ATREVIDA GLACIER 271 



there was distinct evidence of thrust faulting (plate 16, figure 2). At 

 this point the forAvard movement of the glacier is estimated to have 

 pushed the terminus ahead not less than 100 yards. 



It was evident not only that the ice had advanced and become broken 

 during the period of ten months since we last saw it, but that it was 

 even then moving forward. By this movement great crevasses liad been 

 opened and the ice, hitherto blanketed by a thick cover of moraine, with 

 forest superimposed on it in places, was now exposed to rapid ablation. 

 As a result of this marked increase in exposure of ice to melting, great 

 streams of turbid water were running away from the margin and down 

 through the forest, where during the previous summer there had been 

 only trickling brooks of clear water which seeped through the morainic 

 soil as the underlying ice slowly melted. Some of these streams flowed 

 where there had previously been no channel, and these were destroying 

 the forest by the deposit of sediment. Others, occupying former brook 

 beds, had cut them deeply, especially where flowing through loose mo- 

 raine. Altogether it was a wonderful transformation for so short a 

 period, and its effect was shown in Dalton creek, the main glacial stream 

 from the Atrevida, which was greatly swollen as compared to its condi- 

 tion in the previous summer. 



The next day, on ascending the low mountain which borders the east- 

 ern margin of the Atrevida, we had a bird's-eye view of the glacier from 

 the same point where we had looked down on it the previous summer 

 (plate 18). The change observed from this point of view was even more 

 remarkable than that witnessed at the margin. The easily traversed, 

 undulating surface of the moraine-covered outer portion beyond the 

 mountain, and that part farther up the valley wliich in 1905 was clear ice 

 and snow, were now transformed to a labyrinth of crevasses extending 

 from one side of the glacier to the other. Melting liad in places trans- 

 formed the crevassed surface to pinnacled ice reseml)ling that of a typical 

 valley glacier ice-fall. It had also allowed so much of the moraine, 

 veneer to slide into the crevasses that the surface was no longer an undu- 

 lating waste of moraine, but a broken mass, fully half of whose area con- 

 sisted of clear glacier ice. 



The zone of crevassing extended from near the head of the valley far 

 out beyond the mountain base into the alder-covered terminal portion. 

 It gradually died out in a series of great concentric gashes, bringing to 

 view clear ice in the alder thicket (plate 18). The crevasses were cres- 

 centic and roughly parallel to the bulb-like expansion of the glacier be- 

 yond the mountain front. By the breaking of the ice, and by its melting, 



