Sec. 1] THE SHRINES OF KHADALIK 157
abstracting the woodwork. As a sign of the thoroughness with which this quarrying must have been done, I may mention that out of a total length of wall which, if the reconstruction of the
ground-plan is right, must have exceeded nine hundred feet, the only fragment surviving in sitzt consisted of a piece of the eastern cella wall, c, about four feet in length and standing less than a foot and a half in height.
The larger posts and beams must have been worked up on the spot for facility of transport elsewhere ; for again and again I came upon heaps of carpenter's chippings, these often
showing remains of delicate painted figures of Buddhas, etc., with which the surface of pilasters
and other exposed woodwork had evidently been covered. Those engaged in this vandal task seem to have done much of their work within the small rooms viii, ix, and x, laid bare immediately
to the north of the shrine. There great heaps of chippings, as if from a carpenter's shop,
were subsequently discovered. From the fact that these heaps lay only a foot or less above the original floor I conclude that the interval between the abandonment of the shrine and this
quarrying could not have been a very protracted one. The very selection of these rooms for the carpenter's workshop points to the same conclusion ; for evidently they were then still sufficiently preserved to offer some protection from wind or sun.7
In any case it is obvious that the quarrying of the timber and the splitting off of the painted panelling could not have been due to iconoclastic intentions ; for such the burning of the whole
structure would have offered itself as a far more convenient and thorough expedient. On the other
hand, it is equally certain that such extensive destruction was not the work of modern ` treasure-seekers or of Mullah Khwâja's party who would have been only too glad to preserve the painted
panels, etc., for sale. Numismatic evidence fixes the abandonment of the site at the close of the eighth century A.D.,8 so that I am inclined to date the quarrying of the abandoned shrine for its timber from the early Muhammadan period.
The similarity in character and style of the ` finds ' made in this and the other shrines which I excavated close by is so great that it appears from every point of view most convenient to leave my
brief survey of the chief classes of antiques represented among them for one place, and that after
completing my account of the various structures brought to light at the site. The ruin next explored was that of a Buddhist shrine marked by a low débris heap some forty yards to the south of the first
one (see Kha. ii, iii, Plate 6). Here the work proved very interesting and varied in its results,
though the structure was smaller in extent. It consisted of a rectangular cella, measuring twenty-five by twenty-four feet inside and enclosed by a passage eight feet wide on the north and five else-
where, with a large hall (iii) adjoining the south passage. The walls had been broken down in most parts nearly to ground-level ; yet having been built of sun-dried bricks, they could in spite of their relatively slight thickness be traced with greater certainty than those of Kha. i. Wooden posts had served for the strengthening of the walls ; but the wood had completely perished, leaving only empty matrices as traces of its use.
In the layer of rubbish and sand, three to five feet deep, which covered the cella and passages, small pieces of painted stucco were plentiful, showing that the walls must have all borne frescoes. But all the extant remains of walls had lost their stuccoed surface, except only where some portions of wall still rose to a height of about a foot and a half on either side of the