from 0. 035 to 0. 040. The maximum speeds of the floes observed at the same wind force gave a 

 wind factor of 0. 08 - 0. 10. * 



In individual cases the wind drift of ice may be considerably stronger than this. For example, 

 Lavrov pointed out that on 2 July 1928, the icebreaker Mali/ gin caught in the ice of the north- 

 western part of the Barents Sea drifted westward (toward Ostrov Nadezhda) at a rate of 1 knot 

 under an easterly wind of force of 6-7. When the wind, maintaining a force of 6-7, changed to 

 northerly, the /"/ali/c/i Altogether with the ice moved along Ostrov Nadezhda at 3-4 knots and at a 

 distance of 3 or 4 miles to the south of this island (toward the open sea). Calculations of the 

 southward drift yielded a wind factor of 0. 15. Probably this intensive drift of ice was not purely 

 wind drift. Nevertheless, these figures are striking. 



1 have made the following assumptions to arrive at an approximate solution of the problem of 

 the wind drift of an isolated ice floe. First, I have assumed that in the beginning the wind drives 

 the ice floe but the water remains immobile. Thus, at first I examine only the actual wind drift of 

 the floe. This assumption is based on the following: direct observations have shown that an 

 individual floe is set in wind motion at a considerably greater rate of speed than the wind current 

 would allow and that subsequently it moves even faster. 



Furthermore, the wind current in the sea is not created immediately. Struiskii has shown 

 (on the basis of 2836 observations of winds and currents in the Caspian Sea) that frequently there 

 are no currents even during quite strong winds, and sometimes there are currents that run counter 

 to the wind. 



This is explained by the inertia of water masses and chiefly by the presence of residual 

 currents. 



The second assumption I have made to simplify the problem concerns the form of the ice. 

 Actually, since the above water part of a floe is subject to wind action while the underwater part 

 of the floe is subject to the resistance of water, in our theoretical treatment we can select the 

 form of the underwater and above-water parts of the floe such that the floe can move in various 

 directions with respect to the wind, as a sail, set in different ways, can move a sailing vessel in 

 various directions. Therefore, let us select a floe shape which will be indifferent with respect to 

 a saU and resistance of water, namely a cylinder with a vertical axis. 



If we assume that the water is immobile and that the drift will occur after a certain time 

 interval, three balancing forces will act on the floe: i^-wind pressure, /i-hydrodynamic resistance 

 in the direction opposite that of the drift and A-the Coriolis force directed (in the northern hemi- 

 sphere to the right of the drift) perpendicular to the drift, in other words perpendicular to force R 

 (figure 140). 



Under such assumptions, the drift angle can be obtained from the formula: 



tan a = — . (1) 



H 



The Coriolis force is: 



K = m 2(uc sin tp = 8; irr- h2 (oc sin © (2) 



*Once, during a strong wind, an isolated floe was observed to drift at a speed of 120 cm/ sec 

 or 2.33 knots (wind factor 0.12). 



382 



