420 THE CANADIAN NATURALIST. [Vol. Tiii. 



The snow acts by weight, except during a thaw, and in this 

 capacity it would frequently assist the formation of ice above 

 the surface, for when it falls on young ice and presses it to the 

 level of the water, the snow becomes saturated like a sponge by 

 capillary action ; it then freezes and forms a light ice, thus 

 diminishing, if not entirely neutralizing, the effects of precipita- 

 tion under such circumstances. Assuming the precipitation and 

 drainage during the winter months to amount in the aggregate 

 to three inches within the 70th parallel, there would still remain 

 a volume of ice mechanically raised above the mean level of the 

 Polar Sea, in excess of precipitation, equivalent to the entire dis- 

 charge of the Gulf Stream during sixty-three consecutive hours, 

 or two days and fifteen hours. 



But the mechanical up-lifting of 382 cubic miles of ice is only 

 part of the work of winter cold in the Polar Seas. There has to 

 be taken into consideration the enormous quantity of heavy 

 brines squeezed out of the entire body of ice by the process of 

 freezing, and the effect these produce upon the salinity and spe- 

 cific gravity of the Polar waters, hereafter alluded to. The total 

 bulk of the ice formed each year, estimated as before stated, is 

 equal to 3,706 cubic miles, an equivalent to the entire discharge 

 of the Gulf Stream for 34 consecutive days ; and from this vast 

 mass a large percentage of salt is expressed by the freezing process. 



The formation of Polar ice is by no means uniform with the 

 same mean temperature : its daily increase diminishes as its 

 thickness increases. During the first half of the season when 

 fresh ice or floe ice is formed, the quantity raised above the 

 level of the sea is considerably greater than during the second 

 half, especially if snow be absent. The quantity of heavy brines 

 squeezed out is also dependent upon similar conditions ; and it 

 follows that great variation in results from both of these causes 

 must take place during different seasons. 



The currents towards the North Polar circle, to supply the 

 void created by the rising ice, should be greater in October and 

 November, than from December to March, and greater again 

 from March to May,* according to the snow fall and the extent 

 of its retarding influence on the formation of ice. 



Regarding the current in Davis Straits, we find that the north 



* See Koldeway on the protection afforded by snow, and the effect 

 of its disappearance before storms in the early spring. 



