this rise in its bed. Its natural outflow was at the south-east corner, and the little stream, after running southwards for a few yards, quickly turned westwards into a narrow valley which wended towards Fail Loch. Just at this abrupt turning the background rose somewhat steeply to the south, so that the termination of the valley as it entered into the Lochlee basin was very liable to be obstructed by débris washed from the slopes above. Besides being thus favourably situated to catch washed-down materials, it is probable that during the summer the surplus water would be very scanty, and vegetation abundant, so that in the course of time the bed of the outlet would gradually be raised. A section cut across the outlet would readily disclose the sequence of the silted materials, had it not been that the soil was disturbed by a deep covered drain, which was made when the first drainage operations were executed, and ran along its whole course. Also, I understand that previous to this the lake was used as a mill-dam. We cannot therefore get rid of the elements of uncertainty in any calculations which might be based on the change of the level of the lake and the accumulation of silt in its bed and outlet. I may however mention, on the grounds already stated, that since the foundations of the crannog were laid, the increase in the bed of the lake in its vicinity cannot have been less than 10 or 14 feet; and 11 feet is the lowest estimate that I can assign to the submergence of its surface. On the supposition that this apparent encroachment of the water was uniform, and since the last fireplace was about 6 feet above the lowest, and allowing 1-1/2 feet for the time the former was used, we have then the total period of occupancy of the crannog represented by 7-1/2 feet of total submergence. We have no means of comparing this period with its representation in so many feet of lake sediment, but I may state that since the canoe, found about 100 yards from the crannog, was abandoned, no less than 5 feet of this mossy lake sediment accumulated over it. [Illustration: PLATE II. PLAN & SECTIONS OF Crannog at Lochlee.] The composition of the silt forming the bed of the lake, especially near the crannog, as already described at pp. 98, 99, points to the fact that for centuries the increase was due principally to the decomposition of vegetable matters, while latterly it was caused more by a deposition of fine clay; and when excavating along the line of the gangway we had an opportunity of verifying the regularity of this succession. A change so marked in the sediment can only be accounted for by a corresponding change in the surrounding scenery, and no explanation is more likely than that the primeval forests had given place to the inroads of agriculture, when some of the upturned virgin soil would be washed down, as it still is, by every trickling rill that finds its way into this lake basin.