MERGE TO GYRENE. 439 or the ground immediately about the entrance to them*. The cellse were sometimes sunk to a considerable depth below the levels of the chambers, and contained ranges of bodies or cineral urns placed one above another, each division being separated from that above and beneath it by a slab of stone, resting on a projecting moulding which was raised on two sides of the cella. There are also divisions, in many instances, in the length of the cellae, some of them containing three and four places for bodies on the same level, but these are always ranged (to use a naval phrase) head and stern of each other ; and we never saw an instance in which any two of them were parallel. In fact, the width of the cella, which, we have already stated, was regulated by the space between the columns, would have rendered such an arrangement impossible, since it was of the same breadth in aU parts, whatever might be its extent in length and depth. For a more complete idea of these elegant mansions of the dead we refer our readers to the plates containing the ground-plans and elevations of such of them as we had time to secure on paper. It will be seen that the proportions of the several members of the entablature varied considerably in the few instances given ; and indeed, we may say that there are scarcely two facades where the measurements exactly correspond-f.

* All the excavated tombs were not provided with antechambers, and the celloe in such cases commenced from the surface of the external fa9ade.

4 The metopes are often far from being square, and the mutules ai-e placed at different distances from the triglyphs according to the fancy of the architect. The capital of the triglyphs Is very rarely continued, in the same line, across the metopes ; but is almost always deeper in the last-mentioned division, forming a moulding in the space between the triglyphs, which gives an air of finish tp this part of the entablature.