There is also the phenomenon of vertical temperature-control. On this subject Dr A. R. Wallace has written (Ency. Brit., 9th ed., art. "Distribution"): "As we ascend lofty mountains, the forms of life change in a manner somewhat analogous to the changes observed in passing from a warm to a cold country. This change is, however, far less observable in animals than in plants; and it is so unequal in its action, and can so frequently be traced to mere change of climate and deficiency of food, that it must rank as a phenomenon of secondary importance. Vertical distribution among animals will be found in most cases to affect species rather than generic or family groups, and to involve in each case a mass of local details.… The same remarks apply to the bathymetrical zones of marine life. Many groups are confined to tidal, or shallow, or deeper waters; but these differences of habit are hardly geographical, but involve details, suited rather to the special study of individual groups." Temperature-control is therefore mainly a factor which has acted independently in the different zoological regions of the globe, and as such demands little or no further mention in a general sketch of the present nature.
The same remark will apply in the case of the influence of humidity on distribution, and also as regards "station." To illustrate the latter we may take the instances of the European squirrel and the chamois, the former of which is found only in wooded districts and is entirely absent from the open plains, while the latter occurs only in the isolated mountain ranges of the Pyrenees, the Alps, the Apennines and the Caucasus. The distributional area of both may, however, be regarded as including Europe generally, so that these local restrictions of range have nothing to do with the wider problems of distribution.
Very different is the case with regard to geographical barriers to the free dispersal of terrestrial animals. It should be observed, however, that even these act with different degrees of intensity in the case of different groups. From the fact that the great majority of them are oviparous, reptiles, whose powers of dispersal in the adult state are generally as restricted as those of mammals, have an advantage over the latter in that their eggs may be carried long distances on floating timber down rivers and thence across the ocean, or may even be occasionally transported by birds. The eggs of batrachians, like those of fresh-water fishes, will in some cases at any rate withstand being frozen, and hence conceivably may be transported by floating ice. Adult insects may be carried in the same manner as the eggs of reptiles. After all, however, such unusual means of transport are probably of no great importance; and it seems most likely that the varying features in the geographical distribution of different groups of animals are due much more to differences in the dates of radiation, or dispersal of those groups, than to varying degrees of facility for overcoming natural geographical barriers to dispersal.
The greatest barriers of all are formed by the ocean and the larger rivers; and from the former factor it follows that zoological regions coincide to a considerable extent—although by no means altogether—with the main geographical (as distinct from political) divisions of the earth's surface. In the main, mammals and other nonvolant terrestrial animals are debarred from crossing anything more than comparatively narrow channels of the sea, while even these and the larger rivers form a more or less effectual barrier to the dispersal of the great majority of the species. Hence it results that oceanic islands are usually devoid of such forms of life; while it may be laid down, as a general rule, that the existence of nearly allied types of terrestrial animals in countries now separated by stretches of sea implies a former land-connexion between them. There are, however, in many cases great difficulties in determining the nature of such connexions, largely owing to the fact that we are still in the dark as to whether the dispersal of many groups of animals has taken place down the lines of the present continents from north to south or equatorially by means of belts of land long since swallowed up by the ocean. In this connexion it may be remarked, as tending against the old idea of the radiation of all the modern groups of terrestrial animals from the north towards the south, that there is decisive evidence to prove the existence during the Tertiary period (so far at least as mammals are concerned) of certain great centres of development, and in some instances, at all events, also of radiation, in the southern hemisphere; one of these developmental centres being in Africa a second in South America, and a third in Australia.
To the general law that straits and arms of the sea form an effectual barrier to the dispersal of the larger land-animals, and more especially mammals, certain exceptions may be pleaded. Jaguars have, for instance, been known to cross the Rio de la Plata, while tigers constantly swim from island to island in the delta of the Ganges and probably also in the Malay Archipelago, and a polar bear has been observed swimming twenty miles away from land in Bering Sea. Deer, certain antelopes, pigs and elephants are also good swimmers; while hippopotamuses and crocodiles—especially the latter—can cross channels of considerable width. The great tropical and subtropical rivers also carry down masses of floating soil or large trees upon which mammals and reptiles are borne, and although in many or most instances such are swept out to sea and their occupants drowned, in other instances they may be stranded upon the opposite bank or shore where their living freight can effect a landing. Such instances, however, cannot be very frequent, and they cannot affect widely sundered countries, owing to the lack of food supplies. Moreover, supposing a mammal to have reached a new land, unless it happened to be a pregnant female, or unless another individual of the opposite sex be similarly stranded, it would eventually die without progeny. Even in the case of a pregnant female, there is no certainty that the offspring, if but one, would be a male; and even supposing this to be the case, the progeny might perish from the attacks of other animals or from inbreeding. On the whole, it may be said, that instances of such methods of dispersal must be relatively few and can affect only countries not very widely sundered. The most important case that can be cited is the occurrence of a pig and an extinct hippopotamus in Madagascar, which probably reached that island by swimming from Africa. As a rule, a strait like that separating Ceylon from India may be considered an effectual barrier to the dispersal of large land-animals.
Although the Rio de la Plata has effectually prevented the amphibious carpincho from reaching Argentina, deserts form even more impassable barriers than large rivers, the Sahara having prevented the North African fauna from reaching the heart of that continent. High and continuous mountain-ranges are likewise most effective in restricting the range of animals; this being more especially the case when, like the Himalaya, their trend is equatorial instead of, as in the case of the Rocky Mountains and the Andes, from north to south in the direction of the main continental extension. Forests also present great obstacles to animal migration, although this is to a great extent of a local nature and comes, in fact, under the category of "station." Indeed, there appears to be no instance of the separation of one zoological region from another by forest alone.
Lastly it should be mentioned that ice may serve as a factor in the dispersal of animals by acting as a bridge between different land-areas; and at some period this means of communication may have aided in the great migrations of animals that have taken place between the Old and the New World by way of what is now Bering Sea.
I. Terrestrial Distribution
The zoological regions recognized by Dr A. R. Wallace in 1876, which are in the main identical with those Zoological Regions.proposed by Dr P. L. Sclater in 1858, and are chiefly based on the distribution of birds and mammals, are as follows:—
1. Palaearctic, which includes Europe to the Azores and Iceland, temperate Asia from the high Himalaya and west of the Indus, with Japan, and China from Ningpo and to the north of the watershed
