Now the lines in question must in all cases have form ed the shores of this lake ; and as these are found at three distinct elevations in Glen Roy at least, it is plain that this lake existed in these different states, and that the re lative depths of these three accumulations of water may be measured by the critical distances of these lines from the bottom of the valley ; allowing for such waste as the operations of the river may have produced since it was drained. The outline of this lake must, therefore, also have varied at these different elevations, as must the na ture of its communication with the surrounding valleys. Without maps, we could not pretend to convey an idea of these variations and connections. From the certainty that the lake of Glen Roy thus occupied three different depths, the nature of the retaining obstacles becomes more intri cate, as does that of the operations by which they must have been removed.
Now, in examining the correspondence of the pheno. mena in Glen Roy with the theory of a lake, it is ne cessary to clear away some imaginary difficulties from ap pearances that have originated in other causes. Those to which we here allude are the terraces. It was shown that the hills here were covered with a coat of sharp or un transported alluvium, appearing to be merely the ruins of the rocks above, and that, in other places, there was rounded alluvial matter deposited in such a manner as to indicate a transportation from places more distant. The same appearances take place in the other valleys. It was also shown that, in the lower parts of the valleys, vertical ly considered, there were found terraces at different le vels, and often with surfaces of considerable dimensions. These accompany the lateral entrances of the streams, and the principal junctions of waters ; and as they are ne cessarily proportioned to the magnitude and power of these, they are most considerable at the entrances of the Roy and of the waters of Glen Turit. At the upper end of the former glen. the most remarkable of these coin cides with the lowest line, presenting a joint level continu ously prolonged. In these places, they are still subject to the action of the river ; from which cause they gradually waste away, and become diminished in their superficial di mensions. Thus also the lateral wanderings of the river multiplies their number, producing a numerous series at different levels, which accompany the course of the stream.
We may now examine how far all these appearances coin cide with those which are found connected with existing lakes at the present day, and from this examination we shall see that the terraces are phenomena of an indepen dent nature, or different from those which form the lines in question ; connected with, but not originating in, the same actions which produced these.
Where a lake is inclosed by hills of considerable decli vity, which, though formed of rock, is also covered with al luvial matters, we find it skirted by a gravelly shore, which forms an inclined plane, and constitutes a zone at the level of the water. This is of greater or less breadth, according to the declivity of the hill, the quantity of allu vial matter present, and other circumstances, by which these results are occasionally modified. When we examine such a lake by sounding, or when we enter its margin, we find that, after a short time, it deepens suddenly ; and, if the section is carefully examined by the sounding line, it will be found that the declivity from the point downward!: resembles that of the hill above the water. The shallow one is therefore a shore applied to the face of the hill, and not coinciding with its general outline.
If now rocks protrude into the water on the margin of a lake so situated, the shore in question is either imper fectly marked or altogether wanting ; these circumstances being regulated by the particular inclinations of these rocky points, and by other variations, which we need not here notice more minutely. Where the declivity is great est, the shore is not only narrowest, but its inclination is also the greatest ; and, on the contrary, it is most level, and of the largest dimensions, where that is least. Where rivers enter the lake, there also terraces are found, which, at the water level, must necessarily coincide with the shores themselves ; and these terraces, or rather deltas, are of the greatest dimensions at the entrance of the prin cipal stream, where, in progress of time, they even form extensive plains, so as at length to exclude the water.
Now we know not that we could have described any ap pearances more exactly coinciding with those of Glen Roy. If any one line and its terraces be traced, under all the modifications visible, it is in every point a perfect ex emplification of this description of the shores of an exist in lake. It is not, therefore, to be doubted, that, if such a lake were now to be suddenly drained, as we suppose Glen Roy to have been, it would have precisely the ap pearances which occur in that valley, as far, at least, as all the requisite conditions were present. The conditions which have thus led to the present state of Glen Roy, as far as regards the prolonged extent and uniform appear ance of the lines, may be found in the regularity and general similarity of the forms of the including hills, in their uniform and rarely interrupted faces, and in the gen erally equal thickness of the alluvia by which they are covered. Wherever, indeed, these conditions are want ing, those very anamolies occur which we ought to ex pect on the theory thus laid down.