peculiar profiles of typical Alpine valley floors and walls is 

 futile, and would show that the argument had been based 

 on ignorance of stream mechanics. The very fact that the 

 present water streams now occupy tiny canons only, which 

 have been excavated in the flat floors of the old channel, 

 shows first, that the large channel was formed by some 

 other agent much more voluminous than the present 

 ordinary streams which drain these peculiar floors, and 

 second that the water stream has only just commenced 

 action along such floors. 



This slight digression was necessary to make clear to the 

 student that glaciers form large channels as compared with 

 those formed by smaller water streams, and that under 

 certain conditions one must expect to find cirques thou- 

 sands of feet in depth, and also of great length and width. 

 It is simply ;i question of the combination of glacial volume, 

 of the steepness of the old channel floor, and of the width 

 and depth of the old stream formed valley itself. 



To return to the consideration of the formation of the 

 cirque. In the first place since the typical cirque is only 

 developed at a glacial divide (and not some distance down 

 stream nor under an ice-swept col) it is clear that the cor- 

 rasive action of the ice or snow at its lip is a negligible 

 quantity. In the second place it is clear that the cirque is 

 a considerable gathering ground itself. Thirdly, the volume 

 of the ice will increase towards the base of the cirque and 

 therefore its mobility will be much increased. This implies 

 a great increase in corrasive power. Fourthly, since the 



ciples of stream mechanics show that such ice stream must 

 have been in a state of compression in the cirque so long 

 as those conditions of heavy volume obtained. The ice 

 stream must have maintained a continuous surface over 



