42 2 A History of Art in Ancient Egyi'T. Under such conditions we need feel no surprise at finding part of the temple subterranean. In backing" his work against the mountains in this fashion the architect must haYe been partly impelled by a desire to make use of the facilities which it afforded. The mausoleum of Hatasu, unlike the other funerary chapels at Thebes, is, then, a triple hemispeos. At a point immediately opposite to the door in the external pylon, but at the other extremity of the building, a chamber about sixty-five feet deep was excavated in the rock. This must have acted the part of a sanctuary. Right and left of it, and at a shorter distance from the entrance, there are two more groups of rock-cut apartments. The whole arrangement may " be compared to the system of three apsidal chapels which is so common at the east end of European cathedrals. In approaching this temple from the river bank, a dromos of sphinxes had to be traversed of which very scanty traces are now to be found, but in the time of the Institut d' Egypt e there were still two hundred of them to be distinguished, a few of the last being shown in the restoration figured upon the opposite page (Fig. 251). At the end of the dromos, upon the spot where a few traces of the bounding walls still remain, we have placed a pylon with a couple of obelisks in front of it. We have done so not only because nearly all the important temples had such a preface, but also because Sir Gardner Wilkinson says that he saw the foundations of two obelisks and of a doorway. After passing the pylon, a first courtyard was entered, which communicated with a second by an inclined plane stretching almost across its width. ^ Here the arrangements which con- stituted the real originality of Dayr-el-Bahari began. The whole interior of the temple, betw^een the pylon and the commencement of the speos, consisted of four courtyards, rising in terraces one above another like the steps of a gigantic staircase. The walls upon which these inclined planes and terraces were constructed are still to be traced in places. In order to furnish the vast courts, we have supposed them to contain seated statues at regular intervals alonsf the inner faces of their walls ; in such matters of ^ This wide inclined plane agrees better, as it seems to us, with the indications in M. Brune's plan of the actual remains at Dayr-el-Bahari, than the narrow flight of steps given in his restoration ; the effect, too, is better, more ample and majestic.
