Sec. iii] SEARCH AMONG THE RUINS OF TOYUK 613
reference to them is here needed. Two miles farther down, the gorge opens out upon the narrow strip of cultivation known as Lukchun-aghiz, past which the road towards Lukchun descends to the village of Sirkip. In the midst of its houses and gardens rises the imposing Buddhist pile known only by the designation of Sirkip-tura, `the tower of Sirkip '. It has been briefly described already by Dr. Klementz ;' and has, no doubt, been visited also by succeeding archaeological expeditions. But as I can trace no exact account of its structural details, the following record may be useful.
The ` tura ' of Sirkip, like the two corresponding ruins of Taizan at Astâ.na and of At-hayasi (Grünwedel's temple Y) at Idikut-shahri,8 is a terraced shrine planned after the general model of the famous Buddhist sanctuary at Bôdh-Gaya. In all three the successively receding stories of a structure square in its ground-plan were decorated with niches each containing the stucco image of a Buddha seated in meditation. At Sirkip, however, the number of niches and images remains the same on all the stories, instead of gradually diminishing as at At-hayasi, and only their size is reduced. As the sketch-plan in Pl. 28 shows, the pile forms a solid square of 48 feet at the base which is built of stamped clay to a height of Ioi feet. The sides of this base were plain and have completely lost whatever stucco ornamentation they may once have possessed.
Slightly receding stories rise above the base, and each of these, on each of its four sides, originally displayed seven flat-arched niches containing Buddha images. Of these stories or terraces six are still extant, as seen in Fig. 316, where the eastern side, which has suffered less damage than the rest, is represented. From the broken appearance of the top, it is probable that there was originally one more story with niches on it, and perhaps some finishing superstructure besides. The lowest of these stories has completely lost both niches and images. But that the latter originally existed is proved by the holes still to be seen in the masonry which once held beams to which the wooden framework of the stucco images was fastened. Timber was inserted in other places also, to reinforce the masonry and to support the plaster of the niches. Each successive story receded only by two feet or less as compared with that immediately below it. It was consequently impossible to gain access with safety to the outsides of the stories above the lowest, and it is doubtless due to this that the Buddha images in many niches escaped wilful destruction. The total height of the extant structure could not be exactly measured, as the top is no longer accessible ; but it certainly exceeds 5o feet even in its present broken condition. Its south-western corner has suffered much damage, and the bricks quarried from the debris at this point have found their way into many of the houses in the village.
The same destruction has overtaken the flight of stairs that probably once led up to the vaulted passage which, as seen in the plan (P1. 28), traversed the whole pile from north to south and gave access to spiral stairs leading to the top of the structure. This transverse passage had a uniform width of 4 feet and a height corresponding to that of the second and third stories combined. It was examined by Afrâz-gul, the condition of my leg making it impossible to clamber up to it. Its openings both on the northern and southern faces lay in places which otherwise would have been occupied by the third niche from the east. But owing to the damage which these faces have suffered just near the openings it is impossible to say how they were fitted into the general decorative scheme. The more extensive destruction near the opening on the northern face (Fig. 317) makes it appear probable that the proper approach to the transverse passage lay on that side, and that the passage was extended to the southern face merely for the purpose of securing adequate light and air for the spiral stairs. These are 3i feet wide where they strike off from the eastern side of the passage, and winding round a circular core of masonry about 12 feet in diameter gradually grow narrower. Higher up they break off, owing to a fissure which has occurred in the interior
7 Klementz, Expedition nach Turfan, pp. 31 sq. 8 Grünwedel, Idikutsckari, pp. 49 sqq., 173; Figs. 43-6; Ser. iii. Fig. 272.
II 4 K