Grade ProHle in Alpine Glacial Erosion. 275 



ing full width, were level, holding fiord lakes ; or else, more 

 commonly, they had been aggraded to ribbon meadows, 

 on which their streams meandered. Figuratively, I 

 characterized them as valleys down at the heel. But at 

 points far apart compensation of grade was had in steps. 

 Here the draining stream often dropped in a thin sheet, 

 without notch of its own making, or spread in braided 

 cascades. Appeal to rock structure as a determining 

 factor in the development by running water of horizontal 

 elements of such magnitude in the long profile would be 

 futile. The rock of the basins about Mt. Lyell was 

 granitic. Its only large structures were joints — if possible 

 faulting might be disregarded. The joint systems were of 

 all strikes and dips; and some were horizontal. But the. 

 preglacial streams responded to such controls in the same 

 manner that present streams do ; and the present streams 

 were not at grade. At the foot of each step was an 

 alluvial cone and a wash-plain. The streams were 

 engaged in filling. But, obviously, in many instances, 

 they had before them an immense task of grade adjust- 

 ment ; and the measure of that task was the measure only 

 in part of the glacial degradation. 



On the tread of the steps descending to the trough 

 head, where the beginnings of post-glacial refilling were 

 less appreciable, there was persistent occurrence of re- 

 versed grades. In such cases the long floor was ponded 

 backward from the polished brink of a cross-cliff, or 

 cascade descent. In the cirques frequently rock-basin 

 lakes lay deep against the curving talus of the head-wall, 

 visibly shallowing forward. There seemed to be revealed 

 a tendency of the ice-stream vigorously to sink its valley 

 at the head, where its feeding snows gather most deeply ; 



