Mr Milne on the Parallel Ixoads of Lochaber, 349 



the waters could have been dammed up in the valleys to the height 

 of the several shelves. 



Shelf 2 is distinctly marked on both sides of Glen Ivoy, down to 

 a certain point, — and also on both sides of Glen Collarig, down to a 

 certain point. At this period, the water flowed from the east end of 

 Glen Roy into the valley of the Spoy. Something must have ex- 

 isted, therefore, in both glens at the points above referred to, to pre- . 

 vent the extension of the shelf westward. 



Shelf 3, in both glens, extends a little more to the west than 

 shelf 2. We have seen that, whilst Glen Glaster is exempt from 

 shelf 2, it is well marked on both sides by shelf 3. 



To explain these facts, I assume that there was a blockage of 

 some sort, in Glen Roy, which filled the lower part of the valley up 

 to the level of shelf 2, and which blockasje extended a little farther 

 east than the mouth of Glen Glaster. I assume also a similar block- 

 age in Glen Collarig, which filled the lower part of the valley, and 

 as far eastward as the place where shelf 2 stops in that glen. This 

 blockage may have been gravel, clay, or any other detrital matter. 



Such is the supposed state of things, whilst the waters stood at 

 shelf 2 in Glen Roy ; at which period, it will be remembered, they 

 were discharged to the eastward. 



Former writers have assumed, that when the waters sunk from 

 shelf 2, the amount of sinking must have been 82 feet, the distance 

 of shelf 3 below shelf 2 ; and that this sinking had been one act, 

 caused by an earthquake, or other violent operation, which all at 

 once lowered the barrier by that number of feet. But this is a mis- 

 take. MacCuUoch takes notice of a shelf faintly marked on Tom- 

 bhran hill, between shelf 2 and shelf 3, though he expresses after- 

 wards some uncertainty about it. In fact, there are two interme- 

 diate shelves visible there ; and they are also discernible, at precisely 

 the same level on Ben Erin, and also more distinctly near Acha- 

 vaddy, on the south side of Glen Roy ; the one being about 14 feet 

 below shelf 2, and the other about 36 feet lower down.* These 

 two intermediate shelves clearly indicate, that the water which filled 

 the valley, did not all at once sink from shelf 2 to shelf 3. They 

 prove that the water first sunk down 14 feet, and was stationary at 

 this level for some time ; that it then sunk down other 36 feet, and 

 continued at this level for some time ; and that it again sunk other 

 32 feet, at which level it remained for a much longer period, till it 

 formed shelf 3. 



It is evident, from these facts, that the lowering of the barrier (of 

 whatever material composed) which confined the water in Glen Roy, 

 was a process of a more gradual and ordinary description than what 



'•' There are hummocks or knolls of stratified gravel and sand in Glen Glas- 

 ter, the tops of which are all about 36 feet above shelf 3. it is pi'obiibie that 

 they were deposited when the lake stood at one or other of the intermediate 

 }>oints last mentioned. 



