only species found in such localities being either small forms which might have been carried on floating timber, or such as have been introduced by human agency. This absence of mammalian life in oceanic islands extends even to New Zealand, where the indigenous mammals comprise only two peculiar species of bats, the so-called Maori rat having been introduced by man.
One of the leading features in mammalian distribution is the fact that the Monotremata, or egg-laying mammals, are exclusively confined to Australia and Papua, with the adjacent islands. The marsupials also attain their maximum development in Australia (“Notogaea” of the distributionists), extending, however, as far west as Celebes and the Moluccas, although in these islands they form an insignificant minority among an extensive placental fauna, being represented only by the cuscuses (Phalanger), a group unknown in either Papua or Australia. Very different, on the other hand, is the condition of things in Australia and Papua, where marsupials (and monotremes) are the dominant forms of mammalian life, the placentals being represented (apart from bats, which are mainly of an Asiatic type) only by a number of more or less aberrant rodents belonging to the mouse-tribe, and in Australia by the dingo, or native dog, and in New Guinea by a wild pig. The dingo was, however, almost certainly brought from Asia by the ancestors of the modern natives; while the Papuan pig is also in all probability a human introduction, very likely of much later date. The origin of the Australasian fauna is a question pertaining to the article Zoological Distribution. The remaining marsupials (namely the families Didelphyidae and Epanorthidae) are American, and mainly South and Central American at the present day; although during the early part of the Tertiary period representatives of the first-named family ranged all over the northern hemisphere.
The Insectivora (except a few shrews which have entered from the north) are absent from South America, and appear to have been mainly an Old World group, the only forms which have entered North America being the shrew-mice (Soricidae) and moles (Talpidae). The occurrence of one aberrant group (Solenodon) in the West Indies is, however, noteworthy. The family with the widest distribution is the Soricidae, the Talpidae being unknown in Africa. The tree-shrews (Tupaiidae) are exclusively Asiatic, whereas the jumping-shrews (Macroscelididae) are equally characteristic of the African continent. Madagascar is the sole habitat of the tenrecs (Centetidae), as is Southern Africa of the golden moles (Chrysochloridae). It is, however, important to mention that an extinct South American insectivore, Necrolestes, has been referred to the family last mentioned; and even if this reference should not be confirmed in the future, the occurrence of a representative of the order in Patagonia is a fact of considerable importance in distribution.
The Rodentia have a wider geographical range than any other order of terrestrial mammals, being, as already mentioned, represented by numerous members of the mouse-tribe (Muridae) even in Australasia. With the remarkable exception of Madagascar, where it is represented by the Nesomyidae, that family has thus a cosmopolitan distribution. Very noteworthy is the fact that, with the exception of Madagascar (and of course Australia) the squirrel family (Sciuridae) is also found in all parts of the world. Precisely the same may be said of the hares, which, however, become scarce in South America. On the other hand, the scaly-tailed squirrels (Anomaluridae), the jumping-hares (Pedetidae), and the strand-moles (Bathyergidae) are exclusively African; while the sewellels (Haplodontidae) and the pocket-gophers (Geomyidae) are as characteristically North American, although a few members of the latter have reached Central America. The beavers (Castoridae) are restricted to the northern hemisphere, whereas the dormice (Gliridae) and the mole-rats (Spalacidae) are exclusively Old World forms, the latter only entering the north of Africa, in which continent the former are largely developed. The jerboa group (Dipodidae, or Jaculidae) is also mainly an Old World type, although its aberrant representatives the jumping-mice (Zapus) have effected an entrance into Arctic North America. Porcupines enjoy a very wide range, being represented throughout the warmer parts of the Old World, with the exception of Madagascar (and of course Australasia), by the Hystricidae, and in the New World by the Erethizontidae. Of the remaining families of the Simplicidentata, all are southern, the cavies (Caviidae), chinchillas (Chinchillidae), and degus (Octodontidae) being Central and South American, while the Capromyidae are common to southern America and Africa, and the Ctenodactylidae are exclusively African. The near alliance of all these southern families, and the absence of so many Old World families from Madagascar form two of the most striking features in the distribution of the order. Lastly, among the Duplicidentata, the picas (Ochotonidae or Lagomyidae) form a group confined to the colder or mountainous regions of the northern hemisphere.
Among the existing land Carnivora (of which no representatives except the introduced dingo are found in Australasia) the cat-tribe (Felidae) has now an almost cosmopolitan range, although it only reached South America at a comparatively recent date. Its original home was probably in the northern hemisphere; and it has no representatives in Madagascar. The civet-tribe (Viverridae), on the other hand, which is exclusively an Old World group, is abundant in Madagascar, where it is represented by peculiar and aberrant types. The hyenas (Hyaenidae), at any rate at the present day, to which consideration is mainly limited, are likewise Old World. The dog-tribe (Canidae), on the other hand, are, with the exception of Madagascar, an almost cosmopolitan group. Their place of origin was, however, almost entirely in the northern hemisphere, and not improbably in some part of the Old World, where they gave rise to the bears (Ursidae). The latter are abundant throughout the northern hemisphere, and have even succeeded in penetrating into South America, but, with the exception of the Mediterranean zone, have never succeeded in entering Africa, and are therefore of course unknown in Madagascar. The raccoon group (Procyonidae) is mainly American, being represented in the Old World only by the pandas (Aelurus and Aeluropus), of which the latter apparently exhibits some affinity to the bears. The birthplace of the group was evidently in the northern hemisphere—possibly in east Central Asia. The weasel-tribe (Mustelidae) is clearly a northern group, which has, however, succeeded in penetrating into South America and Africa, although it has never reached Madagascar.
The extinct creodonts, especially if they be the direct descendants of the anomodont reptiles, may have originated in Africa, although they are at present known in that continent only from the Fayum district. Elsewhere they occur in South America and throughout a large part of the northern hemisphere, where they appear to have survived in India to the later Oligocene or Miocene.
In the case of the great order, or assemblage, of Ungulata it is necessary to pay somewhat more attention to fossil forms, since a considerable number of groups are either altogether extinct or largely on the wane.
So far as is at present known, the earliest and most primitive group, the Condylarthra, is a northern one, but whether first developed in the eastern or the western hemisphere there is no sufficient evidence. The more or less specialized Litopterna and Toxodontia, as severally typified by the macrauchenia and the toxodon, are, on the other hand, exclusively South American. With the primitive five-toed Amblypoda, as represented by the coryphodon, we again reach a northern group, common to the two hemispheres; but there is not improbably some connexion between this group and the much more specialized Barypoda, as represented by Arsinöitherium, of Africa. The Ancylopoda, again, typified by Chalicotherium, and characterized by the claw-like character of the digits, are probably another northern group, common to the eastern and western hemispheres.
Recent discoveries have demonstrated the African origin of the elephants (Proboscidea) and hyraxes (Hyracoidea), the latter group being still indeed mainly African, and in past times also limited to Africa and the Mediterranean countries. As regards the elephants (now restricted to Africa and tropical Asia), there appears to be evidence that the ancestral mastodons, after having developed from African forms probably not very far removed from the Amblypoda, migrated into Asia, where they gave rise to the true elephants. Thence both elephants and mastodons reached North America by the Bering Sea route; while the former, which arrived earlier than the latter, eventually penetrated into South America.
The now waning group of Perissodactyla would appear to have originally been a northern one, as all the three existing families, rhinoceroses (Rhinocerotidae), tapirs (Tapiridae), and horses (Equidae), are well represented in the Tertiaries of both halves of the northern hemisphere. If eastern Central Asia were tentatively given as the centre of radiation of the group, this might perhaps best accord with the nature of the case. Rhinoceroses disappeared comparatively early from the New World, and never reached South America. In Siberia and northern Europe species of an African type survived till a comparatively late epoch, so that the present relegation of the group to tropical Asia and Africa may be regarded as a modern feature in distribution. Horses, now unknown in a wild state in the New World, although still widely spread in the Old, attained a more extensive range in past times, having successfully invaded South America. On the other hand, in common with the rest of the Perissodactyla, they never reached Madagascar. In addition to the occurrence of their fossil remains almost throughout the world, the former wide range of the tapirs is attested by the fact of their living representatives being confined to such widely sundered areas as Malaysia and tropical America.
The Artiodactyla are the only group of ungulates known to have been represented in Madagascar; but since both these Malagasy forms—namely two hippopotamuses (now extinct) and a river-hog—are capable of swimming, it is most probable that they reached the island by crossing the Mozambique Channel. As regards the deer-family (Cervidae), which is unknown in Africa south of the Sahara, it is quite evident that it originated in the northern half of the Old World, whence it reached North America by the Bering Sea route, and eventually travelled into South America. More light is required with regard to the past history of the giraffe-family (Giraffidae), which includes the African okapi and the extinct Indian Sivatherium, and is unknown in the New World. Possibly, however, its birthplace may prove to be Africa; if so, we shall have a case analogous to that of the African elephant, namely that while
