CLIMATE AND THE POLAR ICE-CAP. 403 



point, and where consequently the ground at some depth is always 

 frozen. Yet so far is this frozen area from being covered with 

 perpetual ice, that there is no continental ice on any part of Asia 

 or America ; and though the interior of Greenland is covered with 

 a true continental ice-sheet, yet this descends to the sea only at 

 particular places, forming glaciers resembling, in all but magnitude, 

 those which descend into the Alpine valleys. In few if any parts 

 .of the northern hemisphere does the ordinary line of permanent 

 snow descend to the level of the sea. The reason of this is, that 

 the short summer is warm enough to melt away the winter's snow. 

 So effectually is this done, that in Siberia trees grow and crops of 

 rye are harvested over a permanently frozen subsoil. 



But let us suppose, what has repeatedly occurred, that during a 

 period of maximum eccentricity, the precession of the equinoxes gra- 

 dually brings the midsummer of the northern hemisphere round until 

 it coincides with the aphelion. The total amount of heat received 

 from the sun at any given latitude will remain unchanged, but, as 

 shown above, the range of temperature will be greatly lessened. 

 Suppose that at the border of the frozen circumpolar area (that 

 is to say, along the isothermal of 32°) the annual range of tempera- 

 ture is reduced to nothing, so that the temperature is always freez- 

 ing, the result will be that over the frozen area the ice and snow 

 will never melt, and a polar ice-cap will be formed. 



In one way this is a little overstated ; for in no region is the 

 climate quite invariable, and even where the mean temperature 

 of the warmest month does not riso above 32°, there will no doubt 

 be some melting. This, however, will have no effect, except slightly 

 to diminish the extent of the ice- cap at its margin ; and it will 

 probably be much more than counteracted by the spreading of the 

 ice-cap, from the same causes that make a glacier descend. More- 

 over the ice-cap will tend to extend itself, in consequence of the 

 effect of masses of ice in preventing the temperature in their 

 neighbourhood from rising much above the freezing point. For this 

 reason the accumulation of ice will depress the temperature, espe- 

 cially the summer temperature, along the margin of the ice-cap ; 

 and this chilling effect will be spread into lower latitudes by means 

 of cold currents and icebergs. 



The ice-cap, as Mr. Croll has elaborately shown, will, by displacing 

 the earth's centre of gravity, draw a greater share of ocean -water to 

 the glaciated hemisphere ; and this will promote glaciation by 

 diminishing the range of temperature; for the range is always 

 least in oceanic climates. Meantime the non-glaciated hemisphere 

 will have a climate of opposite character — a climate of extremes. 

 This is partly because of the withdrawal of ocean water from it, 

 which will increase the area of land and make the climate more 

 continental ; but chiefly because of the nearness of the sun in sum- 

 mer and his remoteness in winter. The instances of Siberia and 

 North America show that such a climate may produce a vigorous 

 forest vegetation, which appears to be injured by no severity of 



