wedge tablet and a wooden fire-stick (Plate xxviii). This was found tied by a loop of goat's-hair rope to a short curved piece of wood resembling a primitive pulley (see N. xxix. ii. ooi. b in Plate XXVIII), of which other specimens are shown by the same plate ; the original use is doubtful. From the long passage, iv, came; a perfectly preserved rectangular under-tablet of respectable size, twelve by five inches, showing eleven lines of clear Kharosthi text with a date which I read as the r ith year of King .ilughavasmana devaj5uira. The large room, v, yielded no relics. Outside it, yet within the fenced enclosure of a court or garden, lay the remains of long, twisted vine branches, once trailing, no doubt, close to the ground, after the fashion of Turkestan gardens. Near a cattle-shed to the north-east a mulberry-tree still raised its gaunt fissured trunk to a height of about fourteen feet.
Among the miscellaneous objects found within, or quite close to, this ruined residence there may be mentioned the fragment of richly coloured ingrain material, N. xxix. oo i. a (Plate xLix) ; a builder's float in wood for plastering wall faces, N. xxix. 002 (Plate XIx) ; and a well-preserved bronze ring, N. xxix. oo6 (Plate XXIX). Two ornaments, the urn-shaped pendant, N. xxix. 005, and the fragment of a large cylindrical bead, N. xxix. 007, both shown by Plate xxix, are of special interest. The material is a blue paste which I was first inclined to take for celadon. Both in shape and material, as Mr. Woolley points out, they closely resemble Western products such as are found in Roman Egypt.
The ruins, which lay to the north, proved, when searched on October 28, the remains of small and poorly built dwellings, as befitted the homesteads of what was evidently the easternmost fringe of the ancient settlement. N. xxx, reached after going over a mile due north from N. xxviii, consisted of three small rooms, the lines of their walls and sitting platforms being just traceable under one or two feet of sand. Two wedge-shaped under-tablets were found here, one completely bleached. N. xxxi was another small dwelling, about a furlong to the north-west. Above ground eroded to a depth of over ten feet there remained here only scanty traces of a room, sixteen by fourteen feet, adjoined by a fenced court to the west. About a quarter of a mile to the west of N. xxx there emerged the remnants of a small structure, probably a cattle-shed, from the side of a tamarisk-cone, rising about eighteen feet above the level marked by the ruin. The ground immediately below the latter was eroded to a depth of fifteen feet. Going for about one-third of a mile to the south-west I found the top of a small eroded ridge occupied by a small ruin, now completely decayed, N. xxxii. The walls of two small rooms were barely traceable, the height of the protecting sand being less than a foot. Much fallen timber strewed the east slope. Finally, a short distance to the south I could trace the remains of a cattle-shed and arbour, N. xxxiii, on two small witnesses', as seen in the photograph, Fig. 67. Close by to the north-east there emerged from the side of a tamarisk-cone the dead trunks of three or four poplars planted in a row. This completed the survey of what structural remains could be found by us northward.
Returning thence to where my last camp had stood, I proceeded southwards in search of the remaining ruins. A strong north-east wind filled the air with a grey dust haze. But even without this the surroundings were growing more and more sombre, in spite of the more frequent appearance of still living scrub. The ruins had to be searched for amidst closely packed sand-cones which raised their heads with tangled masses of tamarisk to heights varying from twenty to forty feet. It was in a singularly gloomy and confined spot of this maze that the ruin N. xxxv proved to be situated, over half a mile to the south of N. xxix. It was that of a relatively large dwelling (see plan, Plate r 7) ; but the exposed western portions had suffered badly by erosion, always particularly effective at the foot of sand-cones, and the rest was made practically inaccessible by the sliding of the sand from the slope above. From the central room, ii, which was still traceable, a complete