It Is clear that if we employ a certain average thickness for the ice carried out, we will then 

 obtain its volume by the equation: 



F (km^/day) = h (km) xq (km^/day) ^^^ 



From the example chosen we see how great is the importance of the neck in the ice behavior 

 of the White Sea. In the winter season, and especially in the second half of winter, southerly winds 

 prevail here, in particular as follows: In Dvina Bay from December through April, southeasterly; 

 in Mezen Bay, southerly with slight deviation to the east; in Onega Bay, southeasterly; in the basin 

 and neck, southwesterly. Such a wind regime results from the synoptic situation which prevails 

 during the winter in the regions adjacent to the White Sea, namely, an area of increased pressure 

 located in the region to the north of the Caspian Sea, and Icelandic lows passing along the southern 

 part of the Barents Sea. 



As we see, the normal winter wind regime favors the transfer of ice out of the White Sea 

 through the neck and out of Mezen Bay into the Barents Sea where the ice melts away. At the same 

 time the north winds do not carry new ice into the White Sea, because before the month of April 

 (when melting has already begun in the White Sea proper) there is not yet any ice in the regions of 

 Barents Sea adjacent to the neck. The role of these v^dnds is therefore confined to redistribution of 

 the ice, its thickening and hummocking. 



Returning to equations (6) and (7), we see that when the isobars are not parallel to the axis of 

 the neck, the ice in its movement should drift into the shore or the fast ice and should form hum- 

 mocks. Obviously, if we have a stationary isobars for a certain interval of time, there should re- 

 sult a more or less steady movement and distribution of ice. 



Let us suppose , for example , that over the entire White Sea the isobars run from southwest to 

 northeast, that is, parallel to the axis of the neck, which as we have seen, is the most natural con- 

 dition for the winter months. With isobars in such a position, the ice should drift from Onega Bay 

 into the basin and from the south shores of Kandalaksha Bay and Dvina Bay towards the northern 

 shore. Only the ice which is carried out of Onega Bay between Zhizhgih Island and Solovetskya 

 Ostrova moves directly into the neck of the White Sea. The ice located east of Zhizhgih Island, 

 approaching Zimnegorski Cape, divides into two parts; one goes into the neck, the other partially 

 hummocks, partially drifts south along the Zimni shore. On the other side the ice which is carried 

 out of Onega Bay between Solovetskya Ostrova and the Karelian shore (an extremely small drift, 

 incidentally) and the ice which drifts from the Karelian shore along the isobars , both run into the 

 Murmansk shore and create here massive hummocks. Thus, with the isobar position as mentioned, 

 along the whole Letni shore, along the north shore of Solovetskya Ostrova, along the Karelian 

 shore, along the south shore of Kandalaksha Bay, and along the Murmansk shore a thinning of ice 

 results and in winter a substitution of new ice for the old. On the other hand, along the 

 Kandalaksha and Zimni shores the ice becomes more concentrated and its movement is conditioned 

 by wind-driving. 



Actually, the ice, in its movement from the south shores of the White Sea, carries with it the 

 surface layer of water. As a result, offshore driving sets in at these shores, which in turn causes 

 compensational currents, forming, together with the offshore currents, whirlpools with horizontal 

 and vertical axes . While the first are not accompanied by horizontal ice movements but only by 

 hummocking, the second, on the other hand, cause a drift of ice sometimes even against the wind. 



Thus, the general circulation of ice in the White Sea is determined by synoptic conditions and 

 is altered by the configuration of shores and by compensational currents. 



443 



