² For the early tradition see article on ‘Aes and Pecunia’, quoted in note 1 above.

³ This is the only true representative of that Romano-Campanian coinage that has bulked so large in modern studies : cf. H. Mattingly, ‘The Romano-Campanian coinage from a New Angle.’ Journal of the Warburg Institute, 1937, 197 ff.

⁴ These bars weigh on the average about 6 pounds of 10 oz. each (of the later Roman pound). Excessively rare as they are to-day, they were not always so, as is proved by the occurrence of fragments, with copper coins going down to the first century B.C., in a number of hoards. One only bears a legend, ROMANO; but the types of all—such as shield and sword, trident, anchor, and caduceus, elephant, sow, ox—seem to fall within or near the Roman sphere. Our tradition, which knows of a coinage preceding the round asses and characterized by types of domestic animals seems to preserve faint but unmistakable traces of these bars. Other bars, with more rudimentary types, such as fishbone pattern or crescents, seem to be nearer to the ‘Aes Rude’, and therefore earlier.

The value of a bar of 6 lb. should be that of a Neapolitan silver nummus (see below).

⁵ See Mattingly, and Robinson, , ‘The Date of the Roman Denarius,’ Proc. Brit. Ac. xviii, 2, 1933, 211 ffGoogle Scholar.

⁶ Nummus is derived from Greek νόμοϛ, ‘standard coin,’ familiar in South Italy and Sicily: cf. op. cit. in note 5, Appendix 1, 254 ff. The derivation of as is unknown. Its by-form assis might link it to asser (or assis) ‘bar’, or to axis ‘(assis) ‘wheel’. As means ‘unit’, ‘ace’ as well as standard coin. Some relation, then, if only by way of false analogy to Greek εἴϛ may be suspected; but the Doric forms, ἄϛ, ἄ ϛ ίϛ quoted in Forcellini's dictionary seem to lack ancient authority.

⁷ For all details see Haeberlin, E. J., Aes Grave (Frankfurt, 1909)Google Scholar: here, only the types of the as are quoted. For the coinages here discussed see also M. v. Bahrfeldt, ‘Le monete romano-campane,’ Riv. It. Num. 1899, 387 ff.; 1900, 11 ff.: Grueber, H. A., Coins of the Roman Republic in the British Museum, 3 vols., 1910Google Scholar: A. Sambon, Les monnaies antiques de l'Italie (incomplete, vol. i only), 1903: Giesecke, W., Italia Numismatica (Leipzig, 1928)Google Scholar; Sicilia Numismatica (Leipzig, 1932)Google Scholar.

⁸ Cf. M. v. Bahrfeldt, op. cit., 1900, 33 f. (wt. 0.65 gm.).

⁹ In Aes, as in silver. Some resemblance between silver and aes may be seen in each mint, but struck silver and cast aes are too unlike to admit of easy comparison.

¹⁰ So too in Sicily the Sicels had a currency of bronze, while the Greek cities gave priority to silver: the litra, the tenth of the Greek nummus, was the silver equivalent of the pound of bronze.

¹¹ The libral as was often equated by later writers with the obol (Aeginetic). The sixth of a didrachm should strictly speaking be called a diobol, but the Aeginetic drachm (over 5½ scruples) was not so very much lighter than the didrachms in question.

¹² A table of weights, compiled for me by my friend, Mr. E. S. G. Robinson, shows the Diana didrachm c. 10 gm. lighter than the reduced Tarentine: itself, it does not quite reach the 6 scruple standard.

¹³ See the introductions to the appropriate sections in Haeberlin's Aes Grave.

^13a It is, of course, impossible as yet to prove that all four issues were precisely contemporary. We go on, in fact, to suggest that they were not so. But that they are approximately contemporary is proved—

(1) by the limiting date, 269 B.C., for the first issue of Roman silver, and

(2) by the lower limits set by the succeeding ROMA issues, themselves followed by the denarius, There is no room in time for the ROMANO-issues to fall far apart from one another.

¹⁴ Hercules of the obverse, founder of the Fabian gens, refers to C. Fabius, the she-wolf and twins of the reverse to Q. Ogulnius, who, as aedile with his brother in 296 B.C., had ‘placed likenesses of the twins who founded the city under the udders of the she-wolf’ (Livy x, 23, 11–12). Groag (P-W s.v. ‘Fabius’, col. 1749) questions whether the connexion of the gens Fabia with Hercules was earlier than Augustus, but he himself cites evidence which might be held to refute his own view.

¹⁵ Jérôme Carcopino (‘La louve du Capitol’— Bull. Assoc. G. Budé, 1925) thinks that the twins, Romulus and Remus, were the founders of the two cities, Rome and Capua, respectively. When Capua revolted and fell, Remus ceased to have any proper function, and fell into neglect. The reverse, she-wolf and twins, recurs on the quadrans (of the 6 oz. standard) probably struck at Capua (see below).

¹⁶ For Cosa see A. Sambon, op. cit., pp. 82 f. For imitations see M. v. Bahrfeldt, op. cit. Riv. It. Num. 1899, 401 ff. For evidence of hoards see Cesano, L., Atti e Mem. d. Inst. It. i, 47 ffGoogle Scholar. Ariminum has light token Aes in a style not unlike that of this mint (A. Sambon, op. cit., 88). The Mars of the obverse owes something to the head of Leucippus at Metapontum, struck for some years down to about 300 B.C. (see W. Giesecke, Italia Numismatica, 95, pl. 13, 1, 2, 6).

¹⁷ See A. Sambon, op. cit., 115.

¹⁸ Over 300 unciae of this series were found at an unknown site in Apulia (L. Cesano, op. cit., in note 16, Tab. i, after p. 82). When Tarentum surrendered to Rome in 272 B.C. it became a socius navalis, but did not retain full autonomy: a Roman legion was stationed in the citadel (CAH vii, 655). From 213–209 Tarentum was in revolt against Rome, but the citadel was still held. In 209 100 lb. of gold from the special reserve was specially allotted to Q. Fabius he consul for the citadel at Tarentum (Livy xxvii, 10, 11–13).

¹⁹ See Th. Mommsen, Römisches Staatsrecht ii², 570 ff. Lydus, Johannes, De mag. i, 27Google Scholar, gives the year 267 B.C.. Mommsen reckons four quaestorships— Ostia, Cales, Gaul (Ravenna or Ariminum), and Lilybaeum ( ? ), possibly included with the strictly Italian quaestorships in Tac. Ann. xi, 22Google Scholar (‘duplicatus numerus’, etc.).

²⁰ Cf. O. Leuze, ‘Das Datum der ersten Silber prägung in Rom,’ ZfN 1920, 15 ff., where the ancient evidence is quoted: the Chronicon Paschale gives the year 273 B.C. marked both by the Roman consuls and by the regnal year of Ptolemy II. The evidence of the coins themselves is entirely consistent with this date. The mummus of Mint C is from the same hand as the local Aes of Beneventum, struck not earlier than 268, when Maleventum received its new name.

²¹ See Tod, M. N., BSA xxxvii, 236Google Scholar.

²² This suggestion, made already by J. N. Svoronos in his Τὰ Νομίσματα τοῦ Κράτους τῶν Πτολεμαίων (Die Münzen der Ptolemäer) 1908, i, 148 ff., 217 ff., iv, 83 ff., 143 ff., has been unduly neglected. Svoronos thinks that Rome and Egypt agreed to issue these coinages, side by side, for a term of fifty years. He himself reckons from 270, the year in which Arsinoe died.

²³ This series, for some reason unknown, is excessively rare.

²⁴ For the coins, see the works quoted in note 7 above.

²⁵ ROMA might seem to indicate the sovereignty of Rome more explicitly than ROMANO; but perhaps the distinction is little more than formal. The small token bronze, with types, Diana-Hound, has ROMA, whilst the nummus, Diana-Victory, has ROMANO.

²⁶ The little pieces, with rev. prow to right, and obv. head of Mercury and head of Bellona, respectively, have usually been called ‘semunciae’, and ‘quartunciae’, and attached to the first and second reductions. They bear no mark of value.

^26a Cf. Haeberlin, Aes Grave, 134 ff. and pl. 56.

²⁷ In both silver and bronze, the standardization is downwards—to the lowest standard in use in the first issue. The triumph of the six-scruple nummus was probably due to its close agreement with the standard of Hiero of Syracuse.

²⁸ See again the works quoted in p. 67, n. 7, and add: P. le Gentilhomme, ‘Les Quadrigati Nummi et le dieu Janus’ in Rev. num. 1934, 1 ff.

²⁹ Pliny, , NH xxxiii, 46Google Scholar; cf. E. A. Sydenham, ‘The Roman Victoriate,’ Num. Chron. 1932, 73 ff.

³⁰ The two reductions are quite clear in the struck issues, though much obscured in the cast, where there seems to be an almost uninterrupted decline in weight. The evidence of the L mint shows conclusively that the cast and struck pieces of the reductions do belong together. The standards, then, were the same in both cases, even if they only appear at all clearly in one.

³¹ The L mint also produces victoriates and quinarii in silver, and Aes, sextantal and uncial. In style, it is very close to two mints that sign respectively CA and Γ (P), Canusium and Paestum (? ). In a series of Aes, shared by this mint with a mint that signs T, types of the Dioscuri are very prominent, One thinks of Tarentum, the colony of Sparta, home of the two gods, and of Locri, the home of their legend as helpers in battle.

The attribution of the L coins to Luceria (Grueber, , BMC Rep. ii, 145, n. 1Google Scholar) rests only on vague reports of finds near that city. Luceria has a coinage of its own, with distinctive style and types.

³² There is nothing in itself improbable in a return after a war to a standard prevailing before it. Such a return must be readily admitted, if a balance of evidence points towards it.

If the return to libral standard is accepted on general grounds, it might be possible to relate to it the remarkable as, of libral standard (10 oz.), with obv. Goddess facing in triple crest, rev. Ox, ROMA, L (or caduceus). This as is unique in bearing the signature of Rome and in having no subdivisions. It seems to be, as Haeberlin has already observed (op. cit., i, 141 ff., ii, pl. 55, 56), a commemorative piece. The goddess would be Ma (Bellona), the war-goddess who attends on Cybele, the giver of Victory over Hannibal, the ox would be the symbol of ‘pecunia’, ‘money’. The caduceus suggests trade: L = ‘Litra’—pound?

³³ viii, 26, p. 1416 C: in 217 B.C. Rome debased (χαλκῷ προσμίξαι) her silver, which before had been pure (ἀμιγὲς καὶ καθαρόν).

³⁴ In 209 B.C., after the defection of the Latin colonies, 4,000 lb. of gold were taken but of the reserve: 500 lb. each were allotted to the consuls, the proconsuls, M. Marcellus and P. Sulpicius, and L. Veturius, the praetor, a special 100 lb. to the consul Fabius for the citadel at Tarentum: the residue was assigned to provision for the Spanish War (Livy xxvii, 10, 11–13). H. Willers in Corolla Numismatica, 310 ff., assigns the ‘Oath-scene’ coinage to this date, noting that the reverse occurs on a denarius struck by Ti. Ve. a member of the gens Veturia (?) about a hundred years later (Grueber, , BMC Rep. ii, 281 ff.Google Scholar). As the six- and three-scruple gold pieces show two distinct styles, it seems better to divide them over two occasions. The six-scruples piece was probably the ‘talent’ of the silver nummus, twelve times its value.

A closer attribution of the piece marked XXX must wait, until we can decide to what unit its mark refers. (For the coins, see M. v. Bahrfeldt, Die römsche Goldmünzenprägung, 12 ff. The coins are certainly genuine.)

³⁵ Cf. Festus, 347 M. The 333, 333⅓ asses voted in 217.B.C. for the ‘ludi Romani’ (Livy xxii, 10,7) must surely be a converted version of a round sum. 333, 333⅓ asses of 6 oz. equal 200,000 asses of 10 oz.

Pliny, (NH xxxiii, 45)Google Scholar thought that on this occasion the as was reduced from being the tenth to being the sixteenth of the denarius: Festus, as far as we can judge from the fragments, shared this view. It is to-day quite certain that the reduction to which they refer is nearly a hundred years later.

³⁶ For the equation, victoriate = 10 asses, note the following evidence.

The Roman series of Aes with mint marks, L and P (Grueber, , BMC Rep. ii, 179 ff., 203 ff.Google Scholar) shows beside the as of 12 oz. a dextans of 10 oz. The silver nummus, then, that contained ten of these asses contained ten of these dextantes. Now twelve to one is the relation of talent to nummus in the system of the ‘little talents’ (see Mattingly, op. cit., in note 1, above), and the only talent small enough to have a nummus expressed by a bronze piece of this size in place of a silver is the ‘Rhegine’ talent, the victoriate. Bronze coins of Venusia and Teate actually mark with N (nummus) bronze coins comparable in size to these dextantes. It appears, then, that at the mints L and P, the victoriate is the ‘denarius’, containing ten asses, of one system of reckoning and the collapsed talent, containing twelve dextantes of the other.

Silver coins of Populonia, weighing c. seven scruples, bear at first the mark X, later the mark XX. The later series may be assigned to the close of the Second Punic War. This doubling of the number of units in the nummus is exactly what we are postulating for the Roman quadrigatus. (A. Sambon, op. cit., in note 7 above, 48 ff.)

³⁷ If, as appears to be the case, the shape of the Roman prow (ending not in a ‘goose's neck’, χηνισκός, but in a solid ball) was of new design, it is much more likely to represent the model of the fleet of the Aegates Insulae (241) than of that of Mylae (260). The first Greek coin to show the new ‘Roman’ shape is the tetradrachm of Antigonus Gonatas, struck to celebrate his victory over the Egyptian fleet at Cos (c. 258 B.C.).

³⁸ That the head is of Saturn, not Jupiter, is proved by a later series of Aes, in which each deity of the obverse has his appropriate emblem on the reverse: the ‘harpa’ (sickle) of Saturn appears for the semis (Grueber, , BMC Rep. ii, 259 ffGoogle Scholar.—coinage of Cn. Domi., M. Sila., Q. Curti.). Cf. T. O. Mabbott, op. cit. in n. 1 above, who has independently reached the same conclusion.

³⁹ For coins of C. Fonteius, c. 105 B.C., see Grueber, , BMC Rep. ii, 292 ffGoogle Scholar. Such Roman deities as did not finally find a place in the Pantheon of the twelve gods—with Roman names co-ordinated with Greek—tended to fall more and more into the background. In many cases glimpses of their early importance can still be caught. The subject deserves further investigation.

⁴⁰ For the style, cf. coinage of Hiero (Giesecke, Sicilia Numismatica, 117 ff., and pl. 24, 25) and, even more, the didrachms (ΣΙΚΕΛΙѠΤΑΝ) struck by Rome for the Sicilians under her rule (Giesecke, op. cit., 147 ff., and pl. 27).

For references to the quadrigatus in literature, cf. Giesecke, Italia Numismatica, 202; Hultsch, Fr., Metrologiae scriptorum reliquiae ii, 76Google Scholar (Festus 98 M) and 80 (Festus 347 M).

⁴¹ If the view here expressed is correct, some hint of the later date of the ROMA issue in D may be looked for in hoards. Also, one might possibly find quadrigati or victoriates showing the style of Mints B and C.

⁴² For Centussis, cf. Festus 54 M; ‘centenariae cenae dicebantur, in quas lege Licinia non plus centussibus praeter terra enata inpendebantur, id est centum assibus, qui erant breves nummi ex aere.’ In Persius, , Satire, v, 191Google Scholar, ‘et centum Graecos curto centusse licetur,’ the Centussis is certainly a coin.

Decussis is more difficult. Festus 335 M seems to make it equal ‘denarius’: ‘sestertius dicitur quarta pars denarii, quo tempore is decusis valebat.’ In another passage (237 M) he makes it equal ten libral asses: ‘Tarpeia lege cautum est ut bos centusibus, ovis decusibus aestimaretur.’ The cistophorus, of c. 190 grains, might perhaps be called Decussis, as being approximately equal to ten sestertii (in the earliest issues, not much inferior to it in weight). For the trinummus see Mattingly and Robinson,. ‘The Date of the Roman Denarius’, 214 ff.

⁴³ For the combination of several nummi in one system, see Mattingly, ‘The Little Talents of the West,’ Num. Chron. 1943, 14 ff.

⁴⁴ The ancient authors, who speak of ‘asses sextantarii’, were possibly thinking not of asses, weighing 2 oz. but of asses, overstruck on sextantes of the first reduction, weighing about 1 oz. These overstruck asses are actually earlier in date than our ‘sextantal’ asses.

⁴⁵ Cf. Sydenham, E. A., ‘Problems of the Early Roman Denarius,’ in Trans. Int. Num. Congress of 1936 (London, 1938), 262 ffGoogle Scholar.

⁴⁶ The evidence of the Prologue to the revival of the Casina of Plautus is decisive. The ‘new plays’, that are there said to be as bad as the ‘novi nummi’, must be those of Terence and his friends, c. 170 B.C. following. At about that time the change of nummi was still recent. If any doubt still remains we note that nummus, which in Plautus means ‘didrachm’, in Terence means ‘drachma’.

⁴⁷ That the ‘denarius’ in 194 B:C. was still the quadrigatus appears from Livy xxxiv, 52, 6 (Attic tetradrachm equals about three denarii). That the as of libral weight was still in use in the same year may be deduced from a comparison of Livy xxxiv, our 46, 3, with Plutarch, Cat. Ma. 10, 4, on Cato's gift to his troops in Spain—270 ‘aeris’ in the one, 1 lb. of silver in the other.

The arguments of Dr. J. G. Milne (see works quoted in note 1) in favour of an earlier date for the denarius—218 B.C., in fact—are individually very ingenious, but they seem to fail in recognition of the solid framework of the new dating. Pliny, (NH xxxiii, 47)Google Scholar does, indeed, state,—if ‘denarius nummus’ is to be preferred as a reading to ‘aureus nummus’—that the denarius was struck fifty-one years after the ‘argenteus’. But Pliny, earlier in the same chapter, had written of a ‘denarius aureus’, and the context shows beyond doubt that a gold coin is in question. For the passage of Pliny and its ‘variae lectiones’, see M. v. Bahrfeldt, Die röm. Goldmünzenprägung, 3 ff., 16 ff.

Again, Livy, in booty lists from Spain from a little after 200 B.C., records the bringing in mass of ‘bigati’. Dr. Milne argues that these bigati must be denarii, with reverses of Diana or Victory in a biga. If these denarii—by no means the earliest—were current in Spain by 200 B.C., the origin of the coin must be some years further back. But Livy in these contexts mentions not only bigati but argentum Oscense. If he means by argentum Oscense what we mean by it to-day—coins of the standard of the denarius—he is simply in error: these coins were not struck until one—perhaps even two generations—later. It is possible at least that Livy meant by ‘bigati’ something quite different. The fragment of Festus S. V. ‘Sesterti’ (347 M) in Lindsay's text (Leipzig, 1913) reads: ‘sesterti not … dupundi et semisis q … tertius; sed aucto sesqu … apud antiquos autem … rant et valebant d … ti bigati quinquessis q‹uin› …’ etc. This tantalizing fragment, certainly spoke of ‘bigati’ and ‘quinquessis’ in immediate conjunction, probably also of ‘[quadriga]ti’ and ‘d[ecussis]’. Where quadrigatus = 10, bigatus then = 5. The bigatus, by a sort of pun, describes the half of a quadrigatus: a biga is half a quadriga. For further consideration of these passages of Livy, see C. T. Seltman, op. cit., in n. 1.

⁴⁸ Op. cit., xiii.