The artificial work, measured from the present surface, was more than five feet six inches in thickness. On the beams (_layer 2 in section_) there seemed to have been a basket flooring, but of this there could not be certainty, owing to the depth of water in the excavation. In connexion with them were observed _wicker-walls_ made of hazel rods, and where these crossed the oak beams there were round holes through the latter for the stakes to pass through. In the vicinity of these beams a small rude stone implement of Silurian grit was found, but so rotten that it broke when handled. The beams in _layer_ 3 were charred on the under surface, as if they had formed the rafters of an edifice destroyed by fire. Near the north of the island, immediately over these charred beams, there was a plank pointing N.W. and S.E.; it was ten feet long, twenty inches wide, and two inches thick; at about one foot nine inches from its N.W. end there were two holes through it, running N. and S. in a line with the north and south piles seen on the N. shore, and on each side of the plank there were upright stakes ranging in the same direction. The conclusion may be drawn that the island was occupied at two distinct periods--the first being when the E. and W. beams formed the floor of the habitations which seemed to have been destroyed by fire: afterwards the oak sticks and stones--irregularly laid--were placed as a floor for new dwellings, and between the periods it must be supposed that the water of the lake had risen considerably.[237] COUNTY WATERFORD. [Illustration: Fig. 212.--Plan of Submarine Crannog at Ardmore. One-half the scale of fig. 213.]