954 IN THE DESERT DELTA OF SISTAN [Chap. XXX
archaeologist of many centuries hence who should cone upon this ` culture stratum ' covered up by alluvium, if the Helmand were to shift its bed for good to this part of its present desert delta!
The other site of prehistoric occupation, and that an important one, lies far away, near the northern extremity of the ancient delta. To the NE. of Hauzdär the steeply scarped edge of the gravel-covered plateau of the ` Dasht ' curves round a large bay liable to inundation from the Hämûn. Where the south-western end of this ancient lake shore breaks up into spits and detached island-like Mesas, there extends along the top of one of the former the debris-covered area called Shahr-i-sôkhta, ` the burnt town '. Its direct distance from Hauzdär is about 7 miles, and as it lies quite close to the line which the caravan route to the south follows during the annual flood season of the Hämûn, the locality is well known to wayfarers. The site of Shahr-i-sôkhta stretches for about Boo yards from NE. to SW. with a maximum width of about half this distance. The plateau tongue occupied by the site rises from about 35 to 4o feet above the plain of bare clay to the south which marks a former extension of the Hämûn. Past the southern foot of the plateau tongue there sweeps a well-marked depression suggesting a bed which a branch of the Rûd-ibiyäbän might have followed at some earlier period. A narrower branch, about 150 to 200 yards across, branches off from this depression near the head of the plateau tongue and turning to the NW. divides it from the wide expanse of the Dasht. The whole of the debris-strewn area on the top is furrowed by small Nullahs, caused by the action of drainage. But the sides of these, as well as the outside slopes to the very foot of the small plateau, are everywhere strewn with a profusion of potsherds. There is thus reason to believe that the debris-strewn area on the top must have formerly been even greater than it is now. Even in its present extent it indicates an ancient settlement much larger than any of the prehistoric sites previously described, and one which, in spite of the total absence of structural remains, may well deserve the designation of ` town ' as implied by the local name of Shahr-i-sôkhta.
The whole of the pottery debris, enough to fill many railway trucks, consists of fragments, both plain and painted, of unglazed chalcolithic ware, as shown by the specimens described in the List. Fragments of stone cups and bowls, mainly of alabaster, all apparently lathe-turned, were also to be found,6a as well as stone beads. A broken lignite seal, S.S. 089, shows an incised kind of key-pattern. S.S. 055, 091 are specimens of small bronze fragments. In the course of my prolonged inspection I failed to find a single piece of glazed pottery of any kind, conclusive evidence, I think, that the site was not occupied again after its abandonment in chalcolithic times. The probing of the ground in half a dozen places was attended by some instructive observations. Underneath the surface layer of pottery debris and fine gravel there was always found a soft stratum of loose disintegrated clay. This soil was often of a reddish colour as if it had been exposed to fire, which accounts for the name Spahr-i-sôkhta, ` the burnt town '. Bone fragments seemed to be frequently mixed with this soil, and in places I noticed what appeared to be the smell of decayed animal or vegetable matter. At a depth varying from 1 a to 18 inches we came upon the very hard clay or sir, which seems to underlie the ` Dasht ' everywhere to the south of the Helmand.
The most probable explanation that occurred to me of the conditions here observed was that human occupation prolonged for a very long period had caused the top of this island-like plateau to be covered with thick ` culture strata ', consisting largely of decomposed mud walls or sun-dried brickwork, kitchen middens, and other refuse. The fine sand, i. e. the particles of alluvial or eroded soil that the north wind blows across Sistän, being caught on occupied ground loess-fashion, may have added its quota to these ingredients. Wind-erosion combined with deflation has been at work
6a Specimens of such fragments of stone jars and bowls Mr. Andrews' List are not accessible at present for inser-
are contained in the collection. Entries of them made in tion.