^{page 513 note 1} An early Khotan king, Vijaya-saṅgrāma, who had slain a man-killing lion, obtained by that feat the soubriquet “lion” (Siṃha): see Tibetan Literary Texts and Documents, i, p. 120.

^{page 513 note 2} See Professor Bailey in New Indian Antiquary, Extra Series, i, p. 1, Vījatta Sagrrauma = Vijita-Saṅgrāma, in a Saka-Khotanī document.

^{page 513 note 3} The Editors' reading, adopted by Professor Konow, Indian Culture, ii (1935–6), p. 193, in place of bis original (Act. Or., ii (1924), pp. 126–7) stanajha ddhe[vabatri].

^{page 514 note 1} In Saka-Khotanī dai means “fire”; the medial h in dhahi points to Sanskrit, as perhaps does also, on Professor Konow's view, the initial dh.

^{page 514 note 2} See Index.

^{page 515 note 1} Indian Culture, ii, p. 196.Google Scholar

^{page 515 note 2} Ada Orientalia, xiv, p. 240.Google Scholar

^{page 515 note 3} I may here take a long sought opportunity of correcting a reading in the document; in text, L 20, read ne(not ste)-źo-myed-par, and in the translation “without illness” (ne-źo), in place of “being penniless”.

^{page 515 note 4} On this king see infra, p. 517.

^{page 516 note 1} (hun);(WSn) (cf. man, Karlgren, no. 595); (Mo).

^{page 516 note 2} Karlgren, no. (in WḊn-ch'Ḋ = Mun-ṛaṅ) 609; Stanislas Julien, Méthode, p. 225; S. Lévi, J.A., XI, v, p. 133; Rosenberg, Introd. to the Study of Buddhism, pp. 226, 281, 368, 382, 411.

^{page 516 note 3} “Tib. Literary Texts and Documents,”JRAS., 1930, pp. 65 sqq.

^{page 517 note 1} See Konow, Ein neuer Saka-dialect, p. 775 (6).

^{page 517 note 2} See Konow, Ostasiatische Zeitschrift, viii, pp. 221–4Google Scholar; Norsk Tidsskrift for Sprogvidenskap, 1938, p. 24.

^{page 519 note 1} Asia Maior, ii, p. 267Google Scholar; Tib. Lit. Texts and Documents, i, pp. 123, 126 n. 1, 132 n. 2. To Professor Bailey we are indebted for some further Viṛa names, ˚Śūra, ˚Dharma, ˚Kirti, mentioned in documents: see BSOS.,VIII, p. 936.Google Scholar

^{page 520 note 1} Abel-Rémusat, Ville de Khotan, p. 30.

^{page 520 note 2} Tibetan Literary Texts and Documents, i, pp. 25, 104.

^{page 520 note 3} By Professor Konow in JBAS., 1914, pp. 344–5. The case of Hsiu-mo-pa (xieu-mâk-pªk, Karlgren, nos. 159, 605, 755), as Chinese equivalent of Sambhava (ibid., p. 345), is different, since Sambhava is regarded as a translation of the native name, and in any case the phonetic resemblance is remote.

^{page 520 note 4} See Chavannes, in T'oung-pao, II. viii (1907), pp. 171, 201–2.Google Scholar

^{page 521 note 1} As regards Biṛa there may be further trouble lurking in the fact that the Khotan kings were said to be of the race of Vaiṛramana, Chin. P'i-sha-men, Saka-Khotanī Biṛa-ramani, Beṛa-ramana, who, in fact, is according to the legend the author of the line: see Beal, Buddhist Records, ii, p. 309, and Tib. LU. Texts and Documents, i, pp. 98–9, 307–8.

^{page 522 note 1} See Böhtlingk and Both's Dictionary.

^{page 522 note 2} See Serindia, p. 332.

^{page 522 note 3} See Stein, Innermost Asia, p. 575.

^{page 522 note 4} See ibid.

^{page 522 note 5} The order of the successioṅ has been established by Rapson (pp. 323–5). Professor Konow's prior suggestions (Act. Or., ii (1924), pp. 135–140 (cf. also vi, pp. 93–6)) making Vaṣmana prior to Aṃgoka and identifying him with the Khotan king Fang-ts'ien (c. A.D. 130), would now probably no longer be maintained.

^{page 523 note 1} Kharoṣṭhī Inscriptions, no. 676 and p. 327. The year 38 of Aṃgoka being later than A.D. 124, all years of Aṃgoka's reign, and the whole compact period of at least 86 (or 96) years, must be later than A.D. 119, when Yu-huan was reigning.

^{page 523 note 2} See the Wei-lüo (Chavannes' trans.) in T'oung-pao, 1905, pp. 535–7.

^{page 523 note 3} See Beal, Buddhist Records, ii, p. 324. I do not find cogency in the argument of Professor Lüders (Acta Orientalia, xviii, pp. 35–7) to the effect that at the date of documents nos. 14 and 367 Niya was already in the possession of the Khotanīs. À priori it is highly unlikely that Niya, situated between Caḍ'ota and Parvata (Sorghak) on the same Niya river, should have belonged to a different state. The indications in the two documents, which dictate arrangements to be made in Niya, seem to accord better with the supposition that Niya was the last place under Shan-shan authority on the particular route to Khotan: and, in fact, in nos. 189 and 518 Shanshan orders are given for measures to be taken in Niya. Most of the references seem to belong to the time of Hahiri. On the other hand, the Wei-lüo, written probably not later than A.D. 265 (Chavannes in T'oung-pao, II. vi (1905), pp. 519–520), assigns to Khotan the old state of Jung-lu, to which Parvata will have belonged. If we suppose Jung-lu, and perhaps also Niya, to have been conquered by the Khotanīs during the reign of Mahiri, this would harmonize with the facts (1) that a war with Khotan and other relations with that state (including the matter of a boundary, no. 86) and fugitives therefrom, belong to the time of Mahiri (Acta Orientalia xii, pp. 41 sqq.), and (2) that references to Parvata, common in the documents of Mahiri's time (Kala Puṃñabala being his son) do not occur at dates in the reign of his successor, Vasmana.

^{page 524 note 1} Conceivably connected with Hindī ḍhalait, ḍhalayat “peon”, Kashmiri dalēḍh “sheriff's officer”, “policeman”, “member of a court”.

^{page 524 note 2} Nos. 571, etc., see Index.

^{page 524 note 3} See Act. Or., xii, p. 69.

^{page 524 note 4} Sc. Taxilā; but also in a Mathurā inscription, and the feminine (nāgarikā) in one from Bharaut; so that a doubt arises. There is too high a percentage of nāgarakas residing away from Nagarahāra. In fact no nāgaraka is ever mentioned in connection with Nagarahāra or the region to which it belongs.

^{page 524 note 5} BSOS., VII, p. 789. Cf. Adhar-narseh in Justi's Iranisches NamenbuchGoogle Scholar.

^{page 525 note 1} For one such party mentioned in a Tibetan document of c. A.D. 800, see JRAS., 1930, p. 291.

^{page 525 note 2} Act. Or., x, p. 74.

^{page 525 note 3} For other cases also of su/ṣu, see nos. 905, 909.

^{page 525 note 4} Shiratori, Mem. Toyo Bunko, ii, p. 99, has Su-Ui.

^{page 525 note 5} Thomsen, , Inscriptions de I'Orkhon, p. 154 (38).Google Scholar

^{page 525 note 6} Tib. Lit. Texts and Documents, i, p. 319, also elsewhere. A Khotanā-Saka form of the name seems not to be known, unless the Sūliya of Leumann, Buddhistische Literatur I Teil: Nebenstücke, p. 169, is such (see Konow, Saka Studies, p. 183), which in the context seems highly questionable.

^{page 525 note 7} Cf. Karlgren, nos. 909, 919, and Professor Shiratori's argument (pp. 99–100) concerning Šu-yu.

^{page 526 note 1} For the Χ see Bailey, loc. cit.

^{page 526 note 2} Of. Gōttingische Gelehrte Anzeige, 1912, p. 555.

^{page 526 note 3} The Candragarbha-sūtra, wherein it occurs, was rendered into Chinese in the period A.D. 550–577 (Lévi, BÉF d'E-O, v, p. 261).

^{page 526 note 4} Die Soghdischen Handschriften des Britischen Museums (Index).

^{page 527 note 1} JBAS., 1930, p. 297.

^{page 527 note 2} Indian Culture, ii, pp. 193–1Google Scholar.

^{page 527 note 3} Berlin Academy Sitzungsberichte, 1913, pp. 406 sqq.

^{page 527 note 4} See Nordisk Tidsshrift for Sprogvidenskap, xi (1938), p. 13, and Act. Or., v, pp. 34–5.

^{page 527 note 5} On identification of “Khotani-Saka” with the language of the coins and on distinctions between them see Konow, Ostasiat. Zeitschrift, viii, pp. 224, 229, and already in the Berlin Academy Sitzungsberichte, 1916, pp. 793 sqq., 811 sqq.

^{page 527 note 6} On the question of such an immigration and on its direction and date Professor Konow's views are stated in Ein neuer Saka-Dialect, pp. 28–32; Indian Culture, ii, pp. 195–8Google Scholar.

^{page 527 note 7} Indian Culture, ii, p. 195Google Scholar.

^{page 528 note 1} Especially the mark distinguishing g and ģ, etc.; see the Kurram casket inscription, edited in Epigr. Ind., xviii, pp. 16 sqq., and Konow, Kharoḍṭhi Inscriptions, pp. 152 sqq.

^{page 528 note 2} Professor Konow's edition, see Index.

^{page 528 note 3} d for t in cadu, t for t in ṛati!

^{page 528 note 4} In Kharoḍṭhī document no. 164 we find pitumadtte.

^{page 529 note 1} Note, for instance, that the prateṛo cited above, as an example of d/t, recurs in document no. 762.

^{page 529 note 2} It is noticeable that nearly all the occurrences of initial dh- for d- are in legal documents, one of which, no. 348, has two together, dhaṃBta dheṛati (but no. 580 dhaṃḍa deyaṃti). It seems possible that the aspirate form in writing was a professional affectation.

^{page 529 note 3} Whether in fact initial d before vowels was pronounced as δ in Khotan or Marālbāshī (Konow, Saka Studies, p. 28, Ein niteur Salca-Dialekt, pp. 8–9, 17) is a problem which I am not competent to discuss. For Saka-Khotanī of the third to fourth centuries A.D., and even of the fifth to tenth centuries, the analogy of nineteenth to twentieth century Pamir dialects (but not of Persian) seems weak. On the other hand, the evidence of transcriptions from and into other languages (Chinese, Tibetan, “Tokharl,” Uigur, etc.) would always, in regard to a difference such as that between d and δ, be open to dispute. Even the surviving ancient names of localities, such as Duwa < Durya, can be put aside on the ground of the changes in population and

^{page 529 note 4} The document 661 has ṛudhi = suṛdhi. And what of udhiṛa = uddiṛya in three documents ?

^{page 529 note 5} It is unmethodical to transliterate in the form cozbo the word which in the Kharoṣṭī Prākrit is spelled cojhbo. The Indian users of the Prākrit, no doubt, pronounced the z in the native word, like any other z, as jh (or j).

^{page 530 note 1} Has it been suggested by European employments of dh and gh to represent spirants ?

^{page 530 note 2} Buddhistische Literatur I Te.il: Nebenstüche, p. 92.

^{page 530 note 3} Apocope of u- (particularly easy before v, in a suffixal word) occurs in Prākrit (see Pischel, § 141) and in Apabhraṃṛa (Jacobi, Bhavisattakaha, p. 29). It occurs also in “Tokharī” (see the Grammatik of Sieg, Siegling, and Schulze, §§ 90, 95).

^{page 530 note 4} Norsk Tidsskrift for Sprogvidenskap, vii, p. 54.

^{page 530 note 5} And probably far earlier in Pāli upanti (in a Jātaka verse, Fausböll, iv, p. 337), also upantika: see the Dictionary.

^{page 531 note 1} BSOS., VII, p. 779; The Language of the Kharoṣṭhī Documents, pp. 77–8Google Scholar. The fact that the prevalent spelling is avana, not avana, which, however, is quite common, seems to be without etymological significance. The variation between v and v, which seem in fact to have differed more or less as expounded by Rapson (p. 309), is frequent in most situations (even initial). If the v was bilabial, the difference was slight indeed, a mere matter of emphasis, as perhaps appears in the fact that “gold” seems, when not svarna, to be suvarna, whereas “of good hue ”in its two occurrences, is suvarna. Cf. avamicae/ava˚, from apamity-.

^{page 531 note 2} Possibly some of the “empty” towns mentioned in Two Medieval Documents from Tunhuang (by F. W. Thomas and Sten Konow), p. 148, were of this nature.

^{page 531 note 3} Saka Studies, p. 92⁵.

^{page 531 note 4} Act. Or., xiv, p. 232.

^{page 531 note 5} Ein neuer Saka-Dialekt, p. 35 (804), 1. 24. But in Norsk Tidsskrift, xi (1938), p. 10, the Khotanī vāna is again “house”, “temple”.

^{page 532 note 1} Nebenstücke, ppx. 42²², 46.

^{page 532 note 2} Very likely the name, Par-ma, of a market-place in the Khotan city (Thomas, Tibetan Literary Texts and Documents, i, p. 134), was this same word.

^{page 532 note 3} See Act. Or., ii, pp. 125–6; Indian Culture, ii, p. 195.

^{page 532 note 4} The forms of the akṣaras and general ductus may be merely a matter of style and date.

^{page 532 note 5} Asia Maior, ii, pp. 270–1.

^{page 533 note 1} The chaos of spellings exemplified in Professor Bailey's article “Hvatanica IV” in BSOS., X, pp. 886 sqq., see especially pp. 898 sqq., may, one imagines, be due in part to manifold provenance of the MSṢ. cited, so that in the MSS. severally there may be a less pronounced inconsistency. Writings from Ṛa-cu/Tun-huang seem to be specially marked by enormities, and variations of spelling, for instance, in the Tibetan transcriptions of Chinese: it is from Ṛa-cu that we get, in Tibetan writing, such forms as Hdre-spe-sad, 'A-ḫdaḫ-bad, representing Sanskrit Duṣprasaha, An[ava]tapta.

^{page 533 note 2} In future the Tibetan evidence in regard to the pronunciation of a number of Khotan words, some native, some loans from Indian sources, may be consulted with profit.

^{page 533 note 3} As was pointed out by Sylvain Lévi (J. As., XI, v (1915), p. 191), and Pischel's Gramm. d. Prākrit-Sprachen, § 217. Lévi's full discussion of the ys is contained in a paper contributed to the Festbundel uitgegeven door het 'KoninUijk Bataviaasch Oenootschap van Kunsten en Wetenschappen…, published at Weltevreden in 1929, vol. ii, pp. 100–8, and reprinted in Mémorial Sylvain Lévi (Paris, 1937), pp. 355–363.

^{page 533 note 4} Berlin Academy Sitzungsberichte, 1913, pp. 406–9.

^{page 534 note 1} But the mere fact that Caṣṭana figures in a collection of Buddhist stories known in Central- Asian languages does not constitute such a connection: what of Aṛoka, Puṣyamitra, Virūḍhaka, and other Indian rulers celebrated in the Buddhist literature of Central Asia ?

^{page 534 note 2} Schwanbeck, Megusthenis Indica, p. 12 (6), and Index: cf. Passalai, Paxalai = Pañcāla, ibid., p. 35 (31).

^{page 535 note 1} Kharoṣṭhī Inscriptions, pp. 61–2.

^{page 535 note 2} Journal of Indian History, xii, p. 28; Acta Orientalia, iii, p. 69; Ostas. Zeit., viii, p. 230; Norsk Tidsskrift for Sprogvidenskap, xi, pp. 10–11; Journal of Indian History, xii, p. 28.Google Scholar

^{page 535 note 3} In “Tokhari”the loan-word āṣānik is used as equivalent to Sanskrit arhant (Sieg, Siegling and Sehulze, Tocharische Grammatik, p. 13, § 22). For book-Pahlavī the meaning “pious believer” is evidenced by Bartholomae, Zum Altiranischen Wōrterbuch, p. 117.

^{page 535 note 4} The word occurs several times in the Mahāvastu with the meaning stated, and is recorded in the Amara-koṛa (I. iv, 25). It is not necessary here to discuss the etymology (see Johansson in Le Monde Oriental, ii, pp. 96–7), or the variation of the initial sibilant, which points probably to some interference.

^{page 537 note 1} To take pirova here as equivalent to pirovala does not seem right.

^{page 537 note 2} Here and in other cases of akṣaras still more or less dubious I follow the spelling of the edition.

^{page 538 note 1} Serindia, p. 240, and photograph 75.

^{page 538 note 2} Particulars of modern routes are given (with mention of wells, etc.) in Dutreuil de Rhins (Grenard), La Haute Asie, iii, pp. 218–19, 221.

^{page 539 note 1} “A well of water in as dry place.”

^{page 540 note 1} Sic in Index.

^{page 540 note 2} Also, as suggested by Dr. Burrow, pp. 106, 109, to the Puṃniyade of no. 554.

^{page 541 note 1} Cf. Yangt'ung (Bushel), JRA8., 1880, p. 527 (9)) for Tib. Byaṅ-thaṅ

^{page 541 note 2} E.g. in Professor Franke's excellent Geschichte des Chinesischen Seiches, ii, pp. 63, 135–6.142, etc.

^{page 541 note 3} I think that Professor Lüders in attributing to me {Acta Oientalia, xviii, p. 34) the view that the substrate language of Shan-shan and other southern parts of Chinese Turkestan was connected with “Tibetan” did not mean anything different from my own expression (Festgabe hermann Jacobi, p. 73; cf. Asia Maior, ii, p. 270), “Tibeto-Burman”

^{page 542 note 1} V.4.75.

^{page 543 note 1} On the spelling see p. 537, n.2.

^{page 544 note 1} See Böhtlingk and Roth's Wōrterbach, s.w. ghuṣ, ghoṣaṇa, paṭaha.

^{page 544 note 2} Similarly we find both pravaṃnaģa lihitaģa 180, 571, and pravaṃnaģaruṃmi lihitaģa, 59, 633.

^{page 545 note 1} “Znr Kontrolle zerschnitten,” sc. cut in two, etc., for verification by subsequent rejoining of the halves, etc., held by different persons.

^{page 545 note 2} .

^{page 546 note 1} .

^{page 546 note 2} .