MR MILNE ON THE PARALLEL ROADS OF LOCHABER. 411 



are incorrect. (2.) That, assuming as true the facts stated by him, they still 

 afford no evidence that glaciers existed in the Lochaber valleys. 



(1.) There are three main facts relied on by M. Agassiz. He states, First, 

 That in Glen Roy, and in that part of Glen Spean, between Bridge of Roy and Loch 

 Treig, there are 3 shelves visible ; Secondly, That these shelves all terminate, on 

 both sides of the valley at or near the Bridge of Roy ; Third, That the bottom of 

 Glen Spean, in front of Loch Treig, is not only polished with that polish charac- 

 teristic of glaciers, but is, moreover, scratched transversely, — that is to say, at 

 right angles to the direction of the valley, by a cause which evidently proceeded 

 from Loch Treig. 



To explain these appearances, it is suggested, that " the supposition of a 

 great glacier descending from Ben Nevis, and shutting up the valley of the 

 Spean, by resting on Moeldhn, which is opposite, combined with the influence 

 of a glacier from Loch Treig, and which would bar the valley a second time at 

 that height, would explain all the facts.'''' 



These facts, for an explanation of which this theory was invented, appear to 

 me not to have been accurately observed. In the first place, the three shelves do 

 not occupy, as M. Agassiz asserts, " all the sinuosities of the lower part of Glen 

 Spean, and of the whole of Glen Roy." It is only the lowest of the three shelves, 

 which occvu-s in Glen Spean and in the lower part of Glen Roy. The two upper- 

 most shelves stop short of the mouth of Glen Roy, by about 2 miles ; so that, if 

 the Lake in Glen Roy was dammed back by a terminal moraine, that moraine 

 could not have rested on Moeldhu, at the foot of Glen Roy ; but must have been 

 pushed up that valley, before the Ben Nevis glacier, 2 miles farther, — an opera- 

 tion which the levels, distance, and direction of the valley would have rendered 

 impossible. 



In the second place, the shelves do not, as M. Agassiz says, " terminate at 

 the same point,'' — viz., at Moeldhu, where he supposes the terminal moraine of 

 the Nevis glacier to have been. The two uppermost shelves (as just stated) do 

 not come within two mUes of this point ; and the lowermost shelf, instead of 

 terminating there, runs, as formerly explained, several miles northwards, on both 

 sides of the valley, towards Unachan, where they are 4 miles apart. It is scarcely 

 necessary to say, that a moraine in this low district, which is not connected with 

 any Ben Nevis valley, and considering its required height and length, is incon- 

 ceivable. 



In the third place, as to the existence of transverse scratches on the rocks in 

 Glen Spean, which are said to indicate the movement of some body from Loch 

 Treig, I could see no such scratches, though I twice sm-veyed the ground, and 

 narrowly inspected the rocks, especially at the outlet from Loch Treig. Indeed, 

 the supposition that any glacier flowed out of Loch Treig seems to be almost ex- 

 cluded by the fact, that a shelf, perfectly horizontal, exists on both sides of the 



VOL. XVI. PART III. 5 M 



