these some 5 species are peculiar 2 belonging to Phrygilus (Emberizidce), 1 to Cinclodes (Dendrocolaptidae), 1 to Musci- saxicola (Ti/rannidce), and 1 to Milvago (Falconidce). Of Water-birds there is a peculiar species of Chloephaga (Anatidce), and there are, or until recently were, about half-a-dozen species of Penguins (Sphmiscidce), some of which, though not resorting exclusively to these islands, may be fairly regarded as finding there their chief breeding- quarters. Of the ornithology of South Georgia, a group of islands lying some 1300 miles east of Staten Island, and nearly in tho same latitude, and doubtless belonging to the Patagonian Subregion, as well as of the South Shet- lands, no particulars are available. lian (2.) The Brazilian Sidrec/ionhgLSordyrecentlyhad removed jgioa. from it the valley of the Amazons and its tributaries. The boundaries, so far as they can be traced, have been already given. This Subregion is not characterized by the pre sence, of any family of Birds peculiar to it alone, but among those families which are found in only two Subregions of the Neotropical Region it possesses 3 in common with the Patagonian, Phytotomidce, Cariamidce, and Rheidce, the first and last only in its southern districts, and 1 in com mon with the Central-American Oxyrhamphidcc. Of the peculiarly Neotropical families occurring in three Sub- regions only, it has but 1, Pteroptochidce, the range of which is shared by the Patagonian and the Subandean. When we come to families of four Subregions, we find the Bra zilian invaded by the almost cosmopolitan Sylviidce, which also inhabit the Subandean, Central-American, and Antil- lean, and possessing in common with the Amazonian, Sub- andean, and Central-American not fewer than 8 Pipridce, Momotidce, Galbididce, Bucconidce, Bkcanphcutidce, Cracidce, Heliornithidce, and Parridce the last of which is, how ever, widely distributed in other regions, besides, Pala- medeidce, which occurs also in the Patagonian. Amazonian, and Subandean Subregions. The chief justification for considering the Brazilian Subregion apart from the Ama zonian is perhaps to be sought in the presence within the limits of the former of 5 families, Sylviida>, Oxyrhamjihidce, Phytotomidce, Pteroptochidce, and Cariamidce, which are not found in the latter, while on the other hand, 5 families, Capitonidfv, Opisthocomidcc, Eurypygidce, Psophiidce, and (Edionemidce, 1 inhabit the latter without occurring in the former ; add to which the fact, that of the families found in only two of the Neotropical Subregions not one is com mon to the Brazilian and Amazonian. In this Subregion we have 42 peculiar genera, 1 belong ing to Sylviidce, 4 to Tanagridoe, 2 to Emberizidce, 3 to Tyrannidce, 2 to Pipridce, 5 to Cotingidce, 6 to Dendroco- laptidce, 3 to Formicariidce, 2 to Pteroptochidce, 7 to Trochi- lidce, 2 to Caprimulgidce, and 1 to each of the families Picidce, Momotidce, Galbulidce, Psittaddce, and Tinamidce. The number of peculiar species is, however, far too great to be here enumerated. There are no islands of any im portance belonging to this part of South America, zonian (3.) T/te Amazonian Subregion, comprehending the val-
- gion. ]ey of the Amazons and its affluents (except their elevated
sources among the Andes) as well as the right bank of the Orinoco and, of course, the intervening country, has been but lately separated from the preceding, and the rea sons for here considering it distinct have just been briefly stated. They are not, indeed, those which first prompted the division, which was established mainly, if not entirely, on account of the peculiarity of most of the species of Birds 2 found within its ill-defined borders as before given; but on whichever ground we proceed we may be pretty sure that its separation is justifiable. We have here 2 1 The propriety of considering the Stone-Curlews to form a family distinct from the other Plovers (CharadriidcE) is very questionable. -C/. Sclater and Salvin, Proceed. Zool. Soc. 1867, pp. 593-596. [DISTRIBUTION; families peculiar Opisthocomidce already mentioned as forming almost a distinct Order, but composed of a singlo species, the Hoactzin, and ihePsopkiidce or Trumpeters, now a-days regarded as distant allies of the Cranes (Gruidce), but presenting many remarkable and unique features. This family also contains but one single genus, including some half-dozen species, the respective range of each appearing, singularly enough, to be separated by rivers. Amazonia pos sesses no family in common with but one other Neotropical Subregion, but it shares 3 with the Central-American and Subandean. These are Capitonidce, Eurypygidce, and GJdicnemidce, of which the second only is peculiar to the Region, the first and last being widely distributed. The families which it shares with three other Subregions have been already enumerated, as well as those which by their presence or absence distinguish it from the Brazilian Sub- region. The peculiar genera remain to be pointed out. These are but 27 in number : 2 belong to Icteridce, 3 to Pipridce and Cotingidce respectively, 2 to Dendrocolaptidcc, 7 to Formicariidce, 2 to each of Trochilidce and Galbulidce, 1 to Ardeidce and Palamedeidce, 2 to Cracidce and the genera Opisthocomus and Psophia before mentioned. There is another genus also which is worthy of remark, Chena- lopex, belonging to the Anatidce, not found elsewhere in the New World, but common to the Ethiopian Region. Space will not admit of our entering further upon the con sideration of the ornithic peculiarities of Amazonia, but perhaps it may be said to form the most self-contained Sub- region of the whole continental area of which we are now treating, and we may expect that with the progress of zoo logical exploration its boundaries may be laid down with tolerable precision. There are no islands which can bo attached to Amazonia. (4.) The Subandean Subregion, from what has beenSuband previously indicated of its extent, will be readily seen to Subreg offer the most varied conditions of existence of any part of the Neotropical Region, and we shall not therefore be sur prised to find its ornis at once rich and remarkable. It might perhaps now, and some day probably will, be broken up into two or more provinces, not to say separated into distinct Subregions, but the means for such partition are at present wanting. Yet there is only one family of Birds peculiar to it, the Steatomithidce, composed of a singlo species, the Guacharo or Oil-bird (Steatornis caripensis), confined to a very few localities in its eastern portion. In common with Central America, but not elsewhere met with in the Region, it has of Neotropical families Cinclidce and Alaudidce, though the latter of these two widely- ranging families is but poorly represented by a single spe cies (Otocorys chrysolcema), apparently the survivor of an old population (all, or nearly all, of which has perished), stranded, as it were, on the high lands of Columbia. Those families which it possesses that are common to but two other Neotropical Subregions have already been named, and it will be enough to repeat that 1 of them extends to the Patagonian and Brazilian, and the remaining 3 to the Amazonian and Central-American. In like manner have been enumerated the families which are also found in but three other Subregions 1 it has in common with the Amazonian, Brazilian, and Patagonian ; 8 with the Central- American, Amazonian, and Brazilian ; while 1 of these (Parridce) is also found in very distant parts of the world, and 1, also a family of extremely wide range, with the Brazilian, Central American, and Antillean Subregions. The genera peculiar to the Subandean Subregion are exceedingly numerous, amounting to no fewer than 72. These may be apportioned as follows : 1 to Troglodytidce, 4 to Cccrebidce, 10 to Tanagridce, 5 to Emberizidce, 1 to Icteridce, 2 to Tyrannidce and Pipridce respectively, 3 to
Cotingidce, 1 to each of Dendrocolaptidce and Formicariidce,