wise possess about an important factor in the ethnic composition of ancient Khotan. In connexion herewith it is interesting to note that, by the side of the numerous Indian names of persons which Dr. Hoernle has recognized in his Brahmi documents, a few others appear which look distinctly Persian 23. One more observation of historical bearing deserves to be mentioned. From the fact that these Brahmi records were found again and again in close proximity to Chinese ones, it may be inferred that the use of the local language for purposes of administration had not ceased during the period of Tang supremacy, in spite of Chinese garrisons and an extensive infiltration of Chinese culture. Seeing that, in addition to these two languages, Sanskrit also and the non-Sanskritic language, to which Dr. Hoernle for the present gives the provisional designation of ` Proto-Tibetan ', must, judging from manuscript finds, have counted students, perhaps even speakers, among those who frequented the little shrine (D. iv) and lived in the closely adjoining house (D. v), we may admire the polyglot faculties of at least the clerical and monastic section of the Khotan population in the eighth century.
Antiquarian The last dwellers in the ruin D. v had left behind some other curious relics of their daily
finds in D. v. life, besides the remnants of manuscripts and records on paper or wood that must have passed through their hands. In the same south-west corner room was found the fragmentary lower portion of a turned wooden bowl, D. v. 9, richly lacquered and painted. The elaborate floral decoration of the inside, as shown in Plate LXV, bears a Chinese look. The wood, too, seems different from any now available in Eastern Turkestan, and points to Chinese origin 24. An article of more humble use, but undoubtedly of local manufacture, is the fairly well-preserved shoe (D. v. i t) made of plaited hemp-string which turned up in the room eastwards next to the corner room. Though the shoe shows signs of much hard wear, Mr. Andrews has been able to ascertain all details of its make. Finally, I may mention a torn piece of fine cord silk, D. v. i 2 (Plate LX XVI), found in the same place, as a sample of contemporary textile work.
Original The variety of the finds made in D. v does not help us, I fear, to form a definite opinion
character of as to the original character of the building. Its vicinity to the temple cella (D. iv), and the ruin D. v.
discovery in it of several fragments of Pôthis in non-Sanskritic (Proto-Tibetan ?), one apparently a Buddhist canonical text, would naturally suggest that the ruined building had served as a convent. But the contents of the Chinese document D. v. 6, and the relatively large size of the ruin compared with that of other monastic dwellings at the site, might make us think rather of the residence of a local official or other person of position to which the cella was attached as a sort of private chapel. The numerous remains of ancient fruit-trees just visible above the sand some fifty yards further to the south-west may have belonged to an orchard connected with the ruined dwelling. Some 150 yards from the latter in the same direction I picked up a well-preserved Chinese ` cash ' with the legend of the K`ai-yüan period (713-741 n. D.).
Ruined About 25o yards to the north-east of D. v. I noticed under a scanty cover of sand, nowhere
structures more than 2 feet, the remains of three small structures (D. xiv). Owing to their exposed D. xiv.,D. xv.
condition they had evidently been thoroughly exploited long before. Turdi remembered to have found here a number of ` Khats ' in the form of crumpled scraps of paper or folded rolls, probably some of the documents which went to Dr. Hoernle's collection. Of the southernmost structure, of which nothing remained but the timber forming the foundation, it was just possible to make out that it had been a cella enclosed by the usual passage. The other two structures still showed scanty remains of walls ; but no finds of any kind rewarded their clearing. The same was the case with the small cella (D. xv) situated about 15o yards to the south of D. v. Here the walls,
23 See Report on C.A. ant., ii. p. 34. and probably of Chinese make, see N. xv. oor. a., Plates
24 For a piece of more ancient lacquer work, also a bowl LXV, LXX.