^{page 77 note 1} “From the hills,” so the Pers., translates a rational gloss, in antithesis to “ground springs“? So B. marks g in a gīrān. “Springs and elevations” are the two obvious sources of water; but only accidentally is this rendering thus rational, gīrān = kūhān = hills.

I hold the θraotōstātas- (?) to be one word, θraota- + the suffix -tāt-, this in an adjective sense. The Pahl. should be deciphered srao(t)-tačešn', not gīrān- (same signs). Otherwise correct the Av. text to θraoštōcāč-after this indication of the Pahlavi translation; see ஸraoštōstōčām in some readings of Vendīdād, xviii, 125 (Sp.).

So the Pers., text and translation, also indicates, from a misunderstanding; see above. We can hardly accede to the old view that stā = “stand” is present; and so “abiding-in-flow”, “steadfast-in-flow,” as gen. sg. or ace. pl. The same word occurs at LXVII, 15, with the same Persian translation. “From the hills” cannot be gloss, for its forms are the same as those for srāō-; see the original. We might indeed as well decipher dērān(g) tačešn = “long flowing,” “flowing from afar,” dērān(g) and gīrān having the same signs. Some might suggest “flowing and standing waters” for the Av. form.

^{page 77 note 2} The Pers. has rīšak = bīχ. The form which one can most naturally read as vaχdūn = “taking” (vebedūn would be too remote) should be associated with rīšak, leaving va-ārādešn for var(e)šhajīš, the external form of the Pahl. word having been accidentally first conformed to the external appearance of var(e)šajīš, and then rationalized to ā-rōdešn = “outgrowth,” a natural process which explains many a curious usage in the Pahl. translations of the Avesta. That the signs which we most naturally read as vaχdūn, vebedōn, or, with the first stroke omitted, sic as āyī; see -jīs-, really mean anything so shadowy as vaχdōn seems to me most uncertain. I can only decipher what appears to be the Pers. trl. of va ārōīšn (ārōdešn) as va + rustah, which is accordant.

^{page 78 note 1} The Persian seems a rištah = —? The Pers. has its Pahl. text asīman. as if “of silver”; so Sp. and so D., a mere mechanical mistake which is said to recur in many good MSS., but see Av. asmanem = Indian açaman.

^{page 78 note 2} B. and C. have curiously patmān; C. has no trl.; whether “all the Heaven” suggested “all the measure (?) of created objects”; patmān rather recalls to us “melodious measure”, as used in the Av. commentaries; again some accidental mistake has occurred.

^{page 78 note 3} The Moon is first masculine, as in Veda, where the Sun was at times feminine, when regarded as a divinity. The prior position of Māh here may have some meaning.

^{page 78 note 4} The Persian translation renders erroneously, but suggestively bi-šumār, “without number.” I should still render the original “without beginning”; “self-determined” is gloss. The fixed stars seem indicated.

^{page 78 note 5} B. has a modern gloss written under, possibly = xχōškīh, χuški, other wise one thinks of χuning (χūnīg) = “having-blood.” Such underlined modern glosses in B. (Pt. 4) seem uniformly taken from the original of C, the Persian MS.

^{page 78 note 6} For ravas-, etc., after Windischmann's rag'ú see Justi's lag'ú, why not simply to var- with a transference, rava- = vara; see urú, also to var, though this analogy may be somewhat far-fetched.

^{page 78 note 7} The Av. reads “which follow grass meadows”; so I would emend my former rendering at p. 329.

^{page 79 note 1} Notice that tē = “thine” is translated as if it were tē, the nom. pl. masc.; the dadāθa, however, determines to “thine”. It is not to be expected that the Pahl. trlr. would he infallibly exact as to either case, or dubious sense, here.

^{page 79 note 2} I think that r of this hvāpar (?) may be a false explanation of any earlier sign, which might equal an r or an ō in the Pahlavi, for the clearly written Pahlavi r of hvāpar causes some hesitation as an explanation of Av. hū + apō; but if this Pahlavi r can be explained as an ō, we should secure at least a transliteration, if not a termination, for we should be able to decipher hū-apō, and this would be not irrational for Av. hvapā.

^{page 79 note 3} That is to say, “to the clean animals” as against the unclean made by Mainyu! Cf. the unclean animals of the Pentateuch. Here we have possibly the explanation: the creatures were clean or “unclean” according as the Deity or Satan made them.

^{page 79 note 4} Man′, “since,” in its sense of “according to which things” = “then”; this to correspond to the pl. in yāš, having nearly that sense.

^{page 79 note 5} “By him”; so, either carelessly or freely; some of the Pahl. trlrs. could not fail to render a second singular perfect active. See my excursus in the Dictionary upon the subject, Gāθes, pp. 862 seq.

^{page 79 note 6} B. has here amat for its own yōi, a somewhat unusual freedom.

^{page 79 note 7} Notice that the adjective “glorious” rather than “comfortable” is indicated here. The trl. of the Persian is indeed āsānī, but the Vedic svar expresses prominently the idea of “shining glory”. Should we not modify the meaning of Persian āsānī; would not “delectable” be a fair compromise between the two ideas?

^{page 79 note 8} One would rather think of “plain”. The Pers. MS. seems to translate kufah, herbage (?), but the writing is obscure, and varayō naturally equals var.

^{page 80 note 1} It is quite impossible that “the truthful speech” can here refer only to the correct intoning of formulas, much less, exclusively, to such a superficial idea. Notice again that we have clearly the presence of the moral idea. The hturgy was not all that was meant, though this follows the allusion to the Fire; for see the striking continuation.

^{page 80 note 2} Aši (arši) hardly means “money” here; the Pers. seems to render tarsakaī (sic) with nēkī = “a benefit.”

^{page 80 note 3} The “perfect mind” strongly implies the moral idea again, both in text and translation, and not the “earth” here for ār(a)maiti; words which are “concurrent” with the “earth” is an impossibility, unless indeed “the earth” was regarded as almost an exclusive synonym for the super-sacrosanct. The apparent long a of Ār(a)maiti may well, as so often in similar cases, have resulted from the misreading of the earlier Av Pahlavi sign in the transitional period, the long a of Pahlavi being equivalent to the short Av. a.

^{page 80 note 4} Hardly “appropriation”, χvēšanīh, with C.; so also at 56; as if “make me your own”, or “that I make you thus mine”. With the yehevūnt of B., C. “it was for my own profit”.

^{page 80 note 5} See buyata, in spite of B.'s translation yehevunt and of C.'s translation būd; yehvūnēt (let us hope it is a genuine K⁵ reading) should certainly be regarded as a second personal plural here for buyata, though doubtless generations of readers have read it as a third singular; from this (?) the yehvunt.

^{page 80 note 6} See the dat. case noticed here, h raī for me; but C. omits rāi (rā) here at 3; h is over, but original, in B.

^{page 80 note 7} Spiegel has no Pahl. text here for “be ye to me for the soul's blessedness”; but the words appear m A, B., and in C. They may well, however, be a later interpolated repetition. B. has χvšanih mē'im yedranam (so B. over, but it seems to be original); by which the last transcriber probably meant “for my personal (interior) interest I bear (them) on” but one suspects that a prior writer meant yedrūn am = “bring (those) to me as my (deepest) interest” (which would more rationally render “be ye to me my soul's benefit”).

^{page 81 note 1} But the translator may also have really meant us to understand that he rendered ruvān as if it stood in the accusative case: “I invoke (his) own soul,”—this notwithstanding the curious combination.

^{page 81 note 2} Not at all improbably meaning the “sacred water and the wood” there present, but yet see above, where the “wood” is especially mentioned. I do not hesitate to leave these words provisionally in their original accusative force; see their terminations in the Av. text.

^{page 81 note 3} Notice the impossibility of any idea of “idolatrous worship” in this sacrificial reverence.

^{page 81 note 4} Bavīhunyih, while possibly a form used for the optative (conj.) of the 3rd(?) sing., as well as of the 2nd, certainly looks more natural as the 2nd personal, and this notwithstanding isōīṭ; for see the (so A.) in 62.

^{page 81 note 5} Man′ should be only a mechanical mistake for valā, or for zag ī man′.

^{page 81 note 6} Even with the texts of A., B., χvēstar- = “(be) an appropriator” seems indicated, and the χvēš- had better be retained, if Spiegel's text really represents K⁵. In which case I think that the āf of āfryeiδyāi may have been mistaken for āp = “to attain to”. The χvēšād of the Pers. text with the trl. χōd kardan is to the same effect; though without the āf, δvēšād as a conj. imperv. might remind us of the infinitive for the imperv.; see the Av. -δyāi; notice also the inf. χōd kardan. Accidental influence is here present doubtless; but it may not be here paramount. The r of χvēštar may well result from the r in āfryƹiδyāi. Read Sp. χvēš-āfr(?). The idea of the original “friend-acquisition” is but dimly adumbrated in an āp = “to obtain”. Why āfyẹiδyāi.

^{page 82 note 1} C. pl., A. singular; see (sic). Pers. “for that man is evil to thee (?) who [gives] the best object to the evil ones.”

^{page 82 note 2} (It may be that the Av. syllable stood earlier in quasi Av.-Pahl. signs resembling Pahl. characters for χvēš) So, not following the pāhrūmīh of A., as I preferred at Gāθas, p. 250. The fravāmešn, (franāmešn) of B., C, and Sp., though it may express “reverential friendship”, and was so understood by the Pers. trl. (see dūstī), is itself, I think, the result of a mistake as to the form of a letter; fryō suggests fry–; but –y– in the original is often mistaken for –v– see, among many other occurrences, mruyẹ, which is nonsense except as a mistake for mruvẹ; so the translator wrote fravāmešn′, or franāmešn′, for a frayešn′ (?) [so B. (Pt. 4), with Spiegel and C.]. It is better not to take refuge in that pāhrūmīh of A., this in view of the dūstī of the Pers.; and this though the pāhrūmīh of A. offers the very easy solution; “He is holy who is the best thing (the beatitude) to the holy,” for this latter solution entirely loses sight of fryō. Perhaps indeed something like p-a-hr-ō was seen in f-r-y, the y being here mistaken for ū as so often, or for v. The Persian enlarges bulandī dehad u dūstī kunad. From 61 to 64 (inclusively) allude to Y. XLVI, 6, occasioned (?) by the occurrence of the adjective “heroic”, which is just before applied to this Haptaṅhāiti.

^{page 82 note 3} The Persian adds the [entire (?)] Yasna; some say “this fargard”.

^{page 82 note 4} As hi seems rendered by mā in Sp. (K⁵¹), so in C. mūn (sic), rendered tā, seems to represent it; mā =, in the old transcription, maman.

^{page 83 note 1} Important alternatives, if the points were really meant—for, first, as to pavan (so E.), “until (so) the last end of life” would imply a prolonged continuous supernatural existence of Zaraθuštra upon earth. Whereas “at (the last ending)” falls in with the usual ideas of “finality”; and, second, the idea of “the last Fargard” (so the Pers. in its gloss) corresponds with that of the last “ending”.

^{page 83 note 2} Whether this means to measure the distance at which Hell is situated from this globe one cannot say; probably not.

^{page 83 note 3} The MSS. B., C. alone furnish us a text for iδa aṅhō a(r)šava.

^{page 83 note 4} Notice -īh as termination. B., C. have here a text -ēnd which we should freely understand as an -ēnī when so indicated; see elsewhere: “will I cause thy soul to pass over the Judge's Bridge” would be an excellent sense; but see the original. Does not this translation, however, suggest a restoration of the original to the first personal, and may not be a first personal like at Y. 28, 11?; also like the first personals in Indian -ase cited somewhat doubtfully by Whitney. “Thou shalt cause thy soul to pass over” seems awkward enough. The Pahlavi translations following our texts sentence by sentence, often, as we should naturally suppose, refer to older texts which have been corrupted to their present forms. Here and there parts of Pahlavi translation were interpolated from older MSS. without revision, or strict comparison with the Avesta text actually under treatment, and in the course of time these translations were actually written mechanically, immediately after the text in the codices, sentence by sentence. See statements to this effect by West upon the introductory portion of B. (D., Pt. 4).

^{page 83 note 5} So A., B. aharūv′, not aharūvān′ with E.

^{page 83 note 6} So again, a present for the imperative. We should not hesitate to render those second singular indicatives in the imperative sense where indicated; see throughout, where they correspond to the imperatives of the original. The Uātavaīt′ was the hymn for the departing soul in its first strophe, suggested here by the closing of the Yasna.

^{page 84 note 1} So A., B. aharūv′, not aharūvān′, with E.

^{page 84 note 2} Notice uštavaiti explained as “the beatific course”; uštā = nēvakīh and (v)-aiti as = “course”. Here, as elsewhere, something like a form of i, aẹ = “to go”, seems to have been noticed in -ātem; as the sign for long ā in Avesta may equal an ae in Pahlavi, while the v may = our Pahl. ′, a perpendicular stroke which equals v, but which is also at times without meaning; something like uštā + āẹ-ti (?) seems to have been thought of.

^{page 84 note 3} These suggestions refer in a somewhat confused manner to the ending in vanhēuš…manaṅhō in Y. XLIII, which begins with uštā ahmāi. At the first glance it looks as if we had an alternative translation in this gloss to 73. Notice desire = var(e)- seen in varežem, a curious error elsewhere recurring.

^{page 84 note 4} “Tears” were forbidden.

^{page 84 note 5} So also the Pers. mūyī (mūyah). Does not this give us a hint toward he restoration of another original text, (to Indian āmīvā) has hitherto sufficed us; but we must not forget that -vā is affix, whereas our vowel ū seems “organic”. If we could emend to an amuvay-, or to some similar form, we secure a more objective rendering; this, first, although mūyī might be a sound-imitation for “murmuring”, “lamentation.” It would be difficult to bring in a form of muš here. Here I would offer to my translation in SBE., xxxi, an alternative in the above sense.

^{page 84 note 6} Shall we decipher vār, or “vahr (?)”. I think now that the long ā of Pahlavi vār has merely the force of short a, Avesta a; the signs are identical, and they are often interchanged. The Pers. has puštī-dādan, “giving protection.”

Here we have another instance of unaccountable mistake. The first words of varedaχem are assigned to var in the sense of “enclose”. “Growth” is not seen, so similarly elsewhere; see Y. LXVII, 3.

^{page 85 note 1} Spiegel, followed by Justi, preferred to decipher ayīkīh (kiī) for the form here used, noticing, however, the obscurity of it; and Spiegel, who was of the highest authority upon such a point, mentions that the word is beyond question used in this sense of “impurity” in the Rivaets. The Persian here in Y. LXXI (Sp. LXX) also writes what might be an ayūdalcī, but its carelessly written translation seems to read ālat i anār = “instrument, utensil, for the incense wood”, its ashes, etc.; and this would point rather to ābōdakīh, “the incense-offering,” same signs as for ayūdakīh. Aēnīkīh, so deciphered, seems, however, rather indicated in the glosses—to the root of aēnah (?).

The aχtiš of Y. XXXVI, 2, is rendered man′ av′ aēnīkīh (or ayūdakīh) aēγ pavan me'im ātaš anākīh vebedūnyēn. This is erroneously placed in section 3, but its forms correspond better to the Av. of XXXVI, 2.

At Y. XXXVI, 3, the same form appears for aχtōyōi (t) = aχtaye; aēnīkīh (is it to ac?, with nasalisation), or ayūdakīh, again glossed in Spiegel with anākīh = “evil”, etc.

Nēr. has first anyāyam = “lawlessness”, this following his dvešam with nigraham meaning “destruction” following his dvešatvam. My Pers. MS. omits the difficulty. So much for aχtōyōi, = aηtaye in XXXVI, 3.

At Vd. ii, 16, we have in my Persian interlinear k-n-k translated by the same word; but Spiegel's copy must have read more plainly kīnah = “revenge”; this is, however, glossed in my Persian MS. with yadūkīh = “sorcery”(?). (Was not ayūdakīh indeed meant (?)). No sign of vištak here, though Spiegel's MS. had it.

At Vd. v, 85, my Persian MS. has ayūdakyā (-ya here equalling īh (?); hardly -īhā as often); and this is translated nā-pāhih = “impurity”; but glossed vištak, which latter is curiously here translated χarābī in my MS., as if an Av. stak + vi = “undone” (cf. staχra(?)) were thought of; whereas viχtahīh, aside from this, more naturally suggests “unfittedness” in another sense, guštāχ = “perverse”, guštāχī, “rudeness.”

At Vd. v, 86, we have again in Spiegel what might equal either an aēnakīh or an ayūkīh (ayūīkik), again also glossed with vištakīh (so— the Persian of Vd. v is not available with me here; nor is that for Vd. vi, 65).

At Vd. vii, 145, the gloss vištakīh appears again in Spiegel's edition; but he reported a χarābī; see above, from his Aspendiārji, as gloss, or translation.

At Vd. vii, 149, aēnīkīh (or ayūkīh (?) ) appears with simply aēnak (or ayūk), my copy of the Persian at Vd. vii not being available in the gloss. (No accustomed searcher needs to be informed that the ultimate etymological origin of a word has often very little indeed to do with its latest point of meaning.)

^{page 85 note 2} The Persian, however, here translates susti (?)=languor.

^{page 86 note 1} A. first read sraoθrem = srāyešn′, but sraoθrem is corrected to avauruš… (or -urus …); the trl. is left srāyešn. So the Persian reads aē srūdešn, as if ava, plus some form of sru = “to hear”,. or “to recite” were seen. But the form of B., whether accidentally or not, certainly allows of a rūdīh to -rud = “hold down”. The aē of the Persian looks as if anā = aē = “this” (!) were accepted. (Notice this decipherment of the original Av. ava, showing that the letter » = v in ava was so written in the Avesta text as to be indefinite, and probably standing in the Pahlavi form of = v or n. All such indications are important, for they tend to prove the lateness of the date at which our Avesta texts in general still stood partly in the Pahlavi forms.)

The Persian expresses this idea of “recitation” in its translation merely with bāz u bāz (as it seems to read), though this is hardly a translation. B. should be deciphered, where possible, in accordance with C., the Persian, with which it stands in the closest relation. But B. might be here more naturally read va aχar(r)ūdīh, which might be defectively meant for aχar-rūsīh = aχar-rūstakī as = “after-growth”, the two r's of aχar and rūdīh having coalesced, or aē = anā, so again chaotically deciphering the Av. form, ava, as if = “this“ + rūsīh, or better, rūs(tak)īh. As to Spiegel's form -ava + —(?), it might strike us at first sight as ava + gir + vaχš = “the down-growth of the chant”, or—?, but the form looks more like a miswriting for ava + sravešn when we glance at the srāyešnā of the sister MS. A. (Ox., C¹, DJ., J²). Otherwise (for Sp.'s form) ava is plainly to be read with the Av. ava-, and then one thinks of ava-drūχs, or, with the aē of the others aē + drūχš(?) = “this falsification” (?), or of an avarōešn (?) = “absence: of creed-impulse”. This all for Spiegel's unexplained form. (Accustomed readers will not need to be reminded that this exhaustive investigation of the Pahlavi translation even here, where the Av, is approximately clear, is still indispensable, as it bears upon other places where, according to some writers, almost an exaggerated dependence is to be placed upon the Pahlavi translation. As of course the Pahl. here is “especial”.)

^{page 86 note 2} Well entitled to this epithet.

^{page 87 note 1} So B. ins.

^{page 87 note 2} Referring to the consecutive arrangement of the various pieces for the ritual. We miss the syllable ham of haçdāitim, which seems in the Pahl. trl. to have been absorbed in the preceding hamāk = haurvām. Otherwise we should have “the entire putting together of the Yasna Yašt-hymns”, as we should say in modern terms, “their edition.”

^{page 87 note 3} B. has with this Av. text yasnya (hardly yẹmya), Stōt Yašk in the trl. only. Did they have an idea of the Yašts which are outside the Yasna ? as well as those within it.

^{page 87 note 4} Not so naturally to “my” soul, or “to my own Frava(r)ši”.

^{page 87 note 5} Notice the gen. pi. for the Av. acc. sg.

^{page 87 note 6} So, the form which looks like Yašt must be rendered yazat, though B., C. were both of them led astray, rendering Yašt. This rendering pavan mēnešn for upamanem (so in Y. I, 44, at Y. II, 58; so in Visp. i, 26 (vii, 28, has no Pahl.)) is defective, being only vaguely significant. Nēr. alone gives us the valuable šāpam = “curse”. For 86, 87, see also Y. VI, 48.

^{page 87 note 7} See Y. XXVI, 1, 2.

^{page 88 note 1} Notice that “for money” or “for wealth” cannot well be the meaning of Aši (Arši) here, except when we regard such “wealth” as being “especially sacred”.

^{page 88 note 2} Probably again mentioned, just here, as if in climax.

^{page 88 note 3} The Zaraθuštra (if so read (hardly -tri (?))) of the original seems to have adjective force.

^{page 88 note 4} Notice the grammatical form of the fut. pass, participle rendered) even with a circumlocution.