¹ This comes out very clearly in Maule, Q. F. and Smith, H. R. W., Votive Religion at Caere: Prolegomena I, Univ. of California Pubs. iv, 1 (1959)Google Scholar. Their discussion of the archaeological evidence seems to the outsider admirably cautious—their treatment of the literary sources is less so.

² F. Lammert, PW IA 1, 546 Reiterei.

³ K. Ziegler, PW XX I 2, 1490 Polybius.

⁴ Meyer, Ed., Das Römische Manipularheer=Kleine Schriften II (1924) I, 225Google Scholar; Walbank, F. W., Historical Commentary on Polybius (1957) I, 698Google Scholar. Toynbee, A. J., Hannibal's Legacy (1965) IGoogle Scholar, chap. III, Annex xii, 505.

⁵ For Strategike as the general Greek term for Kriegswissenschaft see F. Lammert, PW IVA 1, 174 Strategemata.

⁶ A. v. Premerstein, PW IV, 726 Commentarii.

⁷ P. A. Brunt, Italian Manpower (1971), Appendix 19, 625.

⁸ Livy XXV.5.5–9.

⁹ Livy XXVI.31.11 (210 B.C.); note also Varro ap. Nonius 282 (275 B.C.).

¹⁰ None of the passages adduced by Professor Brunt to show a pattern other than that in Polybius VI in operation seems to me to do so conclusively. Polybius IX.6.6, on his own interpretation, shows two legions being raised simultaneously and it is not clear that the men present in Rome do not include more than will be ultimately needed. Livy XLIII.14 seems to me too unclear and uncertain to base much argument upon; but the levy is in difficulties because the consuls will not choose unwilling men, and thus find too few present to choose from.

Liebenam, PW V, 591 Dilectus suggested that the tumultuary levy was on a simplified system and became regular in Polybius' time; but the main feature of this was that vacationes were refused, and this certainly did not become regular. The prevalence, or lack, of volunteers need not affect the choosing ceremony.

¹¹ The Servian classes appear to be of no importance in Polybius' system, though those with less than a certain amount of property serve as light-armed and those with over a certain amount have specially good armour. That this latter category was important in the second century is suggested by Cato's use of the phrase infra classem; cf. classicus. This top class was distinguished from the rest of the people for various purposes, such as the Lex Voconia. An antiquarian source on the army would have tried to bring in the whole Servian system ?

¹² By the military tribunes still, see 20.9: ‘they were accustomed … they make …’

¹³ Hill, H., ‘Census Equester’, AJP lx (1939), 357Google Scholar (cf. The Roman Middle Class (1952) 18–19), to some extent following Mommsen. Gabba, E., ‘Richerche sull' esercito professionale Romano’ Athenaeum N.S. xix (1951) 225Google Scholar and Nota to ‘Le origini della Guerra Sociale’ Athenaeum N.S. xxii (1954) 336Google Scholar. Nicolet, C., L'Ordre Equestre à l'Epoque Républicaitie I (1966) 52 ffGoogle Scholar. believes in a census equester from early times and thinks Polybius is talking of equites equis suis merentes.

¹⁴ See S. Tondo, Il ‘sacramento militiae’ nell'ambiente culturale Romano—Italica (1963), with A. Momigliano's review, JRS lvii (1967) 25Google Scholar.

¹⁵ Vegetius, , Epitoma rei militaris II.2Google Scholar; 15–17; and III.14 especially.

¹⁶ Thus Schenk, D., Die Quellen der Epitoma rei militaris, Klio Beiheft xxii (1930) 236Google Scholar—he believes Cato was only vised via Celsus; Sander, E., ‘Die Hauptquelle der Bücher 1–3 der Epitoma rei militaris des Vegetius’ Philologus lxxxvii (1932) 369Google Scholar; A. R. Neumann, PW Suppl. X (1965) 992 Vegetius. Lammert, F., on the other hand, Die Römische Taktik zu Beginn der Kaiserzeit, Philologus Suppl. xxiii 2 (1931) 30Google Scholar, opts for an early imperial source.

¹⁷ Arrian, Art of Tactics 1.2.

¹⁸ A. Schulten, PW XX 2, 1344 pilum.

¹⁹ The finds from Camp II at Renieblas, which Schulten dates to 153, might suggest that the spears were still in use then: spears were found in the quarters he assigns to the triarii (Schulten, A., Numantia IV 132Google Scholar). But on other occasions in the excavation of these camps weapons turned up in places they could not belong to, so too much weight should not be put on these finds.

²⁰ Bell, M. J. V., ‘Tactical Reform in the Roman Republican Army’ Historia xiv (1965) 404Google Scholar. He does not mention the fragment of a speech of Cato, ORF fragment 35, which is very pertinent to his argument (interea unamquamque turmam manipulum cohortem temptabam quid facere possent). In general writers on the military history of this period tend to neglect the remains of the second century orators and poets.

Troops are not, however, apparently grouped in cohorts in the earlier Spanish camps; in the Scipionic camp, they are grouped, according to Schulten (op. cit. (in note 19), III 134, 164) in bodies of three maniples of the same type, i.e. three hastati or three principes. Schulten believes that this represents eine Zeit des Übergangs to the cohort made up of one maniple from each of three lines. One may note that Polybius XI.23, explaining what a cohort is, simply says that it consists of three maniples, not specifying from which line.

²¹ Varro LL V.91 : quos hi (the decurions) ipsi sibi adoptabant, optiones vocari coepti, quos nunc propter ambitionem tribuni faciunt.

²² Eorum cognitionem esse debere; and a consul regrets that the tribuni plebis prohiberent consulem quem cuique ordinem adsignari e re publica esset eum adsignare. Cf. Cicero, In Pisonem 88.

²³ P. Couissin, Les Armes Romaines (1926) 248. Helbig, W., ‘Zur Geschichte der römischen Equitatus’ Abh. Akad. Munch. I xxiii 2 (1905) 276Google Scholar.

²⁴ Livy II.6, 19; IV.17; VI.37; VIII.39; XXVI.4, etc. Against this a couple of isolated anachronisms reflecting the first century, when the cavalry did wear scuta (XXXV.18; XLIII.6) and an inability to decide whether Badius had a scutum or a parma (XXV. 18) cannot weigh.

²⁵ undique conveniunt velut imber tela tribuno,/ configunt parmam, tinnit hastilibus umbo. (Annals 401–2 Vahlen.)

²⁶ The scholiasts and glossaries call them πϵριϕϵρϵῖς, a word Polybius used for th e πάρμη of the light troops, or στρογγύλα. Cf. also πϵμπτός βοῦς, a crescent shaped cake offered up for every four surely round ones, as the ἔβδομος βοῦς for every six σϵλῆναι, from their name again obviously rounded cakes, not oblong ones.

²⁷ Note the round shield of a horseman on a coin of Larinum, dating from the second Punic War (S.N.G. Copenhagen I Pl. 9, no. 269).

²⁸ Weege, F., ‘Oskische Grabmalerei’ Jahrb. des Arch. Inst. xxiv (1909) 91Google Scholar; Nicolet, C., ‘Equites Campani et leurs representations figurées’ Mél. d'Arch. et d'Hist. lxxiv (1962) 463Google Scholar; and Frederiksen, M. W., ‘Campanian Cavalry: a Question of Origins’ Dialoghi di Archeologia ii (1968) 3Google Scholar. The horsemen carry round or oval shields.

²⁹ Thus Ed. Meyer op. cit. (in note 4). The change was probably made by Pydna, as Reinach probably rightly identifies, on the monument recording the victory, a horseman wearing a cuirass as Roman. This means heavy cavalry, but is not very like Greek wear. (Reinach, A., ‘La frise du monument de Paul-Émile à Delphes’ BCH xxxiv (1910) 433CrossRefGoogle Scholar. Thus also H. Kähler, Der Fries vom Reiterdenkmal des Aemilius Paullus (1965) 34–35.)

³⁰ Fraccaro, P., ‘Polibio e l'accampamento romano’ Athenaeum N.S. xii (1934) 154Google Scholar = Opuscula II 307. For earlier literature see Fabricius, E., ‘Some notes on Polybius' description of Roman Camps’ JRS xxii (1932) 78Google Scholar; and Walbank, op. cit. (in note 4) 712.

³¹ Aulus Gellius, Notes Atticae X.8.

³² His list may not be complete: the removal of the cingulum militiae was perhaps already a great disgrace. On the other hand, the importance of mutatio militiae has perhaps been exaggerated by literary sources, as its effect would only be to make unsatisfactory troops worse by thrusting them into jobs for which they were unfitted.

³³ Pliny, N.H. XXXIII. 38Google Scholar.

³⁴ Livy XXVI.48.

³⁵ But Cato ORF 18 is too general to prove that it had not been introduced in the 190's (maiores seorsum atque divorsum pretium paravere bonis atque strenuis, decurionatus, optionatus, hastas donaticas, aliosque honores).

³⁶ Thus Walbank op. cit. (in note 4) 709.

³⁷ Reinach, A. J., ‘Les Origines du Pilum’ Rev. Arch. 4° série ix (1907) 236 n.2Google Scholar.

³⁸ Below, p. 28.

³⁹ The obvious example is Timaeus, who could even be described as a historian of Rome.

⁴⁰ P. Lévêque, Pyrrhos (1957) 201.

⁴¹ Xystophori were heavy cavalry, sarisophori light; but even the latter must have used both hands, at times at least, if the cavalry sarissa had any resemblance to that used by the phalanx.

⁴² D. H. XIX.12.

⁴³ de Sanctis, G., Storia dei Romani II, 373Google Scholar, n. 26.

⁴⁴ Plutarch, Pyrrhus XXI.6Google Scholar. Plutarch quotes Dionysius as saying Pyrrhus was wounded ὑσσῷ, i.e. pilo. Compare, for southern Italians at this time, Polybius XVIII.28.10 ‘Pyrrhus … employed not only Italian arms but Italian forces, placing alternately Of course Pyrrhus' Italian allies might have developed the manipular formation before Rome herself did–she was in fact said to have taken scuta etc. from the Samnites.

Ennius' lines (Annals 183 Vahlen): proletarius publicitus scutisque feroque ornatur ferro,

which would imply that the phalanx had been given up, are usually supposed to relate to 281 B.C.; but this may be mistaken, or Ennius could well be historically inaccurate—and anyway these proletarii were used, in 281, on garrison duty, not in the line.

⁴⁵ Polybius 11.33; I 40.12 suggests that some pila (ὑσσοί) were in use at Panormus in 250.

⁴⁶ E. J. Haeberlin, Aes Grave (1910) Tafel 30, 32.

⁴⁷ E. g. Marquardt, R. Staatsverwaltung ³ II 358; Fiebiger PW VII 2, 2509 Hastati; Liebenam PW VI, 1593 Exercitus; Meyer op. cit. (in note 4) 45–6; and still A. Neumann, PW Suppl. XI, 1113 Principes, A. Toynbee op. cit. (in note 4) and many others. The confused passages in Vegetius that we have dealt with are not good support for this belief.

⁴⁸ Annals 284 (Vahlen). The early sources suggest that the names for throwing spears (verutum, gaesum, pilum and hasta) are not clearly differentiated. Schulten, A., ‘Der Ursprung des Pilums’ Rh. Mus. lxvi (1911) 572Google Scholar may or may not be right in holding that the true pilum was a Spanish invention. But Ennius can use the word quite loosely in Annals 570 (V)—unless of course he is talking of the Spanish wars: pila retunduntur venientibus obvia pilis.

⁴⁹ Dionysius does not distinguish ἀσπίδϵς and θυρϵοί very clearly, but there is no clear evidence that any of his narrative sources believed in the phalanx (and thus the true ἀσπίς). Even Diodorus, for whose antiquarian pretensions see below, describes Manlius Capitolinus as armed with sword and θυρϵός. (Certainly neither ἀσπίς nor ϕάλαγξ is a safe guide, as both can be used in a general sense.)

For direct evidence about a late annalist, note Claudius Quadrigarius HRR fragment 10b on Manlius Torquatus: scuto pedestri et gladio Hispanico cinctus … scuto scutum percutit (367 B.C.).

⁵⁰ Polybius 11.33.

⁵¹ Varro LL V.89; for the true derivation as from pilus or pila, not pilum, see F. Lammert, PW VIA 2, 2369 Triarii; but see Toynbee op. cit. (in note 4) against.

⁵² A. v. Premerstein, PW XII 1–2, 1133 Legatus. At Cynoscephalae for example (Polybius XVIII.26 ) the military tribunes are still the chief officers. As we saw, the πρεσβεῖς in Polybius VI.35 may be political not military legati.

⁵³ Livy II.19 and 52 (Latins); IX.39 (Etruscans with missiles and swords; cf. their swords in 1.25, II.46); IV.37 (Volscians also with scuta); tripartite organisation IV.28 (Aequi and Volscians); tela II.10.46 and 49 (Etruscans); II.65, V.12 (Volscians). The Samnites however appear in VII.33 with erectae hastae—presumably in phalanx formation ?

⁵⁴ D.S. XXIII.2; Ined. Vat. 3; Sallust Cat. 51; Athenaeus VI.273 e–f; Zonaras VIII.9. It may be repeated here that there is no sound evidence for the view held by Mommsen that Diodorus was using Fabius Pictor (or any other very early source).

⁵⁵ Livy I.43; D.H. IV.16.

⁵⁶ Livy II.28, III.15 and 20.

⁵⁷ It is of course often held, with Nilsson, M. P., ‘The Introduction of Hoplite Tactics at Rome’, JRS xix (1929) 1Google Scholar, that the Servian constitution was specifically introduced in order to introduce, or to reflect, the new hoplite phalanx. See Last, H., ‘The Servian Reforms’, JRS xxxv (1945) 30Google Scholar, on its primary function, military rather than political. F. Cornelius, Untersuchungen zur frühen römischen Geschichte (1940) 80 puts the introduction of hoplite tactics in the mid fifth century, with the introduction of the censors; A. Toynbee op. cit. (in note 4) 510 puts it, with Bloch, Origins of Rome (1960) 105 in the prosperous late sixth century.

⁵⁸ Varro LL. V.89.

⁵⁹ Id. ib. VII.57–8.

⁶⁰ Fraccaro, P., ‘Accensi’ Athenaeum N.S. v (1927) 133Google Scholar=Opuscula II.316.

⁶¹ Id. ib.

⁶² De re militari fragment 6 (Jordan). Compare Vegetius, perhaps using Cato (above p. 18).

⁶³ Marquardt, R. Staatsverwaltung ³ 360 n.l for early literature on this chapter; more recently Ed. Meyer, op. cit. (in note 4).

⁶⁴ A. Klotz, Livius und seine Vorgänger (1940) 203.

⁶⁵ Maule and Smith (op. cit. in note 1), who date their Veian warrior-type statuette armed in the Italic manner to the early fourth century and consider it under Roman influence, argue that Livy's date is right; but they do not make out their case conclusively.

⁶⁶ Agite nunc, subsidite omnes quasi solent triarii (Plautus ap. Varro, LL V, 89Google Scholar).

⁶⁷ Fraccaro, P., ‘Ancora sull'età dell'ordinamento centuriato’ Athenaeum N.S. xii (1934) 57Google Scholar; ‘La Storia dell' antichissimo esercito Romano’ Atti del 20 Congr. Naz. di Studi Romani III (1931) 91Google Scholar = Opuscula II 293 and 287.

⁶⁸ See Fiebiger PW IA 1, 1101 Rorarii on this etymology; he prefers one that means ‘runners’.

⁶⁹ X. 393 (Marx). The chronological and other relation of rorarii to velites, a well-attested name for light troops, is impossible to establish. (There is a tendency to hold, on the basis of two somewhat dubious passages of Livy, that rorarii, who appear in VIII.8, were replaced by velites, introduced according to XXVI.4 in 211 B.C.). All that one can say certainly is that both are attested in writers of the earlier and later second century, who were probably talking of their own time. Lucilius VII.290 (rorarius veles) would suggest that both names applied to the same thing in his time at least. On the disappearance of velites, see M. J. V. Bell, op. cit. (in note 20).

⁷⁰ I am grateful to M. H. Crawford for the numismatic references.