¹ The main discussions of Livy 40.51.9 have been: Lange, L., Röm. Alt. ³ iii. 265, 354Google Scholar; Mommsen, Th.StR ³, iii. 1. 185 (with n. 1)Google Scholar; Smith, F., Die römische Timokratie (Berlin, 1906), 140f.Google Scholar; Botsford, G. W., The Roman Assemblies (New York, 1909), 85 n. 3Google Scholar; De Sanctis, G., Storia dei Romani (Torino, 1916), iii. 1. 380Google Scholar; idem (Torino, 1923), iv. 1. 606–7; McDonald, A. H., ‘The History of Rome and Italy in the 2nd Century B.C.’, Camb. Hist. Jour. 6 (1939), 134Google Scholar; Taylor, L. R., The Voting Districts of the Roman Republic (MAAR, 1960), 139–40Google Scholar; Nicolet, C., ‘La Réforme des Comices de 179 av. J.-C.’, Rev. Hist. 39 (1961), 341–58Google Scholar; Cássola, F., I gruppi politici Romani nel III secolo A.C. (Trieste, 1962), 96Google Scholar; Pieri, G., L'histoire du Cens jusqu'à la fin de la république romaine (Paris, 1968), 155Google Scholar; Treggiari, S., Roman Freedmen during the Late Republic (Oxford, 1969), 44–5Google Scholar; Palmer, R. E. A., The Archaic Community of the Romans (Cambridge, 1970), 73–4Google Scholar; Scullard, H. H., Roman Politics 220–150 B.C. ² (Oxford, 1973), 182–3Google Scholar. These works will be cited usually by the author's name.

² Both censors are the subject of 40.51.8, Aemilius alone of 40.52.1.

³ So McDonald, , ‘The Style of Livy’, JRS 47 (1957), esp. 155–6Google Scholar: ‘We find phraseology in Livy that is technical to the point of legalism; it may also bear the marks of antiquarian research…Livy exploited these elements of archaic formalism, as Cicero accepted them for laws about religion. …Livy's technical vocabulary is revealed in connexion with the despatch of Roman envoys and their reports, the reception of foreign embassies, the records of censors' activities…’; cf. e.g. the war vote of 201 b.c. (31.6.1). In what follows there will be frequent use of expressions such as ‘Livy quotes’ or ‘Livy paraphrases’. These are of course a form of shorthand to avoid endless perambulations concerning which source Livy might or might not be using and how far removed he is from anything that might be called a ‘primary source’. For these matters see the convincing treatment of Luce, T. J. in Livy (Princeton, 1977), ch. 5Google Scholar. The extreme brevity of 40.51.9 might suggest the Annales Maximi as the source. For the nature of these ‘descriptive summaries of events’, ‘short, declarative sentences with no attempt at continuous prose, much less historical narrative’ to provide ‘brief but authoritative information’ for the public, see the comprehensive treatment of Frier, B. W., Libri Annales Pontificum Maximorum: the Origins of the Annalistic Tradition (Rome, 1979), esp. chs. 5 and 8Google Scholar.

⁴ 43.14.5–9.

⁵ Cf. archaism such as the genitive form procum or procum patricium, Cic. Orat. 156, Festus 290L; or olla centuria, Var. LL 7.42; cf. Cic. Leg. 3.11; Lex de viginti quaestoribus, 1.5 (olleisque hominibus). In the following I have depended particularly heavily on the concordances of Merguet and Abbott, Oldfather and Canter for Cicero, Packard for Livy, Lodge for Plautus, McGlynn for Terence, Merguet for Caesar, and on the Thesaurus Linguae Latinae.

⁶ Respectively Sage in the Loeb translation, Botsford and McDonald.

⁷ Kübler, 's ‘Einzelstimme’ (RE iva, 1931, 654–8Google Scholar s.v. suffragium). This is by far the most common meaning of suffragium in both Cicero and Livy. It is also used in both singular and plural for the collective vote of the people: in the singular, Livy 28.27.14, 6.41.6, 4.43.12; Cic. Leg. Agr. 2.31; in the plural, Livy 7.17.12, 5.11.2, 27.30.9; Cic. Plane. 11, Mil. 25, ibid. 96, Imp. Pomp. 58.

⁸ Kübler's ‘Abstimmung’. E.g. Livy 25.3.15, 45.39.20, 6.38.4; Cicero, Sest. 109.

⁹ E.g. Livy 1.43.10, 38.36.7; Cic. Phil. 7.16, Leg. Agr. 2.17.

¹⁰ Kübler's ‘Stimmkörper'. This usage is most familiar in the phrase sex suffragia: see Festus 452L s.v. sex suffragia; Cic. De Re Publica 2.39 (for the greater reliability of this reading see Taylor, L. R., AJP 82 [1961], 337–45Google Scholar, and ibid. 84 [1963], 66–7); Phil. 2.82 (for the text, see e.g. Taylor, L. R., The Roman Voting Assemblies [Ann Arbor, 1966], 96–7 with n. 28Google Scholar); possibly Cic. De Re Publica 4.2 (see Nicolet, C., ‘Le cens sénatorial sous la République et sous Auguste’, JRS 66 [1976], 24–5Google Scholar). Cf. Livy 5.18.1, where a reference to the voting of prerogative suffragia is to be understood (and possibly Cic. Mur. 47; cf. Ps. Sall. ad Caes. sen. 2.8).

¹¹ Cf. Draeger's example, Plaut, , Bacch. 156–7Google Scholar ‘ne Phoenix tuis factis fuam | teque ad patrem esse mortuom renuntiem’, DrDraeger, A., Historische Syntax der Lateinischen Sprache, ii.2 (Leipzig, 1881), -que: 36–46Google Scholar.

¹² Respectively, McDonald and De Sanctis.

¹³ Describere and discribere are confused elsewhere in the Livian manuscripts, e.g. 25.31.9; 30.26.6; cf. 30.37.5. Three times in the early books editors have emended the describere of the manuscripts to discribere: 1.19.6; 1.42.5 (see Ogilvie, Comm. ad locc.); cf. 4.4.2. The only other case relating to discribere in the Codex Moguntinus, the manuscript for 40.51.9, is 33.42.8, where Mog. gives diviserunt as against the discripserunt of B. (This is perhaps an example of Mog. substituting an explanatory annotation for the word in the text; see McDonald, OCT, Bks 31–35, xxxviiiGoogle Scholar.)

¹⁴ For these problems see TLL. v. 1, col. 1354, s.v. discribere.

¹⁵ Sage in the Loeb translation.

¹⁶ TLL, v. 1, coll. 662–3.

¹⁷ E.g. ‘in singulos iudices’, Cic. Cluent. 74; ‘in quinque classes’, idem, R.P. 2.39; ‘suis comitibus compotoribusque’, idem, Phil. 5.22; ‘civitatibus’, idem, 2 Verr. 5.62. The result thus achieved is a discriptio which can take the genitive either of the original group to be divided, as Cic. De Domo 129 ‘servorum omnium vicatim celebrabatur tota urbe discriptio’, or of the resulting groups, as in Livy 4.4.2 ‘census in civitate et discriptio centuriarum classiumque non erat’.

¹⁸ So Kühner, , Ausführliche Grammatik der Lateinischen Sprache ², revised C. Stegmann (Hannover, 1914), ii. 1, 391Google Scholar, who classes them with ablatives accompanying verbs of measurement, judgement, establishing, etc.

¹⁹ So the year could be divided into twelve months (Livy 1.19.6 ‘in duodecim menses discribit annum’); Macedon into four regions (Livy 45.18.7 ‘in quattuor regions discribi Macedoniam…placuit); or land into plots of ten iugera (Cic. Leg. Agr. 2.79 ‘in iugera dens discribat [scil. agrum Campanum]’).

²⁰ E.g. Cic. R.P. 2.14 ‘populumque…in tribus tris curiasque triginta discripserat’; Leg. 3.7 ‘populique partes in tribus discribunto’; Livy 45.15.1 ‘in quattuor urbanas tribus discripti erant libertini’; Val. Max. 2.2.9; Sen. Ep. 89.3; cf. Cic. Leg. 3.44; Comm. Pet. 30.

²¹ It is not clear whether it is describere or discribere that is wanted for the third verb. If discribere, the sense is that the proles is being divided up into two groups, equites and pedites, the genitives taking the place of the indirect object. The same progression is given in the Pro Flacco (15): ‘quae scisceret plebes aut quae populus iuberet, submota contione, distributis partibus tributim et centuriatim discriptis ordinibus, classibus, aetatibus’. The usual punctuation of this passage, placing a comma before tributim, ignores the parallelism with Leg. 3.7 and at the same time produces an incomprehensible text — a man's tribe in no way determined his social, wealth or age group. See also Leg. 3.44.

²² See e.g. Cic. Rab. Perd. 8 (Apulia and Campania); Planc. 22 (defining an area from Sora to Allifae as ‘tota illa nostra regio'); Livy 27.42.17 (Lucani); 27.7.7 (Bruttii); 22.9.5 (Luceria and Apulia).

²³ Cf. e.g. Livy 30.26.6: ‘etiam quod magnam vim frumenti ex Hispania missam M. Valerius Falto et M. Fabius Buteo aediles curules quaternis aeris vicatim populo discripserunt’ and Caes. Bell. Alex. 5.3: ‘distributi [nostri] munitionum tuendarum causa vicatim’. Regionatim does occur once elsewhere in Livy (45.30.2): ‘regionatim commercio interruptis’. The meaning of this phrase is not clear. The force of analogy, however, would suggest an interpretation along the lines of ‘to those region by region (that is, in each region) broken up in their commerce’.

²⁴ As by Botsford, McDonald, Taylor, Treggiari, Scullard.

²⁵ Mommsen saw that genus does not in the main concern what we might call ‘birth’, ‘sondern zunächst das angeborene, dann überhaupt das Merkmal einer Person’, StR ³ iii.1, 9 n. 2Google Scholar. Ample instances of the various meanings of genus are given in TLL vi. 2–3, coll. 1885–97, s.v. genus.

²⁶ E.g. Pl. Trin. 542; Cic. Div. Caec. 28; Livy 36.17.5; 38.17.3; 42.51.8; Paul. exc. Fest. 165L; Sall. Cat. 6.1.

²⁷ E.g. Pl. Ps. 153 ‘plagigera genera hominum’, that is, ‘slaves’; the six kinds of men of Cic. Catil. 2.17–22; idem, Post Red. ad Quir. 21.

²⁸ The most significant passages of this kind are to be found in the Verrines, e.g.: (2.2.137) ‘ordo aliqui censorum est, conlegium, genus aliquod hominum? nam aut publice civitates istos honorer habent, aut, si generatim homines, ut aratores, ut mercatores, ut navicularii: censores quidem qui magis quam aediles?’; (2.2.166) ‘quod genus hominum, quem numerum, quem ordinem proferre possum qui to non oderit, sive civium Romanorum sive Siculorum’; similarly 2.2.149; 2.2.17; cf. 2.3.27 and Leg. Agr. 2.84. Ordo and genus hominum are also used together in connection with the senate, 2 Verr. 1.22, and 5.177.

²⁹ E.g. 27.51.3 and 22.61.14. So also 10.21.3–4; 21.12.8; 10.24.9. Cf. the similar usage in Cic. Sest. 124.

³⁰ Poenulus 831f.:

quodvis genus ibi hominum videas quasi Acheruntem

veneris,

equitem peditem, libertinum, furem an fugitivom

velis,

verberatum, vinctum, addictum: qui habet quod det,

utut homo est,

omnia genera recipiuntur.

Both in sentiment and in the way it is expressed this passage is very close to Trin. 490f., where Plautus explicitly mentions the property assessments of the census in connection with the social division between rich and poor: a division which separated men in different ordines, ibid. 451–3. Cf. also Curc. 499; Mos. 657. With Plautus' genera hominum one might compare Caes. B.G. 6.13.1; idem, B. Afr. 19.1.

³¹ Comparable associations are to be found in Suet. Cal. 34.1 ‘adversus omnis aevi hominum genus grassatus’; Livy 26.9.13 ‘ominium generum atque aetatium’; Caesar, B. Afr. 87.2 ‘cuiusque generis aetatisque’. Genus by itself can also be used for ordo, as equestre genus, Vell. 2.88; libertinum genus, Tac. Ann. 2.85; 4.62; Suet. Aug. 44, although examples tend to be late.

³² The censors did, on one occasion at least, concern themselves with causae militum; see Livy 43.14.9 and 43.15.8 (169 b.c.). However, this task was usually assigned to tribunes of the plebs or consuls, e.g. Livy 34.56.9, 11.

³³ So Botsford, followed by Taylor, Scullard and McDonald.

³⁴ E.g. Pro Balbo 36; ibid. 45; cf. ibid. 35; 2 Verr. 5.49; Caec. 98. Cf. Livy 38.39.5; 38.39.7; 39.26.10.

³⁵ 2 Verr. 1.152 ‘itaque tibi, Hortensi, non illius aetas sed causa, non vestitus, sed fortuna popularis videbatur, neque to tam cornmovebat quod ille cum toga praetexta quam quod sine bulla venerat’ and the explanation: ‘quod ornamentum pueritiae pater dederat, indicium atque insigne fortunae, hoc ab isto praedone ereptum esse graviter tum et acerbe homines ferebant’. Also Cic. Leg. Agr. 3.9. Condicio is used more explicitly in connection with social status; of the equites, Cluent. 154; of slaves, Cat. 4.16. Cf. Livy 23.35.9.

³⁶ Also, e.g., Asinaria 519–20. Cf. Miniconi, P. J., Causa et ses dérivés (Paris, 1951), 29–30Google Scholar. In the Truculentus passage there is a nice pun playing on this sense of causa and on that in ‘causam militis cognoscere’.

³⁷ ‘iam de artificiis et quaestibus, qui liberales habendi, qui sordidi sint, haec fere accepimus. primum improbantur ii quaestus, qui in odic hominum incurrunt, ut portitorum, ut faeneratorum. illiberales autem et sordidi quaestus mercennariorum omnium, quorum operae, non quorum artes emuntur…etc.’ Similar are Livy 21.63.4 ‘quaestus omnis patribus indecorus visus’ and 22.26.1 ‘ex eo genre quaestus pecunia a patre relicta’ (of Terentius Varro, whose father was said to be a butcher, lanius).

³⁸ E.g. Rudens 290–1: ‘omnibu' modis qui pauperes sunt homines miseri vivont, praesertim quibu' nec quaestus est neque didicere artem ullam’; Asinaria 186: ‘…ad suom quemque hominem quaestum esse aequomst callidum’; Persa 53–6. Of course, Plautus also uses quaestus of meretrices and lenones: As. 215, 511; Poen. 1140; Cist. 41.

³⁹ Per. Liv. 20.

⁴⁰ As it stands the text leaves open the question of whether or not the freedman with a child is to be resident in an area belonging to a rural tribe. Most probably the assumption is that the provision applies to those freedmen who, had they been ingenui, would as a matter of course have been registered in a rural tribe.

⁴¹ Genus: see locc. citt. n. 31. Ordo: e.g. Livy 42.27.3; 43.12.9; Cic. Phil. 2.3.

⁴² See Crawford, M. H., Roman Republican Coinage (Cambridge, 1974), ii. 631Google Scholar; cf. Mattingly, H., ‘The Property Qualifications of the Roman Classes’, JRS 27 (1937), 99–107, esp. 104 n. 15Google Scholar.

⁴³ De Sanctis, , Storia, iv. 1.606–7Google Scholar; Livy 41.8.6–12; 41.9.9.

⁴⁴ Livy 38.36.7–9; Palmer, loc. cit. (n. 1).

⁴⁵ For various discrimina, compare 6.35.6.

⁴⁶ Plin. NH 34.1; 35.159; Dio Cassius 38.13; Suet. Jul. 42.3; idem, Aug. 32.1; Livy 1.43.3, and 7; Dion. Hal. 4.17.3; Cic. De Re Publica 2.39–40.

⁴⁷ Cf. Nicolet, C., L'Ordre Equestre (Paris, 1966), 167–9Google Scholar.

⁴⁸ See C. Nicolet, loc. cit. (n. 10). Other familiar ordines such as the tribuni aerarii or scribae, while requiring a property qualification in the later republic, nevertheless did not constitute separate voting categories in the comitia centuriata.

⁴⁹ One thing at least that is known about the enigmatic lex Metilia de fullonibus dicta (220 b.c.) is that it was proposed to the assembly by the censors (Plin. NH 35.197). The censors of 312/11 b.c. took away one of the iura of the tibicines (Livy 9.30.5). The aratores and pecuarii, as might be expected, were of particular concern. According to Varro, a lex censoria required that flocks of sheep being driven from Apulia to Samnium be declared to a publicanus (RR 2.1.16). Cato, 's speech ‘de agna musta pascenda’, ORF frr. 89–92Google Scholar, shows that as censor he was involved in res pecuariae. (Cf. also Gell. N.A. 4.12; Plin. NH 13.24, 14.95.) For quaestus and ordines, cf. the senatus consultum in force in the reign of Tiberius banning young men of the senatorial and equestrian orders from performing as actors and gladiators, Suet. Tib. 35.2; the lex Acilia Repetundarum (FIRA ² i.7) 13 and 16 (restored), excluding former gladiators from elections to the panels of 450 jurors; the Tabula Heracleensis (FIRA ² i.13), which excluded public criers and undertakers from the municipal magistracies and senates (94–6, 105–7; cf. Cic. Fam. 6.18.1) and gladiators, prostitutes, and those who trained gladiators or ran gladiatorial schools or brothels from the senates (111–12, 123, 126–40).

⁵⁰ The seminal discussions are Mommsen, StR ³ iii. 270–9Google Scholar; Tibiletti, G., ‘Il funzionamento dei comizi centuriati alla luce della Tavola Hebana’, Athenaeum 27 (1949), 210–45Google Scholar.

⁵¹ For a detailed argument, see ‘The Reform of the Comitia Centuriata’, forthcoming in Historia.

⁵² See Taylor, L. R., Voting Districts, ch. 7 passim, with a list of the tribes and the areas they probably comprised before the Social War, 95–8Google Scholar.

⁵³ ‘The Centuriate Assembly before and after the Reform’, AJP 78 (1957), 339–43Google Scholar.

⁵⁴ 4.14–15.

⁵⁵ It is to be associated with a number of ancient rites — Paganalia, Terminalia, Faunalia, Ambarvalia; see, under the individual festivals, Warde Fowler, Roman Festivals; Wissowa, RKR ²; Latte, RRG. There are various indications of the great age of these rites. Preserved in Ovid, , Fasti 2.667–70Google Scholar, is the tradition that Terminus was already present on the site of the temple of Jupiter Capitolinus before that temple was built. From Cato, , De Agri Cultura 141Google Scholar, it is known that lustrations of the fields were accompanied, in earlier times, by a prayer to Mars. Originally a god of agriculture, Mars gradually assumed the character of a god of war and was eventually replaced in the area of agriculture by Ceres. Cf. also the so-called Seven Pagi on the right bank, an area of extreme strategic importance to the Romans which, according to the tradition, changed hands several times in the late sixth century (Dion. Hal. 2.55.4; 5.31.4; 5.36.4). Their name would appear to date to well before the establishing of the historical tradition. Dionysius twice refers to them as τοὺς καλουμένους Ἑπτ⋯ πάγους (2.55.4; 5.31.4), and Plutarch once as the land ἣν Σεπτεμπάγιον καλο⋯σιν (Romulus 25.5).

⁵⁶ Ulp., De Censibus, Dig. 50.15.4.

⁵⁷ So Frederiksen, M. W., ‘Changes in the patterns of settlement’, Hellenismus in Mittelitalien, ed. Zanker, P. (Göttingen, 1976), 341–55, esp. 345–6, 351Google Scholar. This was Frederiksen's reconsidered view; cf. his earlier ‘Republican Capua: A Social and Economic Study’, PBSR 27, n.s. 14 (1959), 90Google Scholar.

⁵⁸ Livy 43.14–15.

⁵⁹ According to the tradition, the century of land properly comprised 100 plots of two iugera each, Sic. Fl. 153.26L; Paul. exc. Fest. 47L; Var. RR 1.10.2; Hyginus 110.4L; but cf. Var. LL 5.35. Cf. Dilke, O. A. W., The Roman Land Surveyors (Newton Abbott, 1971), 133Google Scholar. The distribution of such plots, called heredia, was thought to date back to the earliest times (see Paul. exc. Fest. and Varro locc. citt. supra). The term heredium itself was indeed ancient; see Plin. NH 19.50 on its frequent use in the Twelve Tables. Pliny translates heredium as the modern hortus — a small plot of land; see also the definition of Festus — praedium parvolum — Paul. exc. Fest. 89 L. The notion that early Roman citizens were given plots of two iugera each turns up in other contexts also; see e.g. Plutarch, Publ. 21Google Scholar on the division of the land by the Anio among Appius Claudius and his followers. (For πλέθρον as iugerum, cf. Becker, W., RE xxi.1.235Google Scholar s.v. πλέθρον; Taylor, , Voting Districts 48 with n. 3.Google Scholar) There are indications of another belief that the system of centuriation was derived from the Etruscans. For example, it was thought that the limites originated in the disciplina Etrusca, Var. LL 7.7. A case for an Etruscan connection has been made by Dilke, op. cit. supra, 32–3. Such a borrowing is most likely to date to the late sixth century.

⁶⁰ Cic. Phil. 2.82.

⁶¹ If so, the distinction between iuniores and seniores was abolished in the second and third classes, although this did not mean that the censors ceased to record a citizen's age.

⁶² For this reason Palmer may well be right in suggesting that the reform of 179 concerned the enfranchisement of Formiae, Fundi and Arpinum in 188 and the founding of citizen colonies during the 180s.

⁶³ As F. Smith maintained (loc. cit. [n. 1]).

⁶⁴ Propounded in particular by Nicolet. A perennial obstacle to this view remains in the three references in Livy to tribal centuries in the first class at the end of the third century: 24.7.12; 26.22.2–11; 27.6.3. Nicolet's attempts to explain away this evidence (art. cit. [n. 1], 351–3) are not at all convincing. However, to those who still wish to maintain this view it is urged that it can be held without any corruption of Livy's text such as is proposed by Nicolet (357) on the basis of Laelius Felix' schematisation of the three assemblies, ap. Gell. N.A. 15.27.5. (The similarity between the two passages is in fact restricted to the use of the phrase generibus hominum.)

I am greatly indebted to Mr Peter Derow, Mr A. N. Sherwin-White, Professor George Forrest and the Editors of this journal for much helpful criticism of earlier drafts of this paper.