All who confine themselves to the data of observation and experiment will see at once that all animal and vegetable species could not have originated upon any one spot of the globe. The former shows us, in various regions, different types and species, living naturally in countries which present almost precisely the same conditions of existence. The latter teaches us that we can transport the greater number of species from one region to another, and that they will prosper there, if the conditions of existence are the same; that, on the contrary, arctic and tropical species cannot, even temporarily, be submitted to the action of the same conditions; that neither can withstand the action of a temperate climate. It is impossible with all these facts to avoid the conclusion that plants and animals had several points of appearance. But if I accept this doctrine as the only one reconcilable with facts, it is upon the condition of adopting it entirely, and as developed by studies upon the geographical distribution of all living beings. Now, works of this kind are numerous at the present time. For all phanerogamous plants we have the work of M. Ad. de Candolle, which has been a standard work ever since its appearance. Animals have not yet had their de Candolle. The great work of M. Alphonse Edwards will partly fill up this gap for the more southern regions of the globe. In the meantime, important investigations have been made in some of the principal classes. Buffon, by his admirable researches upon the geography of mammals, opened the way, in which he has been followed by the two Geoffroy Saint-Hilaires, Fr. Cuvier, and Andrew Murray; Dumeril and Bibron have studied reptiles from the same point of view; Fabricius, Latreille, Macley, Spence, Kirby, and Lacordaire have done the same for insects; M. Milne Edwards has worked out the distribution of the crustacea; I have endeavoured to do as much for the annelids. Finally, a great number of works bearing upon the lower groups have long been known to science, and Agassiz himself has largely contributed to increase our knowledge in this direction. A certain number of general facts stand out from this mass of research, which we call _laws_. If the theory of Agassiz is true, it ought to agree with these laws. Now the disagreement is apparent from the outset. Let us prove, in the first place, that this theory includes two very distinct ideas: that of the original cosmopolitanism of the human species; and, secondly, that of a geographical connection between the human race and the animal or vegetable groups observed in a common centre. Let us examine the truth and error contained in this last statement.