When the deeper structures of the Lochlee crannog were examined, it was particularly noted that there was no flattening of any of the large logs, as if they had been subjected to great pressure. At Lochspouts and Buston, so far as the water permitted of similar explorations being made, this observation was equally applicable to them. Though quite soft, the logs always preserved their original shape and contour. One day, at Lochspouts, I was greatly puzzled by finding what was evidently portion of a birch-tree, from 6 to 9 inches in diameter, quite flat, and with scarcely any wood left inside the thick bark. In no instance previously had I seen the evidence of pressure on logs of this size; but after carefully considering the point, it was ascertained that such effects occurred only in the upper portion of the mound, and above the log pavement, where the wood had been exposed to atmospheric influences, so that when the woody fibres rotted away, the flattening of the bark was easily produced. All the logs found buried in water or mud retained their original dimensions, and showed no traces of having yielded to superincumbent pressure. In calculating the pressure of the entire crannog on the lake bottom, it is only necessary to take into account the weight of the materials above the surface of the island, as the greater density of the displaced water would act as an upward pressure sufficient, before the mass attained its equilibrium, to allow the surface of the island to project a few feet above the level of the water, the amount varying according to the depth of the lake. After the island grounded, if constructed on the principles suggested at page 262, any additional structures would act as a direct weight on the bed of the lake; but in estimating the final and total submergence due to this element, we must consider the weight of rubbish gradually accumulated during the period of occupancy. As the base of the island at Lochlee was only 4 feet below the level of that of the gangway, it follows therefore that the maximum result due to the weight of the island could not exceed this amount.