¹ The following abbreviations have been used: B. = Babylonian Talmud, cited by tractate and folio; J. = Jerusalem Talmud, cited first by tractate, chapter and section, and then in brackets by folio, column and line as in the Krotoschin edition; T. = Tosefta, cited by tractate, chapter and section, and then in brackets by its page and line in Zuckermandel edition; R. = Rabbi; b. = ben (= son of); d. = denarius; ant. = antoninianus; sol. = solidus. All dates are CE. (= A.D.).

I should like to express my thanks to the following: R. A. G. Carson, whose help and advice is evident throughout this article, Prof. Momigliano, for his kind encouragement, P. Grierson for his astute criticism and valued advice.

² What is meant here is, of course, the Schinus Molle (of the Anacard family) which is not closely related botanically to the true pepper (Piper). The latter in Roman times came from India (Pliny, Hist. Nat. 12, 28, etc.).

³ This term occurs again only in T. Shebiit 5.7 (68.9) and J. Shebiit 7.1 (37B 58). (For the correct reading in the latter source see Bar-Ilan II (Jerusalem, 1964), 128, line 6.) The exact identity of this plant is as yet unknown. See Löw, I., Aramaeische Pflanzennamen (Leipzig, 1881), 319, no. 262Google Scholar, and the same author's Die Flora der Juden, vol. 4 (Wien, 1934), 146, and most recently Lieberman, S., Tosefta Ki-fshutah, Order Zeraim, part 2 (New York, 1955), 551.Google Scholar to line 14. However, it seems clear that it is some kind of aromatic tree. See sources cited above, and also M. Jastrow, A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature, s.v., 1223B.

⁴ Olives were cheap in the Galilee (e.g. Genesis Rabba 20).

⁵ In his edition of Deuteronomy Rabbah (called Midrash Debarim Rabba, 2nd ed., Jerusalem, 1964), 126, n. 2, and in Midrash Wayyikra (= Leviticus) Rabba, ed. Margulies, M., part 4 (Jerusalem, 1958), 879.Google Scholar

⁶ This is the usual Mishnaic meaning of mana. See Jastrow, o.c. 797.

⁷ Jewish Quarterly Review, 34 (1934–44), 481–2.

⁸ In an article in Hebrew to be published in Talpiot, (New York).

⁹ In this I wish to correct much of what I have stated in an article in Archiv Orientalni, 34, part I (Prague), 54–66, touching also upon this subject.

¹⁰ Migne, Patrologia Graeca, XX, col. 816: ὡς ἑνὸς μέτρου πυρῶν διοχιλίας καί πεντακοσίας Ἀττικὰς ἀντικαταλλάττεσθαι (Loeb ed., 352).

¹¹ See Payne-Smith, Thesaurus Syriacus, s.v. mana, 2164. But cf. The Syriac version of Eusebius, etc., from the St. Petersburg manuscript, (dated 462 C.E.), by Wright, W. and McLean, N. (Cambridge, 1898), 369Google Scholar and P. Bedjan's ed. (Leipzig, 1897), 530, where the reading is ‘ma'in’ (= obols). Quite clearly this cannot be the correct reading, as an obol was never equal to an Attic drachma. In Syriac the difference between an ‘n’ and a ‘('ayin) is hardly noticeable at times. Cp., e.g., different readings in Syriac Job 21, 24 cited in Schulthess, , Lexicon Syropalaestinun (Berlin, 1903), 115AGoogle Scholar, s.v. m'a. The St. Petersburg ms. is a copy of an earlier one, now lost. The Greek version was first translated into Syriac in the fourth century. See Lohmann, E., Der textkritische Wert der Syrischen Übersetzung der Kirchengeschichte des Eusebius (Halle, 1894), 10–12.Google Scholar

¹² In my article in Talpiot.

¹³ e.g. Pesikta de R. Kahana, ed. Mandelbaum, 144 and 404.

¹⁴ See Margalioth, M.'s Encyclopaedia of Talmudic and Gaonic Literature (4th ed., Jerusalem 5712, Hebrew), cols. 11–12.Google Scholar

¹⁵ ibid., J. Sota 9.13, (24B 20), B. Nidda 64B, Lev. Rab. 35.12 (but cp. Margulies ed., 832), J. Shabbat 17.7 (cited by Yuchsin). See Yaavitz, Toledot, etc., 7, 157, 11.

¹⁶ Igereth Rav Sherira Gaon, ed. Hyman, A. (London 1910), 96.Google Scholar

¹⁷ He arrived in Israel from Babylon, while R. Hanina (b. Hama) was still alive (B. Shabbat. 121A). As R. Hanina b. Hama died c. 240 C.E. (Hyman, Toldoth, etc., 429A), R. Abba b. Kahana was probably born closer to c. 220.

¹⁸ He lived at least three generations, B. Shabbat 146A.

¹⁹ For the date of R. Jochanan's death see Igereth, etc. (n. 16), 70, and n. 37; idem, ed. B. M. Lewin, 84, n. 3. Halevy, Dorot Harischonim II, 310, etc. That he lived on after the death of both R. Jochanan and R. Zeira I can be seen from Ecclesiastes Rabba 1.4.

²⁰ Byzantion Byzantion 15, 263.

²¹ It is however true that there was one R. Abba b. Kohen, who is at least on one occasion confused with R. Abba b. Kahana (B. Baba Mezia IIB, also J. Challa 2.7, see A. Hyman's Toldoth Tannaim Veamoraim, 48B), and who flourished in Palestine in the fifth generation (c.350–75, ibid.). Thus one might argue that our text is of a later date, c. 370 for example, and mana is here being used in its traditional meaning of 100 d. Hence, we would have an equation of I aureus = 80,000 d., quite a plausible ratio for the later 4th cent. (Segrè, Byzantion I.c.). A careful analysis of certain cognate texts (Yalkut I, 298; 2, 937; Midrash Tillim 9, Buber ed. p. 81; Pesikta Rabbati 23; J. Pe'a 1.1, 15D 14) will however demonstrate conclusively that the reading R. Abba b. Kahana in our text is not to be altered. For there we find statements of a similar exegetical nature (based on the same Biblical verse), yet independent in style and form, all in the name of R. Abba b. Kahana. The last source cited (J. Pe'a) is particularly significant, as it is wholly independent both exegetically and formally, and similar only in idea and content. These texts may therefore be regarded as independent corroborations of the reading ‘Kahana’ (as opposed to the suggested‘Kohen’) in Deuteronomy Rabba. (See also Tanhuma Buber, Deuteronomy 16 and 34.) The equation of I aureus = 80,000 d. could not be prior to c. 340 (Segrè, Byzantion, ibid.), when R. Abba b. Kahana was no longer alive. Hence ‘mana’ cannot here be taken in its more usual sense of 100 d.

²² Bolin, S., State and Currency in the Roman Empire to 300 A.D. (Uppsala, 1958), 278–81.Google Scholar See also C. Oman in Numismatic Chronicle, 4th series (1916), 37 f., and against him Bolin, ibid. 263.

²³ Shortly to be published. I am grateful to Mr. Carson for letting me see his unpublished manuscript and for much useful advice and discussion on the subject in general.

²⁴ Bolin, o.c. 286–7.

²⁵ See West, L. C., Gold and Silver Standards in the Roman Empire (New York, 1941, published as ‘Numismatic Notes and Monographs', No. 94), 165 and 167Google Scholar (table AK).

²⁶ e.g. Roman Imperial Coinage (= RIC), vol. 5, part 2 (by Mattingly and Sydenham), 6.

²⁷ West, ibid., 168.

²⁸ ibid.

²⁹ The Korban Ha'edah (a commentator on the Jerushalmi Talmud) ad loc. reads 1,000 [d.], but has no manuscript basis for this reading. It is merely a hypothetical correction of his own to make the text more ‘plausible’.

³⁰ See above, note 19.

³¹ I have discussed the matter with Mr. Carson, who agrees with me on this point, and further notes the significance of the fact that though the loan was contracted in denarii, it was to be repaid in aurei. This was the only real safeguard against the everprogressing inflation and its effect upon the debt (especially important for Jews in view of the stringency of their anti-interest laws). This new evidence reopens the question of the relationship between the denarius and the antoninianus during this period.

I cannot accept Heichelheim's view (JRS lv, 1965, 251) that ‘Attic’ drachms in certain papyri of c. 300 (Papiri greci e latini 965) were ‘the equivalent of 1000 inflationary denarii during Diocletian's first years, if not under Aurelian, Probus, or Carus already’. Even though the aureus was greatly reduced in weight (to just over 1 gm. towards the end of Gallienus' sole reign) it could never have been equal in value to the Attic drachm. Jones' suggestion (JRS xlix, 1959, 34 ff.) that Attic drachm and inflationary denarius have equal value is far more acceptable. According to this, the νοῦμμος of these papyri, which equalled 25 to 12½ Attic drachms, equalled 25 to 12½ d. (not 12,500 d., as Heichelheim suggests). There is in fact Rabbinic evidence to support Jones' view. For in B. Avoda Zara 34B, a text of c. 320(?), it is stated that 1 xestes wine cost 4 ‘lumma’, and the same amount of muries I lumma. ‘Lumma’ = nummus (see most recently Lieberman, Tosefta ki-fshutah, l.c., p. 229). If a nummus equals 12½ d., then I xestes wine cost 50 d. in Palestine c. 310(?). According to P. Rylands 629, line 41 (c. 317–23), I xestes wine in Antioch cost 75 d. and (ibid. 636, line 267) 100 d. According to Heichelheim, however, the wine should have cost 50,000 d. See also Heichelheim, in An Economic Survey of Ancient Rome IV (Baltimore, 1938), 223.Google Scholar (The ‘lumma’, however, does occur in a Rabbinic passage earlier than Diocletian, i.e. T. Demai 3.12 (50.14), a text of before c. 230.)

³² o.c., 176–7. Out of the weight of 144 gold coins of Probus, two points of concentration emerge, the first around 80 to 82 grains, the second around 98 grains (West, ibid. 176). In fact a careful examination of his table AN (177) shows secondary concentrations at 90 to 93 grains, and around 104 grains. This suggests perhaps that coins were judged by weight, rather than by clear denomination.

³³ For material on the coinage of Carus and Carinus see West, o.c. 178–83.

³⁴ RIC, ibid., part 1, I (Probus), 126 (Carus).

³⁵ ibid. 76, 249, 320.

³⁶ Ed. Graser, E. R., in Tenney Frank's Economic Survey of Ancient Rome, V (Baltimore, 1940), 412.Google Scholar Although the reading was called into question by Mattingly, in his article ‘Monetary System of the Roman Empire from Diocletian to Theodosius I’ (Numismatic Chronicle 1946, 113), he himself later accepted the reading as correct when new corroborative fragments were discovered (see Mattingly, , Roman Coins, London 1960, 217–8Google Scholar). See also Ehrendorfer, F., ‘Die Münzreform des Diocletian’, Num. Zeitschrift 72 (1947), 101.Google Scholar Also West, L. C. ‘The Coinage of Diocletian and the Edict of Prices’ (Studies in Roman Economic and Social History in honour of Allan Chester Johnson, 1951), 290Google Scholar, etc. Most recently F. M. Heichelheim, in JRS lv (1965), 251, basing himself on information supplied to him by Klaffenbach, has shown conclusively that the reading in the Elatea Fragment is not to be called into question (contra West, Mattingly, Pareti, Mazzarino, Bernardi, Ruggini, Skeat, as there cited). Klaffenberg's epigraphic evidence, however, is in itself not enough to render Heichelheim's thesis (ibid.) irrefutable, as West (l.c.), for example, considered the possibility of a mason's error in the carving of the inscription. But his argument from papyrological sources is very convincing, and the Rabbinic evidence here cited would seem to bear it out.

³⁷ Bolin, o.c. 302, notes I, 2. Christ, W., ‘Denar und Follis der späteren römischen Kaiserzeit’, Sitzungsberichte der königl. bayrischen Akademie der Wissenschaft I (1865), 121.Google Scholar Kubitschek, Quinquennium 86. Mickwitz, Gelt und Wirtschaft im römischen Reich des vierten Jahrhunderts n. Chr. (Soc. Seien. Fennica, 1932), 62.

³⁸ Bolin, o.c. 304. Seeck, O., ‘Sesterz und Follis’, Num. Zeitschrift 28 (1896), 176.Google Scholar

³⁹ Mickwitz, o.c. 69.

⁴⁰ Bolin, o.c. 300.

⁴¹ I have discussed this problem in an article to appear in the Journal of the Economic and Social History of the Orient.

⁴² o.c. 309–10.

⁴³ o.c.

⁴⁴ o.c. 315.

⁴⁵ See also Mattingly, in Roman Coins (first edition, London 1928), 226Google Scholar, who reached the same conclusion. Also West, op. cit., 187.

⁴⁶ Bolin, o.c. 332–3 and 285. In this connection it is interesting to note evidence in Rabbinic literature of an increased use of bullion during this period: e.g. J. Ketubot 12.7. (34c 55), c. 290–320, a marriage settlement of one libra aurei; J. Baba Kama 8.8 (6c 15), c. 250–60, a fine of one libra aurei for insulting a Rabbi; Genesis Rabba 63.3 (Theodore ed., p. 681, line 10), a parable of c. 290–320 mentioning the libra aurei. Also Deuteronomy Rabba 1.13, etc.

⁴⁷ RIC, part 2, pp. 210, 235, nos. 151–4 (Rome), p. 246, no. 251 (Siscia), p. 277, no. 51 (Rome). It is, of course, true that these would mainly be below the weight level of a ½ Diocletian post-reform aureus. Nonetheless the denomination as such existed and was known. Many aurei of the earlier part of the century would be equal to ½ a Diocletianic aureus. Bolin, o.c. 252–4, etc.

⁴⁸ RIC, part 1, p. 109, no. 15; p. 123, nos. 4–5; p. 125, no. 21; pp. 138–44, nos. 97–134 (Rome); p. 170, nos. 451–2.

⁴⁹ ibid. p. 33, nos. 147–9 (Rome); (see also p. 202?).

⁵⁰ Unfortunately many of the higher prices in chapters 19, 28, etc. have not survived. In Transactions of the American Philological Association 71 (1940), 157–74, E. R. Graser published two new fragments. In the second one (a short one, which alone has readable prices) the price 8 occurs 13 times, 16 6 times, etc. Cf. West (l.c., note 28): he underestimated the ‘800’. See also West's second chart on p. 300, ibid.

⁵¹ Bolin, o.c. 310. West (o.c. p. 293, note 28) says 21 times.

⁵² o.c. 293–4.

⁵³ RIC, part 2, 207.

⁵⁴ ibid., based on Seeck, in Zeitschrift für Numismatik 17, 36 ff. See also West, o.c. 196–9, table AQ.

⁵⁵ RIC, ibid.