The Theory of Glacier motion. 147 



all possibility of melting and regelation. . . , . That there is 

 such extension, and that it goes on continuously with all 

 :stresses above i kilo, per square centimetre and at all 

 temperatures between — 6° C. and freezing point, is shewn by 

 the above experiments. ... In the discussion, for the most 

 part a priori^ on the extensibility of ice, sufficient im- 

 portance has not usually been assigned to the necessity of 

 •distinguishing between the effect of even a small blow or jar, 

 and that of a much greater force applied gradually and 

 steadily during a long interval. A bar of ice may bear a 

 stress of 4 and 5 kilos per square centimetre if the load is 

 steady, which would fracture at once with a much smaller 

 sudden stress, especially if not uniformly distributed " {Pro. 

 Roy. Soc, XLII. 491-501). . . . 



In the following year similar experiments were repeated 

 iDy Messrs. Mc.Connell and Kidd, " who established," to use 

 their own words, " that not merely . the rate, but even the 

 very existence, of the extension, depends on the structure 

 of the ice." They shewed that when the ice consists of a 

 single crystal, no extension takes place, thus, pro tanto, 

 explaining the results of some of the earlier experiments. 

 Thus clear ice cut from the surface of a bath, which proved 

 to be a rough regular crystal, shewed hardly any extension ; 

 ice of irregular structure made in a mould, shewed consider- 

 ably more. This made it obvious, that for the purpose of 

 testing the question .as applied to the motion of glaciers, 

 actual glacier ice must be used. " And the experiment was 

 next tried upon actual glacier ice, taken from the natural 

 caves at the foot of the Morteratsch Glacier." To use the 

 actual words of the experimenters : — 



" We tested three pieces, which were quite sufficient to 

 disprove the common notions, that glacier .ice. is only plastic 

 under pressure not under tension, and that regelation is an 

 essential part of the process. They showed at the same time 

 the extreme variability of the phenomenon. The first extended 



