^{page 49 note 1} BSOAS, XVIII, 1, 1956, 35,Google Scholar citing E. Sieg, Übersetzungen aus dem Tocharischen, I, 10.

^{page 49 note 2} Khotanese texts, III, 138, P 5535.1.Google Scholar

^{page 49 note 3} Details in BSOAS, XIII, 4, 1951,Google Scholar 921 ff.

^{page 49 note 4} J. Masson, La religion populaire dans le canon bouddhique pòli, 126 ff.

^{page 49 note 5} H. Jacobi, Ausgewählte Erzählungen, 12.16; 21.30, 33.

^{page 49 note 6} Smithsonian Miscellaneous Collections, 1928, Yakṡas.

^{page 49 note 7} Vedische Studien, III, 143.

^{page 49 note 8} L. de la Vallée Poussin, JRAS, 1911, 773.

^{page 49 note 9} Edited by Lévi, Sylvain, Journal Asiatique, 1915, 1,Google Scholar19 ff.

^{page 50 note 1} The passage has long exercised translators. Thus the first edition by R. Gauthiot rendered ‘au dieu des créatures’, and E. Benveniste gave ‘en présence du Mithra des créatures’.

^{page 50 note 2} A. F. R. Hoernle, Manuscript remains, p. 41, Candropama sūtra.

^{page 50 note 3} Edited H. Hoffmann, p. 79, the nom. plur. also on p. 35.

^{page 50 note 4} Sogdian Paris texts with note, p. 215. F. W. K. Müller, Uigurica, II, p. 80, naivaziki tngrilärkä. J. von Klaproth, Uigurisches Vokabular, p. 17. The Oγuz Qaγan text edited W. Bang and G. Rachmati, p. 26.

^{page 50 note 5} BSOAS, X,4,1942, 917.Google Scholar

^{page 50 note 6} BSOAS, X, 4, 1942, 917,Google Scholar and Khotanese texts, III, 45, P 2023.3–4.Google Scholar

^{page 51 note 1} H. Lüders, Weitere Beiträge zur Geschichte und Geographie von Ostturlcestan, 7.

^{page 51 note 2} E. Leumann, Maitreya-samiti, 268.

^{page 51 note 3} This word has been compared with Avestan kaēta-, see Transactions of the Philological Society, 1955,67.

^{page 51 note 4} BSOAS, X, 4,1942,892,Google Scholar and Transactions of the Philological Society, 1954,144 .

^{page 51 note 5} Gauthiot, R., Grammaire sogdienne, I, 39–52.Google Scholar

^{page 51 note 6} Henning, W.B., BSOAS, XI, 4, 1946, 721,Google Scholar and XI, 3, 1945, 470, footnote 1.

^{page 51 note 7} Gershevitch, I., Transactions of the Philological Society, 1945, 139.Google Scholar

^{page 51 note 8} W. B. Henning, Manich. Bet- und Beichtbnch, p. 73.

^{page 51 note 9} J. Markwart, Caucasica, VII, 128. For Armenian -ouhi as a feminine ending contrast Markwart's derivation from Mid.Pers. puhr, and the discussion by E. Benveniste, Trans. Phil. Soc., 1945, 74.

^{page 52 note 1} Recognition that Khotanese has three initials in the word maiz-‘mingere’, as set out in BSOAS, XVIII, 1, 1956, 40,Google Scholar namely, m-, b-, and ph-, has shown that pharsavatā- need not be considered a loan-word in Khotan. It is rather to be taken as containing pharsa- from older fraša- beside the spelling with bras-. The ph- will be the unvoiced ƒ- beside the voiced b- = β.

^{page 52 note 2} See for this and others, BSOAS, X, 4, 1942, 916–7.Google Scholar

^{page 52 note 3} Transactions Phil. Soc., 1945, Asica, p. 35.

^{page 53 note 1} G. Morgenstierne, Norsk Tidsskrift for Sprogvidenskap, V, 40–1; IIFL, II, 274 and 420.

^{page 53 note 2} G. Morgenstierne, IIFL, II, 190. The frequent replacement in Armenian loan-words of Iranian - δ- by -r- may be a feature of the Iranian from which the words come. Then one might expect to find trace of this change in the Persian texts.

^{page 53 note 3} E. Herzfeld, Arch. Mitteil. aus Iran, vn, 97 ff., offered Ind. Jambu.

^{page 53 note 4} R. Gauthiot, Gram, sogd., II, 17.

^{page 53 note 5} Khotanese texts, II, 71, line 11.

^{page 53 note 6} Ch. cvi.001, 29, and b 5 in Khotanese texts, II, 60, 61.

^{page 54 note 1} H. Lüders, Textilien im alten Turkistan, 20 ff.

^{page 54 note 2} According to V. Abaev, Osetinskij jazyh i fol'klor, I, 480, Digor nimät is used only of the ‘burka, felt cloak ’.

^{page 54 note 3} W. B. Henning, Sogdica, 8, and supplement; G. Morgenstieme, IIFL, II, 48٭.

^{page 54 note 4} A fourth word of similar form but unconnected is attested in the Old Indiannamata-‘bent’ from the verb namati‘to bend’.

^{page 55 note 1} If one can rely on this Kalāšā at with short -a- from older ٭arta- as the regular replacement of art- it would show that Pātu is not from ٭part-, see BSOAS, XVI, 3, 1952, 429.Google Scholar

^{page 55 note 2} Asia Major, N.S., I,1,1949,45.

^{page 55 note 3} Asia Major, N.S., I,1,1949,45; Khotanese texts, II, 129.

^{page 55 note 4} Khotanese texts, II, 59. The syllables -rakīnai may contain an older ٭raxtaka- connected with NPers. raxt ‘clothes‘.

^{page 55 note 5} Khotanese texts, II, 103.

^{page 56 note 1} BSOAS, XIII, 4, 1951, 933.Google Scholar For synkt, see I. Gershevitch, Grammar, p. 161, from W. B. Henning.

^{page 56 note 2} This I have from G. Morgenstierne. The name is in Caraka. In IIFL, II, 252, Morgenstierne cited NPers. sunǰāq ‘pin‘beside Oss. syndz.

^{page 56 note 3} Henning, W.B., BSOAS, XI, 4, 1946, 726.Google Scholar

^{page 56 note 4} H. Yule, The travels of Marco Polo, I, 94. Yule and Louis Hambis, La description dumonde, 361, both consider this word of Marco Polo to come from a Persian hinduvāni‘Indian’.

^{page 56 note 5} Pauly-Wissowa, s.v. Chalybes. R. Eisler, Caucasica, v, 82, is exuberant but unhelpful. On the word ayas- see Benveniste, E., Celtica, III, 1955, 279–83Google Scholar

^{page 57 note 1} J. F. Baddeley, Rugged flanks of Caucasus, I, 129; Minorsky, V., BSOAS, XI, 4, 1946, 763.Google Scholar

^{page 57 note 2} S. Feist, Vergl. Wōrterbuch der gotischen Sprache, 353; V. Kiparsky, Die gemeinslavischen Lehnuōrtern aus dem Germanischen, 138–40.