of Valleys and Lakes. 305 



I have said of the confluence of immense glaciers like those that 

 once united in the valley of the Lago Maggiore at what are now 

 the Borromean Isles. But it seems to me that to any one who 

 allows any excavating power to a glacier, it will be evident that 

 when the general inclination of a valley was comparatively steep, 

 a glacier could have had no opportunity of cutting for itself any 

 special basin-shaped hollows. Its course, with a difference, is 

 like that of a torrent. But in a flat-bottomed part of a valley, 

 or in a comparative plain that lies at the base of a mountain- 

 range, the case is not the same. For instance, to take an ex- 

 treme case, if a glacier tumble over a slope of 45°, no one would 

 dream of the ice-flow producing any special effect, except that 

 in the long run, the upper edge of the rock that forms the cata- 

 ract being worn away, its average angle would be lowered. And 

 so of minor slopes ; if the ice flowing fast (for a glacier) rendered 

 the rocky surface underneath unequal, such inequalities could not 

 become great and permanent ; for the rapidly flowing ice would 

 attack the projecting parts with greater power and effect than the 

 minor hollows, and so preserve an approximate uniformity, or an 

 average angle of moderate inclination. But when a monstrous 

 glacier descended into a comparative plain, or into a low, flat 

 valley, the case was different. There, to use homely phrases, 

 the ice had time to select soft places for excavation, and there, if, 

 from the confluence of large glaciers, or for other reasons, the 

 downward pressure of the ice was of extra amount, the excavating 

 effect, I contend, must have been unusually great in special areas, 

 and have resulted in the formation of rock-bound hollows. And 

 though the glacier of Ivrea has been constantly quoted as a case 

 that completely proves the absurdity of my theory, this merely 

 shows the un wariness of those who quote it ; for not only are 

 there a great many rock-basins full of water above Ivrea in 

 among the vast roches moutonnees near the opening of the plain, 

 but, where beyond this point the glacier spread out so wide on 

 the Pliocene plain, it has scooped away so much material that 

 parts of that plain are below the average level of the plains of 

 Piedmont that lie outside the great moraine. Given sufficient 

 time and extension of the glacier, and more matter still would 

 have gone away. The same argument equally applies to the 

 case on the Lake of Zurich, where glacier debris is said to lie on 

 alluvial detritus. In reply to the question why in the actual 

 valley of Aosta there are no lake-basins, I might with equal 

 propriety say, Many contorted regions are much faulted, and 

 there is often an evident connexion between contortion and faults ; 

 but in some contorted regions there are few or no faults, and 

 the reason of their absence remains to be accounted for. I have 

 attempted to explain why the rock-basins are present, and not 



