The only appearance of argument on which this hypo thesis rests is, the coincidence and continuity of the low est line of Glen Roy with the terrace at its upper end. It is not true, as has been said, that all the lines terminate alike in terraces. But the coincidence in question proves nothing respecting the identity of the lines and the ter races; of which a very simple explanation will hereafter-be given, on different principles. In the mean time, we may remark, that various sets of them are seen in Glen Roy, and not those which only coincide with the lines. These are the consequences of the action of the present river, where they coincide most nearly with it ; while those at higher elevations were produced by it when it held higher levels, as in all similar situations. A few are the conse quences of the lateral streams flowing into the valley, and their positions are regulated by the former or present course of these.
It is admitted that the action of the present river flow ing in that part of Glen Roy which has the character of a strath, does produce the abrupt forms of these terraces. But it would require circumstances utterly unintelligible, to permit a similar action, in former times, to have produc ed the lines, without having innumerable intermediate marks and terraces, as is the fact now wherever the river is working. As the action of water on a solid terre plein must have been gradual, these remains could not have been distinguished by such decided intervals. The consequences must have been the same in former times as they are at this day, which is by no means the case; nor is it difficult, on the next hypothesis, as we shall shortly see, to account for the few supernumary and irregular marks of this nature formerly described. These, in fact, will be found chiefly in the vicinity of such torrents as have entered laterally into the valley when it was a lake. They are the remains of deltas or terraces resembling those now at the head of Glen Roy, and at the entrance of Glen Turit, which have been worn down by the action of water, and which have just the same sort of connection with the principal lines lions from a direct line. A current thus charged with gravel and scenes, flowing along a channel thus bent, must necessarily have different impressions on the salient and re-entering angles at every flexure. The former, it is plain, must undergo the greatest change from the action of the water on the side opposed to the current, than on that which declines from it. The same effect will take
place at the re-entering angle, but in a reverse order; and het e also the deposit will be least on the side which op poses the stream. But if the current should have con sisted of water alone acting on an alluvium previously de posited, the effects would be contrary. The lines in this case %% ould be most strongly traced where the action of the stream was must powerful, as in the other they would be best marked where least exposed to it. Similar effects would be found wherein the lines enter into the furrows of the hills ; yet, as we have seen, no regard is paid to this in the disposition and forms of these, and that, under all these varying circumstances, their characters are consis tent. The cul de sac formerly mentioned, at the head of lower Glen Roy, could, on this principle, have shown no traces of these lines.
The argument from the nature of Glen Fintec and Glen Turit is somewhat similar. The bottom of the former is so high as to exclude the lower line ; and that of the lat ter, at no great distance from its junction with Glen Roy, excludes the whole three. No current could therefore have flowed through them, so as to have produced the lines now marked on their sides ; and the rocky nature of the bottom of Glen Turit proves that it has not been ele vated since that period by any deposits of alluvial matter; the only supposition which could be opposed to this rea soning.
A general notion of the different capacities of Glen Roy in different places has already been given ; but, to simplify this argument as much as possible, we shall as sume the lowest line as its groundwork. This cuts the bottom of the valley near its upper extermity, so that there is a point at which the vertical depth of the suppos ed current is nothing. The breadth of the valley in the same place is less than half a mile. But, at its lower end, or when it has joined the valley of the Spean, the breadth is near five miles, where the traces of this line are still visible, while its elevation above the bottom of the valley is not less than 800 feet. It must be superfluous to add to this statement, that no body of water could flow through such a valley so as to preserve a level surface, a condition requisite to the production of the line in ques tion. We need not extend these remarks to the two up per lines, as, with some necessary modifications, they would be the same.