MECHANICS OF GLACIERS. 69 



eaeh glacier-stream must maintain an open channel for itself ; for 

 this reason the same objection applies to it as to a river flowing in 

 an open valley, as an agent of excavation of basin-like hollows. 

 Here and there, where we see the convergent surface-streams of 

 the glacier rushing down the shaft of a movJin, and carrying from 

 time to time earth and stones from the surface-moraines, a certain 

 slight work of excavation is no doubt accomplished, such as we see 

 now exposed to the light of day in the well-known "glacier-garden" 

 at Lucerne ; but this would hardly meet the requirements needed 

 for the excavation of lake-basins. The main glacier-stream, just 

 like any other stream, acts of course as an erosive agent and 

 deepens the trough in which the glacier moves, an excellent ex- 

 ample of which I observed only last summer at the end of the 

 Hochjoch Glacier, at the head of the Rofen Thai, where the ice 

 forms an arch resting upon the sides of the gorge as its buttresses. 



Again it is undeniable that an advancing glacier may do a certain 

 amount of " ploughing " work, such as Prof. Tyndall has described 

 in connexion with the Gorner Glacier *. This, however, is only 

 evidence of the inability of the glacier-ice to move over the super- 

 ficial obstacles which it encounters in its path ; it would seem to 

 tell rather against the notion of the iee being driven up-hill out of a 

 basin, as is assumed by some writers ; and it would be interesting to 

 inquire whether such phenomena have ever been observed where a 

 glacier was not descending a slope. All that has been put forward 

 with reference to the distribution of the potential energy due to 

 gravitation in the procession of the glacier down a valley must tell 

 equally against the propulsion of it up-hill. 



On the other hand, we may, I think, draw a distinction between 

 the " ice plough" and what I may be allowed to call the " ice chisel." 



In cases where successive portions of a glacier descend a vertical, 

 or nearly vertical, precipice, a different set of mechanical conditions 

 is presented to us. Impact may in this case perhaps do the work 

 of excavation to such an extent as is represented in the excavation 

 of many small rock-basins, such as some of those which lie upon 

 the Bernina Pass, or at the foot of the precipices of Snowdon. But 

 when all this is admitted we have no right to reason from " the 

 hundreds of tarns that are found in all glaciated mountain-countries " 

 to the formation of lake-basins which form long depressions in 

 narrow valleys. Many tarns, however, do not occupy rock-basins 

 at all ; some are formed by moraine-heaps left by the retreated 

 glaciers, between which and the mountain-side they may be 

 frequently observed to lie in the Alps ; other tarns simply fill 

 depressions formed by earth-movements on the mountain -slope, 

 where the clayey materials produced by the disintegration of horn- 

 blendic, augitic, and felspathic debris get loosened by water and move 

 unequally downwards, as clay often does on a smaller scale in a 

 railway-cutting. Numerous instances of such could be pointed to 

 in the Alps and in other mountain regions. 



It is not without regret personally that I find myself driven to 

 * Forms of Water. 



