Sec. ii] THE CASTLE OF JAMSHID, GITI 81
kôh, took us first through a gap in the hill chain westwards to the hamlet and fort of Ghazab. These had been abandoned since the river finally changed its course about a generation ago. About a mile to the west of Ghazab there rises in isolation a bold ridge to a height of about 250 feet, bearing on its narrow top the remains of structures known as the Qal`a-i-Naushirwân. As the name shows, local tradition attributes to them considerable antiquity, but I failed to discover definite archaeological proof of this. At the highest point of the ridge stand two small but massive structures built with undressed slabs set in mud plaster. The walls of one still rise to 15 feet and show a large window arched with horizontal courses. Repairs are indicated in places by small stones set in `herring-bone' fashion, a method of masonry very common in modern buildings of Makrân. Lower down, a small fairly level plateau bears a row of small rooms showing the same kind of masonry as the structures above; nearby are the remains of a large enclosure supposed to have been used for stabling. What potsherds could be found were all plain and of coarse fabric, suggesting occupation of comparatively late date and short duration.
Turning thence to the south-west after a march of only 9 miles, we were obliged to halt for the night at the hamlet of Sand, to suit our old Sirdâr's convenience. Fortunately rain held off in the mountains, and next morning the deep-cut flood bed of the river, fully 50 yards wide and quite unfordable after rain, could be crossed without trouble. The height of the banks into which the river has cut its bed since the latest change of its course is here well over 40 feet. This affords some indication of the great depth of alluvium which the river has deposited over this plain; it may also help to explain why our rapid survey failed to trace any relics of prehistoric occupation in this deltaic area.
The ground beyond Dalgân, where the ford lay, is reached in certain years by flood water from the Kâj ii Kaur, a western tributary of the Bâhû, and hence
bears a less barren appearance than the eastern portion of the tract. But the
small village about 14 miles farther on which the map marks as the 'Min Bâzâr', or chief market-place of Dashtiâri, was found quite deserted and the water in its tank too brackish even for the taste of the local people. They are mostly
Jadgâls, speaking an Indian dialect apparently allied to Sindhi. They represent a tribe left behind by some unrecorded migration from the east of Makrân but now strongly affected in physical appearance by mixture with African slave blood.
We found drinkable water 2 miles farther on in a pool left behind by a flood of the Kâj û Kaur. Proceeding from the camp pitched there we succeeded next
day, after a long ride, in visiting two debris areas reported beyond the graziers' huts of Kumb. At the first, situated about 11 miles in a direct line to the north-west, a stretch of ground rising about 2 feet above the level flat and