946 IN THE DESERT DELTA OF SISTAN [Chap. XXX
type constructed with overlapping courses of bricks. The entrance, opening from the south, was vaulted in the same manner. Its width, originally 4 feet, had been reduced by later masonry to 2' 9". Outside it a kind of porch, z2 feet deep, has been cut into the thickness of the wall. The interior was encumbered to a height of about 4 feet with debris which moisture had compacted into a hard mass. At a level of about 3 feet above the present ground four loophole-like openings were traceable outside, at intervals of about II feet. Others are likely to have got blocked up by fallen masonry.
The ground near this rotunda was strewn with pottery fragments, mainly of the ` ribbed ' type, such as were found in such plenty at Ghagha-shahr. Besides plain ware of superior make, some small pieces of glazed undecorated pottery, in green, blue, and white, were also found, and also of coarse green glass and a bluish frit. These potsherds pointed clearly to the place having been occupied or visited during Sasanian times, and this indication received striking confirmation when a silver coin of Oueen Boran (A. D. 63o—I) (P1. CXX, No. zo) was picked up under my eyes at a distance of about 3 feet from the northern foot of the rotunda. It is in excellent preservation.8
It was to the NE. of Hauzdar, at a distance of about 71 miles, that I first came upon ground thickly covered on the surface with debris of painted pottery and similar relics of a prehistoric civilization. As the plateau bearing them lies only some two miles off the high road and is a conspicuous object, it is well known by the name of Shahr-i-sökhta, ` the burnt town '. Reserving a description of it for the next section, we may pass on to the ruins of Machi, which cover a considerable area extending from about 2 miles to the south-east of Hauzdâr.
The structural remains of Machi mark a site which, like Ramrûd in the south of the deserted delta, is known to have been occupied by a large village at the beginning of the nineteenth century.9 This statement is fully borne out by the appearance of most of the ruined buildings and also of the ground itself. When I first visited this in the evening, the slanting rays of the sun clearly showed up not merely the distributaries of the canal which once brought water to the fields here, but also the low earth embankments dividing them. In many places are still to be seen the low trunks of trees, mainly palms, which had been cut down on abandonment. I t all strongly reminded me of the picture presented by the site of ` Old Domoko ' on the desert edge of far-off Khotan to The perfectly level surface covered with a smooth crust of hard mud showed an abundance of potsherds, mostly of very modern appearance. Wind-erosion did not seem to have appreciably affected as yet either ground or buildings. Nevertheless drift-sand, that corrosive agent of the wind, had gathered in sheltered places. These indications of the slight progress of decay have suggested to me that some of the conspicuous ruins of the place may not go back farther than the time when Malik Bahram Khan's efforts had extended irrigation once more to this neighbourhood.
Among such structures may be noted the ` Chigini' shown in Fig. 498, a typical example of the large and skilfully constructed windmills (bad-i-âsiya) also to be seen elsewhere among later Muhammadan ruins of Sistan. The vaulting that carried the floor of the hall in which there were once two large wheels showed the respectable span of nearly 24 feet. Here, as in most other structures examined at Machi and Ramrtid, the vaults and arches were of the regular Western type. Of a large fortified mansion some 30o yards to the SW. of this ` Chigini ', the sketch-plan, Pl. 58, shows the interesting disposition of the apartments, and Fig. 499 the fine ` Aiwan' or loggia, covered with three barrel vaults resting on high pointed arches. The whole structure, impressive by its