294 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 56. 



western coast of South America which promise rich returns to the 

 explorer. From the latter region Cuming many years ago obtained 

 material which kept the British naturalists, Broderip, Sowerby, and 

 Reeve busy for a score of years describing and illustrating his har- 

 vest of new forms. And yet Cuming seems to have collected very 

 few of the smaller species. 



From California northward more exhaustive work has been done, 

 though there are notable gaps, many miles in extent, where dredging 

 has never been attempted and shore collections are practically un- 

 knowii. The character of the coast does not, in such regions, lend 

 itself to easy investigation. Nevertheless we may consider the gen- 

 eral fauna as fairly well known, though still affording the prospect of 

 many novelties. 



If we regard that region where a given genus is represented by the 

 greatest number of specific forms as being probably the center of 

 origin for species of that group, the conclusion is quite obvious that 

 the Oregonian fauna taken in its widest sense is the parental source of 

 the greater part of the boreal mollusks of the world. On the other 

 hand, if the meager fauna of the boreal Atlantic be taken as the 

 focus of origin, the reverse would be true, and the differentiation of 

 specific forms be greatest at the most distant area of migration, 

 other things being equal. 



The Tertiary faunas present much the same problem when ana- 

 lyzed, but in some instances suggest the possibility of reciprocal 

 migration; particular types appearing later in America than in 

 Europe, and the opposite. There can be no doubt, however, that 

 migration of species in the boreal region between the two hemis- 

 pheies was more easy and the routes more open in the Pliocene and 

 ate IMiocene than at any subsequent period. In the tropical region, 

 however, the reverse appears to be true, intercommunication be- 

 tween the two oceans having been finally interrupted at the end of the 

 Oligocene period. In my report on the collections of the Albatross 

 in the Panamic region I have called attention to the remarkable 

 difi'erences which have resulted from this cause in the composition of 

 the Caiibbean and Panamic mollusk faunas. 



The explorations of the Albatross in the Okhotsk Sea and on 

 the coasts of northern Japan and eastern Siberia, taken with those 

 of the Canadian Arctic Expedition and others on the American side, 

 have cleared up many obscurities in our conception of the geographi- 

 cal distribution of boreal mollusks. We know now that the Asiatic 

 fauna, even near Bering Strait, notwithstanding its geographical 

 approximation is measm'ably distinct from that of the American 

 coasts, and that the latter on the Arctic shores extends without 

 Greenlandic admixtm-es far to the eastward of the Mackenzie River 

 delta. Of course there is the expected admixture of characteristically 



