CLOUDS AND RIVERS, ICE AND GLACIERS. 179 



455. It is now our duty to take up the problem, and 

 to solve it if we can. On the neves of the Col du 

 Geant, and other glaciers, we have found great cracks, 

 and faults, and Bergschrunds, exposing deep sections of 

 the neve ; and on these sections we have found marked 

 the edges of half-consolidated strata evidently produced 

 by successive falls of snow. The neve is stratified be- 

 cause its supply of material from the atmosphere is 

 intermittent, and when we first observed the blue veins 

 we were disposed to regard them as due to this stratifi- 

 cation. 



456. But observation and reflexion soon dispelled 

 this notion. Indeed it could hardly stand in the pre- 

 sence of the single fact that at the bases of the ice-falls 

 the veins are always vertical, or nearly so. We saw no 

 way of explaining how the horizontal strata of the neve 

 could be so tilted up at the base of the fall as to be set 

 on edge. Nor is the aspect of the veins that of stratifi- 

 cation. 



457. On the central portions of the cascades, more- 

 over, there are no signs of the veins. At the bases 

 they first appear, reaching in each case their maximum 

 development a little below the base. As you and I 

 stood upon the heights above the Zasenberg and scru- 

 tinised the cascade of the Strahleck branch of the 

 Grindelwald glacier, we could not doubt that the base 

 of the fall was the birthplace of the veins. We called 

 this portion of the glacier a ' Structure Mill/ intimating 



