^{page 495 note 2} This is, of course, not an exactly accurate rendering of the original Avesta text, for a critical free rendering of which see SBE. xxxi, pp. 240–244, now, however, somewhat antiquated.

^{page 495 note 3} Daēvyō (cf. Indian devyás) = ‘demonesses’ is not properly rendered here. The motive to this error, ‘friend,’ is not so clear, yet we must notice that the term -aēevyō- of daēvyo- may have called attention to ‘av,’ the root of ‘friendly approach’, whether because ‘y’ of ‘-vyō-’ was read as often ‘v,’ or simply because an -avyō- was seen; or indeed, again, a possible adjective in -ya, ‘the daēva- devoted one’ (?); but the important word ‘demoness’ was wholly lost, as šedayā-yazak does not reproduce it. Daēvya = Ind. daívya = ‘the divine one’ would be a singular, and not so forcible a reading. Nēr. follows the mistake of his original. Does the Gāthic Pahlavi make worse errors, if it be an error? For others see below.

^{page 495 note 4} The personified ‘obedient hearing,’ the angel of loyal religious acquiescence. But like all these impressive names founded upon the grander abstracts as seen in the Gāthas, in the later degenerated period of the Hōm Yašt, they had largely lost their original significance. See where xšathravairya comes to mean merely ‘the Angel of metals,’ Ār(a)maītī merely ‘the earth,’ etc. x = kh here.

^{page 495 note 5} The word evidently was used to express an entrance upon domicile. The imperative conjunctive mood tense of the original is not expressed. Nēr. follows with this negligence.

^{page 495 note 6} The imperative conjunctive is not expressed here; see also Nēr.

^{page 496 note 1} This ‘production’ (with which I do not agree; see SBE. xxxi) casts light upon that other word rendered by the Pahl. ‘born of aša,’ but which should be more critically read ‘bearing aša’; see below at Y. X, 41 (14). We have only Indian ṛtajā, etc.; ṛtāvan does not occur in composition in old Indian.

^{page 496 note 2} I regard this quasi-adverbial translation for frateremčit as being strictly critical. The original hardly meant ‘I praise thy first pressure.’ I followed this in SBE. xxxi.

^{page 496 note 3} Free grammatical form.

^{page 496 note 4} So, freely critical; the 1st person was not deliberately mistaken, as it is correctly rendered by the Pahl. translators times without number.

^{page 496 note 5} Here evidently the Priest was supposed to crush the Hōm leaves, or stems, in the mortar to extract the juice.

^{page 496 note 6} Again most critically correct, regarding the accusative of the original as quasi-adverbial. Hardly ‘I praise thy final pressure.’

^{page 496 note 7} Possibly ‘I will strike thee’; but the barā seems to express the ‘ni’ of nignē; see also Nēr.'s mhanmi, not a future.

^{page 496 note 8} To extract the juice.

^{page 497 note 1} This gloss refers, I suppose, to the previous verses beginning with ‘I praise.’ Nēr. omits it (this gloss). I suppose it was intended to introduce the praising of the objects next introduced after the Hōm itself, which was directly praised in 4 and 5. We might think that some preparatory ‘praise’ was referred to ‘praising beforehand.’

^{page 497 note 2} So with the text -iša, as a 2nd singular, but as optative, which is not expressed in the translation; the second personal form is alone recognized.

^{page 497 note 3} See the texts for the explanation of this curious error, which recurs at Y. X, 40, Vare- recalled var = ‘to choose,’ ‘to be devoted to’; the -zy- may have recalled -zi-, and -aṅh- an ahu, aṅhu= ‘lord.’ Nēr. follows it. Surely no error of the Gāthic Pahlavi was worse than this.

^{page 497 note 4} The translator seems to have had in his mind the stimulating effect of the extract of the Hōm plant as inducing physical activity. I preferred to take the čar of čarānem in its more immediate sense of ‘movement’ rather than in that of ‘production,’ i.e. as ‘swiftly spreading.’ The quick effect of the drink may possibly have been meant in the original, if not here.

^{page 497 note 5} I nave remarked in writing upon the texts the singular circumstance that Mazdāo is translated masīh there and by A. (DJ.) by ‘mas’ = ‘the great one,’ seeing the ‘maz’ of mazdāo as = ‘great,’ but the gloss, which may, however, have been of later origin, diminishes the force of the point.

^{page 497 note 6} As regards the dimensions of the plant, I am unable to say whether the haoma (soma) bush ever attained to a considerable size. See, however, Nēr.'s mahattarodayaṁ.

^{page 497 note 7} See Nēr.'s pradatto' si ff.

^{page 497 note 8} ‘Many’ is, of course, erroneous for viš-, though Nēr. follows it.

^{page 498 note 1} Possibly ‘sparkling one.’

^{page 498 note 2} Notice that the Parsi idea of ‘increase’ is always associated with moral excellence. The ascetic principle was carefully avoided.

^{page 498 note 3} This is, of course, a rubric.

^{page 498 note 4} I would now avoid dividing ahū + kēnišn, ‘doing despite to life.’ I now recognize the Pahlavi word as ahō, meaning ‘vice,’ etc., so, entirely of itself and without any expression of the ‘malice’ in kēnešn-, we should have a common denominative form from ahō(k), ahō kēnišnīh; the ‘k’ of ahōk would be the frequent quasi-phonetic letter used, as in vohūk for vohū in the citation at Y. IX, 3.

^{page 498 note 5} Mistaking frākeresta for frākereta; see Nēr. also. We could hardly render the Pahlavi: ‘and they would put forth the rest.’

^{page 498 note 6} Nēr.'s somewhat expressive upaçlešyanti may have been induced by the preposition upā in upāzaiti, which may have recalled upaēta. Or was the barā vebedūnāñd, which at the first glance looks so vapid, intended to put force into the foregoing avākēnend, so inducing Nēr.'s rather strong expression?

^{page 498 note 7} Notice that the barā of barā vebedūnāñd is here defined by bāstaǹ, Nēr. nityam, as expressing continued action, and not as meaning ‘they would do away with.’

^{page 499 note 1} This gloss seems extremely natural; and yet it is just possible that the outward shape of the durūstīh, ‘health,’ suggested the word dast, ‘hand,’ which with extra strokes would at least suggest durūstīh.

^{page 499 note 2} As to the distinctions here made, mān, or mihan, would be naturally the more general term. Nēr. pushes the distinction still further with his prāsāde, which might mean even ‘in the palace,’ or ‘temple.’

^{page 499 note 3} The Pahlavi translators omit Zend ‘zi,’ not forgotten, however, by Nēr.; see his yataḥ; the zag Ī, of course, hardly takes its place.

^{page 499 note 4} It is difficult to see the motive for this gloss. It looks like an insistance upon the reading of a text; some variant may have troubled the transcriber. Nēr. knew nothing of it.

^{page 499 note 5} I am not so sure that hūravāxmaǹ, pointing to an origin in vraj for urvāsmana, is critically correct. May not a vrāsman = ‘brahmán,’ lurk here? If such should be the case, we should render ‘Aša, the holy, the brahman-like.’

^{page 499 note 6} Nēr. laghvī.

^{page 499 note 7} Nēr. adds an ityarthah, as if to note that girān was used in some special sense; his *b'āritā, however, conveys the idea of ‘severity.’

^{page 499 note 8} The singular for plural in the verb should hardly be noticed as other than a freedom. The singular anšutā should be taken as either a collective, or as an instance of mere carelessness.

^{page 499 note 9} So, much better than merely ‘accept’; see the Avesta, yet see also Nēr.'s samgṛhṇanti; and read as alternative ‘accept.’ Again, we have a singular for plural with a singular subject. Aside from the original, we should render, ‘forthwith let that which is their bodies (i.e. persons) meet (or ‘accept’) Hōm as healing.’ Another solution would be, ‘forthwith, O Hōm, do thou meet their bodies as healing’ (i.e. healer) (-añd used like -yēn at times for the 2nd singular imperative), but see the original.

^{page 500 note 1} Nēr.'s ‘yaiḥ’ supplies these needed words.

^{page 500 note 2} Nēr. does not render bēš. The Pahlavi translator, followed by Nēr., here distinctly expresses the matured meaning of the original, but only in one item. Literally we might render the original: ‘Haoma, grant to me of the fiendsmiting powers, by which thou art a fiend-smiter.’ But ‘fiend-smiting’ evidently came to mean merely ‘victory’; so the Pahlavi. On the other hand, the translator renders ‘verethra’ with bēš =‘hostility,’ not ‘victory.’ It is often hard to say which we should prefer, ‘fiend-smiting’ or ‘victorious smiting.’ We must note that vṛtrá at times means ‘victory,’ etc., in the R.V., and even ‘defensive victory,’ from var, ‘to enclose.’ But cf. R.V., 1, 91, 5, ‘Tváṁ somá' si sátpatis tvám rājo 'tá vṛtrahā,’ etc. Of course, the original * meaning of vṛtrá (verethra) was ‘the enclosing demon serpent,’ who ‘enclosed’ the clouds and kept off the rains.

^{page 500 note 3} A clearly alternative and improved translation; ‘min lak’ is almost senseless. I would even accept it as a pure genitive; see Nēr., and take mekadlūnam-ī in the sense of vīsāi = ‘I will become,’ if it were possible; see Nēr., however, who has ‘forth I meet thy friendship and thy praise.’

^{page 500 note 4} I have little doubt that Saptīr min aharāyīh means ‘better than A.,’ as in the Semitic languages; see also Nēr. Otherwise we should have ‘to me good has been given from A.’ ‘I have given from A.’ would be very flat.

^{page 500 note 5} So šapir . . . aēϒ means ‘better than.’

^{page 501 note 1} I fear we must report the Pahlavi translator as referring the word rendered ‘well-skilled’ to Haoma, and not to the Deity, with Nēr., who has a nominative form; that is to say, unless we can understand a ‘pavan’ before baϒ = ‘Swift and wise art thou fashioned by God, (He being) well-skilled as regards the spiritual interest, the heavenly world.’

^{page 501 note 2} See Nēr.'s paralokatayā. Even ‘immortality’ was thought to be given through drinking the Hōm; so said one of the commentators; see Y. IX, 5.

^{page 501 note 3} Hardly, ‘Swift, O thou and wise (as) a god, thou hast given. . . . ’

^{page 501 note 4} The remarks upon 26 apply here again, ‘by God art thou created’ would require a ‘pavan’ understood. I should be pleased indeed to hold that ‘wellskilled’ here applied to the Deity as in the original and as in Nēr.

^{page 501 note 5} This is erroneous, of course, as applied to the Amešaspentas.

^{page 501 note 6} The ‘sign’ here should be understood as indicating ‘intelligence’ or ‘instinct.’ In the original I cannot see this simple sense of ‘sign.’ See SBE. xxxi, free and critical. Here the birds were regarded as conveying ‘omens’ with regard to the Hōm, as the word hū-merv means ‘well-omened.’ With more commonplace we may understand that the birds by hovering near pointed out where the Hōm was growing.

^{page 501 note 7} I do not see why ‘apart, apart’ is not a very good rendering for ‘facing apart in different ways.’ I was inclined to render ‘of different kinds’ here in the Pahlavi; but see the original in SBE. xxxi.

^{page 501 note 8} So perhaps, better than ‘beyond the eagles,’ as I formerly rendered the original. Read in my translation there for ‘beyond the eagles,’ ‘where distant eagles circle,’ so with critical freedom for the literal ‘to the mounts having the above (the high-soaring or ‘the encircling’) eagles,’ upairi - saena; though I correct myself with diffidence. Yet see also staẹrō-sāra as a compositum.

^{page 502 note 1} As seen from SBE. xxxi, p. 242, I compare Indian párvata, for paurvatāhva = ‘mountains,’ and not párus, ‘oint of plants.’ But then párus and párvata originate from the same idea, that of ‘swelling’; the Pahlavi translator might so indicate. Nēr. certainly errs here with his pavitra-nikarešu. He seems to have read pāk instead of -tāk, or ‘pūr’ in the sense of pavitra.

^{page 502 note 2} So more literally; but Nēr.'s gaulya may give us the better meaning ‘pleasant’; hardly, however, ‘sweet.’

^{page 502 note 3} ‘Now thou growest,’ twice repeated in SBE. xxxi, p. 242, 12, was inserted as stated on p. 231, “to point the sense and round the rhythm.”

^{page 502 note 4} See the texts and read yēm haē, having an eye upon Y. XXX, 3. Otherwise reading yōm aē, ‘This thy healing is this day . . . .’ Or with yōmīhā, ‘days’ or ‘daily.’ Nēr. favours yēm haē with his arogayukto'si, -yukta' pointing to yēm as referred to yamá = ‘paired’. But what was the motive of the idea ‘joined,’ ‘mated,’ ‘doubled’? I cna only suggest that the idea of flowing may have suggested fusion, or ‘blending,’ and so ‘uniting.’ Of course, we see the ‘doubled’ syllable -rīri- in irīriθare, which might have suggested the idea of ‘pairing’ to the earlier ‘searcher.’

^{page 502 note 5} Patmān = measure; Nēr. pramāṇena is, I think, far too pallid of mayābyō, for which I compare, of course, Vedic māyā = ‘superhuman wisdom’ or ‘cunning,’ i.e. ‘acuteness.’

^{page 503 note 1} I should like to put the sentence in the conjunctive imperative of deprecation to correspond with the original and with Nēr.'s 2nd singular imperative. But the terms do not warrant it. It would not be at all fair to take the bāt (bād), which is a mere variant for asrūnast, as if it were a 3rd singular conjunctive of būdan. B. (D.) omits it. Vaepaya is, of course, totally mistaken here: something in the outward form, probably the syllable ‘pa’ in paya, suggested bāt to someone in the long line of translators. Or does not this bāt of the Pahlavi translators supply us with the proper reading of the original ‘b’ for ‘p,’ so giving us some form of bi = bī, vibāya, ‘Cause us to be in terror (?).’ Vaẹpaya could also mean ‘make us tremble,’ or ‘terrify (us)’: ‘Drive from us through terror,’ or ‘cause them to cast from us.’ Nēr. renders a vaẹpaya well enough with his parivartaya.

^{page 503 note 1} Here the translator supplies the verb as ‘tell me,’ ‘yemalelūn am.’ I did not hesitate in SBE. xxxi, to regard vaepaya as the word to be understood.

^{page 503 note 3} Not, as we might at first sight suppose, ‘He declares the thing which is revolting to be commendable’ (lit. ‘goodness,’ xūbīh for aūvaš). Nēr. knew nothing of xūbīh; and we should beware of smoothing out matters in an ancient commentary like this. See the original, which is ignorant of such an idea.

^{page 503 note 4} Although vebedūnāñd is a 3rd personal, so corresponding to kerenaoiti, yet I think in view of Nēr.'s karoši that we have here a use of the form in -āñd such as when it expresses (though, perhaps, in a roundabout way) the 2nd singular imperative. As forms in -āñd can express the 2nd singular imperative, so this may express it. Read as alternative, ‘thou who wouldest make the poor man's (feeling),’ etc. Notice vebedūnyēn in this sense at 37.

^{page 503 note 6} Have we -yēn for the 2nd singular imperative in conjunctive sense here also? ‘thou, who wouldst make.’ But see the 3rd personal of the original.

^{page 504 note 1} As to xūrsandīh for usnãm, I ventured to differ. That Nēr. omits xūrsandīh, has no force, since he has nothing in its place. It is quite possible that xūrsandīh, was suggested by the appearance of usnãm while the latter stood in the Pahlavi character. With a very simple sign before it, it would be ‘xūrsan-.’ ‘Satisfaction,’ however, renders ‘culmination’ well, if only accidentally.

^{page 504 note 2} See the note above upon 35; the 2nd singular is undoubted here; see kerenūiši; though I am of the opinion that this use for a 2nd singular came from an indefinite sense given to the 3rd plural, ‘they would make’ = ‘persons should make,’ and so used for ‘“you” make’ in the indefinite sense. ‘You should make,’ as equalling ‘one should make,’ is near to ‘make thou.’

^{page 504 note 3} Nēr. omits this repetition, showing that it very probably did not exist in the MS. which he used.

^{page 504 note 4} This error arose from the fact that -vastarem was separately translated, -vastarem being probably referred to a root ‘vas,’ either meaning ‘to nurture’ or ‘to befriend.’ Nēr. follows the mistake with his sahāyinah. A ‘learningfriend’ would be too literal. Nēr. has nirvāṇajñāna- = ‘wise as to nirvāṇa,’ ‘the end,’ or ‘Heaven.’

^{page 504 note 5} We must necessarily render bisrayā as ‘milk’ here; ‘mixed with meat’ would not do. Cf. R.V. v, 37 (391), 4, ná sá rājā vyathate yásminn índras tīvrám sómam píbati gósak'ayam, ‘the King suffers no evil in whose (house) Indra drinks sharp soma mixed with milk . . . .’

^{page 504 note 6} A citation probably from Y. IX, 79. I regard the word pāk = ‘pure’ as a gloss to the arš of aršūxtanām; or it is a piece of débris from another and older translation?

^{page 504 note 7} My word ‘regulated’ is hardly needed. ‘Pavan bahar’ seems to have been added merely for emphasis.

^{page 504 note 8} It is not at all necessary to follow Nēr.'s ridiculous blunder as to the ‘cowbanner.’ The scholars of his day, of course, so read the Pahlavi, but the original, if correctly referred to Indian drapsá, is decisive. The Parsi-Pers. MS. does not translate here.

^{page 504 note 9} It is better not to solve vārom as ‘vār,’ ‘am,’ ‘rain’ . . . ‘from me’; see vārema; it is a transliteration.

^{page 505 note 1} Or, reading mā for lakič: ‘In this place may he stand, for not possibly for the sake of sin may he come’; i.e. ‘may he not for the sake of sin be a guest in my body.’

^{page 505 note 2} So I prefer as to the grammatical form, as against Nēr.'s prabravīmi; see the 2nd singular imperative immediately following. Fravām is probably erroneous, from an erroneously read text; though ‘pour forth’ is a good (accidental) rendering for frayañtu = ‘let them flow forth.’ We should not forget that ‘y’ is often written like ‘v,’ so that frayan- may have suggested fravam-, as the nasals ‘n’ and ‘m’ are often confused, and may have been once expressed by something like the Sanskrit anusvāra; cf. ašhaum, which was impossible as an original and vernacular expression, the word being, of course, ašavan.

^{page 505 note 3} Reading ‘fravām am’ as more in accordance with the original. With fravāmam; see Nēr.'s prabravīmi, read ‘I pour forth.’ Reading franamam, ‘I bend forth in worship toward that which is thy . . . . ’ Nēr.'s prabravīmi, while, of course, looking at first sight more like a translation of fravāmam than of franamam (to-nam), is not yet decisive for ‘-vam,’ as prabravīmi means also ‘I praise,’ which would translate ‘-nam.’

^{page 505 note 4} While ‘cause me to come’ preserves at least the root of jaseñtu, ‘the mastership of thy will’ is almost comical for varez-. The vare- was referred to var = ‘to choose,’ and -āṅhō to ahu, aṅhu, = ‘lord,’ while the ‘z’ of varez is lost sight of. Nēr. follows the mistake also elsewhere. It is possible, but not probable, that -aṅh of varezyaṅhāoṅhō may be meant by -kāmak and -āoṅhō by axūīh, but I hardly think so, unless, indeed, both ‘-āoṅhō’ and ‘-aṅh-’ were meant, which is very possible, by the single translation.

^{page 505 note 5} I could not accede to this opinion in 1887 in SBE. xxxi, nor can I now. ‘Holy-born’ would be more naturally ašaza = ṛtajā than ašavaza-, which would be ṛtāva-ja (sic), or ṛtavan-j-. ṚtāVan occurs seldom, if ever, in composition.

^{page 505 note 6} Avanharezāmi is 1st singular conjunctive.

^{page 505 note 7} Treating the reading -šedkōnāñd (so B. (D)) as representing -ānī (-āñd often represents -ānī); see also the reading šedkōnam-Ī, we could avoid rendering -āñd as a 3rd plural conjunctive. ‘From thee,’ if the thought is of Haoma; ‘thy,’ if the thought is of ‘the evil’ being really addressed.

^{page 506 note 1} Reading ‘gundīn ī marān.’ Heading this ‘gōn-’ we might have ‘the stench (‘gōndin’) of the snakes,’ as a figurative expression, or ‘the evil smelling one of the wretches.’ Reading vāvanēnd as being meant for vāvanēnī, ‘I repudiate and I will conquer.’ Reading xūrsand, ‘I will repudiate the one content with the miscreants’ (or possibly ‘contenting the miscreants’). Exhaustive treatment is absolutely necessary here.

^{page 506 note 2} The translation errs, as I hold, grievously, Nēr. following, first as to janyōiš, then as to ūnām (Nēr. creṇim), while I think evītō xraδayāo is not rendered. Not so Nēr., who gives us: “Cast down in a blow of destruction, the party of homicides, of confused intellect, the set ‘of the low ones.’” As to janyōiš, see SBE. xxxi. I refer it to ‘jánī’ in an evil sense here as elsewhere in the Avesta; not necessarily so, however; see below at 53. As to unãm, I referred it to vedic ánūna (-nās mádās, etc., ‘from which nothing is lacking’; ūnay, to ūná; cf. ūnāyīḥ, ṚV. 53, 3).

^{page 506 note 3} The gloss ‘some say,’ etc., is valuable from the fact that it formally presents an alternative translation of the clearest nature. Whereas avaṅharezāmi had been rendered first as a 1st personal by the MSS. A (DJ), B (D), D (K⁵) (SP), this gloss suggests a reading of šedkonāñd as 2nd singular imperative, which it might express; and that would have in view an avaṅhareza as 2nd singular imperative, erroneous, as I think, of course, but interesting, and distinctly of importance to show the presence of alternative treatment.

^{page 506 note 4} This gloss is natural enough; but does it point to -ēnd in firēbēnd as representing -ēnī, a 1st personal conjunctive, as -ēnd so often represents -ēnī?

^{page 506 note 5} How the 2nd singular imperative came to be thought of here can only be accounted for upon the supposition that that nasyḙ- stood apart from ‘iti’; Nēr. follows.

^{page 506 note 6} Frīf′ (or frēf′) is, of course, not an exact grammatical form of dapta; but the inexperienced reader should understand that I do not carry my polemik against the traditional renderings so far as to expect unvarying exactness in the grammatical forms. Nothing could be more feeble as a procedure than to criticize severely the inexactness of grammatical forms in these ancient renderings, which have come down to us as best they could.

^{page 506 note 7} Did the translator actually see a present participle nom. singular feminine in nižgāoṅheñti; and have we possibly here a case of a present active participle with the nasal preserved. We must remember that all the nom. singular feminine forms of participle present in -atī were possibly (or probably) once ‘-añtī.’ The -eñti as 3rd plural here would have to be corrected. Nēr. did not see a verbal form.

^{page 507 note 1} I cannot accede to ‘puts upon’ for the original, which should mean ‘sits.’ For ‘places’ we should need the causative. Nēr. understood by the word ‘announces,’ whereas the Pahlavi should mean something like ‘dishes’ (the offering).

^{page 507 note 2} Or, ‘the birth of good sons,’ but hū = sū has often the force of ‘many’; cf. suvīrya.

^{page 507 note 3} Possibly meaning ‘beyond any other (woman).’ ‘By another,’ or ‘second’ (husband), is hardly probable. Nēr. has ‘something else besides.’

^{page 507 note 4} Pañč-le-gūń-šēm having the force of a compositum. This word šēm is a refreshing blunder. Sēm translates (?) the end of the genitive plural termination -nãm of pañčanām! Ner. curiously does not follow it. Were his Pahlavi MSS. without it? or have we his extremely interesting refusal to follow the error of his texts?

^{page 507 note 5} See the blunder of šēm perhaps corrected by the aīš of the gloss.

^{page 507 note 6} I do not think that the māhmanīh, or ‘mihmanīh,’ of the gloss is another blunder. My ‘lodging’ would be a ‘chez moi,’ and is quite natural as a rendering of havam = ‘I am’ in the text. The glossist did not see a form of ‘man’ = ‘to dwell’ in -anãm.

^{page 507 note 7} Or ‘which is the good thought’; yet the original may be read as a personal, ‘to the one endowed with good thought.’

^{page 507 note 8} Or ‘which is the evil mind.’ Nēr., however, preferred the personal idea in both cases; see his very clever sumatini, durmatini, from sumati, durmati; cf. ab'imātín from an ab'ímāti = ‘plotting against,’ and atithín from átithi = ‘wandering.’ See the personal concrete ašaunō, i.e. ašavanō, ‘of the saint,’ just on below. Some might, however, think that the personal concrete word ‘saint,’ occurring after, leaves room for the abstracts above. It is, however, on the whole, best to render terms in as personal a sense as possible, avoiding the abstracts.

^{page 508 note 1} As before, these terms may be translated as abstracts. The originals might also be read as abstracts; so I preferred in 1883–6. Nēr. prefers the concrete and personal concept again as before. Dūšhūxt as = dūš + hū + uxt (?) is interesting; or is it ‘dūšāūxt’?

^{page 508 note 2} As before, we might render in the abstract ‘with the good deed’; I now prefer the personal concept. So Nēr. also prefers the personal once more.

^{page 508 note 3} Notice that srōš and asrōš are taken in the gloss wholly as concrete and personal, ‘maintaining,’ or ‘not maintaining the Destoor.’ Nēr. has again the personal concrete form as before.

^{page 508 note 4} Nēr. has, ‘who takes’ or ‘accepts’ the guru. Perhaps yaxsenūnē‴ may merely mean ‘who has a Destoor.’ This, however, would be flat.

^{page 508 note 5} We must understand the expression aharūv (anarūvo), ‘holy,’ as expressing ‘orthodox punctiliousness’ backed by good conduct and principle.

^{page 508 note 6} I think that ‘af,’ usually meaning merely ‘also,’ should be here rendered in the sense of āat, the original.

^{page 508 note 7} At the first glance one would suppose this to be a rubrical direction; ‘again the priest renews, or repeats, the formula of praise.” But if Nēr.'s tarohitaḥ is to be read as ordinary Sanskrit, we have “(he has) ‘vanished’ again upon the completion of the praise.” Hardly “Do you depart again to his praise,” which makes but feeble sense. “Do ye go again to his praise” should be “do ye ‘come’ again.”

^{page 509 note 1} Rōdīt-tāk, or Rōsttāk, is, of course, a blunder for raonãm = ‘of the valleys.’

^{page 509 note 2} So, mistaking the sense of ãzahu, which throws the whole translation into disorder.

^{page 509 note 3} I do not hold this to be correct. Jainī is here used in an ordinary or ‘good’ sense; see my critical free rendering. The word ‘havat’ may be the auxiliary to a past participle understood, or it may introduce the following gloss in the usual sense = ‘(the meaning) is.’

^{page 509 note 4} I do not see what renders derezāhu, unless it be that the word looked in its original form like dregvañt; and so was rendered avarūn', but this occurs only in the gloss. The mistake as to a ãzahu threw all into confusion. Derezāhu I refer to darz = ‘to bind,’ Ind. darh, dŕṅhati.

^{page 509 note 5} Nēr.'s text is here in a hopelessly shattered condition.

^{page 509 note 6} The prohibitive conjunctive form is not reproduced; Nēr. likewise omits it.

^{page 509 note 7} Referring to the stimulus of Hōm as awaking the intellect.

^{page 509 note 8} I hardly think the Gāthas themselves were referred to. The words are here used of the Haoma Hymn, which is, of course, made up of separate pieces of various dates.

^{page 509 note 9} Section.

^{page 510 note 1} Possibly of ‘feastings’ (?), as one might suppose Nēr. to have meant with his āsvādanāḥ (read āsvādāḥ). So indeed most naturally; yet see āsvādaya, used by Nēr. for teaching, elsewhere; see Y. 32, 5c, etc. Even at Y. 32, 8, this should be the meaning.

^{page 511 note 1} This, while affording a good meaning, indicates an error in the text; fravām arose from mistaking fraya- for frava. Recall the frequent interchange of ‘y’ and ‘v.’ Of course, fravām, if it be the true text, is a 2nd singular imperative, see the gloss in the 2nd singular; yet Nēr. seems to have read fravāmam, see his prabruve (always remembering, however, that he may have read franām (am), for prabruve could mean ‘I praise’ as well as ‘I speak forth’).

^{page 511 note 2} Notice the 2nd singular present for 2nd singular imperative. Or shall we clumsily render ‘thou givest’?

^{page 511 note 3} Nēryosangh's ‘sarvaṁ’ is hardly correct for ‘ham’ here; see the haθra of the original.

^{page 511 note 4} Referring to 61, “zag ī pavanič gāsān' gūf‴“This 65 should be considered as gloss, as is indicated by the brackets.

^{page 511 note 5} This should be considered to be gloss; see above, Y. X, 13.

^{page 511 note 6} See Y. 48, 5.