^{page 478 note 1} The gist of this article was presented as a paper to the Twenty-third International Congress of Orientalists at Cambridge, in August 1954.

^{page 478 note 2} Cf. the month-name Biy-Askergiamong the Mountain Tatars who live to the West of Digoria, between the River Urux and the Elburz, Miller, Oc, 9m.in, 9.

^{page 479 note 1} This cup rises of its own account to the lips of the Nart hero who relates his exploits truthfully, but remains on the table if lies are told, cf. Nartl Kadidzltse229 = V. Dynnik, Hapmcnue CKaaanua 298, v. also Dumezil, Legendes sur les Nartes, 136 f. Also Miller, Oc. 9m. i, 161, n. 20, connected the first element of Uacillawith that of Uasiamongmas he spells this name, which only occurs in Digor; the pronunciation is Wa∫ia[, secondary development of Wat222B;ia°].

^{page 480 note 1} Beside fid ‘meat’ and fid‘father’, Miller-Freiman's Dictionary has the entry fid ‘millstone’. This interesting word may be from *piti- ‘crushing, crushing instrument’, and compare with Ps. and Orm. (Waz.) pal ‘millstone’ (cf. Morgenstierne, NTS v, 25, who assumes Indian origin) < *paiti-(hochstufe as in Ved. heti, cf. Wackernagel-Debrunner, Ai. Or. n2, p. 631 c notes; cf. Ps. wala<vaeti-, and, for the ending, NTS xn, 93). This pi-may be the unextended base of Av. pisant-,Skt. pindsti, etc., which can also be recognized in Vend.13, 40: vehrka scaθwa vshrka pθiBwa vvhrka snaezana ‘the slavering wolves are to be killed, smashed, crushed’; this explanation of poiBwa-is to be preferred to Bartholomae's (Air. Wb.1893 on 899) or Geldner's (Studien53). On the other hand, under the entry farktwo homonyms have been thrown together in the Dictionary. Farkmeaning ‘piece, splinter, small plank’ (cf. also II, 810,‘in smithereens’) is connected with NPers. para,while fark ‘blade of a mill-wheel, paddle' belongs to NPers. paru.

^{page 480 note 2} Thus also in the etymologically obscure Digor verb xincun‘to treat hospitably, entertain’, whose past stem is xinst-,while its Iron equivalent xtnein,which in addition means ‘to count’, has as past stem xixt-, xihd-.

^{page 480 note 3} Considerable doubt, however, attaches to Benveniste's identification of Chr. 'wxt with B. 'wyrt.Loss of final rof a present stem before the ending -tof the 3 Sg. occurs in Chr. -d't after a long vowel, but is not attested after short vowels. Also the difference in the construction, which Benveniste has noted, goes against the identification.

^{page 481 note 1} cf.JBAS 1946,181, on 115

^{page 481 note 2} v. Henning, BSOAS xn, 309 n.

^{page 481 note 3} Not from *xsi-,as shown by Digor fexsuyun, fexsud,which preclude Miller's (IF21, 328) and Morgenstierne's (IIFLn, 209, s.v. fsii)connexion of ixsiinwith Skt. ksiydte.

^{page 481 note 4} In common with other Eastern Iranian languages Ossetic has lost Olr. before n.Miller did not mention this development in his grammar, but neither Hiibschmann (Arm. Gr.20) nor Bartholomae (Mitteliranische Mtindartenm, 29) hesitated to derive mxsin,of which sefsinis apparently a side-form, from Olr. *x0B38;aidm-.In addition, Ossetic provides representatives of the usual two test words, Olr. *arani- ‘cubit', and hapani- ‘co-wife ’. The former has been recognized by H. W. Bailey (personal communication) in Digor cieng-serinse,Iron selm-serin, arm-serin ‘cubit’. The latter survives, in my opinion, in binontse‘family, members of a household, relatives, wife ’, which lends itself to the analysis bin(< hapani-) +suffix -an(a)-(cf.binoinag,usually ‘member of a household’, but in Iron Admmon Sfeeldlstad,Ordzonikidze, 1941, 827, ‘wife’) + plural suffix -tie.The original meaning will have been ‘co-wives and their children’, cf. E. H. Minns, Scythians and Greeks,84, on the position of women among the Scyths. In Sogd. pn'nc ’co-wife’ (v. W. B. Henning, Sogdica17 f.) the same suffix -an(a)-has been adapted to the Sogdian personal feminine suffix -anc; this view is preferable to the one taken in GMS§ 1046. The Ossetic palatalization of ain -a(B)ni-compares with the treatment of -anyain (suh)zeerinie‘gold’ from zaranya-(Miller, Oc. 9m.m, 138).

^{page 481 note 5} From my Digor friend I have ristsen use buxsun(or buxsagie dsen) ’I cannot endure pain’, nx ibssl buxsun ‘I miss him, long for him ’ (lit. ‘am impatient for him ’). Oss. buxs-probably continues an Olr. inchoative *buxsa-,the past participle of which survives in Sogd. pwyt'rmyb ‘patient, cf. GMS§469, 1105.

^{page 482 note 1} There three wacilla-angels are referred to as sertse TbauaciUay,viz. uacillasfrom the mountain Tbau; cf. the Song of Acsemsez,line 166 (in Abayev's edition, He ocemunckoso anoca,pp. 54 ff.), where White Yelia ( = the individual spirit Uacilla) is said to reside on mount Tubau.

^{page 482 note 2} An early form of the plural of uac ‘spirit’ (uac ‘news’ has the plur. uactse)survives perhaps in the asseveration uastsen(and uasdeen? v. Diet.), which, as it seems to be used like Engl. ‘by Jove ’, may be a Dat. plur. meaning ‘by the Spirits’. The Dictionary suggests that uastsmcontains the postposition stmn; this is unlikely, for the noun which precedes stsenis always in the Genitive, as the examples collected by G. Morrison, Ricerche Linguistiche11, 79 n., clearly show. Cf. uassenand uadzasnbelow, p. 488, n. 2.

^{page 483 note 1} v. Dynnik, Hapmcvue CriaaaHUa348 ff. = Narii Kadzdiltse261 if

^{page 483 note 2} Dausegis usually translated as ‘seraph, protective genius’. In Iron Adsemon Sfseldistad 275, n. 75, Digor idaugutseis explained by ‘izsedtse ’. Miller remarked, Oc. 9m. n, 240, that to his knowledge neither the individual Uastlrdzi, nor any of the other gods were ever called zasdor dauseg;however, in the Song of AcsemsezUasgergi is referred to as an izaed(line 128), and a group of gods consisting of Uasgergi, White Yelia, Good Nikkola, iEfsati, Fselvaera, and others, is more than once collectively described as izsedtseand idaugutse.–Beside the collectively appearing daudzitsethere exists an individual genius by the name of Dauseg,who was believed to avert the plague. To him a sanctuary was dedicated near Sturdigor (Miller, Oc. 9m.n, 261, 283 f.); cf. Barduagin the next note.

^{page 483 note 3} Beside dauseg(Dig. idauseg)we find duagin an Iron text, Hamsmuukum, 7¹², cf. rini Barduag ‘god of ( = averting ?)illness’ (Dictionary). The development is as in Dig. bauser: Iron buar ‘body’, etc., cf. Miller, Ossetisch17

^{page 483 note 4} Fervista Satana Kürdalsegonmse. Nart semxxrd, semnusezt udtsti daudzitimas, –aercidis Kurdolsegon Satanayt xundmse ‘Satana sent for Kurdalsegon. The Nart were on convivial terms with the dausegs,and K. came at Satana's invitation’, Narti Kadzdiitse64. The name of this smith (Oas. kurd)contains Iseg ‘man’ according to Miller, Oc. 9m. i, 118; on the other hand Abayev, Oc. H3. u .i, 71, starts from the Digor variant Kurd-Ala-Vsergon(Dictionary, 703 f.), which he analyses as ‘the Alanic smith Wsergon ’, Weergoncorresponding in his opinion to Vulcanus.Kurdategon is described by Miller, loe. cit., as living in Heaven or in the realm of the dead, where he forges horseshoes, etc., for the horses of deceased persons.

^{page 483 note 5} The problem of OPers. artavan-was last discussed by J. Duchesne-Giiillemin, ZoroaMre,130 f., who did not consider the Avestan evidence.

^{page 484 note 1} I used to consider, but have abandoned as semantically unsatisfactory, a connexion of idausegwith Sogd. wyt'w- ‘to endure, persevere’. This verbal stem was somewhat misleadingly quoted in QMS§ 216. The passage there referred to, M133, 83, is the one from which Henning, Sogdica,p. 3, had quoted wyt'wp'zny'h.It is only in this compound that wyt'w-occurs in Man. Sogd.; in Buddh. Sogd. we have wyfum'k‘persevering’ in PI, 1130, with which Benveniste rightly compared ptw- ‘to endure’; cf. Chr. pt'wp'zny' ‘Duldersinn’ in ST n, which approximates the meaning of Man. wyt'wp'zny'h.–I take this opportunity to point out that M. Benveniste was the first to notice, in JAs.1939, 275 f., that B. zn'khin VJmeans ‘body’ (cf. TPS1945, 138, n. 4).

^{page 484 note 2} Olr. rta-gives Oss. ard‘oath’ (cf. Bailey apudE. B. Ceadel, Literatures of the East101), which becomes serd-when unstressed, cf. serdxord ‘confederate‘. For the -(a)Jc(a)-suffix added to an -auan-stem Sogd. cn'wq ’thirsty’ can perhaps be compared.

^{page 484 note 3} The form aiiiamdtnthere quoted by Miller is not to be found in the Dictionary, and seems to be due to confusion with aeuuaendin’ to trust

^{page 484 note 4} Trace of a replacement of the pseudo preverb ser-by the preverb (w) i-in the denominative verb *ser-? (Nowadays, according to the Dictionary, Usendtntakes the preverbs a-, ba, s-,and fae-). Such an explanation cannot, of course, account for the i- oiidauseg,since no denominative of this word is attested. At a stretch one may invoke a possible analogical (alliterative) influence of i-zsedtse,with whom the i-dcmgutseare constantly associated in the formula izsedtte asma idaugutse,which for practical purposes is a dvandvaof the mad-zemm-fldtype (cf. Miller, Ossetisch96

^{page 484 note 5} The preverb originally may have belonged to the verb ksenun ‘facere’, with which the corresponding transitive expressions are formed: ni-hmla-kbdta ‘he turned (somebody) into an idiot’, is-haszdwj-kodta ‘he made (someone) rich, he enriched ’, mr-darh-kodta ‘he made(something) long, he lengthened. The pseudo intransitive past participles abstracted from such periphrases were treated like real ones: the 1st Sing. Pret. ishiezdug-dasn ‘I became rich’ was formed on ishsezdug ‘become rich’, as mrcud-teen(<*sercud-dsen) ‘I came, lit. ventus sum ’ on the past partc. sercvd-(Olr. *0050;cyuta-),or MPers. 'md hym,Sogd. “ytymon the past participles 'md, ”yt-(Olr. ag{m)ata-)respectively. In the 3rd Sing, of the Preterit the copula is usually implied (Sogd. yt,MPers., NPers. amad,etc.), but in Ossetic beside Iron sercld ‘he came‘, we more commonly find; sercld-i(s),in Digor always xrcud-sey ‘ventus est’;correspondingly Iron both s-qsezdlgand s-qeezdig-i(s)are possible, in Digor only is-haezdug-eey.These interesting Ossetic periphrastic denominatives, as one may call them, have not yet received the systematic treatment they deserve, which would take into account their complete inflection, and the shades of meaning or aspect conferred by various preverbs; cf. provisionally v. Stackelberg, Beitrdge zur Syntax des Ossetischen 71 f., Miller, Ossetisch§ 87, 2, Abayev apudKasayev, OcemuncKOpyccKuu cjtoeapb492.

^{page 485 note 1} Discounting mssldztg,Dig. mvMzug,‘ant’ < *marui-ca-ka-(cf. Sogd. zm'wrc, OMS§ 247, v. also Bailey, BSOASXIII, 664), where apparently uwas transferred to the suffix by metathesis. The first uof the Digor form will then be secondary, cf. Miller, Ossetisch§ 7, 3, note.

^{page 488 note 1} The loss of the privative particle is known to my Digor friend also in donug(Dictionary only asdonug, idonug,beside Iron doini)‘thirsty’ (lit. ‘without water’), and xormog(Dictionary only sexxormag) ‘hungry’ (lit. ’without food’). Initial seis also lost in csegsey‘in truth’, ii, 9235<

^{page 488 note 2} On p. 1250 the Dictionary has the following enigmatic entry: uadzsen(= uasxn, uasxsen, somieen)adv. ‘not for nothing, if only (< I swear ?)’. Unfortunately no examples are given. Uasxsenand somisen ought to mean ‘for (= upon ?) oath’. Hence uassenand uadzsenare conceivably asseverations in the singular, both inaccurately abstracted from the plural uastsen‘by the Spirits’ (on which v. above, p. 482, n. 2): uassenwith analogical sinstead of expected c(cf. uas,‘sermon’ (A) (a) above), uadzsen‘by the (= my) spirit‘ with a shift to the other connotation of the word ‘spirit’.

^{page 488 note 3} Two further words have been tentatively connected with uac,no doubt wrongly. In Oc. m. J,119, 161, Miller compared the name Uasxo,and his Dictionary has a reference to uadz, uas, under the entry uaz. Uasxowas the name of a sanctuary at Kani. The name of the god to whom the sanctuary was dedicated appears in the formula Uasxo dee uazseg ‘U. is (ormay U. be) your guest’. Miller analysed the name as Uas-xo,but who, or what, is xo? In the Kabard epos oaths are sworn by Uasxo-kan,and in one of the Kabard Nart tales (Eapmu, KadapduHcmiu anoc,Moscow, 1951, p. 334) the god Uasxois himself the curser. One might therefore suppose that the name is connected with the Oss. noun nasxse ‘oath’. However, M. E. Talpa, KaSapdunemiii (fioMKAop, Moscow-Leningrad, 1936, p. 638, offered a Kabard etymology of Uasxo: uafa ‘sky’ -)- sxo ‘blue’. As to uaz,apart from the difficulty of explaining the z, the meaning ‘prayer, admonition’ which the Dictionary assigns to it, scarcely suits the one context in which it is quoted, , 116 29; there, a being who first appears as a bear, later turns out to be a religious man: ye ba Iseg usexsen adtmy sema medavarsey sendeemse uaz Isevardta ‘he was such a man that he gave uazfrom inside the room towards the outside’. At a guess one might say that uaz dsedtunhere means ‘to radiate light’.