heredity we must assume a communication of corporeal elements, vehicles of that heredity, from both sides. And, as Maupertuis also observed, the supposed fact by which Harvey had justified this singular conclusion had already been rendered more than questionable by later investigations of Verheyen.
Having established these essential principles to his satisfaction, Maupertuis proceeded to formulate an hypothesis concerning the nature of the fundamental physical process presupposed by the facts of heredity, on the one hand, and of variation, on the other. The formation of an embryo must, he conceived, be due to the combination in a new organic union, of a great number of living corporeal particles, derived usually from both parents, each of which particles carries with it a sort of organic memory (souvenir) of the life of the organism to which it formerly belonged, and thereby tends to unite with the other particles in such a way as to produce a new organism of the same species. This process of recombination of already living particles was held by Maupertuis to be governed by something more than the laws of mechanism; embryogeny was for him no mere process of juxtaposition, under the laws of gravitation, of so many inert atoms; the ultimate units of the newly constituted living being all possess their own self-contained law of development, and their own distinctive selective affinities for certain other units. None the less, purely mechanical displacements of parts also take place; and to these in part, he supposed, is due the occurrence of monstrous forms, and many of the more ordinary variations from the hereditary specific type. Moreover, the elementary units which, coming from the parents, combine to form the embryonic offspring, in part carry with them a similar sort of organic memory of the particular and individual characters of the parent, and so tend to develop those characters in that offspring; but in part also they are free from this tendency, and carry with them rather the traits of more remote ancestors (atavism), and some of them may even be wholly independent of hereditary predetermination. It is these especially which, in Maupertuis's hypothesis, constitute the explanation of the general tendency to variation in animals, which he recognized to the full. If one must have a further explanation why the transmitted corpuscles tend to reproduce the characters of the parents, Maupertuis suggests that perhaps there enters into the embryo a separate germ from each part of the body of the parent of which the character is reproduced; in other words, he proposes a hypothesis similar to the Dar-
- 'Vénus Physique,' Pt. II., ch. 5. We must suppose 'que la liqueur séminale de chaques espèce d'animau, contient une multitude innombrable de parties propres à former par leurs assemblages des animaux de la même espèce.
- Loc. cit. It is likewise to be assumed 'que dans la liqueur séminale de chaque individu, les parties propres à former des traits semblables à ceux de cet individu sont celles qui d'ordinaire sont en plus grand nombre, et qui ont le plus d'affinité; quoiqu'il y en ait beaucoup d'autres pour des traits différents.'