National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Cuttings made into the lower portion of the mound at three different places
(I–III) disclosed only solid masonry of what may be assumed to have
formed part of the substructure for quarters higher up. It is from these that
the potsherds found along the decomposed mass of sun-dried brickwork below
are likely to have been washed down. Remains of glazed pottery and of glass
vessels turned up with particular frequency at the cutting III made at the
north-eastern end of the mound. There, under the closely adjoining rock face,
some refuse had evidently accumulated. The same observation applies to the
finds of ceramic ware made when clearing a mass of clay debris, IV, showing
signs of burning, which lay between the south-western end of the mound
and the ruined structure below it, V. Two cavities in the latter, indicating what
seemed to have been small rooms, were partially cleared without any finds
being made.

More useful results attended the clearing of the rock-cut trench, VIII, which,
as previously noted, extends along the slope for some 200 feet from the western
end of the wall closing the gap. Though completely filled with earth and
rock debris the straight line of the trench was easily recognized by me when
looking down from the crest of the ridge. The width of the trench at the top
was from 7 to 8 feet; its depth on the average 6 to 7 feet. Pieces of pottery,
both glazed and relief-decorated, together with plain ware and fragments of
glass, turned up in plenty at depths of 1 to 3 feet from the surface. The specimens
of ceramic ware recovered here closely with the pottery found else-
where at the site, whether on the surface or otherwise, that their types may all be
conveniently described here together.

Among these types the most characteristic and best represented is that of
the glazed pottery, which has its decoration, all in line, incised through a light
slip and hence shows dark under the transparent glaze. Numerous specimens
of this ware are reproduced in Pl. IV. As seen in the pieces of the large dish,
Jam. VIII. 169, and in fragments like VIII. 258, 276, 277; IV. 155, and others, the
designs comprise floral shapes, combined with arabesques and often with freely
treated Kufic inscriptions on the rims. The designs are almost invariably shown
on a hachured ground. The glaze varies from pale straw to yellow, brown or
shades of green, often with splashes showing a mixture of these colours or of
purple (manganese) added. Some glazed fragments, like surf. 23 with its
'pearl border', show a different treatment. Painted pottery, without glaze, is
rare.

Relief-decorated ware is well represented by the large jar, VIII. 257 (Pl. V),
cast in two portions from separate moulds and joined before firing. VIII. 202
(Pl. V), part of the handle of a large jar, is decorated with a wriggling snake