Agassiz holds that the influence of the centre of appearance is general and absolute. It extends to all the products of the soil as well as to those of fresh and salt waters. A country is just as much characterised by its plants and animals as by its human beings. In his opinion, an essentially local force seems to have produced all beings, or at least to have imprinted upon them a common mark. This generalisation was inevitable. Any one who wishes to attach a human race to each centre of appearance is compelled to localise in each one of them the original cause of all the animal and vegetable forms which are indigenous in it. For all living beings geographical coincidence must be absolute. Now there is generally no such coincidence. From the waters of a river to the banks which enclose it, the contrast may be striking. This is exactly what was shown by the discoveries of Agassiz himself in the ichthyology of the Amazon. To anyone who accepts the results published by the illustrious traveller, it is evident that this fauna may be divided into groups much more narrowly confined than those of terrestrial fauna. The same fact may be observed upon the shores of two seas separated by even a very narrow strip of land. The terrestrial fauna and flora are the same throughout the whole extent of the isthmus of Suez, whilst M. Edwards has not found a single species of crustacea common to both the Mediterranean and the Red Sea, and the study of annelids has led me to the same result. Moreover, the same region may be the centre of appearance for one class of animals, but by no means for another. Australia, for example, is one of the most characteristic centres for mammals, and stands alone from this point of view among the surrounding countries. With respect to insects Australia agrees, on the contrary, with New Zealand, New Caledonia and the neighbouring islands. I have borrowed this last fact from Lacordaire. It has the more value since this entomologist has multiplied the centres of appearance to a much greater extent than Agassiz, and has, therefore, made their characterisation easier. Thus the coincidence admitted by Agassiz, far from extending to all the organised beings of a region, does not even exist in certain cases between the different classes of animals alone. V. Agassiz divides the entire surface of the globe into nine great regions or _kingdoms_. I cannot here give in detail the numerous criticisms to which the fixed limits and characterisation of these centres are open. I shall confine myself to a few short remarks upon each.