HISTORICAL ZOOGEOGRAPHY 111 



reached Australia and southeast Asia, though these connections were 

 by no means contemporaneous or of equal duration. Africa was con- 

 nected with Europe at times via the Strait of Gibraltar and probably 

 ma Sicily and Italy as well. Repeated connections between Eurasia 

 and North America must certainly have existed, probably across the 

 present Bering Strait and parts of the adjacent seas. This bridge must 

 have been in existence during a warm period, when the Alaskan climate 

 permitted the dispersal of numerous animals. These connections seem 

 to provide for the distribution of all the regions inhabited by animals 

 whose distribution is rigidly dependent upon the existence of land. 

 If all these connections be drawn upon a map regardless of their non- 

 contemporaneity, the result is a connected land mass in the northern 

 hemisphere with three great southward projections. 



Connections between the southern continents through the antarctic 

 land mass must be regarded with suspicion, though such connections 

 have repeatedly been proposed and defended with ability. Many au- 

 thorities, however, have found no insuperable difficulty in the way of 

 a derivation of the southern faunae from the general northern land 

 mass, and so long as this is possible, and in the absence of direct evi- 

 dence to the contrary, the assumption of direct connections between 

 the southern continents is certainly to be avoided. 34 There are theoretic 

 considerations which favor a derivation of the southern faunae from 

 the north, based on general phenomena of climatic change. Land 

 bridges, finally, which extend right across the Atlantic or Pacific, are 

 so dubious even from biological viewpoints that they deserve no con- 

 sideration. The ambitious attempt of Arldt 40 does not carry conviction 

 to a critical reader on account of its failure to weigh the evidence for 

 the permanence of continental outlines against the probability of trans- 

 oceanic land bridges. Wegener's hypothesis of continental drift is not 

 in accord with many geological observations and not only is not 

 needed to explain zoogeographical distribution but actually creates 

 more difficulties than it adjusts. 39 



The distributional relations of the terrestrial animals strongly sup- 

 port the assumption of a northern center as the place of origin of the 

 principal advances in organization in the more important groups of 

 terrestrial forms. The vast land areas surrounding the Arctic Ocean 

 afforded a spacial basis for their development, and the successive 

 periods of cooling and other climatic changes produced a periodic 

 severity of selection and thus favored advances in organization. The 

 groups whose adaptations or changes represented advances were en- 

 abled to enlarge their range, and extend it to the south, at times 

 driving more primitive forms before them. Successive impulses, follow- 



