Mackintosh— The Sea v. Rivers. 



157 



a more correct estimate of the limited extent of fluviatile action in 

 valleys traversed by rivers. The smaller valleys of the Malvern 

 Hills -were referred to during a discussion at the last British Asso- 

 ciation Meeting ; and a hint was then thrown out that they may 

 have been formed by streams arising from melting snows towards 

 the close of the Glacial period. But that they could only have been 

 formed by the sea will, I think, appear from the following consider- 

 ations : — -Their lower entrance is in general narrow, and they widen 

 out very smoothly upwards, uutil they either graduate into the slope 

 of the nearest summit, or into a shallow pass or col. They some- 

 times open at the very summit of the ridge, where no stream from 

 the melting snow could possibly have originated ; and between the 

 highest ground and the lowest of these valleys there is not sufficient 

 space for a stream to have formed. They are of the same shape 

 with the deeper valleys or passes which at intervals divide the hills 



Coves on the Western Coast of the Malvern Straits. 



(such as the valley terminating in a combe behind North Malvern, 

 and the pass leading from Great to West Malvern), only differing 

 from them in extent ; and no one would assert that the latter could 

 have been excavated by any torrent of water derived from the very 

 narrow saddle-shaped ridge of the Malverns. Both passes and 

 shallow valleys present clear indications of their having been 

 scooped out by an agency assailing the hills from without, and 

 operating in an inward and upward direction. That the sea was 

 the denuding agent becomes more apparent when we trace the con- 

 nection between these hollows and the very decided combes which 

 may be seen farther to the south behind Malvern Wells. These 

 smooth hollows have been scooped out in hard Gneissic rocks, 

 composed of Hornblende, Mica, Felspar, and Quartz, and cannot.be 

 explained by the waste of softer materials, for here the Trappean 

 protrusions, which, farther to the north, have guided the excavating 

 agent, are nearly if not entirely absent. Of the three combes 

 represented in the accompanying woodcut, one, I believe, is quite 



