Page:Popular Science Monthly Volume 79.djvu/476

abandoned, and to which he never returned.^[1] Its most characteristic point was the contention that nature knows only individuals and that species are entia rationis merely. The most characteristic point of nearly all his subsequent references to the subject is the contention that species are real entities, definable in exact and strictly objective terms, and necessary to take account of in any study of natural history.

This change already was manifest in the second volume, published in the same year as the preliminary discourse (1749). In this volume Buffon propounded his celebrated definition of species, which was destined to have so great an influence upon the biological ideas of the later eighteenth century.

We should regard two animals as belongng to the same species if, by means of copulation, they can perpetuate themselves and preserve the likeness of the species; and we should regard them as belonging to different species if they are incapable of producing progeny by the same means. Thus the fox will be known to be a different species from the dog, if it proves to be the fact that from the mating of a male and a female of these two kinds of animals no offspring is born; and even if there should result a hybrid offspring, a sort of mule, this would suffice to prove that fox and dog are not of the same species—inasmuch as this mule would be sterile (ne produirait rien). For we have assumed that, in order that a species might be constituted, there was necessary a continuous perpetual and unvarying reproduction (une production continue, perpétuelle, invariable)—similar, in a word, to that of the other animals.^[2]

This language, it will be observed, implies not only that species are real entities, but also that they are constant and invariable entities. The same implication may be found again later in the volume; Buffon thus concludes the exposition of his embryological hypotheses—which embraced a theory of pangenesis:

There exists, therefore, a living matter, universally distributed through all animal and vegetal substances, which serves alike for their nutrition, their growth and their reproduction. . . . Reproduetion takes place only through the same matter's becoming superabundant in the body of the animal or plant. Each part of the body then sends off (renvoie) the organic molecules which it can not admit. Each of these particles is absolutely analogous to the part by which it is thrown off, since it was destined for the nourishment of that part. Then, when all the molecules sent off by all the parts of the body unite, they necessarily form a small body similar to the first, since each molecule is similar to the part from which it comes. It is in this way that reproduction takes place in all species. . . . There are, therefore, no preexisting germs, no germs contained within one another ad infinitum; but there is an organic matter, always active, always ready to be shaped and assimilated and to produce beings similar to those which receive it. Animal or vegetable species, therefore, can never, of themselves, disappear (s'épuiser). So long as any individuals belonging to it

1. Rádl's account, already quoted, of Buffon's attitude towards transformism and towards the conception of species, is apparently based chiefly upon the first volume. For virtually all of Buffon 's views, except his early and quickly repudiated one, Rádl's statement is almost the exact reverse of the truth.
2. "Hist. Nat.," Vol. II., 1749, p. 10.