330 BULLETIN UNITED STATES GEOLOGICAL SURVEY. 



grounds, they are hardly required for the elucidation of the distribution 

 of the mammalia, since they must be mainly characterized negatively. 



Beginning with the Arctic Eegion, we meet, as already shown, and as 

 is almost universally admitted, a continuous homogeneous fauna, of 

 considerable geographical area, but mainly characterized by what it 

 lacks. Its southern boundary may be considered as the northern limit 

 of forest vegetation. Continuing southward, few other than arctopoli- 

 tan genera of mammals are met with north of the mean annual of 36° 

 F. This considerable belt hence includes what may be termed the cold- 

 temperate zone. The American and Europaeo Asiatic portions of this 

 zone are only to a slight degree differentiated, while each is essentially 

 homogeneous. 



Below this, non-arctopolitan genera, or those restricted to more or 

 less limited areas, become more frequent, and, indeed, form a consider- 

 able proportion of the genera represented. This belt occupies the 

 remainder of the north-temperate zone, extending to about the mean 

 isotherm of -TOo F,, and may be termed the warm-temperate zone. Un- 

 like the cold-temperate zone, it is divisible on each continent into sev- 

 eral well-marked minor regions, which are, however, more strongly 

 differentiated, inter se, in the Old World than in the New. 



The tropical zone embraces, of course, in its fullest extension, a much 

 greater latitudinal breadth than the temperate, bub its southern land- 

 border is very irregular, its only considerable development south of the 

 equator being in South America and Africa. It is also so much diver- 

 sified in many parts by mountain-chains that subdivision into secondary 

 zones seems less feasible than in the case with the north-temperate 

 £0ue. A central torrid and a north and a south sub-torrid zones might, 

 however, be readily made, but such a division has not been attempted 

 in the present connection. A northern sub-torrid division may indeed 

 be very conveniently recognized, extending from about the annual 

 isotherm of 67° to that of about 74P F., and including a transitional 

 region consisting of the extreme southern border of what has been 

 above defined as the warm-temperate zone and the northern border of 

 the tropical. 



In like manner, the distribution of life seems to warrant the recogni- 

 tion, in Africa and South America, of a corresponding transitional belt 

 between the two torrid and the southern warm-temperate zones. Aside 

 from these divisions, the Torrid Zone admits of others of a more practi- 

 cal or useful character. These become at once obvious, since they result 

 from the position and configuration of its component land elements. 

 The first is a primary separation into two " realms", an American and an 

 Indo-African. Each of these is again divisible into several minor por- 

 tions or " i^rovinces " ; but the Indo-African admits also of division into 

 two " regions", an African and an Indian, which are divisions of second- 

 ary rank, each having several " provinces". 



The South Temperate Zone has a very limited land-surface, consisting 



