The number of foot soldiers in a Roman legion Although we can determine with tolerable certainty the number of soldiers who, at different periods, were contained in a legion, we must bear in mind that at no epoch does this number appear to have been absolutely fixed, but to have varied within moderate limits, especially when troops were required for some special or extraordinary service. The permanent changes may be referred to four epochs.

Under the Kings.— Varro (L.L. v. § 89) and Plutarch (Rom. 13), both of whom describe the first establishment of the legion, agree that under Romulus it contained 3000 foot soldiers. The words of Plutarch indeed, in a subsequent passage (Rom. 20), would, at first sight, appear to imply that after the junction with the Sabines the number was raised to 6000; but he must be understood to mean two legions, one from each nation. It is highly probable that some change may have been introduced by Servius Tullius, but, in so far as numbers are concerned, we have no evidence.

From the expulsion of the Kings until the second year of the second Punic War.— The regular number during this space of time may be fixed at 4000 or 4200 infantry. According to Dionysius (vi.42) M. Valerius, the brother of Publicola, raised two legions (B.C. 492), each consisting of 4000, and Livy, in the first passage, where he specifies the numbers in the legions (vi.22, B.C. 378), reckons them at 4000, and a few years afterwards (vii.25, B.C. 346) he tells us that legions were raised each containing 4200 foot soldiers, and 300 horse. The legion which possessed itself of Rhegium (B.C. 281-271) is described (Liv. xxviii.28) as having consisted of 4000, and we find the same number in the first year of the second Punic War (Liv. xxi.17, B.C. 218). Polybius, in like manner (i.16), fixes the number at 4000 in the second year of the first Punic war (B.C. 263), and again in the first year of the second Punic War (iii.72, B.C. 218). In the war against Veii, however, when the Romans put forth all their energies, according to Dionysius (ix.13), an army was raised of 20,000 infantry and 1200 cavalry, divided into four legions; and, according to Polybius (ii.24), in the war against the Gauls, which preceded the second Punic War, the legions of the consuls consisted of 5200 infantry, while those serving in Sicily and Tarentum contained 4200 only, a proof that the latter was the ordinary number.

From the second year of the second Punic war until the consulship of Marius.— During this interval the ordinary number may be fixed at from 5000 to 5200. Polybius, indeed, in his treatise on Roman warfare, lays it down (vi.20) that the legion consists of 4200 foot soldiers, and in cases of peculiar danger of 5000. However, the whole of the space we are now considering, was in fact a period of extraordinary exertion, and hence from the year B.C. 216, we shall scarcely find the number stated under 5000 (e.g. Polyb. iii.17, Liv. xxii.36, xxvi.28, xxxix.38), and after the commencement of the Ligurian war it seems to have been raised to 5200 (Liv. xl.1, 18, 36, xli.9, but in xli.21 it is again 5000). The two legions which passed over into Africa under Scipio (B.C. 204) contained each 6200 (Liv. xxix.24), those which served against Antiochus 5400 (Liv. xxxvii.39), those employed in the last Macedonian war 6000 (Liv. xlii.31, xliv.21, comp. xliii.12), but these were special cases.

From the first consulship of Marius (B.C. 107) until the extinction of the legion.— For some centuries after Marius the numbers varied from 5000 to 6200, generally approaching to the higher limit. Festus (s.v. sex milium et ducentorum) expressly declares that C. Marius raised the numbers from 4000 to 6200, but his system in this respect was not immediately adopted, for in the army which Sulla led against Rome to destroy his rival, the six complete legions (ec tagmata teleia) amounted to 30,000 men (Plut. Sull. 9, Mar. 35, but the text in the latter passage is doubtful). In the war against Mithridates again, the 30,000 men of Lucullus formed five legions (Appian. Mithr. 72). Comparing Plutarch (Cic. 36) with Cicero (ad Att. v.15), we conclude that the two legions commanded by the latter in Cilicia contained each 6000. Caesar never specifies in his Commentaries the number of men in his legions, but we infer that the 13th did not contain more than 5000 (B. C. i.7), while the two mentioned in the fifth book of the Gallic war (c48, 49) were evidently incomplete. In Appian, M. Antonius is represented as calculating the amount of 28 legions at upwards of 170,000 men, that is nearly 6100 to each legion, but he seems to include auxiliaries (twn suntassomenwn). During the first century the standard force was certainly 6000, although subject to constant variations according to circumstances, and the caprice of the reigning prince. The legion of Hadrian, if we can trust Hyginus, was 5280, of Alexander Severus 5000 (Lamprid. Sev. 50), that described by Vegetius (ii.6), to whatever period it may belong, 6100, and most of the grammarians agree upon 6000 (e.g. Serv., ad Virg. Aen. vii.274; Isidor. Orig. ix.3 §46; Suidas, s.v. legewn, but Hesychius gives 6666). The Jovians and Herculeans of Diocletian and Maximian formed each a corps of 6000 (Veget. i.17), but beyond this we have no clue to guide us. If we believe the tagmata of Zosimus and the ariqmoi of Sozomen to designate the legions of Honorius, they must at that epoch have been reduced to a number varying from 1200 to 700.

William Smith, A Dictionary of Greek and Roman Antiquities, John Murray, London, 1875. p 489-511.