374 MAMMALIA [DISTEIBUTION. Arctic Eight Whale (Balsena mysticetus) has been conclu sively shown to be limited in its range to the region of the northern circumpolar ice, and no corresponding species has been met with in the southern hemisphere. In this case, not only temperature, but also the peculiarity of its mode uf feeding, may be the cause. The Narwhal and the Beluga have a very similar distribution, though the latter occasionally ranges farther south. The Hyperoodons are restricted to ihe North Atlantic, never entering, as far as is yet known, the tropical seas. Other species are exclu sively tropicul or austral in their range. One of the true Whalebone Whales (Neobalxna marginata) has only been met with hitherto in the seas round Australia and New Zealand, a large Ziphioid (Berardius arnouxi) only near the last-named islands. The Cetacea are not limited to the ocean, or even to salt water, some entering large rivers for considerable distances, and some being exclusively fluviatile. One species of Platanista is extensively distributed throughout nearly the whole of the river systems of the Ganges, Brahmaputra, and Indus, ascending as high as there is water enough to swim in, but apparently never passing out to sea. The individuals inhabiting the Indus and the Ganges must therefore have been for long ages isolated without de veloping any definite distinguishing anatomical characters, for those by which the supposed P. indi was formerly separated from P. g angelica, have been shown by Anderson to be of no constant value. Orcella iiuminalis appears to be limited to the Irawaddy river, and at least two distinct species of Dolphin, belonging to different genera, are found in the waters of the upper Amazon. It is remarkable, how ever, that none of the great lakes or inland seas of the world are, according to our present knowledge, inhabited by Cetaceans. A regular seasonal migration has been ob served in many of the oceanic Cetacea, especially those in habiting the North Atlantic, but further observations upon this subject are still much needed. The great difference in the manner of life of the Sirenia, as compared with that of the Cetacea, causes a correspond ing difference in their geographical distribution. Slow in their movements, and feeding exclusively upon vegetable substances, water-grasses, or fuci, they are confined to rivers, estuaries, or coasts where these grow, and are not denizens of the open sea, although of course there is a possibility of accidental transport by the assistance of oceanic currents across considerable distances. Of the three genera existing within historic times, one (Manatus) is exclusively confined to the shores of the tropical Atlantic and the rivers entering into it, individuals scarcely specifi cally distinguishable being found both on the American and the African side of the ocean. The Dugong (Halicore) is distributed in different colonies, at present isolated, throughout the Indian Ocean from Arabia to North Aus tralia. The Rhytiiia or Northern Sea-Cow was, for some time before its extinction, limited to a single island in the extreme north of the Pacific Ocean. The Pinnipeds, although capable of traversing long reaches of ocean, are less truly aquatic than the last two groups, always resorting to the land or to extensive ice floes for the purposes of breeding. The geographical range of each species is generally more or less restricted, usually according to climate, arj they are mostly inhabitants either of the Arctic or Antarctic seas and adjacent temperate regions, very few being found within the tropics. For this reason the northern and the southern species are for the most part quite distinct. In fact, the only known exception is the case of a colony of the Sea-Elephant (Macrorhimis leoninus), the general range of which is in the southern hemisphere, inhabiting the coast of California. Even in this case a different specific name has been given to the northern form, but the characters by which it is distinguished are not of great importance, and probably,, except for the abnormal geographical distribution, would never have been discovered. The most remarkable circumstance connected with the distribution of the Pinnipeds is the presence of members of the order in the three isolated great lakes or inland seas of Central Asia the Caspian, Aral, and Baikal which, notwithstanding their long isolation, have varied but slightly from species now inhabiting the Polar Seas. 3. Terrestrial Mammals. One of the most important Terres facts connected with the present distribution of terrestrial trial mammals, but one of which the cause is sufficiently a obvious, is their entire absence, except where introduced by the direct agency of man, from all oceanic islands, including even the great New Zealand group. Another, equally striking, but less easily explained, is the very marked isolation as regards its mammalian fauna of the. Australian region of zoologists. When once the narrow neutral ground on the border line between this and the Oriental region is passed, there is not found, notwithstand ing the vast extent of land it comprises, a single indigenous- placental or monodelphous mammal, except a few species of a single family (Muridse) of the very wide spread order Rodentia., the Wild Dog or Dingo having been in all probability introduced. On the other hand, the members of the other two subclasses, the Prototheria or Ornitho- delphia and the Metatheria or Didelphia, are almost entirely restricted to this region. It might have been said entirely, but for the presence of one family (Diddpkidx) of the latter group in America. The Eutheria or Monodelphia are distributed through out the remaining geographical regions, as described in the article DISTEIBUTION, and in many cases form valuable indications by which the natural boundaries of these divisions have been traced. Many anomalies, otherwise inexplicable in the present range of families and. species, have been cleared up by the study of their distribution in former geological epochs, a subject irk relation to which the present facts of distribution should always be studied. There are few lines of evidence so conclusive as this in favour of the existing species bein^ modified lineal descendants of those which have lived iii previous times upon the earth. 1 HISTORY OF THE MAMMALIA IN PAST TIMES. As already intimated, such knowledge as we yet possess of the history of mammals in past times is of very recent growth, and is still extremely incomplete. The very rapid advances which have been made in the last few years, especially in consequence of the explorations of rich fossiliferous beds in North America, have not only com pletely changed the present aspect of the science, but give such promise for the future that any sketch which ve might now attempt of this branch of the subject could only be regarded as representing a transient phase of knowledge. It will be well, however, to gather together in this place the leading facts now ascertained with regard to the mosfc ancient forms, as, owing to the uncertainty of their rela tionship with any of the existing orders, they will be most conveniently treated of separately, while the ascertained, facts relating to the geological history of the forms more nearly allied to those now living will be more appropriately described under the account of the different groups into which the class may now be divided. 1 On this subject, see A. Murray, Geographical Distribution of Mammals, 1866 ; and especially A. R. Wallace, The Geographical
Distribution of Animals, 2 vols., 1876, and Island Life, 1881.