1916] Taylor: Beavers of Western North America -17. > 



merit as over against the other. But the facts of geographical distri- 

 bution of higher vertebrates point unequivocally to figure P as a 

 more accurate illustration of species arrangement than is figure 0. 



Relation of Evidence to Wagner's Theory of Migration and Geographical 

 Isolation 



There is left Wagner's theory, that the differentiation of species 

 has taken place through migration and geographical isolation, the 

 migration and spatial isolation being, according to this view, the con- 

 ditions without which specific differentiation does not take place. 

 With this proposition the evidence from mammals and higher verte- 

 brates in general seems to be, for the most part, in complete harmony. 

 The terms of this theory demand that a group made up of closely 

 related forms be represented by one species or subspecies, only, in 

 each locality, and that the most closely related forms be separated 

 by some barrier which shall keep them from interbreeding to an 

 extent that would bring about the swamping of incipient characters. 

 In other words, on this theory, groups of related species should be 

 arranged as illustrated in figure P. And so they are. 



Reference has already been made (p. 472) to Kofoid 's suggestion 

 regarding the coincident distribution of some of the most closely 

 related species of the fresh-water microfauna and flora. In another 

 paper (1907&, pp. 241-251) the same author considers the distribu- 

 tion of the Chaetognatha with reference to its bearing on the relation 

 of isolation to the origin and preservation of species, coming to the 

 following conclusion : ' ' The apparently wide-spread phenomenon of 

 coincident distribution of related species among pelagic organisms 

 appears to cast some doubt upon the universality of the operation of 

 isolation in the evolution of species as originally maintained by 

 Moritz Wagner (1868) and recently revived by President Jordan 

 (1905)." In the course of this paper Kofoid emphasizes the lack of 

 specific distributional data, especially of that pertaining to vertical 

 distribution. 



Supplementary in a way to this work of Kofoid is an exceed- 

 ingly suggestive paper dealing with the problem of isolation versus 

 coincidence in the same group (the Chaetognatha) recently pub- 

 lished by Michael (1913, pp. 17-50). The extensive collection of 

 specific data with regard not only to latitudinal and longitudinal 

 but also to vertical distribution permits the enunciation of the fol- 

 lowing principle (Michael, 1913, p. IS) : " . . . Pelagic organ- 



