Construction. 79 Pasargadae and Persepolis, at the sides of the pilasters that formed the angles of the buildings (Fig. 23) and the crowning of the bays (Fig. 22), deep grooves extend along the whole length of the block ; ekewhere, in the Propylaea of Xerxes (No. i in plan), the pillars offer saliences that play the part of what our masons call " waiting-stones " (piems daiUnti). The function of these grooves and protuberances is easily grasped : under the pressure exercised by the enormous mass, the penetrated the cavities between the resaults and found itself united in a close embrace to this kind of stone skeleton of which it was the flesh. The fact that we do not find similar pisi walls in place should cause no surprise ; for they were very thin compared with those of the Babylonian and Ninevite palace. It is hard to admit, with one of the explorers, that on either side of the pillars in the Propylaea of Xentes a wall 4 m. 50 c in thickness ran out to meet all the extremities on the main level,* since the greatest depth of the wall — to measure it from the stone frames of its hollows— occurs about the Hall of a Hundred Columns, where it was barely three metres, whilst elsewhere it did not quite reach two metres. Once the buildings were left to themselves, the rev^tement being no longer watched over would soon peel off, and the winter rains, penetrating the core, would turn it into mud and wash it away in the plain. The rubbish we find heaped up to man's height at certain points of the esplanade everywhere corresponds with the interior of the demolished halls ; that is to say, where the attics fell in and carried along with them capitals and broken shafts. Here the beaten earth of the levels, mixed up with fragments ol columns and calcined woodwork, has formed masses of great resisting power, upon which the spade makes but little impression. The recent excavations at Susa have confirmed the above conjectures, for tlie mighty ramparts that surrounded the palaces of the Ach^emenidcC were entirely built of crude brick. Now, the royal architecture at Susa and Persepolis was characterized by features common to both, be it in plan, elevation, disposition, taste, and even style ; whilst the like methods are to be traced evervwhere. Blocks of enamelled frit have been found at Susa ; their function, like the enamelled tiles of Assyria, could only be to act as facings to a solid mass of clay.' To have attempted anything
- DiEui.AFOY, Premitr Rapporti^ 59-6>>
" Ibid.^ pp. 62, 64. Digitized by Google
