^{page 420 note 1} cf. RES, 3695, 1 (Min. from al-'Ela): wm'n/bkln//w'b/bm'mtm/ … ‘et Ma'ān de tout’. bkln should be related to bkland not to kl, which normally is in the construet state with the noun it ‘qualifies’ cf. M. Höfner, ASG, 137. According to the note in RESthe next word appears in the squeeze as wwbl.From the photograph it is possible to read what follows bkln as /wwbln/bm'tm/'n/bkln/wwbln// 'and Ma'īn: the property-owning residenta and the rentpayers (or, the camel-drivers, cameleers)'. For wbln ‘rentpayers’, cf. RES,2695, 3, and Rhodokanakis, Studien, II,152–3; for ‘cameleers’, cf. wabala ‘to strike, to hit, e.g. a horse, with a stick or whip; to drive strenuously’ and wabil ‘a big stick’ used, according to a citation, for driving a she-camel. (Lisān, art.wbl.)In either case wblnwould mean temporary inhabitants, those engaged in caravan transport and trade. Cf. also RES,3285, 2: w'hrhn/wbkln ‘and the ruler and the residente’. (For 'hr ‘ruler’, cf. Arābic 'āhil.)

^{page 421 note 1} For the most recent discussion of Ẓrbcf. Höfner, M., UTQK, 75–9;Google ScholarJamme, A., PEHA,185–9;Google ScholarJamme, A., IHYT, 184–6.Google ScholarRhodokanakis, , Studien, III, 49,Google Scholar compares Ẓrb with Arabic Ẓariba: laṢiqa ‘it stuck ’, and with Ẓurriba,said of a hoof, ‘it became hard and strong’. It is more appropriate to compare with it Arabie darabaand 'iḍtaraba (form VIII): 'Hktasaba ‘he gained, earned, came to possess (something)’ Tāj, art. ḍrb.(According to al-Ṣāghani, quoted in Tāj, it should be read 'iṢtaraba, with Ṣāḍ. Cf. Ṣaraba: kasiba; Tāj, art. Ṣrb. For the interehange of Ẓ and Ṣ, cf. Beeston, A. F. L., ‘Notes on old South Arabian lexicography, III’, Le Muséon, LXIV, 1–2, 1951, p. 132 and n. 26.) One may also compare with Ẓrb here ḍaraba 'alā yadihi ‘he struck his (i.e. another man's) hand; meaning: he struck, or made, tne bargain with him; or ratified the sale with him’ Lane, Lexicon. Compare with this usage of ḍaraba Arabie safqa ‘a striking of the hand of one person upon the hand of another in ratifying a sale or purchase and a covenant’. Ṣafqahas come to mean a final deal; cf. al-bai'u Ṣafqat-un 'aw hiyār-un ‘selling is decisive or with the option of returning’ (Lane, Lexicon).Google Scholar

^{page 421 note 2} Or fa'īl? Cf. Ẓrbt, pl. Ẓrwb, RES, 3854, 2 and 3. This passive participle is known in bḍ ‘wounded’ and ’ẖḏ ‘captive’ cf. Conti Bossini, ‘Glossarium’, 117 and 102respectively.

^{page 422 note 1} To accommodate bḥgwith his translation of Ẓrbas ‘to offer’, Rhodokanakis maintained that bḥg indicated in relation to the gods whose names followed what l indicated in relation to human beings; Studien III, 13, 29. He assigned to it the sense of ‘nach der Gebiihr’, which he explained by saying, ‘D.h. so wie zusteht, gebiihrt dem’ Ibid., 35, n. 8. M. Höfner, who admits that Rhodokanakis's interpretation of Ẓrb as ‘widmung’ can no longer be maintained, says, however, 'An dieser Auffasung des Verbums Ẓrb ändert auch der Passus b-ḥg 'nby nichts, der in keiner dieser Urkunden fehlt. Er bedeutet “nach der Gebühr, nach dem Recht des (Gottes) ‘Anbay”’, UTQK, 80. The latter rendering, ‘nach dem Recht’, seems to be M. Höfner's own suggestion. A. Jamme, apparently following this rendering, translates bḥg in his 118 and 119 as ‘according to the law’ IHYT, 186, 188, 183 f. bhg can only precede a noun, in a genitive relation; cf.RES, 2865, 2: cf. also Jamme, 540, 2–3: ‘jusqu'à ses limites et selon ce qu'a écrit Damarwaqah’, Jamme, A., ‘Inscriptions des alentours de Mareb (Yemen), I’, Cahiers de Byrsa, V, 1955, 269Google ScholarBoneschi, f. P., ‘Tres tituli sabaei iterum interpretati’, RSO, XXXIV, 1–2, 1959, 31, renders wbhg/str in Jamme, 540, 2–3 (above) ‘et sicut delineavit’, making bhghere a conjunction. This is not correct.Google Scholar

^{page 422 note 2} cf. the note on ḥngn, Van Lessen 25, 4–5, infra, p. 428 f.

^{page 424 note 1} It is doubtful whether Ṣrḥtin such inscriptions arways means‘ upper chambers, or super sructure ’. Cf. Arabic Ṣarḥatu al-dār: Ṣāḥatuha wa ‘arasatuha’ the open space, the courtyard, Tāj, art. sr. Cf. also Ṣarḥat [sic] ‘courtyard’, in R. B. Serjeant, ‘ A Judeo- from ḥabbān’, JRAS, 1953, 125 f. The mention of open areas and court is quite common in Islamic legal practice; cf. Adolf Grohmann, Arabic ‘Inscriptions Library, I, Cairo, 1934, Text 67, 12, p. 238. Verbs denoting construction ‘Tres tituli’. for according to Arabic lexica, a Ṣarḥa has to be made level to be called 540, 2–3 (above) art. Ṣrḥ. Nowadays Ṣarḥ;a is the term in Hebron for a piece of levelled,

^{page 424 note 2} cf. the note on ṇ

^{page 425 note 1} A. Jamme reported four big rock inscriptions from ‘Gebel Mġanimeh ’ in his ‘Les inscriptions rupestres de la region de Mukéréâs’, Bulletin de l'Academie Royale de Belgique,5me Sér., Tom. XXXVII, 1951, 308. My colleague Mr. H. T. Norris tells me that it is the same place in Shurjān as that from which some of the Van Lessen inscriptions 21–9 come. I cannot say for certain which of Jamme's individual numbers for the inscriptions correspond with Van Lessen's individual numbers.

^{page 425 note 2} This was before Major Van Lessen sent to me his no. 29, which is a longer text, though not so clear.PP

^{page 426 note 1} Smith, Sidney, ‘Events in Arabia in the 6th century A.D.’, BSOAS XVI, 3, 1954, 439.Google Scholar

^{page 427 note 1} cf. R. LeB. Bowen, ADSA, 45, col. B. Cf. Rhodokanakis, Studien, II, 43, for a discussion of the different senses of grb in the insoriptions. The two words munhama and jurüb occur next to each other twice in two lines of verse by 'Alqama ḋū-Jadan describing the structure of the palace or fortress of Ġumdān. One verse occurs in Ṭabari, Tārīkh, I, 949, which reads: bimunhamat-in w'asfaluhu jurūb-un; the other (Hamdānī, Iklīl,VIII, 15—for mubhama read munhama) reads: 'a'lāhu munhamat-un ruẖām-un 'āl-in wa 'asfcduhu jurūb-u. The latter verse may be translated: ‘Its top is raised high with smooth blocks, munhamat-un, of marble, and its bottom (foundation) is made of jurūb’. This last word is explained as ḥijārat-un maqṬu'a; cf. Rhodokanakis, Ibid. I think that jurūb in 'Alqama's description referred, originally at least, to the terrace shape of the bottom of the walls. A parallel may be found in the method of setbacks used in the foundations of the temple of 'Awwām in Ma'rib; cf. F. P. Albright, ADSA,217, col. B. It seems that the use of roughly dressed stones in the building of this kind of foundation caused the transfer of the meaning of the word to the blocks of stone themselves.

^{page 427 note 2} I owe this information to the kindness of Professor R. B. Serjeant.

^{page 427 note 3} wh]qh:cf. RES, 3854, 2; cf. also Arabie tawqīḥ al-ḥawḍ ‘to repair a tank (or, to perfect it) with mortar, plaster, until it hardens (becomes watertight) so that water would not dry up (leak)’ (Tāj). b'ly ‘above’ cf. Beeston, A. F. L.,‘Notes on old South Arabian lexicography, I’, Le Muséon, LXIII, 1–2, 1950, 53Google Scholar. It is only natural that an irrigation reservoir should be higher than the land it is meant to irrigate. šrgn: cf. RES, 3878, 18: The present-day name of the place. In Hamdānī's Ṣifat jazīrat al-'Arab (ed. D. H. Müller, 1884), p. 96, it is read cf. the variant reading given by the editor, Ibid., II, 98. Hamdānī's text reads: The same inseription, Van Lessen 23, 10, reads: wbql/wynhmw/n'mn.

^{page 428 note 1} cf. CIH, 982, 2: ḥrthw/msqt/ nẖlhu / ‘his canal for the irrigation of his palm garden’ cf. Beeston, A. F. L., ‘Notes on old South Arabian lexicography, VI‘, Le Muséon, LXVII, 3–4, 1954, p. 311, n. 2.Google Scholar

^{page 428 note 2} Landberg enters mājil under mjl; Glossaire , III, Leiden, 1942, 2676.Google Scholar Arabio lexica also mention that some scholars entered it under mjl; cf. Lisan, art. 'ji.

^{page 428 note 3} ‘Notes on old South Arabian lexicography, I’, loc. cit., 53.

^{page 428 note 4} cf. E. LeB. Bowen, ADSA, 8, col. A.Google Scholar

^{page 428 note 5} loc cit., 54 and n. 2; Jamme, A., ‘L'inscription hadramoutique Ingrams I’, Le Muséon, LXIX, 1956, p. 99, n. 3.Google Scholar

^{page 428 note 6} Such a rule of not writing the nūn sākina before another consonant is known in the orthography of the Qur'ān; cf. Rabin, C., Ancient West-Arabian, London, 1951, 146Google Scholar f.; cf. also Kofler, Hans, ‘Reste altarabischer Dialekte’, WZKM, XLVII, 1–2, 1940, 80 f.Google Scholar

^{page 429 note 1} According to al-Laiṯ, quoted in Tāj, and to al-'Azhari, quoted in Lisān, some people of ḥims (in Syria) pronounced ḥaẒẒ as ḥanz but in the plural went back to ḥuẒūẒ. It is interesting and instructive to read in Hamdānī, Ṣifat jazīrat al-'Arab, 132, that ḥimṢ was ḥimyariyya, apparently meaning that the majority of the population (or, of the Arab population ?) were ḥimyarites. (This statement, however, cannot be taken at its face value, since by ḥimyariyya he might have meant that the Arabs were of Banu Kalb of Quḍā'a, whom he mentions in the same line, and who are claimed by some genealogists to be a branch of ḥimyar.)

^{page 429 note 2} cf. Jamrae, A., ‘Les inscriptions rupestres de la région de Mukérâs’, loc. cit., 314.Google Scholar

^{page 430 note 1} cf. Rhodokanakis, , KT, I, 35,Google Scholar n. 4; Ryckmans, J., L'institution, 16Google Scholar, n. 12. The reading mlky/m'n is now established beyond doubt through the text based on M. Tawfik's photograph and edited by Kh. Y. Nāmi, 'NuqūshKhirbat Barāqish ‘ala daw’ majmū'at Muḥammad Tawfīk (al-majmū'a al-ṯāliṯa)', Bulletin of the Faculty of Arts, Cairo University XVII, 2, 1956, 34–6Google Scholar. Professor Nāmi comments on wb/mr's/šhr/ygl/yhrgb/mlk/qtbn as follows: (I translate from his Arabie) ‘Undoubtedly the pronominal suffix in mr's does not refer to the two preceding names, which are: WQH'L YT' and his son ‘LYF’ YŠR, the two kings of Ma'īn. I hold, therefore, that the suffix refers to the dedicant of the inseription, (YŠRḥ'L); and although his name is distant (at the beginning of the inseription), yet it must have been near or present in the mind of the seribe (al-kātib) because he (YŠRḥ'L) was the dedicant of the inseription. It is possible, therefore, to say that the “people of ḍMRN” were Qatabānians who were living in the town of YṮL, and that must have been why they dated their document with the reign of the two kings of the town where they lived and made their living, as well as being the reason why they ended their document with the name of their king to whom they owed allegiance' (Ibid., 36). Professor Nāmi's article was made available to me, through his kindness, only when this article was in an advanced stage in the press. Although I do not find that I can agree with Professor Namī's explanation of the status and circumstanees of the dedicant of RES, 2999, in Yaṯīl, yet it is clear that there is a lot of common ground between his view and mine: in both views it is maintained that RES 2999, did not indicate that a king of Qatabān was then the suzerain of Ma'īn in Yaṯīl.

^{page 431 note 1} See Appendix II, infra, pp. 436–8.Google Scholar

^{page 432 note 1} I am using the term ‘citizens’ in preference to ‘members of the tribe’ because it is becoming clear that a tribe, š'b, need not always mean a kindred. And when the š'b in question happens to be the ruling one, or that of the rulers, first-class members of the community need not all be blood relations. Cf. Sidney Smith in his review of Ryckmans, J., L'institution monarchique en Arabie mériditmale avant l'Islam, Vetus Testamentum, II, 3, 1952, 287.Google Scholar

^{page 433 note 1} A. F. L. Beeston, Sabaean inscriptions, 28, says, in comment on , ‘Marg(oliouth) suspects gilding to be really meaut, on account of the huge expense of thirty golden statues’. This may be preferable to rendering ḍhbas ‘bronze’ cf. Jamme, A., ‘Sabaean inscriptions on two bronze statues from Mate;rib (Yemen)’, JAOS, LXXVII, 1, 1957, 32,Google Scholar n. 10. Ryckmans, G., Le Muséon, LXX, 1–2, 1957, 106Google Scholar, hesitates between bronze and gilded bronze. Kensdale, W. E. N., The religious beliefs and practices of the ancient South Arabians, Ibadan (Nigeria), 1955, p. 6Google Scholar, claims that in ‘Al-Iklīl Book VIII the word ḏahab occurs in many contexts in which it plainly refers to bronze’. Further, Ibid., p. 7, he states that in ‘Al-Iklīl VIII we are told of inscribed bronze … tablets (126, 149, 157), … a bronze lined coffin, … a bronze chest (149), a bronze staff (126, 149)’. (The pagination is supplied by me.) In all these cases ḏahab is used in the Arabie text. All the cases given above occur in legendary accounts of finds in ancient burial-places where the legends could have referred to nothing but gold. Hamdānī even mentions al-ḏahab al-qubūriy ‘burial-places gold’, and names the sites where it was commonly dug out; Ibid., 108. It is true that ḏahab 'aḥmar ‘red gold’ is mentioned twice there, 48 and 155, but that could not exelude ḏahab in the same work from meaning gold. And ḏahab unmistakably meane gold in the contexts on pp. 129, 143, and 183. Besides, mention is made there of statues made of nuḥās 'aṢfar ‘brass, bronze’, 19, and of nuḥās,meaning the same, 19, 20, 204; and we read also of inscribed tablets of Ṣufr ‘brass, bronze’, 94. The use of qiṬr ‘copper’ for joining masonry is mentioned there, 16, 38, 44, 47, and 198. (This practice is confirmed by the report of Ahmad Fakhry, Archaeological journey to Yemen, Cairo, 1952, I, 68.) If the legends, or if Hamdānī had wanted to denote a metal other than gold where ḏahab was used, clearer and less ambiguous terms were available and could have been used.Google Scholar

^{page 433 note 2} CIH, 521, 527, 535, 536; BES, 4142, 4143, 4144, 4145, 4147, 4229, 3902 bis, 138; Ry., 367. Cf. also Ryckmans, G., ‘ Heaven and earth in the South Arabian inscriptions’, JSS, III, 3, 1958, 228Google Scholar f.; cf. also Rossi, , 17 (Jamme, 513), RSO, XXX, 1–2, 1955, 110.Google Scholar

^{page 433 note 3} CIH. 517. 518. 521.

^{page 434 note 1} cf. Nielsen, D., Handbuch der altarbischen Altertumskunde, Kopenhagen, 1927, 81.Google Scholar

^{page 434 note 2} cf. the note in CIH, II, p. 243, on CIH, 528, 3.Google Scholar

^{page 434 note 3} Nielsen, D., Der dreieinige Gott, Kopenhagen, 1942, 131.Google Scholar

^{page 434 note 4} Ibid.

^{page 434 note 5} Le Muséon, LXII, 1–2, 1949, 64.Google Scholar

^{page 434 note 6} ‘Heaven and earth’, JSS, III, 3, 1958, 226 f.Google Scholar

^{page 434 note 7} Der dreieinige Oott,124.Google Scholar

^{page 434 note 8} cf. Ryckmans, G., ‘Heaven and earth’, JSS, 3, 1958, 226.Google Scholar

^{page 434 note 9} Ibid., 227.

^{page 434 note 10} Ibid., 227 and 230.

^{page 435 note 1} Ibid., 230.

^{page 435 note 2} cf. the author's ‘New Qatabani inscriptions’, BSOAS, XXII, 1, 1959, 6.Google Scholar

^{page 435 note 3} cf. Ryckmans, G., ‘Heaven and earth’, JSS, 3, 1958, 230.Google Scholar

^{page 435 note 4} The mention of 'Amm might have been a polite gesture to the ‘host’ community rather than a subordination of ḎSMWY to ‘Amm. This seems to find support in the clause’ nby/wnkrḥm/rd'w, Van Lessen 9, 6. Yet it is clear that the Minaean community in Timnā' must have enjoyed a higher standing than that of ḎSMWY whose members are described in Ry., 367, 1, as "dmn/.

^{page 436 note 1} cf. the note on qbrhmiv, Van Lessen 9, 2, above. Cf. for a comparable case, Hartmann, M., Arabische Frage, Leipzig, 1909, 177, in connexion with the h- and s-dialects in the inscriptions of Kaminahu.Google Scholar

^{page 438 note 1} cf. Akkadian erēbu ‘to enter, to penetrate’. renders, M. Hofner 'rb here ‘er aus Quader-steinen baute’ WZKM, XLIII, 1936, 78.Google Scholar Cf. also Rhodokanakis, , Die Inschriften an der Mauer von Koḥlān-Timna', Wien, 1924, 34, 36, in connexion with 'rb in SE 94, 3, 4. I hold that a sense of ‘went to’ or ‘went into’ should be assigned to 'rb in SE 94 as well. The text of SE 94 reads: I translate: ‘ whereas (ro, since) his chief priest, SRY'M, son of MRQDM, had offered (in fulfilment) (a) a lamb according to the oracular order of 'Amm and ḥWKM, when ‘Amm had given an oracular order concerning a lump of (boneless) meat (b) and ḥWKM (had given his order) concerning a grilled (or, fried) (piece of meat), (c) and had gone with the gifts (in)to (the temple) QLBN (d)and gone (in)to the cella in (the temple) BN'’.Google Scholar

(a) sṬd: cf. Arabie 'Hstanaḍḍa ḥaqqahu: 'istanjazahu ‘he demanded the payment, the fulfil-ment, of his due or right’ tanaḍḍaḍtu ḥaqqī: 'istawfaytuhu ‘I secured the payment, the fulfllment, of my due‘. sṬd here must be form IV from the root nṬd.

(b) n'ḍ: cf. Arabie qabā: ‘meat; a big lump of meat, e.g. a leg; a lump of boneless meat enough for several persons’.

(c) qly: cf. Arabie qala ‘to fry, to grill, meat’. For the offering of parts of a victim cf. RES, 4782, 4784.

(d) qlbn: it is also the name of a temple in Van Lessen 31, an unpublished Qatabāni inseription on a slab of limestone in the possession of Mr. T. W. Hague, by whose kind permission photographs of it were made. It reads:

The names of these deities are repeated in the same words and order in line 3 and again in lines 4–5. The invocation at the end reads: