Hypaethral

[2] However, at the time Vitruvius wrote (c. 25 AD) the cella of this temple was unroofed, because the columns which had been provided to carry, at all events, part of the ceiling and roof had been taken away by Sulla in 80 BC.

The decastyle temple of Apollo Didymaeus near Miletus was, according to Strabo (c. 50 BC), unroofed, on account of the vastness of its cella, in which precious groves of laurel bushes were planted.

The larger the opening, the more conspicuous would be the notch in the roof which is so greatly objected to; in this respect, Jacques Hittorf would seem to be nearer the truth when, in his conjectural restoration of Temple R at Selinus, he shows an opaion about half the relative size shown in Cockerell's of that at Aegina, the coping on the side elevation being much less noticeable.

[2] The problem was apparently solved in another way at Bassae, where, in the excavations of the temple of Apollo by Cockerell and Baron Hailer von Hallerstein, three marble tiles were found with pierced openings in them about 18 inches by 10 in.

[2] In his 1849 Historical Inquiry into the True Principles of Beauty in Art: More Especially with Reference to Architecture, the architect James Fergusson put forward a conjectural restoration in which he adopted a clerestory above the superimposed columns inside the cella; in order to provide the light for these windows, he indicated two trenches in the roof, one on each side, and pointed out that the great Hall of Columns at Karnak was lighted in this way with clerestory windows; but in the first place the light in the latter was obtained over the flat roofs covering lower portions of the hail, and in the second place, as it rarely rains in Egyptian Thebes, there could be no difficulty about the drainage, while in Greece, with the torrential rains and snow, these trenches would be deluged with water, and with all the appliances of the present day it would be impossible to keep these clerestory windows watertight.