Mr Thomson on the Parallel Roads of Lochaber. 55 



mountains than it melts away ; and it is thus usually pre- 

 vented from spreading to any considerable extent over the 

 plains. In the Antai'ctic Continent, on the contrary, the 

 mean temperature is nowhere so high as the freezing point. 

 The ice, therefoi"e, which descends from the hills, unites itself 

 with that which is deposited from the atmosphere on the 

 plains ; and the whole becomes consolidated into one conti- 

 nuous mass, of immense depth, which glides gradually on- 

 ward towards the ocean. The portions which are protruded 

 out to sea break oif, and are floated away as icebergs ; the 

 remainder being left, presenting to the sea a perpendicular 

 face which rises, in insurmountable clitfs, to the height of 

 from 150 to 20U feet above the water, and extends below the 

 water to the depth of perhaps 1000 feet. 



Now, a climate somewhere intermediate between these ex- 

 tremes appears to be that which would be requisite to form 

 the shelves in the glens of Lochaber. The climate of Swit- 

 zerland would be too warm to admit of a sufficient horizontal 

 extension of the glaciers ; that of the Antarctic Continent 

 too cold to allow the lakes to remain unfrozen. If the climate 

 of Scotland were again to become such that the mean tem- 

 perature of Glen Spean would be not much above the freezing 

 point, there seems to be every reason to believe that that glen 

 would again be nearly filled with an enormous mass of ice ; 

 while its upper parts, and also Glen Roy, would be occupied 

 by lakes, which would once more beat upon the ancient and 

 long-deserted beaches, — that the rivers would resume their 

 former channels, flowing out of the lakes by the summit- 

 levels between the glens, — and that the ancient aspect of the 

 country would, in all respects, be again restored. 



It will perhaps be objected, that in imagining the ice to 

 make its way into Glen Roy, we are supposing it to flow up 

 hill. A semifluid mass, however, so long as its upper surface 

 slopes downwards, cannot be regarded as flowing up hill, no 

 matter what may be the form of the bottom on which it rests. 

 If a slightly-inclined trough or channel have an opening made 

 in one side, at the middle of its length, and if a stream of 

 thick mud be kept flowing into it by this opening, the mud 

 will not all turn suddenly round towards the lower end of the 



