Tlic Rate of Ice-Jioiv. 45 



111 addition to the flags, five stakes were planted in a line and 

 about equal distances apart on the eastern side of the glacier, as 

 shown in the map. Their movement was determined from 

 August 6 to 29. The table gives the total movement during 

 that time at right angles to the line of the flags, which was the 

 direction of the slope. The direction of the moraine shows that 

 this is approximately the direction of motion. 



This amounts to al>out 2" a da}^ for the middle flag. 



Conditions holding at the Ends of Glaciers. 



Alpine Glaciers. — It has been long recognized that the com- 

 paratively stationary position of the end of a glacier is due to the 



Figure 1 — E)id uf an Alpine (Tlacier. 



general eciuality between the Cjuantity of ice flowing down and 

 the quantity melted. The mean temperature of a valley increases 

 as we descend ; if, therefore, the end of the glacier should advance 

 l>eyond the point where the rate of melting equals the rate of 

 suj^ply, the ice would melt more rapidly and the end would 

 recede. If, on the other hand, the glacier should not reach this 

 point, ice would flow down faster than it would melt, and the 

 end of the glacier would advance. This point is not merely a 

 point of equilibrium, but a point of stable equilibrium. This 

 explanation is sufficient, so long as we merely look upon the end 

 of a glacier as a whole. But when we consider each part of the 

 end by itself we are met by difficulties which do not seem so far 

 to have been noticed. 



