^{page 281 note 1} Ch. 77, ii, 3 (Roll, 162 × 26 cm., of soiled yellowish-white paper, rather coarse and thick, being for the most part, no doubt, double). The text is in five pages: Nos. 1–3, c. 23–4 cm. wide; No. 4, c. 28 cm.; No. 5, c. 18 cm. The script is a rather formal copybook Dbu-can, not in itself specially clear in regard to the discrimination of certain characters, and now rather faint. The page was prepared by pencilled lines (16 per page) and vertical side lines.

The spaces between pages 3–4 and 4–5 are occupied each by two (superposed) rough miniatures of seated Buddhas, which in relation to the text are upside down. The space before p. 1 (c. 8 cm. wide), and those between pp. 1–2 and 2–3 (c. 7 cm. wide), show in a smaller and neater hand parts of lines of the same text, perhaps the original from which the MS. was copied. Upon separating the paper where it is now double perhaps further parts of the text may come to light, a similar process having in other instances proved fruitful.

On the reverse of the roll are to be seen: (a) thirty-one vertical columns of a Chinese text in large black characters; (b) two horizontal lines of Tibetan, c. 15 and 18 cm. respectively, which may be read and translated as follows:—

[1]ẖdi. yaṅ . de . hi . tshe. | Thu . pod . yaṅ . ẖbrug . gis . bris . pa. ẖo.

[2]gar . soṅ . gar . skyes . kyaṅ. lha . yul. du . skye . bar . smon . no ∥

[1]“This also at that time was written by Thu . pod (. yaṅ), a ẖbrug.

[2]Wherever going, wherever born, to be born in the realm of the gods is his prayer.”

That the ẖbrug people, to whom the writer, Thu . pod (. yaṅ), belonged, possessed the region associated with Śa-cu has been pointed out in JRAS. 1927, pp. 67–8Google Scholar.

^{page 282 note 1} chuṅ . śi . ẖgi seems to be Chinese.

^{page 283 note 1} The word seems to denote alternately the province and its head. Cf. the Turkī cur of the Orchon inscriptions (ed. Thomsen, V., Helsingfors, 1896, p. 155, n. 39a)Google Scholar.

^{page 283 note 2} The serial numbers are inserted with a view to the Index, where the Sanskrit and other meanings will enable the reader to follow the sense with little reference to the Chinese characters.

^{page 284 note 1} There is no syllable corresponding to this character in the MS.

^{page 284 note 2} Twenty-four characters omitted in MS. owing to homœoteleuton.

^{page 284 note 3} No corresponding character in printed text.

^{page 284 note 4} There is no syllable corresponding to this character in the MS.

^{page 285 note 1} No corresponding character in printed text.

^{page 286 note 1} No corresponding character in printed text.

^{page 286 note 2} No corresponding syllables in MS.

^{page 286 note 3} Four syllables omitted, presumably owing to homœoteleuton.

^{page 288 note 1} No corresponding character in printed text.

^{page 290 note 1} No corresponding syllable in MS.

^{page 291 note 1} In these two passages the MS. reproduces a different version from that of the printed text.

^{page 292 note 1} The MS. seems to require here.

^{page 293 note 1} No corresponding character in the printed text.

^{page 297 note 1} Note the final-d, which is found only in these two syllables and “bad”.