¹ For Mommsen's, editions see CIL I (1863), pp. 49–72, no. 198 and pp. 75–106, no. 200Google Scholar: Ges. Schr. I (1905), 1–145Google Scholar (revised text and commentary): CIL I ² (1918), pp. 443–52, no. 583 and pp. 455–64Google Scholar. no. 585 (Mommsen's text, with improved recension of the lost fragment E and minor valuable new epigraphic comments). Unfortunately no measurements are given for any of the surviving fragments. The plates in A. Degrassi, Imagines (1965), pp. 281–95, are useful aids to study, but the editor supplied no scale with any of them—and the degree of reduction clearly varies.

² Fragment Bab preserves much of the top edge and C the left edge of the tablet on the extortion law side. The lower part of fragment A (Ab, now lost) joined the missing fragment E, which in turn was firmly tied to C by direct contact over eight lines. The D fragments are secured in relation to E by a series of short lacunae of certain content (from e 12 to d 4 onwards in the extortion law). Fragments A, E, C and D are precisely fixed in relation to the top of the tablet by the proved correlation of the lines of A and B. See the tables in CIL I² and Mommsen's, comments in Ges. Schr. I, 11—16 and 37Google Scholar n. He estimated the width at a little over 2 metres, the height at under 1 metre.

³ Carcopino divined this, when he attacked Mommsen's restorations in both these passages and line 22 f., but he did not challenge the lacuna lengths. See Autour des Gracques (1928), 220 ff.

⁴ See Athenaeum N.S. XXXI (1953), 21–31Google Scholar: Historia XI (1962), 203Google Scholar. Tibiletti incidentally pointed out (p. 23, n. 2) that in the tables of Ges. Schr. 1 and CIL I² the figures given for the missing letters between A and B (extortion law side) are 20 too many, those to the right of B, 20 too few. We must always go by the figures in the margin of the text.

⁵ Tibiletti's work is fundamental to any study of the lex repetundarum, though Badian, in AJP LXXV (1954), 378–84Google Scholar, disposed of his aberrant dating of the lex Acilia (towards 111 B.C.): he provides full discussion and bibliography of the whole subject.

⁶ Mommsen, himself seems to have been partly aware of this and warned scholars against pressing his figures too closely (Ges. Schr. I, 14 ff.Google Scholar).

⁷ The trial would continue even so: note… eam] rem ab eis item quaerito. Thus Pliny prosecuted the estate of Caecilius Classicus: see A. N. Sherwin-White, The Letters of Pliny (1966), 231 f. on Ep. III 9, 5–6.

⁸ One could read scribundi as in line 18 (b 18: an exact parallel) or omit is.

⁹ For arvorsario perhaps read altero, which Mommsen, had in his first edition (CIL I, p. 59Google Scholar).

¹⁰ Mommsen gives quoiei is quoius nomen etc., but I prefer to model the phrase on quoi is queive ei quei petet in line 22 (a 18).

¹¹ Mommsen's version is inflated by quoi is queive ei quoius nomen delatum erit. For mine compare facito utei is unde petetur (b 24), is quei petet et unde petetur (b 26) and quoi is queive ei qui petet in line 22 (a 18). From gener restoration is certain thanks to the last passage, reproduction of which continues right down to queive in eodem conlegio siet.

¹² For accidental omission of words and even phrases see CIL I², pp. 446, 449 and 451 (nn. on lines 2, 47 and 72), and below p. 138.

¹³ I omit Mommsen's quoius ex h.l. nomen delatum erit after queive eius, also his eius after maiorum (as he did in CIL I, p. 60).

¹⁴ See Ges. Schr. I, 54, against Mommsen's insertion of the ubei… solent clause before aut extra Italiam. If there was no difference between Italy and the provinces, the law could have simply said conquaeri in o.f.c. ubei ioure deicundo praesse solent.

¹⁵ The missing phrase is found in full in line 6, partially in line 22 (a 17).

¹⁶ I have added agrum locum to Mommsen's supplement, which was taken bodily from line 12 (b 12). For mine compare extra eum agrum locum, quei ager locus in line 45 (a 39), or utei pro eo agro loco, quem agrum locum in 83 (d 22).

In line 33 Mommsen has, indeed, completely filled the gap, but Serrao has legitimately objected to the substance of the supplement, both here and in line 34. In particular he wishes to eject ante eidus Martias primas. See La Iurisdictio del pretore peregrino 65 ff.

¹⁷ We could shorten it by three letters, reading in terra Italia (as lines 4, 13, 33) and P. Mucio (as in 15, 27–29, 33).

¹⁸ Mommsen read… indu[xerit ibeique paverit pro eo pecore, quod eius in calli]bus… (45 letters). Photographs suggest the possibility of reading inpu[lserit paverit], which would give a neat balance with pastum inpulsum.

¹⁹ Mommsen gave even less, reading decernitoque as in line 16. But we find neive de [eo agro de]cernito in line 35 (a 30).

²⁰ One might read meilites and in terram Italiam (as in line 1). For this full phrase see line 50.

²¹ In the parallel previous chapter (line 33 f.), which dealt with land made private, nisei cos. pr(aetor)ve comes at the very end of the neive ius deicito section (see b 34). Mommsen gave nisei cos. pr. cens. the same position in this chapter in CIL I, p. 81. I like this balance. An alternative supplement for the b 35–a 30 lacuna would be [v(idebitur) e(sse) neive quis mag. provemag. postea de e]o agro… (31 letters).

²² Mommsen lengthens the supplement by tautologically inserting populo aut publicano dare debeat after aedificia.

²³ Mommsen read quaeque ex eis minus annum gnatae erunt post ea qua[m gnatae erunt] and supplied the same curious phrase in the earlier lacuna, thus achieving a supplement of c. 82 letters. My supplement owes much to the chapter on non-liability to vectigal or scriptura in line 19 f. (partly quoted above). Mommsen reads sati]s dato, but the bronze clearly has ]E. My supplement for the A–B lacuna comes to c. 140 letters as compared with Mommsen's estimate of 146.

²⁴ See Ges. Schr. I, 36 f. (n. on b 53). Orsini was able to add nothing on the other side, but Mommsen came to see that that may have been due to inability to decipher what broken fugitive letters survived.

²⁵ From the end of d 2 to just below sententiam in b 53 there are c. 40 letter-spaces on Mommsen's table. The rest of his rubric takes up 19 of these !

²⁶ See Ges. Schr. I, 37 n. and CIL I², p. 450 n. This could easily be part of a broken R.

²⁷ Any gap—on the evidence here presented—must be a very narrow one indeed. It is hard to find a reading. I am not happy about [quamque in id] iudiciu[m inve r]eum sententia[m. The crucial letters on both fragments are broken and indistinct. All the same I hesitate to press [quamque] iudicis [in r]eum sententia[m, since suspicion should rather fall on the E in Huelsen's EVM. From study of photographs and the bronze itself I feel fairly convinced that d 2 and b 53 should meet at this point. It should be feasible to test it out.

²⁸ For RE ROM see Ges. Schr. I, 37 n. with 86 n. and CIL I², p. 461, n. (Huelsen's verdict). I cannot see them either.

²⁹ Ges. Schr. I, II and 15.

³⁰ 21 surviving letters (15 + 6) + a B–A lacuna of 137: 80 letters (6 + 74) + an A–B lacuna of 90.

³¹ 84 letters (5 + 79) + an A–B lacuna of 104: 16 letters (10 + 6) + a B–A lacuna of 129.

³² 33 letters (28 + 5) + an B–A lacuna of 135; in letters (35 + 76) + an A–B lacuna of 90.

³³ 82 letters (1 + 81) + an A–B lacuna of 100; 14 letters (12 + 2) + a B–A lacuna of 129.

³⁴ 24 letters (13 + 11) + a B–A lacuna of 133; 98 letters (37 + 61) + an A–B lacuna of 85.

³⁵ 165 letters (62 + 103) + an A–B lacuna of 100; 94 letters + a B–A lacuna of 130.

³⁶ Athenaeum N.S. XXXI (1953), 22–25Google Scholar.

³⁷ Occasionally the engraver had to leave blank spaces because of surface flaws in the bronze, as before L.RVBR in line 22 (b 22: c. 9 letters), HACE in line 74 (d 22) and PRAETOR in line 78 (d 26). See Mommsen's, comment in Ges. Schr. I, 31Google Scholar n.

³⁸ Mommsen rightly took the whole passage from queive quaestione to maiorve a in his supplement after b 16 from lines 11 and 13: in line 13 the lacuna (assessed by Mommsen at 104 letters) can be supplied with 5 from line 16 (e merc) and 39 from line 11 (queive quaestione … conde).

³⁹ In line 18 f. we get eight letters from the inevitable supplement de quibus sibei consu[ltum siet and the rest from line 15 (iudices to habeto): for the lacuna of line 15 (given by Mommsen as 100 letters) one letter must be supplied for udices and the rest comes from line 18 (eos to consultum siet).

⁴⁰ From lines 6 to 26 on the side of the extortion law Mommsen's calculations average out at c. 405: from lines 66 to 83 (c 1–18: e 22–38: d 14–31) the average is c. 420. In these two areas there is a measure of control. The right edge of the B fragments, from b 1 to b 26, is roughly parallel to the tablet's right edge, and the missing letters to left of A can be closely established. C (from 1–14) actually preserves the tablet's left edge, and the right side of the D fragments is roughly parallel (and so with constant intervals) to the tablet's right edge. The engraver clearly tended to fit more letters in progressively as he moved down the tablet. For my reconstruction see pls. VII and X, with figs. 12 and 13.

⁴¹ The story can be pieced together from Mommsen, , Ges. Schr. I, 64Google Scholar: CIL I², p. 444: Bormann, Festschrift O. Hirschfeld (1903), 432 ff.: and CIL XI, Add. (1925), p. 1234Google Scholar.

⁴² Rendiconti Pont. Acad. Rom. d'Archaeol. XXIII–XXIV (1947–1949), 13Google Scholar. For Bembo, and Urbino, see Diz. Biog. degli Ital. VIII (1966), 137–40Google Scholar.

⁴³ Ges. Schr. I, 64: ‘Superesse haec apparet ex parte media, COICITO vocabulum respondere ei quod supplevimus v. 52 coniecito. Ea quoque quae sequuntur, quamquam reliquis inseri nequeunt, pergunt in argumento inde noto’. For the correlation of lines on the two sides see ibid. 13 f.

⁴⁴ Basically I follow the text given in CIL I², pp. 444 and 457. The readings for f 1 come from Gamurrini: for f I (agr.) Garrucci and Bormann give VEIL/.

⁴⁵ Thus removing Mommsen's square brackets from the first six letters of a word correctly guessed.

⁴⁶ Mommsen read q[uem agrum locum neque ipse] neque heres… One might suggest q[uod] eius ne[que ipse emptor], since postea is differently placed in line 54.

⁴⁷ I deliberately leave f 1 out of consideration on both sides, since the readings are so dubious. The anchoring of fragment F confirms my view on d 1/b 53 (b 62/d 1 on the reverse)—for which see p. 132— since the gaps between b 60 and 61 (rev.) and f 3 and 4 cannot exceed 30 and c. 20 letters respectively, as will soon appear. Therefore I incorporate my previous new readings for d 1/b 53 (b 62/d 1) and d 2/the lost b 54, whilst fitting in fragment F.

⁴⁸ Mommsen's supplements, though often sound, have had to be modified. The verb coicito can no longer be delayed until the long lacuna after singilatim indices and we need a new verb and subject there. The phrase scriptam atr[amento requires delere/deleto in place of Mommsen's inducere/inducito (conditioned by his guess ceratam). In calculating the missing letters in the lacunae I have, of course, reduced the width of the tablet and the gaps between E and B by 35 letter-spaces. I do the same with the lex agraria.

⁴⁹ For pedites/equites in line 60 see Mommsen, , Ges. Schr., I, p. 122Google Scholar. I hope to justify my supplement in a later article. The letters o quae e in f 4 (line 61) must surely be emended to a quae f: such mistakes are only too common with this engraver (see notes on text in CIL I²). This done, I can find only 13 letters for the gap between b 61 and f 4. Yet between b 60 and f 3 the inevitable supplement amounts to 30 letters. The discrepancy is disturbing, even if we allow that those 30 letters were cramped. The gap between b 61 and f 4 must be of c. 20–25 letter-spaces. The balance could be made up by accidental duplication (note EVM EVM in a 6) or by the engraver leaving a blank space where the bronze was too rough (as after oportuit at the end of c 10?): see n. 37 for this phenomenon even on the obverse.

⁵⁰ Note the recurrence of words like palam and apertus, and the phrase manum demittito et eam devexam popul[o palam ostendito] neiquid l[ateat. There must be no palming or ‘switching’ of votes. Each tablet is double checked after its verdict has been declared.

⁵¹ I accept hesitantly Mommsen's modo de]mum emptor siet, though I have also thought of quei pri]mum e.s.—the clause thus being limited to those who bought direct from a colonist or his legal heir. For the content of this whole section of the Lex Agraria, see Mommsen, , Ges. Schr. I, 119–25 and 141–4Google Scholar.

⁵² See Ges. Schr. I, 142: Cic., , II in Verr. IV, 5, 9Google Scholar: Marcianus, in Dig. XLIX, 16, 9Google Scholar. The immediate context in Cicero, —taken with pro Flacco, 34, 86Google Scholar— shows that in the second century all business dealings by senatorial officials in the provinces were proscribed by ‘unwritten law’.

⁵³ Hence I read kal. Ian. rather than kal. Iun., which might be better with M.Livio L.Calpurnio cos.

⁵⁴ See Ges. Schr. I, 41 n. (on line 72).

⁵⁵ See Mommsen, , Ges. Schr. I, 42 n. and 44 n.Google Scholar: CIL I², p. 451 f. Furthermore maiorei partis satis factum erit (line 75) seems to be corrected in line 82 to maioris partis satis fecerit e…

⁵⁶ Mommsen apparently divined this at the time of his first edition: CIL 1, p. 70 (‘… sive ita ab operis magistro iussus…’). The engraver often confused iudicium/indicium etc., in line 21 he wrote courato for iurato, in line 14 tribunum for tribum. In line 40 utei is has become ut eius, in line 33 um eius fide quaerat was correctly deciphered by Mommsen as [in maior]um eius fide fuerint! See, on the last two, Ges. Schr. 1, 33 f. nn. Daube (Forms of Roman Legislation, 54 f.) has rightly challenged the rubric (line 69) Quaestor moram nei facito because of the imperative. This will be another blunder—the engraver should have cut nei faciat. See also my n. 76.

⁵⁷ Mommsen supplied 14 letters between c 14 and e 35, but between c 13 and e 34 only 9 letters are missing, between c 12 and e 33 only 2. We can form no precise idea of the outlines of the lost fragment E and so surely could assume a smaller gap than Mommsen's between c 14 and e 35. No more than a ceiv needs be restored. Possibly the letters were spaced out a little more than usual here.

⁵⁸ Would the phrase for describing grandsons in the male line be varied at so short a distance in the law, as Mommsen's revised view demands? In CIL I he took eius with militiae and restored quodque eius militiae in sua quoiusque ceivitate (p. 63).

⁵⁹ One could aptly adduce Brisson's errors in transcribing fragment E. He omitted everything from ita planum to de eo agro in e 14 f., when he jumped forward from one agri locei to another. An even worse blunder was his jump from neive (emptum) in e 7 to neive (unius) in e 9. See Mommsen, , Ges. Schr. I, 86 fGoogle Scholar. n.

⁶⁰ For all this see Mommsen, , Ges. Schr. I, 11–13 and 16Google Scholar. Short gaps do occur between chapters throughout the agrarian law.

⁶¹ Autour des Gracques, 228 f.

⁶² For Forum Sempronii see Weiss in P–W VII, col. 73, and Mommsen, , Ges. Schr. I, 106 fGoogle Scholar. The inhabitants of such fora, as Mommsen, noted, were presumably viasii vicani (lex agraria, II ff.Google Scholar).

⁶³ On Bembo and Urbino see my n. 42. On Fossombrone, see Encicl. Ital. (1949) xv, 775 f.Google Scholar: Bembo's patron Guidobaldo died there in April 1508 (ibid. XVIII, 256).

⁶⁴ See Ges. Schr. I, 18 ff. and 22 (Glaucia tribune c. III B.C.): Autour des Gracques, 230 (Glaucia tribune 108 B.C.): Athenaeum, N.S. XXXI (1953), 82Google Scholar and n. 1 (Lex Acilia ‘verso III’).

⁶⁵ PBSR XIV (1938), 114Google Scholar: CRAI 1951, 63 (one of three possibilities).

⁶⁶ See Bartoccini, R., Epigraphica IX (1949), 3–31Google Scholar: A. Piganiol, o.c., 58–63: Luzzatto, G. I., Archiv. stor. pugl. IV (1951), 29–31Google Scholar: Tibiletti, o.c., 73 ff.: Schönbauer, E., Anz. Akad. Wiss. Wien, 93 (1956), 13 ffGoogle Scholar. Schönbauer thinks that it may be Saturninus' treason law.

⁶⁷ This third tablet—marked —ends s(i) s(acrum) s(anctum) e(st) q(uod) n(on) i(ure) s(it) r(ogatum) e(ius) h(ac) l(ege) n(ihilum) r(ogatur). On this sanctio see Cic., , pro Balbo 14, 32–35Google Scholar: lex Gabinia de Delo, line 36 (CIL 1², 2, 2500). It quite certainly comes at the very end of Gabinius' consular law. I would regard tr. pl. in line 26 of the Tarentine fragment as a kind of signature of the board issuing the law. There is space for 10 letters to its right. If it is the start of a new chapter, as Tibiletti argues (o.c., 55 f.), why was this space not used ? For useful photographs see Degrassi, Imagines, p. 297.

⁶⁸ In line 22 Tibiletti prefers Fuchs', H.iouraverit (Hermes LXVIII (1933), 348Google Scholar) to the ioudicaverit of CIL 1², 582, 21, but illogically models his restoration on the CIL version [eius quei ita utei s(upra) s(criptum) e(st) iourarit nomen persc]riptum siet.

⁶⁹ I ignore the uncertain letter-traces in line 1. Tibiletti was puzzled by ‘i segni |>’ in line 2. Study of Degrassi's plate (see n. 67) shows the proper reading to be is (Tibiletti noted the archaic three-stroke sigma: o.c., 38, n. 2). The phrase can now be restored on the model of lex Mamilia Roscia, ch. 5 (FIRA I, no. 12) cuius unius opera maxime is condemnatus erit.

⁷⁰ See Tibiletti, o.c., 39–47: Piganiol, o.c., 59. On p. 47, n. 2 Tibiletti writes ‘Dovevano inoltre esser contemplati, in entrambe le leggi, premi per l'accusatore romano … ma il testo relativo andò perduto’ (my italics). In his table Clause A of the Bembine law is split into a missing Clause A (Offerta della cittadinanza) and a subsidiary Clause C (Privilegi a coloro che accettino la cittadinanza) in the Tarentine law. Is this necessary ?

⁷¹ Did Mommsen think of bere liceto as the end of a rubric before his Sei quis ceivis Romanus ? The subjunctive is required, but he had already admitted Quaestor moram nei facito in line 69 (d 17)— wrongly (see n. 56). Now, however, we must treat [Sei quis cei]vis Romanus … as a rubric itself, or perhaps [De eo quei cei]vis Romanus … may seem preferable. There is no good reason for thinking that a rubric was omitted here. The Bembine text suggests that the Tarentine inhiberi is a mistake for inhibere: the active form anyway suits the context there better.

⁷² 24 letters + 690 (5 lines— 10): 10 letters + 660 (5 lines — 40).

⁷³ The whole section runs from line 85 to line 89 in the lex Bembina. The first passage works out at 18 letters + 326 (361 — 35) + 53 + 326 (361 — 35) + 10 (bere liceto): the second at 40 letters + 327 (362 — 35) + 48 + 382 (417 — 35) + 18 (]regrinos ious [deicet). The […. c. 7 … cei]vis Romanus rubric alone (see n. 71) accounts for at least 40 letter-spaces (6 spaces + 34 surviving letters): another must be supplied before Quoi ex. h.l. (as is shown by the vacant space before quoi), and a third in the long lacuna before pe]regrinos ious in d 37.

⁷⁴ Down to provocare liceto I follow Mommsen fairly closely, fusing the remains of lines 78 and 85 in a single version. For my variant of the quoius eorum opera phrase, see above, n. 69. My sei ceivitatem mutare nolet is based on Val. Max. IX, 5, 1 (de provocatione eorum qui dvitatem mutare noluissent) and lex Salpens, chs. 22 f. (si civitate mutatus mutata non esset: FIRA I, p. 204 f., no. 23). For the rest I note and defend any significant divergences from either Mommsen or from Tibiletti's text of the Tarentine fragment where they occur.

⁷⁵ For te omnium rerum Tibiletti proposed (o.c., 39 f.) ei recte omnium rerum siremps lex … esto. For my version, which makes immunitas precede the double vacatio, one may compare the formulae of the S.C. de Asclepiade (CIL 1², no. 588, 12), ὅπως οὗτοι τέκνα ἔκγονοί τε αὐτῶν ἐν ταῖς ἑαυτῶν πατρίσιν ἀλειτούργητοι πάντων τῶν πραγμάτων καὶ ἀνείσφοροι ὦσιν (non-citizens: 78 B.C.); SC de Aphrodis. (OGIS II, no. 455, 7 and 13 ff.), καὶ τὴν ἀτέλειαν αὐτοὺς πάντων τῶν πραγ[μάτων ἔχειν καρπίζεσθαι] and ὅπως…καρπίζωνταί τε πάντων πραγμάτων ἀτε[λεῖς ὄντες] (civitas libera: 42 B.C.); Edict. Octav. de Seleuco (FIRA I, p. 311, no. 55, 20ff.), πολείτειαν καὶ ἀνεισφορίαν τῶν ὑπαρχόν[των δίδ]ομεν….[…καὶ στρατείας λει]του[ργία]ςτε δημοσίας ἁπάσης πάρε[σις ἔστω] (a new citizen); Edict. Octav. de veteranis (FIRA I, p. 316, no. 56, 8 ff.), ‘ipsis parentibu[s lib]erisque eorum e[t uxo]ribus qui sec[um]que erunt im[mu]nitatem omnium rerum d[a]re … immunes sunto, liberi s[unto mi]litiae muneribus publicis fu[ngend]i vocat[i]o (esto).’

⁷⁶ Here again we get useful help from the SC de Asclep. (FIRA I, no. 35: CIL 1², no. 588, 2–5 (Latin) and 17 ff. (Greek)) and the Edict. Octav. de Seleuco (FIRA I, no. 55, 53 ff.). Tibiletti rightly rejected ioudicium certet—it must be ioudicio—and made the subjunctive depend on conceditor, which he supplied in the lacuna: he further suggested [seiquis … vocabit … in] ioudicium (o.c., 40 f.). I believe with him that the clause covers cases where the beneficiaries are defendants as well as plaintiffs. In Lex Bembina 86 Huelsen read I.PETETVR, noting that the first letter could be T or S. (CIL 1², p. 452, n.). There is little enough of it left and from photographs I judge that it could well be E. As for the Tarentine certet I can only conclude that this is a mistake for certato and so print cert<ato>. Inhiberi and eo (for ei, line 8) show that the engraver was fallible. I take the very apt ne in eius [optione quis eum inhibeat] from Tibiletti—yet another suggestive echo of the Bembine text.

⁷⁷ Mommsen's Quoi ex h[ace lege provo]catio erit … seems somewhat out of place, despite advorsum provocationem a little later. Provocatio was fully safeguarded in Clause B. This is clearly concerned with militiae munerisque vacatio: it concerns the new citizens, the Latin with provocatio and the successful citizen prosecutor alike. My supplements are basically those of Piganiol and Tibiletti (o.c., 44 f.).

⁷⁸ Despite the efforts of Tibiletti (o.c., 47 ff.) I doubt whether any sound or continuous sense can be won from lines 12–19. The passage seems to deal with giving full publicity to the law in Italy, the provinces and the client kingdoms and the setting up of a master-copy in Rome. The responsibility is laid on the peregrine praetor.

⁷⁹ Mommsen's 361—35: 6 letters + 280 (2 lines) + 40.

⁸⁰ c. 40 letters + 327 (362 — 35): 280 (2 lines) + 10 (quei ceivis).

⁸¹ 43 letters (Quoi … oportebit) + 382 (417 — 35) + 18: 280 (2 lines) + c. 100.

⁸² Cic., , pro Rab. Post., 4, 9Google Scholar (hoc totidem verbis translatum caput est).

⁸³ Cic., , pro Cluent. 54, 148Google Scholar with 55, 151 and 56, 154—57. 157: Ewins, U., JRS L (1960), 94–96Google Scholar.

⁸⁴ The phrase is also found in lex Bembina rep., line 70, and lex Gabinia de Delo, line 34 f. (CIL 1³, 2,2500). Dr. Mario Torelli has independently restored Lex Bantina 25 as I do (information by letter).

⁸⁵ The photograph in Degrassi, Imagines, p. 279, suggests that VE is as good a reading as NE.

⁸⁶ See Yarnold, E., AJPh, LXXVIII (1957), 165 f.Google Scholar, n. 7: Lex de Piratis, C II ff. (SEG III, 378: FIRA I, p. 128, no. 9): Cic., , ad Att. III, 23, 4Google Scholar.

⁸⁷ For the oath in leges see Livy XXXI, 49, 7 f. (200 B.C.). For refusal to swear to particular laws (or clauses of laws?) see Lex Agraria line 41 f. Tibiletti thought that the details of the oath could be dispensed with through a simple reference back to his presumed Lex Acilia of c. III B.C. (o.c., 52 f. and 75).

⁸⁸ See the good remarks of Carcopino, o.c., 210 ff.: Piganiol, o.c., 61 f.: Tibiletti, o.c., 57–68.

⁸⁹ JRS, LV (1965), 197Google Scholar.

⁹⁰ See Mommsen, , Ges. Schr. I, 199Google Scholar (text) and 207.

⁹¹ Ibid. 147 with 244 (text), and 158 f. with 263 f.

⁹² See on this above, nn. 56, 71 and 76. The Tarentine engraver missed out poplici after muneris, but the other man in his turn omitted militaria after aera (Lex Bembina, lines 77 and 84). Such variations are common in independent copies of an ancient document: a good example is the Athenian Coinage Decree (Ath. Trib. Lists II, 61 ff., D 14Google Scholar).

⁹³ See Lex Bembina, lines 12 and 89: Lex Tarentina, line 12 f.: Cic., , II in Verr., 2, 49, 122Google Scholar with CIL 1², p. 200. I do not think that Mommsen's, reasonable deduction from Lex Bembina, line 16 (Ges. Schr. I, 51 fGoogle Scholar.), can seriously be challenged. It would seem to be clinched by the phrasing in lines 72/79 sei is praetor quei ex hace lege quaeret sei[ve is quaestor quoi aerarium] vel urbana provincia obvenerit. Bannier's alternative reading (Philol. LXXXIII (1928), 444Google Scholar) has no real attraction or cogency.

⁹⁴ For Caepio see Tibiletti, o.c., 73–75. For Glaucia see Bartoccini, o.c. (n. 66), 28 f.: Piganiol, o.c., 62 f.: Luzzatto, o.c. (n. 66), 29 ff.