ALLEN ON GEOGRAPHICAL DISTRIBUTION OF MAMMALS. 327 



to species. In the foregoing remarks I have had little to say respecting 

 the range of species, and have tabulated merely genera and families. 

 These tables clearly show that a large proportion of the mammalian 

 genera and families of the northern hemisphere have a circumpolar 

 range, the same genera and families occupying the Arctic and Sub-Arc- 

 tic lands in both the Old World and the New, and that only a small 

 per cent, of the whole number found here are peculiar to either of the 

 northern land-areas ; that a large part of the genera and families met with 

 in the temperate and warmer latitudes occur on the eastern continent as 

 well as on the western; that again a considerable proportion of the 

 genera and families met with in the warmer parts of the earth occur 

 also both in the Old World and the New, while many others are well 

 known to have been common to the two during the Tertiary period. It 

 has been further shown that there is a greater diversity of life between 

 contiguous climatic belts of the same continent than between corre- 

 sponding belts of the two continents, especially north of the forty-fifth 

 parallel of latitude, and that any marked faunal differentiation of the 

 two continents begins only in the warm-temperate and subtropical lati- 

 tudes. On each continent, the arctic, temperate, and tropical zones are 

 each marked in their general fades respectively by corresponding phases 

 of life. So obvious is this that we have in current use the expressions 

 '* arctic life", " temperate life", and u tropical life", in recognition of cer- 

 tain common features of resemblance by which each of these regions is 

 distinguished as a region from the others. This is in accordance with a 

 law I have termed the law "of differentiation from the north south- 

 ward",* or in accordance with increase of temperature and the condi- 

 tions resulting therefrom favorable to increased abundance of life. 



In this connection it may be well to recall certain general facts pre- 

 viously referred to respecting the geographical relations of the lands of 

 the northern hemisphere and their past history. Of first importance is 

 their present close connection about the northern pole and their former 

 still closer union at a comparatively recent date in their geological history; 

 furthermore, that at this time of former, more intimate relationship, the 

 climatic conditions of the globe were far more uniform than at present, 

 a mild or warm-temperate climate prevailing where now are regions of 

 perpetual ice, and that many groups of animals whose existing repre- 

 sentatives are found now only in tropical or semitropical regions lived 

 formerly along our present Arctic coasts. We have, hence, an easy ex- 

 planation of the present distribution of such groups as Tapirs, Manatees, 

 many genera of Bats, etc., in the tropics of the two hemispheres, on the 

 wholly tenable assumption of a southward migration from a common 

 wide-spread northern habitat, to say nothing of the numerous existing 

 arctopolitan and semi-cosmopolitan genera. The former greater commu- 

 nity of life in the northern hemisphere in preglacial times is further 

 evinced by the wide spread occurrence there of the remains of Camels, 



* Bull. Mas. Comp. Zool., vol. ii, p. 379. 



