Nor is it less probable that the characteristic types of Australian Mammalia were already developed in that region in Miocene times.

But Austro-Columbia presents difficulties from which Australia is free; Camelidae and Tapiridae are now indigenous in South America as they are in Arctogaea; and, among the Pliocene Austro-Columbian mammals, the Austro-Columbian genera Equus, Mastodon, and Machairodus are numbered.  Are these Postmiocene immigrants, or Praemiocene natives?

Still more perplexing are the strange and interesting forms Toxodon, Macrauchenia, Typotherium, and a new Anoplotherioid mammal (Homalodotherium) which Dr. Cunningham sent over to me some time ago from Patagonia.  I confess I am strongly inclined to surmise that these last, at any rate, are remnants of the population of Austro-Columbia before the Miocene epoch, and were not derived from Arctogaea by way of the north and east.

The fact that this immense fauna of Miocene Arctogaea is now fully and richly represented only in India and in South Africa, while it is shrunk and depauperized in North Asia, Europe, and North America, becomes at once intelligible, if we suppose that India and South Africa had but a scanty mammalian population before the Miocene immigration, while the conditions were highly favourable to the new comers.  It is to be supposed that these new regions offered themselves to the Miocene Ungulates, as South America and Australia offered themselves to the cattle, sheep, and horses of modern colonists.  But, after these great areas were thus peopled, came the Glacial epoch, during which the excessive cold, to say nothing of depression and ice-covering, must have almost depopulated all the northern parts of Arctogaea, destroying all the higher mammalian forms, except those which, like the Elephant and Rhinoceros, could adjust their coats to the altered conditions.  Even these must have been driven away from the greater part of the area; only those Miocene mammals which had passed into Hindostan and into South Africa would escape decimation by such changes in the physical geography of Arctogaea.  And when the northern hemisphere passed into its present condition, these lost tribes of the Miocene Fauna were hemmed by the Himalayas, the Sahara, the Red Sea, and the Arabian deserts, within their present boundaries.  Now, on the hypothesis of evolution, there is no sort of difficulty in admitting that the differences between the Miocene forms of the mammalian Fauna and those which exist at present are the results of gradual modification; and, since such differences in distribution as obtain are readily explained by the changes which have taken place in the physical geography of the world since the Miocene epoch, it is clear that the result of the comparison of the Miocene and present Fauna is distinctly in favour of evolution.  Indeed I may go further.  I may say that the hypothesis