have no hesitation in affirming that it would be a physical impossibility for a river like the Roy to flow over and down such declivities and leave no notch behind. I have seen a stream which had during a flood changed its course, cut out in a few days a channel which centuries would fail to efface. In uncultivated ground such inequalities of surface are very slowly obliterated. In such places 1500 or 1600 years of neglect have done little to efface the old Roman roads or camps ; the still older British hill-circles and vitrified forts are still distinctly marked ; and there is no reason to believe these river-beds more perishable. The Glen-Roy lines are the best proof how durable such markings are in such conditions. In Glen Roy itself there is much curious evidence both of the effects of river-action and the durability of the marks it leaves. The delta-terraces formed where the streams fell into the old bays on the level of the lines, as at the entrance to Glen Turrit, were of course immediately cut into by the rivers when laid dry by the retiring of the waters. Many of these old watercourses, which had been cut out by the Turrit and other streams before they finally settled down into their present beds, are still easily seen. Some of these are nearly as old as the line they accompany, and older than the lower lines — and thus prove that there has been no surface-change here sufficient to obliterate the former river- channels, had they ever existed.
I might now leave the question to be decided on the evidence adduced ; but there are a few other facts corroborating the same views that may be mentioned. And first, though these lines are in some respects unique in character, there are other indications of the former presence of the sea in this region. Thus in the valley of the Spean there is, as it were, a continuous series of terraces continuing the lines at intervals down to the sea-level. Such wide terraces, or shingle beaches, are well seen along the Spean from Roy Bridge downwards. Another similar terrace is seen at the mouth of Loch Treig on the level of the third or lower line. The sea has stood here for a long period, as the hills above this sheet of detritus are washed very bare, and in some places a well-marked shore-cliff has been cut immediately above it. The detritus here, however, has not come down Loch Treig, as might at first sight be imagined, but from the Corry Laire to the west, and has then been swept eastwards by the tidal currents, and even up into Loch Treig, on which it abuts with a bold, almost vertical end. In many other places in this part of Glen Spean there is similar evidence of a cur- rent from the west flowing up the valley. Thus the lower or western sides of the knolls and rocks are bare, and the detritus accumulated in long mounds or tails behind (that is, above them) to the east. This could not have occurred in an inland lake, where there are no currents, but must have taken place in a marine channel open from sea to sea.
Further, there is a vast amount of evidence of the former presence of the sea at various levels between the upper and lower Glen-Roy lines in all the surrounding region. This evidence also is specially distinct in the valley of the Spey, and thus on the other side of the watershed and in the very valleys into which the lakes
