Geology 



At first sight it might appear possible to determine 

 the number of confluent tributaries by that of the medial 

 rock-trains on the surface of the main stream, and this 

 indeed would be so, were it not for the splitting of rock- 

 trains, which we have already observed, and for the fact 

 that a medial train may sometimes arise in quite another 

 manner in the middle of a glacier. Such a case we 

 were able to examine some miles farther up the glacier, 

 and not far below the point where it left the great snow- 

 field which gave it birth. 



Before reaching this point, however, we had to cross 

 a great stretch of broken ice, with numerous deep 

 crevasses and thousands of fantastic pinnacles or seracs. 

 This ice-fall, for such it was, was the result of a steep 

 place in the valley floor, and was produced in much the 

 same way as a rapid or cascade is formed in a stream 

 of water, except that ice broken into fragments, so to 

 speak, by the more violent movement on the steep 

 slope, takes longer to reunite than water under similar 

 circumstances. 



That it does eventually so reunite is proved by the 

 fact that, a mile below, the surface of the glacier showed 

 no signs of its tumultuous passage over the ice-fall, 

 except that there was little rock ddbris on its surface, 

 this having been engulfed by the crevasses amongst the 

 seracs. 



The ice-fall in the brilliant sunlight made a wonder- 

 ful spectacle (Plate XIV.), but its passage was attended 

 by much difficulty and danger. We had to proceed 

 with the utmost caution, roped together, and cutting 

 steps with our ice-axes at frequent intervals. Many 

 times after a laborious ascent of a pinnacle, we found 



270 



