9i6 



HARRY FIELDING REID 



glacier small in comparison to its breadth. After a certain time 

 the sections would be deformed to A' B' and CD', if the ice 

 moved everywhere with its normal velocity. The slope of the 

 deformed section at any point would depend only on the weight 

 of the ice above that point, i. e., on its depth below the surface of 

 the glacier, and therefore the form of the section CD' would 

 be exactly the same as that of the upper part of A'B' . As the 



D D' 



Fig. I. Displacement of sections by flow. 



slope of E'B' is greater than that oi F' B' ,\t is evident that 

 more ice would enter through EB than would leave through FD 

 in the same time, EF' being parallel to the bed of the glacier ; 

 but this is impossible under the supposition of normal flow, for 

 we have only considered a part of the glacier below the surface, 

 where no melting takes place. ^ The tendency of the thicker 

 sections to move faster than the thinner ones, at the same 

 distance from the bed, causes a forward pressure on the latter, 

 increasing their velocity, and a backward pressure on the former, 

 diminishing their velocity. The ice near the lower end of the 

 glacier is thus under a pressure greater than the normal pressure^ 



number of pieces of ice on a slightly inclined rough sandstone slab, and found that 

 the mass slowly slid down the slope with a uniform velocity approximately propoi- 

 tional to the pressure against the slab and to the angle of inclination, for angles 

 between i° and io°. When the sandstone was smoothed but not polished a motion 

 was observed at an inclination of forty minutes. The ordinary laws of the friction 

 between solid bodies do not apply at all to the forces between a glacier and its bed. 



'We here neglect the melting at the lower surface of the glacier which is so small 

 as not to alter the argument. 



^ The normal pressure is the pressure which would exist if the glacier were of 

 infinite length and uniform section. 



