^{page 45 note 1} Ch. Virolleaud, , “La naissance des dieux gracieux et beaux, poème phénicien de Ras Shamra,” Syria, xiv, 128–151Google Scholar. I have adopted for this text the siglum 'INY (= 'Ilm N'mm wYsmm).

^{page 45 note 2} Friedrich, J., “Zu den drei Aleph-Zeichen des Ras-Schamra-Alphabets”, ZA 1933, 305–313Google Scholar.

^{page 45 note 3} See Tarbiz (Hebrew), iv, 383; further examples from 'INY and elsewhere will be pointed out below.

^{page 45 note 4} Albright, W. F., “The North-Canaanite Epic of 'Al'êyân Ba'al and Môt”, JPOS xii (1932), 185–208Google Scholar.

^{page 45 note 5} Tarbiz iv, 380, bottom; OLZ 1933, col. 593, n. 1.

^{page 46 note 1} In the parallel (Syria, xiv, p. 137, n. 2) bkrk dk n'm ';nt n'mh, km tsm 'atrt tsmh, dk is of course not a relative but a demonstrative, like all combinations of the deictic particles d and kin the cognate languages. As a relative, d, alone is used in South-Arabian, Ethiopic, Aramaic, Ugaritic, Hebrew () and Phoenician (). Translate, therefore: “This thy firstborn's beauty is like the beauty of ‘Anat, his goodliness is like the goodliness of’ Atirat.”

Incidentally, can ktmsm, I-'AB 1: 24, the epithet of Bn-Dgn (whom I consider to be identical with B'l), signify “the Crown of Glory” ? I have already conjectured that kt, II-'AB 1: 31–2, may bear the same relation to Heb.-Aram. katr-, kitr- as ḫṭ to Heb.-Aram. ḫuṭr-.

^{page 47 note 1} Rossini, K. Conti, Chrestomathia Arabica Meridionalis Epigraphica, pp. 174a, 175bGoogle Scholar.

^{page 47 note 2} It may yet turn out that the mysterious ksm, RŠ., 1929, 1: 9, 3: 19, is simply the corresponding Ugaritic plural absolute.

^{page 47 note 3} Possibly, however, it was an exaggeration on my part to state in my commentary on II-'AB 5: 72–73, Tarb. v, 86, that conversely the plural never occurs without h: b'ltbtm, RŠ; 1929, 19:4 may be identical with b'lt bhtm, ibid., 1: 21; although btm can equally well be dual, or singular with -ma. Incidentally, ṭhrm ‘iḳn'im, which I formerly left untranslated, means “pure lapis lazuli”: 'iḳn'im = Akkad. 'uḳnu; see Brockelmann, C., Lex. Syr. ², p. 674bGoogle Scholar, s.v. (= “blue”).

^{page 47 note 4} Cf. OLZ 1933, col. 594, n. 1.

^{page 48 note 1} The prevailing metre of these texts is 3 + 3, Albright, W. F., JPOS 1932, 207 (23)Google Scholar.

^{page 49 note 1} Cf. Tarb. iv, 390, bot; v, 75, n. 1.

^{page 49 note 2} This example of hd as a synonym of b'l escaped me at the time when I propounded this equation in Tarb. iv, 385, bot., as also when I applied it to restore II-AB 6: 38–9 as follows: bht[h b]'l y'db, hd 'db 'dbt hklh, Tarb. v, 90.

^{page 50 note 1} The vocable yd “love” was first recognized by me, Tarb. iv, fasc. 1 (October, 1932), 113 [8] and 114 [9], and also extended, ibid. 114 [9], 1. 22 f., to the quotation Syria xii, 354, which has since been republished as II-'AB 8: 16–24.

^{page 50 note 2} Cf. with Montgomery, James A., JAOS 1933, 109 [13]Google Scholar, “the Bab. Ḫasis, genius of personified Intelligence.”

^{page 51 note 1} Albright, W. F., JPOS 1932, 194 (9)Google Scholar.

^{page 51 note 2} On the epithet “Virgin”, see Albright, ibid., and now Tarb. vi, pp. 102 ff.

^{page 52 note 1} As the -y of I-'AB 2 is not the prenominal suffix, there is no reason why it should have been replaced by -h here (against V.).

^{page 52 note 2} Virolleaud cites t[ ]. The parallel text, however, suggests that the t is merely the remainder of a n whose last two strokes have been broken off along with the rest of the word ngš.

^{page 52 note 3} Tarb. iv, fasc. 1 (Oct., 1932), ad loc.

^{page 53 note 1} According to the masoretic pointing, the Ḳals of ysr and yny only occur in the participle, but that is because the other forms, whose consonantal orthography suits the Ḳal and the Ḳiṭṭel equally well, were, in accordance with late Hebrew usage, uniformly pointed as Ḳiṭṭels. For the same reason, the Ḳal of dbr, which must have been fairly frequent in Biblical times, is by the Masoretes only recognized once, viz. in Ps. 51: 6, outside of the participle. This point was brought to my attention in another connection by Dr. D. H. Baneth. Cf. AJSL. xlvi, 53–8, 127–138; Tarb. v, 208–223.

^{page 53 note 2} Féghali, M., Groupe linguistique d'études chamito-semitiques, Paris, Résumé de la séance du 25 Janvier, 1933Google Scholar, is perhaps right in assuming that the Arab, 'inab is a North-Semitic loanword, but there still remains South Arabic '‘nb “vineyards” to be accounted for. It would seem that we have here two distinct roots, though it may some day be possible to show that one of them is the result of a contamination of the other with a third; just as Ugaritic and Targumic dr’ “to sow” is certainly a blend of dr' (South Semitic) and zr' (Common Semitic). It is an interesting fact that dr' (zr') and zr' are employed side by side in Arabic and Ethiopic. Some examples of blending in Arabic will be met with presently.

^{page 54 note 1} Contrast whm 'at (not *whmk) trgmy, RŠ., 1929, 13, 43: 5, 15, 19, P. (É.) Dhorme, , RB., 1931, 48Google Scholar.

^{page 54 note 2} So in Sumerian, but of course derived from the inarticulate babbling of babies, like Heb. dód, Eng. daddy, etc. Cf. Epstein, J. N., Jahrb. d. jued.- lit. Gschft. (Frankfort), ix, 277Google Scholar. Etymologically, however, if not materially, Virolleaud's suggestion that 'ad= 'adn “Adonis” comes near the truth, for “master” is evidently connected with 'ad “father”. See now OLZ 1934, col. 473 f.

^{page 54 note 3} From rt. dwb, Duasaud, R., Rev. de l'Hist. des Religions, cvii, No. 1 (juillet-août, 1933)Google Scholar.

^{page 55 note 1} And nḫt, II-'AB 1: 34, must consequently be a derivative or variant of the root nwḫ “to rest”.

^{page 56 note 1} The most decisive argument is kl'at “two”, 1. 57. So far as the outward form is concerned, ‘attm might be singular; cf. Punic , Chabot, J.-B., Punica, pp. 34, 60, 62Google Scholar.

^{page 56 note 2} Tarb. iv, fasc, 1 (Oct., 1932), ad loc.

^{page 56 note 3} Like Akk. mutu.

^{page 56 note 4} See above.

^{page 56 note 5} Or: “a fowl hast thou roasted.”

^{page 56 note 6} 6 The root ṣḥrr, akin to Arabic ṣḥ “to boil”, is already known to us from I-'AB 2: 24; II, 8: 22, in connection with the sun-goddess Špš As in the preceding line, one might also render “hast thou boiled”.

^{page 57 note 1} Alternatively: “Each daughter of 'Il with her 'lm,” supposing we have the well-known Hebrew distributive construction. As btm certainly means “two daughters”, 'il w'lmh cannot mean “'Il and his armourbearer” (against V.), so that it is no example of '= Proto-Semitic g.

^{page 57 note 2} See Albright, W. F., JPOS 1934, 136Google Scholar. I have, however, hesitated to follow him in emending yš'u, 1. 55, to yšḳ and to restore yš(ḳ) in 1. 49. It is just possible that a root nš' existed as a parallel formation to Hebrew nšḳ, nšb, nšp, nšm.

^{page 58 note 1} Chabot, J.-B., Punica, 159Google Scholar.

^{page 58 note 2} Ibid., 153. I take this opportunity of also calling attention to a number of Punic proper names, occurring in the same collection, in which the name of the god Ktr, who plays such a prominent part in II-'AB (cf. above to 'INY, 1. 7), figures as a theophorous element: (55), (68, 82, 154, 160), (76). Chabot also quotes from as., 3261, .

^{page 58 note 3} As corrected by Bauer, H., ZAW 1933, 101Google Scholar, and apparently confirmed by the photograph of this tablet in Syria x.

^{page 59 note 1} The restoration is mine. All I claim for it is a high degree of probability.

^{page 59 note 2} Tarbiz iv, 38 (2), n. 3.

^{page 59 note 3} Brockelmann, C., Grundriss d. vergl. Gramm. d. sem. Spr., Para. 84aGoogle Scholar.

^{page 59 note 4} The same dissimilation is present in South Arabic trt, Gl. 1000 A, 1. 2, which N. Rhodokanakis, Altsabäische Texte, i, p. 61, renders “canaloutlet” (Rinnenöffnung) and connects with Arab. tugrat(un) “crevice, gap, breach”. In this case Ugaritic does not agree with South Arabic; for in Syria xiv, 143, we find b'ap tg'r “before an opening ()”. If, however, Albright, , JPOS 1934, p. 104Google Scholar, is right in connecting Ugaritic b'r (which in my own opinion never has the sense of “burning” in any of the texts published to date) with Arab, bg'r, and Ugaritic Vs with Arab, galîẓ, then we have here so many further examples of this dissimilation in our dialect.

^{page 60 note 1} Dussaud, R., Revue de Vhist. des religions, cvii, No. 1 (juillet-août, 1933), 34 [30] ff.Google Scholar For wld šb'ny aṭt 'itrḫ, however, he gives the impossible translation “the wife of 'Itrḫ bare Šb'ny”, and equates Šb'ny with the Biblical Ishmael. Apart from the syntactical difficulty of such a rendering, the w of wld is, despite the well-known North-Semitic sound-shift, retained in the Hebrew noun , but not in the verb, which is yld.

^{page 61 note 1} This was brought to my attention by Professor J. N. Epstein.

^{page 61 note 2} Taking wyṣ'i as a haplology for wyyṣ'i, representing the intensive conjugation with causative force; cf. above. Until the existence of a '-causative is proved, the only alternative is to take wyṣ'i as the simple conjugation and render the first line “and Trḫ-Ḥdt went out”. In either case 'i again represents vowelless hamza, and the form is, as in Hebrew after waw consecutive, jussive.

^{page 62 note 1} Dussaud probably has Heb. b'r “to remove” in mind, but I incline rather to take yb'r here as the same word which I rendered by “guide” in II-'AB 4–5: 16 and in Virolleaud's ineditum b'l mdlh yb'r “Baal driveshis carriage”. In my note ad loc. I further compared ba'îr, a Common Semitic word designating in the various dialects various kinds of cattle or beasts of burden which are “driven” or “led”.

^{page 63 note 2} Cf. above, 'rbm wtnnm “incomers and outgoers”. Attention is called, however, to Rs., 1929, 13 + 43 (to be joined with Bauer and Dhorme as in Rev. Bibl., 1931, p. 48), 1. 18, where we read wmtnnšmnk. If šmnk means “thine oil”, one might connect tnn with Arab. ntt “to ooze, anoint”. I am inclined, however, to restore a t in the crack between wmtnn in No. 13 and šmnk in No. 48, and to render wmtnn[t]šmnk “and a messenger we have appointed thee”; cf. whm 'at trgmy (f.) in the next line and twice elsewhere in the text in question. We should then either take this tnn as a by-form of tny (which is employed in this dialect, among others, in the sense of “to repeat”, “to tell”) or else explain mtnnt “messenger” as “one sent forth”, or perhaps as “one who utters” (i.e. “outers”, cf. Germ, äussera).

^{page 63 note 1} See Der Gaonäische Kommentar zur Mischna-Ordnung Teharoth, herausg. vonEpstein, J. N., p. 14Google Scholar.

^{page 64 note 1} —V.

^{page 64 note 2} The form of the verb shows that šlm is here feminine, like Akk. šalmu.

^{page 64 note 3} Optative use of the perfect as in Arabic; frequent in these texts.

^{page 64 note 4} (.) means that two words are apparently divided by a space but not by a vertical bar.

^{page 64 note 5} Arab, zbr = Heb. zmr—V.

^{page 64 note 6} i.e. “train”—all this sentence according to Albright, , JPOS 1934, 134Google Scholar.

^{page 65 note 1} There may be a similar play upon words in Isa. 66: 11.

^{page 65 note 2} Cf. Arabic rt. gzr.

^{page 65 note 3} (.) means that two words are apparently divided by a space but not by a vertical bar.

^{page 65 note 4} Cf. Ex. 23: 19; 34: 26; Dt. 14: 21—V.

^{page 65 note 5} Arab. .

^{page 65 note 6} Read ṯamânê, as indicated by Arab., Aram, and Heb. forms of the cardinal numeral 8; cf. 1. 67.

^{page 65 note 7} See V.

^{page 66 note 1} See now Tarb. vi, p. 105, to RŠ 6: 32–39.

^{page 66 note 2} Reading yâniḳîm(a).

^{page 66 note 3} Reading yᾰnῐḳû-ma (3rd pers. impf. plur. + -ma).

^{page 70 note 1} Apparently scratched out by the scribe himself as a dittography of the next word—V.

^{page 70 note 2} Št, is connected by V. with Arab, sitt “lady”, and supposed to refer to 'Aṯirat.

^{page 71 note 1} Read ṯarnânê, corresponding to the Arabic, Aramaic and Hebrew feminine of the cardinal numeral 8; cf. 1. 19.

^{page 71 note 2} Cf. Isa. 29: 1.

^{page 71 note 3} Imper. of vb. 'wd = “return”.

^{page 71 note 4} On 'm cf. Tarbiz iv, 1 (10, 1932) ad I-'AB. 1: 4Google Scholar.

^{page 72 note 1} Arab, šrb “to drink”; Talm.-Heb. and Aram, ŝrp “to sip, suck”.