length of 2ft. 6in. The first fragment is a cornice, or impost, carved on both sides, in three tiers: the upper, a cima with a leaf; the middle division, a Greek fret, not quite similar on each side the stone, and below is a running ornament. The cornice does not project sufficiently to be the cornice of a building, and, as it is decorated on either side, it could not have been intended for a string-course, as none of the walls are so thin as these stones, although I at first thought it might belong to one of the semi-circular exedrae. The curve is struck with a shorter radius than even the smallest recess. I think it is the capping of the back of one of the semi-circular stone seats, called by the later Romans a stibadium. If this formed the seat in the north-western recess, there would be ample room behind it (3ft. 9in.) to pass by. The next fragment must have been fixed beneath this or a similar capping, and is also carved on each side; the convex side having an adaptation of the well-known honeysuckle fairly drawn, whilst the convex side of it, with the exception of a floriated panelled pilaster in the centre, is the work of an accomplished sculptor. On the right of this pilaster, slightly recessed to admit of relief, is the naked right thigh and leg of a figure that must have stood 1ft. 6in. high. Although only a fragment, this is a most charming piece of work, the action and anatomy of the limb being perfect. On the left side is a similar panel, a headless draped figure, with feet bare, holding a circular shield which rests on the thigh, whilst the limb is bent as if ascending a rock that is slightly indicated. On the third fragment the honeysuckle pattern is on the concave side, whilst the sculpture is on the convex, the arc of which corresponds with the last described. On this there are two niches only, and the figures are much more mutilated. The left figure has a flowing mantle, the only leg remaining being bare from the thigh downwards; the foot and the head are gone. The figure on the right is fully draped, the head is lost, and the right hand much mutilated; a musical instrument, like a guitar,²⁴ or rather a mandolin, rests against the left breast, held in position by the left hand. The fourth fragment has the honeysuckle on both sides, with the flower well carved on one of them. It is a great pity that so little of this superb work is left, and that what there is should be so mutilated.²⁵

This account of the Great Bath will, I hope, be sufficiently complete if I describe the entrances and conclude with a few particulars of the pavement (although many discoveries of considerable interest might be made, I have no doubt, in the latter), omitting a detailed examination as being tedious.

I believe there were five entrances to this bath, two of which remain. In the western wall, on the south, is one leading from other apartments (a hypocaust, hall and bath), which I shall on a future occasion describe. It is 4ft. 3in. wide. Double doors and hinges have been inserted in this doorway, and the base and a portion of a pilaster cut away most barbarously to receive them. On the north, on the same wall, and fronting the northern schola, is a doorway similar to the last, which has been walled up in Roman times, the wall which closed it being covered with the red plaster that covers all the work not being faced freestone. A third doorway, similar in every respect, was at the eastern end of the northern schola, as I infer from the lower paving being much worn in that direction. A fourth doorway was in the eastern wall to the south, but not south enough to face the southern schola, and a fifth was between these two. Of these three doorways, the first of them is still hidden by soil, and the second and third are obliterated with modern walling; a portion of the architrave of one was found near, but their position is well marked by the footmarks in the stone.