We will now take a brief survey of the principal features of these six regions—as shown in the accompanying chart (Plate I., p. 16) and their most characteristic mammal-forms.

1.-AUSTRALIAN REGION

Extent.-Australia, New Guinea, and Moluccas up to Wallace's line, New Zealand, and the numerous islands of the Pacific.

Characteristics.-Absence of nearly all Eutherian Mammals, except a few Rodents and Bats; presence of six distinct families of Marsupials with one hundred species, and the only two known forms of Monotremes.

2.-NEOTROPICAL REGION

Name.-veòs, new, and трóπikos, i.e., tropical land of the New World.

Extent.-America, south of the Isthmus of Tehuantepec, and the West Indies.

Characteristics.-Monkeys of the peculiar families Cebida and Hapalida; absence of Frugivorous Bats, and presence of Vampires (Phyllostomatide); abundance of the Porcupine family; absence of Insectivores and Civets, also of Elephants; presence of Tapirs; no Ruminants except Deer and Lamas; presence of Sloths, Ant-eaters, and Armadilloes; one family of MarsupialsOpossums.

3.-ETHIOPIAN REGION

Name.-'Aillores, ancient name for negroes.

Extent.-Africa, south of the Atlas; Arabia up to the Persian Gulf, and Madagascar.

Characteristics. - Chimpanzee and other Monkeys; absence of Bears and Deer; presence of Lion, African Elephant, Hyrax, Rhinoceros, Hippopotamus, Wart-hog, numerous Antelopes, Giraffe, Pangolin, Ant-bear-general richness in large and highly-organised Ungulates.

4.-ORIENTAL REGION

Extent.-Southern Asia, south of the Palearctic Region, and islands of Indian Archipelago down to Wallace's line, including Celebes.

Characteristics.-Orangs, Gibbons, and other peculiar Monkeys. Flying Lemur, Tiger, and other cats, Indian Elephant, Rhinoceros, Malayan Tapir, Manis.

Generally, it may be said that the peculiar forms of the Oriental Region are fewer than in the Ethiopian Region, and that the Oriental Region has Bears, Deer, and Tapirs, which are wanting in the latter.

5.-NEARCTIC REGION

Name.-veòs, new, and аρктoя, north, i.e., northern district of the New World.

Extent.-North America, down to the Isthmus of Tehuantepec.

Characteristics.-General mammal-fauna, very like that of the Palearctic Region, but mixed up with endemic

forms and intruders from the south.

Bears, Beavers,

Sheep, and Deer similar; Prong - buck, Pouched Mice, and Musquash peculiar; Raccoon and Opossum, probably derived from the south.

6.-PALEARCTIC REGION

Name.-Taλaiòs, ancient, and аρêтоs, north, as embracing the whole northern area of the Old World.

Extent.-Whole of the Eastern Hemisphere north of a line on the south of the Atlas, and running eastward through the south of Palestine and Persia, along the Himalayas, through Central Asia and the centre of China to the Pacific.

Characteristics.-Absence of Monkeys, Lemurs, and Frugivorous Bats; abundance of Carnivores-Ounce, Lynxes, Wolves, Foxes, Bears, and Weasels; RodentsMarmots, Beavers, Pikas; Ungulates - Sheep, Deer, Chamois, and Musk-deer; no Elephants nor Hyrax.

This division of the Earth's surface into six regions was first proposed by one of the authors of the present work in an essay on the distribution of the Class of birds read before the Linnæan Society in 1857 (9). It was further elaborated and upheld in an address given to Section D. of the British Association at the Bristol Meeting in 1875 (10), and in a Paper published in The Ibis in 1891 (11). The same system was adopted by Mr. Wallace in his standard work on "Geographical Distribution" (13), and was there shown to be applicable to the other principal groups of terrestrial animals. Moreover, the names then bestowed on the six great primary Regions are now in general use among naturalists in all countries. Mr.

Wallace, who has devoted many pages to the discussion of this subject,1 has come to the conclusion that, admitting that these six regions are not precisely equal in rank, and that some of them are more isolated than the others, they are in geographical equality, compactness of area, and facility of definition beyond all comparison better than any others which have been suggested for the purpose of facilitating the study of geographical distribution.

Notwithstanding Mr. Wallace's strong support, however, it is right to say that this system has not been universally accepted. Professor Huxley (6) in 1868 proposed to separate the world into two divisions-Arctogæa and Notogæa, the former containing the Nearctic, Palearctic, Ethiopian, and Oriental Regions, and the latter the Neotropical and Australian Regions. He adopted the Nearctic, Palearctic, Ethiopian, and Oriental Regions as subdivisions of Arctogæa, and only stipulated for the formation of a Circum-polar province independent of the Nearctic and Palearctic Regions. Notogæa Professor Huxley divided into three provinces (a) the Austro-Columbian (=the Neotropical), (b) the Australian (=the Australian Region minus New Zealand), and (c) the New Zealand province.

From this it will be seen that Professor Huxley's scheme does not really diverge materially from the system here employed; the chief points of difference being (a) the uniting together of the Australian and Neotropical Regions into Notogæa; (b) the formation of independent Circumpolar and New Zealand provinces. With regard to the first point, almost the only bond of union between the Australian and Neotropical Regions, so far as mammals, at any rate, are concerned, is the presence of Marsupials 1 "Geographical Distribution of Animals," vol. i., chap. iv.

in both regions. But the Marsupials of Australia seem to have but a very remote connection with those of South America, and there is at present no palæontological evidence of the former occurrence of the Australian forms, or of forms allied to them, outside of Australia itself.1 On the other hand, the presence of fossil opossums (Didelphyidae) in the Eocene beds of France, shows that the South American forms were formerly more widely spread.

Professor Huxley has also cited the Parrots (Psittacomorpha) "as helping, together with the three-toed Ratitæ, to bind together the widely-separated portions of the south world." But on referring to the account of the distribution of the Parrots in Salvadori's recently published catalogue (8), it will be found that out of the six families into which he divides the group, five are practically confined to the Australian Region, and that the remaining one is widely spread throughout the tropical regions of both hemispheres. The most recent arrangement of this family, therefore, gives little support to Professor Huxley's arguments.

Looking, again, to the distribution of the Ratita (wingless birds), we find the Neotropical form (the Rhea) more closely connected with the Ostrich, the Ethiopian form, and that they both differ considerably from the Emus, Cassowaries, and Kiwis, the three Australian representatives of this order. Thus, then, there seems to be

1 Recently Señor Ameghino has described from the Santa Cruz beds of Patagonia, which are probably of Eocene age, certain fossil mammals which he has referred to the Dasyurida, one of the Australian families. Again, Mr. Thomas' Canolestes (see P. Z. S. 1895), is also believed to be allied to the Australian Diprotodonts. If these relationships should turn out to be correct, it will indicate further evidence of some connection between South America and Australia, though at a considerably remote epoch of geological time.