One of the most striking features is the extreme paucity of rasorial birds--peafowl, junglefowl, pheasants, partridges, or any of the natural genera into which these divide, and which are all well represented in the Arakan Hills. The next point is the highly specialised character of the ornis, for, excluding waders and swimmers, more than a third of the species are peculiar to the islands; while still more remarkable is the extent to which it is localised in the several groups between which is nowhere a break of more than 80 miles. Even more noteworthy are the details: for instance, the Andaman _Hypothymis_, which, as a rule, is a very distinct form, is replaced in the Nicobars by one which, although not precisely identical with the Indian form, is far more closely allied to this than the Andaman _Tytleri_. Each group has its distinct harrier eagle, red-cheeked paroquet, oriole, sunbird, and bulbul. Two woodpeckers are peculiar to the Andamans, but neither extends to the Cocos or Nicobars. The latter group possesses three distinct but closely allied species of _Astur_, each confined to separate islands. So far as the species not peculiar to the islands are concerned, the influence of the Indian sub-region has vastly predominated; and if we look to the genera the preponderance is still more marked, and thus it seems difficult to avoid the conclusion that the ornis has altogether a very far stronger affinity with that of the Indian region than with those of either the Indo-Burmese or the Indo-Malayan. Yet this involves great difficulties, for if we take Port Blair as a centre we shall find that its average distance in all directions north and east from Tenasserim (where the Indo-Malayan fauna predominates), and north of this from the Indo-Burmese sub-region, is less than half its distance from the nearest point of the Indian sub-region. That so many of the characteristic birds of the Arakan Hills, especially the Rasores, should be entirely wanting, we may partly account for by the supposition that the mountains and the chain of islands never were continuous, and that the same agency that raised the Arakan Hills only raised portions of their continuation above the sea-level, so that, therefore, the islands have never been connected with Pegu. If, however, the groups first appeared and have ever since remained as detached islands, it is inconceivable how the great bulk of the work of colonisation should have gone on from a region so distant while so little should have been done from others less than half as far away. Colonisation in no ordinary sense, however, can explain these facts. But the case of Sumatra, which, although only 80 miles distant from Great Nicobar, and itself the first link of a great chain, teems right up to Acheen Head with species unknown to the Nicobars, is perfectly comprehensible in the light of our knowledge of the deep sea existing between it and these latter islands.[221]