36 ME. J. E. MARR ON THE TARNS OF LAKELAND. [Peb. 1895, 



rapidly until it reached the level of the rock, and then in the 

 majority of instances the stream would cut sideways along the 

 junction between drift and solid rock until, when the stream 

 reached its original position, the lake would be drained. But if a 

 subsidiary ridge of rock lay between the position attained by the 

 stream issuing from the lake and the position of the former valley- 

 bottom, denudation would be retarded to so great an extent (owing to 

 the absence of sediment in the water where it issues from the tarn) 

 that the lakelet would become much more permanent, and its depth 

 would be the difference between the height above sea-level of the 

 bottom of the old moraine-filled valley and that of the present exit 



Fig. 1. — Ideal section across moraine-blocked valley. 



[For Explanation, see p. 47.] 



(see fig. 1). On account of the fugitive character of the moraine- 

 dammed tarn, where the present exit lies over the old valley-bottom, 

 and the comparatively permanent nature of the tarn whose exit 

 does not overlie that bottom, we should expect, even if all the tarns 

 of Lakeland were really moraine-blocked, that the surviving ones 

 should chiefly pour their surplus water over rock, and consequently a 

 mere examination of the exit of a lake gives no proof of the existence 

 of a true rock-basin. To show that such exists it is necessary to 

 prove convincingly that no former valley, now filled with drift, 

 occurs, which might have conveyed the waters away before it 

 became blocked. An examination of a large number of tarns has 

 convinced me that a valley of this kind did exist in a large number 

 of cases, and may have existed in all the others which I have studied. 



(Since writing this paper, I find the following remark on p. 43 of 

 1 The Glacial Geology of Great Britain and Ireland,' by the late 

 Prof. Carvill Lewis, referring to moraine-blocked lakes : — " If the 

 moraine is compact, an outlet for the lake is often cut through 

 the rocks, and a post-glacial gorge is formed." Prof. Carvill Lewis 

 mentions, but does not describe cases of large lakes which he considers 

 due to this cause.) 



I propose now to consider these tarns in detail. In doing so it 

 will be convenient to deal with them in the order in which I 

 examined them, treating them under the following heads: — (1) Tarns 

 near Haweswater ; (2) Tarns between the Grasmere and Langdale 

 Valleys ; (3) Coniston Tarns ; (4) Tarns between the Duddon 

 Valley and Wastdale ; (5) Tarns in the Scawfell Group ; and (6) 

 Tarns between Borrowdale and Thirlmere. 



