Popular Science Monthly/Volume 79/December 1911/Buffon and the Problem of Species II

by Arthur O. Lovejoy

Layout 4

BUFFON AND THE PROBLEM OF SPECIES

By Professor ARTHUR O. LOVEJOY

THE JOHNS HOPKINS UNIVERSITY

II

WE have thus far noted three generally disregarded but fundamental facts concerning Buffon's opinions about the nature of species. The first fact is that in his preliminary discourse in the first volume of the "Histoire Naturelle," in which he sought to apply the Leibnitian principle of continuity to natural history, Buffon's emphasis upon the continuity of the gradations between species probably had no evolutionary implications. The second fact is that the principal doctrine of this discourse is to the effect that only individuals exist in nature, while species exist only by grace of the human imagination, which, aided by human ignorance, sees sharp lines of cleavage among organisms where no such lines are. The third fact is that this doctrine was already tacitly but decisively abandoned in Buffon's second volume, where he represents species as real and well-marked natural entities, their limits being determined by the test of the sterility of the products of cross-breeding. There are, indeed, many later passages where the old phraseology incongruously recurs; but it recurs in contexts in which the reality of species is expressly insisted upon.

2. When the fourth volume of the "Histoire Naturelle"—the first dealing specifically with the lower animals—appeared in 1753, four years after the first three, Buffon's departure from the notions set forth in the preliminary discourse became still more evident. He had by this time, in the first place, been greatly impressed by the homologies in the structure of the vertebrates; he had come to see some significance in those facts of comparative anatomy which his own treatise—though more through the contributions of Daubenton than through his own—was for the first time setting in a clear light. The existence throughout at least all the immensely diverse vertebrate forms of an underlying unity of type, Buffon was, I suppose, the first to bring forcibly to the attention of naturalists and philosophers, as a fact calling for serious consideration and explanation.

If we choose the body of some animal or even that of man himself to serve as a model with which to compare the bodies of other organized beings, we shall find that. . . there exists a certain primitive and general design, which we can trace for a long way. . . . Even in the parts which contribute most to give variety to the external form of animals, there is a prodigious degree of resemblance, which irresistibly brings to our mind the idea of an original pattern after which all animals seem to have been conceived. What, for example, can at first seem more unlike man than the horse? Yet when we compare man and horse point by point and detail by detail, is not our wonder aroused rather by the resemblances than by the differences to be found between them?. . . It is but in the number of those bones which may be regarded as accessory, and in the lengthening or shortening or mode of attachment of the others, that the skeleton of the horse differs from that of the human body. . . . The foot of the horse (as M. Daubenton has shown), in appearance so different from the hand of man, is nevertheless composed of the same bones, and we have at the extremities of our fingers the same small hoof-shaped bone which terminates the foot of that animal. Judge, then, whether this hidden resemblance is not more marvelous than any outward differences, whether this constancy to a single plan of structure—which we can follow from man to the quadrupeds, from the quadrupeds to the cetacea, from the cetacea to birds, from birds to fishes, from fishes to reptiles—whether this does not seem to show that the Creator in making all these used but a single main idea, though varying it in every conceivable manner—so that man might admire equally the magnificence of the execution and the simplicity of the design.

But consideration of the anatomical homologies did not lead Buffon merely to pious reflections. He saw clearly and unequivocally declared that this unity of type forcibly suggests the hypothesis of community of descent. To one who considers only this class of facts, he wrote:

Not only the ass and the horse, but also man, the apes, the quadrupeds, and all the animals, might be regarded as constituting but a single family. . . . If it were admitted that the ass is of the family of the horse, and differs from the horse only because it has varied from the original form, one could equally well say that the ape is of the family of man, that he is a degenerate (dégénéré) man, that man and ape have a common origin; that, in fact, all the families, among plants as well as animals, have come from a single stock; and that all animals are descended from a single animal, from which have sprung in the course of time, as a result of progress or of degeneration, all the other races of animals. For if it were once shown that we are justified in establishing these families; if it were granted that among animals and plants there has been (I do not say several species) but even a single one, which has been produced in the course of direct descent from another species; if, for example, it were true that the ass is but a degeneration from the horse—then there would no longer be any limit to the power of nature, and we should not be wrong in supposing that, with sufficient time, she has been able from a single being to derive all the other organized beings.

Buffon thus presented the hypothesis of evolution with entire definiteness, and indicated the homological evidence in its favor. But did he himself regard that evidence as conclusive, and therefore accept the hypothesis? The passage cited is immediately followed by a repudiation, ostensibly on theological grounds, of the ideas which he has been so temptingly presenting.

But no! It is certain from revelation that all animals have participated equally in the grace of direct creation, and that the first pair of every species issued full formed from the hands of the Creator.^[1]

This repudiation has been regarded as ironical, or as inserted merely pro forma, by those interpreters of Buffon who have made him out a thorough-going evolutionist. Unfortunately, nearly all these writers—dealing somewhat less than fairly with their readers—have failed to mention that his rejection of the evolutionary hypothesis was not put forth by him as resting exclusively upon these religious considerations. If the words just quoted stood alone, it would, indeed, be scarcely possible to take them seriously. But they do not stand alone; they are directly followed by arguments of quite another order against the possibility of the descent of one real species from another; and the essence of the most emphasized of these arguments lies in the Buffonian conception of the nature of species, already expounded in the second volume. In other words, the fact of the sterility of hybrids, and certain other purely factual considerations, were urged by Buffon as conclusive objections against the theory of descent.

Specifically, his arguments against evolution are three: (1) Within recorded history no new true species (in his own sense of the term) have been known to appear. (2) There is one entirely definite and constant line of demarcation between species: it is that indicated by the infertility of hybrids.

This is the most fixed point that we possess in natural history. No other resemblances or differences among living beings are so constant or so real or so certain. These, therefore, will constitute the only lines of division to be found in this work.

But why, it may be asked, should the sterility of hybrids be a proof of the wholly separate descent of the two species engendering such hybrids? This question Buffon does not neglect to answer. An "immense and perhaps an infinite number of combinations" would need to be assumed before one could conceive that "two animals, male and female, had not only so far departed from their original type as to belong no longer to the same species—that is to say, to be no longer able to reproduce by mating with those animals which they formerly resembled—but had also both diverged to exactly the same degree, and to just that degree necessary to make it possible for them to produce only by mating with one another." The logic of this is to me, I confess, a trifle obscure; but it is evident that Buffon conceived that the evolution from a given species of a new species infertile with the first could come about only through a highly improbable conjunction of circumstances. (3) Buffon's third reason for maintaining the fixity of species is the argument from the "missing links."

If one species had been produced by another, if, for example, the ass species came from the horse, the result could have been brought about only slowly and by gradations. There would therefore be between the horse and the ass a large number of intermediate animals. Why, then, do we not to-day see the representatives, the descendants, of these intermediate species? Why is it that only the two extremes remain?

Taking these three arguments into account, then, Buffon arrives at this conclusion:

Though it can not be demonstrated that the production of a species by degeneration from another species is an impossibility for nature, the number of probabilities against it is so enormous that even on philosophical grounds one can scarcely have any doubt upon the point.^[2]

However plausibly Buffon's incidental expressions of deference to the testimony of revelation may be regarded as perfunctory and insincere, it would be absurd to suppose that he was also ironical in these legitimate and ostensibly scientific (however poor) arguments for the fixity of species—arguments which are closely connected with that conception of the nature of species which was perhaps his most influential personal contribution to the biological ideas of his time. We must conclude, then, that, while he clearly envisaged the hypothesis of evolution as early as 1753, and recognized that there was some probable evidence in its favor, he then seriously believed that the preponderance of probability was enormously against it. It is certain that contemporary readers must have understood this to be his position.

The same doctrine—that true species, as determined by the sterility of hybrids, are real natural entities and constant units amid the otherwise infinitely variable phenomena of organic nature—is repeated and emphasized many times in subsequent volumes of the "Histoire Naturelle." Thus in volume five (1755) Buffon—trying to retain as much of the principle of continuity as could be made consistent with his present view—writes as follows:

Although animal species are all separated from one another by an interval which nature can not overstep, some of them seem to approximate one another by so great a number of relations, that there remains between them only so much of a gap as is necessary to establish the line of separation.^[3]

In the same volume he insists upon the equal antiquity of all real species, in the very passage in which he emphasizes the possibility of a wide range of variation within the species:

Though species were formed at the same time, yet the number of generations since the creation has been much greater in the short-lived than in the long-lived species; hence variations, alterations, and departures from the original type, may be expected to have become far more perceptible in the case of animals which are so much farther removed from their original stock.^[4]

This is advanced as a partial explanation of the extreme diversity of breeds in the canine species: the dog is a short-lived animal and has therefore been capable of a relatively great degree of diversification.

A little later (in Vol. VI.,^[5] 1756) Buffon declares that "nature imprints upon every species its inalterable characters." In 1765—that is, at precisely the period during which we are told that Buffon "was expressing very radical views on the mutability of species "—we find him (in his "Second View of Nature," Vol. XIII.) giving his most extreme expression to the doctrine of the reality and constancy of genuine species. Here the language of the preliminary discourse concerning the relative significance in nature of the species and the individual has come to be completely reversed.

An individual, of whatever species it be, is nothing in the universe; a hundred, a thousand individuals are nothing. Species are the only entities of nature (les seuls êtres de la nature)—perduring entities, as ancient, as permanent, as nature herself. In order to understand them better, we shall no longer consider species as merely collections or series of similar individuals, but as a whole independent of number, independent of time; a whole always living, always the same; a whole which was counted as a single unit among the works of the creation, and which consequently makes only a single unit in nature. . . . Time itself relates only to individuals, to beings whose existence is fugitive; but since the existence of species is constant, it is their permanence that constitutes duration, the differences between them that constitute number. . . . Let us then give to each species an equal right at nature's table; they are all equally dear to her, since she has given to each the means of existing, and of enduring as long as she herself endures.^[6]

This sort of rhetoric is not the dialect of an evolutionist; it is almost that of a Platonist. And there is more in plainer language to the same effect:

Each species of both animals and plants having been created, the first indiriduals of each served as models for all of their descendants. . . . The type of each species is cast in a mold of which the principal features are ineffaceable and forever permanent, while all the accessory touches vary.^[7]

Many years later still, in 1778, there appeared the sub-division of the "Histoire Naturelle" which Buffon's contemporaries regarded as his most brilliant and most significant work—the "Époques de la Nature." This was a resumption on a grander scale, and upon new principles, of the task attempted in the "Theory of the Earth" in the first volume, thirty years before—an outline of planetary evolution. To the diffusion of evolutionary ways of thinking in the larger and vaguer sense, this treatise was a contribution of capital importance. Into the details of Buffon's geology I do not wish to enter in this paper. But it is worth while for our purpose to recall one or two striking facts about the "Époques." In it the writer, whom a recent German historian of biology has declared to have had a too little developed sense for the historical or genetic aspect of nature, attempted, in a far more comprehensive, more definite and more impressive way than any of his predecessors, to write the history of the gradual development of our planet from the time when, an incandescent ball, it was separated from the sun. The task was, of course, undertaken prematurely; but Buffon not only made the need of its eventual achievement evident, but also indicated two of the essential means by which it was to be accomplished: the study of present phenomena which can throw light upon the past processes through which existing conditions have been brought about; and the study of those natural "monuments which we ought to regard as witnesses testifying to us concerning the earlier ages." He insisted, moreover, with the utmost plainness upon (as it was then regarded) the extreme antiquity not only of the earth, but also of organic life. And in doing so he showed himself not at all disposed any longer to permit "revelation" to settle scientific questions. "How," he writes, "some one will ask me, do you reconcile this vast antiquity which you ascribe to matter with the sacred traditions, which give to the world only some six to eight thousand years? However strong be your proofs, however evident your facts, are not those reported in the holy book more certain still? "Buffon replies that he has all possible respect for scripture, but that it always pains him to see it used in this way. Doubtless there is no real conflict between its testimony and that of science; and he thereupon introduces what I suppose is the first of the long series of reconciliations of Genesis and geology. The six days were not really days, but long periods of time, and so forth. But in any case, he concludes, the Bible was originally addressed to ignorant men at an early stage of civilization, and was adapted to their needs and their intelligence. Its science was the science of the time, and ought not to be taken too literally. Finally, it is to be noted that in the "Époques" Buffon ceased to talk of the simultaneous creation of all species, and advanced the doctrine of the gradual appearance of different sorts of animals in conformity with geological conditions.

If, then, Buffon was desirous of inculcating the theory of the mutability of species, here was the place in which, above all others, he might be expected to do so fully and unequivocally. But here once more we find him reiterating the substance of his old doctrine:

A comparison of these ancient monuments of the earliest age of living nature with her present products shows clearly that the constitutive form of each animal has remained the same and has undergone no alteration of its principal parts. The type of each species has not changed; the internal mold has kept its shape without variation. However long the succession of time may be conceived to have been, however numerous the generations that have come and gone, the individuals of each kind (genre) represent to-day the forms of those of the earliest ages—especially in the case of the larger species, whose characters are more invariable and whose nature is more fixed.^[8]

By the "larger species" here, Buffon means those of greater size, such as the elephant and hippopotamus; and when he says that these are "especially" invariable, he means, as the whole context shows, not that any other species ever departs from its specific type, but that in these larger creatures even the "accessory touches" have been comparatively little altered.

Thus, in a long series of passages, from 1753 on, we find Buffon reiterating with explicitness and emphasis the same teaching, which has, for him, its principal bases in two of his most cherished conceptions: namely, in his conviction that the sterility of hybrids shows that species are real "entities of nature"; and in his embryological theory of "organic molecules" and of the "internal mold" which "casts into its own shape those substances upon which it feeds" and "can operate in the individual only in accordance with the form of each species." One of the first of modern naturalists to make the idea of organic evolution familiar to his contemporaries and to discuss it seriously, Buffon repeatedly rejected that theory, at all periods of his career; and he did so, not from timidity merely nor from an affectation of deference to scriptural authority, but upon reasoned grounds which he plainly stated and had every appearance of presenting as conclusive. Yet it is also undeniable, as will presently be seen, that he did not maintain this negative position without occasional waverings and doubts and at least one clear, though possibly inadvertent, self-contradiction.

3. In spite of his habitual emphasis upon the constancy of true species, Buffon insisted more than any of his predecessors, and more, perhaps, than any of his contemporaries, except Maupertuis and Diderot, upon the variability of organisms and the potency of the forces making for their modification.

Though nature appears always the same, she passes nevertheless through a constant movement of successive variations, of sensible alterations; she lends herself to new combinations, to mutations of matter and form, so that to-day she is quite different from what she was at the beginning or even at later periods.^[9]

The passage is from one of Buffon's later writings; but its close counterpart is to be found as early as 1756:

If we consider each species in the different climates which it inhabits, we shall find perceptible varieties as regards size and form; they all derive an impress to a greater or less extent from the climate in which they live. These changes are made slowly and imperceptibly. Nature 's great workman is Time. He marches ever with an even pace, and does nothing by leaps and bounds, but by degrees, gradations and successions he does all things; and the changes which he works—at first imperceptible—become little by little perceptible, and show themselves eventually in results about which there can be no mistake.^[10]

For the most part these changes were clearly represented by Buffon as taking place only within the limits of species; they amounted merely to the formation of new "races" or varieties. Since his criterion of identity of species (the possibility of interbreeding) did not essentially depend upon morphological similarity, he could with consistency suppose the descendants of a given pair to have departed to a very great (though not to an indefinite) degree, in the course of ages, from the form and external characters of their ancestors. It was, in other words, characteristic of his biological system that he set up an absolute distinction between species and varieties, gave an extreme extension to the notion of a variety, and sought to reduce the number of separate species as much as possible, by assuming—until the establishment of the sterility of the hybrids should prove the contrary—that most of the Linnæan species were merely varieties descended from a relatively small number of original specific types. Near the close of his essay "De la dégénération des animaux" (1766), Buffon writes:

To account for the origin of these animals [certain of those found in America] we must go back to the time when the two continents were not yet separated and call to mind the earliest changes which took place in the surface of the globe; and we must think of the two hundred existing species of quadrupeds as reduced to thirty-eight families. And though this is not at all the state of nature as we now find it, but a state much more ancient, at which we can arrive only by induction and by analogies. . . difficult to lay hold of, we shall attempt nevertheless to ascend to these first ages of nature by the aid of the facts and monuments which yet remain to us.^[11]

Here, clearly, is an evolutionary program, strictly limited by the assumption that there are irreducible ultimate species, yet tolerably ambitious: to regard all known kinds of quadrupeds as derived from thirty-eight original types, by modification in the course of natural descent; and to determine the general causes and conditions of the production of species in the ordinary sense, i. e., of relatively stable varieties. These ideas occurred to Buffon too late to be made use of in his general plan for the classification of the quadrupeds; that plan, it will be remembered, was formed while he was unluckily under the influence of the principle of continuity. But in the volumes on birds, of which the first appeared in 1770, he had the opportunity for a fresh start; and he took advantage of it to introduce a method of distinguishing and classifying species which—within the limits already indicated—is expressly evolutionary in its principles.

For the natural history of the birds I have thought that I ought to form a plan different from that which I followed in the case of the quadrupeds. Instead of treating of the birds. . . by distinct and separate species, I shall bring several of them together under a single genus. Except for the domesticated birds, all the others will be reunited with the species nearest to them and presented together as being approximately of the same nature and the same family. . . . When I speak of the number of lines of parentage, I mean the number of species so closely resembling one another that they may be regarded as collateral branches of a single stock, or of stocks so close to one another that they may be supposed to have a common ancestry and to have issued from that same original stock with which they are connected by so many points of resemblance common to them all. And these related species have probably been separated from one another only through the influences of climate and food, and by the lapse of time, which brings about all possible combinations and gives play to all the agencies that make for variation, for improvement, for alteration and for degeneration.^[12]

Even the groupings which he gives, Buffon adds, can not be regarded, in the existing state of knowledge, as correctly and exclusively enumerating all the apparent species which are really akin to one another. The number of separate species which he lists, he intimates, is probably much too great. But at all events, he concludes with pride, his work is the first real attempt at an ornithologie historique.

The purpose of the present inquiry does not call for any extended exposition of Buffon's views about the causes of modification in animals and the ways in which quasi-species are formed. In the essay "De la dégénération des animaux" the subject is discussed at the length of over sixty quarto pages; the theories there advanced have been sufficiently accurately summarized by many previous writers. In brief, the factors in modification which he mentions as the most important are changes of climate (in which the most potent influence is temperature), changes of food, and the effects of domestication. But it is evident that he also believed in a general tendency to variation in the germ, and in the influence of acquired habits, of the use and disuse of parts, and of acquired lesions and mutilations. Thus he explains the humps, and the callosities on the knees and chest, of the camel and the llama as due to the habits of those animals under domestication. Similarly, the callosities on the haunches of the baboons arise from the fact that "the ordinary position of these animals is a sitting one—so that the hard skin under the haunches has even become inseparable from the bone against which it is continually pressed by the weight of the body." These theories, of course, take for granted the inheritance of acquired characters, which Buffon also (less cautious here than Maupertuis^[13]) explicitly asserts. It is, I suppose, also well known that Buffon called attention (as Linnæus did independently) to the struggle for existence between species, due to the excessive multiplication of individuals, and pointed out how an equilibrium comes to be established (so long as external conditions remain constant) by means of this opposition.

It may be said that the movement of nature turns upon two immovable pivots—one, the boundless fecundity which she has given to all species; the other, the innumerable difficulties which reduce the results of that fecundity, and leave throughout time nearly the same number of individuals in every species.^[14]

Buffon, in fact, rather over-worked this notion of a stable equilibrium, which rested upon the assumption of an approximate equality among species in their endowment for the struggle for survival. This is perhaps one reason why it did not occur to him to think of that struggle as causing a process of natural selection, or to see in it a factor in the formation of so-called species.

4. It must be evident to the reader from all that precedes that Buffon's mind, throughout nearly the whole of his life, was played upon by two opposing forces. Quite apart from any illegitimate and external influences, such as fear of the ecclesiastics—of which too much has been made—his thought was affected by two conflicting sets of considerations of a factual and logical sort. He saw certain definite reasons for regarding species as the fundamental constants of organic nature; what those reasons were has been sufficiently indicated. But he also saw that there was some force in the argument from the homologies; and—what in his case was still more important—he was committed to the program of explaining the diversities of organisms, so far as might be, by the hypothesis of modification in the course of descent; he was deeply impressed by the fact of variability; and he held to a theory of heredity (namely, of the heritability of acquired characters) which acted as a sort of powerful undertow towards a generalized evolutionism. Add to this that he was little careful of consistency and extremely careful of rhetorical effect—and it is not surprising that he occasionally forgot one side of his doctrine in emphasizing the other. There is one and, so far as I can discover, only one passage in which he seems categorically to contradict his ordinary teaching of the impossibility of the descent of really distinct species, sterile inter se, from common ancestors. This occurs at the end of the chapter on "Animals Common to Both Continents" (Vol. IX., 1761).

It is not impossible that, without any deviation from the ordinary course of nature, all the animals of the New World may be at bottom the same as those of the Old—having originated from the latter in some former age. One might say that, having subsequently become separated by vast oceans and impassable lands, they have gradually been affected by a climate which has itself been so modified as to become a new one through the operation of the same causes which dissociated the individuals of the two continents from one another. Thus in the course of time the animals of America have grown smaller and departed from their original characters. This, however, should not prevent our regarding them to-day as different species. Whether the difference be caused by time, climate and soil, or be as old as the creation, it is none the less real. Nature, I maintain, is in a state of continual flux and movement. It is enough for man if he can grasp her as she now is, and cast but a glance or two upon the past and future, to endeavor to perceive what she may once have been and what she may yet become.

Here Buffon seems either to have forgotten or to have deliberately discarded his own usual criterion of diversity of species. He does not propose to inquire whether the American species are capable of having fertile progeny when mated with their respective congeners in the old world, but predicates difference of species solely on the ground of dissimilarity of form; and to the distinct species so determined he attributes an identical origin. But it is possible that he has here merely lapsed (as he apparently does occasionally elsewhere) into the terminology in which he was brought up, and is using the word species in the Linnæan sense rather than in his own.

More significant, perhaps, than this possibly inadvertent inconsistency is the fact that, in his fourteenth volume^[15] (1766), Buffon seems to raise explicitly the question—though only as a question—whether, after all, descent with modification may not extend to species as well as varieties.

After surveying the varieties which indicate to us the alterations that each species has undergone, there arises a larger and more important question, namely, how far species themselves can change—how far there has been a more ancient modification, immemorial from all antiquity, which has taken place in every family, or, if the term be preferred, in all the genera in which species that closely resemble one another are to be found. There are only a few isolated species which are like man in forming at once a species and a whole genus. Such are the elephant, rhinoceros, hippopotamus and giraffe, which constitute genera, or simple species, and descend in a single line, with no collateral branches. But all other races have the appearance of forming families, in which we may perceive a common source or stock from which the different branches seem to have sprung.^[16]

Even here one can not be wholly sure that Buffon is not referring to Linnæan species, and using the word genera to indicate what he usually means by species in the strict sense. Assuming, however, that he is speaking of "true" species, it must be observed that while he raises the question of their mutability, he does not answer it finally in the affirmative. For the passage is shortly followed by that cited earlier in this paper, in which Buffon, though he derives many species traditionally regarded as distinct from a common stock, yet finds even in "the first ages of nature" thirty-eight irreducible diversities of specific type among quadrupeds.

There is, however, one peculiarly interesting essay in which Buffon shows himself a little dubious even about that "most fixed point in nature" upon which his usual doctrine of the reality and constancy of species was based—namely, the fact of the sterility of hybrids. As I have already mentioned, this seemed to him so central a point in natural history that he for many years assiduously collected data concerning it, and caused experiments bearing thereon to be made and carefully recorded at his own estate at Montbard. The results of these inquiries, which he reports in the chapter "On Mules" (in the third supplementary volume, 1776), led him to the conclusion that hybrids are not necessarily without hope of posterity. On the testimony of an affidavit from a gentleman in San Domingo, Buffon declares that in hot climates mules have been known to beget offspring of mares, and females of their kind to breed with horses. "One was therefore wrong formerly in maintaining that mules are absolutely infertile." Other experiments in the crossing of goats and sheep, dogs and wolves, canaries and goldfinches, are recited; they all go to show that sterility is merely a question of degree.

All hybrids (mulets), says prejudice, are vitiated animals which can not produce offspring. No animal, say reason and experience, is absolutely infertile, even though its parents were of separate species. On the contrary, all are capable of reproduction, and the only difference is a difference of more or less.^[17]

That hybrids are relatively infertile, and probably incapable of breeding with one another, Buffon still maintains; "their infecundity, without being absolute, may still be regarded as a positive fact." Something, therefore, is still left of his test of unity of species. But now that it seemed to be reduced to a mere difference in degree, it was no longer the sharp-cut, decisive, impressive thing that it had at first appeared. And, feeling that his criterion of species had a good deal weakened, Buffon was led—not, indeed, even now to an altogether unequivocal affirmation of the descent of real species from one another—but to a confused, half-agnostic utterance, in which he seems to take at least the possibility of such descent for granted:

In general, the kinship of species is one of those profound mysteries of nature which man will be able to fathom only by means of long and repeated and difficult experiments. How, save by a thousand attempts at the cross-breeding of animals of different species, can we ever determine their degree of kinship? Is the ass nearer to the horse than to the zebra f Is the dog nearer to the wolf than to the fox or the jackal? At what distance from man shall we place the great apes, which resemble him so perfectly in bodily conformation? Were all the species of animals formerly what they are to-day? Has their number not increased, or rather, diminished (sic)?. . . What relations can we establish between this kinship of species and that better known kinship of races within the same species? Do not races in general arise, like mixed species, from an incapacity in the individuals from which the race originated for mating with the pure species? There is perhaps to be found in the dog species some race so rare that it is more difficult to breed than is the mixed species produced by the ass and the mare. How many other questions there are to ask upon this matter alone—and how few of them there are that we can answer! How many more facts we shall need to know before we can pronounce—or even conjecture—upon these points! How many experiments must be undertaken in order to discover these facts, to spy them out, or even to anticipate them by well-grounded conjectures!^[18]

This passage certainly indicates a strong inclination towards an acceptance of a thorough-going doctrine of descent; yet in Butler's lengthy compilation of the evidences of Buffon's evolutionism it is not cited at all! The volume containing it, says Butler, offers "little which throws additional light upon Buffon's opinion concerning the mutability of species"!^[19] In truth, it offers one of the best of the extremely few passages which give some plausibility at least to the theory that Buffon was continuously working towards an unqualified transformism and actually arrived at that position in his later life. But if he reached it (which his language just quoted does not quite justify us in declaring) he did so only in a transient mood. For, as we have already seen, in 1779, in the "Époques de la Nature," we once more find him asserting—though no longer upon the ground of the sterility of hybrids—that the "constitutive form" of each separate species is the same to-day as in "the earliest ages."

5. It is more important, and it is commonly easier, to determine what opinions a man's writings tended to encourage than to determine what opinions he actually held. Mind-reading is perhaps no essential part of the history of science. If, then, in conclusion, we raise the more important question with respect to Buffon, it is evident that his work both fostered and hindered the propagation of evolutionary ideas in biology. Earlier than any other except Maupertuis, he put the hypothesis of organic evolution before his contemporaries in a clear and definite form. He called to their attention, also, the facts of comparative anatomy which constitute one of the principal evidences for that hypothesis. Throughout the rest of the century we never cease to hear about the wonderful "unity of type" characteristic of the vertebrates and perhaps of all living things. It was this consideration which led Kant as near to evolutionism as he ever came; Herder and Goethe are full of it, though the former never admitted its full evolutionary consequences; and all, it is evident, learned it directly from Buffon. He, says Goethe, was the first to recognize eine ursprüngliche und allgemeine Vorzeichnung der Tiere. Buffon, moreover, once and for all inscribed upon the program of natural history, as its primary problems, the reduction of the number of separate species to a minimum, the derivation of highly divergent forms from a common origin through natural descent, and the discovery of the causes and methods of modification. He, finally, did more than any one else to habituate the mind of his time to a vastly (though not yet sufficiently) enlarged time-scale in connection with the history of organic nature, a necessary prerequisite to the establishment of transformism.

These were great steps in the progress of evolutionism. But it is equally true that Buffon probably did more than any other eighteenth century writer to check the progress of evolutionism. He did so partly by the authority which, for his contemporaries, attached to those personal utterances of his favorable to the doctrine of immutability. These utterances were far more numerous and more categorical than those which could be quoted on the other side; and they certainly were not taken as ironical by the average reader of the period. But, what is still more important, Buffon put into currency what passed for a scientific and serious argument against any wholesale theory of descent. In the eyes of many learned men of his own and later generations, perhaps his chief single contribution to science was his definition of species. This, as I have recently pointed out,^[20] was regarded as of immense importance by Kant, and was, indeed, the starting point and the controlling principle of that philosopher's biological speculations. "It is Buffon," wrote Flourens as late as 1844," who has given us the positive character of a species." Now before the Buffonian criterion of species was propounded, there already existed a tendency towards evolutionism, fostered by the principle of continuity and by such embryological conceptions as those of Maupertuis—a tendency to disregard species altogether and to infer from the variability of individuals to an unlimited and rather promiscuous mutability of the successive generations of living things. If it had not been for Buffon, transformism would probably have developed at first^[21] through the increase and diffusion of this tendency; and its development might well, in that case, have been more rapid. But when species came to be regarded as real "entities of nature," determined by the objective criterion of the sterility of hybrids, this somewhat too facile evolutionism received a check, and a certain presumption of the constancy of true species seemed to be created. This presumption had all the more force because it left room for a large measure of mutability in the case of varieties, and thus gave a sort of appeasement to the strong impulse towards genetic modes of thought which was already active in the mid-eighteenth century. But more than all this, Buffon, as we have seen, from the first managed to associate with his definition of species the assumption that the sterility of one kind of animal when crossed with another was a character that (unlike almost all others) could not have been produced in the course of descent with modification. And this supposition that the sterility of hybrids was incapable of an evolutionary explanation long remained a serious obstacle to the acceptance of the theory of descent, even with those little influenced by theological prejudices against the theory. We find even Huxley in 1862 troubled over the difficulty. In his Edinburgh lectures of that year "he warned his hearers of the one missing link in the chain of evidence—the fact that selective breeding has not yet produced species sterile to one another." The doctrine of descent was merely to be "adopted as a working hypothesis, . . . subject to the production of proof that physiological species may be produced by selective breeding."^[22] Since Buffon appears to have been the first to emphasize the notion of physiological species, and to give currency to the supposition that the sterility of hybrids affords a presumption against any thorough-going transformism, he must be regarded as having done more than almost any man of his time to counteract the tendency which he also, perhaps, did more than any other to promote.

1. "Hist. Nat.," IV., 1753, p. 383.
2. These, the most definite and decisive words on the subject to be found anywhere in Buffon 's writings, have been strangely disregarded by most of those who have discussed his attitude towards evolutionism. Samuel Butler can scarcely be acquitted of suppressing the passage, fatal to his theory. For he quotes in full the opening part of the passage, leaving off abruptly at the point where Buffon begins to introduce his serious objections to the theory of descent. Cf. "Evolution Old and New," p. 91.
3. P. 59 (italics mine).
4. P. 194.
5. P. 55.
6. Vol. XIII., p. i.
7. Vol. XIII., pp. vii, ix.
8. "Hist. Nat.," Supp., V., p. 27.
9. Supp., V., 1778, p. 3.
10. Vol. VI., pp. 59-60. I have borrowed Butler's excellent rendering of this passage.
11. Vol. XIV., 1766, p. 358.
12. "Hist, des Oiseaux," I., 1770, preface.
13. Cf. Lovejoy, "Some Eighteenth Century Evolutionists," Pop. Sci. Monthly, July, 1904, p. 248 n.
14. "Hist. Nat.," V., 1755, p. 252.
15. Just a year earlier we have found Buffon using the most exaggerated language possible about the changelessness of species.
16. Vol. XIV., p. 335; italics mine.
17. Supp., III., p. 20; the italics are Buffon's.
18. Supp., III., 1776, pp. 32-33.
19. "Evolution Old and New," p. 165.
20. Popular Science Monthly, January, 1911, pp. 37-38.
21. Cf. Lovejoy, Popular Science Monthly, July, 1904, p. 248.
22. Huxley's "Life and Letters," I., 193.