46 //. F. Reid — Studies of Maiv Glacier. 



As one cipproaclies the end of the glacier the surface of the ice 

 hecomes steeper and steeper, and frequently becomes too pre- 

 cipitous to allow one to stand on it. The diagram (figure 1) 

 shows the form of the surface cut by a longitudinal section. 

 Now, why does the glacier assume this shape ? We know that 

 the surface of a drop of water or of a small quantity of honey on 

 a plate will assume some such shape ; but this is the result of 

 molecular forces which can not have any appreciable effect on 

 large bodies like a glacier. The end of a flowing lava stream 

 Avill have a somewhat similar form, l)ut this is a case of contin- 

 ued flow and not one of equilibrium. These analogies throw 

 no light on the cj^uestion. If we divide the glacier into layers 

 b}^ a series of surfaces parallel to the direction of flow, the con- 

 dition that the end shall be stationary requires that the ice sup- 

 plied by each layer shall be melted at its end. Now, the uj^per 

 layers move more rapidly than the lower 'ones; therefore their 

 ends must melt more rapidly. A glance at the cliagra^n will 

 sliow that, on account of the form of the end of the glacier, the 

 ends of the upper layers expose a larger surface than the lower 

 to the air and sun, resulting in their more rapid melting. This, 

 although undoubtedly a part of the explanation, is not the whole 

 of it, for the form of the glacier's end would be one of unstable 

 equilibrium. If anything should cause the surface to become 

 somewhat steeper, the exposed ends of the upper layers would 

 Ijecome smaller, and these layers would no longer melt away 

 rapidly as they advance ; the surface -would continue to grow 

 steeper until the upper part would break off and thus restore the 

 slope. Although glaciers have been observed to advance I have 

 never heard of it occurring in this manner. A series of meas- 

 urements of the rate and direction of motion, and the rate of 

 melting at the end of some glacier, such as the Gorner or Morte- 

 ratsch, in Switzerland, would undoubtedly throw light on this 

 problem. 



At the end of the valley of Norris glacier, Taku inlet, there is 

 a broad expanse of gravel, etc, on which the glacier, after issuing 

 from its gorge, spreads itself like a great fan, thus presenting a 

 large surface to the air and sun ; so that the melting of the ice 

 is as rapid as the supply.* If it were prevented from spreading 

 it would extend much further than it does, and would undoubt- 



* This level expanse must be either the accumulation of glacial debris 

 or a delta formed Avhen the glacier was less extensive than now. 



