these higher animals become replaced by allied species much more rapidly than the mollusca; and it is also pretty certain that the modification by which this replacement is effected takes place more rapidly when the two sets of individuals are isolated from each other, and especially when they are restricted to islands, where they are necessarily subject to distinct and pretty constant conditions, both physical and organic. It becomes therefore almost a certainty, that Siam and Java on the one hand, and Sumatra, Borneo, and Malacca on the other must have been brought into some close connexion, not earlier than the newer Pliocene period; but while the one set of countries were having their meeting, the other must have been by some means got out of the way. Before attempting to indicate the mode by which this might have been effected in accordance with what we know of the physical geography, geology, and vegetation of the several islands, it will be as well to complete our sketch of their zoological relations to each other, so as ascertain with some precision, what are the facts of distribution which we have to explain.

Malacca, Sumatra, and Borneo.-After having set apart the Philippine Islands and Javą, we have remaining two great islands and a peninsula, which, though separated by considerable arms of the sea, possess a fauna of wonderful uniformity having all the typical Malayan features in their full development. Their unity is indeed so complete, that we can find hardly any groups of sufficient importance by which to differentiate them from each other; and we feel no confidence that future discoveries may not take away what speciality they possess. One after another, species or genera once peculiar to Borneo or Sumatra have been found elsewhere; and this has gone to such an extent in birds, that hardly a peculiar genus and very few peculiar species are left in either island. Borneo however is undoubtedly the most peculiar. It possesses three genera of Mammalia not found elsewhere: Cynogale, a curious carnivore allied to the otters; with Dendrogale and Ptilocerus, small insectivora allied to Tupaia. It has Simia, the Orang

utan, and Paguma, one of the Viverridæ, in common with Sumatra ; as well as Rhinosciurus, a peculiar form of squirrel, and Hemigalea, one of the Viverridæ, in common with Malacca. Sumatra has only one genus not found in any other Malayan district-Nemorhedus, a form of antelope which occurs again in North India. It also has Siamanga in common with Malacca, Mydaus with Java, and Rhizomys with India. The Malay Peninsula seems to have no peculiar forms of Mammalia, though it is rich in all the characteristic Malay types.

The bats of the various islands have been very unequally collected, 36 species being recorded from Java, 23 from Sumatra, but only 16 each from Borneo and Malacca. Leaving these out of consideration, and taking into account the terrestrial mammals only, we find that Java is the poorest in species, while Borneo, Sumatra, and Malacca are tolerably equal; the numbers being 55, 62, 66, and 65 respectively. Of these we find that the species confined to each island or district are (in the same order) 6, 16, 5, and 6. It thus appears that Borneo is, in its mammalia, the most isolated and peculiar; next comes Sumatra, and then Malacca and Java, as shown by the following table.

This result differs from that which we have arrived at by the more detailed consideration of the fauna of Java; and it serves to show that the estimate of a country by the number of its peculiar genera and species alone, may not always represent its true zoological importance or its most marked features. Java, as we have seen, is differentiated from the other three districts by the absence of numerous types common to them all, and by its independent continental relations. Borneo is also well distinguished by its peculiar genera and specific types, yet it is at the same time more closely related to Sumatra and Malacca than is Java. The two islands have evidently had a very different history, which a detailed knowledge of their geology

would alone enable us to trace. Should we ever arrive at a fair knowledge of the physical changes that have resulted in the present condition, we shall almost certainly find that many of the differences and anomalies of their existing fauna and flora will be accounted for.

In Birds we hardly find anything to differentiate Borneo and Sumatra in any clear manner. Pityriasis and Carpococcyx, once thought peculiar to the former, are now found also in the latter; and we have not a single genus left to characterize Borneo except Schwaneria a peculiar fly-catcher, and Indicator, an African and Indian group not known to occur elsewhere in the Malay sub-region. Sumatra as yet alone possesses Psilopogon, a remarkable form of barbet, but we may well expect that it will be soon found in the interior of Borneo or Malacca; it also has Berenicornis, an African form of hornbill. The Malay Peninsula appears to have no genus peculiar to it, but it possesses some Chinese and Indian forms which do not pass into the islands. As to the species, our knowledge of them is at present very imperfect. The Malay Peninsula is perhaps the best known, but it is probable that both Sumatra and Borneo are quite as rich in species. With the exception of the genera noted above, and two or three others as yet found in two islands only, the three districts we are now considering may be said to have an almost identical bird-fauna, consisting largely of the same species and almost wholly of these together with closely allied species of the same genera. There are no well-marked groups which especially characterise one of these islands rather than the other, so that even the amount of speciality which Borneo undoubtedly exhibits as regards mammalia, is only faintly shown by its birds. The Pittida may perhaps be named as the most characteristic Bornean group, that island possessing six species, three of which are peculiar to it and are among the most beautiful birds of an unusually beautiful family. Yet Sumatra possesses two peculiar, and hardly less remarkable species.

In other classes of vertebrates, in insects, and in land-shells, our knowledge is far too imperfect to allow of our making any useful comparison between the faunas.

Banca.-We must, however note the fact of peculiar species occurring in Banca, a small island close to Sumatra, and thus offering another problem in distribution. A squirrel (Sciurus bangkanus) is allied to three species found in Malacca, Sumatra, and Borneo respectively, but quite as distinct from them all as they are from each other. More curious are the two species of Pitta peculiar to Banca; one, Pitta megarhynchus, is allied to the P. brachyurus, which inhabits the whole sub-region. and extends to Siam and China, but differs from it in its very large bill and differently coloured head; the other, P. bangkanus, is allied to P. cucullatus, which extends from Nepal to Malacca, and to P. sordidus, which inhabits both Borneo and Sumatra as well as the Philippines.

We have here, on a small scale, a somewhat similar problem to that of Java, and as this is comparatively easy of solution we will consider it first. Although, on the map, Banca is so very close to Sumatra, the observer on the spot at once sees that the proximity has been recently brought about. The whole southeast coast of Sumatra is a great alluvial plain, hardly yet raised above the sea level, and half flooded in the wet season. It is plainly a recent formation, caused by the washing down into a shallow sea of the débris from the grand range of volcanic mountains 150 miles distant. Banca, on the other hand is, though low, a rugged and hilly island, formed almost wholly of ancient rocks of apparently volcanic origin, and closely resembling parts of the Malay Peninsula and the intervening chain of small islands. There is every appearance that Banca once formed the extremity of the Peninsula, at which time it would probably have been separated from Sumatra by 50 or 100 miles of sea. Its productions should, therefore, most resemble those of Singapore and Malacca, and the few peculiar species it possesses will be due to their isolation in a small tract of country, surrounded by a limited number of animal and vegetable forms, and subject to the influence of a peculiar soil and climate. The parent species existing in such large tracts as Borneo or Sumatra, subjected to more varied conditions of soil, climate, vegetation, food, and enemies, would preserve, almost or quite

unchanged, the characteristics which had been developed under nearly identical conditions when the great island formed part of the continent. Geology teaches us that similar changes in the forms of the higher vertebrates have taken place during the Post-Tertiary epoch; and there are other reasons for believing that, under such conditions of isolation as in Banca, the change may have required but a very moderate period, even reckoned in years. We will now return to the more difficult problem presented by the peculiar continental relations of Java, as already detailed.

Probable Recent Geographical Changes in the Indo-Malay Islands. Although Borneo is by far the largest of the IndoMalay islands, yet its physical conformation is such that, were a depression to occur of one or two thousand feet, it would be reduced to a smaller continuous area than either Sumatra or Java. Except in its northern portion it possesses no lofty mountains, while alluvial valleys of great extent penetrate far into its interior. A very moderate depression, of perhaps 500 feet, would convert it into an island shaped something like Celebes; and its mountains are of so small an average elevation, and consist so much of isolated hills and detached ranges, that a depression of 2,000 feet would almost certainly break it up into a group of small islands, with a somewhat larger one to the north. Sumatra (and to a less extent Java) consists of an almost continuous range of lofty mountains, connected by plateaus from 3,000 to 4,000 feet high; so that although a depression of 2,000 feet would greatly diminish their size, it would probably leave the former a single island, while the latter would be separated. into two principal islands of still considerable extent. The enormous amount of volcanic action in these two islands, and the great number of conical mountains which must have been slowly raised, chiefly by ejected matter, to the height of 10,000 and 12,000 feet, and whose shape indicates that they have been formed above water, renders it almost certain that for long periods they have not undergone submersion to any considerable extent. In Borneo, however, we have no such evidences. No volcano,