Professor Percy Kendall—Glacier Lake Channels, 27 
Professor Bonney has favoured me with a copy of a pamphlet! in 
which he subjects my views to a criticism based upon a careful 
examination of parts of the area. Not content with mere destructive 
criticism, he elaborates in some detail an alternative explanation. 
The whole paper is so moderate, and the author’s appreciation of my 
work so generous, that I must break through my self-imposed rule of 
silence, and this I do the more willingly as no other critic has 
ventured to suggest any other hypothesis. 
I may briefly recite the phenomena to be explained. A great series 
of sharply-cut ravines, in some cases veritable gorges, traverses the 
country, either within the area occupied by glacial deposits or just 
beyond its margin. ‘Their transverse section closely resembles that of 
a railway cutting, with which Professor Bonney aptly compares them ; 
there is the steep ‘batter’, the flat floor, and, let it be particularly 
noted, a comparable sharpness in the angles at the top and bottom of 
the slope. Sinuous channels have a steep batter on the outsides 
of bends, and gentler slope on the inside curve, just as we find in the 
ease of the banks of a river, though not commonly of a river valley. 
I draw two inferences from these features: (1) that the channels were 
excavated rapidly by a volume of water sufficient to occupy the whole 
floor, (2) that their formation was so recent that denudation has done 
little to modify the original contours. 
As regards situation, they are found cutting through watersheds, 
even the main water- parting of the whole district (e.g. Newtondale), 
and sometimes trenching. boldly projecting spurs. Often a succession 
of spurs are cut along a line, as though an originally continuous 
¢ehannel had been segmented up by the development of cross-contour 
valleys of normal type. A feature which opposes this interpretation 
is that, instead of the fall-line constituting a continuous gradient 
from end to end of the series, the outfall level of one segment 
generally coincides with the intake-level of the next, even though 
a distance of a mile or two may intervene. This arrangement is 
explicable on the supposition that the intervening space was occupied 
‘by the standing waters of a lake, but is difficult to reconcile with an 
original continuity of the channels. Similar relations have been 
observed in many parts of Scotland and of the North of England. 
Another significant arrangement is that which I have called 
a ‘* Parallel Sequence ’’, that is, a series of parallel channels cutting 
at successively lower levels a single spur. They fall in the same 
direction, and there is a method indicated by their relative heights ; 
each channel, as a rule, commences to break the ridge-line of the spur 
at almost the exact level of the intake of the next higher channel, as 
though—this is my explanation—the channels drained a lake held up by 
an ice-barrier, and when the ice withdrew so as to uncover a slope 
below the intake level, the drainage was diverted and the formation 
of a new channel begun. In some cases six, eight, or even more 
channels traverse the same spur in a distance of a mile or two. 
No comparable case of river-capture is known to me, and I am unable 
to imagine conditions which would produce such an effect. The 
1 On certain Channels attributed to overflow streams from ice-dammed lakes. 
