36 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [June 4, 



The melting of ice has been suggested as the cause of debacles 

 capable of transporting boulders ; but the melting of ice could never 

 produce a debacle in the rigid sense of the word — I mean such a 

 debacle as would be produced by the bursting of a waterspout, or 

 the head of a reservoir, or an earthquake wave ; the laws of matter 

 prevent it. The conversion of sensible into latent heat is necessarily 

 a work of time : floods, possibly of great violence, might result from 

 such a cause, capable of moving the greatest masses ; but their ac- 

 tion would be continuous, and they would necessarily separate the 

 larger from the smaller fragments, and all of them from the clay in 

 which they are imbedded ; they would be arranged both according 

 to their size and their gravity ; but neither of these is the case, and 

 I must conclude, with Sir James Hall, that such effects are " inex- 

 plicable by any diurnal cause." But even if this difficulty could 

 be got rid of, a more insuperable one meets us ; such floods must 

 necessarily run down the hills, and into the natural lines of drainage 

 of the country. In the west of Scotland the great line of drainage 

 is marked by the course of the river Clyde, and floods poured from 

 the contiguous mountains must inevitably follow its course to the 

 sea; but the very reverse of this is the case. I never yet saw or 

 heard of an erratic block in the valley of the Clyde, where its course 

 could be traced, that did not come in an opposite direction to the 

 flow of the river. We can trace their course, not from the moun- 

 tains to the sea, but from the sea to the mountains. Mr. Milne, in 

 another paper on the geology of Roxburghshire*, after noticing 

 " that the parent rocks were in all cases to the westward of the 

 boulders," contends that they could not be transported by glaciers ; 

 and after noticing the line of drift, he adds, " A glacier which trans- 

 ported (boulders) from Crifi^el to the hills^of Liddesdale, besides being 

 forty miles long, must have crossed the valleys of the Nith, Annan, 

 Esk and Tarras rivers, as well as the high ridges separating them ; 

 it must have done so without having any lateral barriers to retain and 

 guide it ; and lastly, it must have moved up the valley of the Liddel 

 for at least fifteen miles of its course." If this argument is good 

 against the action of glaciers in transporting boulders, it is still 

 better when applied to the case of boulders assumed to have been 

 transported by glacial floods. It appears to me to be perfectly con- 

 clusive against both hypotheses. 



I am very far from contesting the former existence of glaciers in 

 Scotland; on the contrary, I consider it in the highest degree probable 

 that they did exist ; and when we find scratches on the vertical faces 

 of rocks not following a definite course, but having the same direction 

 as that of the valleys, and when we find boulders arranged as they are 

 in moraines, then the action of glaciers affords a natural and satisfactory 

 solution; but these are exceptional cases, and after many years' study 

 of this deposit, I have never met with any. The action of icebergs has 

 also been suggested ; here also I have no difficulty in admitting the 

 probability of their former existence, and there is proof that much of 



* Transactions of the Royal Society of Edinburgh, vol. xv. p. 480. 



