738 BIRDS [DISTRIBUTION. lie between that group and the coast of Asia. All that can be said for certain is, that it does not comprise the Aleutian Islands, the empire of Japan, or the Loochoos. jral Though the characteristic Mammals of the Australian actor- Region are in every way highly remarkable, entirely com prehending as they do one of the three Subclasses (Ornitko- delphia) and nearly all of a second (Didelphia), by far the largest portion of the area it covers is weak if not absolutely wanting in Mammalian life, and the zoological features which mark the Region as a whole are perhaps better exhibited by its Birds than by any other Class of its fauna. This being the case, it may be excusable in this place to dwell longer upon this Region than upon the rest. True it is that we have no Subclass of birds, like the Ornithodelphia among Mammals, which is restricted to the Region ; but, on the other hand, the instance of the Didelphia, to which allusion has just now been made, is here almost exactly paralleled by that of the Struthious Birds the Ratitce, to call them by the name now very gene rally applied to them by the zoologists who recognize them as forming one of the two primary divisions or Subclasses, we may term them, of the Class Aves. All the existing alence Ratita: (with the exception of the species of two forms, the Ostriches of Africa and South America, belonging to the genera Struthio and Rhea, and comprising at most but five species) are found within the Australian Region and nowhere else. But further, the Ratitce of the Region are more widely distributed throughout its area than are the Di delphia, since the former extend from Ceram, in lat. 4 S. and long. 130 E., to New Zealand in lat. 45 S. and long. 175 E., if not a little further, while the Didelphia stop short at lat. 44 S. and long. 155E. 2 But if we take the birds alone, and compare the two Subclasses into which the existing or recent members of the Class are divided, we find the Australian Region remarkable for its ornithic singularity. The smaller of these two Subclasses, the Ratitce, contains six very natural groups which might well be called Orders including, according to the most exaggerated computation of their number, less than 40 species, while the larger Subclass, the Carinatce, though perhaps not including more truly natural groups, comprehends some 10,000 species. Now, out of the six groups of this smaller Subclass, four are absolutely restricted to the Australian Region, and these four groups contain all but, at the highest estimate, as above stated, five of the species known to belong to the Subclass ; thus we should be able to regard the 35 species of recent Ratitce of the Region (a number which is clearly far too large) as the proportional equivalent of an avifauna of more than 8000 species (8750). Leaving, however, such a calculation as this, which indeed cannot as yet be more than an approximation to the truth, we must consider first the remaining ornithic features of the Region as a whole, and then those of its parts. With respect to each of those subjects, it will be evident to every one that a further division is at once incumbent upon us. The prevalent zoological features of any Region are of two kinds negative and positive. It is therefore just as much the business of the zoogeographer, who wishes to arrive at the truth, to ascertain what groups of animals are wanting in any particular locality (altogether independently of its extent) as to determine those which are forthcoming there. 1 Of course, as regards polymorphism, no comparison can be made between the RatiV f, and the Didelphia, the latter presenting a very great variety and the former a very great sameness of structure and habit, though if it be true, as seems to be most likely the case, that Dinornis and its allies were absolutely devoid of wings, we should in them have a divergence from the normal ornithic type which is alto gether unique in the whole Class, and for its singularity might well be set off against the multifariousness exhibited by the Didelphia. Of course, in the former case it would be absurd to regard as a physical feature of any great value the absence from a district of groups which do not occur except in its imme diate neighbourhood ; but when we find that certain groups, though abounding in some part of the vicinity, either sud denly cease from appearing or appear only in very reduced numbers, and occasionally in abnormal forms, the fact obviously has an important bearing. Now, as has been above stated, mere geographical considerations, taken from the situation and configuration of the islands of the so- called Indian or Malay Archipelago, would indicate that they extended in an unbroken series from the shores of the Strait of Malacca to the southern coast of New Guinea, which confronts that of North Australia in Torres Strait, or even further to the eastward. Indeed, the very name Australasia, often applied to this part of the world, would induce the belief that all the countless islands, be they large or small and some of them are among the largest on the globe were but a southern prolongation of the mainland of Asia. But it has been already stated that so A yerj far from this being the case a very definite barrier is l<?tinit interposed. A strait, some 15 miles or so in width, and 1)0uni1 separating the two fertile but otherwise insignificant islands of Bali and Lombok, makes such a frontier as can hardly be shewn to exist elsewhere. The former of these two islands belongs to the Indian Region, the latter to the Australian, and between them there is absolutely no true transition that is, no species are common to both which cannot be easily accounted for by the various accidents and migrations that in the course of time must have tended to mingle the productions of islands so close to one another. The faunas of the two are as absolutely distinct as those of Soxth America and Africa, and it is only because they are sepa rated by a narrow strait instead of the broad Atlantic that they have become so slightly connected by the interchange of a few species and genera. Now, first, of the forms of Birds which are prevalent Imliai throughout the Indian Region, but are entirely wanting in forms the Australian, we have at once the Bulbuls (Ixidce), very | Nantn characteristic of most parts of Africa and Asia, with the allied group of Phyllornithidcc, which is peculiar to the . Indian Region ; the widely -spread families of Barbets (Megalcemidce) and Vultures ( Vulturidoe) ; and the Phea sants (Phasianidce), which attain so great a development in various parts of the Asiatic continent and islands that there must their home be regarded as fixed, though some species are found very far removed from the focus of the family. 2 Some naturalists would add the Finches (Fringillid<x), but the real position of the so-called " Finches " of Australia must at present be considered extremely doubtful, and it may prove that they are the direct descendants of the more generalized group whence sprang both the true Fringillidcc and the Ploceidoe, if, indeed, these can justifiably be kept apart. Then, of forms which are but weakly represented, we have the otherwise abundant Thrushes (Turdidce 3 ), and, above all, the Woodpeckers (Pictdce), of which some 4 species, or at most 5, 4 out of more than 300, just cross 2 The separation of this family from the Tetraonidcc (Partridges and Grouse), though hitherto almost universally recognized, seems to be a very questionable proceeding, and, so far as the present writer is aware, is one that can only be maintained by structural characters, which though patent in their extreme forms, appear to vanish in those which are intermediate (cf. Proc. Zool. Soc. 1868, pp. 300,301) ; but for the purposes of this treatise it is of little consequence, since tho Tdmonidce are but very feebly represented in the Australian Region. 3 It is almost certain that no satisfactory limits can be laid down between this family and the Warblers (Sylviidcc), but, as in the case mentioned in the last note, the result would hardly be affected by combining the two families, since the Sylviida ha.ve comparatively few members in the Region now under notice. 4 There are said to be Dendrotypes analis in Lombok, Mitelleripicus
fuh us and Yungipicus temmincki in Celebes, Y. moluccensis in tl.e