s 
480 . MR. P. F, KENDALL ON A SYSTEM OF [ Aug. 1902, 
(4) Overflow-channels.—The overflows of glacier-lakes have 
attracted but little attention in this country. Those of the Glen- 
Roy Lakes, so far as can be gathered from descriptions, present no 
marked features of eroded channels, but are merely scoured cols 
which have been swept fairly clean by the effluents of the lakes. 
Nevertheless I think that one or two of them show features of 
particular interest, judging by the maps and the behaviour of some 
streams related te them. Dugald Bell appears to have been the 
first geologist to have recognized that an ancient lake might be 
postulated from its overflow with as great certainty as by any other 
sign ; and as I have already remarked, he described ! a lake-overflow 
at Dalry, enforcing his conclusion by the evidence of floor-deposits. 
Prof. James Geikie made similar observations quite independently.’ 
In America such phenomena are among the regular data of 
glacialists. The most noteworthy in the whole.continent is no 
doubt the great overflow of Lake Agassiz, known as the Warren 
River, which is described at length by Mr. Upham in his monograph 
already cited. Mr. Upham, in allusion to these phenomena, says that 
he regards them as evidences of glacial lakes ‘ the most invariably 
recognizable and the most definite in their testimony,’ an pimN 
aan I most cordially endorse. 
As these features are but little known to British pealomeee and 
my experience of them extends to over three hundred examples, 
chiefly in Yorkshire, but also in Northumberland, Westmoreland, 
Lancashire, Cheshire, and Lincolnshire, it seems desirable that | 
should describe the general characters with some care, the more so 
as I have placed my chief reliance upon them in the identification 
of the series of lakes which form the subject of this paper. 
When an ice-sheet or glacier obstructs the drainage of a country, 
the water is impounded so as to form a lake, which will find 
an escape either through, or over, the ice; or, by overflowing, by 
some col] or spur in the surrounding watershed. 
In the former case, the evidence remaining to the geologist after 
the disappearance of the glacier will consist solely of beaches, 
deltas, and lacustrine deposits; but in the latter, and in such a 
country as ours, the more common case, of overflow across a col, or 
spur, the water proceeds to cut for itself a channel. The deepening 
of this channel may be swift or slow according to local conditions. 
Upon the withdrawal of the ice, it may be that a morainic obstruction 
may be left across the valley of such a height that a lake may 
persist for a considerable period, until, in fact, the overflow has 
cut far enough down to drain it, and then remains as the permanent 
course of the drainage. Or the overflow-channel may have cut 
1 Trans. Geol. Soc. Glasgow, vol. iv (1874) p 
2 «Great Ice Age’ 2nd ed. (1877) p. 146. i Be. however, to point out 
that a consideration of the peculiar conditions prevailing in the Clyde area at 
the close of the Glacial Period seems to forbid the ascription of any considerable 
part of the excavation of the valley in question to a lake-overflow of the same 
date as the deposits. 
™ 
