germany, his teacher, Chetverikov, had incorporated his work on the genetic background into a general statement of the evolutionary process which is considered by Th. Dobzhansky28 to be the first that put to rest Jenkin's objections to the theory of evolution by natural selection, and the first of the founding papers of population genetics, preceding those of Wright, Fisher, and Haldane. Chetverikov's reasoning in this paper led to experimental work which has been justdfiably termed "trail-blazing" by I. M. Lerner,29 and hence it will be worth our while to explore this reasonig. The purpose of Chetverikov's major theoretical work, "On Certain Features of the Evolutionary Process from the Viewpoint of Modem Genetics,"30is clearly formulated at the outset: Genetics is in similar contradiction with conventional views of general evolutionary concepts and in this, undoubtedly, lies the reason that Mendelism was greeted with such hostility by many outstanding evolutionists, both here and abroad. The present article sets itself the goal of clarifying certain aspects of evolution in the light of current genetic concepts.3' Chetverikov begins his discussion by treating the "origin of mutations in nature." He argues that the process of mutation observed in the laboratory is also going on under natural conditions, but that the occurrence of such mutations is not evident, primarily because recessive mutants would arise in the heterozygous condition and would "remain hidden from the eye."32 28. Th. Dobzhansky, Mankind Evolving: The Evolution of the Human Species (New Haven and London; Yale University Press, 1962), p. 136. 29. I. M. Lerner's introduction to Chetverikov, "On Certain Features.. . (Lerner trans.). 30. A number of interesting aspects of Chetverikov's paper will not be discussed here, e.g. his use of a reproductive isolation criterion in his definition of the species; his use of calculation and genetic notions in his modified restatement of the theory of speciation by isolation; and a more detailed discussion of his debt to biometrics, genetics, and natural history. We will rather be concerned with those theoretical arguments which lead him to predict a condition of natural populations, which subsequently led to experimental confirmation. (Lerner trans., p. 169; Rus31. Chetverikov, "On Certain Features..." sian reprint, 1965, p. 34). 32. Ibid. (Lerner trans., pp. 170-174; Russian reprint, pp. 3542). In his

33

MARK B. ADAMS

What happens to newly arisen natural mutations? Chetverikov draws on the work of Hardy and of Pearson to show that a "free-crossing" (svobodno skreshchivanyi) or panmictic (randomly mating) population, in the absence of selection, would maintain all genes, including the new mutants, at a constant frequency. Given frequent mutations, then, each would be kept and spread, which leads Chetverikov to conclude that a species, like a sponge, soaks up heterozygous mutations while remaining from first to last externally (phenotypically) homogeneous.83 What role does selection play? Chetverikov cites a table prepared by the English mathematician H. T. J. Norton showing how many generations are required for selection intensities of various magnitudes to alter the relative frequencies of alleles. He observes that the process of the transformation of the species, that is, of the complete replacement of a forner, unadapted form by the more adapted one, always proceeds, practically speaking, to an end.34 But from Norton's table he also concludes that selection, as well as repeated mutation, causes the build-up of hidden recessive mutants in the population, since harmful recessives are selected again more slowly than harmful dominants, which are quickly eliminated. Perhaps the most important feature of Chetverikov's ideas was that they led to the first genetic analysis of a natural population, begun in 1925 and first reported two years later. Notice that the three separate lines of thought outlined above led Chetverikov to conclude that natural populations should contain a large amount of cryptic variabflity. If, because of continual natural mutation, the maintenance of the resultant mutants, and the slower elimination of recessive (hence hidden) mutants by selection, species "soaked up" mutations "like a sponge" while remaining phenotypically uniforn, Chetverikov reasoned that an inbreeding of samples from wild populations should allow these mutations which are masked in the heterozygous condition to become homozygous and thus to be expressed. discussion, Chetverikov also treats other reasons why these mutations would not be noticed in natural populations, e.g. their frequently lower viabilities. The quotation is from Lerner, p. 177. 33. Ibid. (Lerner trans., p. 178; Russian reprint, p. 48). In this and all quotations used in the text, the Italics are those of Chetverikov. 34. Ibid. (Lerner trans., p. 182; Russian reprint, p. 56.)

34

The Founding of Population Genetics To test his reasoning, Chetverikov and his students35 captured 239 wild female Drosophila melanogaster which had already been fertilized in nature, mated the Fl's brother x sister, and examined the F2's. No less than 32 different hereditary characters which had been masked heterozygotically were found.36 Chetverikov understood the evolutionary significance of his results: All these facts confirm the conclusion that the usual wild populations are extraordinarily heterozygous and so at any given time have a rich supply of inherited variations which, with changes in the environment, can be useful and so must play a decisive role for the evolutionary process.37 In a piece of parallel work Timofeev-Resovsky (1927) analyzed 78 females of Drosophila melanogaster from Berlin and found similar results.38 The experimental investigations of the Chetverikov group (1925-1929) had been much more extensive than Chetverikov's brief communication in 1927 before the Fifth International Congress of Genetics had indicated.39 Studies had been made of a whole range of naturally occurring Drosophila species from around Moscow: Drosophila phalerata (by B. L. Astaurov and N. K. Beliaev), Drosophila transversa (B. L. Astaurov), Drosophila vebrissina (E. I. Balkashina), and Drosophila obscura (S. M. Gershenson), and a study had been made of Drosophila melanogaster from Gelendzhik, near the Causasian coast of the Black Sea. But as a result of the breakup of the Chetverikov group which followed his precipitous departure from Moscow, 35. Astaurov, "Two Landmarks.. ." lists the students who participated in "this first work in Moscow": B. L. Astaurov, E. I. Balkashina, N. K. Beliaev, S. M. Gershenson, I. F. Rokitskii, and D. D. Romashov (p. 26). 36. The only report by Chetverikov of this work was given at the Fifth International Congress of Genetics and published in the form: Tschetwerikoff, S. S., "Uber die genetische Beschaffenheit wilder Populationen," Z. Induktive Abstammungs-Vererbungslehre, 46 (1928), pp. 38-39. (The spellings given in the text for Russian authors are generally transliterated from the Russian. Hence "Chetverikov"-though in German sources the name is variously spelled "Tschetwerikoff," "Tschetwerikov," and "Tschetverikov;" likewise, "N. V. Timofeev-Resovsky," instead of the German "N. W. Timof6ef-Ressovsky," under which most of his works published in Germany, 1925-1945, appear; also "Koltsov," instead of "Koltzoff," or other variants.) 37. Ibid., p. 39. 38. H. A. Timofeeff-Ressovsky and N. W. Timofeeff-Ressovsky, "Genetische Analyse einer freilebender Drosophila melanogaster Population," Roux Arch. Entz. Mech. Organ, 109 (1927), pp. 70-109. 39. Astaurov, "Two Landmarks. . ." gives a good description of the work of the Chetverikov group by one of its members.

35

MARK B. ADAMS

these results were only published fragmentarily. Hence the results of the Moscow sampling were first published in 1934 (by Gershenson40) and in 1935 (Balkashina and Romashov41). The analysis of the genetic variability of wild populations of Drosophila, guided by Chetverikov until his departure from Moscow, was taken up in 1930 by Dubinin. A student of Chetverikov's at Moscow University until 1928, Dubinin had worked with A. S. Serebrovsky on the genogeography of domesticated fowl in 1929. Dubinin's first paper on Drosophila population genetics was based on research undertaken with fourteen coworkers.42 Published in 1934, it is especially significant because it yielded a number of surprising and interesting results which stimulated a great number of later studies. Dubinin and his collaborators collected samples from wild populations of Drosophila melanogaster from nine localities in the Caucasus and one in Central Russia in 1931 and 1932. They found some 61 identifiable mutants, which ranged in frequency from 3.9% to 33.1%. The concentrations and nature of the mutants found varied with the geographical source and within one source from year to year. Some of the mutations appeared identical to those obtained in laboratory strains, others were new alleles; some were present in all localities, others only in one. Dubinin's paper is also apparently the first to analyze the chromosomal polymorphism of natural populations, work that was later to be developed by Dobzhansky (beginning some four years later in 1938.) 4 Hence Dubinin's study demonstrated chromosomal and genomic variability, but perhaps the most surprising result came from studies of the frequency of lethal recessives. Chet40. S. M. Gershenson, "Mutant Genes in a Wild Population of Drosophila obscura," Amer. Naturalist 68 (1934), p. 569. 41. E. I. Balkashina and D. D. Romashov, "Geneticheskoe stroenie populiatsii: 1. Geneticheskii analiz Zvenigorodskikh (Moskovskoi oblasti) populiatsii Drosophila phalerata Meig., transversa Fall. i vibrissina Duda." (The genetic structure of populations: 1. The genetic study of ZvenigorodBiologicheskii Zhur. 4, no. 1. skii (Moscow region) populations of...) 42. N. P. Dubinin, M. A. Heptner, S. Iu. Bessmertnaia, S. Iu. Goldat, K. A. Panina, E. Pogossian, S. W. Saprikina, B. N. Sidorov, L. W. Ferry, Analiz Ekogenotipov Drosophila M. G. Tsubina, "Eksperimental'nyi melanogaster," 1 (Experimental study of the ecogenotypes of D. melanogaster), pt. 1, Biologicheskii Zhur. 3 (1934), pp. 166-205. N. P. Dubinin, M. A. Heptner, Z. S. Nikoro, S. Iu. Bessmertnaia, W. N. Beliaieva, Z. A. Demidova, A. P. Krotkova, E. D. Postnikova; ibid., pt. 2, Biologischeskii Zhur., 3 (1934), pp. 206-216. 43. Th. Dobzhansky and M. L. Queal, "Genetics of Natural Populations: 1. Chromosome Variation in Populations of Drosophila pseudoobscura Inhabiting Isolated Mountain Ranges, I and II," Genetics, 23 (1938), p. 239; p. 463.

36

The Founding of Population Genetics verikov, it should be recalled, simply mated brothers and sisters, and analyzed the F2's for mutant traits. Such analysis will tell nothing about the frequency of lethals, however, since flies carrying homozygous lethals will simply not appear in the progeny to be counted. When new techniques of genetic analysis were used (the CIB technique, for example), the frequency of lethal mutations in the 10 natural populations ranged between 0% and 21.4%. In particular, 10-20% of the total number of second chromosomes analyzed carried recessive lethals. This outcome had not been expected at the time-Dobzhansky called it a "novel result-and a very startling one."44 However, followup experiments done by a number of investigators corroborated Dubinin's findings.45 Dubinin's paper, by demonstrating the great allelic and genomic variability present in natural populations, became the first of a long series of such studies, to which he substantially contributed until 1948. The Russian School is important both because of what it ended and what it began. Many authors have alluded to the estrangement between two traditions in biology which characterized its history in the early decades of this century: the "experimentalist" and the "naturalist" traditions.46 It is significant, then, that the Russian School is one of the earliest to draw from both traditions in order to clarify the evolutionary process. Its founder, Chetverikov, was an entomologist, a biometrician, and a geneticist. Indeed, his great theoretical paper set as its purpose the resolution of this split, and it drew heavily on natural history studies for species notions and the theory of isolation; on mathematical studies-for example, those of Hardy, Pearson, and Norton; and on the genetical studies of the Morgan School. And by turning the techniques of genetics onto the problems of evolution in a natural setting, he did much to heal the unfortunate gap between the naturalists and experimentalists in biology-in effect, by creating experi44. Th. Dobzhansky, "Concepts and Problems of Population Genetics," in Cold Springs Harbor Symposia in Quantitative Biology, vol. XX, p. 4. 45. For example, C. Gordon, "The Frequency of Heterozygosis in Freeliving Populations of Drosophila melanogaster and Drosophila subobscura," J. Genet. 33 (1936), 25-60. Sturtevant also did a follow-up study, as he mentions in his A History of Genetics (New York, 1965), p. 110. 46. For example, N. W. Timofeef-Ressovsky, "Mutations and Geographic Variation," in Julian Huxley, New Systematics (Oxford, 1940). See also Julian Huxley, Evolution: The Modern Synthesis (New York: Harpers, 1942), pp. 24-25. Also, Th. Dobzhansky, Genetics and the Origin of Species, lst ed. (New York: Columbia University Press, 1937).

37

MARK B. ADAM S

mental population genetics and making evolutionary theory experimental. The work of Chetverikov and the members of his school had shown the great possibilities in the genetic analyses of natural populations, in particular of Drosophila. The impact of these efforts was blunted by the breakup of the group in 1929 before the bulk of the material had been published, but it came nonetheless. Timofeev-Resovsky continued the work in Germany; Dubinin took up the studies: after the publication of their key works in 1934, together with the belated publication of the findings of the Chetverikov group, population genetics took on a dynamic of its own. In Russia the work proceeded apace: the work was continued by a whole team of investigators until 1948, including N. R. Beliaev, R. L. Berg, S. M. Gershenson, G. D. Muretov, I. M. Olenov, A. N. Promptov, D. D. Romashov, and G. G. Tiniakov, among others. Abroad, Dubinin's work led to confirming experiments in England by Gordon et al.,47 and in the United States by Sturtevant,48 and to the first in a momentous series by Dobzhansky and associates in which he credits Chetverikov, Dubinin, Timofeev-Resovsky, Gordon and Sturtevant with "opening new vistas" by investigations of the genetics of free-living populations-a subject "hitherto almost untouched."49 The ideas of the Russian School on the "genetic background," or the "genotypic milieu," did not have the same kind of immediate impact, at least on theoretical formulation. But their implications are profound. For example, if a gene's effect depends greatly on its genetic and environmental background, then alleles cannot be assigned fixed "fitness" values. It might also be noted that from this work follows the important idea that aberrant phenotypes are not necessarily due to the presence of single mutant genes, but may be rather the result of certain combinations of genes relatively frequent in a population. Thus, the aberrant vti phenotype, which occurs only very rarely in natural populations, is the result of a major gene, vti, which is relatively frequent in natural populations, interacting with a and also later Cecil Gordon, Helen 47. C. Gordon, "The Frequency...," Spurway and P. A. R. Street, J. Genet, 38 (1939). The references listed at the end of the 1939 piece, twenty-five in all, include most genetic analyses eighteen in number. of wild populations done prior to that time-some Significantly, some thirteen of these had been done by members of the or Russian School: Chetverikov or his students, Timofeev-Resovsky, Dubinin and colleagues. 48. A. H. Sturtevant, "Autosomal lethals in Wild Populations of Drosophila pseudoobscura," Biol. Bull. Wood's Hole, 73 (1937), 542-51. 49. Dobzhansky and Queal, "Genetics of Natural Populations, I," p. 463.

38

The Founding of Population Genetics large number of modifying genes (including of course ri). I might add that the enormous implications of this conclusion for eugenics have only very recently been appreciated. To a considerable degree, then, recent investigations on "gene strategy," "genetic homeostasis," and other modem researches on the interrelation of genes in various systems are indebted to the notions of the genotypic milieu, developed by the Russian School. Ernst Mayr has distinguished "classical population genetics" which presented evolutionary change as essentially an input or output of genes, from the "newer population genetics" in which a gene can have a constellation of selective values, depending on its genetic and environmental backgrounds.50 If we accept this distinction, it is clear that conceptually and experimentally the Russian School had laid the basis for the "newer population genetics" even while the "classical" was being enunciated. Acknowledgments Tlhe author would like to thank Dr. Ernst Mayr and Dr. Roger Milkman for reading the manuscript in an earlier form and offering criticisms and suggestions. 50. Ernst Mayr, "Where are We?" Cold Springs Harbor Symposia, vol. XXIV (1959), p. 2.

39

Trigonia and the Originof Species STEPHEN JAY GOULD Museum of Comparative Zoology, Harvard University, Massachusetts

Cambridge,

INTRODUCTION Neotrigonia, a marine clam found only in Australian waters, shares with the coelacanth and Neopilina a special place in the history of zoology. These "living fossils" are the sole survivors of once abundant groups. In each case, the unexpected and accidental discovery of a modem form refuted the accepted opinion that its group had disappeared 70 to 350 million years ago. The discovery of modem trigonians ranks poorly among the three in zoological significance. Of living fish, the coelacanth bears closest affinity to ancestors of man and the higher vertebrates. Neopilina has offered startling evidence of relationships among major invertebrate phyla.' The discovery of modem trigonians, on the other hand, led to few important conclusions beyond those implicit in the mere report of their existence. Yet, as an event in the history of science, the case of Neotrigonia stands out among the three. As discoveries of the last thirty years, living coelacanths and Neopilina provoked no discussion of life's basic nature; in our epoch, evolutionary biology has been operating as "normal science."2 Modem trigonians, however, were found in 1802, at the outset of the greatest conceptual upheaval in the history of biology. All the mutually exclusive yet reasonably consistent pre-Darwinian theories of 1. The body of Neopilina is segmented as in annelid worms and arthropods, thus indicating the relationship of the Mollusca to the articulate phyla. Neopilina is the only known living representative of the molluscan class Monoplacophora. 2. T. S. Kuhn's designation of research within a paradigm accepted by virtually all practitioners of a science. The paradigm, in this case, is the synthetic theory of evolution-Darwinism bolstered by a particulate theory of inheritance applied to the study of natural populations. See T. S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962).

41

STEPHEN

JAY GOULD

life had to accommodateanomalies presentedby these molluscs. Moreover,some of the most illustrious participants in the debates of Darwin's century took special interest in Neotrigonia, for Lamarck first reported the discovery, and Louis Agassiz wrote the definitive pre-Darwinian monograph on trigonians. A compilation of pre-Darwinianinterpretationsof the discovery provides, in mature, a fine summary of three major nineteenth-centuryviews of the history of life: the evolutionism of Lamarck, the nonevolutionary Chrstian progressionism of James Parkinson, and the more static creationism of Louis Agassiz. Beyond this, the account has many aspects of a good story-from the initial unexpected find to a later discovery and "happyending"which demolished the interpretivetactic of antievolutionists and provided,in 1865, one of the early important vindications of Darwin's views on the imperfection of the geologic record.3 THE DISCOVERY The European founders of modern paleontology were well acquainted with Trigonia, a clam with distinctive hinge teeth and triangular shape "recognizedwithout difficulty even by a tyro."4 In William Smith's system, species of Trigonia served as

indices for several of the Secondaryformations,5but were conspicuously absent from all strata above the chalk. Their demise at the top of the Cretaceous System coincided with that of the ammonites and marked a major event in the history of life. For catastrophists, this event was one of the most substantial of pre-Noachianparoxysms. With a good theoretical reason for regarding them as permanently extinct, many naturalists were especially surprised when P. P6ron, "naturaliste 6clair6 et plein d'ardeur,"6found several modem trigonian shells washed up upon the beaches of 3. The history of evolution is usualy told by tracing general ideas as expressed by human participants in the debate. I write this with the conviction that one might also focus profitably on the objects of debate, viz. J. C. Greene's chapter on fossil elephants in John C. Greene, The Death of Adam (New York: New American Library, 1961), pp. 96-133. 4. John Lycett, A Monograph of the British Fossil Trigoniae (London: Society, 1879), p. 1. This is the major nineteenthPalaeontographical century work on trigonians. 5. The name used by Smith for strata now termed Mesozoic. See Table 1 for a modern geological time scale. 6. J. P. B. Chevalier de Lamarck, "Sur une nouvelle espbce de Trigonie, et sur une nouvelle d'Hultre d6couvertes dans le voyage du capitaine Baudin," Annales du Muse6um d'Histoire Naturelle, 4 (1804), 353-354. Hereafter abbreviated as Trigonie.

42

Trigonia and the Origin of Species Southern Australia and Tasmania in 1802. Two years later, Lamarck described these shells as Trigonia margaritacea,7 taking his specific designation from the attractive pearly nacre of the interior (Fig. 1). The inhabitants of Peron's shells were still unknown. J. Quoy and J. Gaimard, naturalists aboard the Astrolabe (1826-1829), made special efforts to secure a complete specimen. For several days they dredged without success until one night, becalmed in Bass Straight, they brought up a very small intact individual: We were so anxious to bring back this shell with its animal that when we were, for three days, stranded on the reefs of Tonga-Tabu, it was the only object that we took from our collection. Doesn't this recall the ardent shell collector who, during the seven years' war, carried constantly in his pocket an extraordinary Phasianella which he had bought for twentyfive louis.8 Two aspects of the story, as known during the years 18041865, commanded the attention of theoretically minded naturalists: Morphology: The modern trigonians differed markedly in appearance from Mesozoic forms, whose triangular shape and discrepant ornament stood in contrast to the ovate profile and simple radial ornament of Lamarck's recent species (Fig. 2). Distribution: The geologic distribution of trigonians seemed to be disjunct; no specimens had been recovered from any Tertiary formation (Table 1). It is in their reactions to these two observations that our three naturalists, Lamarck, Parkinson, and Agassiz, invoked the precepts of their characteristic and contradictory theories of the history of life. CHEVALIER DE LAMARCK Morphology. In his works of 1802 and 1809,9 Lamarck cites two basic 7. The genera of nineteenth-century naturalists were far more inclusive than those used today. Trigonia, as understood by Agassiz, included forms now classified among more than twenty genera, one of which, Neotrigonia, includes all modern forms. This is a convention of naming only and does not indicate any change of view regarding the relative similarity of the forms involved. 8. J. Quoy and J. Gaimard, Voyage de d6couvertes de l'Astrolabe ex&cutE par ordre du Roi pendant les ann6es 1826-29, Zoologie, Tome III (Paris: J. Tatsu, 1834), p. 474. 9. Lamarck became convinced of the transmutation of species during the period 1797-1800 and first expressed his evolutionary ideas as sum-

43

STEPHEN

JAY GOULD

TABLE 1. The Geologic Time Scale ERA

PERIOD

EPOCH

Cenozoic

Quaternary Tertiary

Pleistocene Pliocene Miocene Oligocene Eocene Paleocene

Mesozoic ("Secondary" of early geologists).

Cretaceous("Chalk") Jurassic Triassic

Paleozoic

Permian Carboniferous Devonian Silurian Ordovician Cambrian

"The Tertiary period began approximately 70 million years ago, the Triassic about 225 million, and the Cambrian about 600 million.

causes for the orderand diversityof the animal kingdom. Living matter of simple form, generated spontaneously at the base of l'tchelle des etres, is advanced toward higher structural levels by the "cause which tends incessantly to complicate organization."10This inherent self-elaboratingpropertyof organic matter would produce an evenly graded progressive sequence were it mary propositions in Syst6me des animaux sans vert0bres (1800). The Recherches sur l'organisation des corps vivans (Paris: Maillard, 1802) served as a model for Lamarck's best-known evolutionary work, the Philosophie zoologique (Paris: Dentu, 1809). When PNron discovered modern trigonians, therefore, Lamarck's evolutionary ideas were still in the formative stage. On the chronology of Lamarck's evolutionism, see C. C. Gillispie, "Lamarck and Darwin in the History of Science," in Forerunners of Darwin: 1745-1859, eds. B. Glass, 0. Temkin and W. L. Straus, Jr. (Baltimore: Johns Hopkins Press, 1959), pp. 269-271. 10. La cause

qui tend

sans

cesse

d composer

Lamarck,

l'organisation.

Philosophie zoologique, p. 132. Lamarck's clearest discussion of vertical and tangential forces producing progress and diversity is found in chap. vi of this work, pp. 130-137. The discussion in Recherches sur l'organisation

des

corps

vivans,

the same basic contentions.

44

pp.

3943,

is less

elaborate

but

stresses

Trigonia and the Origi of Species not for "the influence of circumstances," the tangential force which elicits a creative organic response to specific environmental conditions. Reduction of eyes in tunneling moles, loss of teeth in anteaters, and the development of webbing in waterfowl are cited by Lamarck" as deviations from the main course of development. Lamarck's phyletic model is not a tree, but a stick, or rather a column, with myriad "lateral ramifications, the extremities of which represent truly isolated points."12 Lamarck attributes the morphological differences between fossil and modem trigonians to the operation of these tangential forces which produce organic diversity within morphological levels of organization. Of P6ron's shells, he writes:18 They have undergone changes under the influence of circumstances'4 which act upon them and which have themselves changed, such that the fossil remains which we collect of those that lived in the most ancient periods may display several differences from those of animals of the same type which live now but which are nevertheless derived from them. Distribution. It is a consequence of Lamarck's evolutionary views that lineages do not become extinct.'5 There are two main justifications for this contention. First, organisms possess a great capacity for responding creatively to felt needs, and should therefore be able to adapt to the exigencies of a changing environment.'6 11. Lamarck, Recherches sur l'organisation des corps vivans, pp. 54-56. 12. Ibid., p. 42. 13. Lamarck, Trigonie, pp. 352-353. 14. L'influence des circonstances is Lamarck's usual designation of the tangential forces. Lamarck did not believe, as did some of Darwin's forerunners, that organisms were passively modified by direct environmental pressures; it is rather the creative response of organisms to felt needs that effects the transformation through the inheritance of characters acquired by use or lost by disuse. In seeming to imply in the above statement that environment itself produces the morphological change, Lamarck indulges in the same kind of ellipsis that modern evolutionists use in writing, for example, that cave-dwelling vertebrates tend to lose their eyes because light is absent from their environment. This seems to imply the inheritance of characters lost by disuse, but is, by tacit convention, only a shortcut expression omitting statements about selective pressures for reduction of eyes. 15. Extinct, that is, in the sense of termination. Unfortunately, evolutionists also speak of extinction when a species is transformed into another. 16. As an early believer in the uniformity of geologic processes (as expressed in his Hydrogeologie), Lamarck did not believe that environmental changes in times past occurred with the rapidity and magnitude envisioned by catastrophists and diluvialists. To Lamarck, the most

45

STEPHEN

JAY GOULD

Second, extinction is inconsistent with the "dynamic steady state"17 of the Lamarckian system. Simple organic forms are spontaneously generated, progress toward man, are degraded and regenerated. If lineages often became extinct without issue, the amount of organic matter participating in the flux would progressively diminish and lead eventually to the termination of life. In 1804 the debate among naturalists as to the possibility and prevalence of extinction was at its height. Lamarck interpreted the discovery of modem trigonians as a striking confirmation of his view that seemingly extinct marine forms will be found living in the depths of the sea beyond the normal range of human observation. I believe that if any species of the animal kingdom are truly extinct [perdues], these can only be large terrestrial forms. Man, having multiplied and spread to all regions of the globe, has diminished by his presence the number and extent of deserts and virgin lands and has destroyed wild species by hunting them and driving them from their habitats. But small species, especially those which dwell in the depths of the sea, have the means to escape man; truly among these we do not find any that are really extinct.'8 Lamarck ends this section with a prediction, still unfulfilled and unlikely ever to be realized, that the discovery of living ammonites, hippurites, and belemnites will not be long delayed.19 JAMES PARKINSON Lamarck never doubted that the discovery of modem trigonians implied the continuous existence of this group throughout the entire Tertiary period. To a believer in multiple creations, however, the absence of Tertiary forms could be taken at face value and a case built for attributing the production of modern trigonians to a recent exercise of creative power. In 1811 the catastrophic event ever to befall terrestrial organisms was the spread of man, and it is here that he allowed an exception to the proposition that lineages do not become extinct by termination: man may have extinguished some large terrestrial vertebrates. 17. G. G. Simpson, "Three Nineteenth Century Approaches to Evolution," in This View of Life (New York: Harcourt, Brace and World, 1964), p. 47. 18. Lamarck, Trigonie, p. 352. 19. Of these three extinct groups of molluscs, ammonite6, and hippurites are unknown above the Cretaceous, belemnites above the Eocene.

46

Trigonia and the Origin of Species surgeon-paleontolog1st James Parkinson published the third and last volume of his treatise, Organic Remains of a Former World. At this time, Parkinson was a pious progressionist who viewed the history of life as a sequence of creations of continually increasing excellence. "The facts are indubitable," he writes, "and afford a direct proof of the Creator of the universe continuing and superintending providence over the works of his hands."20 He was particularly delighted by the presumed correspondence of organic succession with the Mosaic order of creation. "So close indeed is this agreement, that the Mosaic account is thereby confirmed in every respect, except as to the age of the world, and the distance of time between the completion of different parts of the creation."2' This single discrepancy was removed by an allegorical interpretation of the "days" of Genesis. If modem trigonians were a product of that recent and most excellent creation which first placed man upon the earth, mn what sense could they be regarded as improved upon their Mesozoic counterparts? So acute did this problem appear to Parkinson that of the twelve statements which conclude his book and summarize the argument, the trigonian dilemma is accorded an entire statement, the length of which exceeds all others more than twofold. Parkinson's conclusion, admittedly less than satisfactory, was simply to point out that the modem trigonians belonged to a species unknown in any previous creation. Its superiority, though not at all evident from morphology, is at least known to its maker. Some fossil shells (trigonitae) are found in the Lias and in most of the succeeding strata, and sometimes, but very rarely, in the hard chalk. After this they are not seen in the remaining superior strata, but of late years one species has been found in our present seas. This however requires some explanation . . . this shell, although really of this genus, is of a different species from any shell, which has been found in a fossil state. So that none of the species of shells of this genus, which are known in a fossil state, have, in fact, been found in any stratum above the hard chalk, or in our present seas.22 During the succeeding decade, Parkinson strongly modified 20. James Parkinson, Organic Remains of a Former World, Examination of the Mineralized Remains of the Vegetables and Animals of the Antediluvian World Generally Termed Extraneous Fossils (London: Sherwood, Neely and Jones, 1811), p. xiv. 21. Ibid., p. 451. 22. Ibid., p. 454.

47

STEPHEN

JAY GOULD

his progressionistviews.23We do not know to what extent the persistent trigonian anomaly fostered his receptiveness to nonprogressionisttenets, but the discussion of molluscan history in his Outlines of Oryctologyemphasizes the importance he attached to the fossil record of this phylum. It might have been expected that those beings which had possessed life under its most simple modifications,would be found in the earliest formed strata; and that, in proportion to the lateness of the period at which the strata were formed, would be the degree of complexity in the organizationof the inhabitants whose remains they would contain. But investigation has ascertained . . . that such a conjecture is ill founded. In the carboniferous and the mountain limestone are the remains of shells of the earliest creation, which are unexpectedly found, with hardly an exception, to exceed in complexity of structure, all the shells which have been discovered either in any subsequent formation, or living in our present seas.24 LOUIS AGASSIZ Morphology.

Louis Agassiz also believed that creative power brought species into beimgat many different times in earth history, yet he could discern no general theme of gradual structural advance in successive creations. That animals are constituted according to four permanent structural plans25 which can in no way be one into another was a central contention of Agastransformned siz's static system. In various scattered statements, particularly in early works which bear the influence of his teacher Oken and his early fascination with Naturphilosophie, Agassiz allowed for creative progressionwithin embranchements,26yet through23. Parkinson was acutely sensitive to the pressures of contradictory evidence. F. C. Haber documents an earlier transition from literal belief in the Mosaic chronology to incorporation of Cuvier's work and acceptance of the allegorical interpretation of Genesis. F. C. Haber, The Age of the Johns Hopkins Press, 1959), World, Moses to Darwin (Baltimore: P. 199. 24. James Parkinson, Outlines of Oryctology, an Introduction to the Study of Fossil Organic Remains (London: Sherwood, Neely and Jones, 1822), p. 246. 25. The four embranchements of Cuvier: Radiata, Mollusca, Articulata, and Vertebrata. 26. L. Eiseley, who portrays Agassiz as a progressionist, cites some of these. L. Elseley, Darwin's Century (New York: Anchor Books, 1961), pp. 327-328.

48

FIG. 1. Lamarck's original figure of Neotrigonia margaritacea.

FIG. 2. Comparison of recent and Mesozoic trigonians. Left: the right valve of a recent Neotrigonia margaritacea; actual length 47 mm. Center: Lamarck's original illustration of the left valve of Neotrigonia margaritacea; actual length 42 mm. Right: the left valve of a Cretaceous fossil trigonian from Tennessee; actual length 53 mm. The fossil is more elongate and has discrepant ornament (ribs along upper border radiate in a different direction from those covering the rest of the shell).

Trigonia and the Origin of Species out his career, and especially in his refutation of Darwinism, Agassiz maintained that the geologic distribution of fossils precluded any notion of improvement through time in the history of life. In his early work on fossil fishes, Agassiz denies the claims of transmutationists27 by showing that the earliest representatives of his four great groups of fishes were contemporaneous.28 In refuting Darwin, he argues for an approximately average mean complexity of organization throughout time. He [Darwin] would have us believe that the oldest organisms that existed were simple cells, or something like the lowest living beings now in existence; when such highly organized animals as Trilobites and Orthoceratites are among the oldest known. He would have us believe that these lowest first born become extinct in consequence of the gradual advantage some of their more favored descendants gained over the majority of their predecessors; when there exist now, and have existed at all periods of past history, as large a proportion of more simply organized beings, as of more favored types . . . He would have us believe that the most perfect organs of the body of animals are the product of gradual improvement, when eyes as perfect as those of the Trilobite are preserved with the remains of these oldest animals.29 Finally, in the posthumous essay on "Evolution and Permanence of Type," Agassiz writes: 'The whole history of geological succession shows us that the lowest in structure is by no means necessarily the earliest in time, either in the Vertebrate type or any other."30 Thus, the problem of Parkinson and the progressionists simply did not exist for Agassiz. Holding no a priori expectation that a recreated modem trigonian should be in any way superior to its Mesozoic counterparts, Agassiz viewed the difference between Mesozoic and modern forms as no more important than the differences among Mesozoic species. In his monographic 27. Like the progressionists, many early transmutationists viewed the history of life as a simple sequence of gradual improvements, though produced by secondary causes rather than directly by the First Cause. 28. Louis Agassiz, Les poissons fossiles, I (Neuchatel: Petitpierre, 1840), 172. 29. Louis Agassiz, 'Prof. Agassiz on the Origin of Species," American Journal of Science, 30 (1860), 145. 30. Louis Agassiz, "Evolution and Permanence of Type, Atlantic Monthly, 33 (1874), 101. For an excellent summary of Agassiz' ideas on the geologic record see: Ernst Mayr, "Agassiz, Darwin and Evolution," Harvard Library Bulletin, 13 (1959), 187-190.

49

STEPHEN

JAY GOULD

treatment of trigonian clams,31Agassiz includes all living species in one of his eight sections of the genus Trigonia-les

pectindes. Distribution. While not convinced of its reality, Agassiz was intrigued by the Tertiary gap in trigonian distribution: The absence of Trigoniain Tertiarystrata is a very important fact for discussions of the origin and relationships of species of different epochs; for if it could one day be shown that Trigonia never existed throughoutthe entire durationof Tertiary time, it would no longer be possible to maintain the principle that species of a genus living in successive geological epochs are derived from each other.32 Agassiz goes on to assert that the discoveryof Tertiaryforms would in any event be irrelevant to his belief in the fixity of species. In a fine example of his Platonic contention that the relationshipamong species is akin to that among the component parts of a system of ideas,38Agassiz writes: Nevertheless, although I now invoke this fact [the Tertiary gap] to support my conviction that the different species of a genus are not variants of a single type that have become fixed with the passage of time, the discoveryof a Tertiarytrigonian would still not demonstrate,to my eyes, that the relationship among species of a genus is one of direct descent and successive transformation of original types . . . I certainly do

not deny that natural relationships exist among different species of a genus; on the contrary,I am convinced that species are related to each other by bonds of a higher nature 31. Louis Agassiz, "M6moire sur les trigonies," in Etudes critiques sut les mollusques fossiles (Neuchatel: Petitpierre, 1840). Hereafter abbreviated Mdmoire. 32. Agassiz, Mdmoire, pp. 2-3. The last statement reflects Agassiz' belief in catastrophism. In labeling Agassiz a static creationist, I did not mean to imply a disbelief on his part in multiple creations through time, but rather to indicate his contention that no general theme of temporal improvement could be discerned among the many creations. 33. This is more than an analogy since the forms of species are ideas in the Creator's mind. Therefore, species relationships mirror the way in which God composes units of thought into ultimate idea systems. As Agassiz wrote in his textbook (Principles of Zoology by Agassiz and A. A. Gould): "To study . . . the succession of animals in time, and their distribution in space, is therefore to become acquainted with the ideas of God himself." Quoted in Edward Lurie, Louis Agassiz, A Life in Science (Chicago: University of Chicago Press, 1960), p. 87.

50

Trigonia and the Origin of Species than those of simple direct procreation, bonds which may be compared to the order of a system of ideas whose elements, developed at different times, form in their union an organic whole-although the elements of each time period also appear, within their limits, to be finished products.34 TRIGONIA AND THE ORIGIN OF SPECIES Darwin, Trigonia and Catastrophism. The anti-evolutionary viewpoints of men like Parkinson and Agassiz had to be reconciled with indisputable fossil evidence of "former worlds" inhabited by species no longer living. The paradox of combining a static species concept with a history of change was resolved, as we have seen; by the postulate of multiple creations in time. However, the notion of multiple creations required a mechanism for removing the species of previous worlds to make room for those of the next. Most antievolutionists, Parkinson and Agassiz included, were catastrophists; they attributed these required extinctions to rapid geologic events, often of global magnitude-floods, mountain upheavals, sudden changes in temperature.35 One consequence of catastrophist premises was subject to verification by the fossil record: extinctions should be sudden and affect many groups at once. Charles Darwin, on the other hand, insisted that extinction was a natural consequence of evolution.36 If species disappear as a result of competition and natural selection, then the frequency of extinctions should be relatively constant in time, for these processes are in continuous operation. Moreover, since competition and natural selection operate slowly, the extinction of a species should be proceeded by a gradual dwindling in numbers of individuals and geographic range. These were the testable and contradictory predictions of catastrophism and Darwinian evolutionism. Darwin's task was not simply to provide evidence of change in the fossil record (for this was admitted by all), but rather to show that change occurred in a certain way. He was acutely aware of the need 34. Agassiz, Mt6moire, p. 3. 35. As author of the theory of ice ages, Agassiz referred to his glaciers as "God's great plough." He wrote to William Buckland: "I . . . will have the whole surface of the earth covered with ice, and the whole prior creation dead by cold." Quoted in Edward Lurie, Louis Agassiz, A Life in Science, pp. 98-99. 36. "We may believe that the production of new forms has caused the extinction of about the same number of old forms." Charles Darwin, On the Origin of Species (lst ed., Cambridge, Massachusetts: Harvard University Press, 1964), p. 320.

51

STEPHEN

JAY GOULD

for examples of gradual extinction, especially since the extremely imperfect geologic record could be expected to provide many appearancesof rapid elimination. Darwin therefore focused on a third aspect of Trigonia-its geographic distribution in time. He used Trigonia as a prime example of a once-dominantgroup which had been drastically restricted im diversity, abundance, and geographic range: "We have reason to believe that . . . the numbers of the species decrease till finally the group becomes extinct . .. The Palaeother-

ium was extinct much sooner in Europe than in India: the Trigonia was extinct in early ages in Europe, but now lives in ChapterV of this 1844 essay ("Gradual the seas of Australia."37 Appearance and Disappearance of Species") is devoted to refuting the catastrophist predictions. We may assume that Tngonia was again in his mind when he wrote: The view entertained by many geologists, that each fauna of each Secondaryepoch has been suddenly destroyed over the whole world, so that no succession could be left for the production of new forms is subversiveof my theory, but I see no grounds whatever to admit such a view . . . As far as is

historically known, the disappearance of species from any one country has been slow-the species becoming rarer and rarer, locally extinct, and finally lost . . . It has happened,

also, that shells common in a fossil state, and thought to have been extinct, have been found to be still living species, but very rare ones."38 The Discovery of Tertiary Trigonians.

Two erroneousaccounts of Tertiarytrigonianswere published during the years intervening between Agassiz's M6moiresur les trigonies and Darwin's Origin of Species. In 184239 Alcide 37. Charles Darwin, The Foundations of the Origin of Species, Two Essays Written in 1842 and 1844, ed. Francis Darwin (Cambridge University Press, 1909), pp. 198-199. Palaeotherium is related to the ancestors of horses. 38. Ibid., pp. 145-147. Francis Darwin has here appended the following note in this 1909 edition of his father's work: "The case of Trigonia, a great Secondary genus of shells surviving in a single species in the Australian seas, is given as an example in the Origin, ed. i. p. 321." The statement in the Origin is made in a slightly different context-to demonstrate that the complete extinction of a large group is generally a slower process than its origin. Three species of modem trigonians had been described by 1859. Darwin cited but one in the first edition of the Origin, but rectified his error in the third and subsequent editions. 39. A. d'Orbigny, Voyage dans l'Amirique mridionale . . . exicutd

52

Tngonia and the Origin of Species D'Orbigny described Trigonia hanetiana from supposed middle Eocene strata in South Amenica, but by 185040 he had decided that these rocks were Senonian (upper Cretaceous) in age. C. G. Giebel described a single poorly preserved mold and shell fragment from Oligocene rocks in central Germany as Trigonia septaria in 1852, but the delicate, truly indigenous molluscan fossils of this formation are extremely well preserved and the Trigonia fragment was probably eroded from older rocks and redeposited during Oligocene times.4' No other Tertiary trigonian has ever been reported from Europe. This and other similar anomalies led Darwin to write in his chapter, "On the Geologic Succession of Organic Beings": A group does not reappear after it has once disappeared; or its existence, as long as it lasts, is continuous. I am aware that there are some apparent exceptions to this rule, but the exceptions are surprisingly few, so few, that E. Forbes, Pictet and Woodward (though all strongly opposed to such views as I maintain) admit its truth, and the rule strictly accords with my theory.42 Darwin might well have cited the catastrophist D'Orbigny among nonevolutionist supporters of this principle, for d'Orbigny had written of his supposed Tertiary trigonian: "It is the only species that has been found in Tertiary strata. This fact, although new, is not extraordinary since there exists a living species."43 Although Darwin was unperturbed by this particular problem, general difficulties of reconciling the observations of paleontology with a theory of slow and gradual evolutionary change troubled him deeply, for the geologic record of life, as known in 1859, was one of discontinuity and absence of transitional forms. With characteristic candor, he writes: "All the most eminent paleontologists . . . and all our greatest geologists . . . have unanimously, often vehemently, maintained the immutability of species"44 pendant les anndes 1826 d 1833, PalMontologie, voL III, pt. 4 (Paris: P. Bertrand, 1842), pp. 127-128. 40. A. d'Orbigny, Prodrome de palsontologie, U (Paris: 1850), 240. 41. A. Briart, "Sur le genre Trigonia et descriptions de deux Trigonies nouvelles des terrains supra-CrEtac6s de Maestricht et de Ciply," Annales de la SociMtWMalacologique de Belgique, 3 (1888), 325-339. This is not special pleading on Briart's part. Remani6 fossils are very common in the geologic record. 42. Charles Darwin, On the Origin of Species, 1st ed. facsimile (Cambridge, Massachusetts: Harvard University Press, 1964), p. 316. 43. A. d'Orbigny, Voyage dans l'Am6rique m&ridionale, p. 128. 44. Darwin, Origin of Species, p. 310. In the following sentence, he cites the vacillation of Lyell as a single point in his favor.

53

STEPHEN

JAY GOULD

Darwin rests the case, and indeed his entire theory,45on the proposition that geologists have studied carefully only a minute part of an extremely imperfect record. For my part, following out Lyell's metaphor, I look at the natural geological record, as a history of the world imperfectly kept, and written in a changing dialect; of this history we possess the last volume alone, relating only to two or three countries. Of this volume, only here and there a short chapter has been preserved; and of each page, only here and there a few lines.46 Although Darwin prevailed primarily because he had provided a new rubric under which a vast amount of common knowledge could be synthesized, new facts also helped to establish his evolutionary principles. Paleontological discoveries of the 1860's filled some gaps in the record and convinced many skeptics that the multitude of remaining discontinuities might be equally artificial. Although the discovery of Archaeopteryx surpassed all others in significance, the first indisputable Tertiary trigonians were also found in this decade. Trigonia subundulata, found by F. McCoy in Australia, was described by H. M. Jenkins47 in 1865. Jenkins saw this discovery as a typical example of events that would vindicate Darwin's views on the geologic record and establish the compatibility of gradualistic evolutionary notions with the data of paleontology: Every palaeontologist believes that, when a genus of animals is represented by species occurring in strata of widely different ages, it must have been perpetuated by some one or more species during the whole of the intervening period . . . The only rational meaning that has ever been attached to this presumed general law (for it is incapable of proof in many cases) is, that the perpetuation of the genus, species, family, &c., as the case may be, has been due to 'descent with modification.' The accident of the intermediate links being unknown in such cases, when everyone believes them to have 45. "He who rejects these views on the nature of the geological record, will rightly reject my whole theory." Darwin, Origin of Species, p. 342. 46. Ibid., pp. 310-311. a leader in Australian science, 47. Sir Frederick McCoy (1823-1899), was the founder and head of the Museum of Natural History and Geology in Melbourne and first Professor of Natural Science at the University of Melbourne. H. M. Jenkins (1841-1886), a minor figure in mid-nineteenthcentury British geology, edited the Geological Society of London's Quarterly Journal from 1862-1865 and spent the remainder of his life as Secretary to the Royal Agricultural Society.

54

Trigonia and the Origin of Species existed, supplies an excellent parallel to the frequent absence of the fossil remains of the 'connecting links' between groups of animals, especially as this absence has so frequently been a stumbling-block in the path of the students of Mr. Darwin's theory . Trigonia subundulata is one of the links hitherto wanting; first, in explanation of the existence of the genus Trigonia in the Australian seas of the present day; and secondly, as showing that the great gap which before existed in its lifehistory was no proof of the falsity of the postulate of palaeontology to which I have referred, but was simply a consequence of the imperfection of our knowledge of the geological record; for, though the record itself is imperfect enough, our knowledge of it is still more so.48

CONCLUSION While the Trigonia story is a microcosmic representation of nineteenth-century evolutionary debates, it also serves as a model for assessing the impact of new empirical material upon a controversial issue potentially explained by several intemally consistent but contradictory theories; for there can be no fact quite so pristine as a discovery anticipated by no one. The reaction to modem trigonians was, I suspect, completely typical; all parties to the dispute managed to incorporate the new datum into their systems. Evolutionists emphasized the morphological differences between Mesozoic and modern forms and assumed that the disjunct distribution was an artifact of an imperfect record. Agassiz cited the known distribution in support of special creation, but announced that the discovery of a Tertiary species would discredit none of his ideas. Parkinson could not readily encompass the difference without evident improvement in his progressionist synthesis, but invoked almighty wisdom in his ignorance. I do not doubt that all these naturalists proceeded properly in refusing to yield to the anomalies of a single fact which destroyed no deductive sequence in any of their theories. When one considers the stupendous amount of misinformation current in early nineteenth-century scientific circles,49 it is easy to 48. H. M. Jenkins, "On the Occurrence of a Tertiary Species of Trigonia in Australia," Quarterly Journal of Science, 2 (1865), 363-364. 49. The attempts by Linnaeus to include travelers' reports of semihuman troglodytes and tailed lucifers in his classifications, and the volumes of fiction that Darwin received from animal breeders, come to mind.

55

STEPHEN

JAY GOULD

appreciate the salutary aspects of stubbornnessin the face of inevitable contradictorycitations. De Beer5 has marveledat Darwin'sability,in the 1844 sketch, to work his way through a mire of misinformation: "It is a matter for wonder that with the meagre materials at his disposal he was able to steer a straight course across a largely uncharted ocean of ignorance, with rocks of falsehood right across his path." Yet Darwin approachedthese rocks with the idea of natural selection already firnly in mind. Any pure empiricist would have surrenderedto confusion long before 1859. 50. Gavin de Beer, Charles Darwin (London: Thomas Nelson and Sons, 1963), pp. 131-132.

56

Sherrington'sConceptof IntegrativeAction JUDITH P. SWAZEY Biomedical Sciences Group, University Program on Technology and Society, Harvard University, Cambridge, Massachusetts

In one of the most sustainedly productive careers in the annals of science, his writings covering a span of sixty-nine years, the English neurophysiologist Sir Charles Scott Sherrington (18591952) "almost singlehandedly crystallized the special field of neurophysiology."l Sherrington's classic investigations dealt primarily with reflex motor behavior in vertebrates, and with the nature of muscle management at the spinal level. The data, terms, and concepts which he introduced have become such a fundamental part of the neurosciences that it is perhaps not surprising their authorship is often forgotten. The neurophysiologist works daily with terms such as proprioceptive, nociceptive, recruitment, fractionation, occlusion, myotatic, neurone pool, motoneurone, and synapse, and with concepts such as the final common path, the motor unit, the neurone threshold, central excitatory and inhibitory states, proprioception, reciprocal innervation, and the integrative action of the nervous system. But seldom is he aware that these core contributions to his discipline were largely the work of one man, Charles Sherrington. The span of Sherrington's career and the scope of his empirical and conceptual contributions offer the historian a broad canvas for tracing the development of present-day knowledge about the physiology of the nervous system. The two major concerns of this study are, first, to analyze the development of the integrative action concept in Sherrington's work from 1884 to 1906 and, secondly, to consider the significance of the integrative action concept for the development of neurophysiology. These lines of inquiry have raised many of the types of questions to which biologists or historians of biology want answers. First, and perhaps most obviously, just what does the phrase "the integrative action of the nervous system" denote? 1. John F. Fulton, "Sir Charles Scott Sherrington, 0. M. (1857-1952)," J. Neurophys., 15 (1952), 168.

57

JUDITH P. SWAZEY

A preliminary answer may be given from the introductory pages of Sherrington's Silliman Lectures, delivered at Yale University in 1904 and published in 1906 as The Integrative Action of the Nervous System. "Integrative action," Sherrington wrote, is the action in virtue of which the nervous system unifies from separate organs an animal possessing solidarity, an individual . . . The due activity of the interconnecting function of the nervous system resolves itself into the coordination of the parts of the animal mechanism by reflex action.2 The foregoing definition, in turn, raises the question of what was new about Sherington's concept. An extensive body of experiments on reflex action was available when he began his researches, and Sherrington himself credited the noted French investigator Pierre Flourens (1794-1867) with "formally" introducing the idea of nervous coordination into physiology during the 1820's. Precursors may be unearthed for virtually every discovery or concept in the history of science, for each advance in knowledge usually is a mixture of the old and the new. In this light it is meaningful to ask when and where the new outweighed the old in Sherrington's work. How did he arrive at the integrative action concept, and why does his work, embodied in that concept, mark a watershed in the history of neurophysiology? In 1880 Charles Sherrington entered Cambridge's Gonville and Caius College and began to study physiology in Michael Foster's laboratory. By that date the centuries-old development of data and theories about the nervous system's structure and functions presented a piecemeal state of affairs. Controversy was rampant in almost every area of inquiry and, apart from some textbook presentations, few attempts had been made to correlate structural and functional data within a given field of study, much less to interrelate the various separate channels of nervous system researches. The study of reflex actions, for example, was prosecuted almost independently of concurrent work on problems such as the structure and interconnection of nerve cells, the differentiation of the spinal cord's sensory and motor functions, and the determination of brain structure and function. Investigators like the English physician Marshall Hall (1790-1857) had discovered a vast number of specific reflexes and had begun to define the internal and external variables affecting their 2. Charles Sherrington, The Integrative Action of the Nervous System, 2nd ed., paperbound (New Haven: Yale University Press, 1961), pp. 2, 5.

58

Sherrington's Concept of Integrative Action occurrence, such as the role of the stimulus. Experiments were in progress and theories being advanced, too, about some of the phenomena recognized as basic events in a reflex action, such as inhibition, excitation, and "Bahnung" (facilitation). A fairly extensive pool of techniques, data, and theories was available to a researcher in the last decades of the nineteenth century. But, as illustrated by the profusion of conflicting reports about the nature of such phenomena as the knee jerk,3 the whole field of reflexology was in need of restructuring. Techniques generally were imprecise, a sounder foundation of anatomical knowledge was needed, and, above all, there was a singular lack of experimentally based concepts with which to interpret the known facts of reflex action and evaluate their role in the animal economy. The resolution of this piecemeal state of affairs was largely effected by Charles Sherrington's work over a period of twentyfive years. "The mission of Sherrington's life," John Fulton aptly stated, "was the turning of anatomical facts of the nervous system into physiological language."4 Two important elements in the success of this "mission" were the nature of Sherrington's scientific training and the men who influenced him during its course. From his work in Michael Foster's Cambridge laboratory and in the European laboratories of Friedrich Goltz,5 Rudolph Virchow, and Robert Koch, Sherrington received a superb grounding in physiology, morphology, and microscopy and its allied sciences of histology, pathology, and bacteriology. The way in which he chose to apply his training and talents in these fields was due in large measure to the influence of John Newport Langley and Walter Gaskell, under whom he worked at Cambridge. Langley and Gaskell shared a dominant interest which they imparted to their student-how anatomical knowl3. For good summaries of the debate over the nature of the knee jerk at the end of the 1880's, see W. P. Lombard, "On the Nature of the Knee-jerk," J. Physiol., 10 (1889), 122-148, and A. D. Waller, "On the Physiological Mechanism of the Phenomenon Termed the Tendon Reflex," J. Physiol., 11 (1890), 384-395. 4. Fulton, "The Historical Contribution of Physiology to Neurophysiology," in Science, Medicine, and History. Essays on the Evolution of Scientific Thought and Medical Practice Written in Honour of Charles Singer, ed. E. A. Underwood (London: Oxford University Press, 1954), II, 543. 5. Sherrington's time in Goltz's laboratory was devoted largely to the study of secondary and tertiary nerve tract degenerations, work which deepened his interest in problems of the central nervous system's structure and function.

59

JUDITH P. SWAZEY

edge reflects, or is expressed in, physiological function. This facet of his career has been summarizedwell by RagnarGranit: Sherrington,the trained microscopistturned physiologist . . . was ready at the starting-linewhen the great era of creative cellular histology began and problems concerning the structure of the central nervous system rose to actuality . . . he

brought to the subject [of the physiology of the spinal cord] fresh insight into the necessity of knowing its cellular organization before further advance could be made. The nerve cell with its interconnectionsbecame his analytical unit. This was Sherrington'sfirst great notion and one of his main contributions to neurology.6 Sherrington decided to concentrate on neurophysiological researches rather than pursue his original interest in pathology, after he returned to England from Koch's Berlin laboratoryin 1887, and he became a lecturer in systematic physiology at London's St. Thomas's Hospital. He began his major contributions to knowledgeof vertebratemotor behaviorwhile serving as Physician-Superintendentof the Brown Institution, a London animal hospital, from 1891 to 1895. In a letter to Dr. Henry Head on November 18, 1918, Shenringtoncredited Walter Gaskell with directinghis attention from his first neurophysiological researches into brain-spinalcord connections to the physiology of the spinal cord. [Gaskell]was always an inspiration to me and to any work I was able to try . . . My own work began by chance at the

wrong end-the cortex-pyramidaldegenerations, etc. It was certainly through Gaskellthat I very soon felt that. One could not talk with him long without realizing that the cord offered a better point of attack physiologically.7 "A PRELIMINARYSTEP": ANATOMICALSTUDIES When Sherringtontumed his attention to the spinal cord the first object of his investigations was the knee jerk, a phenomenon which had first been described, independently and in the same joumal, by W. H. Erb and C. I. 0. Westphal in 1875.8 6. Ragnar Granit, Charles Scott Sherrington. An Appraisal (London: Thomas Nelson, 1966), pp. 28-29. 7. E. G. T. Liddell, "Charles Scott Sherrington, 1857-1952," Obituary Notices of Fellows of the Royal Society, 8 (1952), 244-245. 8. W. H. Erb, "Ueber Sehenreflexe bel Gesunde und Ruckenmarkskranken," Arch. Psychiat. Nervenkr. 5 (1875), 792. C. I. 0. Westphal, an gelihmten Gliedern," ibid., 803. "Ueber einige Bewegungserscheinungen

60

Sherrington'sConcept of Integrative Action From the beginning, opinions were sharply divided over the nature of the knee jerk. Erb and his supportersasserted that it was a "true tendon reflex" (involving neural conduction to the cord and back), while Westphal and his followers argued that the jerk was a direct mechanical twitch of the muscle (an idiomuscular contraction), with a state of tonus in the muscle a necessary condition for its occurrence. Sherringtonfirst directed his attention toward the basic question of the muscles and nerves upon which the jerk depends, a question which had not particularlyengaged the clinicians who were using the jerk as a diagnostic tool and studying such factors as its' variability. Working with the monkey, rabbit, cat, and dog, Sherringtonreportedthe results of his first analyses of the muscles and nerves controlling the jerk, and functional factors affecting its occurrence, in papers of 1891 and 1892.9 In these first studies, however, he found that he could not deal satisfactorily with functional problems in the face of a major gap in neuroanatomicalknowledge-the distributionof the spinal cord's sensory and motor fibers. At the commencement of some observations on the reflex mechanisms of the spinal cord in the Monkey, I was met by some difficulties which made it desirable to attempt for that animal a somewhat particularexamination of the distribution of the efferent and afferent spinal nerve-roots belonging to the lower half of the body ... For the study of the functions of the spinal cord it is of importance to know accurately the positions of the central and peripheral structures between which the fibres of the spinal nerves constitute links.10 At the outset of his investigations of spinal cord function Sherringtonthus found that he had to take a "preliminarystep." For a decade of what seemed to him often "boring"and "pedestrian" research he surveyed the whole field of distribution of each spinal root in order to create a sound anatomical foundation for physiological studies. He began his decade-long attack upon these anatomical problems with an examination of motor pathways, chiefly those in the lumbo-sacral plexus, publishing his findings in an 1892 "Note"of 150 pages.-1 Two years later, in 1894, he published a fundamental paper establishing the 9. "Note on the Knee-jerk," St. Thomas's Hospital Reports, 21 (1891), 145-147. "Note Toward the Localization of the Knee-jerk-Addendum," Brit. Med. J., 1 (1892), 545, 654. 10. Sherrington, "Notes on the Arrangement of Some Motor Fibres in the Lumbo-sacral Plexus," J. P4ysiol., 13 (1892), 621. 11. Ibid., pp. 621-772.

61

J UDITH P. SWAZEY

existence of sensory nerves in muscles, the second facet of his three major contributions to neuroanatomy.12His final step, "preliminaryto some observationson the functions of the spinal cord,"was detailed in two long papers of 1894 and 1898, which provided,respectively,maps of the cutaneous distributionof the thoracic and post-thoracic and of the cranio- and cervicobrachial posterior spinal roots.13 If one reads through the approximately700 pages of publications from 1884 to 1898 which bear Sherrington'sname one will appreciate why their author found his anatomical researches laborious and frequently dull. But a reader will appreciateeven more the quantitative and qualitativemagnitude of those necessary preliminarysteps that Sherringtontook in pursuing knowledge of the reflex functions of the spinal cord. To summarize the most important results of his anatomical studies, Sherrington found that 1) most muscles are multiply innervatedand can be stimulatedreflexlyfrom differentlevels of the spinal cord; 2) each afferent spinal root contributesto several nerve trunks and thus produces overlappingof skin fields; 3) each afferent spinal root has a continuous and self-contained segmental skin field; 4) the segmental sensory supply of skin may differ greatly from the motor supply of underlying muscle; 5) the "formationof functional collections of nerve-fibres(peripheral nerve trunks) out of morphological collections (nerve-roots) . . . is the expla-

nation-the meaning-of the existence of nerve plexuses";14 6) sensory fibers exist both in and from muscles, and muscles possess sense organs functioning independentlyof the overlying sldn.

Finally, in his 1897 Croonian Lecture on "Spinal Reflex Action," Sherrington presented the following short summary table of spinal organization.15

be likened to a The spinal apparatus of the limb may... funnel, the wide entrant mouth of which is represented by sensory nerves, the narrow end of exit by the spinal motor roots to the musculature. In the upper limb the sensori-motor funnel has the following segmental extension: 12. "On the Anatomical Constitution of the Nerves of Skeletal Muscles with Remarks on Recurrent Fibres in the Ventral Spinal Roots," J. Physiol., 17 (1894), 211-258. 13. "Experiments in Examination of the Peripheral Distribution of the Fibres of the Posterior Roots of Some Spinal Nerves (I)," Phil. Trans. Roy. Soc., 184B (1894), 641-763; "Experiments in Examination . .. (II), ibid., 190B (1898), 45-186. 14. Sherrington, Phil. Trans., 190B (1898), 152. 15. Ibid., 153.

62

Sherrington's Concept of Integrative Action

Number of spinal nervewhich evokes reflex movement

of limb

II.

III.

Cervical IV. V. VI. VII. VIII.

I.

Thoracic II. III. IV.

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

which supplies sense-endings in skin of limb which supplies motor innervation to muscle of limb

and in the lower limb the following: Number of spinal nervewhich evokes reflex movement of limb which supplies sense-endings in skin of limb which supplies motor innervation in muscle of limb

Thoracic XI. XII.

+

V.

++

Post-thoracic V. VI. VII. VIII. IX.

I

II.

II.

IV.

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

X.

+

Here, simply and clearly given, was the anatomical foundation upon which Sherrington built the physiological structure of his integrative action concept. REFLEX FUNCTIONS: PROBLEMS AND METHODS Concomitant with the anatomical researches which occupied most of Sherrington's time until 1898, and often deriving from them, came a profusion of ideas and observations on the reflex functions of the spinal cord. The two central, intertwined lines of these researches were the analyses of antagonistic muscle action and of larger "pieces" of reflex action such as the extension, flexion, and scratch reflexes of the hind limb. Out of these studies emerged Sherrington's conviction that the "main secret of nervous co-ordination . . . lies in the compounding of reflexes,"16 a compounding built up by the play of reflex arcs about their "common paths." Behind this play, he demonstrated in turn, lie the key processes of inhibitory and excitatory actions at the junctional regions between nerve cells-at the synapse. A fuller appreciation of the fusion of theory and experimental data in the integrative action concept may be gained by briefly examining the development of ideas about antagonistic muscle action, inhibitory processes, and the concept of the synapse during the nineteenth century, and, of equal importance, the types of methods which Sherrington used to study these phenomena in relation to the reflex functions of the spinal cord. 16. Sherrington, Integrative Action, p. 8.

63

JUDITH P. SWAZEY

From antiquity on, biologists have realized that for animal movement to be coordinatedand economical, opposing muscles must not work against each other. During the nineteenth century there was a vigorous and protracteddebate over the functional relations between antagonistic muscles. One school of thought, which held as Descartes had in the seventeenth century that muscle antagonism is an active process, received its first experimentalsupportfrom CharlesBell. In 1823, describing an experiment in which flexor contraction coincided with imposed relaxation of its antagonistic extensor, Bell wrote the followig striking anticipation of Sherrington's principle of reciprocal innervation: The nerves have been considered so generally as instruments for stimulating the muscles, without thought of their acting in the opposite capacity, that some additionalillustrationmay be necessary here. Through the nerves is established the connection between the muscles, not only that connection by which muscles combine to one effort, but also that relation between the classes of muscles by which the one relaxes and the other contracts.l7 A second widely held nineteenth-centuryview derived from the writings of the eighteenth-centuryanatomist J. B. Winslow, who suggested in his 1733 Anatomical Exposition that antagonistic muscles contract concurrently, the contraction of one muscle offering a "moderatingresistance" to that of the other. Winslow'sview received its first extensive laboratorytesting and supportin the researches of Beaunis (1885, 1889) and Demeny (1890). In the main, their work showed that under normal conditions there is a simultaneous contraction of antagonistic muscles during voluntary movement.18 A new chapter in the investigation of antagonistic muscle action was opened in 1897, when Charles Sherringtontold his audience at the Royal Society's Croonian Lecture: My own observations lead me to believe that inhibito-motor spinal reflexes occur quite habitually and concurrently with 17. Charles Bell, "On the Nerves of the Orbit," Phil. Trans., 113 (1823), 289. Sherrington appears to have been both excited and impressed when he discovered Bell's statement, and quoted it as a "remarkable passage" in an addendum to his second note on reciprocal innervation ("Further Experimental Note on the Correlation of Action of Antagonistic Muscles," Proc. Roy. Soc., 53 11893], 407420.) 18. For an excellent review of these and other studies of muscle antagonism, see F. Tilney and F. H. Pike, "Muscular Coordination Experimentally studied in Its Relation to the Cerebellum," Archiv. Neurol. Psychiat., 13 (1925),

64

289-334.

Sherrington'sConcept of Integrative Action many of the excito-motor 'reciprocalinnervation.'"9

.

.

This co-ordination I term

With these words he linked the frequently observed but little understood phenomenon of inhibition with the equally debated topic of antagonistic muscle action.20 Behind Sherrington'sfirst researches into the nature of inhibition as seen in antagonistic muscle action lay half a century of growing speculation and experimentationon the phenomena of peripheraland central inhibition, stemming from the Webers' classic demonstrationof cardiac arrest by vagal stimulation in 1845. The second half of the nineteenth century saw an increasing awareness of the modifiabilityof reflex action and of the unitary functions of the nervous system. Prior to the concept of integrative action, investigation of these topics had begun to crystallize around the study of central inhibition and facilitation of reflex responses. The possibflity that inhibitory phenomena in reflex actions have a central rather than a peripheral seat had been implicit from the time of RobertWhytt (1714-1766). He and numerous later workers had commented upon the ability of an individual to inhibit voluntarily certain reflexes such as sneezing, and upon the increased activity of the spinal cord after removal of the brain. Equally prevalent, however, was the idea that the inhibition seen in a process such as muscle antagonism is a peripheralevent-one occurringin the muscle itself rather than in the "nervecenters" of the brain or cord. The whole problem of reflex inhibition, as Sherringtonpointed out in 1913, moved to the central nervous system because investigators failed to find specific inhibitory peripheral nerves for vertebrateskeletal muscle. "As a working physiological thesis," he wrote in 1900, the notion of central inhibition "only became accepted doctrine after Setchenov."21In 1863 Johann Setchenov, a professorof physiology at MoscowUniversity,pub19. Sherrington, Phil. Trans., 190B (1898), 178. 20. Anticipating later discussion, we may note that Sherrington's analysis of reciprocal innervation clarified the nature of antagonistic muscle action primarily at the spinal level. In a normal, intact animal, with the "higher centers" operative, antagonistic muscles may be simultaneously Beaunis and Demeny showed. contracted during voluntary movement-as But even in this case the contractions occur in a reciprocal relation: i.e., if an extensor is contracted to 9/10 of its maximum, the opposing flexor may contract to only 1/10 of its maximum, thus maintaining steadiness of movement without antagonism. 21. Sherrington, "The Spinal Cord," Text Book of Physiology, ed. E. A. Schiifer (Edinburgh: Y. 3. Pentland, 1900), p. 838.

65

JUDITH P. SWAZEY

lished a monograph on inhibitory mechanisms controlling spinal reflexes in the frog.22 Setchenov ran three main series of experiments on various parts of the brain and brain stem, involving sectioning, chemical (salt crystal), and electrical stimulation, and physiological tests. The data from these experiments confirmed his belief that the brain possesses "centers" inhibiting spinal reflexes, and indicated that the centers are located in the optic thalamus, corpora quadrigemina, and medulla. Setchenov's experiments were cited widely as the demonstration of central inhibitory processes, but there were prompt and strident objections to his data and to the vast number of specific centers which his theory demanded. His controversial "center theory" thus led to the formation of numerous competing theories about the nature of and processes involved in central inhibition.23 It was against the background of these theories, and the emergence of the synapse concept, that Sherrington began to work out his ideas on the roles of central inhibitory and excitatory states in motor behavior, and on the reflex nature of inhibition itself. In the last decade of the nineteenth century the dominant view of nervous transmission was the reticular theory, championed by Joseph von Gerlach and Camillo Golgi, which held that nerve impulses are transmitted throughout the body over a continuous network, or "reticulum," of anastomosing nerve processes. At an 1889 meeting of the German Anatomical Society at Berlin the foundations of the reticular theory were sharply undermined by the report of an unknown Spanish neuroanatomist and histologist, Santiago Ram6n y Cajal (1852-1934).24 Cajal's theory, labeled the "neurone theory" by H. W. G. Waldeyer in 1891, derived from his histological studies of the embryonic bird and mammalian cerebellum. His preparations showed that although nerve cells often do make contact with other nerve fibers coming from many sources, definitely 22. Johann Setchenov, Physiologische Studien ueber die Hemmungsmechanismen fur die Reflexthatigheit des Ruckenmarks im Gehirne des Frosches (Berlin, 1863). 23. For discussions of late nineteenth- and early twentieth-century theories of inhibition, see Raymond Dodge, "Theories of Inhibition," Psychol. Rev. 33 (1926), 106-122, 167-187; Fulton, Muscular Contraction and Reflex Control of Movement (Baltimore: Williams and Wilkins, 1926), chaps. xiii-xiv; Franklin Fearing, Reflex Action. A Study in the History of Physiological Psychology (New York: Hafner, 1964), chap. xii. 24. Santiago Ram5n y Cajal, Recollections of My Life. Trans. E. Home Craigie. Edited in 2 vols. as Memoirs of the American Philosophical Society (Philadelphia: American Philosophical Society, 1937), chaps. v-vii.

66

Sherrington's Concept of Integrative Action lbmited conduction paths exist in the gray matter, and nerve impulses are transmitted by contact or contiguity, not by the reticularists' continuity. Although the reticularists clung tenaciously to their views, the neurone theory had a rapid impact upon ideas of the central nervous system's structural and functional architecture. E. A. Schafer, writing in 1893 on "The Nerve Cell Considered as the Basis of Neurology,"25 dismissed the reticular theory with the words, "It was formerly supposed that all nerve-cells were united with one another by distinct processes," and went on to present the new "general conclusions" of the neurone theory. As Schafer's article shows clearly, many details in the neurone theory were uncertain, especially concerning the junction between nerve cells: what are the exact anatomical boundaries between the cells, and how is their "physiological continuity" effected? Understanding of these relationships began to emerge in 1897 when Hans Held announced his discovery of "Endfusse" (end feet) and Sherrington introduced the term "synapse." From that time discussion of the "central machinery" controlling reflex functions began to be cast in the clearer terms of the neurone theory, of events at the synapse, rather than in the shadowy terms of "nerve centers" and "barriers of resistance." Hans Held's discovery of the terminal branches of nerve processes ironically served to strengthen his adherence to the reticular theory. As reported in his 1897 paper, he observed axonal nerve endings in adult nerve tissue broken up into "nleurosomes," coarsely vacuolated and granulated protoplasm, which appeared to be anatomically continuous with the dendrites or the body of an adjacent nerve cell.26 Terming the axonal endings Endfusse, Held hypothesized that they were the zones for transfer of stimuli between nerve cells. But, he emphasized, this transfer takes place across a network of fine protoplasmic "concrescences," not by "contact." In the same year that Held announced his discovery of Endfusse within the "reticular network," Charles Shernington, writing in Michael Foster's Textbook of Physiology, stated: So far as our present knowledge goes we are led to think that the tip of a twig of the [axon's] arborescence is not continuous with but merely in contact with the substance of the dendrite or cell body on which it impinges. Such a specialized 25. Brain, 16 (1893), 134-169. 26. Hans Held, "Beitrage zur Struktur der Nervenzellen und ihren Forsatze," Arch. Anat. Physiol. Wiss. Med. Leipzig (1897), 204; Suppl. (1897), 273.

67

JUDITH P. SWAZEY

connection of one nerve-cell with another might be called a synapsis.27

The circumstances behind Sherrington'sintroductionof "synapsis" were related by him in a letter to John Fulton on December 25, 1937: M. Foster had asked me to get on with the Nervous System part . . . of a new edition of his "Textb.of Physiol."for him. I had begun it, and had not got far with it before I felt the need of some name to call the junction between nerve-celland nerve-cell (because that place of junction now entered physiology as carrying functional importance.) I wrote him of my difficulty, and my wish to introduce a specific name. I suggested using syndesm . . . He consulted his Trinity friend

Verrall, the Euripidean scholar, about it, and Verrall suggested 'synapse', . . . and as that yields a better adjectival

form, it was adopted for the book. The concept at root of the need for a specific term was that, as was becoming clear, 'conduction'which transmitted the 'impulse' along the nerve-fibrecould not-as such-obtain at the junction, [because] a "membrane"there lay across the path, and "conduction"per se was not competent to negotiate a 'cross-wise"membrane.28 The significance of Sherrington'schoosing to work with the neurone rather than the reticular theory and his coining of "synapse"has been clearly stated by Granit: When Sherrngton decided in favor of nerve-cell contacts he refashioned thinking in this field along lines that determined its future course for all time and also tied it to the newly born science of electrophysiology . . . Only a contact theory could

bridge the gap between reflex transmission and electrophysiology; such is the power of a fundamental concept like the synapse.29

Ideas about synaptic properties and their relation to reflex phenomena were explored by Sherrington, in detail and upon an experimentalbasis, in the Silliman Lectures.The "characters distinguishing reflex-arc conduction from nerve-trunkconduc27. Sherrington, "The Central Nervous System," in Sir Michael Foster's A Textbook of Physiology, 7th ed. (London: Macmillan, 1897), III, 929. 28. John F. Fulton Papers. Yale Historical-Medical Library, New Haven, Connecticut. 29. Charles Sherrington, p. 43.

68

Sherrington'sConcept of Integrative Action tion," he demonstrated, "may be largely due to intercellular barriers,delicate transversemembranes, in the former."30 The barrageof claims and counterclaimsregarding the reflex functions of the brain and spinal cord during the nineteenth century clearly derived, in large measure, from an inadequate base of anatomical knowledge combined with often imprecise and haphazard physiological methods. Sherrington, like other investigators of the anatomy, histology, and physiology of the nervous system, was thereforeconfrontedby the task of devising techniques for reducing and controlling the nervous system's structural and functional complexity to the point where meaningful data could be obtained. His first steps were 1) to concentrate on the reflex functions of the spinal cord rather than on the more complex field of the brain; 2) to choose an appropriateexperimental animal, the monkey, and run parallel experiments on a variety of lower forms for controls and comparisons; 3) to establish the necessary points of anatomicalknowledgeon which to rest functional studies. A complex of additional factors further combined to make Sherrington's career so manifestly successful. Many of his techniques, skills, and work habits were developed and sharpened during the course of his anatomical studies: operative procedures, means of stimulus standardization, ways to exclude extraneous variables from experiments, methods of recording data, and the thorough testing of any hypothesis, however labonrousand monotonous the labor involved. In contrast to today's highly refined microtechniques and physicochemical ways of analyzing biological phenomena, Sherrington's basic method was to study simple motor acts which could be made to occur in isolation, such as antagonistic muscle action and, on a more complex level, the simple and compound reflexes of the hind limb. An integral part of this method was to correlate the exacting analyses of the inputoutput relations of reflex responses with anatomical and histological data. To study simple acts of muscle management at the spinal level in the living animal, Sherrington eliminated the higher controllingregions of the brain and brain stem. He reduced his experimental animal to a functioning, isolated spinal cord or cord segment, able to execute limb reflexes in response to given stimuli. Approachingstill closer to the experimentalideal of one variable at a time, he could further simplify these simple reflexes by isolating single muscles and their nerve pathways. 30. Integrative Action, p. 17.

69

J UDITH P. SWAZEY

Analyzing the simple reflex contractions of muscles in vivo and then comparing them to such phenomena as the twitch of an isolated nerve-muscle preparation and conduction in a peripheral nerve trunk, Sherrington gradually unveiled the characteristic properties of reflex pathways through the spinal cord. By studying the interplay of these pathways in successively larger and more complex reflexes, he slowly built up a picture of the pattern of integrated motor behavior in an intact, norimal animal. Sherrington primarily used two types of experimental animals: the classic "spinal animal" whose cord has been permanently transected above the lumbo-sacral enlargement, and the "decerebrate animal." Decerebration, by actual removal of the cerebrum or by transection of the brain stem, was an old operation and its effects had been described, in part, by many earlier workers, including Magendie, Longet, Flourens, Bernard, Fontana, and LeGallois. To Sherrington, however, must go the credit for having named and established "decerebrate rigidity" both as a phenomenon in its own right and as a major tool for examining the reflex functions of the spinal cord, particularly the nature of inhibition and reciprocal innervation. In 1868 H. Sanders-Ezn had used a decerebrate frog with the cord cut in the lower thoracic region to study the role of motor roots in reflex responses. Following Sanders-Ezn's method, Sherrington's successful use of a decerebrate mammal with the cord transected and anaesthesia then remitted provided him with an ideal experimental preparation-an isolated length of spinal cord, with a good blood supply free of anaesthesia, in which he could analyze reflex paths at will. Sherrington probably used a decerebrate monkey to study spinal reflexes as early as November 1894,31 and during the summer and fall of 1896 he began to explore more fully the effects of decerebration. The first fruits of his study appeared in January 1897, as a description of "Cataleptoid Reflexes in the Monkey."32 These reflexes, he wrote, were of "extremely prolonged duration, and absolutely devoid of clonic character and of alternating character," quite in contrast to the reflexes usually elicited from the isolated cord. "Not their least interesting part," he promptly realized, "is a remarkable glimpse which they allow into the scope of reflex inhibition as regards the coordinate movement of the limbs."33 Moving toward his classic definition of decerebrate rigidity, 31. dress, 32. 33.

70

Figure 1, Plate 17 of Sherrington's 1897 Marshall Hall Prize Adshowing a decerebrate monkey, is dated November 18, 1894. Proc. Roy. Soc., 60 (1897), 411414. Ibid., 413.

Sherrington's Concept of Integrative Action Sherrington described the effects of transecting the neural axis at the level of the crura cerebri in a second paper of January 1897. He used the terms "decerebrate rigidity" and "decerebrate tonus" to describe the decerebrate animal's marked extensor rigidity, in his Croonian Lecture in April 1897,34 prior to publishing his fundamental paper on the subject in 1898.35 In subsequent studies by Sherrington and by other workers the decerebrate animal served as a major element, both conceptually and technically, for deciphering the reflex functions of the spinal cord and the levels of control operative in the central nervous system. RECIPROCAL INNERVATION The most important theme in Sherrington's functional researches up to 1900, both for his understanding of the operations of spinal reflexes per se and for his comprehension of the mechanisms of nervous coordination, was his analysis of the reciprocal innervation of antagonistic muscles. For it was the principle of reciprocal innervation, as Lord Adrian has commented, "which opened the way to the further advance from the simple to the complex. It was the clue to the whole system of traffic control in the spinal cord and throughout the central pathways."36 The results of Sherrington's exhaustive study of reciprocal innervation are found chiefly in his fourteen classic "Notes" in the Proceedings of the Royal Society from 1893 to 1909. The first Note-"On the Correlation of Action of Antagonistic Muscles"37-stemmed from observations he had made on the knee jerk while studying the motor filaments in the lumbo-sacral plexus. The work reported in this Note provided the first sound experimental support of Erb's thesis that the knee jerk is a reflex phenomenon. Sherrington showed, in essence, that the jerk can be inhibited centrally, as can other spinal reflexes. The debate over the knee jerk, however, was far from settled, for supporters of the "idiomuscular theory" continued to raise issues such as reflex time and to argue that Sherrmngton had not directly inhibited the jerk but rather the mysterious "tonus"'of the muscles on which the jerk is dependent. Each of Sherrington's experiments on a specific problem 34. Phil. Trans., 190B (1898), 159, 161. 35. "Decerebrate Rigidity and Reflex Co-ordination of Movements," J. Physiol., 22 (1898), 319-332. 36. Edgar D. Adrian, "The Analysis of the Nervous System: Sherrington Memorial Lecture," Proc. Roy. Soc. Med., 50 (1957), 993. 37. Proc. Roy. Soc., 52 (1893), 556-564.

71

JUDITH P. SWAZEY

tended to produce a host of other issues which engaged his attention-a fact which accounts for the profusion of data and ideas crowding the pages of his papers. The manner in which his knee jerk studies led him into an analysis of antagonistic muscle action typifies this aspect of his work. Data in the first Note suggested that antagonistic muscle action-and the inan active and central phehibitory processes it involves-is nomenon. For in the case of the knee jerk, at least, he had found that excitation "of the afferent fibres from one set of the antagonistic muscles induces reflex tonic contraction of the opposing set with extreme facility, despite the fact that the opposing muscles are not innervated from the same spinal segment."88 Sherrington's second Note39 continued to examine the "qualities of alteration restraining or abolishing the jerk," adding evidence for its being a reflex mechanism dependent on a central spinal mechanusm. And, his work on the knee jerk muscles having suggested a "search for instances of analogous correlation elsewhere," he reported the results of his first experiments on antagonistic muscle action in the eye muscles and palpebral (eyelid) apparatus. Apart from an 1894 paper dealing further with eye muscle antagonism, the bulk of Sherrington's publications from 1894 to 1896 dealt with his anatomical researches. He returned to the analysis of antagonistic muscle action in 1897, the most productive year of his life in terms of number of publications sixteen original papers. The term "reciprocal innervation" was first used in the title of Sherrington's third Note, read before the Royal Society on January 21, 1897.40 The term, he explained in the Note, denoted the "particular form of correlation" in which one muscle of an antagonistic couple is relaxed as its mechanical opponent actively contracts. Four months later, as the Royal Society's Croonian Lecturer, he proposed his classic definition of reciprocal innervation as that form of coordination in which "inhibitomotor spinal reflexes occur quite habitually and concurrently with many of the excito-motor."" From 1897 to 1900, as reported in "Notes" three through six, and in other papers, Sherrington continued to explore the 38. Ibid., 563. It was this finding that led Sherrington to his study of sensory nerves in muscles. 39. "Further Experimental Note on the Correlation of Action of Antagonistic Muscles," Proc. Roy. Soc., 53 (1893), 407420. 40. "On Reciprocal Innervation of Antagonistic Muscles. Third Note," Proc. Roy. Soc., 60 (1897), 414-417. 41. Phil. Trans., 190B (1898), 178.

72

Sherrington's Concept of Integrative Action nature of reciprocal innervation and to extend demonstrations of the range of its occurrence. By the turn of the century he had firnly established reciprocal innervation as a widely occurring phenomenon, showing that it may be produced by excitation of skin and peripheral nerves, muscles and their afferent nerves, the dorsal spinal column, the anterior surface of the cerebellum, and the crusta cerebri, pyramidal tract, internal capsule, optic radiations, Rolandic cortex, and occipital cortex. Conceming the mode of operation of reciprocal innervation, Sherrington's first studies had shown that the process must be viewed as an active and central one. As for the levels of nervous system control, he demonstrated, it is not essential that the "high level" centers be active for the type of "elementary coordination" effected by reciprocal innervation. His finding that reciprocal innervation can operate by a simple reflex mechanism at the spinal level derived from his study confirming the existence of sensory nerves in muscles. Applying his knowledge that a reflex may be elicited by a purely muscular reaction, he showed in 1899 that electrical stimulation of the central cut end of an exclusively muscular nerve (the hamstring's) inhibits the tonus of its antagonist (the knee extensor muscles). Moreover, he found, the same effect is produced when the flexors are exposed and detached from the knee, making them mechanically incapable of affecting the position of the joint, and then are stretched or kneaded. Such data, he reasoned, show that reciprocal innervation may be secured by a simple reflex mechanism, an important factor in its execution being the tendency for the action of a muscle to produce its own inhibition reflexly by mechanical stimulation of the sensory apparatus In its antagonist.42 By 1900, in summary, Sherrington's researches had revealed a great deal about the nature of antagonistic muscle action. He had found it to be a widely occurring phenomenon, active rather than passive in its operation, brought into play through a central rather than a peripheral mechanism, dependent on the integrity of afferent nerves from muscles, and at the spinal level involving the reciprocal processes of inhibition and excitation. Of fundamental importance to his understanding of the principles of muscle management, he had shown reciprocal innervation to be a basic element in motor coordination. "Reflexes obtained from the decerebrate animal," he wrote in 1898, "exhibit . . . 'reciprocal innervation' . . . in such distribution and sequence as 42. "Inhibition of the Tonus of a Voluntary Muscle by Excitation of Its Antagonist." J. Physiol., 23 (1899), 26.

73

JUDITH P. SWAZEY

to couple diagonal limbs in harmonious movements of similar direction."43

In view of the ways in which Sherrington'sdefinition of the mode and role of antagonistic muscle action reshaped understanding of the nervous system's coordinative activity, the following words from his Croonian Lecture in retrospect must rank as a classic understatement in the annals of neurophysiology: In short, my observationsprove the existence of "reciprocal innervation"of antagonistic muscles as part of the machinery of spinal reflexes, and point to it as possibly a widely extensive part of that machinery.44 With reciprocalinnervation as the key phenomenonrevealing patterns of "trafficcontrol"in the central nervous system, Sherrington by 1900 had begun to formulate a comprehensivepicture of the motor functions of the spinal cord. His conception of these functions, the "rules"which govern them, their mechanisms of control, and their role in the unitary functioning of the nervous system, were developedbetween 1897 and 1900 in his Croomianand Marshall Hall Lectures and in the pages of E. A. Schafer's Text Book of Physiology.

'The individual,"Sherringtontold his Marshall Hall Lecture audience on May 23, 1899, "is a mass of living units, their activity co-ordinated together by conductive strands (nerve cells) reacting to the environment."45When one turns to the physiological analysis of the nerve cells in the spinal system, he affirmedin 1900, "the unit reached is the 'reflex.'Upon it as a basis our existing notion of the reaction of the nervous system is built."46 For analyzing the functions and mechanisms of spinal reflexes Sherringtonhad at hand the myographic records of his spinal-root mapping work with the monkey, Macacus. These records, he emphasized in appraising their usefulness, possess points of interest, but it must not be forgotten that the value of a reflex obtained by exciting the afferent spinal roots... is but slight as regards the light thrown by it on the normal workingsof the cord.47 Because, as he had shown, the spinal root is a morphological 43. 44. 45. 46. 47.

74

J. Physiol., 22 (1898), 332. Phil. Trans., 190B (1898), 180. "On the Spinal Animal," Med.-Chirug. Transactions, 82 (1899), `The Spinal Cord," Text Book of Physiology, ed. Schafer, p. 784. Phil. Trans., 190B (1898), 133.

462.

Sherrington's Concept of Integrative Action rather than a functional unit, Sherrington realized that an analysis of spinal reflex action must collate several sets of data: the morphology of the spinal roots, and the reflexes obtained from the roots, peripheral nerves, and skin-spots. By pooling and analyzing these four sets of data he formulated a series of "rules" regarding the irradiation or spread of impulses seen in two broad classes of spinal reflex action: short spinal reflexes, in which the muscular response occurs in the same region as the stimulus, and long spinal reflexes, in which a stimulus applied in one region evokes a response in another.48 Sherrington's "rules of spinal reflex action" were forged from his knowledge of sensory and motor pathways and of antagonistic muscle action. They were rules, he stated repeatedly, which did much to explain the fact of coordinated muscle management at the spinal level. With this perception, as summarized in one brief sentence from his Croonian Lecture, we see him moving steadily toward the concept of integrative action: "In this way reflex action, by its 'spread' develops a combined movement, synthesizes a harmony."49 Prior to Sherrington's work, four interacting factors had been isolated as the chief determinants of purposeful, coordinated spinal reflex movements: the character of the afferent impulse, stimulus intensity, stimulus locus, and the intrinsic condition of the spinal cord. Sherrington subsequently showed that, at a gross function level, reflex movements are effected by the reciprocal innervation of antagonistic muscles. By 1900 he had pinpointed three key mechanisms which in turn effect and affect the operation of reciprocal innervation: the muscular sense and the processes of central inhibition and facilitation. The fact that organism can function without the guidance of its "head pole" indicated to Sherrington, first, "that some set of spinal sense organs forms the chief basis of the mechanism ensuring such elemental co-ordination."50 On the basis of his anatomical and physiological experiments, he designated the sense organs in the "musculo-articular apparatus" as the basic mechanism of muscle management by the cord or cord segment.51 Although guided by information from these sense organs the isolated cord needs additional means to prevent a discord 48. Sherrington's "rules" are presented in Schafer's Text Book, pp. 819-845, and in his Croonian Lecture, Phil. Trans., 190B (1898), 145-177. 49. Phil. Trans., 190B (1898), 162. 50. Sherrington, "The Spinal Cord," Text Book of Physiology, ed. Schafer, p. 844. 51. Ibid.

75

JUDITH P. SWAZEY

of movement. These means, Sherringtonsaw, were providedby the fact that "reflex arcs, both higher and lower, are interconnected. The condition of one is partly dependent on the condition of many others."52 As many investigatorshad pointed out by 1900, the effect of this interdependencemay take two forms: the depression of local reflex reactions-Hemmung, or inhibition; or the promotion of a local reflex reaction-Bahnung, or facilitation. The causes of central inhibitory and facilitory processes, as we have noted, were a subject of controversyat the turn of the century. Shenington was less concerned, however, with their basic causes than with their roles in motor coordination. His studies of irradiation had shown clearly that facilitation plays an important role in harmonizing the spread of impulses throughoutthe spinal cord. Similarly, in view of the widespread occurrence of antagonistic muscle action in the body-action which involves the reciprocalprocesses of inhibition and excitation-he judged that 'it is probablethat inhibition of one spinal 'centre' by another plays a great part in the elementary coordination of actions executed under spinal mechanisms."3 In retrospect the historian can see that by 1900 Sherrington had assembled the major ingredients of the integrative action concept. Proceeding from his knowledge of sensory and motor paths and the existence of the musculo-articularsense organs and nerves he had applied his basic, exacting "input-output" analysis to a host of reflex phenomena. From a study of the seemingly simple anatomy and physiology of the knee jerk he had become engaged with a series of broaderproblems, such as the nature and mechanisms of antagonistic muscle action, the production and maintenance of decerebrate rigidity, and the nature and significance of spinal shock. Out of the mountains of data collected in the course of these researches, Sherringtondeveloped a number of basic functional principles: reciprocal innervation, interaction between higherand lower-level centers of motor control, and the muscular sense, inhibition, and facilitation as three key mechanisms of muscle management at the spinal level. And, recognizing the importof the neurone theory for his work, he had perceivedthat many of the characteristicpropertiesof reflex pathways might, at root, be explicable by the events at the synapse. By the time Sherringtonsat down to write on the spinal cord in Schifer's Text Book these facts and principles clearly had begun to coalesce in his mind. From his discussion of the 52. Ibid., p. 837. 53. Ibid., pp. 840-841.

76

Sherrington's Concept of Integrative Action "influence of associated parts on the central organ" flashes the basic perception of integrative action: to say that "the nervous system is a unity" is to say that each even local reflex action is in truth a reaction to all sensifacient stimuli incident on the individual at the moment in sum both as to space and as to time. And those activities of the central nervous system which are customarily spoken of as autochthonous, have also, on account of the inter-connection of the arc, influence upon the reaction of other arcs. The influences arising either way may take expression in . . . inhibition [or] facilitation.54 Taking the final steps from this perception to the full-grown integrative action concept, Sherrington determined how interdependent reflex arcs combine to form successively larger and more complex reflex pattems, such as those of the scratch reflex. THE SCRATCH REFLEX PARADIGM Sherrington's analyses of the scratch and other hind limb reflexes, engaging many more muscles than does a relatively simple reflex such as the knee jerk, confirmed his earlier observations and theses: the same functional principles, he found, obtain in both the simple and more complex reflex actions. And, because of its very complexity, the scratch reflex further illuminated a wide range of central nervous system phenomena underlying motor coordination, such as inhibition, facilitation, spinal induction, and the events at the synapse. Reflex scratching must have been recognized for centuries, particularly by any dog owner, but it had received scant attention in experimental literature until Sherrington tumed to it. He began to study the scratch reflex while extending the analysis of short and long reflex paths first made in his 1897 Croonian Lecture. In a 1903 paper he and E. E. Laslett defined the skin field in the spinal dog from which reflex scratching can be elicited, the rhythm and other properties of the reflex's apparatus, and its probable trineuronic intraspinal pathway.55 Sherrington continued his definition of the reflex's effective stimuli, the mechanism of its motor response, and factors affecting its occurrence, in another paper of 1903 and one in 1904.16 54. Ibid., p. 837. 55. Sherrington and E. E. Laslett, "Observations on Some Spinal Reflexes and the Inter-connection of Spinal Segments," J. Physiol., 29 (1903), 58-96. 56. "Qualitative Difference of Spinal Reflex Corresponding with Qualitative Difference of Cutaneous Stimulus," J. Physiol., 30 (1903), 3946; "On Certain Spinal Reflexes in the Dog," J. Physiol., 31 (1904), xvii-xix.

77

JUDITH

P.

SWAZEY

By 1904, from investigations of reflex actions ranging from the knee jerk to more complex patterns, such as reflex scratching, Sherrington had forged the detailed and comprehensive picture of reflex motor behavior which he presented as Yale's Silliman Lecturer in April 1904. In the same year he enunciated the nucleus of the integrative action concept in a presidential address to the Physiological Section of the British Association for the Advancement of Science. Published in the British Association Reports on August 18, 1904, "The Correlation of Reflexes and the Principle of the Common Path" stands as Sherrington's most important published conceptual statement before The Integrative Action of the Nervous System (1906). The physiology of the nervous system, Sherrington began his address, may be studied from three main viewpoints: 1) the life of the individual nerve cell; 2) the specific functional and 3) the integrative property of nerve cells-conductivity; by nerve cell conducfunction of the nervous system-how, tivity, "the separate units of an animal body are welded into a single whole, and from a mere collection of organs there is constructed an individual animal."57 The main theme of Sherrington's address was an analysis of the reflex chain of the synaptic system, in which he introduced his "principle of the common path." He pointed out first that a reflex chain's receptive neurone forms a "<privatepath" into the brain or spinal cord, a path which is "exclusive for a single receptive point." WVhenthe receptive neurone passes into the "great central organ" it forms numerous connections with a vast network of conductive paths. Within this complex, locally fluctuating conductive network, Sherrington explained next, many private paths converge at an internuncial neurone to form a second type of reflex pathway, a public or common path. Internunical common paths, in turn, run to the third link in the reflex chain, the motor neurone. From the motor neurone, finally, the impulses travel over a final common path to converge upon the effector organ. Sherrington presented the results of his work on the scratch reflex to the members of the British Association as an excellent example of how the paths of a reflex chain interact to effect motor coordination. By his analysis of the scratch reflex he had mapped a fairly complex reflex action, explaining its occurrence in terms of the play of competing reflex arcs about a common path. The seat of this competition between rival flexor and extensor arcs, he surmised, seems to lie not in the motor 57. British Association Reports, 74 (1904), 2. Sherrington also began his Silliman Lectures by delineating these three lines of investigation.

78

Sherrington's Concept of Integrative Action nerve but in the gray matter, where the arcs impinge together on the motor neurone. That is equivalent to saying that the essential seat of the phenomenon is the synapse [between] the motor neurone and the axone-terminals of the penultimate neurones that converge upon it. There some of these arcs drive the final path into one kind of action, others drive it into a different kind of action, and others again preclude it from being activated by the rest.58 Summarizing his conception of the common path and its role in the integrative functions of the nervous system, Sherrington likened the central organ (brain and cord) to a telephone exchange, where connections between starting and end points are changed from moment to moment to suit passing requirements. To realize the "exchange" at work, he suggested, one must see it as a combination of spatial plan and temporal data, whose incoming paths offer connections with varying degrees of resistance. It is temporal variability, Sherrington realized, which provides the synaptic system with a mechanism for higher integration, synthesizing an individual animal from a "mere collection of tissues and organs." "Expressed teleologically," said Sherrington, the common path is a coordinating device which although economically subservient for various purposes, is yet used for only one purpose at a time . . . In this way the motor paths at any moment accord in a united pattern for harmonious synergy, co-operating for one effect.59 At the spinal level, as seen in reciprocal innervation, this coordination covers merely a limb or a pair of limbs. But the same mechanism of the common path extended to the play of the great arcs arising from the projicient organs of the head engages the conductive patterns of the total organism. "The singleness of action from moment to moment thus assured" by the principle of the common path, Sherrington declared, is a keystone in the construction of the individual whose unity it is the specific office of the nervous system to perfect . . . Correlation of the activities of arcs from receptive points widely apart is the crowning contribution of the brain toward the nervous integration of the individual.60 58. Ibid., p. 7. 59. Ibid., p. 12. 60. Ibid., pp. 12, 10.

79

JUDrrH

P. SWAZEY

Prior to 1900, as we have noted, Sherrngton had focused on facilitation, inhibition, and the sense organs and nerves in the musculo-articular apparatus as three basic mechanisms underlying the "elemental coordinations" effected by reciprocal innervation. In four papers of 1905-1906 he further examined these mechanisms in the more complex patterns of hind limb reflexes, with the new insights gained from his concept of "the correlation of reflexes and the principle of the common path." This stage in his analysis of the mechanisms controlling reflex actions focused on spinal induction, inhibition, and the proprioceptive system. From the time of his observations of the knee jerk, Sherrington had catalogued numerous cases of a temporarily depressed spinal reflex entering into a state of "exaltation" manifested by a more active, intense motor discharge. In his seventh and ninth Notes61 he analyzed instances of this intraspinal rebound effect supervening on a state of spinal inhibition in hind limb reflexes. He recognized the rebound effect as a form of facilitation and felt that it was "fundamentally akin" to the phenomenon of visual contrast that Hering had termed "induction." On these bases he designated the phenomenon of reflex-rebound as "successive spinal induction." The word "successive" was used to differentiate the rebound type from a form which Sherrington had found to occur without previous inhibition, via a summation of subliminal stimuli mutually reinforcing reflex arcs from two receptive points. He had described this latter species of induction in his papers on the scratch reflex and the common path prior to naming it "immediate spinal induction" in 1906.62 In spinal induction Sherrington realized that he had found a process "obviously qualified to play a part in linking reflexes together in a co-ordinate sequence of successive combinations."83 Induction, he saw, is a form of facilitation especially fitted to combine the successive opposite phases of cyclic reflexes. As such, it was a phenomenon whose definition did much to explai the altemating discharge of antagonistic muscles which Sherrington had shown to be so characteristic of the mammalian spinal cord's locomotor activity. By 1893 Sherrington's study of antagonistic muscle action 61. "On Reciprocal Innervation of Antagonistic Muscles. Seventh Note," Proc. Roy. Soc., 76B (1905), 160-163; "On the Innervation of Antagonistic Muscles. Ninth Note. Successive Spinal Induction," Proc. Roy. Soc., 77B (1906), 478-497. 62. Integrative Action, p. 121. 63. "Seventh Note," Proc. Roy. Soc., 76B (1905), 162.

80

Sherrington'sConcept of Integrative Action had convinced him that inhibition is part of the normal reflex process, actively going on side by side with excitation. In his eighth Note (1905) he retumed to this thesis, examining "in some particulars the conditions attaching to the initiation, and the course run by [inhibition and excitation] under comparable circumstances."64To define the conditions, he made a detailed study of the activities of all muscles involved in rm)ovingthe knee joint, carefully analyzing the reciprocal activities of each of the antagonists. On the basis of data reportedin his eighth Note and earlier studies, Sherrington formulated his own "inference"about inhibitory and excitatoryprocesses in spinal reflexes. He preceded his hypothesis with the cautionary statement that in hind limb reflexes, as elsewhere, "the intimate nature of the process which reveals itself as inhibition is admittedly obscure,"65although many theories have been advanced. In the spinal cord, he went on to propose, individual afferent fibers from the receptive field of a reflex each divides into end-branches (collaterals). When the nerve fiber is active, one set of these collaterals normally produces excitation in the efferent neurones of muscle A, while another set always produces inhibition in the efferent neurones of A's antagonist, muscle B. In brief, he suggested, the single afferent nerve-fibrewould therefore be in regard to one set of its central terminal branches specifically excitor, and in regard to another set of its central endings specifically inhibitory. It will in this respect be duplex centrally.66

From the time he had demonstratedconclusively that muscles possess sensory organs and nerves, Sherrington had shown their importance for the production and maintenance of a host of reflex phenomena. In the 1906 issue of Brain dedicated to John Hughlings Jackson he generalized these findings in his concept of the proprioceptivesystem. At the base of his concept were certain inferences about the scope of reflexes, their relations to one another and to the broader functions of the central nervous system, drawn from the distribution of nerve receptors.The receptorsof the body, he pointed out, fall into two great groups: the surface field, subdividedinto the exteroceptive and interoceptive,and the deep field. The deep field presents many points of contrast to the surface 64. "On Reciprocal Innervation PTOc.Roy. Soc., 76B (1905), 269. 65. Ibid., p. 283. 66. Ibid., p. 287.

of Antagonistic

Muscles. Eighth Note,"

81

J UDITH P. SWAZEY

field. Most notably, Sherrington felt, the stimuli effective for the two fields are fundamentally different. The deep receptors, he stated, appear adapted to "excitation by changes going forward in the organism itself" -changes working largely through the agency of mass, with its mechanical consequences of weight and inertia, and through the mechanical strains and alterations of pressure from muscular contraction and relaxation. On the bases of their mode of excitation, Sherrington proposed that "the deep receptors may be termed proprio-ceptors,and the deep field a field of proprioception."87

After defining proprioception Sherrington considered some of its characteristics and functions. He felt that perhaps the most importantattributeof the proprioceptorsis "theirtendency to induce and maintain tonic reactions in the skeletal musculature."68Fundamentally,he held, "reflextonus is the expression of a neural discharge concerned with the maintenance of attitude"-the maintenance, that is, of posture rather than movement.89The production and maintenance of posture as well as of locomotion, Sherringtonrealized, involves an integration or unificaidonof reflex actions: Reflex arcs of analogous function belonging to successive segments unite to a homogeneous reflex system extending more or less continuously through the length of the animal. Thus it is that the proprio-ceptors and their reflex arcs have in their sum total, to be treated as a proprio-ceptive SYSTEM.70

Had Sherringtonnever delivered and published the Silliman Lectures, the corpus of his research papers, and such general addresses as "The Spinal Animal,""The Principle of the Common Path," and "The Proprio-ceptiveSystem,"still should have marked him as biology'sforemost analyst and interpreterof reflex motor behavior. These abilities are nowhere more evident than in his definitive paper on the scratch reflex (1906),71 in which the propertiesof reflexes as found in the isolated spinal 67. "On the Proprio-ceptive System, Especially in its Reflex Aspects," Brain, 29 (1906), 472. 68. Ibid., p. 473. 69. Ibid., p. 474. Sherrington's view that tone is a postural reflex was generally accepted during his lifetime. "Today," Granit observes, "while one would still be willing to give the postural reflex a rBle to play in tone, it seems impossible to identify the two concepts" (Charles Sherrington, p. 155). 70. Brain, 29 (1906), 475. 71. "Observations on the Scratch-reflex in the Spinal Dog," J. Physiol., 34 (1906), 1-50.

82

Sherrington's Concept of Integrative Action cord were described minutely and fully as never before in the annals of neurophysiology. Some of the data already had been reported in earlier papers before Sherrington brought them together, supplemented them, and gave them physiological meaning in 1906. His intensive study of reflex scratching enabled him to isolate and describe almost every physiological characteristic of reflex muscle management. He saw, for example, that the steady frequency of the reflex means that so many of the stimuli in a continuous series fall in a period when the reflex is refractory to stimuli, and that the reflex has its own characteristically frequent after-discharge. The fact that single stimuli fail to elicit scratching told him of the importance of summation of stimuli in reflex response. The nature of immediate spinal induction in reflex scratching suggested to him that facilitation is a special form of summation. The paths of the scratch reflex, he affirmed, are truly intraspinal. They begin and end in the cord, with no essential anatomical processes going up to higher centers. Combining his anatomical and physiological data, he attempted to assign various characteristics of reflex scratching to specific places in its reflex chain: the seat of local fatigue "probably" is in the first synapse of the disynaptic arc; refractory phase is referable either to the descending propriospinal neurone internuncial between the receptive path and the final common path, or to the synapse between the propriospinal neurone and the motor neurone of the final common path; immediate spinal induction is referable to the internuncial propriospinal neurone. Sherrington's analysis of the scratch and other hind limb reflexes, such as flexion and extension, inaugurated a new and fuller era of neurophysiological research. By following a simple but rigorous course of methodical tests, he had, in the words of E. G. T. Liddell, "minutely cross-examined the spinal cord on its evidence,"72 unveiling and charting the characteristics of its reflex pathways. The scratch reflex was the last such large piece of reflex "machinery" that Sherrington analyzed. The data it yielded, showing how many nerve units are coordinated for a single, complex, purposeful neuromuscular act, made him see more clearly than ever the complex nature of reflexes and the intricacy of the central physiological mechanisms behind them. A new conception of the nature of reflexes and their role in the animal economy clearly was demanded by Sherrington's data, and was duly supplied in the concept of integrative action. After 1906 and the publication of the Silliman Lectures, problems which the scratch 72. The Discovery of Reflexes (Oxford: Clarendon Press, 1960), p. 136.

83

JUDITH P. SWAZEY

as how and other hind limb reflexes had raised afresh-such excitation is built up in spinal centers until the final motor to provide a paradigm that neurone is discharged-continued analysis of the motor and functional isolation guided him to the unit in 1930. THE INTEGRATIVE ACTION OF THE NERVOUS SYSTEM In 1904 the forty-seven year old Charles Sherrington, with his wife, joumeyed from the University of Liverpool to New Haven, Connecticut, to become Yale University's second Silliman Memorial Lecturer. Before departing for Yale Sherrington wrote with characteristic modesty to Harvey Cushing: "[I] possess misgivings of a serious character as to my shortcomings for the Silliman Lectureship. Mais il faut de l'audace, et encore de l'audace, et toujours de l'audace."'78 Sherrington delivered ten lectures on "Integrative Action by the Nervous System" at Yale's North Sheffield Hall between April 22 and May 4. Anyone who had known him might have predicted the mixed reactions of his audiences at the lectures. For, as Lord Cohen succinctly remarked, throughout his long career Sherrington "was not classed among the more successful lecturers." He eschewed dogmatism, he fought for the . . . word or phrase to convey precisely his meaning so that he appeared unduly hesitant; he qualified his statements, he inserted parentheses; indeed he gave the impression that his thoughts were not really in the lecture room but already contemplating and designing another experiment.74 Despite the mixed reactions which Sherrington's delivery of the Silliman Lectures evoked, their publication in 1906 was recognized as an epochal event in the development of neurophysiology. Dedicating his book to David Ferrier, whose researches had helped to interest him in neurophysiology, Sherrington entitled it The Integrative Action of the Nervous System. His choice of a title came after great deliberation, for as E. G. T. Liddell commented, Sherrington "used to spend a lot of time in choosing words."75Thus, he considered and abandoned the lectures' original title, "Integrative Action by the Nervous System," and a second alternative, "Integration in the Nervous System."78 73. Sherrington to Harvey Cushing, February 24, 1904. Fulton Papers. 74. Lord Cohen of Birkenhead, Sherrington, Physiologist, Philosopher, Poet. The University of Liverpool Sherrington Lectures, vol. II. (Liverpool: University of Liverpool Press, 1958), p. 14. 75. Personal communication. 76. Ibid.

84

Sherrington's Concept of Integrative Action before selecting the words which are now synonymous with his name. In the sixty years since its publication The Integrative Action has been reprinted six times, in 1911, 1914, 1916, 1947, and 1961. A revised second edition was suggested by the Yale University Press to Sherrington's former pupil and close friend, Dr. John Fulton, in 1932, when the book was out of print, but Fulton reacted strongly against the idea of textual revision. "I ventured to explain [to the head of the press]," he wrote to Sherrington on March 3, 1932, that revision of the book after so long a period would be exceedingly difficult but that I thought you might possibly be willing to prepare a preface which would give some idea of the relation of recent work to the concept of integration. To touch the original text would be like attempting to rewrite [Harvey's] De Motu Cordisl The demand for the book continues to be very considerable and it would be a great pity if the book should become no longer available to students.77 When The Integrative Action was reissued in 1947, it differed from the original only in its reset type and illustrations and the addition of a foreword by Sir Charles dealing with his more recent views on the concept of integrative action. The lasting value of this work for students of the nervous system is reflected clearly in the numerous reviews of the 1947 edition. F. M. R. Walshe, writing in the British Medical Journal, said: I have called it 'an imperishable work,' for it is one of those works, rare in science, the permanent value of which is unquestionable, and I believe that future generations of physiologists will so acclaim it. In physiology, it holds a position similar to that of Newton's Principia in physics ... For it is more than an orderly record of precise observations: it is a product of sustained thought upon what is essentiallythough only his genius revealed it as such-a single problem-namely, the mode of nervous action.78 The book which Walshe termed the Principia of physiology consists in essence of a synthesis of Sherrington's own researches and concepts and a chronicle of relevant work by other investigators. In its latest issue (1961) the book's ten Silliman Lectures total 390 pages, with 85 illustrations. A bibliography of 314 items, 22 of them Sherrington's publications, ranges from 77. Fulton Papers. 78. "A Foundation of Neurology. The Integrative Action of the Nervous System, by Sir Charles Sherrington," Brit. Med. J., 2 (1947), 823.

85

JUDITH P. SWAZEY

the writings of such seventeenth-century figures as Descartes to research papers published in 1905. The Integrative Action is not an easy book to assimilate, first and foremost because of the vast amount of material it contains. Moreover, Sherrington's style, characterized by long and seemingly complex sentences, is hard to follow at first. In writing, as in lecturing, he sought to convey his meaning precisely and unequivocally, a doubly difficult goal to realize when gearing a work for both a lay and professional audience. For analyzing the structure and major concepts of The Integrative Action the book may be divided into six parts: Sherrington's definition of his topic in the first seven and one-half pages of Lecture I; Lectures I-III: coordination in the simple reflex; Lectures IV-VI: coordination between reflexes-their interaction and compounding by simultaneous and successive combination; Lecture VII: reflexes as adapted reactions; Lectures VIII-IX: the brain's roles in integrative action; and Lecture X: sensual fusion. Sherrington's written analysis of the integrative action of the nervous system followed basically the same pattern as his research work. He began by delineating the characteristics of the simple reflex, the smallest functional unit of integrative action as seen in the spinal animal, and then built towards the complex patterns of reflex muscle management, guided by the brain, in the intact animal. For a person unfamiliar with Sherrington's writings the first seven and one-half pages of Lecture I provide an essential and succinct definition of the meaning and scope of the integrative action concept. As he had in his 1904 "Common Path" address, Sherrington began by defining three main lines of neurophysiological research. Nerve cells may be studied: 1) like all cells, as living units; 2) as regards their characteristic feature, conduction; and 3) in their integrative functions. Sherrington broadly defined the "integration of the [multicellular] animal organism" as the welding together of a "mere collection of organs" into an "animal individual."79 He observed that such integration is possible in part because the structure of multicellular organisms permits the operation of intercellular activities in addition to intercellular and cell surface ones, and by a brief comparison of the modes of intercellular integration he clearly depicted the uniqueness of the nervous system's structure and functions. It is the living bond of the nervous system, he emphasized, which is employed when the organism needs precise speed and fine temporal and spatial adjustments for such activities as movement. 79. Integrative Action, p. 2.

86

Sherrington's Concept of Integrative Action It is in view of this interconnecting function of the nervous system that the field of study of nervous reactions which was called at the outset the third or integrative, assumes its due importance. The due activity of the interconnexion resolves itself into the co-ordination of the parts of the animal mechanism by reflex action.80 The reflex arc is the unit mechanism for integrative nervous functions, he went on to explain, "because every reflex is an integrative reaction and no nervous action short of a reflex is a complete act of integration."81 As Sherrington's study of vertebrate motor behavior had demonstrated, however, not every reflex is a unit reaction, for some reflex actions are compounded of simpler reflexes. Therefore he distinguished between two grades of coordination. The first grade is that by the simple reflex: a given effector responds to a given receptor with all other parts of the body indifferent to and unnecessary for the reaction. He defined the second grade of nervous coordination as the coordination of reflex actions, a grade involving two phases. First, a number of simultaneously occurring simple reflexes are coadjusted into an orderly reflex pattern. Secondly, there must be orderly changes from one reflex pattern to another, a "co-ordination of reflexes successively proceeding."82 In his brief introduction to The Integrative Action, Sherrington has laid down three central propositions: 1) the nervous system is a, if not the, major integrating agent in complex multicellular organisms; 2) the reflex is the unit reaction in nervous integration; 3) there are two grades of reflex coordination-that effected by the simple reflex and that by the simultaneous and successive combination of reflexes. Working from these premises, he proceeded to demonstrate in meticulous detail how "the main secret of nervous co-ordination lies . . . in the compounding of reflexes."" The breadth and depth of The Integrative Action, the way in which Sherrington developed his analysis of reflex muscle management, his major conceptual statements and the researches on which he based them, can be fully appreciated only by studying the work itself. The reader will find that the Silliman Lectures are a brilliantly constructed study, drawn from contemporary knowledge of vertebrate evolution, anatomy, histology, and physiology, and, above all, encompassing and crystallizing the vast 80. 81. 82. 83.

Ibid., p. 5. Ibid., p. 7. Ibid, p. 8. Ibid.

87

JUDITH P. SWAZEY

yield of Sherrington's first quarter-century of researches. Moving step by careful step from the nature of coordination in the simple reflex to the physiological position and dominance of the brain in motor behavior, Sherrington delineated the central premise of his integrative action concept: "The nervous synthesis of an individual from what without it were a mere aggregation of commensal organs resolves itself into co-ordination by reflex action."84 NEUROPHYSIOLOGY'S WATERSHED From his first researches with John N. Langley in the early 1880's through the publication of The Integrative Action in 1906 Sherrington was engaged in a vast, multifaceted program of researches: brain-cord connections and spinal degenerations, distribution of motor and sensory roots, the proprioceptive system, the characteristics of synaptic reflex arc conduction, reciprocal innervation, the reflex patterns of the spinal, decerebrate, and intact animal, and the motor cortex. By examining the antecedents and tracing the course and contents of these researches, one sees how strikingly the new outweighed the old in Sherrington's work: in instance after instance he himself "made the time ripe" for answering a given problem. He pioneered new techniques and apparatus, such as the method of successive degenerations for tracing reflex pathways, surgical procedures for mammalian decerebration, and the use of the myograph for reflex recordings. He established new methodological canons with his meticulously designed and executed experiments. He marshaled and culled extant facts and theories and added a host of new ones about each topic he studied, emphasizing particularly the correlation of structural and functional data. The scope of his specific contributions clearly marks Sherrington as a major figure in the history of the neurosciences. The greater significance of his work, however, lies in his "synthetic attitude," in the fact that he perceived the interrelatedness of his varied researches. It seems clear that one of his goals as a researcher was to explain the functional unity of motor behavior, primarily by interpreting central nervous system function in histological terms. At what point he set this goal, and when he realized that all his researches were pointing toward its attainment, are questions which only Sherrington could answer precisely. From the contents of his 1897 Croonian Lecture on "The Mammalian Spinal Cord as an Organ of Reflex Action" it ap84. Ibid., p. 7.

88

Sherrington's Concept of Integrative Action pears that he was moving toward the integrative action concept by that date. At least in retrospect, Sherrington's work up to 1906 is resolvable into a three-fold study of reflex actions: their gross and histological architecture; the spinal and higher-level mechanisms controlling them; and their functions in vertebrate motor behavior. From this study, using the nerve cell and its interconnections as his basic analytical unit, he in turn wove the revolutionary fabric of the integrative action concept. The idea of nervous integration did not originate de novo with Sherrington. The fact of motor coordination and the participation of the nervous system in its operation had been recognized since antiquity, and prototypes of the integrative action concept may be unearthed from the ancient idea of "sympathy" through Pierre Flourens' studies of how specific brain regions effect an animal's functional unity. The significance of Sherrington's concept lies in the fact that it provided the first comprehensive, experimentally documented account of how the nervous system, through the unit mechanism of reflex action, produces an "integrated," or "coordinated," motor organism. In essence, then, I believe that the paramount contribution of Sherrington's work from the 1880's to 1906 is the unified meaning it gave to a host of phenomena previously discussed in isolation. The integrative action concept, and all that it embodied, brought together in a "final common path" many hitherto unconnected channels of neurophysiological, anatomical, and histological research. By this unification Sherrington provided investigators of the nervous system with their first major paradigm, and concomitantly laid down fruitful guidelines which refashioned the course of their researches after 1906. It was this watershed achievement, synthesizing the work of one era and opening a new one, that led Sherrington's peers to designate him as "'themain architect of the nervous system," "the supreme philosopher of the nervous system," "the author of the Principia of physiology," and the man who almost singlehandedly crystallized the special field of neurophysiology." ACKNOVVLEDGMENT The research on which this paper is based was supported by a National Institute of Health Predoctoral Fellowship from 1963 to 1966.

89

AugustWeismannand a Breakfrom Tradition FREDERICK B. CHURCHILL Department of the History and Philosophy of Science, Indiana University, Bloomington, Indiana

In 1892 August Weismann (1834-1914) presented to the scholarly world a remarkable theory which in a single sweep explained the diverse biological phenomena of heredity, development, regeneration, sexual reproduction, mitotic division, and Darwin's theory of evolution by natural selection. Only after a seasoned career had Weismann arrived at this comprehensive master plan-his germ-plasm theory.' He had completed his medical studies at G6ttingen in 1856 with a prizewinning essay on the source of hippuric acid in herbivors; he had worked briefly as an assistant at a chemical laboratory in Rostock; he had served as a military and private physician; he had returned to histological studies under Rudolph Leuckart (1822-1898) at Giessen and, after the completion of his "Habililationsschrift" on the metamorphosis of insects in 1863, he had become docent in zoology and comparative anatomy at Freiburg im Breisgau. At this university in the southwestern corner of Germany, Weismann had quickly risen through academic ranks to full professor and director of the zoological institute. Despite sustained eye trouble, which seriously curtailed his microscopic studies, he had completed by 1876 a series of monographic studies on metamorphosis in butterflies and in Axolotl with the object of examining the efficacy of natural selection. Furthermore, he had devoted considerable effort to investigating the ecology of and the parthenogenesis among the Clodocera of the Bodensee, and by 1880 he had begun studying the origin of the sexual cells in Hydrozoa. It was during the 1. Ernst Gaupp, August Weismann, sein Leben und sein Werk, (Jena: Gustav Fischer, 1917), is a full biography with a detailed description of Weismann's scientific work. Waldemar Schleip, "August Weismanns Bedeutung fur die Entwicklung der Zoologie und allgemeinen Biologie. Zu seinem hundersten Geburtstag am 17. Januar 1934," Naturwissenschaften, 22 (1934), 33-41, is a useful synopsis of Weismann's scientific accomplishments, but despite its title does not examine Weismann's influence on the later development of biology.

91

FREDERICK B. CHURCHILL

1880's that Weismann turned his attention to more general questions about the nature of the hereditary material. It is the development of these more general ideas, culminating in the germ-plasmtheory of 1892,2 that I intend to examine in detail in this paper. In doing so, I wish the reader to bear in mind a number of queries to which I will allude but will not discuss in detail. The first has been suggested by a number of contemporaryhistorianss and deals with the extent to which Weismann's ingenious synthesis helped prepare the way for twentieth-centurygenetics; the second touches upon the attitude of late nimeteenth-centurybiologists toward the role of speculation in their profession; the third revolves about the nature of mechanistic explanations in biology. At the moment I wish to put a more restricted question to Weismann's work namely, what did Weismann think he had accomplished when in 1892 he proposed such an elaborate and highly speculative theory? To put the question in a slightly different way, what sort of solution about development and heredity did Weismann wish to avoid? To answer this I must return to the earlier decades of the nineteenth century and review a particular brand of biological explanation which was inspired by Schwann's cell theory. Cranefieldhas describedthis approachon the part of "the 1847 group"of Germanphysiologists as a desire "toreduce physiology to physics and chemistry,"4 and Mendelsohn has emphasized that with Schwann's vision of cell formation and growth the reduction took the form of an analogy to the crystallizationof inorganic compounds.5Point by point, in fact, Schwann com2. August Weismann, Das Keimplasma. Eine Theorie der Vererbung (Jena: Gustav Fischer, 1892). The English translation appeared the following year: The Germ-Plasm. A Theory of Heredity, trans. W. Newton Parker and Harriet Ronnfeldt, The Contemporary Science Series, ed. Havelock Ellis (London: Walter Scott, Ltd., 1893) 3. For example, see Paul Ostoya, Les theories de l'evolution, origines et histoire du transformisme et des idees qui s'y rattachent (Paris: Payot, 1951) pp. 159-163; L. C. Dunn, A Short History of Genetics. The Development of Some of the Main Lines of Thought, 1864-1939 (New York: McGraw-Hill, 1965) p. 40, and Hans Stubbe, Kurze Geschichte der Genetik bis zur Wiederentdeckung der Vererbungsregein Gregor Mendel#, (Jena: Gustav Fischer, 1963) pp. 184-185. 4. Paul F. Cranefield, "The Organic Physics of 1847 and the Biophysics of Today," J. Hist. Med. 12, (1957), 407-423; for quotation see p. 423. Hermann von "The 1847 group" consisted of Carl Ludwig (1816-1895), and Emil du Ernst von Brucke (1819-1892), Helmholtz (1821-1894), Bois-Reymond (1818-1896). 5. Everett Mendelsohn, 'Physical Models and Physiological Concepts: Explanation in Nineteenth-century Biology," Brit. J. Hist. Sci. 2 (1965), 213.

92

August Weismann and a Break from Tradition pared the two processes, and although he found minor differences, he concluded: The view then that organisms are nothing but the form under which substances capable of imbibition crystallize, appears to be compatible with the most important phenomena of organic life, and may be so far admitted, that it is a possible hypothesis or attempt towards an explanation of these phenomena.6 With this attitude in mind I turn to Emst Haeckel, for it was he who promoted a conception of development and heredity which was fully within the spirit of Schwann's inorganic analogy. Ernst Haeckel (1834-1919) was a dramatic and controversial figure even in his youth.7 During his fifty year career at the University of Jena he crusaded for Darwin and relentlessly promoted a mechanistic interpretation of life. There is no question that Haeckel's numerous works had a strong influence on later generations of zoologists; both Hans Spemann (1869-1941) and Richard Goldschmidt (1878-1958) testify to this.8 Before turning his attention to more popular and discursive texts, Haeckel completed a careful study of Mediterranean Radiolaria and compiled a curious and speculative synthesis of contemporary views in anatomy, cell theory, and systematics. Published in 1866 as his Generelle Morphologie,9 this synthesis established particular notions of development and heredity against which I wish to measure Weismann's later achievements. It was in this work, moreover, that Haeckel made a speculative suggestion about the role of the nucleus as the bearer of hereditary material that was reflected in the later research of Edward Strasburger (1842-1912), Oscar Hertwig (1849-1922), and Weismann himself.10 6. Theodor Schwann, Microscopial Researches into the Accordance in the Structure and Growth of Animals and Plants, trans., Henry Smith (London: Sydenham Society, 1847), p. 215. 7. For Haeckel's meteoric rise in Jena, see Georg Uschmann, Geschichte der Zoologie und der zoologischen Anstalten in Jena 1779-1919 (Jena: Gustav Fischer, 1959), pp. 46-63. 8. Jane M. Oppenheimer, in Essays in the History of Embryology and Biology (Cambridge, Mass.: The MIT Press, 1967), cites passages from both Spemann and Goldschmidt which verify this (see pp. 209-210). 9. Ernst Haeckel, Generelle Morphologie der Organismen, allgemeine Grudziige der organischen Formenwissenschaft, mechanisch begrundet durch die von Charles Darwin reformirte Descendez-theorie, 2 vols. (Berlin: Georg Reimer, 1866). Its very title in fact, bespoke its comprehensive import. 10. William Coleman, "Cell, Nucleus, and Inheritance: An Historical Study," Proc. Amer. Phil. Soc., 109, (1965), 146-147, 157.

93

FREDERICK B. CHURCHILL

The Generelle Morphologie was a highly schematic work. In the first two hundred pages, which compared the various branches of science and defined the subject matter of morphology, Haeckel introduced an extensive, personally forged technical vocabulary and drew up a classification of anatomy which strikes the modern eye as excessively cumbersome for an active morphologist. Again and again Haeckel stated in subcategories the points which he had already made for the major categories. This was true when he discussed his classification of structures, or the reproduction of different orders of organic individuals, or differentiation of those organic individuals. The vocabulary and the repetitions added to the dryness of the text, but Haeckel's explanations of current biological problems were alluring because of the utter simplicity of their mechanisms. For Haeckel there was no basic difference between the material of the organic and inorganic world. The substrate of both, being atomistic, showed the same basic characteristics of extension, elasticity, inertia, and weight-properties which were innate to matter. All chemical elements found in the organic realm were also found in the inorganic, and although organic compounds were perhaps larger and more complex, Haeckel cited Wohler's synthesis of urea as proof that there was nothing unique about them. Stoff and Kraft were inexorably bound together, and since there was no special Lebensstoff, there could be no specially associated Lebenskraft." To be sure, Haeckel claimed organic bodies did exhibit a few unique qualities, but considering that he was trying to order the entire universe, these differences were minor. As he elaborated, the basic plasma, or material of life, possessed a lesser propensity for crystallization and a greater capacity for imbibing water, a characteristic which Schwann had also considered.'2 This plasma aggregated into clumps, or Plasmaklumpen, which Haeckel designated as Cytoden. By definition these cytodes were homogeneous in composition, which implied that it was impossible for an investigator to analyze them further into distinguishable subparts by microscope or chemical reagents."s Although customarily studied in communities as the structural parts of larger organisms, Haeckel felt these cytodes often existed as separate organic individuals, such as bacteria and protoamoebae. In this latter state Haeckel classified them as Moneren.14 The phylogenetic and structural appearance of the nucleus 11. 12. 13. 14.

94

Haeckel, Generelle Morphologie, I, 113-122. Ibid., 122-130; Schwann, Microscopical Researches, Haeckel, Generelle Morphologie, 1, 135-136. Ibid., 11, xxii-xxiii.

pp. 211-212.

August Weismann and a Break from Tradition marked for Haeckel the first stage in the differentiation of the homogeneous plasma clumps. The resulting unit he considered equivalent to the traditionally understood cell; the cytodes and cells together he designated Plastiden, or plastids, the first order of individuality.15 As schematic as this may seem, the transition between the inorganic and organic spheres and between homogeneous plasma clumps and differentiated cells needs emphasis, for it was around this hierarchial development that Haeckel constructed his morphology and explained growth, differentiation, and inheritance. To become attuned to Haeckel's comprehensive scheme requires as well an understanding of a double-faced concept of the organic individual. On the one hand, he argued, there existed the structural individual to be understood in terms of the plastids.16 The cytodes and cells, already mentioned, constituted the first order among such structural units; cell fusions, tissues, and organs-those structures which normally make up the subject matter of descriptive anatomy and embryologycomposed the second order of individuality; higher orders simply consisted of a further process of summation. Thus the third order, or Antimeren, existed in the homologous, or opposed parts; the fourth order, Metameren, consisted of the repeated segments along the main axis; the fifth order, Personen, comprised the customarily understood organism, and the final order, Stocke, denoted colonies. Each order (with the exception of the sixth) could exist as a subdivision of a higher one; yet each order (with the exception of the first) existed as the sum of lower biological individuals. Each level, therefore, must be interpreted eventually in terms of the first order, or the plastids. On the other hand, Haeckel viewed the individual as part of a continuous spiral of successive generations. Each generation, in completing its single cycle of genesis, development, and decline, depicted a genetic individual of the first order; while the summation of such cycles comprised the second order, that of the species. The products of reproduction, Haeckel explained, show the rhythmic, repetitive developmental unity out of the multiplicity of which the species builds itself. Both generation cycles together, the asexual division cycle and the sexually differentiated egg cycle, we can denote generally as the generation product or germ product, still better perhaps as the reproduction cycle. This Cyclus Generationis is our geneological individual of the first order. In conclusion, Haeckel described the third and highest genetic individual as the Phylon or stem.17 15. Ibid., I, 269-279. 16. Ibid., 289-331. 17. Ibid., II, 26-31. Quotation appears on p. 30.

95

FREDERICK B. CHURCHILL

This pluralistic image of the individual, seen both in its morphological sense as a summation of plastids and in its genetic sense as a cyclic succession of organisms, was characteristic of Haeckel's endeavor to unify all fields of biology. This juxtaposition of structural and phylogenetic units was of course the foundation of Haeckel's controversial and widely popularized biogenetic law. The point that I am emphasizing, however, before reviewing Haeckel's conception of growth, differentiation, and heredity is that he stressed a continuity of homogeneous material. This was so whether he thought of the organizational development within a single organism or thought cyclically from one biological individual to another. The continuity lay in the plastids, and it directly reflected Haeckel's first law of nature which depicted cause and effect as inseparably bound to the structural material.18 With this double image in mind, consider then what Haeckel meant by ontogeny, a term which he coined in the Generelle Morphologie. As defined by Haeckel this term embraced both embryogenesis and morphogenesis in adult life. The study of ontogeny implied the study of an organism from the moment of its existence to the moment of its demise.19 In a single word Haeckel united the problems associated with reproduction, growth, differentiation, and degeneration.20 He devoted many pages to analyzing these functions separately, but his emphasis was on them in sequence and brought into focus the entire life history of an organism. As Haeckel set about explaining development and heredity he did so within the unifying features of the four ontogenic functions. Haeckel considered growth2l as the most fundamental of the four. Organic growth resembled inorganic growth insofar as the individual behaved as a center which attracted to it from the surrounding fluid medium select molecules and incorporated them into its aggregate self. The only difference between the growth of a crystal and that of a plastid lay in the fact that the crystal increased in volume by apposition, whereas the plastid, because of its semifluidity, increased by intussusception. Growth of orders above the plastid level was simply considered 18. For Haeckel, the "first and highest of all natural laws" was the law of causation, which he considered simply a statement of the constant relationship between causes and effects. Haeckel equated a "causal" could understanding with a mechanistic one, although he did not-nor that such a correspondence necessarily followed from not-demonstrate the "first and highest" law. It was an article of faith which, once accepted, led him to a mechanistic biology. Ibid., I, 97-98. 19. Ibid., II, 1; 21-23. 20. Ibid., 72. 21. Ibid., 73-74.

96

August Weismann and a Break from Tradition a compound growth which combined the reproduction and growth of the constituent plastids. Haeckel identified reproduction with propagation.22 Again, espousing the basic case of the plastids, he argued that reproduction was simply a matter of division after a certain determined point of increase in size. In a critical sentence Haeckel claimed that "reproduction is a maintenance and a growth of the organism over and beyond the individual mass, one part of which is elevated to the whole." 23 I have singled out this "growth . . . over and beyond the individual mass" as the most conspicuous example of Haeckel's concern with material continuity. This concept I shall designate as Haeckel's "overgrowth" principle, and I consider it momentous, for it obliterated with a single stroke any distinction between growth and reproduction; Haeckel could now conveniently explain both in telrms of material accretion. With this single phrase Haeckel also erased from his mind any special consideration of sexual reproduction-a process which in the 1880's and 1890's returned to the forefront of fruitful research.24 In the two volumes of the Generelle Morphologie the most searching analysis of sexual reproduction appeared in the following meager passage: Sexual . . . reproduction . . . may be characterized only this way, that we hold as its criterion the mixture of two different stuffs, which are produced by two different individuals or by two different parts (sexual parts) of one and the same individual.25 Here again was Haeckel's overriding concern for the continuity of material (or of two materials). He reduced sexual reproduction to a special case of the "overgrowth" principle, and the example of the plastids, as usual, made his case appear selfevident. Some protozoans and algae exhibited an intermediate stage between asexual and sexual propagation since they enacted a partial or total fusion of plasma and nuclei before division (conjugation). For Haeckel this was "the first step toward sexual reproduction" 26 and rendered the "overgrowth" principle all the more credible. 22. Ibid., 16-17. Both to propagate and its German equivalent, fortpflanzen, imply an extension in space as well as time which is not true of to reproduce and zeugen. It is significant that Haeckel should consider fortpflanzen as an exact equivalent of zeugen. 23. Ibid., II, 16. 24. It is interesting to note Olby has recently claimed that Darwin was opposed "to the distinctive nature of sexual reproduction." Robert C. Olby, Origins of Mendelism (New York: Schocken Books, 1966), p. 91. 25. Ibid., 58. 26. GeneTelle Morphologie, II, 119.

97

FREDERICK B. CHURCHILL

Associated with the process of reproduction was the question of inheritance or the transmission of parental characters. Within the framework of the "overgrowth" principle this posed no problem at all; it simply ceased to exist. Haeckel made this clear in his more popular History of Creation as he described reproduction and inheritance in Monera: Now, when one examines this simplest form of propagation, this self-division, it surely cannot be considered wonderful that the products of the division of the original organism should possess the same qualities as the parental individual. For they are parts or halves of the parental organism, and the matter or substance in both halves is the same, and as both the young individuals have received an equal amount and the same quality of matter from the parent individual, one can but consider it natural that the vital phenomena, the physiological qualities, should be the same in both children. In fact, in regard to their form and substance, as well as to their vital phenomena, the two produced cells can in no respect be distinguished from one another, or from the mother cell. They have inherited from her the same nature.27 Haeckel's explanation of differentiation depended on the same obsession with growth and material continuity. Haeckel was a self-confessed proponent of epigenesis,28 or the belief that organisms differentiate from a homogeneous substrate. "One normally understands in this important process," he wrote, "generally a production of dissimilar parts out of a similar matrix." But the differentiation of one order of individuality, he elaborated, could only be understood in terms of the differentiation process in the lower orders which made it up. This was equally true for the plastids, where "we trace back with one word morphological and physiological differentiation to the chemical division of labor of the molecules." 29 This ultimate resort to chemical heterogeneity presented a contradiction in Haeckel's own epigenetic convictions, for he had earlier described the plasma of the cytodes as homogeneous as far as the microscope (Berlin: Georg 27. Ernst Haeckel, Naturliche Schopfungs-Ceschichte Reimer, 1868). This was intended to be a more popular version of the Generelle Morphologie. I have had access to the English translation, The History of Creation: or the Developnment of the Earth and Its Inhabitants by the Action of Natural Causes. 2 vols., trans. E. Ray Lankester (2nd ed.; London: Henry S. King, 1876), I, 191. The first emphasis is mine; the second is Haeckel's. 28. Haeckel considered himself in the epigenetic tradition of Wolff, Germans he proudly noted. Pander, von Baer, Remak, and Schleiden-all Generelle Morphologie, II, 13-15. 29. Ibid., 74.

98

August Weismann and a Break from Tradition or reagents could determine. Did this mean that, in fact, the homogeneous plasma was chemically differentiated and that organic differentiation must be explained by inorganic differences? At this critical point Haeckel remained unclear. In emphasizing Haeckel's "overgrowth"principle and his belief in epigenesis, I do not wish to accuse Haeckel of consciously constructing his biological theories within the strictures of a few convenient phrases. I have extracted these expressions because they offer a suitable shorthand for a combination of notions which the eclectic Haeckel stuffed into the Generelle Morphologie. They implied a commitment to the "reductionistic" side of Schwann's cell theory in that Haeckel understood growth simply as a crystal-like accretion of material. They stressed Haeckel's endeavor to formulate a unified theory of heredity and development and at the same time indicated his confusion of both processes. Epigenesis implied that embryological differentiation began in a homogeneous substrate. "Overgrowth" minimized the importance of sexual reproduction, and it is clear that this was no accidental phrase on Haeckel's part. He returned to a similar expression a few years later in his History of Creation when describing division of Moneren. Thus he concluded: "Here it is evident that the process of propagation is nothing but a growth of the organism beyond its own individual limit of size." 30 Nor was the phrase considered accidental by his contemporaries, for, as I shall show, Weismann singled it out for special attention. In fact, combined together, epigenesis and "overgrowth" may very well express a mid-nineteenth-century attitude which implied a common set of assumptions for many investigators who examined the same combination of problems. In his early works Weismann appears steeped in just those notions which Haeckel combined in the Generelle Morphologie of 1866.31 In an essay on the mechanical conception of nature, Weismann agreed with Haeckel explicitly that nutrition, growth, and heredity were analogous mechanistic processes and must be understood in terms of molecular motions.32 30. The History of Creation, 1, 187. Emphasis is Haeckel's. 31. The recently published correspondence between Haeckel and Weismann indicates the high regard Weismann had at first for the Generelle Morphologie. See especially "Weismann and Haeckel, 21 Mai 1867" and Haeckel and Weismann, 28 Juni 67," in Georg Uschmann and Bernhard Hassenstein, eds., "Der Briefwechsel zwischen Ernst Haeckel und August Weismann" in Kleine Festgabe aus Anlass der hundertjahrigen Wiederkehr der Grundung des Zoologischen Institutes der FriedrichSchiller-Universitat Jena im Jahre 1865 durch Ernst Haeckel, ed., Manfred Gersch, (Jenaer Reden und Schriften; Jena: Friedrich-SchillerUniversitiit, 1965), pp. 6-68. 32. August Weisman, Studies in the Theory of Descent, trans. and ed.

99

FREDERICK B. CHURCHILL

Several years later, in 1883, in fact, he directly employed the idea of "overgrowth." "Haeckel," Weismann declared, was probably the first to describe reproduction as "an overgrowth of the individual," and he attempted to explain heredity as a simple continuity of growth. This definition might be considered as a play upon words, but it is more than this; and such an interpretation rightly applied, points to the only path which, in my opinion, can lead to the comprehension of heredity.83 The "rightly applied" interpretation-namely, the close identification of growth and propagation-became immediately apparent as Weismann sought to elucidate the above statement with examples of fission among Infusoria. "Among these unicellular organisms," he concluded, "heredity depends upon the continuity of the individual during the continual increase of its body by means of assimilation." 34 Weismann viewed more complicated forms of propagation as the result of specialization of this simple case-first with the appearance of homogeneous cellcolonies and then with division of labor within the colony. This was a familiar line of reasoning used earlier by Haeckel. The years between 1878 and 1883, however, have their interest for the student of Weismann. It was during this time that he carefully studied the development of the reproductive cells in Hydrozoa.35 It was also a time when he gave considerable thought to the biological significance of senescence and death.86 Both lines of investigation led directly in 1883 to his famous inaugural address, "On Heredity," delivered at the University of Freiburg, in which Weismann advanced his own unique idea of continuity. Where Haeckel had emphasized a continuity beRaphael Meldola. (London: Sampson Low, Marston, Searles, & Rivington, 1800-1882), pp. 666-668. This book appeared originally as Studien zur and consisted of five essays of Deszendenztheorie (Leipzig, 1875-1876) which the last was entitled "The Mechanical Concept of Nature." 33. August Weismann, "On Heredity," republished in Essays upon Heredity and Kindred Biological Problems, trans. Edward B. Poulton, Selmar Schonland, and Arthur E. Shipley, 2 vols. (2nd ed.; Oxford: ClarI, 72, endon Press, 1891-1892), 34. Ibid., 73. 35. Weismann's Die Entstehung der Sexualzellen bei den Hydromedusen. Zugleich ein Beitrag zur Kenntnis des Baues und der Lebenserscheinungen dieser Gruppe. (Jena, 1883) is his definitive statement of this line of research. N. J. Berril and C. K Liu, "Germplasm, Weismann, and Hydrozoa," Quart. Rev. Biol. 23, (1948), 124-432, have shown in this excellent article Weismann's dependence upon Haeckel's biogenetic law in establishing a continuity of the germ-plasm. 36. See for example, August Weismann, "The Duration of Life," in Essays upon Heredity, I, 1-66, written originally in 1881.

100

August Weismann and a Break from Tradition tween generations of individuals, Weismann asserted that the key to both heredity and development was the continuity of the substance of the germ cells within the life span of a single generation. He suggested that a continuous and distinct germ-track coursed its way through the life history of the individual, and since it passed to the next generation, it alone survived the death of the somatic cells. Weismann distinguished two lives, the somatic life and the life of the germ-plasm, and he considered this distinction a morphological reality: We cannot explain this fact except by the supposition that each reproductive cell potentially contains two kinds of substances, which at a variable time after the commencement of embryonic development, separate from one another, and finally produce two sharply contrasted groups of cells.37 Schleip has pointed out how essential this distinction was to the further development of Weismann's thought.38 The important question at the moment, however, is the extent to which Weismann's idea of continuity challenged Haeckel's "overgrowth" principle and belief in epigenesis. From an historical distance it might be argued that to distinguish between the germ-plasm and the somatoplasm was to clarify Haeckel's confusion between the transmission of heritable characters on the one hand and assimilation, growth, and differentiation on the other. Curiously enough, this was not the case. A previous quotation (see p. 100) has already indicated that Weismann could favorably refer to "overgrowth" in the same Freiburg address. Further direct references could be made.39 There are other indications of the compatibility between the somato- and germ-plasm distinction and "overgrowth." In 1881 Weismann contrasted the death of multicellular animals and the apparent immortality of single-celled organisms which re37. Weismann, "On Heredity," p. 74. 38. Schleip, "August Weismann," p. 37. Weismann himself recognized that other investigators, such as Francis Galton, Gustav Jager, August Rauber, and Moritz Nussbaum, independently arrived at somewhat similar ideas of continuity, The Germ-Plasm, pp. 198-212. 39. At the end of his 1885 address Weismann returns even more strongly to Haeckel's position ("On Heredity," p. 105. The emphasis is

mine): If, as I believe, the substance of the germ-cells, the germ-plasm, has remained in perpetual continuity from the first origin of life, and if the germ-plasm and the substance of the body, the somatoplasm, have always occupied different spheres . . . we can, up to a certain point, understand the principle of heredity; or, at any rate, we can conceive that the human mind may at some time be capable of understanding it ... for we can thus trace heredity back to growth; we can thus look upon reproduction as an overgrowth of the individual.

101

FREDERICK B. CHURCHILL

produced by division.40 Multicellular organisms, Weismann continued, were nothing more than communities of cells which had increased in complexity during evolution and which had become ever more specialized. As the course of evolution proceeded, Weismann visualized that only certain cells and their descendants retained the capacity to reproduce the entire organism. The process entailed a segregation of reproductive cells from somatic cells. The idea of continuity of the germ-plasm was inherent in this argument, for with the archaic metazoan all the cells were potentially germ cells and therefore potentially immortal. An internal continuity of the germ-plasm by the nature of this simplest case was merely a conjunctive concept. With a division of labor or a differentation into somatic cells, such a continuity was never interrupted; it merely became more restricted as phylogenetic history progressed. Somatic specialization simply signified a reduction of the number of reproductive cells. From this point of view Weismann's concept of continuity seems hardly a substantial innovation; in fact, it was nothing more than Haeckel's "overgrowth" principle carried to its logical conclusion. An "overgrowth" was valid simply for fewer cells. By 1883 Weismann had convinced himself that there was a continuity of the germ track throughout the life of the individual. It is fair to ask, however, what this continuous substance was. Besides making a distinction between somatic cells and germ cells, Weismann could not be precise about the nature of the reproductive or developmental material. Recognizing this failure, he projected the ultimate goal for "geneticists" and embryologists past the bounds of the gross cellular elements to the molecular structure beyond. "I am unable," he insisted to indicate the molecular and chemical properties of the cell upon which the duration of its power of reproduction depends: to ask this is to demand an explanation of the nature of heredity-a problem the solution of which may still occupy many generations of scientists.41 In a memorable work of 1885,42 Weismann was hardly clearer about the nature of the germ-plasm. He considered it confined to the nucleus and to consist of a chemical mixture of 40. Weismann, "The Duration of Life," in Essays upon Heredity, I, 28-29. 41. Ibid., 29. 42. August Weismann, "The Continuity of the Germ-Plasm as the Foundation of a Theory of Heredity," in Essays upon Heredity, 1, 163-255. This monograph presented a much stronger statement of Weismann's understanding of continuity.

102

August Weismann and a Break from Tradition ancestral germ-plasms. When he turned to elucidate the process of differentiation, he reasonably retained the same germ-plasm, conceived in the "reductionist" spirit of chemical and physical properties and described as an undefined molecular structure. This image, after all, had served him well in working out inheritance in terms of continuity; there was no need for a reconsideration before applying it to embryogenesis. Taking advantage of recent descriptions of mitosis, however, Weismann made one elaboration: during the process of histological differentiation, he argued, cell divisions must be qualitatively unequal: I therefore believe that we must accept the hypothesis that, in indirect nuclear division, the formation of unequal halves may take place quite as readily as the formation of equal halves, and that the equality or inequality of subsequently produced daughter-cells must depend upon that of the nuclei.43

The decade between 1875 and 1885 was a remarkable one in the history of cytology.44 It opened with the publication of Ueber Zellbildung und Zelltheilung by Edward Strasburger (18441912), in which the author described the chromatin movements in indirect cell division of Spirogyra. It progressed with a series of papers by Walther Flemming (1843-1915), in which he traced the nuclear material from one interphase to the next and described the longitudinal splitting of the chromatin threads, which are now known as chromosomes. The decade terminated with a paper by Karl Rabl (1853-1917), in which the author demonstrated the constancy of chromosomal number. Furthermore, is was a decade in which Oscar Hertwig (1849-1922) observed the fusion of male and female pronuclei in fertilized sea urchin eggs; when Edouard van Beneden (1846-1910) suggested that maturation division consisted in a halving of the chromosomal number, and Wilhelm Roux (1850-1924) speculated that the nucleus might divide qualitatively as well as quantitatively. 43. Ibid., 193. Although Weismann cited Wilhelm Roux's 1883 paper on mitotic division elsewhere in this monograph, he did not give Roux credit for the idea of qualitative nuclear division until 1887; see Weismann, "On the Number of Polar-bodies and Their Significance in Heredity," in Essays upon Heredity, I, 369-371. 44. Good sources for the history of cytology of this period are: Arthur Hughes, A History of Cytology (London and New York: Abelard-Schuman, 1959), pp. 62-73; John R. Baker, "The Cell-theory: a Restatement, History and Critique. Part V. The Multiplication of Nuclei." Quart. J. Microscop. Sci. 96 (1955), 465-472, and Coleman, "Cell, Nucleus, and Inheritance: an Historical Study," Proc. Amer. Phil. Soc. 109 (1965), 124-158.

103

FREDERICK B. CHURCHILL

Weismann contributedto the increasing understandingof cytological events, too, although, as I shall show, his immediate interpretationsconflicted with the later judgment of his peers. The combination of this research helped steer Weismann and others away from Haeckel'simage of a homogeneoushereditary and developmental substance to a belief in discrete nuclear units. In 1886 Weismann turned his attention to the meaning of sexual reproductionand by so doing began an assault on one of the critical assumptions of Haeckel's work. Lecturing before the Associationof GermanNaturalists, Weismann concentrated on showing that sexual reproductionintroduceda large number of variations upon which natural selection could operate.45It is clear that he was here attacking the established conclusion that "each organism is capable of producinggerms, from which, theoretically at least, exact copies of the parent may arise."46 By contrast, Weismann argued that sexual reproductionwas not a blending of two parental gern-plasms, as Haeckel's "amphigomc" reproduction implied, but the combining of many discrete ancestral germ-plasms, each of which maintained its identity. By accepting discrete ancestral germ-plasm units, Weismann broke away from Haeckel's "overgrowth"principle. In Weismann's mind sexual reproduction was now not just the

ill-definedblending of two "stuffs,"but a unique combinationof material worthy of investigation in its own right. "Sexualreproduction,"he pointed out, "notonly adds to the number of existing differences, but it also brings them into new combinations, and this latter consequence is as important as the former." 47

Weismann, however, still clung to the associated embryological idea of epigenesis, and since his germ-plasmis so commonly identifiedas an aggregateof preformedparticles, the trail of this inconsistency is worth tracing. "Thedevelopmentof the nucleoplasm during ontogeny," Weismann declared in 1885, in an interesting but to the modern ear an awkward analogy, may be to some extent compared to an army composed of corps, which are made up of divisions, and these of brigades,

and so on. The whole army may be taken to represent the nucleoplasm of the germ-cell: the earliest cell-division . . . may be representedby the separation of the two corps, similarly formed but with different duties; and the following cell-divisions by the successive detachment of divisions, bri45. August Weismann, "The Significance of Sexual Reproduction in the Theory of Natural Selection," Essays upon Heredity, 1, 257-342. 46. Ibid., 275.

47. Ibid., 279, 283-284. Quotation appears on p. 287.

104

August Weismann and a Break from Tradition gades, regiments, battalions, companies, etc.; and as the groups become simpler so does their sphere of action become limited. It must be admitted that this metaphor is imperfect ... because the quantity of the nucleoplasm is not diminished, but only its complexity... And we must also guard against the supposition that unequal nuclear division simply means a separation of part of the molecular structure, like the detachment of a regiment from a brigade. On the contrary, the molecular constitution of the mother-nucleus is certainly changed during division in such a way that one or both halves receive a new structure which did not exist before their formation.48 Here was a beautiful description of unequal nuclear division within the framework of epigenesis. What was the ultimate content of this nucleus? Not preformed units ready to come into play, but a complex molecular structure which upon division rearranged itself into two different and less complex structures. Was this explanation of differentiation compatible with the idea of discrete ancestral germ plasms? For seven years, between

1885 and 1892, Weismann felt it was.49 When, in 1892 Weismann published his best-known work, The Germ-Plasm, a Theory of Heredity, he did more than gather under one cover a decade of tempered ideas. It was here for the first time that he discarded the Haeckelian tenet of epigenesis and accepted the viewpoint that ontogeny as well as heredity must be explained in terms of a particulate hypothesis. Weismann's decision may have been precipitous; only once had there been an inkling of a possible conversion. The suggestion came in 1890 when in a general defense of his ideas to date he reviewed Darwin's "provisional" theory of pangenesis.50 Weismann had always objected to the "throwing off, circulation, and collection" of gemmules which was paramount in Darwin's theory, but he looked with interest at Hugo de Vries' (18481935) peculiar version described in the Intracellular Pangenesis of 1889.51 In this short but provocative work, de Vries had maintained that each heritable characteristic became manifest 48. "Continuity of the Germ-Plasm," Essays upon Heredity, I, 195. Emphasis is mine. 49. Weismann recognized the discrepancy himself in the introductory chapter of The Germ-Plasm when he confessed that up until 1892 he "tried in several ways to arrive at a satisfactory epigenetic theory" (pp. 4-5). 50. August Weismann, "Remarks upon Certain Problems of the Day," in Essays upon Heredity, II, 71-97. 51. Hugo de Vries, Intracellular Pangenesis (Jena, 1889), reprinted in Hugo de Vries Opera e Periodicis Collata, 7 vols. (Utrecht: A Oosthoek, 1918-1927), V, 1-149.

105

FREDERICK B. CHURCHILL

through the action of corresponding living units. By definition, these units, or pangenes, could self-replicate and hence comprised a higher order of organization than inorganic molecules. Furtherxnore, these pangenes existed in either one of two states, depending upon whether or not they were immediately involved in the expression of a characteristic: the active pangenes, multiplying in the cytoplasm, imprinted on the cell its specific character; the inactive ones, remaining in the nucleus, awaited their tum to migrate into the cytoplasm of the appropriate daughter cell. 2 Here was a scheme which was totally different from the epigenesis offered by Haeckel and Weismann. "The future will decide," Weismann asserted, "whether the assumption of modified gemmules furnishes a better explanation of the facts of heredity than my hypothesis." 63 It is difficult to pinpoint the cause of Weismann's change of mind. The cogency of de Vries' own arguments may simply have required three years to prevail over Weismann's ill-defined belief in epigenesis. There are, however, several events which helped Weismann reform his opinion. One in particular is worth describing in detail. Since the third decade of the century, investigators had noted the presence of polar bodies in association with the development of the ovum.54 Because of the primitive state of cytology and the great variability in the sequence of events of maturation among different animal species, the relation between fertilization, nuclear division, and the number of polar-body expulsions remained confused until the 1880's. The Belgian zoologist Edouard van Beneden (1846-1910) renewed the effort to clarify these events. In 1883, in a detailed study of the maturation of the ova of the nematode Ascaris megalocephala (bivalens),55 van Beneden established that the chromatin material of the male and female pronuclei failed to fuse at fertilization. Instead, each of the four "chromosomes" split longitudinally so that the first two blastomeres received four "halves" apiece, two traceable to the father and two to the mother. It was evident that the ovum had to halve its original compliment of four "chromosomes" to prepare for this pairing process, so van Beneden argued that this reduction took place somehow during the expulsion of two polar bodies. Weismann found that van Beneden's work on polar-body ex52. Ibid., 132-138. 53. "Remarks upon Certain Problems of the Day," Essays upon Heredity, II, 81. 54. Hughes, History of Cytology, pp. 67-70. 55. I have here abbreviated the account given by Coleman in "Cell, Nucleus, and Inheritance," pp. 140-141.

106

August Weismann and a Break from Tradition pulsion corroborated his own distinction between germ-plasm and somatoplasm. After studying the polar bodies in Daphnidae, he decided that the involved chromatin material corresponded to the ovogenetic somatoplasm which was necessary for the formation of the egg but which hindered future embryonic development. "The expulsion of the polar bodies" he concluded, "is nothing more than the removal of the ovogenetic nucleoplasm from the egg-cell." 56 Further study of the phenomenon, however, forced him in 1887 to modify this stand.57 With the aid of his student, Chiyomatsu Ishikawa, Weismann discovered that parthenogentic eggs uniformly produced only one polar body. To be consistent, Weismann had to consider the first and second polar bodies as performing different functions. "Only one polar body can signify the removal of ovogenetic nucleoplasm from the mature egg, and the second is obviously a reduction of the germ-plasm itself to half of its original amount." 58 This second conclusion beautifully enhanced his broader theory of heredity of 1886, for Weismann recognized that a reduction of the ancestral germ-plasms was needed to compensate for the doubling which took place at fertilization. He remained unsure, however, how a similar reduction took place in spermatogenesis.59 It was Oscar Hertwig's study of 1890 which clarified the events of spermatogenesis and forced a further revision of Weismann's explanations.60 In this extensive monograph Hertwig reviewed all the phenomena of egg and sperm development in Ascaris. He discovered that he was able to compare point by point the two processes from oocytes and spermatocytes to the completion of maturation division. He recognized that both processes included the same unusual sequence of two nuclear divisions and involved a reduction in the same number of "chromosomal" elements. Hertwig reasoned that since the obcytes and the spermatocytes pass through the same process of nuclear division and include in exactly the same way all the peculiarities which deviate from the norm, the products of division must also possess the same morphological value... The polar bodies therefore have the morphological value of egg cells."6' 56. Weismann, "Continuity of the Germ-Plasm," Essays upon Heredity, I, 218. 57. August Weismann, "On the Number of Polar Bodies and Their Significance in Heredity," in Essays upon Heredity, I, 343-396. 58. Ibid., 357. 59. Ibid., 385-388. 60. Oscar Hertwig, "Vergleich der Ei-und Samenbildung bei Nematoden. Eine Grundlage fur celluliire Streitfragen," Arch. Mikroskop. Anat. 36 (1890), 1-138. 61. Ibid., 71.

107

FREDERICK B. CHURCHILL

Weismann found that this was a convncing analogy;62yet it put him in an awkward position. If the polar bodies were merely undeveloped eggs, he could not insist that any of them contained expelled ovogenetic somatoplasm. "My previous interpretationof the first polar body as the removal of ovogenetic nucleoplasm from the egg," he lamented, "must fall to the ground."16 How then was he to explain away the fate of this somatoplasm after it had accomplished its mission of forming the egg? De Vries'idea of pangenes came to his rescue. Weismann reasoned that the ovogenetic somatoplasm must migrate from the nucleus into the cytoplasm. 'In this way alone can we account for no trace of it remaining in the nucleus, and for embryonic developmentnot being subsequentlyimpededby its presence."6' In the process of fulfilling its mission of molding the cytoplasm of the egg, the ovogenetic somatoplasm must be "used up"-a concept which was entirely alien to the progressivetransformation suggested in Weismann's metaphor of marching divisions and brigades. This "using up," Weismann wrote, "indicates, if it does not prove, that the control of a cell by a determinant is accompaniedby the absorptionof the latter, and a further support may thereby be obtained for the hypothesis of emigration."85 To "use up" the directing germ-plasmcould only imply discrete units, and once Weismann had employed such units to form the somatoplasmof the egg, he could equally well use them for other stages of development. It is worth noting that this pertinent four-page section in The Germ-Plasm was labeled "Proof that the Determinants become Disintegrated into Biophors."86 Weismann's units for ontogeny began with these biophors, the ultimate particles in the architectureof Weismann's germ62. August Weismann, "Amphimixis, or the Essential Meaning of Conjugation and Sexual Reproduction," in Essays upon Heredity, II, 117. 63. Ibid., 122. 64. Weismann, The Germ-Plasm, p. 350. 65. Ibid. The emphasis is mine. The entire sentence in the German edition is worth repeating to indicate Weismann's own choice of words: Diese Determinante wird verbraucht und verschwindet als solche, und darin, dass thatsachlich keine Ausstossung derselben aus dem Ei erfolgt, liegt, wenn nicht ein Beweis, so doch ein starker Hinweis darauf, dass die Behrrschung der Zelle durch eine Determinante mit ihrer Vernichtung einhergeht, woraus dann weiter eine Stutze fur die Auswanderungs-Hypothese entnommen werden darf (Das Keimplasma, p. 459). Here the emphasis is Weismann's. The reference to "die AuswanderungsHypothese" was, of course, a reference to de Vries, Intracellulare Pangenesis.

66. The Germ-Plasm, p. 348.

108

August Weismann and a Break from Tradition plasm and a key to his new concept of differentiation. These biophors were first and foremost units of life, and as such represented the complete break from the "reductionist" attitude of the Generelle Morphologie. As basic vital units they consisted of complex chemical compounds which possessed unlimited arrangements and combinations. The only qualification rested on the capacity of these units to assimilate additional chemical material, increase in size, and replicate themselves. The combined activities of the biophors, which compose the basic structural unit of all protoplasm, added up to the known vital phenomena of the individual cells. Since these phenomena were not properties of ordinary chemical molecules, Weismann argued that a complex unit must exist which possessed these capacities. It is for this reason, he asserted, that "the biophors are not, I believe, by any means mere hypothetical units; they must exist, for the phenomena of life must be connected with a material unit of some sort." 67 This all sounded very much like de Vries' pangenes, as Weismann himself confessed.68 The biophors were cytological units responsible for the functions and structures of individual cells. The nucleus itself, Weismann reasoned, "must be in a sense a storehouse for the various kinds," 69 and his ensuing elaboration entailed the complex hierarchy of self-replicating units. Weismann considered that each independent somatic character was controlled by a higher vital unit, or determinant. Corresponding to the ancestral germ-plasm, the continuity of which he had so laboriously worked out, Weismann proposed yet a higher unit, the id. Each id contained a full complex of determinants for the development of an entire organism, but since any given individual represented a combination of ancestral traits, the nuclear "storehouse" contained many ids. These in turn were arranged into clusters, or idants, which perhaps, though not definitely, could be identified with chromosomes. Development therefore consisted in a sequential segregation by means of qualitative division. Differentiation was the result of a translation of ancestral traits into histological phenomena by the timely release of biophors.70 One additional trimming to this "storehouse" of particles is worth mentioning. Weismann had to explain the capacity of some animals to regenerate lost or destroyed parts. A clarification of this question, after all, was in particular demand after experimental embryologists, such as Wilhelm Roux (18501924) and Hans Driesch (1867-1941), had focused attention 67. 68. 69. 70.

Ibid., Ibid., Ibid., Ibid.,

44. 42. 48. 60-75.

109

FREDERICK B. CHURCHILL

on these very phenomena. Weismann solved this problem to his satisfaction by proposing that a reserve germ-plasm with a full complement of accessory ids, determinants, and biophors was carried along in embryogenesis to the appropriate peripheral parts. This implied that as the somatoplasm segregated during differentiation, the reserve germ-plasm accompanied it on a parallel path with the difference that the germ-plasm divided in a qualitatively equal rather than unequal fashion. In an emergency, such as an amputation of a limb, the reserve germ-plasm lying dormant on the edge of the wound could discharge the necessary determinants to regenerate the missing part.71 Weismann's solution was far from convincing. Oscar Hertwig was quick to point out that this post hoc explanation destroyed the very advantage which Weismann had hoped to build into his system, for if there was any appeal in qualitatively unequal nuclear division during ontogeny, Weismann had annulled it by introducing this reserve germ-plasm which necessarily divided equally.72 Hertwig's criticism was more than just, but it is to my point to discover the type of solution Weismann wished to avoid. While discussing alternative explanations of regeneration, Weismann concluded that the organism appears to us like a crystal whose broken points always complete themselves again from the mother-lye after the same system of crystallization . . . But the difference between the organism and the crystal does not-as people have been hitherto inclined to believe-lie only in the fact that the crystal requires the mother-lye to complete itself, while the vital unit itself procures the material for its further growth; it lies also in the fact that such regeneration is not possible in every organism and at every place, but that special "primary constituents" are necessary, without which the relevant part cannot arise. The indispensableness of these primary constituents, the determinants, seems to me to depend on the fact that the new structure cannot be built up simply by procuring organic material, but that specially hewn stones, different in every case, are necessary, which can only be supplied in virtue of an historical transmission.73 The contrast between broken crystals assimilating new material and "primary constituents" emphasizes the complete change of 71. Ibid., 140-141. 72. Oscar Hertwig, Zeit-und Streitfragen der Biologie. Vol. I. Praformation oder Epigenesis? Grundziuge einer Entwicklungs theorie der Organismen (Jena: Gustav Fischer, 1894), pp. 72-75. 73. Weismann, The Evolution Theory, trans., J. Arthur and Margaret R. Thompson, 2 vols. (London: Edward Arnold, 1904) U, 36. Emphasis is Weismann's.

110

August Weismann and a Break from Tradition mind which Weismann had experienced. Here, too, with reference to this special problem of regeneration, Weismann was responding to his earlier ideas of "overgrowth" and epigenesis; in fact he was implicitly responding to the "reductionist" biology of Schwann and Haeckel. Weismann broke from certain assumptions contained in the Generelle Morphologie when he adopted a particulate theory of heredity and development. His acceptance of such units was important, not so much because it added hypothetical molecular structures to the spectrum of matter lying between the nucleus and the chemical atom, but because it meant restoring a special biological explanation to certain phenomena of life. This may seem an odd attribute for Weismann, who emphasized time and again his own bias for mechanistic biology. But mechanistic biology did not always mean a total reduction to matter in motion. Weismann defined his hypothetical living units by their special functions of assimilation, growth, and replication, and he claimed to recognize their presence by the somatic events. In insisting on such fundamental units, he denied that an unorganized, homogeneous mass passed on in a gamete could sufficiently account for the phenomena of heredity. It may have been just a matter of time and the prompting of some cytological discoveries which forced him to apply the same conclusion to development as well. Weismann's germ-plasm theory of 1892 incited considerable adverse professional criticism. As with the remark of Oscar Hertwig, given above, many of these comments were merited. But many attacks focused on the abstract nature of the ids, determinants, and biophors, which raises an interesting issue in the methods of science. Hertwig, for example, comparing his own and Weismann's work, concluded that the germ-plasm theory proffers us directly a closed system in which appears a formal explanation for each and everything. Indeed, when examined in the light, this form of explanation is more of a renunciation of an explanation. For it is explained with formulas and symbols which evade observation and experiment and therefore cannot be the subject of an objective research. With this explanation nothing more is given than a description of that which occurs in development before our eyes.74 The American cytologist, Edmund Beecher Wilson (18561939), wrote in even more scathing terms as he linked Weismann's germ-plasm theory with Wilhelm Roux's mosaic theory of development: "From a logical point of view the Roux-Weis74. Hertwig, Zeit-und

StTeitfragen, pp. 136-137.

ill

FREDERICK B. CHURCHILL

mann theory is unassailable. Its fundamental weakness is its quasi-metaphysical character, which indeed almost places it outside the sphere of legitimate scientific hypothesis."75Weismann had an answer to this attack which at least revealed that he was aware of the tentative nature of many of his ideas: To go on investigating without the guidance of theories, is like attempting to walk in a thick mist without a track and without a compass. We should get somewhere under these circumstances,but chance alone would determinewhether we should reach a stony desert of unintelligible facts or a system of roads leading in some useful direction; and in most cases chance would decide against US.76 What constituted a 'legitimate scientific hypothesis" about heredity and development? This is a question to which historians, as well as biologists, have yet to find a suitable answer. If it be found that Weismann's germ-plasmtheory contributedin a significantfashion to the subsequentgrowthof ideas, this itself may be a sufficient-although not sole-criterion. In this paper I have analyzed a fundamental change in Weismann'sthinking. It is worth suggesting that Weismann's gradual rejection of the "reductionist"assumptions of Haeckel's "overgrowth"principle and patronage of epigenesis constituted the real value of the germ-plasm theory of 1892. 75. Edmund B. Wilson, The Cell in Development and Inheritance. Columbia University Biological Series. IV. (New York: Macmillan, 1896), p. 306. By the third and definitive edition, now entitled The Cell in Development and Heredity, Wilson considerably softened his tone, although not his objection to Weismann. Essays upon 76. Weismann, "Significance of Sexual Production..." Heredity, I, 305. This particular statement was actually made in 1886, but there are examples expressing the same idea which follow the publication of The Germ-Plasm. See, for example, August Weismann, The Evolution Theory, I, 396-397.

112

ThomasHunt Morganand the Problemof NaturalSelection GARLANDE. ALLEN Department of Biology, Washington

University, St. Louis, Missouri

INTRODUCTION At the beginning of the present century a number of prominent biologists were voicing strong doubts about the Darwinian theory of natural selection. Despite some historians' statements to the contrary,' the literature of this period indicates that the idea of natural selection had by no means gained wide or whole-hearted acceptance even as late as 1915. Among the particularly outspoken critics were a number of experimental biologists, including William Bateson, T. H. Morgan, Ross G. Harrison, and E. G. Conklin. Members of a younger generation born after 1860, these men viewed the theory of natural selection as incomplete and of little value in understanding the origin of species. Both a failure to solve some of the long-standing problems which Darwin had left behind, as well as the appearance of new evidence which threw doubt on some of the basic assumptions of natural selection, were the chief causes for this growing skepticism. By the 1930's, however, through new work in heredity, plant and animal systematics, and morphology, many of the doubts were resolved, and a new, more comprehensive concept of natural selection emerged. The present paper will indicate some of the arguments brought to bear against Darwin's theory during the early years of the present century, and suggest, through a specific case study, some of the factors which may have been responsible for its ultimate acceptance. The version of Darwianian theory with which most biologists 1. See, for example, R. Hofstadter and W. Smith, eds., American Higher Education, A Documentary History, 2 vols. (Chicago: University of Chicago Press, 1961), II, 841-842. The statement is made: "Academic scientists were quick to embrace Darwinism, to which almost all had been converted by the mid-1870's. "See also F. Rudolph, The American College and University (New York: Alfred Knopf, 1962), pp. 346-347:" "The scientists, young and old, accepted Darwinism with remarkable readiness."

113

GARLAND E. ALLEN

were familiar at the turn of the century had been set forth in the sixth and final edition of The Origin of Species. Since more offspring were always bom than the environment could support, Darwin maintained, some would have to perish. The slight heritable differences among organisms gave some individuals of a species a slightly greater survival value than others. Those with favorable variations survived longer than those with unfavorable variations. However, the key to evolution by natural selection was not simply long survival, but rather what Darwin referred to as "differential fertility," the ability to leave a greater number of offspring to the next generation. Selection preserved favorable variations by allowing those organisms which carried them to pass these variations on to a greater number of secondgeneration offspring. The one persistent problem which Darwin had left unsolved was the mechanism by which new characters could arise and be passed on from one generation to another in a population of organisms.2 Lacking a workable explanation of the origin of variations, Darwin and his followers had frequently fallen back on a revised version of Lamarckian theory. They explained the appearance of specffic adaptive variations as the result of environmental influence on the tissue of the organism, or as the result of use and disuse of parts. In earlier editions of the Origin, most notably the first, Darwin had consciously avoided reliance upon Lamarckism. But criticism from a variety of directions had forced him to adopt more of this older idea about how variations could arise. In reading Darwin, as in reading Shakespeare or the Bible, it is possible to support almost any viewpoint desirable by focusing on certain isolated passages or ideas. To his credit, Darwin was not definite and dogmatic about issues where he felt the evidence was not conclusive, but, by emphasizing one factor or idea over another in discussing evolution, many workers in the post-Darwin period often violated this principle. Claims that natural selection itself caused variations, that variations in certain directions established an evolutionary momentum, that all variations were the results of mechanical pressure of parts of an organism, abounded among naturalists of all sorts. These 2. Gavin DeBeer, Charles Darwin (New York: Thomas Nelson, 1963), pp. 175, 188, 203; also, the same author's "Mendel, Darwin and Fisher," Notes and Rec. Royal Soc. 19 (1964), 210-211. In an article which points up more the contemporary status of evolutionary and hereditary theory than its historical development, Th. Dobzhansky points out that Darwin "acknowledged that the origin of variation was unknown." "Mendelism, Darwinism, and Evolutionism," Proc. Phil. Soc. 109 (1965), 205. Also, R. G. Swinburne's analysis of Galton's laws of inheritance: "Galton's Law-Formulation and Development," Ann. Sci. 21 (1965), 15.

114

Thomas Hunt Morgan and Natural Selection interpretations, often masquerading under the cloak of Darwinism were, in reality, depressing exaggerations of Darwin's own thought. As a result, it was often difficult to know exactly what Darwinian theory had come to mean, so loosely had the term been used. Although few biologists at the turn of the century seriously questioned Darwin's intellectual integrity, or his genius in reasoning from the data, a large number appear not to have read his work very thoroughly. Gaining their ideas from the promient neo-Darwinians, they did not always understand what the theory of natural selection was about. Opposition to Darwinian theory seems to have resulted as much from a lack of understanding of Darwin's ideas as from over-reaction to the exaggerated claims of his followers. With no clear evidence to suggest (1) that some, or all individual variations are heritable, and (2) that such variations are rigorously acted upon by selection, many biologists were unwilling to fully support the Darwinian theory. It was thus the lack of a workable theory of heredity which made it difficult to get around some of the arguments brought against the mechanism of natural selection. What the strict Darwinians needed was a workable theory of heredity to help convince their skeptical colleagues of the validity of the selection theory. In turn, those who criticized Darwinian theory could have used some corroborative evidence from another area of biology (such as the study of heredity) that complemented the idea of natural selection. In the latter part of the nineteenth century no such theory was available, but after 1900 this was no longer the case. With the rediscovery of Mendel's laws in that year, it would seem that the way would at last have been cleared for showing how variations were inherited in a definite and predictable manner; this was not the case, however. Those who objected to Darwinian theory before 1900 continued to object to it for a decade or more, even with full knowledge of Mendelian concepts before them. Conversely, those who expounded either orthodox Darwinism or some modification of it tended to see little applicability to the mechanism of natural selection. Yet today we view the two theories as complementary, each contributing to and illuminating the other. It is thus instructive to ask why the two viewpoints remained so far apart for over a decade, and what factors ultimately determined their merger. The work of Thomas Hunt Morgan (1866-1945) provides a particularly good focal point for a study of the relationship between these two fields during the period of 1900-1915. Throughout his long career, Morgan was always concerned with the problem of natural selection, while at the same time becoming, after 1910, one of the leading proponents of the Mendelian

115

GARLAND E. ALLEN

theory. At first hostile to natural selection as a means of explaining the origin of species, Morgan eventually became a champion of an integrated theory in which Mendelian inheritance became the chief means of explaining the origin of heritable variation. His change in viewpoint was not quick or easy, and his intellectual struggles indicate clearly many of the real problems which his contemporaries also faced. As a strong proponent of the experimental tradition, Morgan came to accept Darwinian selection only after he could see how the central assumption on which it was based - the origin of heritable variation - could be rigorously tested. He thus helped to lead a new movement in biology by insisting on the application of rigorous methods to traditionally descriptive and qualitative subjects. MORGAN AND NATURAL SELECTION The view which a biologist holds about the nature of species will determine much of what he believes about the mechanisms of evolution - the means by which species have arisen. Following what Ernst Mayr has called the "nominalist" tradition,8 Morgan held that species- were only arbitrary units created for convenience by taxonomists. The only real unit in nature, he felt, was the individual: We should always keep in mind the fact that the individual is the only reality with which we have to deal, and that the arrangement of these into species, genera, families, etc. is only a scheme invented by man for purposes of classification. Thus, there is no such thing in nature as a species, except as a concept of a group of forms more or less alike.4 "Adaptation" was the key word in Morgan's view of species: "a group of forms more or less alike" because they shared a number of common adaptations - not because they were alike in trivial details such as bristle number. When taxonomists spoke of species, Morgan held, they spoke of groups differentiated on the basis of nonadaptive characters, and their concept of evolution thus had no relation to the development of adaptations.5 Like Darwin, Morgan considered adaptation the final out3. E. Mayr, The Species Problem (Washington, D.C.: AAAS Symposium, 1957), pp. 4-7. Although given its greatest boost by Darwin, who, despite the title of his book, did not believe species groups were real in nature, the nominalist tradition also found strong advocates in Lamarck and Hugo de Vries. 4. T. H. Morgan, Evolution and Adaptation (New York: Macmillan, 1903), P. 33. 5. T. H. Morgan, "Chance or Purpose in the Origin and Evolution of Adaptation," Science 31 (1910), 203-204.

116

Thomas Hunt Morgan and Natural Selection come of evolution, and so he rejected taxonomic species concept because it seemed to have so little to do with these important differences between groups of organisms. As he put it in 1910: If, then, the systematists' definition of species is what we mean when we speak of species, and this definition does not concern adaptive characters (or only incidentally), clearly it is futile to atempt to explain the origin of species by the theory of natural selection.6 Morgan became interested in the question of evolution through his own work on regeneration. In 1903 he wrote: One of the general questions that I have always kept before me in my study of regenerative phenomena is how such a useful acquirement as the power to replace lost parts has arisen, and whether the Darwinian hypothesis is adequate to explain the results.7 "The conclusion I have reached," he continued, "is that the theory is entirely inadequate to account for the origin of the power to regenerate; and it semed to me, therefore, desirable to reexamine the whole question of adaptation for might it not prove true here, also, that the theory of natural selection was inapplicable? This was my starting point." 8 While Morgan always believed in evolution, he felt that the theory of natural selection had many flaws and did not provide an adequate mechanism for the evolution of adaptations. What were these inadequacies, and why was Morgan so adamant in rejecting natural selection? SELECTION DOES NOT ACT ON CONTINUOUS VARIATION Among Morgan's most general objections was that Darwin was confused about the types of variations on which natural selection operated. Darwin distinguished between two types of variation: large variations which appeared suddenly ("sports"), and the small, minute variations ("individual differences") on which he felt that natural selection primarily acted. Objections to this view had gained considerable prominence in England after the publication of William Bateson's Materials for the 6. Ibid., p. 203. 7. Evolution and Adaptation, pp. ix-x. 8. Ibid. For Morgan's views on regeneration as explained by natural selection as early as 1898, see "Some Problems of Regeneration:" Biological Lectures Delivered at the Marine Biological Laboratory of Woods Hole in the Summer Session of 1897 and 1898. Tenth Lecture, pp. 193-207.

117

GARLAND

E.

ALLEN

Study of Variation in 1894.9 Bateson, a proponent of discontinuous variation, had suggested that evolution occurs by definite, discrete steps. Discontinuous variations were much larger than individual differences, occurring at one jump and showing no intermediates between themselves and the parental form. Discontinuous variations, Bateson claimed, were the only ones which were inherited and therefore the only ones upon which natural selection could act. In the period before 1910, Morgan supported the idea that natural selection acted upon definite or discontinuous variations. He felt there were two reasons why continuous (sometimes called fluctuating) variations could not be considered sources of evolutionary advance: first, there was no proof that continuous variations could be inherited;10 second, continuous variations could not occur in a large enough single step to form a new species. To Morgan, continuous variations represented quantitive rather than qualitative differences among members of organisms, that is, a normal distribution about a mode. Selection of one extreme or the other among continuous variations could not ultimately produce new species. According to Morgan, the most that selection could accomplish would be to keep the type to the upper limit attained in each generation by these variations. Despite the fact that Morgan maintained species to be arbitrary units, his thinking was nevertheless tinged with a typological prejudice.'1 He thought of a species as having a fixed mode, a form like rubber ball which could be altered temporarily by an outside agent (in this case selection) but which always returned to its original shape. Only distinct breaks with the type could produce a new species.'2 INSUFFICIENCY OF IDEAS ON DISCONTINUOUS VARIATIONS While he could not accept evolution by continuous variations, Morgan was also skeptical about the idea of evolution by discontinuous variations. He saw the same objection which Fleeming Jenkins had seen in 1867: the loss of a new variation (in Darwin's term, "sport") through swamping. The fact that Morgan's view of swamping was developed with Mendel's laws available 9. London: Macmillan & Co. 10. Evolution and Adaptation, p. 267. 11. Ernst Mayr has pointed out in his introduction to The Species Problem (Washington, D.C.: AAAS, 1957, pp. 5-7), that many biologists, alike, held a typological view of the naturalists, and experimentalists species during the earlier part of the present century. 12. As we shall see later, Morgan was a strong supporter of De Vries' mutation theory in the period 1900-1910.

118

Thomas Hunt Morgan and Natural Selection to him implies that he had not yet come to understand or accept the new concepts of heredity as of 1903. To Morgan, discontilnuity could only account for the origin of species if the new variant were infertile with its parent form. As he wrote in 1903: We see then that discontinuity in itself, unless it involved infertility with the parent species of which there is no evidence, cannot be made the basis for a theory of evolution, any more than individual differences, for the swamping effect of intercrossing would in both cases soon obliterate the new form. If, however, a species begins to give rise to a large number of individuals of the same kind through a process of discontinuous variation, then it may happen that a new form may establish itself, either because it is adapted to live under conditions somewhat different from the parent form, so that the dangers of intercrossing are lessened, or because the new form may absorb the old one.13 There was one way around the dilemma which swamping posed to those who tried to understand heredity and evolution. Morgan assumed that if enough of the same discontinuous variations occured at one time, the possibility of two organisms with the same variation mating would be increased. Thus, the variation could be passed on in full strength to the next generation. Morgan admitted to having no evidence that large numbers of the same mutation could occur at once in a given population, so that his suggestion remained only as assumption. NATURAL SELECTION ONLY A NEGATIVE FACTOR To Morgan, natural selection was only a negative factor in origin of adaptations. It could select out the unfit, but could not create the new variations which could be favorably selected.'4 Selection could not explain the origin of the fit, but only the failure of the unfit to propagate. Thus, for him, the theory of natural selection was incomplete on the very point that Darwin had devised it to explain: the origin of species. The problem of how distinct and heritable variations arose and were inherited was to Morgan still unsolved in 1903. Morgan's attitude regarding this negative aspect of selection was revealed by his reaction to the experiments of W. E. Castle and H. MacCurdy at Harvard's Bussey Institution.'5 Strict 13. Evolution and Adaptation, pp. 286-287. See also the same objection to sexual selection, p. 195. 14. Ibid., p. 462. 15. H. MacCurdy and W. E. Castle, Selection and Cross-breeding in Relation to the Inheritance of Coat-pigments and Coat-patterns in Rats and Guinea Pigs. (Carnegie Institution of Washington, Publication No. 70, 1907).

119

GARLAND E. ALLEN

Darwinians and selectionists, MacCurdy and Castle were investigating the nature of inheritance of spotted and albino patterns in rats and guinea pigs to discover whether the albino condition was simply one extreme of the spotted coat pattern (that is, of a fluctuating variation) or a separate factor altogether. Breeding results indicated that albino was indeed a separate factor. During these experiments, however, they made another important discovery: while selection of the least-spotted offspring in each generation could not fix a pure-breeding line of albinos, it could produce a different degree of spottedness which bred true. Thus, selection and inbreeding tended to fix individual variations to any degree desired. These results seemed to indicate that selection could, indeed, produce new lines or races which could breed true even when selection was relaxed.'8 Morgan was skeptical of MacCurdy's and Castle's experiments, arguing in 1907 that even though spotted and albino were separate factors, there might be "several or even many semi-stable states of the coat" which selection could seem to fix as true-breeding forms.17 In other words, the spotted-coat gene could exist in a variety of states, which could be affected by selection. Thus, MacCurdy's and Castle's interpetation was not the only possible conclusion from their results.'8 SELECTION CANNOT ACCOUNT FOR INCIPIENT STAGES OF USEFUL ORGANS Morgan's fourth objection to natural selection was essentially that voiced by St. George Mivart (1827-1900) in 1871:19 if selection acted on small individual differences which occur by chance, fine levels of adaptation such as the vertebrate eye could never have become established because their incipient stages would not provide enough advantage for favorable selection. Darwin had successfully refuted this argument in the sixth edition of The Origin of Species by pointing out that no matter how little advantage an incipient structure provided, if 16. It is important to note that these workers were selecting continuous variations and drawing from their results a conclusion essentially in agreement with orthodox Darwinism. 17. T. H. Morgan, "Review of Selection and Cross-Breeding in Relation to the Inheritance of Coat-pigments and Coat-patterns in Rats and Guinea Pigs, by H. MacCurdy and W. E. Castle," Science 26 (1907), 751-752. 18. At this time, neither Morgan nor Castle and his associates were aware of the concepts that later came to be known as multiple alleles, or modifier genes. The development of these concepts played an important role in the history of genetic and evolutionary theory, and will be dealt with in a later section of this paper. 19. St. George Mivart, The Genesis of Species (New York: Appleton, 1871), esp. Chap. II, on the development of incipient structures.

120

Thomas Hunt Morgan and Natural Selection it were of any advantage at all it would definitely be favorably selected.20 Morgan, however, maintained that he could not see how, by random variations acted upon by natural selection, something as complex and wonderfully adapted as the vertebrate eye could develop from an initial bit of light-sensitive tissue.2' He compared this situation to the problem which had originally brought him to the question of evolution-regeneration-asking how partial regeneration of a limb could provide any selective advantage. To be adaptive, regeneration must be complete.22 THE OBJECTION TO LAMARCKIAN PRINCIPLES Morgan's fifth objection to natural selection was that Darwin himself had leaned too heavily on Lamarckian principles, relying in many places on the principle of use and disuse to explain certain evolutionary novelties. As Morgan wrote of one example: "By falling back on the theory of inheritance of acquired characters Darwin tacitly admits the incompetence of natural selection to account for the evolution of the flatfish." 28 This indicated to Morgan that the whole idea of selection acting on minute variations was insufficient, even in Darwin's eyes. What he did not seem to realize was that Darwin had not seen natural selection and Lamarckian use and disuse as an either-or choice, but had opted for the middle road and accepted both. Morgan was disconcerted by the school of neo-Lamarckism which had arisen in the latter part of the nineteenth century, feeling that there was no evidence to suggest that the effects of use or disuse of a part could be passed on to the next generation. In 1903 he discussed the trend toward Lamarckism in the thinking of American naturalists, A. D. Cope, Alphaeus Hyatt, J. A. Ryder, and others: A number of evolutionists, more especially of the American school, have tried to show that the evolution of a number of groups can best be accounted for on the theory of the inheritance of acquired characters... Despite the large number of cases that they have collected, which appear to them to be most easily explained on the assumption of the inheritance of acquired characters, the proof that such inheritance is possible, is not forthcoming. Why not then spend a small part of 20. Charles Darwin, The Origin of Species (New York: Appleton, 1889), esp. pp. 177-181. 21. Evolution and Adaptation, pp. 131-132. 22. Ibid., pp. 380-381. 23. Evolution and Adaptation, p. 138. Morgan had not followed the various editions of The Origin of Species, but he was fully aware that Darwin had moved more and more closely to Lamarckian views between 1859 (lst ed.) and 1872 (6th ed.).

121

GARLAND E. ALLEN

the energy, that has been used to expound the theory, in demonstrating that such a thing is really possible: One of the chief virtues of the Lamarckian theory is that it is capable of experimental verification or contradiction, and who can be expected to fumish such proof if not the neoLamarckians? 24 Neo-Lamarckians spent a good portion of their time devising elaborate hypotheses, and trying to explain away objections to their theory.25 To Morgan, all of this was beside the point, for neo-Lamarckians failed to ask verifiable questions and to put their theory to experimental test: We may fairly sum up our position in regard to the theory of the inheritance of acquired characters in the verdict "not proven." I am not sure that we should not be justified at present in claiming that the theory is unnecessary and even improbable.26 THE NATURE OF SCIENTIFIC METHODOLOGY Perhaps the most general area of criticism which Morgan leveled at the Darwinian theory concerned scientific methodology. Morgan was an empiricist, concerned (first and foremost) that hypotheses should be testable and should be in agreement with all facts. He felt that many of the strict Darwinians, neoDarwinians, and particularly the neo-Lamarckians, had engaged in flights of fancy and speculation which had no basis in fact.27 An even greater offender along these lines was the German biologist August Weismann (1834-1914), who was particularly notorious, from Morgan's point of view, because of his widespread popularity and enormous reputation. Weismann had tried to unite three important areas of biology-evolution, 24. Evolution and Adaptation, p. 260 (see also pp. 230-231). 25. For a discussion of the neo-Lamarckian movement in America, see the recent study by Edward J. Pfeifer, "The Genesis of American neoLamarckism," Isis 56 (1965), 156-167, where the works of Hyatt, Cope, Dall, LeConte, and Clarence King are treated at some length. For European neo-Lamarckism, one of the best sources for the nineteenth century is Herbert Spencer's The Principles of Biology (New York: Appleton, 1866) I, 184-200; 402-431. See also P. G. Fothergill, Historical Aspects of Organic Evolution (London, 1953), and Y. Delage and M. Goldsmith, The Theories of Evolution, trans. A. Tridon (New York: B. W. Huebsch, 1912). 26. Evolution and Adaptation, p. 260. 27. See, for example, T. H. Morgan, "Regeneration and Liability to Injury," Science, 14 (1901), 235-248; "The Origin of Species through Selection Contrasted with Their Origin through the Appearance of Definite Variations," Pop. Sci. Monthly, 67 (1905), 54-65; Evolution and Adaptation, pp. 126, 163, 165ff, 171-172, 180.

122

Thomas Hunt Morgan and Natural Selection heredity, and development-into one comprehensive view. By inventing a hierarchy of hereditary units which were parceled out during development, and which, by their state of nourishment gained greater or lesser prominence in the next generation, he developed an enormously elaborate conceptual scheme. Weismann's greatness lay in his ability to see biology as a whole and to unite its many disparate branches. To a large extent his speculative approach to biology set the pattern, especially for evolutionary work, in the decades after 1885. To Morgan and his generation, however, Weismann's work presented more problems than it could solve. Morgan objected to the speculative and fanciful nature of Weismann's theories28 and he chastized him for failing to distinguish between an assumption and a fact,29 and to be specific and accurate in reporting experiments.30 But it was on the nature of theory formation that Morgan leveled his most vehement attack: Thus Weismann has piled up one hypothesis on another as though he could save the integrity of the theory of natural selection by adding new speculative matter to it. The most unfortunate feature is that the new speculation is skillfully removed from the field of verification and indivisible germs (particles) whose sole functions are those which Weismann's imagination bestows upon them, are brought forward as though they could supply the deficiencies of Darwin's theory. This is, indeed, the old method of the philosophizers of nature. An imaginary system has been invented which attempts to explain all difficulties, and if it fails, then new inventions are to be thought of. Thus, we see where the theory of selection of fluctuating germs has led one of the most widely known disciples of the Darwinian theory.31 Weismann's conception of the role of theory (in the period after 1900), and Morgan's were indeed antithetical.32 To Weis28. Evolution and Adaptation, pp. 165-166. 29. Ibid., p. 163. 30. "Regeneration and Liability to Injury," Science, 14 (1901), 235248. 31. Evolution and Adaptation, pp. 165-166. 32. Weismann had done some important cytological work in the 1860's predicting the occurrence of meiosis, among other things, before it was actually shown microscopically. For an evaluation of Weismann's work, see Arthur Hughes, A History of Cytology (New York: Abelard-Schumann, 1959), esp. Chap. IV; see also William Coleman, "Cell, Nucleus and Heredity." Proc. Amer. Phil. Soc. 109 (1965), 124-158, esp. pp. 151-154. The best available source on Weismann's life and works is E. Schleip, "August Weismanns Bedeutung fur die Entwicklung der Zoologie und 22 (1934), 33-41. A number allgemeinen Biologie," Naturwissenschaften of Weismann's own writings have been translated into English-a testi-

123

GARLAND E. ALLEN

mann, working in the tradition set by Darwin in The Origin of Species, a theory served primarily to gather in the facts in an umbrella fashion.33 To Morgan, on the other hand, a theory served primarily as a means of giving direction to further research, so that for a theory to be of any value in science, it had to be framed in such a way as to be testable. TELEOLOGYAND EVOLUTION Philosophically,Morgan found that one of the most difficult problems in evolutionarytheory was the question of chance or purpose. Always opposed to teleological explanations in science, he maintained firmly that variations in organisms did not arise because they were needed but occurred solely by chance.84In 1910 he pointed out that chance operates at two differentlevels in evolution, and that confusion of these levels had led many writers to adopt some sort of teleological explanation.85The first level was that of the occurrenceof one out of many possible events: for example, of the many possible variations which could occur, a particularone actually did take place, and whether it was variationA, B, or C was purely a matter of chance. The second level was that of the occurrence of a chance event at a particular time and place (unconnected with other events at that time or place) such as "I chanced to be there."In evolutionary terms the one chance event, a single variation, will be favorably selected only if it occurs in a particularenvironment at a particular time. The occurrence of the variation and the set of conditions composing the environment are two independent sets of chance events, both of which must be considered in discussing the role of chance in evolution. Morganmade this distinction in order to get to a more basic point: that the concept of "chance" did not imply anything mysterious, or outside the realm of cause and effect.36In opposing those who held that chance events could not be studied scientifically, he pointed out that, in fact, it was the opposing idea-that of purposefulness-which was obscure and metaphysical.87After all, Darwin himself had used the term "chance mony itself to his widespread influence. See The Evolution Theory, trans. J. Arthur and Margaret P. Thomson (London: Edward Arnold, 1902), and Essays upon Heredity and Kindred Biological Problems, trans. E. B. Poulton and Arthur Shipley, 2 vols. (Oxford: Clarendon Press, 1892). 33. A. Weismann, Vortrage uber Descendenztheorie, 2 vols., (Jena: Gustav Fischer, 1902), II, 4. 34. Evolution and Adaptation, pp. 391-394. 35. T. H. Morgan, "Chance or Purpose in the Origin and Evolution of Adaptations," Science, 31 (1910), 201-210. 36. Ibid. 37. T. H. Morgan, "For Darwin," Pop. Sci. Monthly 74 (1909), 367-380.

124

Thomas Hunt Morgan and Natural Selection variation" as synonymous with "fluctuating variations."38 Although he did not know the causes of these variations, his use of the term "chance" did not imply that they had no cause; but only that they did not originate for the purpose of fulifiling a specific need: The origin of adaptive structure and the purpose it comes to fill are only chance combinations. Purposefulness is a very human conception for usefulness. It is usefulness looked at backwards. Hard as it is to imagine, inconceivably hard it may appear to many, that there is no direct relation between the origin of useful variations and the ends they come to serve, yet the modern zoologist takes his stand as a man of science on this ground. He may admit in secret to his father confessor, the metaphysician, that his poor intellect staggers under such a supposition, but he bravely carries forward his work of investigation along the only lines that he has found fruitful.39 Despite the many objections to Darwinian theory which he brought forward, Morgan nevertheless was able to see the value which Darwin's work held-both at that particular time as well as historically. Both men were concemed with the origin of adaptations.40 Morgan also felt that Darwin has served the admirable function, historically, of putting the question of evolution on a sound scientific basis, collecting many observations and arguing his theories directly from them. Morgan could only object that many of the neo-Darwinians had not followed their leader's example in this regard, but had wandered too far from the facts. Furthermore, Darwin was concerned with a mechanism of evolution that was not teleological and that did suggest a direction for further research. MENDELIAN PRINCIPLES, MUTATIONISM AND EVOLUTION The writings of both Hugo De Vries (1848-1935) and Gregor Mendel (1822-1884) emphasized the discontinuity of variations in nature. Their investigations came to the attention of biologists at a time (1900) when the problems of heredity as they related to evolution were assuming great importance, providing a new look at old subjects. In 1903 Morgan indicated the great influence these works had in the early years of this century: 38. "Chance or Purpose . . .," pp. 201-210. 39. "For Darwin," p. 380. 40. Evolution and Adaptation, p. ix.

125

GARLAND E. ALLEN

The whole question of inheritance has assumed a new aspect; first on account of the work of De Vries in regard to the appearance of discontinuous variation in plants; and secondly, on account of the remarkable discoveries of Gregor Mendel as to the laws of inheritance of discontinuous variations.4' De Vries' work was concerned with the origin, Mendel's with the inheritance, of variations. These seemed to be two areas of investigation which complemented Darwin's theory of natural selection. While greatly attracted to De Vries' theory, Morgan and many of his contemporaries could not accept Mendelism as throwing much light on the Darwinian theory. Morgan himself found that "the chief value of Mendel's results lies in their relation to the theory of inheritance rather than to that of evolution."42 Morgan's failure to see that Mendelian principles could fill in the weak or missing parts of the theory of natural selection arose out of a single problem: strange as it may seem in light of his later work, it is apparent that he did not understand Mendelian principles in the period before 1909 or 1910. In a previous paper I have dealt with his many objections to Mendelism in the first decade of this century.43 Some of these objections were valid at the time, since the Mendelian theory still lacked general applicability; but others were the results of Morgan's failure to understand the Mendelian laws in their full significance. What were some of his reasons for saying that Mendel's results did not apply to evolution? First, Morgan claimed that Mendelian variations (contrasting factors) were not exactly the small individual differences on which Darwin claimed natural selection acted. To Morgan, Mendelism indicated that Darwin had drawn too sharp a distinction between the results of the inheritance of continuous and those of discontinuous variation.44 The problem was that Morgan, between 1900 and 1910, thought of Mendelian variations as too definite to be the type of which Darwin spoke, and too slight to be the mutations which De Vries emphasized. Mendelism seemed to be between both current theories of evolution, and thus to apply to neither. Second, he thought that Mendelian variations (such as a recessive character, once it appeared) would be subject to 41. Ibid., p. 278. 42. Ibid., p. 286. See also T. H. Morgan's Poultry" by C. B. Davenport, Science 25 (1907), 43. G. E. Allen, "Thomas Hunt Morgan and mination," Proc. Amer. Phil. Soc. 110 (1966), 44. Evolution and Adaptation, p. 286.

126

review of "Inheritance in 465. the Problem of Sex Deter48-57.

Thomas Hunt Morgan and Natural Selection swamping in the same way as minute individual differences. Unlesss self-fertilization occurred, or unless a number of the same variations appeared simultaneously, Morgan felt that a given variation of this sort would disappear from the population within a few generations. In 1903 Morgan felt that the Mendelian concept was no better than Darwinism in getting around the problem of swamping. The fundamental problem here lay in Morgan's suspicion about the cornerstone of the Mendelian concept-the purity of gametes.45 Before 1909 or 1910 Morgan saw purity as only an assumption unsupported by any direct evidence. His own view of the union of egg and sperm was that the hereditary material of the two fused as completely as two drops of water.46 The idea that the factor from one parent did not chemically influence the corresponding factor from the other parent was foreign at this time to his own strongly physiological orientation. Failing to accept the idea of "purity of gametes" would necessarily throw the alternative concept of blending (of which swamping is the result on a population scale) into stronger relief. This, added to his confusion about the nature of species, led him to miss completely the possible application of Mendelian principles to the problem of evolution by natural selection. Far more influential in Morgan's thinking at this time was De Vries' "mutation theory." Morgan had visited De Vries' laboratories at Hilversum in Holland, and had been greatly impressed with the examples he had seen of the mutating plant Oenothera lamarckiana, the evening primrose.47 He wrote: Unexpectedly, new light has been thrown on the questions of variation and of evolution by the immensely important experiments of Hugo De Vries of Amsterdam. No one can see his experimental garden, as I have had the opportunity of doing, without being greatly impressed, for here on all sides are new species that have suddenly appeared, fully 45. T. H. Morgan, Sex Determining Factors in Animals," Science 25 (1907), 382-384. 46. T. H. Morgan, "Chromosomes and Heredity," American Naturalist 44, 449-496. 47. There seems to be some uncertainty as to when Morgan actually did visit De Vries. E. A. Carlson, in his book, The Gene: A Critical History (Philadelphia: W. B. Saunders, 1966, p. 40), states it was in 1903. However, A. H. Sturtevant, in his biographical sketch in Nat. Acad. Sci., Biog. Mem. 33, p. 290, says it was probably 1900, but does not cite any evidence as to how that date was determined. Morgan was in Europe in 1900 and 1902, attending the Naples Marine Station for research work, but, as far as records indicate, was not in Europe at all in 1903. See C. A. Kofoid, The Biological Stations of Europe (Washington, U.S. Bureau of Education Bulletin No. 4, 1910), pp. 16-17.

127

GARLAND E. ALLEN

equipped, from a known original parent form, living now side by side with its group of descendants.48 Morgan could accept De Vries' mutation theory as a means of evolution because it showed how species might originate without invoking the haphazard and problematical theory of selection. Thus, where Mendelism threw light on inheritance of definite variations, it still required selection for evolution to occur. De Vries' theory avoided selection and all of its attendant problems and allowed Morgan to avoid the rigid but tenuous distinction between continuous and discontinuous variations. To Morgan, the mutation theory did not exclude the role of natural selection or the existence of fluctuating variation. Each mutant form represented a mean about which a range of fluctuation occurred. However, the important point was that mutations were definite and were inherited with great constancy. Morgan was even willing to compromise, for he did not maintain that all adaptations arose by mutation.49 Furthermore, he agreed that mutations could be jumps that were no greater than the extreme of fluctuating variation. Thus, he saw that evolution could be a gradual process while still taking place by specific and definite mutations.50 This new approach gave him a rounded-out concept of evolution which was far more successful in his own mind than natural selection had ever been.8' FACTORS RESPONSIBLE FOR MORGAN'S CHANGE OF VIEW ON THE EFFICIENCY OF NATURAL SELECTION Before 1910 Morgan opposed both Darwinian natural selection and Mendelian genetics, but he eventually came to accept and even champion both. His own work with Drosophila began in about 1908 with the purpose of trying to demonstrate De Vriesian mutations in animals. When his experiments began to yield results consistent with the Mendelian scheme, Morgan quickly became a strong advocate of this new theory. Not long 48. T. H. Morgan, "Darwinism in the light of Modern Criticism," Harper's Magazine 106 (1903), 476-479; see esp. p. 477. 49. Ibid., p. 357. 50. T. H. Morgan, "The Origin of Species through Selection as Opposed to Their Origin through the Appearance of Definite Variations," Pop. Sci. Monthly 67, 54-65, esp. p. 63. De Vries had originally made this point, according to Morgan. 51. De Vries' views on the exact nature of mutations seem to have changed slightly during the period 1900-1910. In the later years he seemed to see mutations as affecting only specific characters, but not necessarily the whole constitution of the organism. A thorough study of De Vries' ideas is certainly needed to understand his exact role in the history of genetic and evolutionary thought.

128

Thomas Hunt Morgan and Natural Selection after that, he also began to change his views on natural selection; and the publication in 1916 of A Critique of the Theory of Evolution52 showed that he had become a convinced Darwinian. Examination of some of the specific factors which seem to have played a part in his change of view helps resolve a major problem in understanding not only why Morgan objected to natural selection before 1910, but also how he could eventually come to accept it. THE MUTANT DROSOPHILAS Probably the most significant evidence which gave Morgan insight into the mechanism of natural selection was his own work with mutant forms of Drosophila, particularly important in emphasizing the role of small variations in evolution. In 1909, as the Drosophila work was getting under way, Morgan began to have doubts about the evolutionary role of discontinuous variations in general, and De Vries' mutations in particular. In his article of that year titled "For Darwin," he wrote: It has often been urged, and I think with much justification, that the selection of individual, or fluctuating variations could never produce anything new, since they never transgress the limits of their species, even the most rigorous selection-at least the best evidence that we have at present seems to point in this direction. But a new situation has arisen. There are variations within the limits of Linnaean species that are definite and inherited, and there is more than a suspicion that by their presence the possibility is assured of further definite variations in the same direction which may further and further transcend the limits of the first steps. If this point can be established beyond dispute, we shall have met one of the most serious criticisms of the theory of natural selection.53 The occurrence of small but definite variations which seemed to be inherited in a predictable manner gave Morgan a new insight into natural selection-an insight he had previously lacked. While he still recognized that many continuous variations are indeed not influenced by selection because they are nonheritable, he now saw that seemingly continuous variations were actually the result of several small but definitely inherited differences. These small but distinct differences from the wildtype form Morgan referred to as mutations, a borrowing of the De Vriesian term. He did not really think of the Drosophila mu52. Princeton University Press. 53. Popular Science Monthly 74 (1909),

367-380.

129

GARLAND E. ALLEN

tations as different qualitatively from those De Vries had discussed. To Morgan, variations such as "bar eye" or "dumpy wing" which he observed in the laboratory were simply less extreme examples of large-scale De Vriesian mutations.54 As he wrote explicitly in 1925: Darwin knew of cases of sudden mutation and called them sports or monstrosities. He thought that they could seldom supply materials for evolution because they changed a part so greatly as to throw the organism as a whole out of harmony with its environment. This argument for rejecting extreme or monstrous forms seems to us today as valid as it did to Darwin; but we now recognize that sports are only extreme types of mutation, and that even the smallest changes that add to or subtract from a part in the smallest measurable degree may also arise by mutation. We identify these smaller mutational changes as the most probable variants that make a theory of evolution possible both because they do transcend the original types, and because they are inherited.55 This quotation presents a clear statement of the new view which Morgan had developed of the role of small variations in evolution. Yet the point should be made that most of the variations which he observed in his laboratory cultures (such as white-eye or vestigial wing) were indeed larger and more definite than those of which Darwin and his followers had spoken. Despite the fact that he had divested De Vries' term "mutation" of some of its extremist connotations, Morgan's new use of the term after 1910 was still a far cry from what most naturalists were used to seeing as the small individual differences found in nature. It is perhaps unfortunate that Morgan chose to retain the word "mutation" in referring to those small Mendelian differences he observed, for it further confused the issue of variation by failing to distinguish between large-scale variations which De Vries considered new species and small but definite variations which occurred within the limits of an existing species. Morgan himself included a whole range of variations within the term "mutation." But many of his readers, familiar with the term only as De Vries had used it, were understandably confused about what type of variations Morgan actually thought were acted upon by natural selection. As Morgan's work with Drosophila proceeded, a vast number 54. Ibid. 55. T. H. Morgan, Evolution and Genetics (Princeton University Press, 1925), pp. 129-130.

130

Thomas Hunt Morgan and Natural Selection of mutations of all sizes and degrees appeared, ranging from an easily detectable difference, such as appearance of white-eye, to a very minute difference, such as scute abdominal and thoracic bristles ("scute" mutants have certain bristles missing). Morgan was well aware that these mutations were inherited in accordance with Mendelian laws, and that though the white-eye mutation was definite and heritable, it did not represent a species difference. By 1912 De Vries' theory of mutation was beginning to draw serious criticism, and although Morgan never really abandoned it, he was aware of the difficulty which it encountered. In 1912 he wrote that the English edition of De Vries' Mutationslehre was already behind the times,56 referring specificaly to the work of Bradley M. Davis, who had shown the possible hybrid nature of Oenothera. This new evidence indicated that what De Vries had taken to be "mutations" in the evening primrose were actually recombinations of hybrid characters from the parental forms, in agreement with Mendelian principles.57 Thus, the newest evidence on the genetics of Oenothera seemed to indicate that the sudden new species which De Vries had thought he observed were not really new species at all, a finding which undermined the whole mutation theory. It was at the same time that Morgan began observing mutations in Drosophila, providing him with substitute evidence by which the mutation theory could be modified to a more empirically accepted form. He included within the framework of the mutation theory both large- and small-scale variations.58 It is evident that by 1912 Morgan did not think of "mutations" as being large-scale jumps which created new species instantaneously. However, he still considered that discontinuous variations-that is, mutations of greater or lesser magnitude-were the raw material on which evolution acted. To Morgan, both the mutation theory and the Mendelian theory were attractive in regard to evolution because they appealed to the same scientific methodologies. Unlike other fields supporting natural selection, such as embryology, paleontology, and taxonomy, both the Mendelian and De Vriesian theories 56. T. H. Morgan, "Some Books on Evolution," Nation 95 (1912), 543544. 57. Bradley M. Davis, "Genetical Studies of Oenothera." I. "Notes on the Behavior of Certain Hybrids of Oenothera in the First Generation," Amer. Naturalist, 44 (1910-1912), 108; II. "Some Hybrids of Oenothera biennis and 0. grandiflora that Resemble 0. lamarckiana," ibid., 45, 193223; III. "Further Hybrids of Oenothera biennis and 0. grandiflora that ibid. 46, 377-427. Resemble 0. lamarckiana," 58. T. H. Morgan, "Concerning the mutation theory," Sci. Monthly 6 (1918), 385-405.

131

GARLAND E. ALLEN

were experimental and were subject to rigorous proof or disproof. Morgan wrote: "The Mutationist and the Mendelian are twin brothers because they appeal to the same scientific procedure, viz. analytical experimentation." 59 Thus, he did not have a difficult time altering De Vries' concept to fit the newer Mendelian discoveries.60 INFLUENCE OF THE IDEA OF CHARACTER RECOMBINATION Morgan's work on Drosophila led to another idea which may have been important in allowing him to apply Mendelian principles to natural selection. When he began to collect information on large numbers of mutant loci in Drosophila, linkage groups began to become evident. While this finding was in accord with the chromosome theory, which he had accepted by 1912,61 it 59. Ibid., p. 46 60. Although instrumental in the rediscovery of Mendel's work, De Vries himself was, especially between 1900 and 1910, somewhat ambiguous on the relation of the Mendelian laws to the mutatio-n theory. In Die Mutationstheorie, 2 vols. (Leipzig: Von Veit, 1901, 1903) he spoke of three types of mutation: progressive, retrogressive, and degressive (II, 637-642). Progressive mutations occur when a character emerges from its original latent condition and becomes active; retrogressive mutations are just the reverse: when a character passes from the active to latent condition. In either case, the organism has the same number of internal factors, whether active or latent. Degressive mutations, on the other hand, represent new characters which have been added to the germ plasm. In 1903 De Vries felt that Mendel's law applied only to retrogressive and degressive mutations, but not to progressive (II, 641). Later his views seemed to shift slightly, and he applied the term "progressive mutation" to what originally had been designated "degressive." See De Vries, "Luther Burbank's Ideas on Scientific Horticulture," Century Mag. (1910), pp. 674-681. Even after Morgan's work had begun to elucidate the chromosome theory, De Vries maintained that inheritance in OenotheTa mutants did not follow the Mendelian pattern. Progressive mutations (as in the newly-appearing forms of Oenothera) were not of the type to follow laws of segregation. The new "mutants" bred true, showing that they were new species rather than Mendelian hybrids (H. De Vries, "Uber monohybriden Mutationen," Biol. Zentralblatt 37 (1917), especially p. 140. Also H. De Vries, "Mutations of OenotheTa suaveolens Desf.," Genetics 3 (1917), 1-26. De Vries had did not accept the conclusions of Bradley Davis, who (1910-1912) shown rather conclusively that mutations in Oenothera were really complex recombination phenomena which could be explained in Mendelian terms (H. De Vries, " The Probable Origin of Oenothera lamarckiana Sr.," Botan. Gaz. 17 (1914), 345-361). A helpful summary of De Vries' views on evolution, mutation, and Mendelism in the period after 1910 can be found in the text of an address of 1915: "Mutations in Heredity," Rice Institute Pamphlet No. 1, pp. 339-391. 61. Linkage was first reported in a paper by Correns in 1900, although he thought only in terms of complete dominance. Incomplete linkage was first given for the sweet pea by Bateson, Punnett, and Saunders, in "Ex-

132

Thomas Hunt Morgan and Natural Selection was also noted that the linkage groups seemed to change positions between the two homologous chromosomes. Thus, new combinations of linkage groups would sometimes be observed. To explain these results, Morgan appealed to the chiasmatype theory first proposed by the Belgian cytologist, F. A. Jannsens, in 1909,62 which showed that during meiosis in certain amphibians there occurred cross-figures among the chromosomes in synapsis. Jannsens interpretated these as the result of a fusion at one point between two of the four strands of a tetrad, followed by breakage and reunion between complementary parts of the two chromatids. In relation to evolution, the work on linkage and crossing-over showed Morgan and his students that an almost infinite number of possible combinations of characters could and probably would be obtained in the course of time. For any given set of characters, crossing-over could give all the possible intermediate combinations between the two parental forms. Although H. J. Muller maintains that Morgan did not see immediately the implications of crossing-over for evolution,63 it is clear that by 1916 he understood the relationship quite thoroughly.84 It was new combinations of old characters, as well as the appearance of additional mutants and their successive recombination that provided the raw materials on which natural selection could act. INFLUENCE ON THE CONCEPT OF GENE INTERACTION The mutations which Morgan observed in Drosophila, although not so momentous as those of De Vries, were nevertheperimental Studies in the Physiology of Heredity." Reports to the Evolution Comm. Royal Soc. 3 (1905), 1-131. Called "coupling," this phenomena was discussed by Bateson and his co-workers totally outside the context of chromosomes. It was thus the work of Morgan's group after 1911 which established securely the chromosome hypothesis in relation to Mendelian characters. For a more complete discussion, see A. H. Sturtevant, A History of Genetics (New York: Harper & Row, 1965), Chap. VI, 39-44; also, L. C. Dunn, A Short History of Genetics (New York: McGraw-Hill, 1965), pp. 113-115. The history of the cytological relationship between Mendelism and the chromosomes is partly outlined in two recent works: H. J. Muller's introduction to the reprint of the first edition of E. B. Wilson's The Cell (New York: Johnson Reprints, 1966), pp. xxxi-xxxv; and M. R. Green, The Development of a Chromosmal Theory of Sex Determination (unpublished A. B. thesis, Harvard College, Cambridge, Mass., 1966). 62. F. A. Jannsens. "La Th6orie de la chiasmatypie, nouvelle interpretation des cin6ses de maturation." La Cellule 25 (1909), 387-406. See also L. C. Dunn, A Short History of Genetics, p. 114, 144. 63. Personal communication. 64. T. H. Morgan, A Critique of the Theory of Evolution, p. 162.

133

GARLAND

E.

ALLEN

less more definite than those Darwin had discussed under the heading of individualvariation.As long as variation was regarded in classical Mendelian terms, which emphasized discrete unit characters between which no intermediates could exist, it was difficultto reconcile the new laws of heredity with Darwinian theory. However, work in several fields-plant and animal breeding as well as breedingresults in Morgan'sown laboratory -began to suggest that certain hereditary pattems could be explained by assuming that one set of alleles might affect the way in which another was expressed. By 1914 or 1915, the idea of genic interactionbegan to gain considerablesupport,especially with the Morgan group.65 By arguing that selection accumu-

lates more or less modified genes, rather than actually changing genes, Muller in particularstressed the importanceof genic interaction. Here was an area where the relationship between genetics and evolutionbecame quite clear-a relationshipwhich Morganno doubt picked up from Mullerdirectly. In A Critique of the Theory of Evolution,'6 Morgan devotes considerablespace to showing how modifiergenes could account for the productionof many fine gradationsbetween the definite unit charactersof the Mendelians.Modifiergenes were shown to exist in Drosophila,and seemed to be inherited in a Mendelian fashion. The strength of expression of any character depended upon the modifying factors present in the germ cell; hence, selection could produce almost any intermediate form by reducing or increasing the number of modifier genes. Selection could stabilize the number of modifiersin a population,it could increase them to their maximum, or it could eliminate them altogether. The concept of modifying factors allowed Morganto meet the objection to natural selection deriving from Johannsen'spureline experiments of 1903.67 Johannsen's work had suggested that selection could only separate out the various pure lines in a heterogeneouspopulation, but that it could not produce anything new. Because his results were experimentallyverffied and quantitative, they had a particularly strong effect on experimentalists such as Morgan.68Within the concept of modifier 65. See A. H. Sturtevant, A History of Genetics (New York: Harper & Row, 1965), pp. 60-61. 66. Page 165. 67. W. Johannsen. Uber Erblichkeit in Populationen und in reinen Linien (Jena; Gustav Fisher, 1903). 68. This can be observed by reading through the pages of the American Naturalist, or other publications dealing especialy with evolution and heredity during this period. See also Arne Muntzing, "Genetics and Plant Breeding," in Genetics in the Twentieth Century, ed. L. C. Dunn (New York: Macmillan, 1950), pp. 473-492. In the same volume, see H. J.

134

Thomas Hunt Morgan and Natural Selection genes, however, limits to the effects of selection were attributable to the fact that modifying factors had been accumulated to their maximum (for the population) or reduced to zero. Selection had then to await more mutations or the appearance of more modifiers before it could produce further change. Johannsen's experiments did not really contradict the idea of natural selection in the way Morgan had originally (1905) thought. It is curious in this regard that Morgan did not mention in his 1916 book on evolution the work of H. Nilsson-Ehle, which had been published in 1908 and 1909.69 Particularly interested in problems of practical breeding, Nilsson-Ehle had undertaken a factorial (Mendelian) analysis of the inheritance of "quantitative characters" (that is, measurable characters) in wheat. Studying grain color in several races, he had shown that the color of any particular group of seeds could be explained as resulting from three different gene pairs, each pair contributing in an additive fashion to the phenotype. In Europe this idea became known as "polymeric inheritance," and in the United States as the "multiple-factor theory."70 According to Dunn, it was Nilsson-Ehle whose work showed most conclusively the validity of Mendelian interpretation for all cases of quantitatively (as opposed to qualitatively) varying characters. Further work was carried out along these lines in America by E. M. East and H. K. Hayes in 1911.71 Morgan does cite the latter work in 1916, so that his omission of Nilsson-Ehle is perhaps more a matter of citation than unfamiliarity. Nevertheless, this omission does suggest that Morgan was less adept in making some of the conceptual bridges between specific work in heredity and the concept of natural selection than his own writings suggest. Exactly how much of the connection between work on multiple factors (or modifiers) and selection of slight individual differences Morgan himself made, and how much he derived from those around him, is of course impossible to say. But Muller, for instance, was especially interMuller, "1The Development of the Gene Theory," pp. 77-99. For two examples of work influenced by Johannsen's experiments, see H. S. Jennings, "Heredity and Variation in the Simplest Organisms," Amer. Naturalist 43 (1909), 321-337; and Raymond Pearl, "Seventeen Years' Selection of a Character Showing Sex-linked Mendelian Inheritance," Amer. Naturalist 49 (1915), 595-608. 69. H. Nilsson-Ehle, "Kreuzungsuntersuchungen an Hafer und Weizen." Lunds Univ. Arsskr. 5 (1909), 1-122. This work was known to Morgan by 1914 or 1915, as he cites it in The Mechanism of Mendelian Heredity (New York: Henry Holt, 1915), pp. 178-179. 70. L. C. Dunn, A Short History of Genetics, p. 100. 71. E. M. East and H. K. Hayes, "Inheritance in Maize." Conn. Agr. Exp. Sta. Bull. 167 (1911), 1-141. Cited from Dunn, A Short History of Genetics, p. 100.

135

GARLAND E. ALLEN

ested in the multiple-factor hypothesis, as evidenced by his strong criticism of the work of Castle and Phillips. Muller records that he and Sturtevant talked with Morgan on this and other evolutionarymatters at considerablelength on many occassions.72 It is quite possible that in this, as in other instances, Morgan'sviewpoint was strongly influenced by his remarkable group of students. It was they who often saw a relationship which Morgan,by virtue of his particularprejudices,was unable to perceive directly. THE ROLEOF MORGAN'SSTUDENTS Morgan'sstudents, particularlyH. J. Muller and A. H. Sturtevant, had considerable influence in bringing about a revision in their teacher's thinking on evolution. Muller had entered ColumbiaCollegeas a freshman in 1906 and Sturtevantin 1907. In 1908 Muller had taken a course in general biology under E. B. Wilson in which R. H. Lock's Variation, Heredity and Evolution,judged by many at the time to be the most up-to-date work on evolution and heredity,73 was assigned reading. In Variation, Heredity and Evolution, Lock tried to show how Mendelian genetics and Darwinian natural selection fitted together to form a complete mechanism for evolution. From the outset, then, Muller and other students in Wilson's course were exposed to the idea that there was no fundamental conflict between the theory of particulate inheritance and the theory of natural selection. Muller reports many long discussions with Morgan in the years before 1912 on the subject of natural selection.74 Accord-

ing to him, Morganwas stubborn;he would not accept the idea that natural selection could create new species, and continually regarded it as he had done in Evolution and Adaption, as a purely negative force. As Muller put it: "All of us [himself, Sturtevant,and Bridges]arguedwith Morganon that. . . Morgan would come back and back... it seemed to us as if he somehow couldn't understand natural selection. He had a mental block which was so common in those days."75 Muller feels that Sturtevant probably had the most influence on Morgan along these lines, as he was Morgan'sspecial student. Continual talks and debates with Morganin a half-joking,half-seriousway, may well have been an important factor in changing his mind, al72. Personal communication. 73. R. H. Lock, Variations, Heredity and Evolution (New York: E. P. Dutton, 1907). 74. Personal communication. 75. Ibid.

136

Thomas Hunt Morgan and Natural Selection though this could never have happened had the Drosophila work not taken the course which it did after 1910. Yet, even then, Muller points out that Morgan was not quick to pick up the immediate relationship between recombination, modifying genes, and so on and natural selection. He seems to have seen it only after continual debate, after it was suggested to him by more than one person, and after it had had some time to settle into his thoughts. The same impression is given by Sturtevant, who feels that Morgan was such an empiricist that he always remained skeptical of any broad theoretical interpretation unless it lent itself to experimental testing.76 THE EXPERIMENTALIST-NATURALIST DICHOTOMY Experimentalists such as Morgan were hostile to Darwinian natural selection for several reasons which lie beneath the surface of objections to the theory itself and to its mode of operation. In his reactions to the theories of Darwin, Weismann, or the neo-Lamarckians, Morgan displayed his impatience with a tradition in biology which he felt was outmoded. The day of the sweeping speculation, the generalized theory based on only a few facts, he felt, was over. Where had this method brought biologists but to monumental paper castles such as Weismann's hierarchial system? Darwin's work had been the crowning achievement, but he had usually stuck close to the facts. His followers, however, tried to carry the method beyond its limitations, attempting to develop theories without facts to back them up. Worst of all, they proposed theories which could not be tested-in Morgan's mind the greatest sin a scientist could commit. As part of a new generation of experimentally trained biologists, Morgan was reacting to both the ideas and the methodologies of his teachers. In a personal way this is shown in his view of his own graduate professor at Johns Hopkins, W. K. Brooks. According to Sturtevant,77 Morgan felt that Brooks did far too much philosophizing and far too little science, and although Brooks had a reputation as a great teacher78 Morgan was singularly unimpressed with his attempts to formulate speculative theories of heredity or evolution. Morgan's reaction to the naturalist methodologies was widespread among the younger generation of experimentalists such as Conklin, Harrison, Bateson, and Loeb. The prominent neo-Darwinians, paleontolo76. A. H. Sturtevant, personal interview, July 1965. 77. Personal communication. 78. He also taught, as graduate students or research fellows, Bateson, E. G. Conklin, Ross Harrison, and E. B. Wilson.

William

137

GARLAND E. ALLEN

gists, and neo-Lamarckians in 1900 were all members of an older generation which had followed in the wake of the Darwinian success. According to Morgan and his contemporaries, they had carried Darwin's methods as far as, or even further, than was permissible. The remaining problems of evolution would have to be solved by a new approach, that of rigorous experimentation. Around the turn of the century a number of advances in experimental biology began to give those fields greater prestige than they had previously enjoyed. Developments in experimental embryology, in cytology, and the rediscovery of Mendel, all began to attract students away from the descriptive fields and into laboratory biology. After 1900 the growth of Mendelism, and studies of chromosome behavior in particular, showed what experimental biology could do. With the joining of Mendel's laws and the chromosome theory by Morgan and his associates between 1910 and 1915, the experimentalists had a conceptual scheme to their credit which they considered equal, if not superior to, Darwinian natural selection. Many experimentalists after 1905 were sometimes arrogant in their attitudes toward the naturalists and their methods. Flushed with success, they challenged their elders on the very grounds which the latter had come to take for granted: the efficacy of natural selection, and the descriptive method as a means of analysis in biology. The old problem which Darwin and the naturalists had not been able to solve-the origin of variations, and proof of the type of variations on which selection acted-was answered by the experimental method. In defense, the naturalists either ignored or criticized the new Mendelian principles. In many cases they rejected Mendelism as merely another of the several current theories of discontinuous variation. Their reaction was defensive, however, for they could not always offer viable alternatives to explain how variations had originated. In this sense they had not progressed beyond Darwin. Yet a few saw the objections of the experimentalists in a clear light. For example, W. K. Gregory of the American Museum of Natural History challenged the experimentalists to examine the naturalists' collections to see some of the problems confronted in taxonomic, paleontological, or evolutionary work.79 Gregory's point was well made and applied to both sides of the controversy. Too frequently, the experimentalist was unfamiliar with what the naturalist's method really was, and, conversely, the naturalists did not understand the value of 79. W. K. Gregory, "Genetics versus Paleontology," 51 (1914), 626-627.

138

Amer. Naturalist

Thomas Hunt Morgan and Natural Selection laboratory work in helping to solve their own problems. The difference of opinion about the exact mode of evolution so apparent in those early years was eventually resolved in the late 1920's by approaches developed through population genetics. After the definitive work of Haldane, Wright, and Fisher, the genetical theory of natural selection was established in a quantitative and rigorous way. But the rift between the naturalists and the experimentalists remained. In conclusion, two questions should be posed concerning the specific study of Morgan's evolutionary views. First, what did Morgan's discussion of the factors involved in evolution, and especially in natural selection, contribute to his understanding of the evolutionary process? It is evident that his work on evolution per se did not increase his understanding of the mechanism of natural selection in any substantial way. It was really his own work in genetics which, in large part, helped him to understand how selection could operate to bring about the gradual production of new species. Morgan's unawareness of many areas of biology directly related to evolution, such as systematics or field collecting, was one of the factors responsible for his failure to understand selection for such a long period of time. A second question is in what way his work in the area of evolution contributed to our present-day understanding of the process of selection. In actual substance the answer is: virtually nothing. He did not add any significant ideas to the field of evolutionary theory; and, indeed, his constant failure to appreciate the importance of such concepts as isolation or speciation may have been something of a hindrance. Historically, however, Morgan's work on evolution was more important. He was well respected, especially among experimentalists, as both an embryologist and, after 1910, a geneticist. His advocacy of Mendelian laws as complementary to natural selection in 1916 did a great deal to bring experimentalists into the Darwinian camp. Although not a particularly original thinker in this area, Morgan's many writings and lectures on evolution helped to lead otherwise skeptical biologists to see the way in which Mendelian genetics could be applied to a fuller understanding of the theory of natural selection. ACKNOWLEDGMENT Preparation of this paper was financed in part by NSF Grant No. GS-1832.

139

First Steps in ClaudeBernard's Discoveryof the GlycogenicFunctionof the Liver M. D. GRMEK Centre National de la Recherche Scientifique,

Paris

The Archives of the College de France in Paris are in possession of a very large and impressive collection of the notebooks, laboratory journals, and other scientific manuscripts of Claude Bernard. These papers are now classified and available for scientific research.' Some notebooks and papers give significant documentary information on Bernard's philosophical background and his position between the materialistic doctrine and the vitalistic conception of life.2 For the historian of science, however, more interesting perhaps are Bernard's laboratory journals and day-by-day reflections on physiological problems.3 In his famous Introduction4 he accords to his own discoveries 1. M. D. Grmek, Catalogue des manuscripts de Claude Bernard. Avec la bibliographie de ses travaux imprimgs et des etudes sur son oeuvre. Avantpropos par M. Bataillon et E. Wolff. Introduction par L. Delhoume et P. Huard. Paris: Coll6ge de France and Masson & Co., 1967, 419 pp. 2. Cf. C. Bernard, Philosophie. Manuscrit inedit. Texte pr6sent6 par Jacques Chevalier. Paris: Hatier-Boivin, 1937, XIV, 63 pp. See also M. D. Grmek, "Quelques notes intimes de Claude Bernard," Arch. Intern. Hist. Sci., 1963, 16, pp. 339-352. Many important materials are still unpublished. Thus, for example, in the first draft of his acceptance address upon his election to the Acaddmie Francaise, Bernard expresses some very interesting thoughts which in fact were not intended to be disclosed publicly, and which subsequently were omitted from the final lecture. 3. Only a small part of these journals is published. The so-called "Cahier rouge" represents a curious mixture of "philosophical" and technical notes; cf. C. Bernard, Cahier de notes 1850-1860. Pr6sent6 et comment6 par M. D. Grmek; preface de R. Courrier. Paris: Gallimard, 1965, 315 pp. In one of my recent publications I have used Bernard's manuscripts for a detailed analysis of the genesis of an important scientific concept with complicated "philosophical" involvements; see "Evolution des conceptions de Claude Bernard sur le milieu int6rieur," Philosophie et m6thodologie scientifique de Claude Bernard. Paris: Masson & Cie, 1967, pp. 117-150. 4. C. Bernard, Introduction a 1'etude de la m6decine exp&imentale. Paris: Bailli6re, 1865, 400 pp. An Introduction to the Study of Experimental Medicine. Translated by H. Copley Greene, with an Introduction by L. J. Henderson and a Foreword by I. B. Cohen. New York: Dover, 1957, 226 pp.

141

M. D. GRMEK

the dignity of paradigms. Thus a detailed study of all the steps in his creative activity is a necessary condition for acceptance of his findings as epistemological examples. An analysis of his laboratory journals reveals in many cases an important historical inconsistency. On the one side, his original manuscripts suggest a very complicated gradual development of his discoveries, while on the other side, his published works show a tendency toward a secondary rationalization, that is, a very strong post hoc simplification of facts. If the examples quoted in his Introduction are all logically consistent, many of them are chronologically incorrect and simplified to the point that some very important steps are masked.5 I will try to illustrate this point by an example which seems minor at first glance, but which actually is of extremely great importance in Bemard's research work. The unexpected result of one experiment changed the whole direction of his investigation of the destination of sugar in animal organisms. When he began his experiments with sugar, Bemard shared the view of Dumas and Boussingault that it was formed by green plants, introduced in animals by alimentation, and destroyed in them by a special process of combustion.6 Animals were supposedly able only to break down sugar supplied by vegetables. Bernard accepted Liebig's opinion that sugar was the fuel of life, and he believed that the action of combustion took place either in the lungs (Lavoisier's initial hypothesis) or in the general capillaries (hypothesis of Lagrange and Hassenfratz). In 1843 Bemard discovered that an aninal organism could directly utilize only sugars of the so-called "second species" (for example, grape sugar) and that sugars of the "first species" (cane sugar), when injected into the blood of animals even in very weak doses, passed into the urine. He noted, too, that gastric juice could transform cane sugar into a form capable of assimilation, that is, of destruction in the animal organism.7 The next step in Bernard's work was later summarized by him in the following way: Thereupon I wished to learn in what organ the nutritive sugar disappeared, and I conceived the hypothesis that sugar 5. Cf. M. D. Grmek, "Examen critique de la genese d'une grande d6couverte: la piquire diabetique de Claude Bernard." Clio Medica, 1 (1966), 341-350. 6. J. B. Dumas and J. B. Boussingault, Essai de statique chimique des Chemical etres organis6s. 3d ed. Paris: Fortin, Masson et Cie, 1844.-The and Physiological Balance of Organic Nature. London: J.-B. Bailli&re, 1844; New York: Saxton, 1844. 7. C. Bernard, Du suc gastrique et de son r61e dans la nutrition. These pour le doctorat en medecine. Paris: Rignoux, 1843, 34 pp.

142

The Glycogenic Function of the Liver introduced into the blood through nutrition might be destroyed in the lungs or in the general capillaries. The theory, indeed, which then prevailed and which was naturally my proper starting point, assumed that the sugar present in animals came exclusively from foods, and that it was destroyed in animal organisms by the phenomena of combustion, i.e., of respiration. Thus sugar had gained the name of respiratory nutriment. But I was immediately led to see that the theory about the origin of sugar in animals, which served me as a starting point, was false. As a result of the experiments which I shall describe further on, I was not indeed led to find an organ for destroying sugar, but, on the contrary, I discovered an organ for making it, and I found that all animal blood contains sugar even when they do not eat it. So I noted a new fact, unforeseen in theory, which men had not noticed, doubtless because they were under the influence of contrary theories which they had too confidently accepted. I therefore abandoned my hypothesis on the spot, so as to pursue the unexpected result which has since become the fertile origin of a new path for investigation and a mine of discoveries that is not yet exhausted.8 This famous text emphasizes an invaluable general recommendation. But in some details it seems quite vague, even to the point of obscurity. With his sentence "But I was immediately led to see . . ." Bernard gives the impression, and wrongly so, that he changed his mind very quickly after the beginning of his experiments on sugar destruction in an animal organism. And what is more important, he does not really explain abandons the prevalent why-at which concrete occasion-he theory. Curiously enough, he never really elucidated this point,9 8. C. Bernard, An Introduction. Transl. by H. C. Greene, New York: Dover, 1957, pp. 163-164. 9. It is only in his thesis for the doctorate in science that Claude Bernard gives some valuable information concerning the first steps of his discovery of the glycogenic function of the liver. He states that his aim was to follow closely the sugar which was absorbed from the food. He wanted to know if it was destroyed in traversing the liver; then what happened after the passage of the blood stream with sugar through the lungs, and so on. For this purpose a dog which had been fed on carbohydrate food for seven days was killed during the digestion, and Bernard was able to show that the blood of the hepatic veins, where they join the inferior vena cava, contained a large amount of glucose. This seemed to be an experimental proof that the liver did not destroy the sugar. As a counter-proof, Claude Bernard performed a similar experiment on a dog which had been fed exclusively on meat, and to his surprise found again that the blood of the hepatic veins contained a considerable amount of sugar, although there was no sugar in the intestines. He found also that

143

M. D. GRMEK

and in his fundamental publication on the discovery of the glycogenic function of the liver, he presented his experiments without chronological order, without dates, and following a logical development completely independent of the historical linkage of his experiments and the real evolution of his thought.10 He was not "immediately led to see that the theory about the origin of sugar in animals . . . was false," because he started his experiments in 1843, increased their number and perfected them from 1844 to 1847, and finally understood that he was on the wrong track in August 1848. His notebooks contain descriptions of large numbers of experiments concerning the search for the location and mode of destruction of carbohydrates after ingestion, intravenous injection or other introduction into an animal organism.1' These experiments have never been published because Bernard was fully aware that they brought forward nothing new and represented only a failure. The positive side of all of this lengthy previous work is that Bernard elaborated-with his friend, the young chemist Charles-Louis the chemical testing of sugar, Barreswil (1817-1870)-on further that he understood better the first phases of the digestion of starch, and that he observed the influence of the nervous system on the presence of sugar in blood and urine. Certainly, by his long and numerous experiments he became sensitive to all the possible physiological implications of the "the blood of the portal vein contains no sugar before it enters the liver, whereas on leaving that organ the same blood contains considerable amounts of glucose." Cf. C. Bernard, Recherches sur une nouvelle fonc. tion du foie considgr6 comme organe producteur de mati.6re sucr6e chez l'homme et les animaux. These pour le grade de docteur Als-sciences naturelles. Paris: Martinet, 1853, 97 pp.-In spite of some simplifications and errors (for example, the omission of the fact that in the experiment with the dog on a meat diet, Bemard discovered sugar in the blood of the portal vein), this story is basically correct, especially in its emphasis on Bernard's astonishment after the unexpected results of the counter-proof experiment. The best historical presentations in English of Bernard's discovery of the glycogenic function of liver follow the text of his thesis. For example, D. Wright Wilson, "Claude Bernard," Pop. Sci. Monthly, 84 (1917), 567-578; F. G. Young, "Claude Bernard and the Theory of the Glycogenic Function of the Liver," Ann. Sci., 2 (1937), 47-83; J. M. D. Olmsted, Claude Bernard, Physiologist (New York: Harper, 1938, 272 pp.). In most other publications, the story is very distorted. 10. C. Bernard, "De l'origine du sucre dans l'6conomie animale," Arch. g6n. M6d., 18, 4th ser. (1848), 303-319. Published also in M6m. Soc. Biol., 1 (1849), 121-133. An English translation ("The Origin of Sugar in the Animal Body") is published in Kelly's Medical Classics (1939), III, 567-580. 11. Claude Bernard's unpublished papers in the Coll6ge de France, Ms. 7b, 7c, and others.

144

The Glycogenic Function of the Liver presence or absence of sugar in various parts of the animal circulatory system. Beginning in 1845 Bernard became interested in the clinical problems of diabetes. He observed patients'2 and formulated his first theory of the pathogenesis of this disease.1I According to Bernard's first opinion, diabetes is "a nervous affection of the lungs." For the modem reader this theory is very surprising, but actually it was a very logical conclusion from these four premises: 1) sugar cannot be synthetized in the animal body; 2) it is normally destroyed in the lungs; 3) the principal symptom of diabetes is the presence of undestroyed sugar in the urine; and 4) the nervous system controls the breakdown of sugar in the lungs. Claude Bernard discovered that after cutting the pneumogastric nerves in rabbits the pulmonary functions are affected and glucose passes undestroyed into the urine.14 One important problem was whether or not the blood of diabetic patients actually contained sugar. Thomas Willis was the first to believe it. In the eighteenth century, Dobson, Cawley, and Rollo tried to extract sugar or sugar-like substances from the serum of diabetic patients. They wished to demonstrate that glycosuria is merely a sequence in glycemia. But none of these attempts produced any definite conclusion. The sweet taste of blood was not sufficient proof, and chemical analysis by alcoholic fermentation gave generally negative results. Thus P. F. Nicolas and V. Gueudeville, Soubeiran, Vauquelin, and other authorities on this subject at the beginning of the nineteenth century were not ready to accept the theory of diabetic glycemia. Their negative results can probably be explained by the fact that the analyses were performed on old blood, after glycolysis. In 1835 an Italian chemist, F. Ambrosioni, was the first to give definite proof of the presence of sugar in the blood of a diabetic person. His demonstration was based on the alcoholic fermentation by yeast of blood sugar.15 Even before it was definitely proved that sugar could be found in the blood of persons with diabetes, it was known that this substance could be present in the blood of healthy animals, at least in some animals under special conditions. In 1826 F. Tiedemann and L. Gmelin demonstrated the presence of fer12. He worked in Rayer's and Andral's departments in the well-known hospital, La Charit6, in Paris; cf. Ms. 7b, pp. 246 and 249-250; Ms. 15i and Fasc. 25b, f. 370. 13. Ms. 7b, p. 133. 14. Ms. 7b, p. 130. 15. Ann. univ. di med. e chir., 74 (1835), 160.-See introductory chapters in R. Lkpine, Le diabeRte sucr6 (Paris: Alcan, 1909).

145

M. D. GRMEK

mentable glucose in the intestines and venous blood of healthy dogs after ingestion of starch.'8 In England, MacGregor (1837) confirmed the observation of Ambrosioni, and Thomson, a chemist of Glasgow, found (1845) that chicken blood normally contained a certain amount of sugar.-T In France, F. Magendie discovered, independently of the aforementioned authors, that sugar can be found in the blood of normal rabbits and dogs after they had been fed on starch or potatoes.'8 After Magendie's experiments, performed during his lectures at the College de France in 1846, the majority of physiologists and physicians agreed in supposing an alimentary origin of sugar and considering glycemia as a physiological a phenomenon compatible with health, but inconstant-being result of ingestion of special kinds of food. Thus the presence of sugar in blood was considered to be either a pathological or an accidental fact. It was Bernard who discovered that glycemia was a normal and constant phenomenon, largely independent of alimentation.'9 One unpublished manuscript permits us to have a real understanding of how this discovery occurred: this is Bernard's laboratory journal Ms. 7c, compiled from 1846 to 1848. The beginning experiments are of no interest, because, having taken a wrong turn, Bemnardwas unable to progress. Until May 1848 he attempted to answer a badly formulated question. Yet at this time he believed that sugar must be destroyed somewhere within the organism. His main attention was evidently directed toward the lungs, and in the last week of May he observed what happened when grape sugar was exposed in vitro to the pulmonary tissue of freshly killed animals. After ten to twelve hours the sugar disappeared, and Bernard concluded that the lungs contained a special ferment for the destruction of glucose.20 But following faithfully his method of experimental 16. F. Tiedemann and L. Gmelin, Die Verdauung nach Versuchen, vol. I, Heidelberg and Leipzig: Groos, 1826. 17. LUpine, Le diabate sucrT. See also Berard's historical sketch in Lepons sUT Le diabte (Paris: Baillire, 1877), pp. 142-161. 18. C. R. Acad. Sci., 23 (1846), 189. 19. In one of his last works, at the very end of his life, Bernard states proudly: "Je montrai .. . que la glyc6mie est ind6pendante de l'alirnentation; qu'elle se rencontre chez l'homme et chez les animaux nourris de viande ou sounis k l'abstinence. Je prouvai que la pr6sence du sucre dans le sang est un fait normal coincidant toujours avec l'6tat de sant6 et ne disparaissant que lorsque la nutrition Ftait arret6e. De sorte qu'au lieu d'admettre, comme mes pr6d(cesseurs, que la glyc6mie fut un fait pathologique ou accidentel, je fis voir que la proposition contraire 6tait vraie, et que c'6tait l'absence de sucre dans le sang qui constituait le v6ritable fait anormal" (Lefons sUT le diab&te, 1877, pp. 127-128). 20. Ms. 7c, p. 308.

146

The Glycogenic Function of the Liver research, he proceeded to a counter-proof. This he did by mixing sugar with tissues of liver and other organs. He obtained positive results, and the problem became even more obscure than before he started his experiments. On the last day of May, he injected one gram of grape sugar into the jugular vein of a dog, extracting at the same time blood from the carotid artery. This blood contained a large amount of sugar. The conclusion was evident: glucose is not destroyed in the lungs, because blood must pass by these organs in order to move from the jugular vein to the carotid artery.2' Bernard guessed that perhaps the combustion of carbohydrates took place in the general capillaries or in the liver.22 Numerous experiments, carefully executed by Bernard during June and July of 1848, were strongly opposed to the theory of pulmonary combustion. Grape sugar was injected into the jugular veins, or starch introduced into the stomachs of rabbits and dogs. Then either blood was taken from various parts of the animals or they were killed and blood extracted separately from different organs. Sugar was present in all samples. Bernard was unable to find a rule for its quantitative distribution. His friend Quevenne, pharmacist at the Ho6pitalde la Charite, extracted and purified blood sugar from a diabetic patient, and Bernard showed that in physiological experiments there is no difference between grape sugar and the sugar of diabetics.23 In July 1848 Claude Bernard discovered some important facts: 1) the transformation of cane sugar into grape sugar by the action of gastric juice is not performed by gastric acid, but by some special "organic matter"24; and 2) sugar is always present in the vitreous humor of the eye of a dog and also in the white of a chicken egg.25 This last finding was not immediately published, and when in March 1849 an Irish physician named Aldridge discovered this fact independently of Bernard, he claimed priority.26 Preserved notebooks provide proof that Bernard's claim was well founded. This discovery is not without historical significance, because it invalidated the older demonstration of the presence of sugar in the blood of diabetic persons. Actually Ambrosioni and MacGregor added egg white to blood before testing its sugar content. 21. Ms. 7c, p. 311. 22. He presumed that "chez un animal en digestion d'amidon, il devra y avoir du sucre dans le sang de la veine porte arteriel et pas dans le sang veineux de retour" (Ms. 7c, p. 311; note dated May 31, 1848). 23. Ms. 7c, p. 307. 24. Ms. 7c, p. 312. 25. Ms. 7c, pp. 354, 358, and 363-366. 26. C. Bernard and Ch. Barreswil, "Du sucre dans l'oeuf," C. R. Soc. Biol., 1 (1849), 64.

147

M. D. GRMEK

Bernard observed that Barreswil'scopper reagent did not react well with sugar in the presence of fibrin. He imagined a new theory of the pathogenesis of diabetes. Thus sugar was supposed to be destroyedin blood, fibrin having some important function in this destruction.And diabetes was nothing more than a stopping of this destructive process probablyby some chemical disorders involved in the synthesis and distributionof fibrin.27 This original point of view gave Bernard the possibility of foreseeing new experiments. Analyzing the blood in the vessels before and after each single organ, Bernardwished to eliminate, step by step, the ancient theory that sugar combustionis located in a particular area of the organism. The first experiments seemed to confirm his new working hypothesis. In dogs fed on a carbohydrate-richdiet, the blood from the hepatic veins and vena cava contained sugar; thus it was not destroyed in the liver. Sugar was also present in both ventricles of the heart, meaning that it had not been destroyedby the lungs. From many laboratorynotes it is clear that Bernardattained these results and was very happy that they correspondedto prediction. But as in all cases he wished to assure the results by counter-proofs.One dog was submitted to a noncarbohydrate diet, then killed by section of the spinal bulb. Blood was taken from the portalvein, from both ventricles, and from a peripheral artery.The results were completelyunexpected, astonishing and puzzling. The blood of the portal system contained enormous quantities of sugar; the blood from the heart contained sugar, but in a small amount; and the arterialblood showed only traces. Chyle had no sugar whatsoever. It is perhapsuseful to publish the exact text of Bernard'slaboratorynotes concerning this crucial experiment: Rue Dauphine 32. Aou't 1848. Exp6riences sur le sucre dans le sang. Mouvements p&istaltiques. Sur un chien d jeun

depuis 6 jours de tout aliment solide et n'ayant bu que de l'eau, on retire du sang par la veine jugulaire; le s6rum de ce sang tout frais, aussit8t apres sa coagulation, donne par le liquide bleu des traces de reduction. Le lendemain il n'en donne plus. Une partie de ce mAmes6rum trait6 par l'alcool, puis 6vapore et repris par l'eau, mis avec la levure de biere ne donne que quelquesbulles de gaz de sorte que la fermentation est A peine sensible. Infaut dire qu'il y avait fort peu de serum, environ 15 grammes,et que la quantite de sucre devait etre fort peu de chose, s'il y en avait. Sur le meme chien, dans une autre circonstance, lorsqu'il 27. Ms. 7c, p. 338.

148

The Glycogenic Function of the Liver n'dtait A jeun que depuis 2 jours, le s6rum frais donne 6galement de la reduction par le tartrate de cuivre. Sur le meme chien remis i la nourriture de la viande et ayant mange pendant 8 jours consecutffs uniquement des debris de viande crue pris chez le boucher; il est tue par la section du bulbe pour recueillir les sangs differents. Aussit6t le bulbe coupe, les mouvements respiratoires furent comme -. l'ordinaire completement arretes. L'oeil est rest6 sensible d'un c8te et pas de l'autre. J'ouvris aussitot le ventre, le coeur allant encore. Les chyliferes 6taient pleins de chyle blanc, les intestins contenus dans le ventre par la position de l'animal qui &tait couch6 sur le dos n'etaient le siege d'aucune contraction peristaltique. Alors je comprimai l'aorte dans la poitrine entre les deux doigts, et aussit6t les contractions p6nrstaltiques eclaterent avec imp6etuosite pour ne plus cesser. Seulement elles me semblaient etre un peu moins violentes quand je cessais de la comprimer, pour augmenter quand je la reprenais. Du reste, cela a ete de peu de duree, car bientot les mouvements du coeur ont cesse. Ce qu'il y a eu de saillant, c'est le depart des contractions peristaltiques intestinales au moment oiu j'ai comprim6 l'aorte thoracique. Extraction des sangs. On a retir6 d part: 10 du sang de la veine porte i son entr6e dans le foie; 20 du sang du coeur dans les ventricules droit et gauche; 30 du sang provenant de la plaie faite 'a la nuque pour couper le bulbe. Ces trois sangs ont ete laiss6s en repos pour les laisser coaguler. Tous se sont coagules au bout de quelques instants en presentant un serum blanchatre lactescent. (Le sang de la veine presentait egalement cet aspect; le chyle y avait-il penetre ou bien etait-ce un retour par le sang veineux?) Du chyle blanc a egalement 6te extrait du canal thoracique. II s'est coagule au bout de quelques instants. On a essaye avec le tartrate de cuivre les 3 serums tout frais et le serum de chyle 6galement tout frais. 10 Le chyle de la veine porte donne une reduction 6norme en le traitant directement. En precipitant par le sulfate de soude sec de fagon a obtenir une liqueur incolore, la reduction se falt 6galement tres abondamment. 20 Le chyle du coeur trait6 directement par le tartrate de cuivre donne une reduction tr6s nette mais moins abondante que le sang de la veine porte. Traite egalement par le sulfate de soude, la reduction est tres nette mais toujours moins abondante que dans le sang de la veine porte. 30 Le serum du sang de la nuque trait6 directement par le tartrate de cuivre donne A peine des traces de reduction. Apr6s traitement par le sulfate de soude, la reduction est tou-

149

M. D. GRMEK

jours 6quivoque. 40 Le serum du chyle trait' directement par le tartrate de cuivre ne donne pas de reduction sensible. Comment se fait-il donc qu'il y a du sucre (ou une matiere qui r6duit) dans le sang de la veine porte? On recherche dans l'intestin: 1? Le liquide de l'estomac qui contenait de la viande en voie de digestion ne r6duisait aucunement le tartrate de cuivre. 20 Le liquide intestinal bilieux ne r6duisait pas du tout le sel de cuivre. 30 L'urine traitee pr6alablement par le sulfate de soude ne r6duit aucunement le sel du cuivre. Du serum du coeur conserve jusqu'au lendemain contenait encore du sucre, c'est-a-dire r6duisait le tartrate. Cette experience est fort singuliere. C'est d n'y rien comprendre. II se formerait du sucre dans la veine porte. Par quel organe, par quel mecanisme? II faudra prendre le sang de la veine porte d'un chien a jeun et voir si l'on y trouvera cette matiere qui reduit. S'il se forme du sucre dans une autre alimentation que celle de l'amidon, la question des diab6tiques est singulierement compliqu6e. II faudra voir si cette mati6re reduisante (sucre ou autre) disparaitra assez vite car le sang du coeur en contenait moins et le sang de la nuque d'une maniere tr6s 6quivoque. Quel est donc l'organe qui formerait ce sucre ou cette matiere r6duisante.28 Where did the sugar come from in an animal without an alimentary supply of carbohydrates? Bemard wrote in his journal, as if crying in surprise: "It is absolutely incomprehensible!" This experiment was done in the laboratory of Theophile-Jules Pelouze (1808-1867), a famous chemist and Bemard's mentor. The exact date of the experiment is not stated, but its location in the laboratory journal places it between the 10th and the 17th of August 1848, probably closer to the latter date. It is significant that iM this case Bernard conducted at the same time and on the same animal two different experiments, one conceming the regulation of conditions of peristaltic movements in the intestines, and the other concerning the metabolism of sugar. It is clear enough that the second part was considered only as a routine counter-proof of previous experiments. Bernard's notes express his astonishment. Actually, the discovered facts completely contradicted his working hypothesis and the generally accepted ideas on animal physiology. The presence of sugar in the blood of an animal without an alimentary supply of carbohydrates was such an incredible finding 28. Ms. 7c, pp. 379-382.

150

The Glycogenic Function of the Liver that Bemard, as we see from his journal, doubted the specificity of the copper reactive. This "reductive matter"-was it really sugar? He decided to repeat the experiment, using other methods of chemical analysis. Bernard understood immediately that the presence of sugar in the portal vein of a dog without sugar in the chyle had farreaching consequences, and that it would completely change existing theories of the pathogenesis of diabetes. He decided to repeat the experiment on a starving animal, and he posed the crucial questions: 1) where and by what mechanism was sugar formed in animals? 2) which animal organ performed this "vegetable" function? Within a few days the basic fact, namely, the presence of non-alimentary sugar in mammalian blood, was confirmed by new experiments. On August 21 Bermard obtained positive evidence of the presence of glucose in the blood of a dog fed on lard and tnrpeexclusively. Thus he no longer hesitated to affirm that "there is a formation of sugar at the expense of fat."29 But the great innovation of this expeniment lies in the results of chemical examination of tissues taken from various abdominal organs. I cannot resist the temptation to quote the findings in Bernard's own words: J'ai pris les tissues 10 de la rate, 20 des ganglions m6senteriques, 30 du foie, et j'y ai recherche le sucre. Les tissus des ganglions et de la rate ne montraient pas nettement du sucre, mais le tissu du foie en contenait enorm6ment.30 There was no sugar either in the spleen or in the lymphatic ganglia, but the substance appeared in "enormous" quantity in the liver. What a surprising resultl Many questions assailed Bernard's mind. Was the presence of sugar in liver tissues a physiological phenomenon? Was it exclusive to the liver or was it a property shared by other organs? Was it a characteristic of dog's liver only, or was it common to all animals? After a few days of feverish research, Bernard had found the answers to all these questions. Some extracts from his laboratory journal give good evidence of his investigations: Le 22 aoiut 1848. J'ai f ait acheter chez un tripier du foie de veau et de boeuf. J'y ai trouve enormement de sucre dans l'un et dans l'autre par le reactif et par la fermentation. Le 23 aou't 1848. J'ai pris a l'Hopital de la Charit6 3 morceaux de foie. 10 Un morceau de foie granuleux, chez un veillard, mort tres amaigri de je ne sais quoi; je n'y ai 29. Ms. 7c, p. 387. 30. Ms. 7c, p. 390.

151

M.

D. GRMEK

pas trouv6 de sucre par le reactif. 20 Un morceau de foie tres ramolli chez une femme tres grasse, morte de je ne sais quoi; je n'y ai pas trouv6 de sucre par le r6actif. 30 Un morceau de foie d'apparence saine chez un homme empoisonn6 par l'acide arsenical auquel on avait donn6 du peroxyde de fer. J'y ai trouv6 6normement de sucre par le r6actif et par la fermentation. Le 24 aoiut. Du foie du chien de l'exp6rience de la page 387, mort depuis trois jours, contient encore au r6actif 6norm6ment de sucre. J'avais mis le foie avec de l'eau et c'est dans cette eau 'a odeur forte, paludineuse, qui j'ai agi. Le 25 aotut. Foie d'un albuminurique avec maladie organique du coeur, maladie chronique tres longue; tres infiltree. Le foie est tres congestionne; par tartrate, il y a

des traces de sucre. Sur le foie d'un homme mort de maladie du coeur il y a des traces de sucre. II faudra doser dans les cas diff6rents.31 Thus the presence of sugar in human and bovine liver was clearly demonstrated. On August 25th Bernard found sugar in the livers of a frog, a rabbit, a capon, and two veal fetuses killed in the slaughter house at Popincourt, but he was not able to find sugar in the liver of a ray or a lizard. These phases of Bernard's work culminated in the presenting of a note to the Academy of Science on August 28, 1848. In this communication, signed by Claude Bernard and Charles Barreswil as co-authors, it was stated that sugar extracted from the liver is glucose by chemical nature, that it cannot be found under physiological conditions in any other organ, and that its presence in the liver "Is a physiological fact which is completely independent of the nature of alimentation."32 Bernard and Barreswil presented as evidence to members of the Academy a sample of alcohol originating from the fermentation of liver sugar. The crucial experiment quoted in this paper corresponds to the first experiment of the second series in Bernard's classical memoir on the origin of sugar in the animal body, presented on October 21, 1848, before the Societe de Biologie.33 But there are significant differences between the original experiment and the published text. In August 1848 Claude Bernard was able to demonstrate only the first of the four conclusions quoted in his October 31. Ms. 7c, p. 392. 32. C. Bernard and Ch. Barreswil, '"Du presence du sucre dans le fole," C. R. Acad. Sci., 27 (1848), 514-515. 33. Quoted in note 10 above.

152

The Glycogenic Function of the Liver memorandum. He was sure that "in the physiologic state, there exists constantly and normally the sugar of diabetes in the blood of the heart and in the liver of man and animals," but he only conjectured without any experimental proof that "the formation of this sugar takes place in the liver." Actually, Claude Bernard was immediately led to see that the old theory about the nutritive origin of animal sugar was false, and that the liver produced sugar. But this "immediately" meant immediately after his crucial experiment in August 1848, and not immediately after the beginning of his researches concerning the metabolism of sugar. Of course, it was easy to explain the presence of sugar in the blood of the portal vein in spite of the fact that the current of the blood should carry in opposite direction all substances found in the tissue of the liver. Bermard supposed that the blood rich in sugar had seeped back into the portal vein when, by the opening of the abdomen, the pressure on the viscera ceased. The real demonstration of the glycogenic function of the liver was accomplished during September and October 1848 with a series of experiments characterized by ligatures of the blood vessels of living animals and determination of the sugar in different parts of the circulatory system. Bernard showed that in properly performed experiments there is no sugar in the blood of the portal vein of an animal which is starved or fed on meat. His main argument in favor of the theory of the glycogenic function of the liver was precisely the absence of sugar in the portal vein and its presence in the suprahepatic veins and in the arterial blood. How lucky he was to ignore some facts. First of all, there is in every case some amount of sugar in the blood of the portal vein. It was only by a special property of his chemical test that a gradual difference was transformed into an all-or-none reaction. The large amount of sugar in the blood of Bernard's dogs resulted from the manner of killing (section of the medulla oblongata) and should be interpreted as an exceptional, pathological condition. It is astonishing "how much instinctive judgment and even sheer luck contributed to a discovery which Bernard, with a good deal of justification, believed to be based upon the strictest experimental proof."34 And how interesting it is to measure the extent to which a great scientist reconstructs The his own previous thoughts to fit his later point of view.35 34. J. M. D. Olmsted and E. Harris Olmsted, Claude Bernard and the Experimental Method in Medicine (New York: Schuman, 1952). 35. Cannon's book The Way of an Investigator, New York, 1945, offers many excellent examples of "deductive," historically wrong approaches to

153

M. D. GRMEK

next steps in Bernard'swork on the glycogenic function of the liver are from this point of view even more illuminating. In this case, as probablyin the historical analysis of all other scientific discoveries,it is of invaluablehelp to resort to original first-hand documents. The importance of systematic conservation of this kind of document, especially of laboratoryjournals, can hardly be overestimated. the analysis of scientific discoveries. Cannon states (p. 65) that Claude Bernard "in testing the blood for its sugar content at various points after its departure from the intestine, where sugar is absorbed . . . found less in the blood of the left side of the heart and in the arteries than in the veins. He drew the erroneous conclusion that the sugar was consumed in the lungs. Then Bernard's interest in the metabolism of sugar in the body led him to examine persons suffering from diabetes, and he was struck by the evidence that the output of sugar in the urine of diabetics is greater than that represented in the food they take in. There sprang into his mind a guiding idea that sugar is produced in the organism." This is, I guess, the way in which Cannon would discover animal glycogenesis, but it has no connection with the historical reality.

154

ESSAYREVIEW

PHAGE AND THE ORIGINS OF MOLECULAR BIOLOGY Reviewed by R. C. Lewontin, Department of Biology, University of Chicago Edited by J. Cairns, G. S. Stent, and J. D. Watson. Cold Spring Harbor Laboratory of Quantitative Biology, Cold Spring Harbor, New York. 1966. xii + 340. When Alexis de Tocqueville sat in the Chamber of Deputies on February 24, 1848, watching the dissolution of the bourgeois monarchy, he detected a certain lack of spontaneity in the proceedings. The men of the first revolution were living in every mind, their deeds and words present to every memory. All that I saw that day bore the visible impress of those recollections; it seemed to me throughout as though they were engaged in acting the French revolution rather than continuing it. Nor was this simply the prejudice of a man himself conscious of history. The political factions of Lamartine and Caussidiere called themselves "Girondins"and "Montagnards," the latter taking their seats in the highest benches of the assembly in conscious imitation of the National Convention. De Tocqueville found "the imitation so evident that the originality of the facts remained concealed beneath it." What De Tocqueville had discovered first hand is what is known to all historians: that a consciousness of past events and their relation to the present, that a "sense of history," has a profound effect on the rhetoric and actions of men in the present. The events of 1789-99 were the models, not only for the Revolution of 1848, but for 1917 as well. Trotsky's chief source of allusion was the French Revolution, and in the organization and tactics of the Military Revolutionary Committee conscious account was taken of the 18th Brumaire and the lessons it held. The identification of present events and movements with

155

ESSAY REVIEW

those of a past golden age, and especially the overt self-identification of participants with dead heroes, has been characteristic of the history of political institutions, but not of science. Whatever the form of their personal fantasies, scientists have not written of their work as part of an historical movement nor have they preened themselves self-consciously before the mirror of history. Individual scientists have written memoirs, and a few have considered the philosophical implications of scientific discoveries of which their own work has been a part. They are, after all, as vain as other men, but their concern for their place in history has not been flaunted. This is only partly because of the myth of modesty-what Andre Lwoff calls "the Golden Rule of intellectual hygiene," that "a scientist should never attempt to judge the value of his own achievements." Very few scientists really believe the myth, certainly not Andre Lwoff, whose contribution to "The Origins of Molecular Biology" bears the remarkable title "The Prophage and I." The chief reason for the lack of self-conscious historical orientation among scientists is their belief in the objective historicity of science. That is, science is regarded as discovery of relations between objects, relations that exist apart from the knowledge of them. Thus, the development of scientific knowledge is inevitable, provided enough effort is put into it and only a single truth eventually emerges. Moreover, the course of scientific discovery, in this view, is rather inflexible since the discovery of one relation is the necessary condition for the discovery of the next. Looked at thus, the history of science is of antiquarian interest only. It is a history of the correction of error and of the inevitable passage from one discovery to the next. A knowledge of history cannot then illuminate the present or guide the future. Political institutions, on the other hand, are seen as having subjective historicity. Even for an orthodox Marxian historian the relation between two events in time is not a necessary one. No historical sequence is inevitable, although the eventual outcome may be, and political institutions are not "discovered" but "created." Since human political history may have had many outcomes never realized, the individual acts of men take on a great significance, and the past becomes both a caution and a model for the present. For the scientist the past is a textbook; for the political man it is a scenario. Phage and the Origins of Molecular Biology is a unique work. It introduces into the literature of science, for the first time, a self-conscious historical element in which the participants in scientific discovery engage in writing their own chronicle. As

156

ESSAY REVIW

such, it is an important document in the history of biology. Whereas The Origins of Molecular Biology gives little insight into the way in which science is done, it is tremendously revealing of the sociology of science and of the self-image of successful biologists. The book, consisting of a preface and thirty-three essays contributed by thirty-five molecular biologists, is ostensibly a Festschrift for Max Delbruck. But it is a strange Festschrift. Delbruck is only sixty and still active. One of the articles is by Delbruck himself; another, by Werner Reichardt, reports work "interrupted" by a short sojourn in Delbriuck's laboratory; and a third, by Aaron Novick, turns out to be a celebration of Delbruck's chief competition for godhood, Leo Szilard. The editors in their preface offer us the explanation: Besides paying homage to Delbruck as a prime mover and arbiter of nascent molecular biology, this book is an attempt therefore to write a history of a bygone age and put on record the network of interactions, foldlore, and method of operation of the Phage Group that had Delbruck as its focal point. What the editors have complied in their attempt to write "a history of a bygone age" is a kind of Gesta Philosophorum, with photographs of knights in full panoply of sport shirt, shorts and sandals, including one of James Dewey Watson transfixing with his lance the coils of a dragon DNA. Nor are the holy places lacking. California Institute of Technology and the Cold Spring Harbor Laboratory, called the "Mecca and Medina of the Phage Group" by the editors, are part of the iconography. There are even relics-three pages reproduced from Szilard's notebooks in his very handwriting. And what of the grail? The search for it is the central theme of the book. It is called The Paradox. In his introductory essay, "Waiting for the Paradox," Gunther Stent with characteristic intelligence has described the basic epistemological position of molecular biology, at least at its beginnings. Thus it was the romantic idea that "other laws of physics" might be discovered by studying the gene that really fascinated the physicists. This search for physical paradox, this quixotic hope that genetics would prove incomprehensible within the framework of conventional physical knowledge remained an important element in the psychological infrastructure of the creators of molecular biology. Stent does not discuss the origins of this desire for paradox although he draws attention to the essay-book What is Life by

157

ESSAY REVIEW

Erwin Schrodingeras having an influential role. For the history of science, however, it is precisely the origin of this desire that is of most interest, for it is the real origin of molecularbiology. Modem science has undergoneone real revolutionsince Newton, a revolutionthat began in 1905 with Einstein'spaper on the photoelectric effect and reached its culmination in 1925-27 in the quantum theory of De Broglie, Heisenberg,and Schr6dinger. That revolution did nothing less than overturn the Cartesian world view of cause and effect and replace it with the fundamentally antimaterialisticprinciple that chance is an ontological propertyof matter and energy.For scientists broughtup since that time (including Max Delbruck, who was a student in Tubingen,Berlin, and Bonn in 1925 and 1926) the development of quantum physics has been the model, the "firstrevolution" whose men are "livng in every mind, their deeds and words present in every memory."It is a result of that revolution that modern scientists, especially physicists, have become acutely conscious of epistemologicaland ontologicalproblemsof science. Moreover,that revolutionwas a challenge to the objectivehistoricity of scientific development.One does not speak of the discoverers of quantummechanics (as even scientists will often speak of Newton "discovering"the Law of Gravitation), but of the inventors of quantum theory. The invention by De Broglie and Heisenbergof alternative equivalent systems to resolve the particle-wave paradox represents the triumph of human ingenuity over the limitations of the mind to cope with "paradoxical" behavior of natural objects. Quantum theory in itself is not a revelation of one of the secrets of nature, but a reorganization of man's attitudes about nature. It is then a human revolution with human heroes and, as such, calls forth the same reaction from its admirersthat the events of 1789 called forth from the activists of 1849. Some membersof the 'Thage Group,"at least, saw themselves as the Heisenbergsand Schrodingersof biology. It is Stent's opinion that Delbriickmoved away from molecular genetics when he found no paradox and turned deliberatelyto that field of biology that was most likely to result in one-the higher nervous system. Aside from Delbruck himself, it would appear from their contributions to Phage and the Origins of Molecular Biology

that many others see themselves and their work in an historical perspective, a perspective they have gained, I claim, from their reading of the history of physics. First, there is the intensely personal and anecdotal nature of some of the contributions.Of the thirty-threeessays, ten are reminiscences in which personal history strongly predominates over the development of ideas.

158

ESSAY REVIEW

Some reveal themselves in their titles: "The Prophage and I" (Andre Lwoff), "Adventures in the rII Region" (Seymour Benzer), "Growing up in the Phage Group" (J. D. Watson). Indeed, the editors requested that the contributors make their essays "highly personal." Now reminiscence is a favorite after-dinner occupation of scientists; it is their innocent and sometimes not-so-innocent form of gossip. But the deliberate and serious publication of reminiscences implies that some circle, wider than the group of intimates who have shared the experience, will care, and that implies, in tum, that the contributors and their friends have performed acts of historical note. Bus conductors do not write their memoirs. Second, the rhetoric of these pieces is, like the psychopathology of everyday life, marvelously revealing of the attitudes and perspectives of their authors. There are many references to persons and places without further identification, as if these names, so familiar to the Phage Group, would call forth a sure response from every reader. And there is a great deal of the little-did-Irealize-that-morning school of writing. Rhetorical devices that juxtapose the trivial against the weighty as a means for creating relief depend for their success on the obvious magnitude of the contrast. Lwoff's "Golden Rule of intellectual hygiene" has not been universally applied. Third, there is everywhere in the book that feeling of commitment to an ideal, to a movement, that is usually characteristic of political activists, but not scientists. Perhaps the closest parallel can be found in mathematics where the Bourbakists have been so committed to their group ideal that they have submerged their individual identities in a mystical union of intellects. While the followers of Delbriuck did not go so far, there was, in fact, a cadre that met in Cold Spring Harbor to parcel out problems and experiments. This willing submission to party discipline on the part of ambitious, brilliant individualists testifies to the sense of historic mission they must have felt. Apparently there would be enough glory to go around, as indeed there has been. Not all the articles are highly personal. A few, like Reichardt's summary of his work on the insect optomotor response, Hershey's discussion of the injection of DNA from phage into bacteria, and Doermann's article on the eclipse period in the phage life cycle, are straightforward summaries of the development of particular areas of biology. Most of the articles are of this kind with a mixture of personal reminiscence and scientific report. Only one seems to me to have been completely successful in giving the intellectual history of important discoveries in molecular biology in a way that is directly illuminating for the

159

ESSAY REVIEW

historian of Ideas. That is the article "Gene,TransformingPrinciple, and DNA"by Rollin Hotchkiss, one of the subtlest interigences in biology. Lightly salted with anecdote so as to enhance the flavorrather than dominate it, the article relates all the concepts of formal genetics to the discoveries of molecular biology through the medium of the author's own work and that of his intellectual predecessors.It is the model for such essays. What did the Phage Group create? Is molecular biology a scientific revolution? There is not the slightest question that more progress has been made in our understandingof the fundamental physicochemicalmechanisms of life in the last twenty years than in all the previoushistory of biology. Nor is there any doubt that the movement that calls itself "molecularbiology"is in large part responsible for this progress. J. D. Watson and Francis Crickhave earned themselves an importantplace in the history of biologyby making a truly seminal discovery,a discovery that produceda sudden order in a chaos of phenomenology and providedfor a vast amount of future discovery and understanding. The elucidation of the structure of DNA is the most important event in biology since the rediscovery of Mendel's work in 1900. It is also the paradigm of molecular biology for it illustratesthe differencein attitude that separatesthat science from classical biochemistryand cellular physiology. In What is Life? Schrodingerpointed out that the mechanism of inheritance requiredthe seemingly contradictoryattributesof extremely high precision of a chemical mechanism and very low concentrations of molecules. That is, the molecular mechanism of heredity must be based not on the Laws of Mass Action, on statistical properties of ensembles of molecules, but on the individually repeatable behavior of separate molecules. This behavior is a result of the structure of the molecules themselves rather than the thermodynamicpropertiesof their milieu. The entire process of reproduction in turn must be explained in terms of a chain of molecular events in proper temporal order and with the molecules occupying specific sites or moving in specific pathways. Always the emphasis is on individual molecules in space and time rather than ensembles averaged over space and time. Molecularbiology, unlike classical biochemistry, places the emphasis on discreteness rather than continuity, on deterministic rather than average statistical behavior. An example of this last emphasis can be found in the interpretations of the experiments assigning particular polynucleotides as codons for particularamino acids in protein synthesis. The actual experimentalresults show that a particulartriplet of nucleotide bases will code for more than one amino acid, but one more than

160

ESSAY REVIEW

others. The classical biochemical view of such results would be that this is a reaction with one main product and numbers of side products and that chemical reactions are, after all, never pure. For the molecular biologist, however, "we see through a glass, darkly."The ambiguity arises from the imperfectionwith which the in vitro experiment mimics the living system. He therefore makes an unambiguous assignment of a triplet of nucleotides as the codon for a particular amino acid with the faith that in the living cell the reaction has only one product. As a result, molecular biology has built up a picture of the assembly of proteins based upon the movement and fitting together of specific molecules in specific combinations. It is a mechanical system with gears and levers, and resembles nothing so much as an automobile assembly line. Ironically, molecular biology, originating in the hope of a paradox and a revolution, ends by being utterly mechanistic and the greatest triumph of the bete machine that Descartes could have imagined.

161