This last consideration leads us up to the argument from Geographical Distribution. In considering the various local faunas and floras on the face of the globe, no point strikes one more forcibly than the fact that neither their similarities nor their dissimilarities can be accounted for by climate or physical conditions. The animals of South Africa do not in the least resemble the animals of the corresponding belt of South America; the Australian beasts and birds and trees are utterly unlike those of France and Germany; the fishes and crustaceans of the Pacific at Panama are widely different from those of the Caribbean at the same point, separated from them only by the narrow belt of intervening isthmus. On the other hand, within the same continuous areas of sea or land, however great the differences of physical conditions, we find everywhere closely related types in possession of the most distinct and varied situations. On the burning plains of La Plata we get the agouti and the bizcacha as the chief rodents; we ascend the Cordillera, and close to the eternal snows we discover, not hares and rabbits like those of Europe, but a specialised chilly mountain form of the same distinctly South American type. We turn to the rivers, and we see no musk-rat or beaver, but the coypu and capybara, slightly altered varieties of the original bizcacha ancestor. Australia has no wolf, but it has instead fierce and active carnivorous marsupials; it has no mice, but some of its tiny kangaroo-like creatures fulfil analogous functions in its animal economy. Everywhere the evidence points to the conclusion that local species have been locally evolved from pre-existing similar species. The oceanic isles, of which Darwin had had so large an experience, and especially his old friends the Galapagos, come in usefully for this stage of the question. They are invariably inhabited, as Darwin pointed out, and as Wallace has since abundantly shown in the minutest detail, by waifs and strays from neighbouring continents, altered and specialised by natural selection in accordance with the conditions of their new habitat. As a rule, they point back to the districts whence blow the strongest and most prevalent winds; and the modifications they have undergone are largely dependent upon the nature of the other species with which they have to compete, or to whose habits they must needs accommodate themselves. In such cases it is easy to see how far Darwin's special conception of natural selection helps to explain and account for facts not easily explicable by the older evolutionism of mere descent with modification.