106 



New England, and a little smaller than the combined area of England 

 and Scotland. 



Without attempting to discuss the vertical circulation that takes 

 place within the radius of a mile or so from a berg melting under the 

 varying weather and water conditions that it encounters during its 

 life span in the Grand Banks region, it is certain that the bergs do 

 chill sea water there in melting. It makes no difference in a discussion 

 of their total chilling effect whether they affect surface or subsurface 

 layers. From whatever stratum the heat is chiefly drawn, the total 

 amount consumed will be the same. 



Let us assume that the bergs directly affect a layer of water averag- 

 ing over the 74,000 square sea-mile "melting area" 50 feet thick. 

 This is a minimum thickness to expect them to affect and a convenient 

 one for calculations. It gives 133,200 billion cubic feet of water in 

 the "melting area" to be affected by the disintegration of the total 

 of 26 billion cubic feet of glacial ice. Simplifying the problem, 

 it is found that for each 5,123 cubic feet of water there is 1 cubic 

 foot of ice. 



The latter is very close to 32° F. in temperature when it crosses the 

 forty-eighth parallel. Neglecting the lightness of the glacial ice, and 

 the salinity of the sea water, and generously allowing that 80 cubic 

 feet of the latter can be chilled 1° F. by the melting of 1 cubic foot of 

 the ice, it follows that the melting of the total amount of glacial ice 

 present throughout the whole ice season in the region under discussion 

 would only counteract the normal seasonal warming of a 50-foot layer 

 of w^ater in the melting area south of the forty-eighth parallel by 

 0.0156° F., an insignificant amount. 



It should be borne in mind that the 1929 ice year, on which the 26 

 billion cubic-foot berg ice total used above is based, was an ice year 

 about three times as heavy as the normal one. It is safe to say that 

 during the normal year, when less than 400 bergs come south of the 

 forty-eighth parallel, supplying the heat requirements of the glacial 

 ice disintegrating in the "melting area" does not chill any 50-foot 

 stratum of water in the cold current there by more than 0.01° F. 



It has been assumed in arriving at the above estimate that no chill- 

 ing effect from the bergs is lost directly to the air. This loss exists, 

 but it is doubtless extremely small. It has been further assumed that 

 no locally ice-chilled water is lost from the "melting area" to the 

 westward past Cape Race, to the westward to form bottom water over 

 the Grand Banks, or to the southeastward past Flemish Cap. Analy- 

 sis of berg drifts shows that, during the ice season at least, there is 

 but little push of cold water to the westward past Cape Race or onto 

 the Banks. Some losses in these directions occur, however, and even 

 more ice and cold water are lost to the southeastward past Flemish 

 Cap. The sum total of losses in all three directions probably amounts 



