430 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [3Iay 28, 



boulders, and drift deposits. Ice in some form has had much to do 

 with most of them ; but in what form has it acted ? 



The direction of the ice-scratches, the way they run along with 

 the main valleys, although sometimes systematically crossing low 

 watersheds, at once suggests sheet or glacier ice originating within 

 the district ; floating ice would not produce this regular rock- 

 grooving. The moraines, for the most part, bear witness to glaciers 

 each confined to its own valley and of late date. 



The boulders give evidence in two directions. First, it is evident 

 that the direction in which they have travelled agrees with that of 

 the uniform ice-scratches ; and since these last seem due to a more or 

 less general ice-sheet, there is presumptive evidence that many of 

 the boulders have been moved onwards by the same. Secondly, there 

 are some facts pointing to the transport of boulders in directions 

 other than that of the ice-sheet, and into positions into which 

 neither ice-sheet nor smaller glaciers could have carried them ; this 

 second class of facts points to the agency of floating ice. 



With regard to the drift deposits, while the Till of the district 

 probably represents a moraine profonde, the drift-gravel looks like 

 moraine -matter and Till remodelled and partly rounded beneath 

 water ; and the mounds of stratified sand and gravel free from 

 boulders seem to indicate currents of water meeting at certain 

 points and forming sand bars when the climate was milder and 

 there was no floating ice. 



But before ice-sheet and floating ice can have their limits assigned 

 to them, particular notice must be taken of what must have been 

 the configuration of the land at various stages of submergence, and 

 we must consider whether ice would be likely to float in certain 

 directions or not. 



YI. Land -contour at various Stages of Submergence. 



In figs. 1-5 the land-contour is given at various points of sub- 

 mergence. 



Fig. 1 shows that if the land were submerged to a height of 1000 

 feet, that part of the lake-district north of the main watershedding 

 line would communicate with that south of it by only one channel, 

 which we may call the " Straits of Dunmail liaise." The height of 

 this pass (Dunmail Raise) is 783 feet. AW. the other valleys of the 

 area under consideration would be closed fiords, except the great 

 east and west vale of Keswick. 



At the 1250 feet submergence the straits of Dunmail Raise would 

 still be the only through passage (fig. 2). But instead of each of 

 the other valleys — except the vale of Keswick — being simple fiords, 

 the Buttermere and Borrowdale would communicate by the Honister 

 Straits, and the range of mountains on the east side of the Butter- 

 mere and Lorton Valley be split up into three large islands parted 

 from one another by ISTewlands Straits and Whinlatter Straits. 



At 1500 feet (fig. 3), Dunmail Raise would still be the only 

 through strait; but now the Ennerdale fiord would communicate 



