242 THE ANTARCTIC. 



dries up and forms water-vapour. Thus the cold 

 ground alternately deprives the air of its moisture by 

 condensation, and returns it again by evaporation. The 

 rents and gaps in the ice covering of the sea, forming- 

 large open spaces, constitute in winter an additional 

 source of moisture, whilst from the open sea beyond, 

 the supply of moisture cannot be of any consequence — 

 at least not that supplied by winds which were blowing 

 polewards at no great height above the ground, as in 

 consequence of the distribution of barometric pressure 

 these winds are of rare occurrence. This does not, 

 however, in our opinion, exclude the possibility of 

 moisture being supplied by the upper air currents, which 

 feed the polar anticyclones. Assuming the temperature 

 of the ascending air in the zone of depression round the 

 Pole to be 32° F., and the heights of ascent to be about 

 6,000 feet, there would result from it a reduction of the 

 temperature of the air of some 35° to 36° F., that is to 

 say, that by far the greater part of the precipitation must 

 be in a solid form. But now it must be assumed that 

 the precipitation of moisture out of the air must, at 

 least in the higher regions, be in the shape, not of 

 large flakes of snow, but of very minute ice-needles, 

 which are carried mechanically towards the interior of 

 the country by those stormy winds which at a great 

 height flow off to the anticyclone. As the air gradually 

 descends and thus grows warmer, a portion of the con- 

 densed moisture brought in may evaporate, only to be 

 immediately again condensed by the colder strata of air 

 resting over the surface of the land, since it is most 

 probable that over the Antarctic land, in winter at least, 

 there exists on the whole the well-known reversed state 

 of the temperature of the atmosphere, which on ascending 

 from the ground first shows an increase to be turned into 

 a decrease at great heights. The purely hypothetically- 

 assumed height of the in-streaming winds is probably 



