The Dutch naturalist Zollinger, who resided a long time on the island of Bali, informs us that its productions completely assimilate with those of Java, and that he is not aware of a single animal found in it which does not inhabit the larger island. During the few days which I stayed on the north coast of Bali on my way to Lombock, I saw several birds highly characteristic of Javan ornithology. Among these were the yellow-headed weaver (Ploceus hypoxantha), the black grasshopper thrush (Copsychus amoenus), the rosy barbet (Megalaema rosea), the Malay oriole (Oriolus horsfieldi), the Java ground starling (Sturnopastor jalla), and the Javanese three-toed woodpecker (Chrysonotus tiga). On crossing over to Lombock, separated from Bali by a strait less than twenty miles wide, I naturally expected to meet with some of these birds again; but during a stay there of three months I never saw one of them, but found a totally different set of species, most of which were utterly unknown not only in Java, but also in Borneo, Sumatra, and Malacca. For example, among the commonest birds in Lombock were white cockatoos and three species of Meliphagidae or honeysuckers, belonging to family groups which are entirely absent from the western or Indo-Malayan region of the Archipelago. On passing to Flores and Timor the distinctness from the Javanese productions increases, and we find that these islands form a natural group, whose birds are related to those of Java and Australia, but are quite distinct from either. Besides my own collections in Lombock and Timor, my assistant Mr. Allen made a good collection in Flores; and these, with a few species obtained by the Dutch naturalists, enable us to form a very good idea of the natural history of this group of islands, and to derive therefrom some very interesting results.
The number of birds known from these islands up to this date is: 63 from Lombock, 86 from Flores, and 118 from Timor; and from the whole group, 188 species. With the exception of two or three species which appear to have been derived from the Moluccas, all these birds can be traced, either directly or by close allies, to Java on the one side or to Australia on the other; although no less than 82 of them are found nowhere out of this small group of islands. There is not, however, a single genus peculiar to the group, or even one which is largely represented in it by peculiar species; and this is a fact which indicates that the fauna is strictly derivative, and that its origin does not go back beyond one of the most recent geological epochs. Of course there are a large number of species (such as most of the waders, many of the raptorial birds, some of the kingfishers, swallows, and a few others), which range so widely over a large part of the Archipelago that it is impossible to trace them as having come from any one part rather than from another. There are fifty-seven such species in my list, and besides these there are thirty-five more which, though peculiar to the Timor group, are yet allied to wide-ranging forms. Deducting these ninety-two species, we have nearly a hundred birds left whose relations with those of other countries we will now consider.
If we first take those species which, as far as we yet know, are absolutely confined to each island, we find, in:
Lombock 4 belonging to 2 genera, of which 1 is Australian, 1 Indian.
Flores 12 " 7 " 5 are " 2 "
Timor 42 " 20 " 16 are " 4 "
The actual number of peculiar species in each island I do not suppose to be at all accurately determined, since the rapidly increasing numbers evidently depend upon the more extensive collections made in Timor than in Flores, and in Flores than in Lombock; but what we can depend more upon, and what is of more special interest, is the greatly increased proportion of Australian forms and decreased proportion of Indian forms, as we go from west to east. We shall show this in a yet more striking manner by counting the number of species identical with those of Java and Australia respectively in each island, thus:
In Lombock. In Flores. In Timor.
Javan birds... . 33 23 11
Australian birds.. 4 5 10
Here we see plainly the course of the migration which has been going on for hundreds or thousands of years, and is still going on at the present day. Birds entering from Java are most numerous in the island nearest Java; each strait of the sea to be crossed to reach another island offers an obstacle, and thus a smaller number get over to the next island. [The names of all the birds inhabiting these islands are to be found in the "Proceedings of the Zoological Society of London" for the year 1863.] It will be observed that the number of birds that appear to have entered from Australia is much less than those which have come from Java; and we may at first sight suppose that this is due to the wide sea that separates Australia from Timor. But this would be a hasty and, as we shall soon see, an unwarranted supposition. Besides these birds identical with species inhabiting Java and Australia, there are a considerable number of others very closely allied to species peculiar to those countries, and we must take these also into account before we form any conclusion on the matter. It will be as well to combine these with the former table thus:
In Lombock. In Flores. In Timor.
Javan birds........ ... 33 23 11
Closely allied to Javan birds.. 1 5 6
Total.............. 34 28 17
Australian birds......... 4 5 10
Closely allied to Australian birds 3 9 26
Total..... ......... 7 14 36
We now see that the total number of birds which seem to have been derived from Java and Australia is very nearly equal, but there is this remarkable difference between the two series: that whereas the larger proportion by far of the Java set are identical with those still inhabiting that country, an almost equally large proportion of the Australian set are distinct, though often very closely allied species. It is to be observed also, that these representative or allied species diminish in number as they recede from Australia, while they increase in number as they recede from Java. There are two reasons for this, one being that the islands decrease rapidly in size from Timor to Lombock, and can therefore support a decreasing number of species; the other and the more important is, that the distance of Australia from Timor cuts off the supply of fresh immigrants, and has thus allowed variation to have full play; while the vicinity of Lombock to Bali and Java has allowed a continual influx of fresh individuals which, by crossing with the earlier immigrants, has checked variation.
To simplify our view of the derivative origin of the birds of these islands let us treat them as a whole, and thus perhaps render more intelligible their respective relations to Java and Australia.