MOLLUSK-FAUNA OF ALASKA AND SIBEEIA. 363 



5. The Columbian. — This includes the coast drainage of British 

 Columbia and Southeastern Alaska, the basins of the Fraser and 

 Columbia rivers, the coastal part of the State of Washington, and the 

 northern part of Idaho and Montana, west of the Selkirk Range and 

 its more southern equivalents in the Eocky Mountain region. The 

 northwestern extension of this system in Alaska, between the Coast 

 Ranges and the sea, including the Alaska Peninsula and the Aleutian 

 Islands, being really an extension northwestward of the Columbian 

 system, might perhaps for convenience be called the Alaskan sub- 

 region. 



6. The Yukonian. — This system includes the entire drainage of the 

 Yukon river, the tundra north of it and the basin of the Kuskokwim; 

 or all of Alaska north, northwest and west of the Alaskan range, as 

 well as that of the basin east of the Coast ranges drained by the 

 Yukon and its tributaries. 



We know through the researches of General G. K. Warren, IT. S. 

 Engineers, that a considerable portion of the original Mississippi 

 drainage near its former headwaters has, through changes of level 

 of the earth near the 49 th parallel, been captured by the Hudsonian 

 drainage system, and it is not unreasonable to suppose, that a certain 

 proportion of the Mississippi fresh-water fauna was captured with the 

 streams in which it lived. Some such supposition seems necessary to 

 explain the far northern extension of certain Mississippi Valley species 

 in the Hudsonian region, while other drainages equally suited to their 

 inhabitation are destitute of them. 



Another factor in the distribution of land as well as fresh water 

 shells is the former existence of a continental ice-cap by which the 

 entire mollusk fauna of the region occupied by ice would be ex- 

 terminated; though, in the interglacial periods, the external fauna 

 might advance as the ice retreated, only to be driven out again on the 

 return of the ice. Such fluctuations have been well shown, by the 

 researches of Dr. Coleman, of Toronto, to have occurred in the valley 

 of the Don, near that city. 



The vast territories included in these drainage systems are, it is 

 true, only partially and imperfectly explored for mollusks. Yet cer- 

 tain portions of them are tolerably well known and the uniformity im- 

 posed on the fauna by its high northern position and unvaried condi- 

 tions, leads to the belief that, while much is yet to be known in trac- 

 ing out the details of distribution, little is to be expected in the way 

 of absolutely new species, even from this immense territory yet to be 

 explored. 



It would be rash to conclude that nothing new remains to be found, 

 but our expectations should certainly be moderate. 



There are a few characteristic fresh-water forms in Greenland, 



