The order Passeres, is the most extensive among birds, comprehending about 5,700 species grouped in 870 genera, and 51 families. The distribution of the genera, and of the families considered individually, has been already sufficiently given, and we now have to consider the peculiarities of distribution of the families collectively, and in their relations to each other, as representing well-marked types of bird-structure. The first thing to be noted is, how very few of these families are truly cosmopolitan; for although there are seven which are found in each of the great regions, yet few of these are widely distributed throughout all the regions, and we can only find three that inhabit every sub-region, and are distributed with tolerable uniformity; these are the Hirundinidæ, or swallows, the Motacillidæ or wagtails and pipits, and the Corvidæ or crows,--but the latter is a family of so heterogeneous a nature, that it possibly contains the materials of several natural families, and if so divided, the parts would probably all cease to be cosmopolitan. The Sylviidæ, the {300}Turdidæ, and the Paridæ, are the only other families that approach universality of distribution, and all these are wanting in one or more sub-regions. If, now, we divide the globe into the New and the Old World, the former including the whole American continent, the latter all the rest of the earth, we find that the Old World possesses exclusively 23 families, the New World exclusively 14, of which 5 are common to North and South America. But if we take the division proposed by Professor Huxley--a northern world, comprising our first four regions (from Nearctic to Oriental), and a southern world comprising our last two regions (the Australian and Neotropical)--we find that the northern division possesses only 5 families exclusively, and the southern division 13 exclusively, of which not one is common to Australia and South America. This plainly indicates that, as far as the Passeres are concerned, the latter bipartite division is not so natural as the former. Again, if we compare temperate with tropical families (not too rigidly, but as regards their general character), we find in the northern hemisphere only two families that have the character of being typically temperate--the Cinclidæ, and in a less degree the Ampelidæ--both of small extent. In the southern hemisphere we have also two, the Phytotomidæ, and in a less degree, the Pteroptochidæ; making two wholly and two mainly temperate families. Of exclusively tropical families on the other hand, we have about 12, and several others that are mainly tropical.