Conclusions 323 



tions, partly as remnants of a broken-up area of circumpolar type. From a bio- 

 geographical point of view, the necessity for any kind of land-connection in the 

 past, directly joining the European with the North American continent, is thus 

 denied. Like Simpson (1947, p. 666, footnote) I started my investigation "with a 

 feeling that a north Atlantic bridge was probable" and like him I dare hope that 

 abandoning this prejudice reduces the subjectivity of opinion. 



On the other hand, it seems unevitable to assume that part of the fauna of 

 Greenland and Iceland immigrated across a Pleistocene land-bridge from the 

 European mainland. An interesting point is that this invasion did not at all affect 

 the fauna of Baffin Island, which is completely devoid of a European element. 

 Actually, the narrow strait between Greenland and Baffin Island is the most 

 pronounced north-south barrier within the arctic region of the Holarctic area. 



Since the Bering land-bridge, repeatedly at work during Tertiary and Pleistocene 

 time, was the most important link toward evolution of the circumpolar fauna, a de- 

 tailed comparative study of the Alaskan and Northeast Siberian faunas is the most 

 important task for the future. 



