MERGE TO GYRENE. 451 and the places for the bodies were sunk in the floor itself and covered with tablets of stone. In such cases we often see that two, or more, bodies have been ranged parallel with each other round the sides of the chamber, in the manner represented in the ground-plans (page 464), a circumstance which never occurs in the cellm, as we have already stated above.

The galleries which are formed in one side of this ravine lead round the cliff into another valley, somewhat broader, in which are also several excavated tombs. In one of these, which has been furnished with a Doric portico, Mr. Campbell discovered the suite of beautiful little subjects which we have given with all the fidelity we could command in the plate (page 456). They are painted on the zophorus 'of an interior facade, of which we have given the elevation ; and each composition occupies one of the metopes, the pannel of which appears to have been left plain in order to set off the colours of the figures. The outline of these highly finished little groups has been very carefully put in with red: the local colour of the flesh and draperies have then been filled in with body colour, and the lights touched on sharp, with a full and free pencil, which reminded us strongly of the beautiful execution of the paintings at Herculaneum and Pompeii.

There is no other attempt at light and shadow in any of them but that of deepening the local colour of the drapery in two or three places, where the folds are intended to be more strongly marked than in others ; the flesh being left (so far as can at present be ascertained) with no variation of the local colour produced either by light or 3 M 2