748 BIRDS [DISTRIBUTION. than the latter. But the affinities of its Reptilian fauna point to a connection, however remote in point of time, with South America, and accordingly the Galapagos are here left in that Region to which they have been commonly assigned, tral- (5.) The Central- American Subregion is the next to be erican considered, and in treating of it we become aware of a dis- region. t ui .bi n g force which renders impossible the laying down for it of anything like a definite frontier. This disturbing force is the entrance, as before intimated, of a Nearctic fauna which runs along the backbone, so to speak, of the Sub- region to an unknown but variable extent ; for part of this Nearctic fauna ebbs and flows according to the season of the year, in winter possibly creeping down the mountain sides, and being strongly reinforced by immigrants from the north, but in summer retiring northward and perhaps up ward, so as to occupy only the most lofty ridges. Yet that two Subregions here unite and inosculate is certain ; but in considering the Central- Am erican avifauna, we have to guard ourselves against this periodic stream of northern immigrants, and cannot deal with it precisely in the same way as we have done those Subregions further removed from the influence which is here so strongly manifested. In CentiisJ. America, though its ornis is of the richest, we find not a single peculiar family of Birds, and those which it, to a more or less limited extent, shows with the other Subregions of the Neotropical Region have been already named, except the Ampelidce, a small but widely-ranging family of the northern hemisphere, which it has in common with the Antillean Subregion. 5 other families, however, Paridce, Sittidce, Certhiidce, Laniidce, and Meleagridce, be longing also to the Nearctic Region, occur here. Of genera which a v e not found elsewhere in the Region, it seems to have 93, but 47, or just more than half, of them are also found in the Nearctic Region ; and therefore to obtain any thing like a true notion of the Central- American ornis, it will be necessary to keep the two categories apart. Taking first those which are absolutely peculiar, we have 2 belong ing to Turdidce, 1 to Troglodytidce, 2 to Mniotiltidce, 1 to each of Vireonidce, Ampelidce, and Tanagridce, 3 to Emberi- zidce, 1 1 to Icieridce, 2 to Corvidce and Tyrannidce respec tively, 1 to Cotingidce, 2 to Formicariidce, 19 to Trochi- lidce, 2 to Momotidce, 1 to each of Trogonidce, Cuculidce, and Psittacidce, 2 to Cracidce, and 1 to Tetraonidce. Then, taking those not found elsewhere in the Neotropical Region, but inhabiting the Nearctic, we have, as occurring in Central America, 1 belonging to Turdidce, 2 to Syl- viidce, 3 to Paridce, 1 to each of Sittidce and Certhitidce, 2 to Troglodytidce and Mniotiltidce respectively, 1 to Laniidce, 2 to Ampelidce, 14 to Emberizidce, 1 3 to Fringillidce, 2 to Icteridce, 1 to each of Tyrannidce, Trochilidce, Picidce, and Cuculidce, 2 to Strigidce and Anatidce respectively, 1 to each of Columbidce and Melea- gridce, 2 to Tetraonidce, and 1 to Charadriidce. Nicely balanced as these numbers are, they shew a result which might well have been expected from the physical and geographical configuration of the country, while the numbers of other families peculiar to the Neo tropical Region, though shared by some of its Subregions, as already given, prove incontestably the propriety of including Central America with that Region ; and this would come out even more plainly did our limits permit of the investigation being extended to species, though so many northern forms here find their winter-quarters. It remains to remark that almost the only island of any im portance belonging to the Subregion is Socorro, the largest of a small group lying to the westward of Mexico in lat. 18 30 N., and long. 111 W. Here out of 9 species of Land-birds, 4 have been described as peculiar, 2 others are 1 See preceding footnote. elsewhere known as occurring only on the Tres Marias, a little group some 250 miles nearer the mainland, and 1 is regarded as a local race of a continental species, leaving but 2 (both Birds-of-prey) which cannot be deemed auto chthonous. The still more remote Cocos Island, lying in lat. 5 33 N., and long. 87 W., from which one peculiar species of Coccyzus (Cuculidce) is known, may belong just as likely to the Subandean as to the Central- American Sub- region. (6.) The Antillean is the only one of the Neotropical Sub- Antill regions the precise boundaries of which can be definitely Subrq laid down ; and it is in many respects one of the most sug gestive and interesting, comparatively small though it be. Extending from Cape San Antonio de Cuba in the west to Barbadoes in the east, its greatest length is only about 1700 miles, and from Abaco, one of the Bahamas, in the north to Grenada in the south, it does not cover 15 degrees of latitude, while within these limits the proportion of land to water, being less than 98,000 square miles, is very in considerable. The unbroken chain of islands which are commonly known as the " West Indies " though that term rightly includes not only all of the " Spanish Main," but an indefinite extent of coast lying both north and south of the ancient dominions of the Catholic King in the New World forms, geographically, a second line of connection between the two halves of the American continent, sepa rated from the great western isthmus by the deep waters of the Caribbean Sea and the Gulf of Mexico, and at once suggests a former communication by land with Yucatan at the one extremity and with Venezuela at the other, to say nothing of a possible junction with Florida. Yet, as will presently be shewn from a consideration of the peculiar forms of Bird-life which have grown up along the chain, any such communication, if it ever existed, must have been exceedingly remote in point of time ; for narrow as are the channels between Cuba and the opposite coast of Central America, between the Bahamas and the south-western peninsula of North America, and between Grenada and Tobago (the last belonging zoologically, as has been already demonstrated, to South America), the fauna of the Antillean chain, instead of being a mixture of that of the almost contiguous countries, differs much from all, and exhibits in some groups a degree of speciality which may be not unfitly compared with that of oceanic islands. Except such as are of coral formation, the Antilles are hilly, not to say mountainous, their summits rising in places to an elevation of 8000 feet, and nearly all, prior to their occupation by Europeans, were covered with luxuriant forest, which, assisting in the collection and condensation of the clouds brought by the trade winds, ensured its own vitality by precipitating frequent and long- continued rains upon the fertile soil. Under such condi tions we might expect to find an extremely plentiful animal population, one as rich as that which inhabits the same latitudes in Central America, not many degrees fur ther to the west ; but no instance perhaps can be cited which shows more strikingly the difference between a con tinental and an insular fauna, since, making every allowance for the ravages of cultivation by civilized man, the contrary is the case, and possibly no area of land so highly favoured by nature is so poorly furnished with the higher forms of animal life. Here, as over so large a portion of the Australian Region, we find Birds constituting the supreme class the scarcity of Mammals being accounted for in some measure as a normal effect of insularity. Glancing at the entire chain, we may first set aside the Bahamas, a succession of emerged coral-reefs founded on, and to the south and east surrounded by, shoals or banks, broken only here and there by deeper channels ; and then
by drawing a line to the south of the islands of St Croix