220 ME. E. J. GAEWOOD & DE. J. "W. GEEGOEY [May 1 898, 



terraces at the mouth of the Sassendal similarly descend to the 

 south-east. 



The elevation was most rapid on the south side of Ice Fiord, especi- 

 ally near its mouth, which is the line of greatest recession of ice. 

 This coincidence is in harmony with Baron de Geer's ^ theory that 

 depression and elevation have been often determined by the 

 formation and disappearance of an ice-load. 



Hence the weight of an ice-cap on Snowdon might itself lead 

 to a depression of the area. This would lessen the height up which 

 the Moel Tryfaen shells would have had to be uplifted by a glacier 

 on the one theory. But, on the other hand, it would weaken the 

 argument against the marine origin of the lloel Tryfaen sands 

 based on the absence of evidence of submergence in other districts. 



We therefore express no opinion as to the origin of these drifts, 

 as we see no impossibility either in the local subsidence of the 

 Snowdon area or in the uplift of the shells by land-ice. 



(5) The Mow of Glaciers. 



The internal processes in a glacier by which this uplift of 

 material by ice is possible are intimately connected with the flow or 

 advance of the glacier-ice. The mechanical processes are of three 

 kinds. 



There is first a simple flow like that of the Swiss glaciers, which 

 takes place mainly in the upper layers which are free from intra- 

 glacial material. The ice in these upper layers appears more flexible 

 than that of Switzerland, for the ice in the terminal cornice is often 

 bent and curved. The greater flexibility may perhaps be due to the 

 larger size of the glacier- grains, some of which are enormous. Some 

 of them, in a block which had fallen from Booming Glacier, were 

 4 inches in diameter. These were much larger than any we had 

 seen in Switzerland ; and the biggest that we remember recorded 

 thence were some found by Porel on the Aletsch Glacier, which 

 were as much as 3 inches in diameter. 



The nature of the ice-movement in the lower debris-laden por- 

 tions of the glacier appeared to be very different from that of the 

 upper part, for the ice behaves as if it were rigid and inflexible. 

 The mere existence of vertical ' Chinese walls ' shows that the ice 

 is not capable of rapid change of shape. The advance of the ice is 

 effected not by a flow like that of a viscous substance, but by a 

 continual series of deformations, like a rock which is yielding under 

 great strains. The ice is crushed and fractured, and successive 

 slices are thrust forward along shearing-planes. A certain amount of 

 melting and regelation also takes place, but these processes are much 

 less important than in the case of Alpine glaciers or of the layers 

 free from debris. 



A third type of advance occurs in cases where the upper layers 



1 G. de Geer, ' Quaternary Changes of Level in ScandinaTia,' Bull. Gaol. 

 Soc. Amer. vol. iii (1892) p. 67 ; ' On Pleistocene Changes of Level in Eastern 

 North America,' Proc. Eost. Soc. ]N'at. Hist, vol.xxv (1892) pp. 459, 473-474. 



