Sec. iii] CONCLUDING STAY AT BUSHIRE 239
beyond the last shaft a shallow surface channel runs on to a patch of grass-covered ground, part of the terrace it was meant to irrigate. On the opposite (western) side of the road, which here runs along a small but distinct ridge rising up to 134 feet above sea-level, the continuation of the same ganât can be traced in the shape of a straight line of hollows marking shafts, or where the rock crust covering the underground channel between them has fallen in. This line, about 180 feet in length, ends at the head of a ravine ( Fig. 87) lined with eroded slabs of rock and draining into the nullah below.
Descending into the nullah to the south-west and crossing the drainage bed at its bottom, I found small terraces rising between deeply eroded small ravines and covered with low remains of walls built with roughly cut stones in regular courses but without mortar. These terraces extend for about 300 yards from north to south and are obviously `witnesses' marking the original ground level, which has been preserved there by the protection the ruined structures above it afforded. The fragments of turquoise-blue pottery with low relief decoration picked up among these decayed walls resembled those common at Sirâf and suggested occupation in early Muhammadan or possibly Sasanian times.9 Erosion around these terraces had obviously been facilitated by the absence within the wide nullah of the overlying cretaceous rock which forms the surface on the flanking plateaux. This had evidently become denuded at some earlier period, thus permitting of cultivation on some patches of ground within the depression.
In a broad bay on the western side of the nullah, about 300 yards to the north of the ruins just noted and not far from the series of ganâts first observed, there rises a curious tower-like structure ( Fig. 88 ) known -as Chàh-buland ( `the high well') . At first sight this was rather puzzling; but close examination soon showed that the lower portion of the structure, for a height of about 9 feet, consists of the rough stonework forming the lining of a circular well, with an outside diameter of about 17 feet. This had once been below the ground, but had been laid bare by erosion of the surrounding soil. That this erosion had proceeded even farther was shown by the small mound, about 4 feet high, on which the well foundation now rises. On top of the thick masonry that once lined the top of the well rises a circular wall about 6 feet high and measuring approximately 10 feet in diameter, which probably enclosed the well above ground. The interior of the well is filled with earth and rubble. From the top of the masonry which lined the well a hard earthen ramp leads down on the south side to the present ground level. This obviously represents the slope, probably once roughly paved or
9 These remains of old walls appear to be assumes to be probably the site of a Portuguese referred to by Lieutenant G. Pézard in the brief fort. There is no evidence to support this attribudescription he gives, loc. cit., p. 37, of what he tion.