laid-out fields which a small channel, provisionally dug, had just irrigated. It was said to be surplus water from the Domoko-yâr, available after satisfying the needs of Malak-âlagan, the new colony first visited in 1901. Thus I had ocular proof that the fresh extension of Domoko cultivation northward, of which I have tried above to explain the reason and special interest,4 was still in steady progress.
After securing at Malak-âlagan, the first permanently inhabited place reached since our start from Shahyâr, an adequate number of labourers, I proceeded on March 2 to the site of Farhâd-Beg-
yailaki, which my guides had failed to locate before when coming from the east. Its remains
proved to be scattered over a relatively extensive area, stretching, as the site-plan, Plate 56, shows, for about 4 miles from south-east to north-west and with its nearest point about 6 miles north of
the central homesteads of Malak-âlagan. This area, as seen from Map No. 31. A. 4, lies between two
well-marked old flood beds which the stream now descending in the Domoko-yâr was said to have reached in years of heavy floods of ak-su from the mountains, before the opening of the new colony
below Kara-kir caused these to be diverted north-eastwards for the cultivation of Achma.5 A tiny
watercourse, carrying some 6 cubic feet per second, was then taking the surplus of the water from the springs of the Domoko-yar, not needed by Malak-alagan cultivation, down towards the
southern end of the site. Before the end of my stay it had managed to make its way between the
tamarisk-covered sand hillocks even close to ruin 1. But this was the season when the flow from springs (kara-su) is amplest, and a little later on the stream was said to dry up at about 2 miles
from Malak-âlagan. The temporary overflow suffices to produce ample growth of reeds and scrub along' its course. This extends also for a mile or so beyond a conspicuous high tamarisk-cone, known from a rough signal-post on its top as Farkad-Begning ilesi, where my Camp 343 was pitched. The grazing around is called Farhâd-Beg-yailaki after a local Bag who in Yâgttb Bag's time used it also for getting saltpetre from the neighbouring ruins.
A first rapid inspection of the ruins sufficed to show me that they had all suffered badly through this and earlier exploitation by villagers and wood-cutters such as the vicinity of ` Old Domoko '
and other inhabited ground rendered easy. The portions of the structures that had originally been
above ground were nowhere apparently covered by more than 2 or 3 feet of sand. Yet, on the other hand, I noticed with satisfaction that moisture could not have caused so much damage to the
remains as might have been feared otherwise from the present close approach of water. Almost all the ruined structures traceable were found to occupy erosion terraces, rising up to about 15 feet above the adjoining ground-level. This particular feature of the site seemed to suggest from the first that these ruins of what manifestly were ancient dwellings and shrines, resembling those of Dandân-oilik and Khâdalik, might yet possibly be of somewhat earlier origin. I shall have occasion below to mention other indications pointing in the same direction.
The construction of the walls, which was easily recognized as being mainly in timber and plastered wattle or else in sun-dried bricks, showed no clearly datable features. The wattle was
made either of horizontal reed-bundles or else of thin tamarisk branches placed vertically and
secured to the timber framework by cross-pieces. Besides these methods, construction with mere plastered rush walls and masonry of hard and more or less flat clay clumps was also represented.
That the number of ruined structures was much smaller than Badruddin Khan's report had suggested was a disappointing observation, made subsequently elsewhere also. Its explanation lay in the fact that my worthy friend, who never visited any old site himself, had put a wrong interpretation on the statements of the ` treasure-seekers ' he sent out to search. They meant rooms, not houses, when reporting to him how many of they had traced in a particular locality.
' See above, pp. 203 sqq. 5 Cf. Ancienl Kholan, i. p. 468.