of the rivers had been made to flow east, or out of the upper ends of the supposed glacier-lakes, instead of escaping at the lower ends, in a westerly direction, where the great blockages of ice are assumed to have occurred.
In addition to these arguments of Mr. Jamieson, I may mention that in Switzerland, at present, no testacea live in the cold waters of glacier-lakes; so that the entire absence of fossil shells, whether marine or freshwater, in the stratified materials of each shelf, would be accounted for, if the theory above mentioned be embraced.
When I examined 'the parallel roads' in 1825, in company with Dr. Buckland, neither this glacier theory nor Mr. Darwin's suggestion of ancient sea-margins had been proposed, and I have never since revisited Lochaber. But I retain in my memory a vivid recollection of the scenery and physical features of the district, and I now consider the glacier-lake theory as affording by far the most satisfactory solution of this difficult problem. The objection to it, which until lately appeared to be the most formidable, and which led Mr. Robert Chambers in his 'Sea Margins' to reject it entirely, was the difficulty of conceiving how the waters could be made to stand so high in Glen Roy, as to allow the upper most shelf to be formed. Grant a barrier of ice in the lower part of the glen, of sufficient altitude to stop the waters from flowing westward, still, what prevented them from escaping over the 'col' at the head of Glen Glaster? This 'col' coincides exactly in level, as Mr. Milne Home first ascertained, with the second or middle shelf of Glen Roy. The difficulty here stated appears now to be removed by supposing that the higher lines or roads were formed before the lower ones, and when the quantity of ice was most in excess. We must imagine that at the time when the uppermost shelf of Glen Roy was forming in a shallow lake, the lower part of that glen was filled up with ice, and, according to Mr. Jamieson, a
