438 



MR. BERXAKD SMITH ON THE 



[Sept. 191 2, 



rocks when the western margin of the ice, west of the Duddon 

 Valley, rested against the rising slope of Lathrigg at nearly 

 800 feet. A similar but smaller valley occurs in parallel sequence 

 with Windy Slack about 200 yards farther east. 



The head of the Knott-Gill Channel is about a mile to the souths 

 and lies just below the 600 -foot contour-line, that is to say, a 

 little lower than the lower end of Windy Slack. It may, therefore, 

 have been in operation at the same time as Windy Slack, the water 

 crossing the gap between them by traversing the glacier that 

 occupied the Elack Beck valley. 



The marginal channel takes a south-westerly trend across the 

 south-eastern shoulder of Knott Hill. It is a Y-shaped valley, 

 000 yards long, with a rather uneven Hoor descending about 

 150 feet, that is, at about the same rate as Windy Slack. It is 

 excavated in volcanic rocks, but ends in boulder- clay on the lower 

 ground. jN'ear both its head and foot there is a little gravel 

 (fig. 15, below). 



At that time the Whicham Lake was probably in existence. 



Fig. 15. — The Knott-Gill marginal channel, loohing 

 nortli-north-eastwards. 



,-- .3.S. 



[This view shows the steep gradient common to the marginal channels which 

 are associated with the steeply-descending inland glaciers.] 



The ground on the west of the Duddon Valley north of Windy 

 Slack was not examined in detail, but a dry valley was visible 

 between the 500- and 600-foot contours on the eastern slope of 

 Penn Hill, north of Logan Beck ; and I have little doubt that 

 others Avould be found in similar situations as far north as 

 Seathwaite. 



(3) Sand and Gravel of the Plain. 



The coast-sections of sand and gravel which form so large a 

 ])art of the plain have been already described (p. 416), and we 

 have seen that these beds are to some extent stratified. Inland 

 sections confirm this conclusion. The drift between Bootle and 



