3. Spiegel, Bartholomae, Geldner, and others, abandon the significant rendering of the neut. dual (with vwhyéd akemchd) for an adverbial loc. = ‘in the former time’. Objection is made to the neut. as being improbable, whereas, if a neut. were certain, it would be of the last importance, as agreeing with the two neuters, vahyéd akemchd, to prove a striking depth of conception. An adverb in the loc. would be hardly less valuable than a neut. dual. ‘There were two spirits at first’ would show a close habit of logical procedure in statement, beginning with the beginning, but I prefer a more
natural solution. It might be said that the loc. corresponded to manahichd etc. But manahicha defines vahyd akemchd, ‘a better being (nt.), or an evil, in thought - - - etc.’ ‘Thus were the two first spirits’ either presupposes one of the many gaps of lost matter which described the two spirits more in detail, or else, what amounts to the same thing, it presupposes a thorough familiarity on the part of the masses addressed with the doctrine of the ‘two first spirits’, Moreover the adjective is essential to the force of the expression; ‘the jirst’ spirits meant the ‘original’ spirits, the sources of all later phenomena (see v. 4); and the neut. dual carries the two all-impor- tant elements with it, the abstract conception, ‘things’, ‘existences’, ‘a better existence, and an evil one’, and the expression of their originality. (There was no beginning before them in which they could exist. They were without beginning , creating life, (the beginning), and non-life.) The true adverb may be seen in paowrvém in the next verse; not that paouruyé(é) cannot bean adverb; see it elsewhere so translated by me; I am only loath to give up the neut. dual, even for the adverb.
But, for those who cannot see the neut. dual, we can fall back on what is stronger, and that is, a nom. dual masc. For I have an explanation which will have weight with many who study the Pahl. Alphabet. The word is properly
paouruyyd (sic) for paourvyd. The letter 49 (for which yy and S are at
written) is a well-known Pahl. combination, equalling, among other things, ya. The entire Avesta Alphabet having been originally Pahlavi, we can understand that some of its letters still retain their original application to
widely different sounds. I have had occasion to refer to this letter 4 (0 ) more than once in reference to this fact, and of all the Zend-pahlavi letters, this is the most important in this connection. we is here, as elsewhere, a lengthened 4, and equals y@; and the first y is one of those superfluous insertions which are so familiar; cp. also the frequent omissions of u, as in hvarezdo for huvarezdo etc. The awkward insertion of the »» here was caused by the equally awkward omission of one half of the letter »», » = wu being most certainly miswritten for »» = » (or English w (?)). As >» was
defectively written », so 99 was clumsily overwritten before = W=yd; 2999) bas 4s = paouruy® should have been sa») bargy = paourvy®. When therefore 445999) bares was written, the »9 of the 44999 (wo) constituted
the overwriting of one half of the Zend-pahlavi letter 4, which was actually and literally only a freely written ss joined into 4, in Pahl. = yd. We have then the doubly falsely written’ 442») ba0gy = paouruyyd, and we should correct by adding to the », and rejecting the 9); that is, we should correct the clumsy pen-strokes y» into »y, and neglect the half y, 9.
yon rbargy = woo») bargy = paourvyd, Sk. *pirvy® nom. dual masc. The proof of the correctness of these suggestions may be seem in the termination you, which is absolutely senseless and impossible when read
ahé(é), jand which is simply youe = ahya = asya. So also in the word
Yo}r4g 5 when transliterated kainé(é), it is impossible as a nom. sg. fem. of
the a stem. The word is kainya = kanyd. (See similar remarks else- where.) Justi, with keen insight, long since saw that é(¢) and ¢(¢) must
often represent ya, yd, but the fact is that yw does not merely represent
ya (yd); it is often actually and literally ya, or yd (see S. B. EK. XXXI. p. xxxiv), while even the fact that yy ( ) at times represent ya, yd, is
often totally missed; *see my explanation of thwdi = twé = thwayd, ¥o51, 9.
Regarding then the word as a nom. dual masc., or neut., I would compare the parallelism in Y. 45, 2, 3, 4. See paouruyé(é) (sic) ending the first line in vy. 2, then paourvim in the next verse, then vahistem in the following verse; no adverbs appear. I would cite also a rendering which has circulated: ‘Und die beiden uranfdnglichen Geister -- -’. One scholar follows the Pers. trl. of the Pahl. yémd?, comparing Sk. y@ma = ‘nightwatch, time’, and ydman (loc.) = ‘at the time’. I think the yémdi an ineffectual
transliteration of yémd. The pep = yo is simply miswritten for 99 which may well represent yé in transliterating, or else the Gatha-text was read ydma = yamd; cp. jydtim = jydtim (jivdtim, jivdtim). The letter av which, as Pahl., we should read d@, is simply the Zend wm = d; see #85 also partly in Zend letters (in Y. 46, 18). In the gloss of the Pahl. vinas va kirfak have the place of yémé@ before Lenafshman, obliging us to recognize the probability that yémd? (sic), or yomd, was understood as mean- ing ‘the pair’, and this notwithstanding the zak @ kolé II (dé) = td in
the former part of the verse. This yomat (yoma) was supposed by Ner., and by the Pers. trlr. to be a corrupt form of yém, or, more probably, it stood divided (yém @) in Ner.’s MSS. Then the trirs. translated what stood before them by rézhd and bhiimandale (‘ world’ or ‘day’). Yemé = yamd (cp. the Agvins, and Yama’ and Yami’) is nom. dual masc. It is a curious expression here, but its occurrence possesses authority. Vafend == ba + apand? = sva + apna* = ‘self, or independently (that is, originally, as Deities) operating’, One writer suggests the meaning proper in hvapdo, Y..44, 5; cp. sodpas = ‘skilful’. This is however not very probable as applied to tbe Iranian Demons. Their chief was often emphasized as the ‘unintelligent Aiigra Mainyu’. Another suggests ‘in sleep’: ‘The two were made known to me in my dream’. Another suggests the reading adafnd = ‘sleepless’ (cp. Yt. 10,7. etc.). Asrvdtem = ‘have been heard *; 1. e. ‘widely known’, as a 3"? dual mid. in a passive sense, so the most. (* Vafend defines paouruyé).
But I hear of a gallant effort to break this spell: ‘And the two original spirits understood well (not a mid. passive) the right and wrong in thought, word, and deed; and by them both were the righteous approved of (sic), but not the evil’.
Vis(k)yatd is to be referred to chi ; See the Pahl. vijid, possibly vijéd (same signs), from ow (vt) + chi; see alsoNer. and the Pers. As to the s(k)
(the & merely transcribes the outward appearance of a part of 49 = 5
before y) it represents ch before y (Haug); see ash(sk)yd translated saritdrth = ‘greater wickedness’ (Y. 51, 6), aS comparative of aka (acha) which has the superlative achista. So also S(k) represents Sansk. ch in §(k)yaothna = chyautnd.. Vis(k)yaté = yt + chydta = vi + chi + dtd. After the Pahl. trlr. has positively deciphered the word for us by his vijid(°éd), the ques- tion remains whether we shall refer it to a root conjugation of chi (comp. cheti beside chindti), or to a root aor. As Whitney reports achidhvam as a root aor. (of ché), it would seem better to call it a root aor. But the two chi’s were originally one. Except for the long ¢, we have no difficulty in recognizing the word as a of + chi + até (= °ata), a 874 pl. mid. pret., or improper conj. But then the long 4 is significant, so the MSS., (see also vy. 6). Is this ¢ a miswriting?, or does the long @ represent an in the termination, as, for instance, long ¢ in the root Ja, jan, etc.? If so, it is possible that we may have here an unstrengthened conjunctive 374 pl. o@ + chi + até = vt + chi + antd = * vichyanta; cp. bhuvdni. Surely, if we accept dtd as equalling a 3°¢ pl. ending, it is better to give it some explanation in har- mony with known laws, even if, at the same time, we must accept an ex- ceptionally unstrengthened root,
The Pahl. trlr. seems to have understood a 3r4 ge. mid., but his word may also be rendered 24 pl. imper. ‘do ye choose’. It would offer a fine thought if we dared follow his possible hint, and read véS(k)yatd as an imperative of the a@ aor.; cp. ahyat from hi (ep. a(d)vaé(é)natd in the prece- ding verse); yet see vis(k)ydtd in verse 6.
The meaning ‘selves’ was first seen by the Pahl. trir. With glosses dropped the effect is better. As ordinary Pahl. read line 4 (with glosses): ‘Both thought thoughts, spoke words, and did deeds, the one who is good, and the one who is evil []’. Possibly a: ‘Thus the two spirits, the first, were themselves heard, (or known) - - -’.
4. There is not much difference of opinion here. Some would however take ajyditim in the sense of ‘the continuance beyond this life’: ‘And, as these two spirits of old counselled together, to establish the earthly life, and also the continuance beyond, and that, at the end, the worst condition should be the portion of the wicked, but for the righteous the highest rejoicing of the soul’,
There is some question as to whether we may take paourvim as quali- fying jasaé(é)tem, or dazdé(é). Looking at its position I should say dazdé(é); but this would be a little too minute, paourvim and apémem in- antithesis. Paourvim qualifies the whole verse, and also v. 5.
The Pahl. trlr. renders dazdé as an infin. (* dadh + ¢é(é)). Otherwise one is strongly tempted to regard the word as a 3r4 dual with Justi, Hiibsch- mann, and Geldner, and as a perf. with the last two, daddte (cp. dadhite), shortened by later corruption to fit the metre. As it stands, it is a regular 3rd so. pres.; ‘each establishes for himself’. The accusatives forbid a passive sense, and an infin. leaves the sense uncompleted. ‘Some finite verb ex- pressed, or understood, is needed. The next verse may however possibly supply one. One writer would narrow the sense of the ‘best mind’ to the ‘ood conscience’. It was not only ‘conscience’, but correct mental insight as to all particulars, as subjective beatification in the future state , apema anhu; see vispdi yave(é) Vanheus Mananhd, Y. 28, 8.
Reading va after Aharmano (with Sp.), ¢ in the Pahl. might be ren- dered: ‘The worst characteristic of the wicked, [and the devastation owing to Aharman and the wicked are fully seen], and so, on the other hand, the best thought of the pure accompanies [A. everlastingly|’; so I rendered formerly more naturally as Pahl.; but see the Gatha and Ner. The last line needs the force of some form of khaditéintand understood.
5. Varatéd, Haug as imperative (?); the Pahl., followed by Spiegel and others, gives the proper indication. Some would regard verezyé as an infin., or would even change the text to verezaydi. Others, abandoning an
infin., hold to an acc. pl. of an adj. veree?, yet not reading vereeayd with two MSS. I have very little doubt that verezyé is an adjective in the sense of ‘effecting’. Its position at the end of the sentence highly favours a nominative form, as does the symmetry of the construction, varatd - +» - verezyd ++ - spenistd . - - vasté(é). Another form of verezyO occurs in mathra verezya _ in a good sense, here the sense is evil.
For vasté(¢) the Pahl. trlr. renders néhifts, which gives the idea of ‘covering’, ‘clothing on’; cp. Yt. 18, 3. Haithydi§ must have a meaning approximating ‘good’, or it may mean ‘real’ merely, and find its associate in fraor(e)d(f). This latter was considered the weak form of the present participle of var + fra taken adverbially, by Haug, Justi, and others, and the metre favours three (?) syllables. But it is better taken as = fravrit, adv. neut. adj. from vi (var). This suggestion had long circulated, and was later published with, however, the meaning potius, lieber. I take the mean- ing to be that indicated in frd té(e) verené(e) etc., and so ‘piously’, ‘beliey- ingly’; see the Pahl. rendering in Y. 53, 2, Srandmishno,