MERGE TO GYRENE. 535 disposition of the columns is by no means evident in this temple, and the number is very uncertain ; but the ambulatory must have been a good deal below the pavement of the aedes, since there is no space allowed for it upon the basement we have mentioned ; and it must consequently have been upon the terrace beneath it, which a})})ears to have been left for that purpose. The columns must therefore have been unconnected with the roof, and have merely supported the covering of the ambulatory. Indeed, it seems likely that the portico was altogether detached from the aedes, and judging from the remains of a wall, which appears to be part of the original plan, and the position of a column without it, we may perhaps suppose that it was situated on the edge of the terrace above-mentioned ; and that the whole space of four-and-twenty feet between this wall and that of the fcdes, was a space between the portico and the body of the temple, which does not appear to have been covered in. In this disposition we imagine the wall just alluded to to have formed the back of the portico, and the column, still remaining, to have been one of the range which supported its roof in front. Immediately below this column the ground descends, and traces may be observed of steps leading up to it.

In the aedes there seem to have been only apronaos and cella; and in the latter is a detached mass of building, raised above the level of the other parts of it, for which we are wholly at a loss to account, there being no analogy between its disposition and that of any part of a cella in its usual arrangements.

The length of the aedes is a hundred and eleven feet, and its