THE MECHANICS OF GLACIERS 923 



of the dissipator the stratification is but slightly, if at all, 

 indicated by dust bands. The strata should be well defined 

 at the lower end, but the large amount of debris on the 

 surface and in the crevasses, would make them difficult to 

 recognize. 



The debris in the lower layers of some glaciers may have a 

 similar origin, and may not have been picked up from the 

 ground beneath. Indeed, in general a downward rather than an 

 upward motion probably occurs near the bottom, and some 

 of the material forming the ground-moraine may have been 

 originally dust or rock fallen on that part of the reservoir's 

 surface, near the encircling cliffs, which afterwards formed the 

 lower layers of the glacier. For, on account of the heat derived 

 from the earth, two feet of ice would be melted from the glacier's 

 under surface in a hundred years, and would deposit its debris 

 under the ice. 



For7n of the glacier' s surface. — For equilibrium the melting of 

 the ice at every point of the surface must exactly equal the 

 supply ; this is the necessary condition ; and since the upper 

 layers move fastest, their melting must be fastest ; this is brought 

 about by the slope of the surface being least where the motion is 

 most rapid, thus exposing to melting a larger surface of the layer. 

 Or we may say that the angle between the direction of motion 

 and the surface is greatest where the motion is smallest and the 

 melting most active. At the end of the glacier, therefore, and 

 near the ground, we find the slope steepest ; and as we pass up 

 along the surface of the ice the motion becomes faster and the 

 melting less and consequently the slope diminishes. This also 

 explains the rounded plan of the glacier's end and the depressed 

 sides. If we knew accurately the distribution of velocity and 

 of melting we could calculate the shape of the surface for 

 equilibrium ; but without this our approximate knowledge of 

 these quantities enables us to see that the general form of the 

 dissipator's surface agrees with the theory. 



Glaciers are continually changing in size, so that the con- 

 ditions for equilibrium can be looked upon as approximately 



