^{page 641 note 1} J.A., 1930, ii, 193 sqq.; 1932, i, 245 sqq.

^{page 641 note 2} The Babylonian (not Assyrian) tree.

^{page 641 note 3} This seems to me a wrong form altogether.

^{page 641 note 4} J.A., 1932, i, 276 line 7, 278 u.

^{page 641 note 5} As asp and asps, J.A., 1932, i, 280; xyōnān and xiyōnān, 286; ast , 286; dast , 287; pošt , 278; even ī-m , 274, and ōυš, 270 (for μš); mazdēsnān, 274, but mazdayasnān, J.A., 1930, ii, 194 sq.

^{page 642 note 1} The old word for “neck” still persists in modern dialects, e.g. that of Sangisar (Zhukovskiy ii, 314). One does not see why it should be changed to gardan (Unvala, BS0S., ii, 645, followed by Benveniste, J.A., 1930, ii, 194).

^{page 642 note 2} māz- is apparently a dialect word, from mārz-, in the sense of Persian mālīdan, to which it belongs by etymology; another example of the loss of -r- in this position is Persian māze, māzū “spine”, from *mārzu- “the place of the vertebrae” (derived from Av. m r zu “vertebra ”, cf. JRAS., 1942, 242). I do not think that there is any talk of “kissing” in this line (as Mr. Unvala suspected, loc. cit.).

^{page 642 note 3} This Parthian word is common throughout the text (= MPers.rāy). Failure to recognize it has produced some interesting misunderstandings, cf. Benveniste, p. 200.

^{page 642 note 4} Probably belongs to Persian āzīdan/āžadan (the forms are not clear); cf. ču dībā ba-zarr āzade (āžade) in the Shahname.

^{page 642 note 5} Although many forms and words are Parthian (or Median), many others belong to the Southern dialect. The confusion reminds one of that familiar from the later FaKlavnyyât; how much of it belonged already to the original text is not by any means clear. It would be easy enough to harmonize.

^{page 642 note 6} “A leather cloth (a sufre) on which they serve the dinner.”

^{page 642 note 7} This is cast in the form of a riddle. The reader or listener is left to guess, from the description, that the date-palm is meant. “Its leaves resemble the (leaves of) reeds”—in shape, of course, not in taste (as Bartholomae said, Mir. Mund., iv, 24). That the last two words belong to this paragraph is shown by para. 28 (see below).

^{page 642 note 8} Junker has two ideograms for šakar “sugar” in the Frahang, both of them due to misunderstanding. “HLY’” is sometimes = HLB’; = šir “milk” vii 4 (LBN’ also occurs, see p. 67, n. 20), sometimes = HL’ (i.e. ballā) = sik “vinegar” v 2 where the correct word is relegated to the variants (sik and sirke); in Pahl. Texts 30, 6 = Husrav ud Rēdak 31, “HLY” ī truš is “astringent vinegar” = Arab. xall iqqif (Tha'alibi). The other series, v 2 = xxxi 2–3, contains the Semitio word for “beer ” (or date-wine), Aram. šikrā, etc. The Persian words are hur and another that I cannot read (it occurs in the Karnamag vii 8 “As they had no wine, they offered him beer”).

^{page 643 note 1} dārōgān wasnād would make better sense.

^{page 643 note 2} “Migrants” or “tramps”.

^{page 643 note 3} “Date-stones.”

^{page 643 note 4} “If the people leave (the young shoot) alone, so that they refrain from hurting me, my crown will be green till the end of the days.”

^{page 643 note 5} The identical, wholly Parthian, phrase recurs in the Ayādgār-i Zarērān para. 93 (p. 13, line 5). Parth. yad is found several times in the text under review. The explanation given in BSOAS., xii, 52, cannot be fully maintained in view of the ideogram HN = yad in the Parthian inscriptions, see ibid. 54, 66. The ideogram is = “if”, which corresponds best to OIr. yadi; several interrelated forms may have coalesced in yad. [Cf. even in the colophon to the Book of Zarĕr, P.T., 16¹⁶, yad ō rōz fraškerd.]

^{page 643 note 6} “ Until they have had their fill.” Instead of ōšt one could also read awišt-, more closely conforming to Man. Parthian ’wyšt-.

^{page 643 note 7} “HT,” if not simply a mistake for yt, is presumably the ideogram for that samo word. If the Pahlavi HT (already in inscriptions) is an ancient mistake for HN = hēn, it may have been used in this text to represent the Parthian ideogram HN = hēn = yad. Cf. above. At any rate, it does not correspond to ag “if”, to judge by para. 25, HT ‘L’ YK = yad ō kū.

^{page 643 note 8} The Parthian form is preferable on account of para. 19 (see above), where hirz- seems better than arẑ-. Perhaps one should replace all ideograms by strictly Parthian or Median forms; I fear I have not been sufficiently consistent.

^{page 643 note 9} Here the full Parthian form is spelled out, not too correctly. In the same line the apparent ’yě represents ayi or a'i “you are”, Parthian ’yy So also in para. 53, tū kust a'i ēdar “you are affixed here” (you are stuck here).

^{page 643 note 10} “Until you can bear fruit for men, they have to lead a male to you, as they do with cattle. I would even go so far as to suspect that you were born out of wedlock.”

^{page 643 note 11} = which.

^{page 643 note 12} Uncertain. Kyč may be = Parthian kyč, which equals Persian kas. On the other hand, the rhythm (cf. kardan nē šahēd at the end of lines) may favour the explanation proposed by Bartholomae, loc. cit., 26, line 5.

^{page 644 note 1} Doubtful. MSS. ‘Lc.

^{page 644 note 2} “Satchel”, = Persian bārjāme.

^{page 644 note 3} = which.

^{page 644 note 4} “Of morocco leather”. saxtag belongs to Pers. saxtiyan; cf. also the Sogdian form mentioned in BS0AS., xi, 714, n. 6.

^{page 644 note 5} “The archer's thumb-stall [not ‘gloves‘] for the illustrious companions of the king.”

^{page 644 note 6} Cf. Sb.P.A.W., 1934, 33, n. 4 (Man. MPers. rby ).

^{page 644 note 7} Or dibēwān (not, of course, to be read dabīrān), the ancestor of Persian dīwān. Originally dipt + pāna, hence “where one keeps and looks after the documents, writings, etc”. The word was early shortened to dēwān (by diwēwān). In this form it is attested (apart from Arm. divan) as the name of one of Mani's books, his Epistles. Each epistle was called a dyb = dib in Middle Persian, e.g. Muhr Dib “the Epistle of the Seal”; the whole collection was a dipi-pāna-. Surely the oldest example of the use of this word for the collection of a man's writings.

^{page 644 note 8} Var. pist. Both pronunciations existed also in Persian.

^{page 644 note 9} MSS. HRWNN = ? Scarcely = mēš (Fr.P., vii 3) or ālā(y) (ibid., iv 6 note 32). Cf. P.T., 16¹⁶.

^{page 644 note 10} Rōγn-x^wardig, literally “butter-food”, means “sweetmeats” in Pahlavi. The “Southern” form rōwn-x^wardig in the Husraw ud Rēdag, para. 37, corresponds to halāwī in the Arabic version. In spite of the help afforded by Tha'alibi, Mr. Unvala misunderstood the word as “side-dish ”.— The Man. MPers. form is rwyyn, see BSOAS., xi, 57, n. 56.—Possibly the line ran originally rōΓn ud rōγn-x^wardig.

^{page 644 note 11} “Tokharian marten-furs” ? IIz = xaz occurs in the list of fur-animals in GrBd., 96¹².— Or should one read mušk syā(w) [ĉē] buz tuxārīg “black musk [of] the ‘Tokharian’ goat” (= Musk deer ?) ?—Neither xaz nor muŝk are articles usually associated with the name of Tokhāristān/Balkh.

^{page 644 note 12} Fr'č is used in Man. Parthian, but in the Parthian inscriptions there is prhŝ instead, e.g. HN prhŝ ‘L =yad frāxš ō. This curious form represents the ancient nominative, i.e. frānxš = Av. frąš = Skt. prān, with loss of the nasal; it shows that Bartholomae's rule, Grdr. Ir. Phil., i, 1, p. 11, § 24, is not correctly formulated.

^{page 645 note 13} Thus rather than āyēd.

^{page 645 note 1} It is strange that all students of this text, even Bartholomae (loc. cit., 27), have stumbled over the perfectly ordinary ideogram for “date”.—Amrāw is the appropriate Parthian form (Man. ‘mr'w, against Arm. armav); however, at the end of the text, para. 54, xurmā is written in clear (hwlm'y, “hwlm'k”).

^{page 645 note 2} The goat predicts that the hopes which the date-palm put on its seeds (in para. 19, see above) will come to nought. Hence, “may your pips and stones end up in (lit. go forward to) the alley of the dead”? I feel rather uncertain of the reading of the last two words, kōy murdān; may one compare the Persian phrasekūĉe-yi xāmūŝān “cemetery”? At any rate, Mr. Unvala's version has little to commend itself, “Wounded to the life thon willst be destroyed exterminated by the spiritual leaders”!

^{page 645 note 3} The first word remains uncertain.

^{page 645 note 4} So, of course.

^{page 645 note 5} “The rest-house of the Jewels is a flowery place.”

^{page 645 note 6} “On the great New Year's day.”

^{page 645 note 7} So to be read.

^{page 646 note 1} “Selected by sifting”, Pers. bixtan. Cf. below str. 5. In MPers. ‘zwyxtn occurs (same meaning).

^{page 646 note 2} This transcription of nmstyg is indicated by Parth. inscr. nymstyk (Inscr. of Shapur, Line 4, cf. also Sprengling, AJSLL., lviii, 169 sq.), which is rendered by παράκλησις in the Greek version, i.e. “appeal, request”. This meaning fits the Manichæan texts far better than “adoration”. Both spelling and meaning are at variance with the derivation from Olr. namah-, which has to be abandoned.

^{page 646 note 3} Here spelled gwmdg, but elsewhere gwyndg. “Failings, offences.”

^{page 646 note 4} nywryd (ni + var-).

^{page 646 note 5} 1 do not regard as a poem the passage from the Great Bundahishn (p. 10) to which M. Nyberg has given such prominence (ZDMG., lxxxii, 222 sqq.). The wording indicates clearly that the passage is merely a Pahlavi version of an Avestan text (quite possibly of an Avestan poem).

^{page 647 note 1} Benveniste, M. quoted two passages to prove the existence of rhyme in Man. Parthian (J.A., 1930, ii, 223).Google Scholar In the first, the words at the end of the lines should be read (a) framanyög, (b) abēstaft, (c) wilāstīft, (d) mārtag. In the second, we have (a) wasnad, (b) astâd, (c) mardōhmān, (d) paidāg; ‘st[’;d], in the place of ‘st[d], is wrongly restored (the next word is [mdy]n). There are no strophes in the second passage, which is an “alphabetical” hymn. Far better accidental rhymes can be found in most Parthian poems.

^{page 647 note 2} The words I have added are in square brackets.

^{page 647 note 3} A gloss:spâhbedan “generals”.

^{page 647 note 4} Another gloss: ku amāh mihtar hēm andar gēhän “thinking ‘we are the greatest in the world’”. Presumably to explain wēš-mēnīdār.

^{page 647 note 5} MSS. ’pi's (= āfrāh) instead of ’pyl's.

^{page 647 note 6} A variant: āsmān (which also would make sense of a sort). I prefer the rarer word, a-sāmān, which here, in conjunction with abē-rāh, probably had the meaning of Pers. bīsā mān

^{page 647 note 7} Not spōz here. Cf. P.T., 56¹⁸-57¹gētī pad spranj dār ud tan pad āsān.

^{page 648 note 1} Or “self-abandoned”. The word had both meanings. Cf. the passages collected by M. F. Kanga, The Testament of Khusrav I, p. 3, n. 4 (add Pahl. Texts, 143, 5).

^{page 648 note 2} The terminus ante quem is A.D. 956 (if the figure—324—in the first colophon, P.T., 83, deserves to be trusted).