540 THE PUR SEALS OF THE PKIBILOF ISLANDS. 



the arctic water of the vicinity of Bering Strait. In winter the water temperatures 

 range but a few degrees above the freezing point of salt water, about 28° P. 



The topography of the Bering Sea basin has been elucidated by the work of the 

 United States Pish Commission and the Eevenue Marine, though much remains to be 

 done. In general, the southwestern part of the sea is deep on both sides of the 

 Aleutian Islands. The northern and eastern parts of the sea are relatively shallow, 

 forming a large submarine plateau covered by less than 100 fathoms of w^ate^, and 

 over a great part of its extent by less than 50 fathoms. The western edge of this 

 plateau is some distance west and south of the Pribilof group, and the margin between 

 them and the peninsula is somewhat deeply embayed, so that the plateau joins dry 

 land on the south not far from the western end of the peninsula. 



By platting the positions where the edge of the pack ice was met in early spring 

 from the log books of a large number of whale ships, I have been able to determine, 

 approximately, the usual extent of the pack in winter.' It must be stated that the 

 margin of the ice pack is not a strictly determinate line, but is fringed with more or less 

 floating and broken ice, which varies in position with the prevailing winds. Occasions 

 have been known when long-continued northerly winds in Pebruary and March have 

 carried the loose ice as far south as the northern border of the eastern Aleutians, filling 

 the bays of Unalaska with the drift and obstructing navigation, but this is very excep- 

 tional. Usually the water about the islands is free of obstructive ice throughout the 

 year. The Pribilof group, however, lie so much nearer the average limit of the winter 

 pack that few winters pass when the shores are not, for at least a short period, sur- 

 rounded by the floes, and sometimes they are icebound for one or two months. 



The presence of floe ice is destructive to a littoral fauna unless the animals can 

 retreat into the depths of the sea beyond the reach of ice. Por this reason the Arctic 

 shores and the beaches of the Pribilof Islands are poorly supplied with living mol- 

 lusks. The shores of the Aleutians have a fairly rich littoral fauna, though for some 

 reason, perhaps the scarcity of the red and green seaweeds noted by botanists all 

 through this region, there is a general scarcity of the minuter, mostly phytophagous, 

 forms of mollusks, such as Bissoidae and the like. The dredgings of the United States 

 Fish Commission have chiefly been made with the beam trawl, which does not retain 

 the more minute species; but my own, done with the dredge, also failed to obtain 

 any large number of small mollusks, so we may regard the fact that the fauna is rather 

 deficient in them as fairly well proved. 



The shallow waters around the Aleutians possess a well-marked and pretty 

 uniform fauna of some two or three hundred species of mollusks, a certain number of 

 which are common to the adjacent continental shores and the Pribilof Islands. This 

 fauna comprises a number of characteristic species, together with a contingent of 

 Arctic forms and a certain number of Oregonian types. 



The work of the Albatross, however, has shown that another fauna exists in 

 Bering Sea largely distinct from that of its shores and widely spread over the great 

 plateau to which reference has been made. This is not a deep-sea fauna, for the water 

 is no deeper, and often shallower, than that of the bays and harbors of the Aleutians, 

 in which but few of its component species are found. It is marked by a rather limited 

 number of bivalves and an unusual predominance of species of Buccinum, Ghrysodomus, 

 and Stromhella (or Volutopsis). My study of this fauna has not progressed far enough 



' See accompanying map. 



