6io AT RUINED SITES OF TURFAN [Chap. XVIII
Oasis of opportunity for useful archaeological work, I may briefly mention such old remains as I was able
Lamjin. to visit on this short tour. The bed of the stream which descends from Örtang-aghzi and lower
down irrigates the Lukchun oasis is wide but usually for the most part dry ; where this approaches the outer hill range before breaking through it, there lies along the northern foot of those hills and on either side of the stream the small oasis of Lamjin. It is reckoned at about four hundred households and forms part of the Pichan district. From the house of its headman or ` Shangye ', situated near where the high road to Pichan crosses the stream bed, I visited on November i8th a group of old remains situated to the south-east along a tributary stream. This is fed by springs rising near the neighbouring oasis of Khand6, and joins the Lamjin stream some three miles to the south of the high road. The remains were said to have remained unexplored except for some diggings by Ilyâs, a native dealer in antiques ; some graves had also been searched on behalf of a Pichan ` Amban '.
Mazar of After proceeding to the south-east for about two miles to the edge of the main cultivation and
z_,zz_then crossing a bare gravel plateau, we dropped down to a narrow strip of fields tilled by the dozen . households of Yutôgh and situated along the deep-cut bed of the stream coming from Khand6.2
Close above its left bank a gravel plateau rises very steeply to a height of about eighty feet, bearing on its top the Mazâr, known as Yetti-kiz-khôjam, ` the Seven Holy Maids ', and visited as a place of pilgrimage. Near the shrine stand seven domed tombs of small size, a mosque and some roughly built shelters for pilgrims, and along the edge of the cliff extends a Muhammadan graveyard. I was not able to obtain any clear account of the legend of the ` Seven Holy Maids ' ; but pious eyes recognize them in seven boldly eroded rock pinnacles standing on a crest of the rugged hillside to the south, about six hundred feet above the stream. This makes it clear that the place owed its sanctity to worship as a kind of svayambhû-tirlha, worship obviously going back to pre-Muhammadan times.
Pre-Muham- Evidence of this is furnished by crumbling walls and vaults, manifestly old, which line the
madan side of the cliff eastwards immediately below the shrine and tombs, and by three small cemeteries,
cemetery. undoubtedly pre-Muhammadan. As the sketch-plan, Pl. 26, shows, they are found on a continuation of the terrace beyond a small ravine to the east. The low mounds of rubble forming rectangular enclosures around these cemeteries, and the roughly circular heaps of stones and gravel that mark the position of the tombs cut into the ground below, are of exactly the same type as those found at the extensive ancient burial-grounds near Astâna which I shall have occasion fully to describe below.3 In the largest of the Yutôgh cemeteries we found one among six tombs opened, and guided by the indications which its construction afforded were soon able to trace the narrow trench which formed the approach to the tomb nearest to it.3a It proved to be two feet wide, and at a depth of z 2 feet led to a small tunnel of similar width, closed by brickwork at its inner end. The condition of the closing wall showed that the tomb chamber had not previously been entered ; nevertheless, the interior, which measured 8 feet by 9 when cleared, yielded no finds whatever. The bottom of the chamber was covered with damp earth which must have fallen from its ceiling, and the penetration of moisture explained the complete decay of what the tomb had once contained. But at the outer end of the approach trench to the nearest tomb on the south we came upon a roughly cut wooden stick, about two feet long, showing on its flattened side a line of Chinese characters, evidently a funeral record. [Dr. L. Giles kindly informs me that the inscription mentions ` the grave site of Chao Chin-hsiang ' and a date corresponding to A.D. 67 r .]