^{page 95 note 1} Without Iranian, Indo-Iranian studies must always limp along. Neglect of Iranian impoverishes the Indianist: note the omission of Ossetio läsäg “salmon” (salmon are fished in the Caspian sea), and of Khot. rrāṣa-, NPers. raxš, and Armen. loan-word erašx used of a dark colour in P. Thieme's discussion of Old Ind. la眱sa and lākṣā in Zeits. vgl. Sprachf. 69 (1951), 209 ff., and in M. Mayrhofer's reference to that study in Studien zur indogerm. Grundsprache, 1952, 45. Similarly a connection of Old Ind. bráhman- with Greek μορφή would surely be excluded by using the Iranian evidence (Thieme, P., ZDMG., 102, 1952, 127)?Google Scholar It would be profitable in discussing Old Ind. Śaśa- “hare” to consider the Iranian Parāči sahōK and Khot. saha- “hare”, see JRAS., 1831, 424, and G. Morgenstierne, IIFL., 1.287; see M. Mayrhofer, loc. cit., 27 ff. Three similar deficiencies were noted in JRAS., 1951, 193–5.

^{page 95 note 2} As often in Later Khotanese i is replaced by ai. The sound intended may have been ι, since ai is put for Turkish ï in, for example, Staël-Holstein Roll 22: cąmaiḍa baḍaikä for Čamïl balïq.

^{page 95 note 3} 64 v 3.

^{page 95 note 4} For the importance śruti, bahuśruta, see Przyluski, J., Rocznik Orient. 8.15Google Scholar.

^{page 96 note 1} The phrase ḥphags-pa slob-pa-rnams “the venerable preceptors” is given in Jäschke's Dictionary. In the Uigur colophon of Or 8212.109.46 a occurs ary-a ačari tükällig bilig isdon-pa baxšï, see the facsimile in M. A. Stein, Serindia, plate clxv.

^{page 96 note 2} Full references can be found in PW.

^{page 96 note 3} See Konow, Sten, NTS. 11.83Google Scholar.

^{page 97 note 1} Jātaka-stava 21 v 4, Avestan an-, Sogd. βrn “breath”.

^{page 97 note 2} N. 115.19.

^{page 97 note 3} Saka Studies, 125.

^{page 97 note 4} Thieme, P., ZDMG., 91.107; 102.121Google Scholar.

^{page 97 note 5} See Nehring, O. Schrader-A., Reallexikon der indoger. Altertumskunde, ii, 464Google Scholar.

^{page 97 note 6} Bahman Yašt, ed. Anklesaria, 7.22Google Scholar. In Avestan we have Y 68.12 hašąmča hāvištanąmča aēθrapaitinąmča aēθryanąmča.

^{page 97 note 7} A similar hā- occurs in Av. hākur∂na- “co-operation”. Below, p. 115.

^{page 97 note 8} Thomas, F. W., Tib. Lit. Texts, i, 42Google Scholar. In E it is the tale of Dusprasava.

^{page 98 note 1} E. Leumann's rendering “durchdrungen” evidently took its meaning from Sansk. āviṣṭa- from viś- “penetrate” and is unequal to the situation in the Kali Age.

^{page 98 note 2} The Poona Critical edition has adopted narāvṛṣṭibhir arditāḥ, and provides many other variants, Araṇyaka-parva 3.188.70–1.

^{page 98 note 3} Wackernagel, J. (Zeits. vgl. Sprachf. 67 (1942), no. 42)Google Scholarwrongly compared viṣṭi- “forced labour” with the vic- “turn” in Lat. vicissim. It is not a matter of regularity. Similarly the vāra- of vāra-yoṣit “courtesan” does not mean “turn”, see Thieme, P., ZDMG. 91.108Google Scholar.

^{page 98 note 4} Weller, Fr., Zum sogd. Vim. pp. 60–4Google Scholar.

^{page 98 note 5} W. B. Henning had suspected a miswriting of *āyōštak, see Sogdica 51.

^{page 99 note 1} Connection with Ind. viṣṭi- was proposed by Müller, Fr., WZKM. 6.268Google Scholar; repeated by Hübschmann, H., Armen. Gram. 515Google Scholar, and by H. Adjarian, Armen. Etym. Dict., s.v. višt.

^{page 99 note 2} Variants bejs-, baijs-, and bej- indicate -i-.

^{page 99 note 3} BSOS. 6.265 ff.; Trans. Phil. Soc. 1945, 75. On Avestan vōiγnā “famine”, see Henning, W. B., BSOAS. 11.717Google Scholar.

^{page 99 note 4} Both E. Leumann and Sten Konow traced this word to bis- “enter”.

^{page 99 note 5} NTS. 11.69.

^{page 99 note 6} Khot. Bud. Texts 34.70. There is also būka- “food”, būkaja “steward”.

^{page 99 note 7} Saka Studies 125.

^{page 100 note 1} Printed b[ä]ṣỖa.

^{page 100 note 2} E. Leumann in his translation of E; my own note in BSOS. 9.73, and Sten Konow, NTS. 11.69.

^{page 100 note 3} P 4099.134–5 in Khot. Bud. Texts 119.

^{page 100 note 4} On Sogdian bš see Henning, W. B., Sogdica 43Google Scholar.

^{page 101 note 1} Above, p. 96.

^{page 101 note 2} E 25.431; 18.27; Sanghāṭa-sūtra 80 b 2; 117 a 2; Vajr. 43 a 2. Similarly kä ta 2nd plur. “think” occurs beside kei ta.

^{page 101 note 3} Pali tiropākāra, Mahāvy. 219 tiraḥkuḍyaṁ tiraḥprākāram.

^{page 101 note 4} E. Leumann gave “bindet”. Konow, Sten, NTS. 7.17Google Scholar, proposed to trace a form of bād- “thrust”. If the subscript hook's absence separates bettä from betta, a derivative of bād- would be possible.

^{page 102 note 1} See Hilka, A., Die altindischen Personennamen, 119Google Scholar.

^{page 102 note 2} BSOS. 9.82.

^{page 102 note 3} Dātastāā ī dēnīk 4.4; 30.9; Kārnāmak 15.22; Dātastān ī dēnīk 36.17. See also BSOS. 7.73.

^{page 102 note 4} Of the two Avestan forms with vī- and vī- I accept that with vi- since later the name became Guštāsp.

^{page 102 note 5} I had in JRAS. 1939, 117, tried to find a word višta- “bound”, from the base ṷeiĉ- attested in Lat. vincīre and Old Ind. paḍ-vīśa- “fetter”, as C. Bartholomae, ZAIW. 27 had done. I do not think this solitary evidence for viś- “to bind” necessarily excludes this word from Indo-Iranian, but one would certainly have liked to find more of it. On Mid.Pers. gyšt, see below, p. 111.

^{page 103 note 1} JA. 1936, 1.229.

^{page 103 note 2} Beiträge zur Namenforschung 2.165 ff.

^{page 103 note 3} Henning, W. B., Sogdica 43Google Scholar “open-minded”, like NPers. gušādah-dil.

^{page 103 note 4} Morgenstierne, G., NTS. 12. 264Google Scholar.

^{page 103 note 5} Zātspram 1.1 (K 35, 233 v 16); Gr Bd. 3.5. Pāzand vašādaī occurs in Škandgumānīk vičār 16.52.

^{page 103 note 6} Sanghāṭa-sūtra 83 a 6 ṭä naḍe is in Tibetan de “he”. References apud Konow, Sten, Salca Studies 158Google Scholar.

^{page 104 note 1} Konow, Sten, NTS. 7.13Google Scholar. The name of the Pāṇḍus was used for the number “five”, Bühler, G., Ind. Palaeographie, p. 81Google Scholar. Among the Kāfirī speakers the group was enlarged to seven, Morgenstierne, G., Acta Orient. 21.165Google Scholar.

^{page 104 note 2} piśkala may be another form of piṣkala “section”; ttāra is “that”, “darkness”, “forehead”, and perhaps “cord”; gvays- “to split, separate”.

^{page 104 note 3} ed. E. H. Johnston 7.45. Pāṇḍu occurs also in 4.79, and Saundarananda 7.45.

^{page 104 note 4} Tocharische Sprachreste 134 and 135.

^{page 104 note 5} Müller, F. W. K., Uigurica ii 24–6Google Scholar. According to Kāšγarī tonga meant “tiger”. He knew it also in proper names, Brockelmann, C., Index, pp. 213, 250Google Scholar.

^{page 104 note 6} P 2801.34, 35, BSOAS. 10.367.

^{page 104 note 7} Konow, Sten, Saha Studies 158Google Scholar; E 14.77; 6.2; BSOAS. 10, Rāma text six times; P 2834.17, P 2957.141, P 2834.23 (in Khot. Bud. Texts 45; 38; 45), Jātaka-stava 13 r 2.

^{page 104 note 8} Salemann, C., Manichäische Studien 57Google Scholar.

^{page 104 note 9} Lüders, H., Varuṇa i 13–27Google Scholar; Thieme, P., Bráhman 110; BSOAS. 14Google Scholar.

^{page 105 note 1} Šāhpuhr i insc. line 12; Pers. line 14 wlty č. baḍa- in Asia Major, n.s., i 44.

^{page 105 note 2} Siddhasāra 17 r 3, translating Sansk. vartaka, Tib. bartaka.

^{page 105 note 3} See Morgenstierne, G., Etym. Voc. Pashto, p. 54Google Scholar; IIFL 2.548. Indo-Aryan Khowar bertī, Nep. baṭṭāi, Sansk. vartikā and others.

^{page 105 note 4} So in Asia Major, n.s., ii 38. Note paḍe “axes”, Osset. färät.

^{page 105 note 5} ttu näjsaḍu = Tib. dper-na.

^{page 105 note 6} Vajr. 24 a 3 rendering Bud. Sansk. prāmuncat.

^{page 105 note 7} Siddhasāra 148 v 5.

^{page 105 note 8} Siddhasāra 103 v. 1 = Sansk. unmārgī; Sudhanāvadāna P 2025.114, Khot. Bud. Texts 15.

^{page 105 note 9} Konow, Sten, Saka Studies 158Google Scholar, traced the word to *nartavan-.

^{page 105 note 10} For rare verbal bases see above, p. 96.

^{page 105 note 11} Morgenstierne, G., IIFL. 1.276Google Scholar. Yidγa in IIFL. 2.233.

^{page 105 note 12} Morgenstierne, G., NTS. 12.265Google Scholar.

^{page 106 note 1} n∂r∂-myazdana- adj. from n∂r∂-myazda- may have nṛ- in verbal meaning “strong” or “making strong”, see JRAS. 1939, 116.

^{page 106 note 2} The hymn will be edited in Miss M. Boyoe's study of the Hymn Cycles, and she has kindly allowed me to quote her interpretation. The text is paraphrased in Reitzenstein's, R.Das iran. Erlösungsmysterium, p. 22Google Scholar.

^{page 106 note 3} The transitive-causative form is wyn r-, Zor.Pahl. vīnārtan “to arrange, establish”, which could be understood to derive from “be well placed”.I. Scheftelowitz's comparison with the Indo-Eur. base sner- “noose” hardly seems suitable.

^{page 106 note 4} See BSOS. 8.343 and Iwanow, W., Gabrī Dialect, p. 70Google Scholar.

^{page 106 note 5} Mededeelingen d. K. Ned. Akad. 1951.

^{page 106 note 6} JRAS. 1943.2.

^{page 106 note 7} Against such procedure we are warned by Dumézil, G., “Maiestas et Gravitas,” Revue de Philologie, 26 (1952) 7 ffGoogle Scholar.

^{page 107 note 1} Charpentier, J., Le Monde Oriental 25.1 ffGoogle Scholar. after H. Jacobi.

^{page 107 note 2} Les dieux indoeuropéens, p. 59.

^{page 107 note 3} F. B. J. Kuiper, loc. cit. 24.

^{page 107 note 4} See for example Vendryès, J., Etudes celtiques 3.237Google Scholar.

^{page 107 note 5} Williams, I., Canu Aneirin, p. 167Google Scholar.

^{page 107 note 6} Hessen's, Irish Dictionary ii 149Google Scholar.

^{page 107 note 7} Dumézil, G., Loki, p. 170Google Scholar.

^{page 108 note 1} Others are quoted by Dumézil, G., Les Légendes sur les Nartes p. 2Google Scholar. Recent explanations of the name are Trubetzkoy, N., MSL. 22 (1921) 252Google ScholarRemarques sur quelques mots iraniens: nart traced to *nartama-. Alborov, B. A.Termin “Nart”, 1930Google Scholar, who proposed a connection with the ancient Nairi land; Rkliokij, M. V., K voprosy o nartax i nartskix skazaniax, reviewed by Tybylty, A., Fidiŭæg 11–12 (1927) 64–5Google Scholar, explained by a social class.

^{page 108 note 2} Osset. Wörterb. s.v. Abaev, V. I., Oset. Jazyk i Fol'klor i 316Google Scholar.

^{page 108 note 3} Pamiatniki ii 37, 5, 25, 60, 26, 28; Osset. Dict. s.v. gūtondžyn and Abaev, V. I., Iz oset. ëposa 32, line 6Google Scholar.

^{page 108 note 4} Both Digor and Iron have plur. -tä, see “Asica”, Trans. Phil. Soc. 1945, 24 ff.

^{page 108 note 5} Blake, R. P., Byzantion 16 (1942–1943) 236Google Scholar; used beside bumberaz-i (Arabic mubāriz) “champion”, as an epithet of the legendary Vaxtang Gorgaslan. My inquiries have failed to find whence thia word is quoted.

^{page 108 note 6} Abaev, , Oset. Jazyk i 304Google Scholar, who reports Svanetian tales on all Nart subjects, printed by A. Šanidze and V. Topuria, 1939.

^{page 108 note 7} Trubetskoy, N., Caucasica 11.13Google Scholar.

^{page 108 note 8} Sbornik Mat. Kavkaza. 21, p. 258. The obliq. collective plur. is nart∂m.

^{page 108 note 9} Bouda, K., ZDMG. 93 (1939)Google Scholar, song no. 3, line 3. Here also t = t .

^{page 108 note 10} Bouda, K., ZDMG. 94 (1940) 237Google Scholar. Abaev, V. I., Oset. Jazyk 277Google Scholarnart∂xw∂.

^{page 108 note 11} Dumézil, G., La langue des Oubykhs, p. 5, 38Google Scholar.

^{page 108 note 12} Abaev, V. I., Oset. Jazyk i 316Google Scholar.

^{page 109 note 1} Dumézil, G., La langue des Oubykhs, 120, 136, 115Google Scholar.

^{page 109 note 2} Žirkov, L. I., Avaro-Eusskij Slovar' (1936) p. 89Google Scholar: skazočnyj velikan.

^{page 109 note 3} Sbornik Mat. Kavkaza 14, p. 46.

^{page 109 note 4} Soslan occurs as one of the names of the second husband of Tamara, the Georgian queen of the twelfth century. The story of the Alan princess Satlinik is discussed by Dumézil, G., Lea Légendes sur Us Nwrtes, pp. 167 ffGoogle Scholar.

^{page 109 note 5} See above, p. 108.

^{page 109 note 6} Osset. Digor färnug “fortunate”; beurä “many”; -äfs “horse”. See Abaev, V. I., Oset. Jazyk i 164Google Scholar; 159.

^{page 109 note 7} The t is aspirated t . In Caucasian languages three series exist: aspirated t , voiced d, and ejective t . The aspirate and the voiced interchange more easily than the aspirate and the ejective.

^{page 109 note 8} It is likely that in disyllables the older -ă- had been preserved as -ä- instead of passing to -a- as in monosyllables. Newly created derivatives would then be regulated according to the inherited scheme. Note the following cases: mard “dead, dead body”; märdä “death” (from *mṛti-) in compound märd-sintä “bier”; marγ “bird”, plur. märγtä; fars “side”, plur. färstä. In Digor farnä, farn, and Iron farn are found, but the Digor abl. sing, färnäj and the adj.färnug, Iron färnyg have -ä-, as also in the compound färnägun, Iron färndžyn.

^{page 110 note 1} For other developments of ātar- see G. Morgenstierne, IIFL. 2.273, Yidya yūr < *ārt, Parāčī âr, Bud. Sogd. 'rδ *ārθ, and also Kušan AΘþO and NSogd. Yaghnābī ǡl, cited in Zoroast. Problems, p. 67.

^{page 110 note 2} Abaev, V. I., Oset. Jazyk i 260 ffGoogle Scholar.

^{page 110 note 3} The names of the principal Nart heroes are transferred to the giants Ǟrxstuaj by the Čeč.ns, see Dumézil, G., Loki, p. 217Google Scholar.

^{page 110 note 4} As in Benveniste, E. and Renou, L., Vṛtra and Vrθragna, p. 13Google Scholar, and Morgenstierne, G., NTS. 12.268Google Scholar.

^{page 110 note 5} Outside Indo-Iranian other words are cited in Walde-Pokorny, , Vergl. Wörterb. i 280 ffGoogle Scholar.

^{page 110 note 6} On fravaši- see now Barr, K., Festshrift til L. L. Hammerich, 1952, p. 35Google Scholar; earlier Zoroast. Problems 109.

^{page 110 note 7} Written with -yh for yk, see GrBd. 172.14, Mēnōk ī Xrat 43.7 and the Pahlavī translation of Av. vāe∂θman-. Pāzand has gurdī.

^{page 110 note 8} Przyluski, J., La Légende de l'empereur Açoka, p. 401Google Scholar, now in Taishō Issailcyō

^{page 111 note 1} kai used in the Mahāvyutpatti 6073 for varma-saṃnāha-.

^{page 111 note 2} Mid.Pers. hašā- in Zor.Pahl. hš grt, hš grt *hašākart, Mid.Pers. hašāgerd, see Bartholomae, C., Zum altiran. Wörterb. 38Google Scholar; Benveniste, E., Trans. Phil. Soc. 1945, 69–70Google Scholar. Armen. ašakert, NPers. šāgird.

^{page 111 note 3} Vajr. 4 b 2. Wrong readings ad loc. by Konow, Sten and Hoernle, A. F. R. are given in Manuscript Remains of Buddhist Literature, p. 241Google Scholar. The aksara ṭhy was not recognized but it is quits clear.

^{page 111 note 4} See “Asica”, Trans. Phil. Soc, 1945, p. 4; Morgenstierne, G., Etym. Voc. Pashto, s.v., IIFL. i. 394Google Scholar; Walde-Pokorny, , Vergl. Wörterb. i 593Google Scholar.

^{page 111 note 5} Henning, W. B., in Ghilain, A., Index des termes en moyen-iranien, p. 377Google Scholar.

^{page 111 note 6} 2 Chron. 13.20, in Girk Mnac ordac , ed. Grigor Xalateanc , 1899, p. 10. No decisive etymology is offered in H. Adjarian, Armen. Etym. Dictionary.

^{page 111 note 7} The -g- kept, as -k-and -g- kept after pat- in, for example, patkan-, patkar-, and patgam.

^{page 111 note 8} Similar -i- umlaut before two consonants in paltcīmphākä “chedaka” in the Suvarnabhāsa 27 v 5, Khot. Texts i 236.

^{page 112 note 1} In a Hedin military document 21.3 (now with the printer) occurs pūṃna hu śtä be sa ttyāṃ va ādarä yanīrau “arrows, spears, be sa, take care of them”. In the context be sa- may best be taken as “shield”. That gives be sa- from *vartsa-, as we have bei'sa-, ābei sa- “whirlpool” from the verb vart- “turn”. In be sa- “shield” we should then have a double suffix from var- “protect”: -tsa- from Indo-Eur. -tes-: -ts- with -o-. The suffix -tes- is well known in OPers. rautah- “river”, Av. θraotah-, Old. Ind. srótas-.

^{page 112 note 2} BSOAS. 14.422. Both b- and v- can represent both b- and v- of Old Iran. The third replacement by g- is also known in Khotanese. To the note on Kharoṣṭhī vaka, vaga, add the baga of the coin legend baga-vharna- in Rapson's, E. J. paper, JRAS. 1905, pp. 790–1Google Scholar.

^{page 112 note 3} Morgenstierne, G., NTS. 12.268Google Scholar.

^{page 112 note 4} But before and after u we have ur- in urz “finger”, urs “stallion” from *ṷršan-, urdug “upright” from *ṛdṷua- (contra C. Bartholomae, Mittelir. Mund. 6.7).

^{page 112 note 5} -t- comes from -θ- in fätän “wide”, Av. paθana-, and in äzmäntun “stir up”, from uz-manθ-, as maṃth- in Khotanese, Old. Ind. manth-.

^{page 112 note 6} Miller, Vs., Oset. ëtiudy i 30–4, no. 6Google Scholar.

^{page 112 note 7} Grégoire, Mélanges H., A propos de “V∂r∂θraγna”, 223–6Google Scholar.

^{page 113 note 1} 5.109.2.

^{page 113 note 2} In Pamiatniki 2.111.32 ämbal jezdon is rendered by tovarišč blagorodnyj. I hope to write on its connections with Av. vazd- later.

^{page 113 note 3} Miller, Vs., GIP., Ossetisch, pp. 4 ff.Google Scholar; Abaev, V. I., Oset. Jazyk 284Google Scholar.

^{page 113 note 4} See Dumézil, G., Les Légendes sur les Nartes, p. 115Google Scholar.

^{page 113 note 5} Abkhaz a-g°aban, quoted by Abaev, V. I., Oset. Jazyk i 314Google Scholar.

^{page 113 note 6} Iron differs: cydystäm, farstam.

^{page 113 note 7} Abaev, V. I., Oset. Jazyk i 269Google Scholar.

^{page 113 note 8} Soska is gen. sing, in -a from the name in Oss. Sosäg, the father's name.

^{page 114 note 1} Batradz and Soslan are traced to old myths of gods by Dumézil, G., Les Légendes sur les Nartes 178 ffGoogle Scholar.

^{page 114 note 2} Pamiatniki 2.5.

^{page 114 note 3} Similarly in Ubyx närt -λäpq-na zä-laxwa (with Čerkes λäpq) “one family of Narts” occurring in Dumézil, G., La Langue des Oubyhhs, p. 136Google Scholar.

^{page 114 note 4} Pamiatniki2.25: narti guppurgintä Bestaui bärzondmä bärgä ärämburdäncä and semuncä kafuncä narti guppurgintä.

^{page 114 note 5} See I. Gershevitch, BSOAS. 14.493, who traces the word also in Oss. Digor igurun, Iron gŭryn “be born”.

^{page 114 note 6} See for example Hittite -uar- beside gen.sing. -unas.

^{page 114 note 7} See Abaev, V. I., Iz oset. ëposa, p. 8Google Scholar, where Greek ππ was used.

^{page 114 note 8} Oss. tuppur “tomb, kurgan” can be explained as from *tau-θuar- to the Indo-Eur. teu- base found also in Lat. tumulus.

^{page 114 note 9} Trans. Phil. Soc. 1952, p. 63.

^{page 115 note 1} Rather than *gavaθra-, as in Zoroast. Problems, p. 83, since Old Ind. gotrá- “family” is probably the same word, perhaps originally distinct from gotrá- “cattle-stall”. The Sogd. ywš is in Henning, W. B., Sogdica 17Google Scholar; 20.

^{page 115 note 2} Bartholomae, C., Zum altiran. Wörterbuch 234Google Scholar; Junker, H., Ein Bruchstück der Āfrīnaghān ī Gāhānbār 21 (top)Google Scholar; Dhabhar, B. N., Zand-i Khurtak Avistāk, p. 155Google Scholar, line 3, with note p. 376.

^{page 115 note 3} Andreas-Barr, , Iranische Dialektaufzeichnungen, p. 307Google Scholar.

^{page 115 note 4} -rn- is represented by long vowel followed by -l-in dālman, Avestan zannumainiš (see Zoroast. Problems 230), in the name of Erzerum in Arabic script Qālīqalā, the city of Karin (qālī < karni), and in Gēlān if this is from *varnya-. It may be due to the presence of -i-, -y-, that -l- not -r- resulted. In Ossetic -l- occurs in fäl- from pari- and other words.

^{page 115 note 5} See Zoroast. Problems, p. 89, for Mid.Pers., Mid.Parth. zhg “child”. NPers. zaq “young animal”, Armen. loan-word dzag.

^{page 115 note 6} Cited from a Khotanese text BSOAS. 14.427.

^{page 115 note 7} Lith. kuriù, kùrti “to build” is cited in Walde-Pokorny, , Vergl. Wörterb. i 518Google Scholar under k^wer-, to which Old Ind. kṛṇnotiis also traced. Possibly kur-in hākurdna- and igurun is a secondary form from *kar-ṷ-, as we find tar-ṷ-. But in Avestan the form expected would have been *kaurv-. A base kur- seems preferable.

^{page 116 note 1} Iran, āvan- is found as -aim, -ôn also in the loan-word paštaun, paštôn “service” from *parištāvan-. The -āvansurvives in vtauan, vteuan “bowshot”, Mid.Parth. wt'wny in the Hajjiabad inscription, ed. Nyberg, H. S., Øst og Vest, p. 66Google Scholar.

^{page 116 note 2} Bigveda 4.16.7 vṛtrárṃ, vavrvāiṃsaṃ.

^{page 116 note 3} Revue de l'Histoire des Religions 117 (1938).

^{page 116 note 4} In E. Benveniste and L. Renou, Vṛtra et Vṛθfragna.

^{page 116 note 5} Rev. Hist. Rel. 133, 1–18, La promotion de Bahram.

^{page 116 note 6} Macdonell, A. A., Vedic Mythology, 158 ffGoogle Scholar., and Hillebrandt, A., Vedische Mythologie iii 260 ffGoogle Scholar. treated of Vṛtra, Vala, and the fight with Indra. Brown, W. N., Creation Myth of the Rig Veda (JAOS. 1942, 85 ffGoogle Scholar.) uses neither E. Benveniste and L. Renou's nor G. Dumézil's studies. Eliade, M., Rev. Hist. Bel. 134 (1948), p. 12Google Scholar, explained vṛrtra- as a “binder”, but var- does not mean “to bind”. See also Lommel, H., Der arische Kriegsgott, 46 ffGoogle Scholar.