^{page 18 note 1} Dr. J. Walker, Keeper of the Department of Coins and Medals in the British Museum, has been kind enough to discuss with me some of the numismatic difficulties which have arisen in connection with this article. I should like to express my gratitude to him. I am no less indebted to Professor A. F. L. Beeston of St. John's College, Oxford, who read the typescript of the article and made some valuable suggestions on it, many of which have been incorporated.

^{page 18 note 2} See Hill, G. F., Catalogue of the Greek Coins of Arabia, Mesopotamia and Persia, British Museum, London 1922, pp. xliv–lxxxivGoogle Scholar. According to Hill the oldest pieces can scarcely be earlier than the third century b.c. (p. xlviii) while his latest he would assign to the period circa 50–150 a.d. (p. lxv). For some later items of Ḥaḍrami origin, all copper, see below, p. 28, n. 4.

^{page 19 note 1} This seems to me the most feasible interpretation of this difficult text. Earlier ones varied considerably; cf., for example, Rhodokanakis, N., Der Grundsatz der Öffentlichkeit in den südarabischen Urkunden, SBWA 177/2, Vienna 1915, pp. 7–8Google Scholar, and Beeston, A. F. L., Sabaean Inscriptions, Oxford 1937, pp. 44–6Google Scholar. The text is unlikely to be a statement of debt (so Rhodokanakis) since a loan as such is a temporary arrangement destined to be terminated at a later date. We may however see the text as a title to land dependent on some monetary arrangement. The withdrawal of 'B'LY is not explicitly stated but may be inferred from the absence of his (or her) name from the transaction which gave rise to the inscription, cf. 11. 1–4 as against 11. 4–9.

^{page 19 note 2} Paléographie des inscriptions sud-arabes, Brussels, 1956, p. 207Google Scholar. A partial photograph may be seen in the same work, plate XXXI.

^{page 19 note 3} Unfortunately neither CIH 376 nor CIH 73 mention any royal names and dating must be done by external criteria. Hommel, F. (in Nielsen, D., Handbuch der altarabischen Altertumskunde, Copenhagen 1927, p. 96, n. 2Google Scholar) notes two sphinxes in the panel above the latter text, which suggest to him an Achaemenid influence. (See also Grohmann, A., Göttersymbole und Symboltiere auf südarabischen Denkmälern, DKWA 58/1, Vienna 1914, pp. 67–70 andfig. 181Google Scholar.) However, Hommel's suggestion of a date around 400 b.c. is certainly too early. Pirenne, , La Grèce et Saba, Paris 1955, p. 57Google Scholar, argues that the panel is Phoenico-Persian in inspiration but she does not hazard a date. I can lay no claim to expertise in the field of palaeography but the reproduction of CIH 73 provided in the Corpus shows a ductus somewhat later than those described by Pirenne. In general it is rather similar to those in vogue in the period of the kings of SB' and ḏRYDN.

^{page 19 note 4} Chrestomathia arabica meridionalis epigraphica, Rome 1931, p. 114A.Google Scholar

^{page 20 note 1} See also D. H. Müller in ZDMG 29 (1876) p. 612, J. H. Mordtmann in ZDMG 46 (1892) p. 323, and Beeston, op. cit., p. 12. J. Halévy, in JA 1873 pp. 344–6, had suggested for the passage in CIH 73, “en poids juste, mot à mot en pierres acceptables”, after the sense of the Arabic balāṭt “pavement, paving stone”, but Mordtmann dismissed this on the grounds that one does not weigh with stones. Be that as it may, caution must be exercised in comparing balāṭ since it may be a loan-word from Latin platea; cf.Landberg, C., Glossair daṭinois, Leiden 1920–1942, p. 206Google Scholar.

^{page 20 note 2} Op. cit., p. 613.

^{page 20 note 3} Ḳatabanische Texte zur Bodenwirtschaft, II, SBWA 198/2, Vienna 1922, p. 25, n. 6Google Scholar.

^{page 20 note 4} Op. cit., p. 26.

^{page 20 note 5} Op. cit., p. 180A. In RSO IX (1923) pp. 603–4 he further defines it as “moneta a pieno titulo”.

^{page 20 note 6} Op. cit., p. 44. One might note in passing a suggestion made by D. S. Margoliouth and recorded by Beeston, op. cit., p. 45, to the effect that mṣ'm might be related to mṣ' of the Aramaic inscriptions in the sense of “middle”, i.e. “a coin of ‘normal’ weight (like [Arabic] mu'addal”. This word however seems to have a purely spatial reference, cf. Jean, C. F. and Hoftiyzer, J., Dictionnaire des inscriptions sémitiques de l'ouest, Leiden 1960–, p. 164Google Scholar.

^{page 20 note 7} These are discussed in detail by Beeston, op. cit., p. 12.

^{page 20 note 8} For myrn Beeston, loc. cit., notes Arabic mīra “corn”, imtāra “to purchase corn”, and māma “to supply with corn”. Hence he translates, “the purchasing (of corn)”. But “selling (price)” would be more appropriate in the context. Various other derivatives of the root abound in the texts published by Jamme, A., Sabaean Inscriptions from Maḥram Bilqîs (Mârib), Baltimore 1962, but in his note on p. 113BGoogle Scholar Jamme ignores the present text, brm is probably not a monetary unit, as is suggested by those editors who read ḏḏhbn. There is no satisfactory etymology for br or brm in this sense and the reading of ḏ is doubtful, br however occurs elsewhere, e.g. CIH 540/40, in the sense of “wheat”. For ṭmn, “an eighth”. i.e. a technical measure, cf. Arabic tumn and Hinz, W., Islamische Masse und Gewichte, (Handbuch der Orientalistik, hrsgb. Spuler, von B., Ergänzungsband I, Heft I, Leiden 1955), p. 52Google Scholar.

^{page 21 note 1} Cf. Beeston, , A Descriptive Grammar of Epigraphic South Arabian, London 1962Google Scholar, §39: 4. To take blṭtm as, for example, a plural of abundance would deprive the statement of any point. In RES 4765/5, wb/blṭ/sṭry/bhw/'lm/w[., “ and with the blṭ, in respect of which they (dual) have written a document of (?) …”, blṭ appears to have a masculine singular ‘āid in bhw. If it is the same word as that under discussion, the gender and number concord will be irregular for Old South Arabian. Granted that blṭt has a monetary reference, I am inclined to see biṭ in RES 4765/5 as a masculine singular form with slightly different meaning from blṭ, plural of blṭt, possibly a collective form with the sense of “money” or the like.

^{page 21 note 2} Cf. Mordtmann, and Müller, , Sabäische Denkmäler, Vienna 1883, p. 76Google Scholar: “rrḍy … heisst eigentlich ‘gangbar, annehmbar, gut’”.

^{page 21 note 3} See Hill, op. cit., pp. 54–63 and plates VIII. 1–X. 11. The type with Arab head is dated circa 100–24 b.c. and that with Augustan head circa 24 b.c.-A.D. 50.

^{page 21 note 4} Cf. RSO IX (1923) p. 603.

^{page 21 note 5} He seems however to read too much into Rhodokanakis's remark (op. cit., p. 26, n.4). The latter merely considers that the verb ballaṭa may be established as native to Arabic on the evidence of the Old South Arabic noun, not, I think, that blṭ produced ballaṭa. That the coin name was known to Arabic at some period is however suggested by the verb ablaṭa = aflasa “to go bankrupt”, i.e. “to be deprived of biṭ or fals”. I am grateful to Mr. M. A. Ghul for bringing this to my attention. Similarly belṭā is alleged by Castell to occur in Syriac as “moneta”. I suspect a non-Semitic origin for the root, see below p. 23.

^{page 21 note 6} Hill does not specify the nature of the wreath. Schlumberger, G., Le trésor de Sar'â, Paris 1880, p. 10Google Scholar, states that it is an olive wreath. The oak tree is not to my knowledge mentioned by any writer on South Arabia.

^{page 22 note 1} Cf. Talmudic bālūṭ “acorn, nut; oak” (Jastrow, M., Dictionary of the Talmud, New York/Berlin 1926, p. 171B)Google Scholar.

^{page 22 note 2} The general appearance of the coinage is not Nabataean and any similarities in the heads need not be explicable in numismatic terms.

^{page 22 note 3} Much controversy has raged around the interpretation of this legend, which is not confined to the later Attic type but occurs on some of the older Attic coins as well, cf. Hill, op. cit., pp. 53–6. The same author, op. cit., pp. liii–liv, likens the script to Characenian Aramaic, but J. Walker, RSO XXXIV (1959) pp. 77–81, has conclusively shown that it is Lihyanite and should be read šhr hll. These particular coins may thus be Qatabanian and would presumably have been used in the northern trade.

^{page 22 note 4} e.g. Mordtmann, op. cit., p. 323, “eine Münzgattung”, and most recently Jamme with “biṭ, . . . a South-Arabian coin” (op. cit., p. 124A).

^{page 22 note 5} Jastrow, op. cit., p. 172A.

^{page 22 note 6} Cf. Schrötter, F. von, Wörterbuch der Münzkunde, Berlin 1930, p. 481Google Scholar: “wegen des Athenakopfes auf der Vs. hiessen . . . die athenischen M., doch wohl vorzugsweise die AR-Tetradr., Pallades.” He gives a reference to the Onomasticon of Pollux, IX, 75. The Greek-English Lexicon of Liddell and Scott (9th ed., Oxford 1940) cites only Eubulus, an Attic comedian of the fourth century b.c., but even that is sufficient to support the word's currency in colloquial Greek.

^{page 23 note 1} And also of the Arabic and Syriac words cited above, p. 21, n. 5. R. F. Burton mentions finds of imitations of the Attic tetradrachms in North Arabia, cf. The Land of Midian (Revisited), London 1879, I, pp. 92–3Google Scholar. The coins are discussed by Hill, op. cit., pp. lxxxv–lxxxvi.

^{page 23 note 2} Gold coins from South Arabia are excessively rare and Hill lists one specimen only (op. cit., pp. lvii and 54), though noting a second in Vienna. A further specimen has since come into the British Museum. The coin described by Hill is of the later Attic type and the one acquired subsequently is of the same type but a double of the former. C. M. Kraay has kindly informed me that there is a similar piece in the Ashmolean Museum, Oxford. The Vienna coin is of the late series with profile head on obverse and reverse. Despite this apparent shortage, of course, it cannot be assumed that gold was uncommon in ancient times. At a relatively late date the Martyrium Sancti Arethae mentions a Himyarite gold coin, λ;ϰσ or δλϰάσ. The actual coin in this case is unknown but it may have been Aksumite. See generally Schlumberger, op. cit., p. 6, n. 3, and Ryckmans, J., La persécution des Chrétiens himyarites an sixième siècle, Istanbul 1956, p. 5, n. 21Google Scholar.

^{page 23 note 3} Much as the Latin nummus could refer to the sesterce, though blṭ may have retained its associations with the tetradrachm.

^{page 23 note 4} Note that payment in any of the lower denominations would result in an overall loss in weight.

^{page 23 note 5} Le Muséon 69 (1956) pp. 91–4Google Scholar. Another version is offered by Beeston, , Le Museon 65 (1952) pp. 142–7Google Scholar, but it has little effect on the validity of my argument. This text also has not been dated but the ductus is not dissimilar to that of CIH 73, cf. p. 19, n. 3. A poor photograph may be seen in Nami, Kh. Y., Našr nuqūš sāmīya qadīma min janūb bilād al-'arab wa-šarḥu-hā, Cairo 1943, p. 95Google Scholar.

^{page 24 note 1} One cannot, of course, exclude the possibility of copper in this context. But it would not be suitable elsewhere and ritual offences were regarded as very serious in South Arabia.

^{page 24 note 2} blṭ occurs again in each of two texts engraved on one stone, Fakhry 30 and 30 bis, cf. Fakhry, A., An Archaeological Journey to the Yemen, II (= Ryckmans, G., Epigraphical Texts), Cairo 1952, pp. 20–3Google Scholar. The first text is a record of the extinction of a debt (cf. Ryckmans, J., Le Muséon 67 (1954) pp. 342–3)Google Scholar, though the purpose is not stated, and the second is probably of similar purport. According to G. Ryckmans the sum involved (ml') in Fakhry 30 was thirty blṭm/ḏdrḍym/ (11. 1–4). But if blṭt are silver coins this sum is curiously small, and in Fakhry 30 bis it is stated to be 200 blṭm[/]drḍym/ (11. 2–3). Similarly in RES 2724/10 the ml' is only 40 rḍym. I suspect however that Ryckmans may be mistaken in his rendering of ml'as “montant” and that in fact it refers to the interest alone, i.e. that which fills, makes up, and thereby extinguishes the debt. The “montant” I should prefer to see in the term šnqt. The coins could thus well be of silver.

^{page 24 note 3} Published by Jamme, op. cit., pp. 123–4.

^{page 24 note 4} Jamme, op. cit., pp. 390–1, would date this NŠ'KRB to about a.d. 10–20, thus in the period of the later Attic style coinage with Augustan head.

^{page 24 note 5} ṣrfn. Jamme renders this “brass”. Similarly ḏhbn is always translated “bronze”. But in no language does *ḏhb refer to bronze and I feel that whatever the metal of the object in question may in fact have been, the word should nevertheless be translated “gold”. Similarly ṣrfn should be translated “silver”. On ḏhb see also Ghul in BSOAS 22 (1959) p.'433, n. 1.

^{page 24 note 6} The supplementation is certain, cf., for example, Jamme 608/5–6 and 609/5. Another statue is mentioned after this one and although there is a lacuna, there is not enough room for its weight to have been given as well. It was probably of gold, cf. RES 4191 and Jamme 669 where the mdlt of silver objects is given but not that of gold. This may have been because “gold” statues, etc., were actually gilt.

^{page 24 note 7} Jamme interprets rḍy in these texts as a weight.

^{page 25 note 1} Cf. Dillmann, A., Lexicon Linguae Aethiopicae, Leipzig 1865, col. 1082Google Scholar.

^{page 25 note 2} A monetary context is unlikely in view of the qualifying phrase, “in red gold”. Beeston has suggested to me that ḥlfm might mean something like “standard ingot”. One might then compare the root ḥlf, cf. Hebrew ḥәlīfāh “substitute”, and Arabic ḥalaf “restitution; payement d'une dette” (Dozy, R., Supplément aux dictionnaires arabes, Leiden 1881, I, p. 397A.)Google Scholar

^{page 25 note 3} Cf. Hinz, op. cit, p. 51. The ṣā' is 3·24 kg.

^{page 25 note 4} It occurs in the Minaean RES 3695 from el-Öla where there is reference in line 4 to ṣ‘/’ḥly, “une ṣâ' de dattes douces”. But the context is very obscure and in any case the text's provenance is in North Arabia, ṣā' as a measure does not seem to be recorded in the South.

^{page 25 note 5} Op. cit., p. 2156. Note also ṣāḠa “tourmenter, agiter, remuer”.

^{page 26 note 1} Beeston, Descriptive Grammar, §4:11, admits this only where an etymological Ḡ and an r are in close contact. However, it seems to occur in Dathîni Arabic without the presence of an r, cf. the examples in Brockelmann, C., Grundriss der vergleichenden Grammatik der semitischen Sprachen, I, Berlin 1908, pp. 120–1Google Scholar, though in both cases '/Ḡ is initial.

^{page 26 note 2} Cf. II Chron., 3, x. The context of that passage suggests that the cherubim were not made of gold.

^{page 26 note 3} The passage runs as follows: wkwn[t]/ḏt/hqnytn/bwrṠ[/ḏ]dṯ'/ḏḠrf/s'd'ṯtr/bn/hwf'ṯt/ṣw'm”, and this dedication was in the month of DṮ' of the year of S'D'ṮTR bn HWF'ṮT ṣ;w'm”. The offering consisted of a golden bull, cf. line 3.

^{page 26 note 4} Sabaean Inscriptions, pp. 44–5. See also p. 41 on CIH 343/13.

^{page 26 note 5} Beeston's translation in Sabaean Inscriptions is thus far incorrect with “one thousand blṭ, good coins of ḤY'L”. Rather, “one thousand coins of legal standard, ḤY'L-coins”.

^{page 26 note 6} See Beeston's version, op. cit, pp. 52–5.

^{page 26 note 7} wdmwm/bšy'hw/ (11. 3–4). This translation suggested by Beeston seems preferable to any alternative. Cf. Ethiopic śu' “linen”.

^{page 26 note 8} Cf. also Müller, loc. cit.: “eine Nisbe von ḥy'l, einem Könige, dessen Bildnis oder Name auf den Münzen vielleicht ausgeprägt war.” However, no known South Arabian king has this name.

^{page 27 note 1} ḥy'l occurs in RES 3571 where it is the clan name of a Minaean, apparently resident in Upper Egypt. It may possibly be read in CIH 614/1–2. The Thamudic Jaussen-Savignac 521 attests a curious combination of bḥcyw'lḥbṣ.

^{page 27 note 2} The name hḥiwdin RES 2745/1–2 might be so explained but it is very doubtful.

^{page 27 note 3} See the items described by Hill, op. cit., pp. 61–3. These are of the later Attic, Augustan type. Hill suggests that the monogramme be read slḥn, i.e. the place of minting (op. cit., p. lxii), but the form of the monogramme does not lend itself to this. However, it is preferable to the suggestion of P. Boneschi, RSO XXIX (1954) pp. 20–1, that it be read ḥl'and taken with the accompanying monogramme to produce ḥl' [w]ynḠfš, “il s'affablit [et] fut battu”. Boneschi's entire approach to these monogrammes raises problems of a numismatic, historical and linguistic nature, not least that a propaganda legend of that sort would hardly be rendered into the comparative obscurity of monogrammes.

^{page 27 note 4} Unfortunately no reproduction of CIH 548 is available. If however the types chosen by the Corpus editors for printing the text provide any criterion for judging the ductus of the copy (Halévy 152), the inscription must be of relatively late date. Note in particular the forms of š and b. Both are much later than Pirenne's last stade. Yet there are so many seemingly early types used that it is difficult to arrive at any firm conclusion.

^{page 27 note 5} Schlumberger, op. cit., p. 27, suggests that the monogrammes on the later Attic coins may represent the names of agoranomes or city magistrates, much as is the case on the Athenian prototypes with monogrammes. It is to be regretted that there is no information in the inscriptions on the subject of monetary officials.

^{page 28 note 1} On the significance of this see Beeston, , Epigraphic South Arabian Calendars and Dating, London 1956, pp. 5–7Google Scholar.

^{page 28 note 2} Op. tit., pp. 25–6.

^{page 28 note 3} Rather, it may be a gold coin.

^{page 28 note 4} Copper coins from South Arabia in general are very rare. One Minaean specimen, actually of bronze with traces of silvering, is noted by Hill, op. cit., p. 76. It is an imitation of an Alexandrian prototype and probably of the third or second century b.c. (op. cit, p. lxxxiii). More recently a small selection of Ḥaḍrami pieces of various weights and sizes, some cast, has been acquired by the British Museum. All are certainly late. They are described by Walker in BSOAS 14 (1952) pp. 623–6. He argues that the earliest date to which they can be assigned is the second century a.d.

^{page 28 note 5} The only attempts known to me are dismissed by Jamme, op. cit., p. 343, n. 197.

^{page 29 note 1} Cf. Mordtmann, J. H. and Mittwoch, E., Sabäische Inschrifien, Hamburg 1931, p. 233Google Scholar. Not all their examples however are valid.

^{page 29 note 2} My argument here is based largely on an interpretation communicated to me by Beeston. It is perhaps more readily reconcilable with what is known of the Qatabanian social system than is that of Rhodokanakis.

^{page 29 note 3} Chrestomathia, p. 154A. Note that the Arabic verb had already been adduced by him as an etymology for ḥbṭ “fecit (nocturnam?) incursionem”.

^{page 29 note 4} RSO IX (1923) p. 603.

^{page 29 note 5} Cf. Hill, op. cit., pp. 45–53.

^{page 29 note 6} See the article by Walker cited above, p. 22, n. 3. He dates the coins in question to about the end of the second century b.c.

^{page 29 note 7} See Hill, op. cit., p. 75. They are of the latest type with profile head on each side. Three curious coins of possible Qatabanian origin are described on pp. lii and 52.

^{page 29 note 8} Le royaume sud-arabe de Qatabân et sa dotation, Louvain 1961, p. 62.Google Scholar

^{page 30 note 1} Lane, op. cit., p. 697C: ḥabaṣa “he turned over and mixed and made [ḥhabisḥṣ” (i.e. a kind of food made of crushed dates and clarified butter). Landberg, op. cit., p. 559, lists ḥabaṣa “intriguer”, II “écraser avec le pied”.

^{page 30 note 2} Hill, op. cit., pp. 46, 50 and 51, lists six base specimens, all of the old Attic series. But he adds that base coins are common in the “bucranium” series as represented in Vienna. However, one would hardly be likely to stipulate these in payments.

^{page 30 note 3} Particularly if used concurrently with the later, smaller Qatabanian coins.

^{page 30 note 4} Sabaean Inscriptions, p. 53.

^{page 30 note 5} The actual phrase is nṭš/'št/mḥrmn/ (1. 10) and Beeston compares Hebrew nāṭaš “to abandon” [actually “expel”], despite the phonological problem of the sibilant correspondence, and Akkadian ištu “from”. However, von Soden, W. relates the latter to *wiśtum “middle” (Grundriss der akkadischen Grammatik, Rome 1952, p. 165)Google Scholar. I should rather compare Arabic 'iša “way of life” and naṭīš, which the Qāmūs defines as ḥaraka “movement”, for the translation, “leave the community of the sanctuary”.

^{page 30 note 6} Normally Old South Arabic s corresponds to Hebrew š, yet the sibilant of sl'yn is definitely a samek, which should answer to ò. This may be sufficient to render the comparison doubtful, even allowing that 'sl'm would be a loan-word. There is however one certain case of the correspondence in the verb s'd: Hebrew s'd.

^{page 31 note 1} Bascially the Hebrew word means “rock, clod, boulder”, then “weight, coin”. The latter corresponds to one sacred or two common shekels and is defined in Roman terms as 48 dupondii (Bekh. 50^a) or four denarii (Y. Kidd. 1, 58^d); cf. Jastrow, op. cit., p. 996A. This does not reflect the Nabataean standard noted below and since Jastrow notes that sela' can refer to gold, silver or copper coins, it cannot be assumed to have developed far from the idea of a weight.

^{page 31 note 2} Thus the Nabataean text RES 1103/9 mentions sl'yn 'If ḥrty, “one thousand sela' of Aretas”. Cf. also RES 1108/8 and 1144/9. The form si'n occurs in RES 1148 and 1187/2, 3. All are of the reign of Aretas IV.

^{page 31 note 3} A Text-Book of North Semitic Inscriptions, Oxford 1903, p. 223Google Scholar.

^{page 31 note 4} The standards of Nabataean silver are given by Hill, op. cit., pp. xx–xxii. The average weight of the coins of Aretas IV is 4–204 gms. Those of his successors Malichus II and Rabbel II average 3·52 and 3·40 gms. respectively. de Morgan, J., Manuel de numismatique orientate, Paris 1923–1936, p. 255Google Scholar, compares the former to the Attic drachm and the latter to the pre-Neronic denarius (3·90 gms.). The “unit” in South Arabia at that time was a heavier coin but of less attractive appearance. Schlumberger, op. cit., p. 9, notes that many of the coins in the later Attic series were crudely struck, however well the dies may have been cut. But the weights are remarkably uniform.

^{page 31 note 5} Schlumberger claims to have seen in a bazaar in Constantinople an Attic tetradrachm with a South Arabian n countermarked on Athene's cheek (op. cit., p. 22). Pirenne comments (La Grèce et Saba, p. 79) on the absence of foreign coins in South Arabia. Their lack is all the more remarkable when one considers the composition of other eastern coin hoards, and leads me to suggest that, as in South India in Roman Imperial times, imported gold and silver coins may have been used as bullion, and perhaps also melted down to provide metal for native coins. One might recall the general complaints voiced by the Roman Emperors that Roman money was flooding out to the East without any return. According to Strabo (XVI, iv, 22) one of the motives of Aelius Gallus' expedition was to find a remedy for this situation.

^{page 32 note 1} Op. cit., pp. 1966–7.

^{page 32 note 2} The word is quoted from the 'Anaza dialect on the authority of J. L. Burckhardt. I have been unable to verify the meaning in any other authority and I suspect that Burckhardt may have confused the principle of the penalty with the material content.

^{page 32 note 3} Cf. Coulbeaux, P. S. and Schreiber, J., Dictionnaire de la langue Tigraī, Vienna 1915, p. 208Google Scholar.

^{page 32 note 4} Cf. also Conti Rossini, op. cit., p. 604. In addition to CIH 73 we might note RES 4772/5 with t/blṭm/r(ḍ)ym/ṣbbm/rmym/ḏš'm/mk[, which I would tentatively render, “current coin, paid down as interest, which … bought . . . ”. For ṣbbm cf. Arabic ṣabba “he paid down a price or sum of money”. Various other analogous senses are found, rmym may represent Arabic ar-ramā” “interest, usury” and colloquial rimāya, with the same sense.

^{page 32 note 5} Cf. Ryckmans, J., Le Muséon 69 (1956) pp. 94–8Google Scholar.

^{page 33 note 1} See generally Rhodokanakis, , Studien zur Lexikographie und Grammatik des Altsüdarabischen, II, SBWA 185/3, Vienna 1917, pp. 141–2Google Scholar. The ductus of this text is similar to that of Nami 74.

^{page 33 note 2} Ḳatabanische Texte, II, pp. 25–6. He compares the use of ϰαλóν and ṭayyib on certain Arab coins, “in Bezug auf Gewicht und Reinheit des Metalls” (p. 25, n.4). For these coins, from Emesa, cf. Walker, , A Catalogue of the Muḥammadan Coins in the British Museum, II, London 1956, p. xcviGoogle Scholar.

^{page 33 note 3} Cf. Ryckmans, J., Le Muséon 67 (1954) p. 343 and n. 11.Google Scholar

^{page 33 note 4} Cf. above, p. 24, n. 2.

^{page 33 note 5} Cf. Jamme, op. cit., pp. 106–7.

^{page 33 note 6} Jamme's comment on the term is that it is “a measure of weight”. He is at fault however in comparing Arabic rḍḍ “to break into large fragments”, etc., since this seems to presuppose a term other than the monetary one, which is quite unnecessary.

^{page 33 note 7} Cf. p. 23, n. 2.

^{page 34 note 1} The most recent version of the text is that of Beeston, in his pamphlet, Qahtan. Studies in Old South Arabian Epigraphy, I, London 1959Google Scholar. His is the text I follow.

^{page 34 note 2} Wärq is the name of the standard gold coin in modern Ethiopia. That country however has no continuous tradition for a coinage.

^{page 34 note 3} Cf. RES 3910/7, 3951/3, and perhaps 4130/1. The case of Fakhry 55/7 is doubtful since the reading of the passage is obscure, but if G. Ryckmans is justified in interpreting the preceding clause as referring to the payment of rent on land, a monetary context is not to be ruled out.

^{page 34 note 4} Paléographie, p. 172.

^{page 34 note 5} Op. cit., p. 211.

^{page 34 note 6} Fakhry 55 can be dated approximately from the mention of the king Š'Rm ‘WTR of SB’ and ḏRYDN in line 3. Jamme places this king around 65–55 b.c. (op. cit., p. 391), thus about the date of the known gold coins of later Attic style.

^{page 34 note 7} Cf. Hill, op. cit., p. lxxvi. The coin is in Vienna.

^{page 34 note 8} See above, p. 22, n. 3. The gold coin recorded by Hill belongs to the first sub-class of the later Attic series and some of the silver in this sub-class carry the Aramaic legend.

^{page 34 note 9} Op. cit., p. 12. The w is an abbreviation for wrq while the n probably had a numerical value based on its position in the alphabet. See generally Beeston, op. cit., pp. 8–9.

^{page 35 note 1} This factor must of course be understood in the light of my remarks on p. 31, n. 5, above.

^{page 36 note 1} That is, on to the standard of the Neronic denarius.

^{page 36 note 2} Pirenne argues that the latest pieces were minted about a.d. 300, cf. Le royaume sud-arabe de Qatabân, pp. 61–5.