X.J ICE AND ITS WORK. 155 



air entangled among its crystals. In squeezing a handful 

 of snow into a snowball, some of this air is forced out, and 

 the loose crystals begin to adhere to one another ; while, by 

 compressing snow very tightly in a hydraulic press, it may 

 be rendered almost homogeneous and thus brought nearly 

 to the condition of ice (see p. 158). In this way, the great 

 pressure exerted by the piles of snow in the Alpine snow- 

 fields compresses the lower layers, and converts them more 

 or less completely into ice. The imperfectly consolidated 

 substance, partly snow and partly ice, is known in Switzer- 

 land as Neve or Firn. Moreover, the water produced by 

 temporary thaw, during sunshine, becomes frozen into ice ; 

 and, in these and other ways, the water, which fell on the 

 mountain-top as loose white snow, is ultimately sent down 

 into the valleys in the form of solid ice. The river of ice 

 which thus drains the high snow-fields is termed a glacier 



(Fig- 39)-' 



Although we have just spoken of a " river of ice," it is not 

 easy, at first, to believe that a substance so solid and rigid can 

 really move in any way like a mobile liquid. Yet the fact 

 that the glacier does so move can easily be demonstrated. 

 Drive a row of stakes firmly into the ice across a glacier and 

 opposite to some well-marked point, as at a in Fig. 40, so 

 that you may know exactly their position. If you examine 

 these stakes a week or two afterwards, you will find that they 

 are no longer at a, but at some point lower down the glacier, 

 say opposite to b. The ice has therefore moved during this 

 time from a to b, carrying the stakes with it. 



From this experiment it is seen that the ice really moves. 

 But the experiment teaches something more than this ; for 

 it will be observed that the stakes have not only moved 

 down, but have changed their relative positions. Instead of 



' From Agassiz's EhicUs sur les Glaciers. Neuchatel, 1840. 



