338 POLAR PROBLEMS 



to distances up to several hundred miles. In the Antarctic the ice 

 covering is maintained in spite of the meager snowfall of a desert 

 climate, because of the smallness of the denudational and distributive 

 forces operating to clear the continent of ice. Surface melting seems 

 to be unknown except in the immediate neighborhood of rock, and 

 the rate of movement of the glaciers whose movement has been 

 measured is far. less than in Greenland, for example. The tendency 

 to form floating extensions to the ice covering of the land results 

 partly from the low rate of melting in the cold waters that bathe the 

 seaward terminations of the glaciers and partly from the fact that the 

 snowfall added annually to their surface is probably often greater 

 than that lost through evaporation. One of the most difficult of the 

 problems awaiting the future scientist-explorer is the measurement of 

 the rate of melting of such floating ice tongues in water and the eval- 

 uation of the factors which control their denudation. The importance 

 of a complete knowledge of these various factors for our present purpose 

 lies in the possibility of determining whether the greater extension of 

 ice on this continent in the past was associated with greater or less wind 

 velocity and snowfall and with a higher or lower temperature. On 

 these points we have no reliable information; nor are there many 

 useful data regarding absolute or relative humidity, which, together 

 with wind velocity and temperature, determine the rate of evapora- 

 tion of the snow and ice surfaces. There are no accurate data regard- 

 ing either snowfall or radiation to and from the earth. It is this last 

 factor, radiation, that must, in the main, determine the activity 

 of the glacial anticyclone or outflow of air cooled by contact with the 

 snow surface. To what extent this outflow is controlled by the snow 

 slopes and to what extent by the contrasting meteorological conditions 

 at the boundary of the continental ice remains uncertain. Still another 

 factor of unknown degree of importance in shaping the glacial anti- 

 cyclone is the contrast of conditions across the boundary which 

 separates a loosely coherent snow surface from a compact icy surface. 

 For the greater part of the year this boundary lies outside the Antarctic 

 Continent, but it is possible that it penetrates inland to some extent 

 in summer and in some areas. 



To appreciate the importance of the type of snow surface it is 

 necessary to refer to the fact that the cooling of the air above the 

 surface takes place by contact with the snow. Where the surface 

 consists of loose snow, it cools rapidly through the small heat capacity 

 and low heat conductivity of the snow, whenever outward-directed 

 radiation exceeds that directed inwards. As it ages, the snow surface 

 becomes more coherent by the slow growth in the mean size of indi- 

 vidual crystals, until finally it becomes a compact and dense mass 

 of interlocking grains. The absence of fresh snowfall or drift may 

 thus modify the surface so as to prevent a close correspondence 



