places 20 feet or more, owing to the scouring effect of the sand driven by the north wind, which blows over Sistàn with more or less violence during four months of spring and summer.'
The difference in the height of these debris-strewn terraces, true ` witnesses ' of a period preceding all historical records, is easily accounted for by the varying conditions which the neighbouring area has probably undergone since their original occupation. Wherever water from the Rûd-i-biyàban made its way from time to time, whether through inundation or by irrigation,
vegetation for the time being afforded protection from the erosive force of the wind, and the lowering of the ground level was in consequence retarded. In the same way the height retained by those
` witnesses ' of ancient occupation must also in a measure depend on the thickness of the culture
strata which had accumulated by the time when that occupation ceased and wind-erosion began its work. For not until the layer of hard debris, winnowed out as it were by the wind from the
embedding masses of loose earth and refuse, has attained a certain thickness and consistency can
its protection become fully effective. It is obvious that this stage would be reached sooner at points where human occupation had been dense and continuous for a long time than at others where
it may have been only intermittent and dependent on the seasonal movements of small semi-nomadic
communities. Nor ought it be overlooked that reoccupation of such sites at much later periods would necessarily affect their level as well as the character of the relics to be found there. It is
equally easy to realize that relics of the same early epoch may elsewhere also be brought by wind-erosion to lie exposed on the surface, side by side with objects of much later times, just as they are to be found at ` Tati ' sites of the Taklamakàn or in the wind-eroded wastes of the Lop Desert.
Apart from the last two observations, it can safely be asserted that the vast majority of the abundant remains preserved for us on those Mesas of the southern Sistàn delta are remarkably
uniform in character, and date from a single and evidently prolonged epoch of civilization. Having
regard to the occurrence along with them, though rarely, of small bronze objects, and to the remarkably close agreement between the decorated pottery from these sites and certain ceramic ware
found in other regions, as widely distant as Thessaly and Western China, this early Sistàn culture
may justly be described as ` chalcolithic '. Before proceeding to a succinct analysis of characteristic types represented among the specimens of its relics, as described in the List below and
illustrated in Pl. CXII—CXIV, I may record brief observations regarding the sites from which they were collected and the circumstances attending their discovery. Our notice of these sites may conveniently start from the southern area to which the account of later remains given in the preceding section has brought us. It was there that I was first able clearly to recognize the peculiar conditions explaining the survival of these relics.
Mound near A little over half a mile to the N. of Kalat-i-gird a mound, about 8o yards long and 66 yards
Kalät i across, rises to a height of roughly 24 feet above the bare level ground. Its longitudinal axis lies
in the direction of the prevailing wind, and it slopes down gently towards the SSE. An identical observation was made at the other mounds visited, and its explanation is the same as that which applies to the bearing of the ` tail ' invariably shown by the Mesas of the Lop basin.la The whole of the flat top of the mound and most of its slopes are thickly covered with potsherds of the chalcolithic type, plain, incised, or painted. A complete search for them might take days or weeks, and their removal en masse would fill many cart-loads. Among the numerous specimens of this
stone ' that he notes as covering ` the rounded mounds '. The terraces I was able to examine looked, indeed, dark from a distance. But I did not find any pottery which can be called black.
la See above, e. g., i. pp. 226, 263, 310.