Mr Thomson on the Parallel Roads of Lochaber. 53 



the bottom of it, does not appear to have been visited, and 

 certainly has not been correctly described by any former 

 observer. But this is not all. Mr Milne has also traced 

 the channel of an ancient river, proceeding from the water- 

 shed in question, down into Glen Spean, and there termi- 

 nating in a huge delta, or alluvial deposit, at the only shelf 

 which winds round the sides of the latter glen, thus marking 

 the point where the turbid waters of the river were swal- 

 lowed up under the stagnant surface of the lake which, by 

 these same indications, is palpably shewn to have stood in 

 Glen Spean on a level with the lowest shelf, at the time when 

 Glen Roy was occupied with water to the height of the shelf 

 next above. 



In connection with these circumstances, Mr Milne finds 

 that the uppermost shelf of Glen Roy does not, as was erro- 

 neously indicated on Sir Thomas Dick Lauder's map, run 

 round the sides of Glen Glaster, but that it suddenly stops 

 short in Glen Roy, just above the entrance to that smaller 

 tributary glen. 



From this we conclude, that the barrier which blocked up 

 Glen Roy, so as to occasion the formation of its highest 

 shelf, must have disconnected it from Glen Glaster, and thus 

 forced it to discharge its surplus water into the valley of the 

 Spey by the summit-level at its head, instead of permitting 

 it to discharge by the lower summit-level at the head of Glen 

 Glaster, and down by the ancient rivei"-channel into Glen 

 Spean, — a course which must have been followed by any 

 water occupying Glen Glaster, or communicating with it 

 uninterruptedly. 



Now, to explain the formation of the highest shelf of Glen 

 Roy, Agassiz supposed one glacier, in the lower part of Glen 

 Spean, to have extended across from Ben Nevis to Moel 

 Dhu ; and another, farther up that glen, to have issued from 

 the valley of Loch Treig ; the two being sufficiently high and 

 extensive to maintain the water between them, and, of course, 

 also the water in Glen Roy, at the level of the shelf under 

 consideration. In confirmation of this supposition, he stated 

 that the shelf is marked on the south side of Glen Spean, be- 

 tween the sites of the two supposed glaciers. Were the sup- 



