being due to a radical difference of type, and therefore not indicative of climate. The early European flora seems to have been a portion of that which now exists only in the tropical and sub-tropical lands of the Eastern Hemisphere; and, as much of this flora still survives in Australia, Tasmania, Japan, and the Cape of Good Hope, it does not necessarily imply more than a warm and equable temperate climate. The early North American flora, on the other hand, seems to have been essentially the same in type as that which now exists there, and which, in the Miocene period, was well represented in Europe; and it is such as now flourishes best in the warmer parts of the United States. But whatever conclusion we may arrive at on the question of climate, there can be no doubt as to the distinctness of the floras. of the ancient Nearctic and Palearctic regions; and the view derived from our study of their existing and extinct faunasthat these two regions have, in past times, been more clearly separated than they are now-receives strong support from the unexpected evidence now obtained as to the character and mutations of their vegetable forms, during so vast an epoch as is comprised in the whole duration of the Tertiary period.

The general phenomena of the distribution of living animals, combined with the evidence of extinct forms, lead us to conclude that the Palearctic region of early Tertiary times was, for the most part, situated beyond the tropics, although it probably had a greater southward extension than at the present time. It certainly included much of North Africa, and perhaps reached far into what is now the Sahara; while a southward extension of its central mass may have included the Abyssinian highlands, where some truly Palearctic forms are still found. This is rendered probable by the fossils of Perim Island a little further east, which show that the characteristic Miocene fauna. of South Europe and North India prevailed so far within the tropics. There existed, however, at the extreme eastern and .western limits of the region, two extensive equatorial land-areas, our Indo-Malayan and West African sub-regions--both of which must have been united for more or less considerable periods with the northern continent. They would then have received

from it such of the higher vertebrates as were best adapted for the peculiar climatal and organic conditions which everywhere prevail near the equator; and these would be preserved, under variously modified forms, when they had ceased to exist in the less favourable and constantly deteriorating climate of the north. At later epochs, both these equatorial lands became united to some part of the great South African continent (then including Madagascar), and we thus have explained many of the similarities presented by the faunas of these distant, and generally very different countries.

During the Miocene period, when a subtropical climate prevailed over much of Europe and Central Asia, there would be no such marked contrast as now prevails between temperate and tropical zones; and at this time much of our Oriental region, perhaps, formed a hardly separable portion of the great Palearctic land. But when, from unknown causes, the climate of Europe became less genial, and when the elevation of the Himalayan chain and the Mongolian plateau caused an abrupt difference of climate on the northern and southern sides of that great mountain barrier, a tropical and a temperate region were necessarily formed; and many of the animals which once roamed over the greater part of the older and more extensive region, now became restricted to its southern or northern divisions respectively. Then came the great change we have already described (vol. i. p. 288), opening the newly-formed plains of Central Africa to the incursions of the higher forms of Europe; and following on this, a still further deterioration of climate, resulting in that marked contrast between temperate and tropical faunas, which is now one of the most prominent features in the distribution of animal as well as of vegetable forms.

It is not necessary to go into any further details here, as we have already, in our discussion of the origin of the fauna of the several regions, pointed out what changes most probably occurred in each case. These details are, however, to a great extent speculative; and they must remain so till we obtain as much knowledge of the extinct faunas and past geological history of the southern lands, as we have of those of Europe and North

America But the broad conclusions at which we have now arrived seem to rest on a sufficiently extensive basis of facts; and they lead us to a clearer conception of the mutual relations and comparative importance of the several regions than could be obtained at an earlier stage of our inquiries.

If our views of the origin of the several regions are correct, it is clear that no mere binary division-into north and south, or into east and west-can be altogether satisfactory, since at the dawn of the Tertiary period we still find our six regions, or what may be termed the rudiments of them, already established. The north and south division truly represents the fact, that the great northern continents are the seat and birth-place of all the higher forms of life, while the southern continents have derived the greater part, if not the whole, of their vertebrate fauna from the north; but it implies the erroneous conclusion, that the chief southern lands-Australia and South America-are more closely related to each other than to the northern continent. The fact, however, is that the fauna of each has been derived, independently, and perhaps at very different times, from the north, with which they therefore have a true genetic relation; while any intercommunion between themselves has been comparatively recent and superficial, and has in no way modified the great features of animal life in each. The east and west division, represents-according to our views-a more fundamental diversity; since we find the northern continent itself so divided in the earliest Eocene, and even in Cretaceous times; while we have the strongest proof that South America was peopled from the Nearctic, and Australia and Africa from the Palearctic region: hence, the Eastern and Western Hemispheres are the two great branches of the tree of life of our globe. But this division, taken by itself, would obscure the facts-firstly, of the close relation and parallelism of the Nearctic and Palæarctic regions, not only now but as far back as we can clearly trace them in the past; and, secondly, of the existing radical diversity of the Australian region from the rest of the Eastern Hemisphere.

Owing to the much greater extent of the old Palearctic region (including our Oriental), and the greater diversity of

Mammalia it appears to have produced, we can have little doubt that here was the earliest seat of the development of the vertebrate type; and probably of the higher forms of insects and land-molluscs. Whether the Nearctic region ever formed one mass with it, or only received successive immigrations from it by northern land-connections both in an easterly and westerly direction, we cannot decide; but the latter seems the most probable supposition. In any case, we must concede the first rank to the Palearctic and Oriental regions, as representing the most important part of what seems always to have been the Great Continent of the earth, and the source from which all the other regions were supplied with the higher forms of life. These once formed a single great region, which has been since divided into a temperate and a tropical portion, now sufficiently distinct; while the Nearctic region has, by deterioration of climate, suffered a considerable diminution of productive area, and has in consequence lost a number of its more remarkable forms. The two temperate regions have thus come to resemble each other more than they once did, while the Oriental retains more of the zoological aspect of the great northern regions of Miocene times. The Ethiopian, from having been once an insular region, where lower types of vertebrates alone prevailed, has been so overrun with higher types from the old Palearctic and Oriental lands that it now rivals, or even surpasses, the Oriental region in its representation of the ancient fauna of the great northern continent. Both of our tropical regions of the Eastern Hemisphere possess faunas which are, to some extent, composite, being made up in different proportions of the productions of the northern and southern continents,-the former prevailing largely in the Oriental, while the latter constitutes an important feature in the Ethiopian fauna. The Neotropical region has probably undergone great fluctuations in early times; but it was, undoubtedly, for long periods completely isolated, and then developed the Edentate type of Mammals and the Formicaroid type of Passerine birds into a variety of forms, comparable with the diversified Marsupials of Australia, and typical Passeres of the Eastern Hemisphere.

It has, however, received successive infusions of higher types from the north, which now mingle in various degrees with its lower forms. At an early period it must have received a low form of Primates, which has been developed into the two peculiar families of American monkeys; while its llamas, tapirs, deer, and peccaries, came in at a later date, and its opossums and extinct horses probably among the latest. The Australian region alone, after having been united with the great northern continent at a very early date (probably during the Secondary period) has ever since remained more or less completely isolated; and thus exhibits the development of a primeval type of mammal, almost wholly uninfluenced by any incursions of a later and higher type. In this respect it is unique among all the great regions of the earth.

We see, then, that each of our six regions has had a history of its own, the main outlines of which we have been able to trace with tolerable certainty. Each of them is now characterised as it seems to have been in all past time of which we have any tolerably full record-by well-marked zoological features; while all are connected and related in the complex modes we have endeavoured to unravel. To combine any two or more of these regions, on account of existing similarities which are, for the most part, of recent origin, would obscure some of the most important and interesting features of their past history and present condition. And it seems no less impracticable to combine the whole into groups of higher rank; since it has been shown that there are two opposing modes of doing this, and that each of them represents but one aspect of a problem, which can only be solved by giving equal attention to all its aspects.

For reasons which have been already stated, and which are sufficiently obvious, we have relied almost exclusively on the distribution of living and extinct mammalia, in arriving at these conclusions. But we believe they will apply equally to elucidate the phenomena presented by the distribution of all terrestrial organisms, when combined with a careful consideration of the