This page needs to be proofread.

370 History of Art in Antiquity. them betray aught Persian about them. But as it is quite im- possible at this date to distinguish between successive recon- structions, the historian must needs carry the whole structure to the account of the last people who helped to constitute and pre - serve it. The notion gained by Dieulafoy respecting the Susian defences, such as the Macedonian conquest left them, is summed up in the following words: — " The fortification works consisted iirstof a deep broad ditch full of water, communicating with the Shftttr and a double rampart. The external or first wall was massive and. built of crude bricks, in width 35 metres by 22 metres in height Against the inner lining of the wall— separated from the masonry by a trail of small pebbles or gravel ^leant a mass of earth beaten into a compact mass, 27 metres thick and 18 metres high. On this platform stood two groups of buildings which served at once as barracks and ivalk round, where, shielded by the earthworks, the defenders could circulate without danger, even when the first rampart ¥ras already in the grasp of the besiegers. The second rampart, 14 m. 70 c. broad, was constituted by two walls of unbaked brick, in thickness 3 m. 50 c. to 4 m. 60 c, between which damp earth was beaten down. Behind the second rampart ran a path the extent of which I was unable to determine. Broadly stated, the enceinte was not furnished with bastions ; its tracing in plan is in the shape of a saw with teeth at right angles. It is the indented line described by Philo.* At one point only of the external wall, in the middle of fragments of masonry, I detected a vaulted gallery lined with baked bricks. " Besides other information, the clearing of the walls of the for- tress has clearly shown that the roughnesses of the ground, no matter their apparent complication, corresponded with the salience of ancient fortifications. Due allowance, however, should be made for the direction of prevailing rains, and the very different damage they inflict upon the walls, according as these face this or that point of the compass. We knew at once from the mere aspect of the ground the situation the towers had occupied. Toweis had been distributed at the crenelated summits of the fortress, and its tracing had been so contrived that the towers of the second ram- part struck the middle of the curtains of the exterior wall. From

  • FHlLOb FtrHfittttiim TlttaHie^ viii 13 (tnuulated by RochM d'Aiglun in

Principts de ia/M^ttiHcn antipie. Pans, Ducher, 8vo, 1884). Digitized by Gopgle