174 



Messrs. R. M. Deelev and P. H. Par 



r on 



Between the 2610 and the 2520 contours, near the middle 

 of the glacier, are some crevasses which are nearly parallel 

 with the centre line of the glacier. Here, again, thev are in 

 positions where the ice shear is small and quite unable to 

 produce appreciable viscous-flow stresses. 



Fiff. 12. 



\«, 











^^>rr^ 





2700 TKeters. 





1 ^S^ 2 6 SO ■■ 







I M^ 



2 600 - 



i 



In all cases where the ice slope is becoming steeper 

 crevasses are formed, bat where the slope is becoming less 

 steep they do not appear. However, the crevasses do not 

 run straight across the glacier, but point upwards. There 

 must, therefore, be some other stress set up when a change 

 o£ slope takes place, which gives the upward direction to the 

 crevasses. 



One of the principal causes of the changes of slope which 

 a glacier undergoes at the sides is due to the presence of 

 shoulders of rock, separated by bays, which run down the 

 mountain sides. Where the glacier is mounting such a 

 shoulder it is also getting narrower, and two compression 

 stresses are produced at about right angles to each other. 

 AY hen the glacier has mounted the shoulder, the increasing- 

 angle of slope brings into play a tensile stress tending to 

 produce fractures normal with the valley sides. At the 

 same time the sides of the glacier sink into the bay suc- 

 ceeding the shoulder, and a tensile stress is set up whose 

 direction tends to produce longitudinal fractures. The 

 actual crevasses are along the resultant line of maximum stress 

 produced by these two forces. Although the differential 



