OF THE EARTH'S SURFACE. 193 



I beg leave, in the first place, to avail myself of this testi- 

 mony, in proof of the fact, that a stream, at least equal in force 

 to the Garvalt, has, at one period, flowed over the rock at the 

 summit of Corstorphine Hill. 



It now remains a question of importance to say, from what 

 source such a stream could have flowed, and how so much wa- 

 ter could be collected. The spot under consideration, is not 

 above twenty feet perpendicular below the extreme summit, so 

 that the heaviest thunder shower, could scarcely produce there 

 a sensible run of water ; and if such a run were produced, it 

 would have nothing to carry, the whole consisting of one hard 

 and firm rock. The dressing is of large extent in itself, as the 

 bare rock shews, and it is also connected, more or less intimately, 

 with the dressed spots on the western face, Nos. 6, 7, 8, 9, 12, & 

 14., so that a stream of considerable breadth as well as power 

 would be required to fulfil the conditions of the case. The diffi- 

 culty is rendered more pressing still by two cases already men- 

 tioned ; one at No. 11., and the other at No. 13., in both of 

 which cases the rock is visibly dressed, in a spot which con- 

 stitutes the local summit; the last of these, No 13. being the 

 actual summit of this hill, and the highest point within the dis- 

 tance of many miles ; and the other being separated from any 

 higher ground by ravines, so that, in the present state of things, 

 no water could flow towards either of them. 



The only explanation that could be given of these facts, by 

 means of the diurnal agents alone, must be obtained by 

 going back (as has been done in a theory discussed in an ear- 

 ly part of this paper), to some very remote period, when this 

 hill was much higher, and more extensive than at present, and 

 when a stream equal to the Garvalt could be collected above 

 those spots, and could run over them, so as to produce the 

 dressings under consideration ; and, by supposing, that in sub- 



Vol. VII. B b sequent 



