III. In spite of these dogmatic assertions, when it comes to application of any kind whatever, Agassiz, like Lamarck in former times, and Darwin in our own day, is obliged to use the word _species_ in the sense in which it is employed by so many others. In the memoir, from which I have already quoted, animal and vegetable _species_ are constantly being discussed. Their geographical distribution serves as a foundation to the theory of human origins. The author admits that they could not have arisen upon one and the same point of the globe; that the centres of creation were numerous, and that the species diverging from these centres give to the actual flora and fauna all their characteristic features. Up to this point Agassiz has only accepted the doctrine of centres of creation, a doctrine entirely French in origin, having been formulated by Desmoulins and developed by M. Edwards. What is due to Agassiz is the reproduction, in the name of science, of a theory at first proposed by La Peyrère in the name of theology: giving to man the whole world as his original home: the admission that the human races originated in the same places as the groups of animal and vegetable species, and the connection of one of these races with each centre of creation; the multiplication of the number of human creations to such a degree as to profess that “man was created by nations,” endowed from the first with all their distinctive characters, and each speaking its own special language. There is, at first sight, no absurdity in the idea itself, nothing at all contradictory to anything which we have as yet met with. We have seen above that physiology leads to the conclusion that “human groups are _to all appearance_ descended from one primitive pair.” It goes no further than that. Anyone who confines himself to inferences drawn from this order of facts might, therefore, accept the theory of Agassiz as, it is true, a very gratuitous hypothesis, but convenient in order to account for the distribution and actual diversity of human types. This is no longer the case when we turn to another branch of the natural sciences, _zoological_ and _botanical geography_. We then can easily prove that the theories of Agassiz tend to make an exception of man, to place him at variance with the general laws of the geographical distribution of all other organised beings, and, consequently, that they are false. IV. I fully agree with the views of Agassiz, as far as _centres of creation_, or rather _centres of appearance_ are concerned.