^{page 599 note 1} In a three-line bilingual Khotanese-Chinese, of which Professor Haloun, kindly read for me the difficult cursive Chinese, Domoko C 1, thaunaGoogle Scholar corresponds to Chinese (for ) “embroidered silk cloth”. Also thauna = Skt. paṭṭa.

^{page 599 note 2} hvāṣṭa renders P 3513. 53 r 4 jyeṣṭhaku; Konow, Suvarṇabhāsa 63 a 6 hvaṃdānu hvāṣṭu = Skt. manujeśvara, ibid. 35 a 4 hvāṣṭye rrīñe = Skt. agramahiṣī.

^{page 599 note 3} The second akṣara is broken. Suhadi, Svarrnadi and āskvīra are personal names.

^{page 599 note 4} K refers to Karlgren, B., Analytic Dictionary of Chinese and Sino-Japanese.Google Scholar

^{page 599 note 5} For the sounds K 898 ṣou < “head, a chief”, would suit equally well.

^{page 600 note 1} In A Volume of Eastern and Indian Studies presented to Professor F. W. Thomas (1939), 146–7.Google Scholar

^{page 601 note 1} tcaisyāṃ P 5538 a 9 mistä tcaisyāṃ, P 2786.221 tcaisyau, may, according to a suggestion of Professor Moule, represent Chinese K 1023, 786 tsai-siang < tsâi– “prime minister (Giles no. 11,490)”. The name paḍa-ttana is probably Tibetan dpal-brlan “having secure fortune”.

^{page 601 note 2} Hoernle, , Manuscript Remains of Buddhist Literature found in Chinese Turkestan, plate xviiGoogle Scholar.

^{page 601 note 3} Rendering uncertain, aiśta, also written éśta, aủ'śte, aiśca, in the Siddhasāra, renders Skt. punarnavā (pw. Boerhavia procumbens). kūsa is frequent as a mureeas of grains, ṣaṃga = 8 śiṃga (śiṃga = Chinese K 873 ṣing < ś “litre”). Ch ii 003. 21 v 4 śau ṣagą “one ṣaṃga” renders Skt. prasthaṃ ca caturguṇa “4 prasthas”, Ch ii 003, 33 v 2 haṣṭa śiṃga “eight śiṃga” = Skt. caturguṇa “4 (prasthas)”, see also BSOS., VIII, 920.

^{page 602 note 1} Konow has edited these verses in A Volume of Eastern and Indian Studies, cited above. Note that in line 1 gyastä is to be read: the gya as often is written with an ornamental flourish. It can be seen for example in the Staël-Holstein roll 32, 40. In line 5 haspisyari is 2 plur. imperative. For Kāṣ– “to protect”, cf. the derivative noun of agent kā'ka: Ch c. 001.954 jsīñi kā'ka vījya = Tib. ćhe skyoṇ-baḥi rig-sṇags “life-preserving spell”; P 2929.5 hvana kṣīrä hīyi “Of the Khotan land the protectors, local spirits (naivāsika, see BSOS., viii, 902, note 2), guardian (āysdagaraa–) deities, wardens (paripālaka)”.

^{page 602 note 2} The name is written with rṇ and rn. For the adjectival suffix -āṃga < ānaka, cf. also Ch ii 002, 100 v 1 hīṃdvāṃgą, Ch 1.0021a, b 19 hīdvāṃga “Indian”. The Skt. nalada appears as a Khotanese loan-word Kha 0013 d, 1.2 naladä bhāgä IV II kuṣṭi bhāgä IV II “nard four parts, costus (kuṣṭha) four parts” in a medical prescription; ąlso Ch ii 003, 1 r 5 nalądha. Laufer, , Sino-Iranica, 455 fGoogle Scholar., discusses the provenance of nard:it was known in Sogdiana and India.

^{page 602 note 3} Tibetan Literary Texts and Documents concerning Chinese Turkestan, vol. I (1935), 151Google Scholar ff., where earlier references are to be found.

^{page 603 note 1} The Bhadracaryā-deśanā in the Gaṇḍa-vyūha, ed. Suzuki, and Idzumi, , 1934–6, vol. iv, 544, verse 18.Google Scholar

^{page 603 note 2} Ṣacū-pavāna– is adjective to ṣacū-pata, plur. to *ṣacū-pa (cf. Or 8212 (162) *153 kamacū-pa “inhabitant of Kam-ṭṣou”), where -pa is probably Tibetan -pa, as in lha-sa-pa “inhabitant of Lhasa”. Between a final -a and the Khotan. plural -a a t or v is inserted pava, pata. This -pata should probably also be found in Ch 00269, 78 dūṃva, u cahä:spata u sūlya, BSOS., VIII, 883 (and correct BSOS., VIII, 918).

^{page 603 note 3} See Thomas, F. W., Tibetan Literary Texts and Documents, vol. i, 132, 133; also JRAS., 1939, 89.Google Scholar

^{page 603 note 4} An edition is now in the press in Dr. Henning's book Sogdica, of which I am in his absence reading the proofs.

^{page 604 note 1} In the royal name Väśa’ vāhaṃ, see Konow, , JRAS., 1914, 342 fGoogle Scholar.

^{page 604 note 2} The same can be established for Khotanese transcriptions of Turkish, of which many examples can be seen in JRAS., 1939, 85 ff. For Tibetan only two sounds were needed for a and o, since long ā was not distinct in Tibetan words from a.

^{page 604 note 3} Edited BSOS., IX, 521 ff.

^{page 604 note 4} The interpretation of ttāha'ta by *tofat offered BSOS., IX, 542, due to an attempt to find a labial in the word, now seems to me doubtful. We know h < f in Old Iranian words (saha– “hoof” = Av. safa–) and to represent initial Middle Chinese –, – (Mod. Chin. f–) we find hv–, ZDMG., 91 (1937), 34–5Google Scholar, but -h– is not assured for foreign medial -f–. For h = γ χ there is full evidence, see ZDMG., 92 (1938), 590Google Scholar, and the words quoted JRAS., 1939, 85 ff.