[[p. 365]] The two most remarkable and most isolated groups of fruit-eating birds--the Parrots and the Pigeons--attain their maximum development, as regards beauty, variety, and number of species, in the same limited district, of which the great island of New Guinea forms the centre, and which I have proposed to call the Austro-Malayan subregion. It extends from the island of Celebes on the west to the Solomon Islands on the east, and includes the Moluccan and Timor groups. Its actual land-area is less than one-sixth that of Europe, yet it contains more than one-fourth of all the species of Pigeons that are known to exist. The islands west of Celebes, as far as Malacca and the Nicobar Islands and including the Philippines, are also rather rich in this family of birds. They form the Indo-Malayan subregion; and by combining the two we have in the Malay Archipelago considerably more than one-third of all the Pigeons that inhabit the earth. We can only vaguely speculate on the causes that have led to this peculiar distribution, since it would seem, at first sight, that the forests of Africa, of India, and especially of South America, would be equally well adapted to the development and support of these beautiful birds; and the fact that [[p. 366]] fruit-eating birds, as a whole, are more abundant in South America than in these islands, proves that their comparative scarcity cannot be attributed to a deficiency of appropriate food. It is to be noted, however, that the most striking superabundance of Pigeons, as well as of Parrots, is confined to the Austro-Malayan subregion, in which, although the most luxuriant forests everywhere clothe the country, and fruit-bearing trees, especially those of the Fig tribe, are very abundant, yet all the forest-haunting and fruit-eating mammals, such as Monkeys and Squirrels, are totally absent. But Monkeys, besides consuming vast quantities of fruit, are exceedingly destructive to eggs and young birds; and Pigeons, which build rude, open nests, and whose young are a long time helpless, must be more particularly exposed to their attacks. This is no doubt the reason why, in the dense forests of the Amazon, where Monkeys are most abundant, Pigeons are scarce or almost entirely absent; and in South America generally, it is to be observed that by far the larger number of the Pigeons inhabit the districts where Monkeys are almost or quite wanting--the mountains of Chili and of Mexico, the open plains of the Orinoko and La Plata, and the savannas of Central Brazil. The South American Pigeons are mostly ground-feeding species, and build in low bushes and thickets to which Monkeys rarely descend. In India and Africa, where Monkeys, especially the smaller kinds, are less abundant, true fruit-eating Pigeons occur, feeding and building on lofty trees, and protected to some extent by the green tints of their plumage. They form, however, in these countries but a small portion of the group, whereas more than two-thirds of the Pigeons of the Malay Islands are fruit-eaters of the genera Treron, Ptilonopus, and Carpophaga, which never descend to the ground, and are true denizens of the dense virgin forests. We may also remark that in these regions there are no great families of fruit-eating Passeres like the Tanagers and Chatterers of tropical America, whose place seems to be in some measure supplied by the Fruit-Pigeons, which, being generally larger birds, consume a vast quantity of fruit. The great development and rapid increase of these, unchecked by the competition of fruit-eating mammals, or by the attacks of arboreal carnivora, would [[p. 367]] perhaps, in the struggle for existence which is always most severe between creatures of a similar mode of life, prevent the increase of the smaller fruit-eaters; and we may thus understand how it is that in many of these islands Parrots and Pigeons form such a large proportion of the avifauna, and are by far the most prominent and characteristic of the living creatures that inhabit them.

    The classification of the Pigeons is a very difficult subject, and can probably only be satisfactorily effected by an examination of the anatomy of all the genera. They may, however, be very conveniently grouped into three great families:--1st. The Treronidæ, or Fruit-Pigeons, which have short legs with broad-soled, grasping feet, feed entirely on fruits, and never descend upon the ground; 2nd. The Columbidæ, or true Pigeons and Doves, which have larger feet and slenderer toes, and feed either on trees or on the ground; 3rd. The Gouridæ, or Ground-Pigeons, which have generally longer legs, feed always on the ground, run quickly, and only ascend trees to roost.

    The Treronidæ are entirely confined to the eastern hemisphere. A few species of the genus Treron are found in Africa, but the greater portion inhabit India and the western Malay Islands. These are beautiful birds, almost always of a yellowish or ashy-green colour, variegated with patches of bright yellow, purple, or chestnut, which are less vivid or altogether absent in the females. This genus may be considered to be almost confined to the Indian region, fourteen species occurring in India, ten in the islands of Java, Sumatra, and Borneo, and three in the Philippines; but on passing into the Australian region they diminish rapidly, two, which scarcely differ from those in the other islands, being found in Celebes, one in the Moluccas; and two in the islands between Java and Timor. In the next genus, Ptilonopus, the distribution is reversed, since New Guinea is their metropolis, whence they diminish in every direction, only one species occurring in Borneo and Sumatra, and the utmost limits of the genus being reached in the southern part of the Malay peninsula. In the Pacific islands and in the Moluccas they abound, many even of the smallest islands having their peculiar species. These are the smallest and most beautiful of [[p. 368]] the Fruit-Pigeons; their ground-colour is generally of a rich grass-green, diversified with bands and spots, caps, and shoulder-patches of the most vivid colours--crimson, pink, purple, white, or yellow in endless diversity. The genus Carpophaga, on the other hand, comprises the giants of the family. They also have their metropolis in New Guinea, but they extend further westward, two species occurring in India. Some of these birds have a deep booming note, which might almost be taken for the roar of a wild beast. Their gape and throat are so extensile that they can swallow very large fruits. In the Moluccas they devour the nutmegs, as soon as the fruits open, for the sake of the mace, which is digested off in the bird's stomach, and the seed disgorged entire.

    Looking at the whole family of Fruit-Pigeons, we find that fifty-four species are confined to the Austro-Malayan subregion, while twenty-eight inhabit the Indo-Malayan district, only three species (one of each genus) being common to the two. Beyond the Archipelago, fourteen species (all of the genus Treron) are found in India, and six of the same genus in Africa; thirty (of the genera Carpophaga and Ptilonopus) inhabit the various islands of the Pacific, and eight have been found in Australia and New Zealand. Even with our present imperfect knowledge of New Guinea we have fourteen species from that island, a larger number than are known to inhabit any other single tract of land, and plainly marking it out as the focus of the group. Of all the other islands Celebes is by far the richest in Fruit-Pigeons, containing ten species, seven of which are peculiar to it.

    The family Columbidæ is chiefly represented in the Archipelago by the genus Macropygia, which extends from the Himalayan Mountains to Australia and the Pacific Islands. Rarely is more than a single species found in any island, except in Java, which has three if not four species, and may therefore be considered the headquarters of the genus. These birds feed on the ground or on low bushes; and all are more or less of a chestnut-brown colour, and have long and graduated tails. They are of a weak structure and seldom take long flights. Turacœna and Reinwardtœna are two genera so closely allied to Macropygia that they have been often combined with it. The former, [[p. 369]] however, is characterized by the shorter tail, the bare orbits, and the black colour of the plumage. Three species only are known, which are widely scattered over the Austro-Malayan subregion, though each species is very local--one being confined to Celebes, another to Timor, and the third to the Solomon Islands. In the Moluccas, which occupy the space between the widely scattered localities of Turacœna, is found the single species of Reinwardtœna, which has a much stronger bill and longer tail, and has the whole under surface white, while the back is rich brown. The presence of these birds, so closely allied to Macropygia, in the Austro-Malayan subregion only, would lead us to suppose that this peculiar form of Pigeon is really most characteristic of that district, and that the preponderance of the species of Macropygia in Java is only due to some favourable local conditions. The beautiful metallic Pigeons forming the genus Ianthænas, and which seem to form a transition from the Macropygiine form to that of the true Pigeons, are found also in the Moluccas, New Guinea, and Timor, extending to the Pacific islands, and one species to Japan. The old-world genus Turtur has a few representative species in the Indo-Malay islands, but does not properly extend to the Australian region, as only stragglers have reached Timor along the chain of islands from Java, and those found in the Moluccas may perhaps have been introduced, as they have not extended to the easternmost islands or to New Guinea.

    The Gouridæ, or Ground-Pigeons, seem especially to abound in the Australian and American regions. Of the seven genera found in the Archipelago only two extend on to the continent of Asia (one species of each), while five are confined to the Austro-Malayan subregion and three to New Guinea, and several other peculiar genera inhabit Australia and the Pacific islands. Of the seventeen or eighteen species in the Archipelago no less than fourteen inhabit the Austro-Malayan subregion, and seven are found in New Guinea itself, although so little is yet known of that great island. Some of these are among the most remarkable of Pigeons. Trugon terrestris, by its stout hooked bill and strong legs, shows some approach to the wonderful Didunculus, the existing representative of the Dodo. [[p. 370]] Henicophaps has a long, straight, and powerful bill, like that of some of the larger Plovers. Calœnas nicobarica I believe to have spread westwards from New Guinea as far as the island where it was first found and from which it has derived its name. It has a massive body, with immense pectoral muscles, and very stiff and ample wings, and is thus capable of passing from island to island; and it is a remarkable fact that it is found almost entirely on small uninhabited islands, scattered at intervals over the four thousand miles of ocean between New Ireland and the Nicobar Islands. Over this wide range it presents no perceptible differences of form or colouring, which may be considered to indicate that migration still takes place at intervals, and by crossing the breeds in distinct islands, checks the formation of local races. Phlegœnas is another beautiful genus, scattered sparingly over a wide area; but here each island has a distinct species, showing that the causes that once favoured the distribution of the form have now ceased to act. Accordingly we find these birds to have a much weaker structure than Calœnas, and limited powers of flight. The magnificent Crown-Pigeons, the largest and most majestic of the whole order, are confined to the Papuan islands, where they take the place of the Curassows of South America. I have often seen these fine birds walking along the forest-paths in New Guinea, where the absence of carnivorous mammals, and the scarcity of large reptiles and of birds of prey, permit them to multiply unmolested. When disturbed, they fly up into the lowest branches of the nearest tree, in which situations they roost; but they spend the greater part of the day upon the ground, feeding on fallen fruits. The Geopeliæ are small, long-tailed Ground-Doves closely resembling Turtledoves in appearance, but having their nearest allies in several Australian species. They appear to have passed from Australia into Timor, and thence along the chain of islands into Java, as they are not found in any of the other parts of the archipelago. The green Ground-Doves of the genus Chalcophaps are the only ones which have a more extended distribution. All the species, however, are very closely allied; and the one which is found in India is so very similar to that of the western Malay islands, that [[p. 371]] its extension on to the continent may probably not have been of very ancient date. Indeed we have so many instances of the larger animals multiplying rapidly and becoming thoroughly acclimatized in countries very remote from their original home and often differing very widely from it in physical conditions, that I should be inclined to think that in this case, as in many others, the distribution of species has been modified by the agency of man. From a very remote date there must have been communication between Java and India, since the Hindoo religion had been established in the island for an unknown period when it was subverted by Mahommedanism in the fifteenth century; and it is highly probable that a bird so beautiful, and so easily caught and preserved, as the Chalcophaps javanica, should have been often carried to the continent, where a few escaping would soon stock a wide extent of country. The fact of this being the only Ground-Dove in all India, and that it so closely resembles the Javan bird that great doubts are entertained of its specific distinctness, renders the supposition of its recent introduction highly probable, since, in most other cases, the species of Java and those of India offer well-marked differences.

    If we now turn from the consideration of the separate families, genera, and species to the distribution of the Pigeons as a whole, we shall discover facts not less interesting. The total number of Pigeons now known to exist is about three hundred, or perhaps a few more; and of these the Malay Archipelago possesses no less than one hundred and eighteen. This number will seem especially large if we compare it with that representing the species of Pigeons in other countries. According to Jerdon's work on the Birds of India, only twenty-eight Pigeons are found in that country, exclusive of Ceylon and the countries east of the bay of Bengal. Australia possesses twenty-three species, Africa less than forty, while the vast continent of America has not more than eighty of these beautiful birds. These numbers show that the Malay Archipelago is preeminently the metropolis of the Pigeon tribe. It is now well known, however, that this part of the world belongs to two distinct zoological regions--the Indian and the [[p. 372]] Australian: and in these the Pigeons are very unequally distributed; for the western and larger portion (the Indo-Malayan subregion) contains nine genera and forty-three species, while the eastern and smaller portion (the Austro-Malayan subregion) has fifteen genera and eighty-four species. Here, therefore, the species of Pigeons become more condensed and more varied than in any other part of the globe: here is the focus of the order; and it was probably from this part of the world that the original dispersal and modification of the group chiefly took place. This condensation is carried to its greatest height in New Guinea, in which, although only a few points on its coast have been visited, no less than twenty-five species of Pigeons have been obtained.

    I believe, therefore, that the distribution of Pigeons in the Malay Archipelago fully confirms the results I have already arrived at from the study of other groups of birds, mammalia, and insects. These are, briefly, that this district is not one of the primary divisions of the globe, but that while one-half of it belongs to the Indian region, the other forms part of that of Australia; that the whole district may be further divided into groups of islands, the productions of which have a very close affinity--the Moluccan group being a satellite of New Guinea, while the Timor group is more closely connected with Australia; that Celebes is a very isolated and remarkable island, which, from the variety and peculiarity of its productions, appears to be the remnant of some more extensive land, which existed anterior to the present distribution of land and water in the surrounding regions; and that New Guinea must be looked upon as the remnant of a vast continent, now sunk beneath the waves of the Pacific. We find, also, that among the Indo-Malay islands (Sumatra, Java, and Borneo) Java is far the most isolated, possessing a considerable number of species peculiar to itself, while almost all those of Sumatra and Borneo are common to those two islands. We learn from this that what at first sight seems a very probable tradition of the Javans, the very recent separation of their island from Sumatra, is the reverse of truth since the evidence of the distribution of the Pittidæ, of the Parrots, and of the Pigeons among birds, of the Squirrels [[p. 373]] among mammals, and the Papilionidæ among insects, distinctly proves that, while all these islands have at no very remote geological epoch been united to the continent, yet the separation of Java was the earliest event, long subsequent to which a land communication existed between Sumatra and Borneo, although a far wider sea now separates them than the narrow strait which divides Sumatra from Java.

    In the following list of the Malayan species of Pigeons, which I have endeavoured to make as complete as possible, I have thought it necessary to refer, in most cases, only to Bonaparte's 'Conspectus,' where a full synonymy is given, and to a good figure. Wherever practicable, I have given the colour of the eyes, bill, and feet from my own notes, as well as the dimensions, taken in the flesh and thus indicating the true size of the bird. For the localities I have chiefly depended on my own observations, indicated by "(Wall.)" after the names of places where I myself observed the species; but I have also given such other localities as appeared to me trustworthy, with an indication of the authority. I have added a list of such Indian and Chinese species as belong to genera occurring in the Malay islands. Descriptions of four species, which seem to me to require separation, have been given, and a few notes on habits and synonymy are occasionally inserted. The table of the distribution of the species has been found useful in the preceding generalizations, and will enable the ornithologist to see at a glance what species have as yet been ascertained to inhabit each island. I have adopted the limits of the Archipelago which are pointed out in my paper "On the Physical Geography of the Malay Archipelago," printed in the Journal of the Royal Geographical Society for 1863. . . . [[concluding portion on species-level systematics omitted]]