94 GEOLOGY OF ARRAN. 



We have already attempted to account for the origin of the 

 glens (Art. 32). Denudation by water, acting along fractures 

 when the land was rising from beneath the sea, seems the 

 most probable cause. Some geologists would ascribe their 

 origin wholly to scooping out by ice; M. Necker, to their 

 being in the line of great dikes, along which, as in many 

 cases already cited by us, disintegration would be more rapid. 

 Whatever the cause, there must have been some original 

 difference of surface to determine the first action. The 

 origin of the corries is still more difficult to account for. It 

 has been supposed, that if such a bowl-shaped hollow existed 

 when the snow and ice began to become permanent on the 

 summits, ice would depend in great sheets from the rim, and 

 work all round on the rock to its disintegration and removal, 

 so as gradually to widen the corrie. But what gave the rim 

 in the first instance ? It must have been of considerable 

 size to bring into play this supposed ice action, much larger 

 than any hollow that could have been formed like those on 

 the Corriegills shore, already alluded to (Art. 34), as exca- 

 vated by the constant rolling about of large stones : so small 

 a depression could scarcely determine the beginning of ice 

 action. Volcanic action has been suggested but may be 

 dismissed as inadmissible, since we know that granite was 

 not erupted as volcanic matter. We can then only 

 conclude that the same cause which removed the mantling 

 slate from over the granite (Art. 13), and began to mould 

 the peaks, may have determined the first formation of the 

 hollows. 



As regards the after work, we are inclined to adopt the 

 view very ably argued by the Rev. T. G. Bonney in a recent 

 paper on Alpine cirques, which are very similar to the corries 

 of our mountain tracts, that water, not ice, has been the agent 

 in their formation (Jour. Geol. Soc., Aug., 1871). Streamlets 

 from the upper heights trickling over the rim would "notch 

 the edge, seam the cliffs, and undermine the base with their 

 spray." In this way a number of small streams towards the 



