The migrations of mammals are now well understood and an account of them will be found in the article ZOOLOGICAL REGIONS. The time taken in migration must vary enormously for different groups and only one case in which it can be determined has been described. W. D. Matthew has pointed out that when at the end of Meiocene times, North and South America became connected, northern animals, horses and mastodons, migrated into South America, and southern forms, such as the giant ground-sloths and glyptodonts, moved into the United States. It is then possible by comparing Pleistocene North and South American faunas, to show that during the period of their migration these animals underwent little if any evolutionary change, and that in South America Plio cene forms of horses and elephants are associated with Pleistocene ground-sloths, whilst in North America ground-sloths of Pliocene type are found with Pleistocene elephants. The implication is that the migration took a period comparable in length with the interval between the Upper Pliocene and the Lower Pleistocene. If it should prove to be generally true that whilst migrating an animal exhibits little evolutionary change it should follow that at any time the members of a group at its home will be more advanced than those at the periphery of the area of distribution. This conclusion seems to be justified by the condition of affairs at the present day.
the mammal-like reptiles, of the truth of Reichert's theory of the derivation of the mammalian ear-ossicles. (See REPTILIA.) The most striking case, however, is that which arises from the work of E. A. Stensio on Cephalaspis. In this animal, Stensio has been able to show that complete cranial nerves associated with gill pouches exist in the orbital region, and that the mouth is a small hole lying quite anteriorly in front of the profundus nerve. It had long ago been concluded by students of embryology that the ancestral vertebrates must have possessed these characters but the evidence on which they had relied was inconclusive and its validity had been denied by competent workers. Now in the light of Stensio's work it is quite clear that the original interpreta tion of the embryological data was sound. (See CYCLOSTOMATA.) The study of fossils is thus seen to be of great practical utility to geologists and is capable of affording information about the course of evolution which can be obtained in no other way. Its evidence is clear that evolution provides an adequate account of past and present faunas and the facts with which it deals serve as a con trol of evolutionary theory. Palaeontology forms the base of a rational zoo-geography and it is capable of giving invaluable assistance in the solution of many morphological problems.