G. W. Buhnan — Glacial Geology. 339 



(2) Tbe products of stream and ice action would necessarily be 

 mingled during the continuance of the cold period. For, during 

 suunner, ice action would give place to that of water near the 

 termination of the ice-sheet or glacier ; and thus truly aqueous 

 deposits would be formed over the glacial, to be in turn overlaid 

 by the latter. 



A succession of exceptionally warm summers, too, might cause 

 the temporary retreat of the ice — ^just as in recent years a succes- 

 sion of exceptionally severe winters caused the formation of a 

 veritable glacier in one of the Scotch valleys — and permit of the 

 formation of a thicker series of sands and gravels. 



And as near the extremity of a glacier or ice-sheet stream action 

 takes the place of that of ice every summer, and to a more marked 

 extent during a succession of warmer seasons, so we should expect 

 to find the retreating ice, as the glacial epoch passed away, followed 

 by a series of mixed flu via tile and glacial deposits. 



And as oscillations in the extent of existing glaciers are well 

 known, so we naturally infer it was in the past. 



As regards the present day, Prof. Prestwich quotes the case of 

 a glacier which has advanced a mile in a century, and that in a 

 region where there is a genei'al retreat of the ice. 



And Mr. Lydekker, in his account of the Geology of Baltistan, 

 Kashmir, describes one glacier which is retreating, and another 

 which is advancing. Thus of the Tapsa glacier he writes : 



" A comparatively modern terminal moraine, forming a dome- 

 shaped hill covered with cypress, bounds the cultivated ground 

 superiorly ; above this old moraine is the present terminal moraine 

 of the glacier. The above appearances seem to indicate that the 

 Tapsa glacier has receded by small gradations, pausing here and 

 there, until it finally attained its present shrunken dimensions." 

 But of the Palma glacier : 



" The present Palma glacier shows pretty evident signs of being 

 on the increase, since it terminates inferiorly in an abrupt and 

 precipitous wall of ice, with but comparatively little debris and no 

 distinct terminal moraine, which seems to have been overflowed and, 

 so to speak, swallowed up by the glacier." ' 



The possibility of lacustrine action in glacial times is also sug- 

 gested by the usually received interpretation of the parallel roads 

 of Glen Eoy. These are held to have been formed by a lake caused 

 by the blocking up of the lower part of the valley by ice. While 

 the lake stood at different levels, the action of the water formed a 

 series of shelves on the mountain side along the margin of the lake. 

 But it is difficult to resist the conviction, that if the lake was fed by 

 streams, more or less lacustrine deposits would be formed in it : in 

 other words that gravel, sand, and silt would be laid down, and 

 afterwards overlaid by typical glacial deposits. 



And stream action also, it is well known, occurs beneath the ice- 

 sheet and glacier. Springs, for example, issuing from the ground 

 beneath the ice, will cut for themselves channels through it. The 

 1 Records Geol. Surv. India, vol. xiv. pp. 44-45. 



