^{page 32 note 1} BSOAS, XV, 538–9.

^{page 32 note 2} mū has the -u- from -i. or -ä- after the labial m-. The transcription with the macron indicates a particular shape of akṫara, but in later Khotanese cursive script may mean no ṁore than a variety of the akṫara mu.

^{page 32 note 3} BSOAS, XV, 538.

^{page 32 note 4} tgl- ‘arrow’, later NPers. tīr, with āha- ‘throw, shot’ from ah- ‘throw’ in Avestan, and Khotanese, OInd. ásyati. This rather than the verb ās- ‘come’ offered by H. S. Nyberg, Øst og Vest, 66.

^{page 32 note 5} ed. H. S. Nyberg, Øst og Vest, 66.

^{page 32 note 6} Jamasp-Asana, Pahlavi Texts, p. 69, §§ 123–5, and 70, § 138; edited in transcription and with translation by H. S. Nyberg, Texte zum mazdayasnischen Kalender, 48–51.

^{page 33 note 1} ywy *yōy ‘channel, stream’ with HPL- ‘dig’, kandan. Related words were quoted in Trans. Philol. Soc., 1954, 139: OPers. yauviya-, NPers. Rigvedic Nyberg gave yavē ‘Korn’. Usually we have ywrt'k *yav-artāk, Turfan Mid. Pers. yw'rd'w. The yav- seems to occur in BSOS, VI, 583, družz-ē ī vat-yavakān xvānīhēt ‘demon called producer of bad crops’.

^{page 34 note 1} Jamasp-Asana, Pahlavi Texts, 155.2.

^{page 34 note 2} Different from the Portuguese mesa ‘table’ which later became familiar to Asia, see the forms quoted in S. Rodolfo Dalgado, Influência do vocabulário português em linguas asiáticas, 109.

^{page 34 note 3} Sten Konow, Saka Studies, 70 (the Tibetan corresponds cd + ab).

^{page 35 note 1} Thomas, F.W., Ada Orientalia, XII, 38;Google ScholarBurrow, T., Language of the Kharosthī documents, 111; BSOAS, XV, 538–9.Google Scholar

^{page 35 note 2} The Iranian and Armenian -ak was represented in Georgian loan-words by -ak'-, -ag-i, and -a; thus ešmak'-i, ešma ‘demon’ corresponds to Mid. Parth. ‘šmg *ešmag, cited in Boyce, M., BSOAS, XIII, 912.Google Scholar The form in -ag- is known in Georg. šmag-i ‘mad’, ušmago ‘sane’, see Deeters, G., Caucasica, III, 81.Google Scholar

^{page 35 note 3} Deeters, G., Caucasica, IV, 3–4.Google Scholar

^{page 35 note 4} Sieg, E., Übersetzungen aus dem Tocharischen, I, 10.Google Scholar

^{page 35 note 5} Thomas, F.W., Tibetan Literary Texts, II, 342.Google Scholar

^{page 35 note 6} JAOS, LXXV, 33. P. Poucha, Institutiones linguae tocharicae, 377, 397.

^{page 36 note 1} BSOAS, XV, 539.

^{page 36 note 2} The -š- not -šx- in Armen. mšak indicated the not K.

^{page 36 note 3} The var. is likely to refer to the chanting of the enthusiastic Kavi during the celebration of his cult. I propose to see in this word the verbal base ang- ‘tell, celebrate, sing, make poems’ which occurs elsewhere in OInd. ángiras-, originally the ‘singer’, and the ‘announcer’ in Greek â¼¼єλos (as we have OInd. kār- ‘ singer ‘, and Greek Kâpuξ ‘ herald ‘). It was in this ąnman- that the traditional learning knew Ahura Mazdā (Yasna 45.10, mazdȧ srāvī ahurō). On this more must be offered later. From the same ang- came Rigvedic āngūṫā- ‘ song ‘, see Trans. Phil. Soc., 1955, 78.

^{page 37 note 1} This Truth, әrәta-, arәta-, aša-, has become Zoroaster's new interpretation of the ancient doctrine that the potent fitting of facts by recognition of ‘truth’ was a cosmic force.

^{page 37 note 2} J. Bloch, BSOS, VIII, 414.

^{page 37 note 3} Avesta, 78. This same plej is given also for Yasna 48.5 fšuyō.

^{page 37 note 4} ‘Note sur l'agriculture dans l'Avesta’, Revue de l'Histoire des Religions, LIX, 1909, 333–7.Google Scholar

^{page 37 note 5} ‘The Indo-European root meik-: meiĝ- and Avestan ’, Language, IV, 178–80.

^{page 37 note 6} Die Zoroastrische Religion, 3: ‘begehren’.

^{page 37 note 7} F. C. Andreas in Gāthā's des Zarathustra, übersetzt und erklärt von H. Lommel, 85 ‘strebten’.

^{page 38 note 1} JBAS, 1952, 177.

^{page 38 note 2} dāng is explained in the Frahang ī Pahlavīk by mĞvak ‘fruit’.

^{page 38 note 3} Zand ī xvartak apastāk, Māh nigāyišn, ed. Dhabhar, p. 30, 6.

^{page 39 note 1} For mȧnhəm tāpayeiti note the remarks of V. Pisani, Rivista degli Studi Orientali, xv, 362.2

^{page 39 note 2} Zum Altiranischen Wōrterbuch, 207.

^{page 39 note 3} H. Hübschmann, Armen. Gram., 194, had thought of this possible connexion.

^{page 39 note 4} Die Yäšt's, 43, 46.

^{page 39 note 5} A similar list also in Gr. Bund., 89–90, gives 17 liquids.

^{page 39 note 6} So in the Altiranisches Wōrterbuch.

^{page 39 note 7} Bartholomae, Zum Altiran. Wōrterbuch, 207. Andreas-Henning, Mitteliran. Manichaica, II.

^{page 40 note 1} Some cases were quoted (in an article rather remote) in the Silver Jubilee Volume of the Zinbunkagaku-kenkyusyo, Kyoto, 1954, 5–7, particularly loan-words. The same alternation lies behind the comparison of NPers. bādrīsah and Rigvedic mātaríśvan-, W. B. Henning, JRAS, 1946, 13.

^{page 40 note 2} Jīvaka-pustaka 81 r 2 in Khotanese Texts, I, 165, 30 (where the word was misread; corrected in Khotanese Texts, II, 134).

^{page 40 note 3} So in the Siddhasāra 102 v 5. The word was wrongly rendered in E by ‘Galle’.

^{page 40 note 4} So Siddhasāra 18 r 2; 122 v 4.

^{page 40 note 5} Khotanese Texts, 1, 165.

^{page 40 note 6} Siddhasāra 9 v 2; 131 r 1; beśi 20 v 5.

^{page 40 note 7} Siddhasāra 131 r 1; P 2893.230.

^{page 41 note 1} Walde-Pokorny, Vergleichendes Worterbuch, II, 245, quoting Lidén, E., Studien zur altind. und vergl. Sprachgeschichte, 1897, 41.Google Scholar Vigfússon and Cleasby's Dictionary quotes mysa from Króka Refs saga.

^{page 41 note 2} J. Pokorny, Indog. etym. Wórterbuch, 722.

^{page 41 note 3} None in the Petersburg Lexicon (O. Bóhtlingk and R. Roth, Sanshrit-Wórterbuch); Monier-Williams's Dictionary (‘of doubtful origin’); Lanman's Sanskrit reader; omitted in Walde-Pokorny. Uhlenbeck's Etymological Dictionary gives no connexion outside Indo-Iranian.

^{page 41 note 4} is not in Steingass's Persian Dictionary nor in Vullers’ Lexicon. Both are also given by W. Geiger, Etymologie des Balᥫčī, 12.

^{page 41 note 5} La langue marathe, 376.

^{page 41 note 6} J. Filliozat, Fragments de textes koutchéens, 113.

^{page 41 note 7} T. Burrow, Trans. Phil. Soc., 1946, 10, 22.

^{page 42 note 1} Ossetic zar-, see Bailey, H.W., ‘Indo-Iranian studies, III’, Trans. Phil. Soc.1, 1955, 55 ff.Google Scholar

^{page 42 note 2} Walde-Pokorny, II, 687.

^{page 42 note 3} Like the *nira- in Mid. Pers. panīr ‘cheese’, discussed in the S. K. Chatterjee volume, 1956.

^{page 42 note 4} W. B. Henning, Manich. Bet- und Beichtbuch, 129.