Page:Archaeological Journal, Volume 29.djvu/398

 332 ODIIIAM CASTLE, HANTS. gone, but the mortar has preserved the beds of the stone more or less perfect. The same stone Nvas used in the interior fur dressings for the openings, and t»r a band about 4 ft. high at the base of the wall, and for the groining of the internal angles. In these two latter positions some of the ashlar has been left undisturbed. About one-third of the tower, including most of the two western faces, has fallen, but though the remainder is very rough, and a mere ma.ss of flint conglomerate, held together bv the excellence of the mortar, the cores of the buttresses remain, and enough of the recesses of the window openings to show something of their original form and dimensions. The tower is composed of a basement and two stories. The basement floor is about G ft. below the exterior ground level. It was about 12 ft high, and six of its eight faces appear to have been }»ierccd. Of tlicsc openings five within commence at 4 ft. from the floor. They were round-headed, and 4 ft. witle. TliL-y converged ujion an ordinary loop, and as tho sill rose by six steps, the base of the loop was about a foot above the exterior ground level. Three of these recesses are tolerably perfect. The sixth 0])ciiing, judging from an appearance in the wall above, m;iy have been the door into the base of a well-stair, ascending in the wall to the summit. ►Such a stair there was likely to have been, and the hollow in the wall is mure like that for a staircase than for a chimney shaft or a garderobe vent, and the weakening of the wall by such staircase would account for its having fallen on this side. The stair, if such it was, occu})icd the south end of the south-west face. Two of the openings in tho basement have been calle<l doorways of entiance from with- out. What remains scarcely leads to this conclusion, and it is exceedingly improbable that there should have been a door on the ground level, when there certainly was one on the firstf floor. iSir K. Home's plan, mentioned below,' shows a sort of 8tairca.se in the centre of tho towei-, as tlioiiL-ii descendinji: to a sub-basement floor. Of this n(»t a (race is visible, and in so wet a soil a chamber mucli iielow the surface would bo usually full of water. As regards the ground-floor entry, it is very po.ssibh; that heri', as usual, a basement wind(»w may, ' III Iho ArcluKologin, vol. xxix. |ilnto without iit tcconiiianicd by uo duMcriptiun, uxl iiiO.