Ch. XVI.] 



GKOUND-ICE AND GLACIKKS. 



367 



lias attained a temperature of 32° Fahr. it sets into a sheet 

 of ice. It would seem therefore impossible, according to this 



Ration, that ice should ever form at the bottom 

 of a river ; and yet such is the fact, and many speculations 

 have been hazarded to account for so singular a phenomenon. 

 M. Arago is of opinion that the mechanical action of a 

 running stream produces a circulation by which the entire 

 body of water is mixed up together and cooled alike, and the 

 whole being thus reduced to the freezing point, ice begins to 

 form at the bottom for two reasons, first, because there is 

 less motion there, and secondly, because the water is in 

 contact with solid rock or pebbles which have a cold surface.* 

 Even in the Thames we learn from Dr. Plott that pieces of 

 this kind of ice, having gravel frozen on to their under side, 

 rise up from the bottom in winter, and float on the surface. 

 In the Siberian rivers, Weitz describes large stones as having 

 been brought up from the river's bed in the same manner, 

 and made to float, f It is a common remark in Russia that 

 where the bottom of the stream is muddy, ground-ice forms 

 less readily, and that it is produced most freely when the 

 sky is cloudless. In that case, stones lying in the channel 

 part with their heat by radiation more rapidly. By an ad- 

 mirable provision of nature, it is in those countries where 



r 



most 



from 



ground-ice comes to the aid of the carrying power of running 



water. 



Glaciers. 





—As the atmosphere becomes colder in proportion 

 as we ascend in it, there are mountainous heights even in 

 tropical countries where the heat of summer is insufficient 

 to melt the winter's snow. But to reach the lower limit of 

 perpetual snow at the equator, we must rise to an elevation 

 of about 16,000 feet above the sea (see above, p. 211). In 

 the Swiss Alps, in lat. 46° K, we find the line of perpetual 

 snow descending as low as 8,500 feet above the sea the 

 loftier peaks of the Alpine chain being- from 1 2.000 fn 1 R ooa 



* M. Arago, Annuaire, &c. 1833 ; and 

 Eev. J. Farquharson, Phil. Trans. 1835, 

 p. 329. 



t Journ. of Eoy. Geograph. Soc. 

 vol. vi. p. 416. 



