Sec. v] DISCOVERY OF DATED DOCUMENTS 269
Seeing the roughness of the material, and the apparent purport of the legend which curiously recalls what might be inscribed on a mere tally, this find alone would scarcely have suggested that wood was used as a regular writing material at the time when this site was deserted. Yet the fact of such use was clearly established by the discovery in the same place of two small wooden tablets bearing writing in cursive Brahmi characters. Both tablets (D. v. 3, 4) are oblong and rounded off at the left end. D. v. 3, as seen in Plate CV I, was provided with a small string-hole at the same end, and shows writing only on one side, arranged in four lines ; it measures 9 by 22 inches. D. v. 4, measuring 74 by 2 inches, is inscribed on both sides with four and three lines, respectively, parallel to the length of the tablet. From the transcription and notes with which Dr. Hoernle has favoured me for the writings, it results beyond all doubt that their language is the same early Eastern Iranian dialect which Dr. Hoernle has recognized in the paper documents with cursive Brahmi writing previously secured from Dandan-Uiliq. A number of words which occur in Dr. Hoernle's extracts from the latter reappear in these tablets, and so also the peculiar dating by day and month 12. Signatures, too, are found attached in the form of curious marks or monograms. Seeing that the script presents no apparent differences, it may be safely assumed that these written tablets cannot be far separated in time from the corresponding paper documents of the same site.
A third and somewhat larger tablet (D. v. ro), 131- inches long and 3 inches broad, which turned up close by, at once attracted my interest by showing in its form and lozenge-shaped handle the closest resemblance to the takhti, that traditional wooden board which in all native schools of Northern India takes the place of the slate 13. This tablet was found blank, but there are plentiful marks of scraping to show that it had once been used for writing on. I shall have occasion to refer again to the antiquarian interest of these finds ; for my subsequent discoveries at the Niya Site amply proved that we must look upon these few tablets, like the Indian ' Takhti' itself, as quasi-archaic survivals from the time when paper was unknown, and wood was the general writing material of Eastern Turkestan. But at the time of their discovery I little suspected what much more extensive finds of records on wood were awaiting me elsewhere.
The document in cursive Brahmi (D. v. 8), written on a sheet of thin paper, roâ by 72 in., turned up almost complete in the form of a crumpled small roll ; when opened out by me, with fingers half-benumbed through cold, it broke into two pieces. It was the first find of this class made in the immediate proximity of a record of ascertained date, and I may hence appropriately discuss here what observations I have to offer as regards the probable character of these documents and their antiquarian bearing. The total number of manuscript pieces in cursive Brahmi recovered by me at Dandan-Uiliq was six, to which must be added the small fragment brought to me by Turdi previous to my start for the site". Out of the six pieces three were found in ruined dwelling-places, viz. D. ni, D. v, D. viii, while the rest turned up in shrines D. iv and D. vi, which were closely adjoined by residences 16. All fragments were found either crumpled like bits of waste paper or folded into narrow rolls just like the Chinese
discerne la fin d'une évaluation en mesures de capacité :
- fi. â " un cheng et cinq ko",;--le ko est la dixième
partie du cheng.—Immédiatement après ces mots, on lit:
:i &F )I " sixième année, neuvième mois " ; l'indica-
tion du jour est illisible ; d'ailleurs, comme la période d'années n'est pas exprimée ici, la date ne peut . être connue.'— E. CHAVANNES.]
12 The words particularly noted by Dr. Hoernle with
references to their occurrence in previously published documents are : spâtâ, comp. Report on C.-A. ant., ii. p. 37 ; sidaki, comp. loc. cit., p. 36 ; hvartide, comp. p. 36 ; 14 mye hadai ' on the 14th day ', comp. p. 35.
13 For ' Takhtis ' of exactly similar shape found at the Niya Site, comp. Plate CI.
14 See above, p. 236.
16 The pieces are numbered D. m. 13 ; D. iv. 6 ; D. v. 8 ; D. VI. 4, 6 ; D. vIII. 2.