192 ON THE REVOLUTIONS 



surface, and carrying large blocks of stone along with it. The 

 number of examples concurring in the same indications of di- 

 rection, might easily have been greatly augmented, as any person 

 will see who examines the summit and western face of this hill ; 

 but I have judged that number to be already sufficiently great. 

 I have taken care to denote each spot alluded to, in a manner so 

 particular, that it might easily be found at this moment : I say 

 at this moment, because, as is well known to every geologist, 

 nothing is more precarious than the existence of such specimens, 

 they being liable perpetually to be overgrown with vegetation, 

 or covered with rubbish, or perhaps completely annihilated by 

 the progress of the same operations which brought them into 

 view. These specimens are by no means all equally interesting. 

 Some of them are brought forward as simple indications of di- 

 rection, and owe their value to the force which the argument 

 derives from the concurrence of such a number of results. 

 Those which are deserving of notice on their own account, are 

 Nos. 5, 8, 10, & 15. 



By considering together four or five of these examples, 

 which lie at a small horizontal distance from each other, we 

 shall be led to a conclusion which could not be deduced from 

 any one of them singly. I have said, that close to the summit 

 of the hill upon the west, an extensive space (No. 10.) of 

 rock occurs nearly bare, displaying a succession of parallel 

 grooves, pointing in the direction of the supposed stream, 

 and upon which some of the minute dressings also are ap- 

 parent. On examining this spot, two members of this Socie- 

 ty, whose ideas on this subject are hostile to mine, immediate- 

 ly remarked, that the appearance reminded them of what they 

 had seen in Arran, in the bed of the Garvalt, where the gra- 

 nite has been dressed by that torrent ; and they alleged as a con- 

 sequence, that the phenomenon was reduced to a common event. 



I 



