² For a recent survey of the subject cf. Grohmann, A., Arabien (Handbuch der Altertumswissenschaft, iii.1.3.3.4), München, 1963, 132–1.Google Scholar

³ For the position under Islam cf. Anderson, J. N. D., ‘Homicide in Islamic law’, BSOAS, xiii, 4, 1951, 811–28; in Lihyan, W. Caskel, Lihyan und Lihyanisch, Köln, 1954, 51; amongst Beduin communities, E. Gräf, Das Rechtswesen der heutigen Beduinen (Beiträge zur Sprach- und Kulturgeschichte des Orients, Bd. 5), Walldorf-Hessen, n.d. [1952 ?].CrossRefGoogle Scholar

⁴ cf. N. Rhodokanakis, Die Inschriften an der Mauer von Koḥlān-Timna' (Sitzungsber. Akad. Wiss. Wien, cc, 2), 1924, 13–25. A photograph of the text appears in Pirenne, J., Paliégraphie des inscriptions sud-arabes, Bruxelles, 1956, plate xxxivGoogle Scholar

⁵ Along with several other administrative texts. For the significance of the city gate in this respect cf. Grohmann, op. cit., p. 135 and n. 2.

⁶ cf. Pirenne, op. cit., 207.

⁷ op. cit., 211.

⁸ cf. Rhodokanakis, op. cit., 17, and his Ḳatabanische Texte zur Bodenwirtschaft, ii (Sitzungsber. Akad. Wiss. Wien, cxcviii, 2), 1922, 7.

⁹ op. cit., 234–5.

¹⁰ However, they clearly relate to one another. The second is decreed by the king as mlkn/bn/hgr/ 'š'b/'m/ and not with his full titles as in the first. Moreover, although the first is dated, the second is not but has a long list of witnesses (11. 13–21) which must apply to both clauses.

¹¹ Die Inschriften, 15–16

¹² op. cit., 18–19.

¹³ cf. C. Conti Rossini, Chrestomathia arabica meridionalis epigraphica, Roma, 1931, p. 152a, and add Jamme, no. 723, 1. 4 with bn/ḥrmn/ḥrm ‘from the interdict he incurred’.

¹⁴ cf. Conti Rossini, op. cit., p. 119b.

¹⁵ cf. Rhodokanakis, Ḳatabanische Texte zur Bodentvirtschaft, I (Sitzungsber. Akad. Wiss. Wien, cxciv, 2), 1919, 111.

¹⁶ For ‘Sühne’ Rhodokanakis adduces Hebrew 'br ‘auferlegt werden’, in Deut. xxiv, 5. Cf. Die Inschriflen, 23

¹⁷ cf. Arabic 'abara ‘examinavit (nummos, merces), cuius ponderis et qualitatis essent’. m'br may be equivalent to the modern South Arabian or trial by ordeal. The term recurs in Jamme, no. 669, below.

¹⁸ The palaeography is of type El according to Pirenne, op. cit., 206. Cf. also H. von Wissmann, Zur Oeschichte und Landeskunde von aU-Siidarabien (Sitzungsber. Öster. Akad. Wiss., ccxlvi), Wien, 1964, ‘Genealogische Tafel II’, who dates YD''L BYN about 235 B.C.

¹⁹ ‘A Sabaean penal law’, Le Muséon, LXIV, 3–4, 1951, 305–15. [M. Höfner's definitive edition of this inscription, published in C. Rathjens, Sabaeica, iii, Hamburg, 1966, 21–6, reached the author too late for consideration here. To judge from her convincing interpretation, however, CIH, no. 126 reflects a situation similar to that of RES, 3878.]Google Scholar

²⁰ In the photograph there are traces resembling tg before the s. I can think of no immediate supplementation.

²¹ cf. Exod. xxi, 14 where a man who has killed by accident may seek asylum to escape talion.

²² cf. Gräf, op. cit., 46 where the of the ' can deprive a person found guilty of breaking a tribal law of his rights so that anyone killing him is not liable to talion.

²³ The word does occur in the Sabaean CIH, no. 347,11. 6–8, but in a not very illuminating context. Cf. Conti Rossini, op. cit., p. 258a.

²⁴ cf. Anderson, op. cit., 814.

²⁵ cf. Anderson, op. cit., 811.

²⁶ An examination of the contexts where hrg is used elsewhere, all warlike, suggests that it would imply deliberate killing.

²⁷ cf. Anderson, op. cit., 812.

²⁸ Moreover, as with wrongful and deliberate murder in Islam, seemingly only the killer and not his family would be held responsible. Cf. Anderson, loc. cit.

²⁹ op. cit., 16.

³⁰ An obscure term, cf. Rhodokanakis, Ḳatabanische Texte, i, 70 ff.

³¹ 23. §

³² Deut. xxi, 1–9. The elders must perform a ritual cleansing over a slaughtered heifer.

³³ cf. A. F. L. Beeston, A descriptive grammar of epigraphic South Arabian, London, 1962, § 32: 4.

³⁴ Die Inschriften, 25.

³⁵ On nl cf. Beeston, op. cit., § 43:5.

³⁶ On SM'Y generally cf. von Wissmann, op. cit., 271–384; on Hajar Ẓahra, 302 and 304–5.

³⁷ Around 240 B.C. according to von Wissmann, op. cit., 277–8 and ‘Genealogische Tafel’ opp. p. 280. The roughly contemporary RES, 4626 is dated somewhat earlier by Pirenne, op. cit., 120, 122.

³⁸ ‘Altsabäische Texte, ii’, WZKM, xxxix, 1932, 173–226.Google Scholar

³⁹ Sabaean inscriptions, Oxford, 1937, 75–82.

⁴⁰ basically is to kill by cutting the throat. Nothing of the sort need be supposed here. This verb has in fact become the standard word in Eastern Arabic for ‘to kill’, supplanting qatala which now means ‘to hit’. (Verbal communication from Dr. T. M. Johnstone.)

⁴¹ Rhodokanakis's suggestion that a final b has been omitted in t'l, i.e. the name of the god, is ingenious but unnecessary.

⁴² hrd' is basically ‘to help’. For a parallel semantic development of. Amharic arādda, (c) ‘to apply for compensation (for damage done to one's property)’. Cf. C. H. Armbruster, Amharic-English vocabulary, Cambridge, 1920, 316.

⁴³ Rhodokanakis translates as ‘zehn Jahre’.

⁴⁴ a hendiadys, literally ‘consequences and disputes’. It would seem that should be related to the earlier mḥr/'ršwt/tr't/, so that Rhodokanakis's interpretation, which starts a fresh clause a t is doubtful.

⁴⁵ cf. the Sabaean regulation on sales, BES, 3910, where slaves are on a par with cattle. The interests of the purchaser are paramount. Cf. Grohmann, op. cit., 135.

⁴⁶ So Beeston.

⁴⁷ cf. A. Jamme, Sabaean inscriptions from Mahram Bilqîs (Mârib), Baltimore, 1962, 174–5.

⁴⁸ Jamme dates them to ± A.D. 350–65 (op. cit., 393), von Wissmann to A.D. 370 (op. cit., ‘Genealogische Tafel III’).

⁴⁹ A statue in silver and an inscription, their combined weight being 'sym, and a statue in gold. 'sym may be a weight as Jamme interprets it, but hardly after Arabic ‘usūm 'stones’. (In his ‘Glossary’ it is listed under root 'sy.) I prefer a comparison with Arabic ‘asⁱⁿ 'aptus, conveniens’, thus ‘verified’ or the like.

⁵⁰ Better than Jamme's ‘entered’. 'dw/'ly- implies a hostile act.

⁵¹ Jamme translates, ‘he was left dying’, presumably with myt as passive, as indeed it is listed in his ‘Glossary’. But in the commentary he notes that myt = mwt. This last is preferable, cf. Beeston, Descriptive grammar, § 12:1 for occasional interchange of w and y. māta in the passive could only mean ‘there was a dying’.

⁵² Lit. ‘at the hand of their son’, i.e. an unspecified member of the tribe. Jamme takes bn as plural.

⁵³ Jamme, ‘condemned them publicly’. This specification is not contained in which merely means ‘to proclaim’, cf. Beeston in Le Muséon, LXV, 3–4, 1952, 265–0.

⁵⁴ Jamme renders ‘explanation’ after Arabic mu'abbar.

⁵⁵ cf. Arabic wasaala ‘rem peregit, qua ad Deum propius accederet c. [ilā] p., eique gratior fieret’. For the loss of the initial w see below on Jamme, no. 700, 1. 6, hẒmn. Jamme's ‘drive’ is difficult with the preposition b.

⁵⁶ See also below on the .

⁵⁷ In CIH, no. 126, there is a provision for a proclamation of outlawry in a temple.

⁵⁸ cf. Jamme, op. cit., 190–1.

⁵⁹ cf. the forms of š, m, r, etc., as exemplified in the photographs of the texts in Jamme, op. cit., plate 19 (Jamme, no. 669) and plate 27 (Jamme, no. 700).

⁶⁰ For the correspondence of ESA : Arabic ḥ, cf. Beeston, Descriptive grammar, § 5:2.

⁶¹ Erroneously for 'ṭṭen (<ṭan), cf. Müller, W. W. in Archiv für Orientf., xxi, 1966, 108.Google Scholar

⁶² cf. Müller, op. cit., 106.

⁶³ Unless from Arabic ṣāmma ‘abstinuit cibo, potu, …’, when it would be interesting, if hazardous, to compare Qur'ān, iv, 94.

⁶⁴ cf. Beeston, op. cit., § 23: 4.

⁶⁵ cf. Beeston, op. cit., § 2: 4, who feels that Old South Arabian may have been vocalized after the pattern of and the modern South Arabian languages rather than Arabic.

⁶⁶ The Tāj alone gives as an authority Ibn al-A'rābī.

⁶⁷ cf. R. Dozy, Supplément aux dictionnaires arabes, Leiden, 1881, s.v., who notes that the word properly means ‘vituperavit’.

⁶⁸ laḥama and alhama, ‘iunxit rent rei’.

⁶⁹ The Qāmūs defines as qaḍīb qdbl an yuṣlaḥ, but they cannot be synonymous here.

⁷⁰ cf. in Liḥyanite ḥqwy ‘die beiden Einschnürzungen, Seiten (des Einganges), eines Stollengrabes’, Caskel, op. cit., 134.

⁷¹ The term also occurs in a very obscure context in Nami, no. 74. Cf. Beeston, in Le Muséon, LXV, 1–2, 1952, 146, who translates it ‘sacrilege’.

⁷² For a general discussion of asyndetically linked clauses in Semitic, cf. Brockelmann, C., Gruniriβ der vergleichenden Grammatik der semitischen Sprachen, Berlin, 1908–1913, ii, 471ff.Google Scholar

⁷³ For -hw as feminine, cf. Beeston, Descriptive grammar, § 37: 3.

⁷⁴ Or ‘by’ with Jamme.

⁷⁵ In CIH, no. 619, they exclude their 'dm/w'mh/ from burial in the family tomb.

⁷⁶ On the mqtwy cf. J. Ryckmans, L'institution monarchique en Arable méridumale avant l'Islam, Louvain, 1951, 145–6.

⁷⁷ For NŠN cf. von Wissmann and M. Höfner, Beitrāge zur historischen Geographie des vorislamischen Südarabien, Wiesbaden, 1952, 14–16.

⁷⁸ A woman described as 'mt is always 'mt of t h e whole tribe, never of a n individual, and so far as one can judge, she is never a member of that tribe originally.

⁷⁹ In Islam diya, as opposed to talion, was the responsibility of the killer's 'āqila in cases of wrongful, non-deliberate homicide. Cf. Anderson, op. cit., 814.

⁸⁰ Several of the penitential texts from Haram are related to blood pollution, for example CIH, nos. 548 and 568.

⁸¹ Caskel, nos. 31 and 82.

⁸² Published in Boissonade, J. F., Anecdota Graeca, v, Paris, 1833, 63–116. For a summary cf. R. Dareste, Lois des Homérites, in Nouw. Rev. Hist, de Droit Français et Étranger, LIX, 1905, 157–70.Google Scholar

⁸³ cf. Dareste, op. cit., 160. The punishments at least are typically Arab.

⁸⁴ Boissonade, op. cit., 78.

⁸⁵ The term itself appears simply to refer to homicide.

⁸⁶ Boissonade, op. cit., 73–4.

⁸⁷ See my ‘Four Sabaean texts in … Istanbul’, Le Muséon, LXV, 3–4, 1952, 279.Google Scholar

⁸⁸ w and consecutive imperfect, see my Descriptive grammar, § 52:9.