Vol. 54.] ON THE GLACIAL GEOLOGY OF SPITSBEKGEN. 211 



appeared to play in the economy of the ice-sheet. He found no 

 stream flowing over the surface of the ice, and accordingly scouted 

 the idea that kames and eskers could have been formed in channels 

 upon the surface of the ice-sheet. The observations of other explorers 

 have, however, shown that the conditions of the ice-sheet at the 

 time of Nansen's march must have been exceptional, Jensen, for 

 example, has published views of streams flowing over the surface of 

 the ice-cap.^ 



We were not surprised therefore to find that the Spitsbergen 

 glaciers are intersected by river-channels, sometimes so deep as to 

 be impassable for the sledges, and that numerous streams dis- 

 charge from the terminal face of the glaciers. The stream that we 

 examined most carefully was one on Plower Glacier, the course 

 of which we followed during an ascent of Mount Lusitania. The 

 channels both of the stream and its tributaries were generally quite 

 free of debris. In fact all the supraglacial streams flowed with 

 such velocity that they kept their channels quite clear, except of an 

 occasional boulder. "Whether subglacial streams deposit debris 

 along their courses we could not of course directly see. But their 

 velocity, so far as could be judged from the behaviour of the streams 

 at the mouths of their tunnels, is probably even greater than that 

 of the superficial streams. The latter flow in open channels, whereas 

 the subglacial streams are forced through pipes, the size of which 

 is no doubt kept at the minimum by the weight of ice upon the 

 roof. It is only natural, therefore, that the power of the subglacial 

 currents should be great, and it is not likely that deposits would 

 be formed beneath them. 



We at least found no trace of gravel-deposits resembling eskers 

 formed either in subglacial or supraglacial streams. The moraines 

 of the Ivory Glacier were in places kame-like in form, but they were 

 certainly not formed subglacially. On the plain at the foot of 

 Booming Glacier is a small esker-like ridge, but it is very low, and 

 there is no available evidence as to its formation. 



We were, however, so fortunate as to find a fairly typical esker 

 in a position which left no doubt as to its origin. It occurred 

 along the floor of a valley which descended from Brent Pass to 

 the Sassendal. The valley is marked on Sir Martin Conway's map 

 as the Esker Valley (fig. 1, p. 201 ). 



The esker is about | mile in length, and its course is slightly 

 sinuous. It varies in width at the base from 15 to 40 yards. Its 

 upper end is semi-cylindrical in transverse section, but at the north- 

 eastern end its summit is flattened into a platform 25 yards wide. 

 The summit of the esker is almost horizontal, but the height of the 

 ridge gradually increases from 20 to 35 feet owing to the slope of 

 the valley. The esker rises from a flat, which at the time of our 

 visit was ice-covered ; a stream flowed across the plain on each 

 side of the esker. Beyond the flat a terrace of gravel, similar to 



1 J. A. D. Jensen, ' Expeditionen til Syd Gronland i 1878,' Medd, om 

 Gronl. vol. i (1879) p. 61 & pi. iii. 



