ALLEN ON GEOGRAPHICAL DISTRIBUTION OF MAMMALS. 329 



based on it can have " little bearing on the great features of zoological 

 geography the limitation of groups of genera and families to certain 

 areas ". 



II. MAMMALIAN REGIONS OF THE GLOBE. 



The influence of temperature as a limiting agent in the distribution 

 of life, as well the "law of the distribution of life in circumpolar zones", 

 was fully recognized by Humboldt nearly three-fourths of a century 

 ago, and later, practically if not explicitly, by Ritter, De Candolle, 

 Agassiz, Wagner, Forbes, Dana, Giiuther, Meyen, Middendorff, and 

 many other leading zoologists and botanists. While this law must 

 incontrovertibly underlie every philosophic scheme of lief-regions, the 

 number of zones to be recognized, as well as their boundaries, must in 

 a measure be open to diversity of opinion. Professor Dana, in 1852, 

 recognized five primary zones for marine animals, namely, a torrid, a 

 north and a south temperate, and a north and a south frigid. The torrid 

 and temperate were subdivided, the first into three, the others each into 

 five sub zones, the two frigid being left undivided. Mr. A. Agassiz, in 

 treating of the distribution of the Echini,* recognizes also five zones, a 

 torrid, two temperate, and two frigid. These five primary zones prove 

 to be applicable also to the mammalia, and even their subdivisions may 

 be readily traced, but ape rather too detailed for practical use. Owing 

 to the irregular surface of the land-areas, occasioned by elevated pla- 

 teaus aud mountain-chains, these zones of distribution have of course 

 a less regular breadth and trend than they preserve over the oceans. 

 Their boundaries, however, approximate to the courses of the isotherms, 

 by certain of which they may be considered as in a general way limited. 



In recognition of these zones, and also of the law of differentiation 

 of life with the relative isolation of the principal land-areas, I proposed 

 in a former paper (I. c., p. 380) a division of the land-areas into eight 

 "Realms", namely: I, Arctic; II, North Temperate; III, American 

 Tropical; IV, Indo- African; V, South American Temperate ; VI, Afri- 

 can Temperate; VII, Antarctic; VIII, Australian. A subdivision of 

 most of these primary regions was provisionally suggested, but only 

 the North American was treated with any degree of detail, and this 

 mainly with reference to the birds, and more especially those of its 

 eastern portion. Subsequent study of the distribution of mammalian 

 life over the globe has led me to modify some of the views then ex- 

 pressed, especially in relation to the divisions of the Australian Realm, 

 and to unite the South African Temperate with the Indo-African, as a 

 division of the latter, and also to recognize Madagascar and the Masca- 

 reue Islands as forming together an independent primary region, in 

 accordance with the views of Sclater, Wallace, and others. Whether 

 or not the Arctic and Antarctic Regions should stand as primary divi- 

 sions seems also open to question. While perhaps tenable on general 



* Ihustr. Cat. Mas. Comp. Zool., No. vii, 1872, pis. A-F. 

 Bull. iv. No. 2 2 



