398 The Parallel JRoads of Glen Boy. 



The barrier was then partially removed or breached, and 

 the waters subsided to the level b B, where they con- 

 tinued stationary till a new deposit of alluvium collected 

 at the margin of the loch, and formed the shelf b. The 

 barrier was again lowered or breached ; the waters subsid- 

 ed to the level a A, where they remained stationary till 

 the shelf a was formed. When this took place, the val- 

 leys of the Roy and the Spean formed one loch ; for the 

 shelf a passes from the Roy into the Spean on the same 

 level, and extends along both sides of the latter for a space of 

 20 miles. On the other hand, no trace of the higher shelves 

 b and c are found in the valley of the Spean. Sir Thomas 

 therefore supposes, that when this lowest shelf a was formed, 

 the two lochs were united, and shut in by a common barrier, 

 which was below the junction of the valleys, about High- 

 bridge, near the Great Glen. This barrier in its turn gave 

 way, and the waters subsided gradually to their present level, 

 leaving both valleys dry, and merely traversed by rivulets. 



Glen Gloy in the same way formed a separate loch, shut in 

 by a bai'rier neai* Loch Lochy, which held up the waters 12 

 feet above the greatest height of those in the Roy, till the 

 shelf existing there was formed ; after which the barrier open- 

 ed, and the waters subsided till the valley became dry, with- 

 out forming any other shelves. 



Dr MacCulloch's theory is similar to the above, but he as- 

 sumes the existence of one or two baiTiers only, instead of 

 three. 



The insuperable objection to both theories is, that no trace 

 of the barriers is visible. They must have consisted either of 

 walls of solid rock, or mounds of clay and gravel, stretching 

 across a valley one, two, or three miles wide ; and whether 

 we suppose that they were burst through at once from top to 

 bottom, or worn down by successive stages, it is utterly incon- 

 ceivable that every trace of such walls or mounds could have 

 disappeared. If the nature of the process left this doubtful, 

 the analogous cases of the lake of Subiaco, and of the Swiss 

 lakes which have burst their barriers, would prove the fact. 

 Neither can we assume that a transient deluge had swept 

 away the remnants of the barriers ; for a deluge sufficient to 

 accomplish this would have obliterated every trace of the 



