portions of the earth's surface, and one which, as far as the Tertiary period at least is concerned, was of an exceptional character. Such a cause exists in the great and recent physical change known as "the Glacial epoch." We have proof in both Europe and North America, that just about the time these large animals were disappearing, all the northern parts of these continents were wrapped in a mantle of ice; and we have every reason to believe that the presence of this large quantity of ice (known to have been thousands of feet if not some miles in thickness) must have acted in various ways to have produced alterations of level of the ocean as well as vast local floods, which would have combined with the excessive cold to destroy animal life. There is great difference of opinion among geologists and physicists as to the extent, nature, and duration of the Glacial epoch. Some believe it to have prevailed alternately in the northern and southern hemispheres; others that it was simultaneous in both. Some think there was a succession of cold periods, each lasting many thousands of years, but with intercalated warm periods of equal duration; others deny that there is any evidence of such changes, and maintain that the Glacial epoch was one continuous period of arctic conditions in the temperate zones, with some fluctuations perhaps but with no regular alternations of warm periods. Some believe in a huge ice-cap covering the whole northern hemisphere from the pole to near 50° north latitude in the eastern, and 40° in the western hemisphere; while others impute the observed effects either to glaciers from local centres, or to floating icebergs of vast size passing over the surface during a period of submersion.

Without venturing to decide which of these various theories will be ultimately proved to be correct, we may state, that there is an increasing belief among geologists in the long duration of this ice-period, and the vast extent and great thickness attained by the ice-sheet. One of the most recent, and not the least able, of the writers on this question (Mr. Belt) shows strong reasons for adopting the view that the ice-period was simultaneous in both hemispheres; and he calculates that the vast amount of water abstracted from the ocean and locked up

in mountains of ice around the two poles, would lower the general level of the ocean about 2,000 feet. This would be equivalent to a general elevation of the land to the same amount, and would thus tend to intensify the cold; and this elevation may enable us to understand the recent discoveries of signs of glacial action at moderate elevations in Central America and Brazil, far within the tropics. At the same time, the weight of ice piled up in the north would cause the land surface to sink there, perhaps unequally, according to the varying nature of the interior crust of the earth; and since the weight has been removed land would rise again, still somewhat irregularly; and thus the phenomena of raised beds of arctic shells in temperate latitudes, are explained.

Now, it is evident, that the phenomena we have been considering of the recent changes of the maminalian fauna in Europe, North America, South Temperate America, and the highlands of Brazil-are such as might be explained by the most extreme views as to the extent and vastness of the ice-sheet, and especially as to its simultaneous occurrence in the northern and southern hemispheres; and where two such completely independent sets of facts are found to combine harmoniously, and supplement each other on a particular hypothesis, the evidence in favour of that hypothesis is greatly strengthened. An objection that will occur to zoologists, may here be noticed. If the Glacial epoch extended over so much of the temperate and even parts of the tropical zone, and led to the extinction of so many forms of life even within the tropics, how is it that so much of the purely tropical fauna of South America has maintained itself, and that there are still such a vast number of forms, both of mammalia, birds, reptiles, and insects, that seem organized for an exclusive existence in tropical forests? Now Mr. Belt's theory, of the subsidence of the ocean to the extent of about 2,000 feet, supplies an answer to this objection; for we should thus have a tract of lowland of an average width of some hundreds of miles, added to the whole east coast of Central and South America. This tract would, no doubt, become covered with forests as it was slowly formed, would enjoy a perfectly

tropical climate, and would thus afford an ample area for the continued existence and development of the typical South American fauna; even had glaciers descended in places so low as what is now the level of the sea, which, however, there is no reason to believe they ever did. It is probable too, that this low tract, which all round the Gulf of Mexico would be of considerable width, offered that passage for intermigration between North and South America, which led to the sudden appearance in the former country in Post-Pliocene times, of the huge Megatheroids from the latter; a migration which took place in opposite directions as we shall presently show.

The birth-place and migrations of some mammalian families and genera.-We have now to consider a few of those cases in which the evidence already at our command, is sufficiently definite and complete, to enable us to pronounce with some confidence as to the last movements of several important groups of mammalia.

Primates.-The occurrence in North America of numerous forms of Lemuroidea, forming two extinct families, which are believed by American paleontologists to present generalized features of both Lemurida and Hapalidæ, while in Europe only Lemurine forms allied to those of Africa have occurred in deposits of the same age (Eocene), renders it possible that the Primates may have originated in America, and sent one branch to South America to form the Hapalida and Cebidæ, and another to the Old World, giving rise to the lemurs and true apes. But the fact that apes of a high degree of organization occur in the European Miocene, while in the Eocene, a monkey believed to have relations to the Lemuroids and Cebidæ has also been discovered, make it more probable that the ancestral forms of this order originated in the Old World at a still earlier period. The absence of any early tertiary remains from the tropical parts of the two hemispheres, renders it impossible to arrive at any definite conclusions as to the origin of groups which were, no doubt, always best developed in tropical regions.

Carnivora. This is a very ancient and wide-spread group, the families and genera of which had an extensive range in very

early times. The true bears (Ursus) are almost the only important genus that seems to have recently migrated. In Europe it dates back to the Older Pliocene, while in North America it is Post-Pliocene only. Bears, therefore, seem to have passed into America from the Palearctic region in the latter part of the Pliocene period. They probably came in on the north-west, and passed down the Andes into South America, where one isolated species still exists.

Ungulata.-Horses are very interesting. In Europe they date back under various forms to the Miocene period, and true Equus to the Older Pliocene. In North America they are chiefly Pliocene, true Equus being Post-Pliocene, with perhaps one or two species Newer Pliocene; but numerous ancestral forms date back to the Miocene and Eocene, giving a more perfect "pedigree of the horse" than the European forms, and going back to a more primitive type-Orohippus. In South America, Equus is the only genus, and is Post-Pliocene or at most Newer Pliocene. While, therefore, the ancient progenitors of the Equidae were common to North America and Europe, in Miocene and even Eocene times, true horses appear to have arisen in the Palearctic region, to have passed into North America in the latter part of the Pliocene period, and thence to have spread over all suitable districts in South America. They were not, however, able to maintain themselves permanently in their new territory, and all became extinct; while in their birth-place, the Old World, they continue to exist under several varied forms.

True tapirs are an Old World group. They go back to the Lower Miocene in Europe, while in both North and South America they are exclusively Post-Pliocene. They occur in France down to the Newer Pliocene, and must, about that time, have entered America. The land connection by which this and so many other animals passed between the Old and New Worlds in late Tertiary times, was almost certainly in the North Pacific, south of Behring's Straits, where, as will be seen by our general map, there is a large expanse of shallow water, which a moderate elevation would convert into dry land, in a sufficiently temperate latitude.

The peccary (Dicotyles), now a characteristic South American genus, is a recent immigrant from North America, where it appears to have been developed from ancestral forms of swine dating back to the Miocene period.

Antelopes are an Old World type, but a few of them appear to have entered North, and reached South America in late Pliocene times. Camels, strange to say, are a special North American type, since they abounded in that continent under various ancient. forms in the Miocene period. Towards the end of that period they appear to have entered eastern Asia, and developed into the Siberian Merycotherium and the North Indian Camelus, while in the Pliocene age the ancestral llamas entered South America.

Cervidae are a wide-spread northern type in their generalized form, but true deer (Cervus) are Palearctic. They abounded in Europe in Miocene times, but only appear in North and South America in the later Pliocene and Post-Pliocene periods.

True oxen (Bovina) seem to be an Oriental type (Miocene), while they appear in Europe only late in the Pliocene period, and in America are confined to the Post-Pliocene.

Elephants (Elephantida) are an Old World type, abounding in the Miocene period in Europe and India, and first appearing in America in Post-Pliocene or later Pliocene times. Ancestral forms, doubtfully Proboscidean (Dinocerata), existed in North America in the Eocene period, but these became extinct without leaving any direct descendants, unless the Brontotheride and rhinoceroses may be so considered.

Marsupials are almost certainly a recent introduction into South and North America from Asia. They existed in Europe in Eocene and Miocene times, and presumably over a considerable part of the Old World; but no trace of them appears in North or South America before the Post-Pliocene period.

Edentata. These offer a most curious and difficult problem. In South America they abound, and were so much more numerous and varied in the Post-Pliocene and Pliocene, that we may be sure they lived also in the preceding Miocene period. A few living Edentates are scattered over Africa and Asia, and