SECTION IV.—EXPLORATION OF N. XXVI AND OF THE SOUTH-EASTERN
GROUP OF RUINS
After completing the excavation of N. xxiv, I turned my attention again to the ruined dwelling, N. xii (for plan see Plate 8), to the north-west, which there had not been time on the occasion of my former visit to search with all needful care. The thorough clearing now effected served to reassure my conscience ; for apart from a quantity of beads, metal fragments, and similar small relics picked up on eroded ground near the ruin, the only finds made consisted of two small tablets and the double-bracket in wood, N. xii. i. 2. This is a fine piece of wood-carving, nearly five feet long and in very good preservation, as seen in Plate XVIII, with floral designs in Gandhâra style on the under-surface and a very effective saw-tooth ornament on the sides. The same inner room, i, where this double-bracket was found, still retained at its north entrance the ornamented wooden door-frame with slanting jambs seen in the photograph (Fig. 56). Its decorative patterns, as already noted in my former Report, closely recalled the Kingri work still common in the decorative wood-carving of the Indus region and derived from Gandhara art.'
To the east of N. xxiv three ruins could be traced in a row lying close together and all half-smothered by dunes. The nearest, N. xxv, proved that of a small and fairly well-preserved dwelling, but yielded no finds of any sort. The next, N. xxvi, was of larger size, and owing to the covering of sand, which rose in parts to a height of eight feet, its clearing cost nearly a day's hard work. The arrangement of its rooms, as seen in the plan (Plate 15), showed several interesting features. The construction of its walls in timber and plaster, with a core of horizontal reeds or else of diagonal matting, as well as that of its doors, roofing, etc., could be traced with ease, as the several photographs show. In the set of rooms to the north, i—iv, ix, we have probably the public apartments of the house, including those which would correspond to the modern ` Aiwân' and ` Mihmânkhâna' of this region. In illustration of these the plan of the ` Aiwan', with guest-rooms adjoining, as seen in a Beg's house at Bagh-jigda, near Yârkand, has been reproduced in Plate 12.2 On the south the family quarters may safely be recognized in rooms v—viii, communicating with each other, and approached from the rest of the house by a separate passage. In the north-west corner room, i, partially eroded, there were found eight Kharosthi tablets, including the rectangular covering-tablet, N. xxvi. i. 4 (Plates xx, xxiii), still retaining its seal impression in clay ; also the curious coarsely cut bone seal, N. xxvi. i. ooi (Plate XXIX).
The large room, iii, in the centre, was once, no doubt, used as the main public apartment. It is seen in the photographs, Figs. 63, 64, and was provided with a raised sitting platform on three sides as well as an open fire-place in the centre where the roof probably once showed a clerestory opening. With the passage, iv, on the east, this central apartment communicated by the door, and also by two side windows, A, B, clearly seen in Fig. 64 and once probably latticed, which