the terraces, he saw that if the mouth of Glen Gluoy were stopped by a barrier the waters from the surrounding mountains would be collected in the valley until they had reached the water-shed, when any further rise would be prevented by the branch-valley, which would carry the additional water off to Glen Roy. As long, then, as the barrier remained, there would be a lake in Glen Gluoy, at the exact level of the road, which, by constant action upon the loose drift,
would be sufficient to produce the road. Now, if the mouth of Glen Roy should also be barred at the same time by a sufficiently high barrier, the waters would be collected behind it, the surface of the lake would rise till it reached the water-shed dividing Glen Roy from Glen Spean, when the superabundant water would flow into the latter valley. In this way the highest shelf of Glen Roy would be formed. If its barrier were now to be partly removed, so as to establish a connection between it and the upper part of Glen Spean, while the lower part remained blocked up, upper Glen Spean and Glen Roy would then be occupied by a continuous lake, the level of which would be determined by the water-shed discovered in Glen Spean. The water in Glen Roy would take a level corresponding to its new place of escape, and the lowest parallel road would be formed. The conclusions thus drawn would be strictly logical, if proof could be offered as to the existence of the barriers.
In Glen Spean there is a large quantity of detritus, and Sir Thomas Dick-Lander supposed that this had at one time been heaped up by some unknown convulsion. As he could not account for the middle