the ends of which crossed each other and were kept in position by the uprights. One group was carefully inspected. The first or lowest beam observed was right across, the next lay lengthways and of course at right angles to the former, then three or four spread out diagonally like a fan, and terminated in other groups at the opposite side of the gangway, and lastly, one again lay lengthways. Thus each beam raised the level of the general structure the exact height of its thickness, though large lozenge-shaped spaces remained in the middle quite clear of any beams. The general breadth of the portion of this unique structure examined was about 10 feet (but an isolated pile was noticed farther out), and its thickness varied from 3 to 4 feet. A large oak plank, some 10 feet long, showing the marks of the sharp-cutting instrument by which it was formed, was found lying on edge at its west side, and beyond the line of piles, but otherwise no remains of a platform were seen. All the beams and stakes were made of oak, and so thoroughly bound together that, though not a single joint, mortise, or pin was discovered, the whole fabric was as firm as a rock. No relics were found in any of the excavations along the line of this gangway." (B. 373, pp. 99-101.) _Structure of the Island._--Having now collected the chief facts regarding the log-pavement, its surrounding structures and superincumbent materials, we determined to sink a shaft about the centre of the crannog for the purpose of ascertaining, if possible, the thickness, composition, and mode of structure, of the island itself. This shaft was rectangular in form, and large enough to allow three men to work in it together. (See plan and sections, page 416.) After removing the three or four layers of oak planks which constituted this portion of the log-pavement, we came upon a thin layer of brushwood, and then large trunks of trees laid in regular beds or layers, each layer having its logs lying parallel to each other, but transversely or sometimes obliquely to those of the layer immediately above or below it. At the west end of the trench, after removing the first and second layers of the log-pavement, we found part of a small canoe hollowed out of an oak trunk, evidently part of an old worn-out one, thus economised and used instead of a prepared log. Much progress in this kind of excavation was by no means an easy task, as it was necessary to keep two men constantly pumping the water which copiously flowed from all directions into the trench, and even then there always remained some at the bottom. As we advanced downwards we encountered layer upon layer of the trunks of trees with the branches closely chopped off, and so soft that the spade easily cut through them. Birch was the prevailing kind of wood, but, occasionally, beams of oak were found with holes at their extremities, through which pegs of oak penetrated into other holes in the logs beneath.