A Dictionary of Greek and Roman Antiquities (1890), EBUR, EXAUGURA´TIO, EXE´RCITUS

Home Collections/Texts Perseus Catalog Research Grants Open Source About Help

Hide browse bar Your current position in the text is marked in blue. Click anywhere in the line to jump to another position:

This text is part of:

Greek and Roman Materials

View text chunked by:

EXE´RCITUS

EXE´RCITUS (στρατός), army.

Greek

The earliest notices which we possess of the military art among the Greeks are those contained in the Homeric poems. The unsettled state of society in the first ages of Greece led to the early and general cultivation of the art of arms, which were habitually worn for defence, even when aggressive warfare was not intended (Thuc. 1.6). But the Homeric poems contain an exhibition of combined military operations in their earliest stage. Warlike undertakings before the time described in them can have been little else than predatory inroads (βοηλασίαι, Il. 11.672). A collection of warriors exhibiting less of organisation and discipline than we see depicted in the Grecian troops before Troy, would hardly deserve the name of an army. The organisation which we see there, such as it was, arose, not from any studied, formative system, but naturally out of the imperfect constitution of society in that age. Every freeman in those times was of course a soldier; but when all the members of a family were not needed to go upon an expedition under the command of their chieftain or king, those who were to go seem to have been selected by lot (Il. 24.400). As the confederated states, which are represented as taking part in the Trojan war, are united by scarcely any other bond than their participation in a common object, the different bodies of troops, led by their respective chieftains, are [p. 1.767]far from being united by a common discipline under the command-in-chief of Agamemnon. A common epithet for allies is “called from afar” (τηλεκλειτοί, Il. 5.491, 6.111). Each body obeys its own leader, and follows him to the conflict, or remains inactive, according as he chooses to mingle in the fight or not. Authority and obedience are regulated much more by the nature of the circumstances, or by the relative personal distinction of the chieftains, than by any law of military discipline. Gifts (δῶρα) were given to them at the end of service; and such may be considered as the beginning of pay being given to soldiers (Il. 17.225). Agamemnon sometimes urges the chieftains to engage, not by commands, but by taunts (Il. 4.338 ff., 368 ff.). Accordingly, nothing like the tactics or strategy of a regularly disciplined army is to be traced in the Homeric descriptions of battles. Each chieftain with his body of troops acts for himself, without reference to the movements of the rest, except as these furnish occasion for a vigorous attack, or, when hard pressed, call for assistance from the common feeling of brotherhood in arms. The wide interval which in the Homeric age separated the noble or chieftain from the common freeman, appears in as marked a manner in military as in civil affairs. The former is distinguished by that superior skill and prowess in the use of his arms, which would naturally result from the constant practice of warlike exercises, for which his station gave him the leisure and the means. A single hero is able to put to flight a whole troop of common soldiers. The account of a battle consists almost entirely of descriptions of the single combats of the chiefs on both sides; and the fortune of the day, when not overruled by the intervention of the gods, is decided by the individual valour of these heroes. While the mass of the common soldiers were on foot, the chiefs rode in chariots [CURRUS], which usually contained two, one to drive (ἡνίοχος) and one to fight (παραιβάτης). In these they advanced against the antagonists whom they singled out for encounter, sometimes hurling their spears from their chariots, but more commonly alighting, as they drew near, and fighting on foot, making use of the chariot for pursuit or flight. The Greeks did not, like the ancient Britons and several nations of the East, use the chariot itself as an instrument of warfare. Cavalry was unknown at that time to the Greeks, and horsemanship but very rarely practised; the ἱππῆες of Homer are the chieftains who ride in chariots. These chiefs are drawn up in the front of the battle array (Il. 4.297, 505, πρόμαχοι, προμάχεσθαι); and frequently the foot-soldiers seem to have done nothing but watch the single combats of their leaders, forming in two opposite, parallel lines, between which the more important single combats are fought. How they got the chariots out of the way when the foot-soldiers came to close quarters (as in Il. 4.427 ff.) is not described.

Though so little account is usually made of the common soldiers (πρυλέες, Il. 11.49, 12.77), Homer occasionally lays considerable stress on their orderly and compact array; the Atreidae are honourably distinguished by the epithet κοσμήτορε λαῶν (Il. 1.15). Nestor and Menestheus were also skilled in marshalling an army (Il. 2.553, 4.293 ff.). The troops were naturally drawn up in separate bodies according to their different nations. It would appear to be rather a restoration of the old arrangement than a new classification, when Nestor (Il. 2.362) recommends Agamemnon to draw the troops up by tribes and phratries. Arranged in these natural divisions, the foot-soldiers were drawn up in densely compacted bodies (πυκιναὶ φάλαγγες)--shield close to shield, helmet to helmet, man to man (Il. 13.130, 16.212 ff.). In these masses, though not usually commencing the attack, they frequently offer a powerful resistance, even to distinguished heroes (as Hector, Il. 13.145 ff., comp. 17.267, 354 ff., 13.339), the dense array of their spears forming a barrier not easily broken through. The signal for advance or retreat was not given by instruments of any kind, but by the voice of the leader. A loud voice was consequently an important matter, and the epithet βοὴν ἀγαθὸς is common. The soldiers advanced and engaged in battle with loud shouting (ἀλαλητός, Il. 4.436, 14.393). The trumpet, however, was not absolutely unknown (Il. 18.219). Respecting the armour, offensive and defensive, see ARMA No engines for besieging are found. There were in the army, besides the hoplites, light-armed troops, archers and slingers (Il. 13.767).

Under the king or chieftain who commands his separate contingent we commonly find subordinate chiefs, who command smaller divisions. It is difficult to say whether it is altogether accidental or not, that these are frequently five in number. Thus the Myrmidons of Achilles are divided into five στίχες, each of 500 men. Five chiefs command the Boeotians; and the whole Trojan army is formed in five divisions, each under three leaders. (Il. 4.295 ff., 16.171-197, 2.494, 495, 12.87-104.) The term φάλαγξ is applied either to the whole army (as Il. 6.6), or to these smaller divisions and subdivisions, which are also called στίχες and πύργοι (Il. 11.90, 4.333).

When an enemy was slain, it was the universal practice to stop and strip off his arms, which were carefully preserved by the victor as trophies. The division of the booty generally was arranged by the leader of the troop, for whom a portion was set aside as an honorary present (γέρας, Il. 1.118, 368, 392). The recovery of the dead bodies of the slain was in the Homeric age, as in all later times, a point of the greatest importance, and frequently either led to a fierce contest (Il. 16.756 ff.), or was effected by the payment of a heavy ransom (Il. 24.502). (Köpke, Krieqswesen der Griechen in heroischen Zeitalter; Wachsmuth, Hellen. Alterthumsk. vol. 2.110; Grote, History of Greece, vol. ii. p. 106; Buchholz, Die Homerischen Realien, 2.1, 303-331.)

After the heroic age considerable impulse was given to the cultivation of the military art by the conquests of the Thessalians (the first Grecian people, apparently, that employed cavalry, to the use of which their conquests were probably in great part owing) and Dorians, among the latter of whom the art of warfare was earliest reduced to system. The distinction of heavy and light armed foot-soldiers of course [p. 1.768]took its rise with the beginnings of military service, the poorer class being unable to provide themselves with the more efficient, but more costly weapons of those who were better off than themselves. Political considerations tended to make the distinction more marked and systematic. The system of military castes was indeed unknown among the Greeks, though something answering the same purpose existed in the earliest times, when the nobles and their more immediate dependents and retainers, having greater leisure for the cultivation of skill in the use of arms and greater means for procuring them, were separated in that respect by a wide interval from the lower class; while conversely, military superiority was the most direct means of securing political supremacy. Hence, as soon as the distinction between the nobles (the privileged class) and the commonalty (demus) was established, it became the object of the former to prevent the latter from placing themselves on a par with them in military strength, and so the use of the full armour of the heavy-armed, infantry was reserved by the former for themselves; and when, in times of distress, it was found necessary to entrust the demus with full armour, the result was not uncommonly a revolution (as was in some degree the case at Mytilene, Thuc. 3.27). But in the democracies this distinction as regards the kinds of service depended merely upon the greater or less ability of the citizens to procure arms. In the Greek commonwealths all those who enjoyed the privileges of citizens or freemen were held bound to serve as soldiers when called upon, and were provided with arms and trained in military exercises as a matter of course. The modern system of standing armies was foreign to Greek habits, and would have been dangerous to the liberties of the different commonwealths, though something of the kind may be seen in the body-guards, usually of mercenary troops, kept by tyrants. The mercenaries in the pay of Alexander of Pherae formed a considerable army. Practically too, from the continuity of the warlike operations in which they were engaged, the armies of Philip and Alexander of Macedon, and their successors, became standing armies. The thousand λογάδες at Argos (Thuc. 5.67; Diod. 12.75), the sacred band at Thebes (Plut. Pel. 18), and the Arcadian ἐπάριτοι (EPARITI) were not considerable enough to be called armies. The employment of mercenary troops might have led to the use of standing armies, had it not been that the use of them characterised the decline of the Grecian states, so that the circumstances which led to their employment also rendered it impossible to provide the resources for their maintenance, except when they were immediately needed. Still, as in the case of the Scythian bowmen at Athens, individual corps of mercenaries might be regularly maintained. Slaves were but rarely trusted with arms; and when it was the case, they were usually manumitted. The Greek armies accordingly were national armies, resembling rather the militia than the regular armies of modern times. Their smallness in comparison with modern armies must be noticed. The largest Greek armies we know of as having operated in Hellas proper were, at Plataea, 38,700 hoplites and 69,500 ψιλοί (Hdt. 9.28 ff.); in the first invasion of Attica by the Lacedaemonians, 70,000 (Plut. Per. 33); in the invasion of Laconia by Epaminondas, 70,000 (Plut. de Glor. Ath. 2). At Mantineia, in 362 B.C., 33,000 Thebans fought against 23,000 Lacedaemonians, according to Diodorus (15.84).

In all the states of Greece, in the earliest as in later times, the general type of their military organisation was the phalanx, a body of troops in close array with a long spear as their principal weapon. It was among the Dorians, and especially among the Spartans, that this type was most rigidly adhered to. See Tyrtaeus passim, who insists on the especial duties of fighting ἐν προμάχοισι, and each keeping his place in the phalanx. The strength of their military array consisted in the heavy-armed infantry (ὁπλῖται). They attached comparatively small importance to their cavalry, which was always inferior (Xen. Hell. 6.4, § 10). Indeed, the Thessalians and Boeotians were the only Greek people who distinguished themselves much for their cavalry; scarcely any other states had territories adapted for the evolutions of cavalry. The Spartan army, as described by Xenophon, was probably in all its main features the same that it was in the time of Lycurgus. The institutions of that lawgiver converted the body of Spartan citizens into a kind of military brotherhood, whose almost sole occupation was the practice of warlike and athletic exercises. The whole life of a Spartan was little else than either the preparation for or the practice of war. The result was, that in the strictness of their discipline, the precision and facility with which they performed their military evolutions, and the skill and power with which they used their weapons, the Spartans were unrivalled among the Greeks, so that they seemed like real masters of the art of war (τεχνίτας τῶν πολεμικῶν), while in comparison with them other Greeks appeared mere tyros (αὐτοσχεδιαστὰς τῶν στρατιωτικῶν, Xen. Rep. Laced. 13.5; ἄκροι τεχνῖται καὶ σοφισταὶ τῶν πολεμικῶν, Plut. Pel. 23). The heavy-armed infantry of the Spartan armies was composed partly of genuine Spartan citizens, partly of Perioeci (e. g. Thuc. 4.8; comp. Grote, Hist. of Greece, vol. ii. p. 458). In later times, as the number of Spartan citizens decreased, the Perioeci constituted the larger portion, a fact which renders nugatory all attempts to connect the numbers of the divisions of the army with the political divisions of the Spartan citizens. The proclamation of the levy was called (φρουρὰν φαίνειν (Xen. Hell. 3.5, 6; 4.2, 9, &c.); but if the service was abroad, it was called στρατιὰν ποιεῖν (ib. 5.2, 20): cf. A. Bauer in Müller's Handbuch, 4.247. Every Spartan citizen was liable to military service (ἔμφρουρος) from the age of twenty to the age of sixty years. In later times the father of three sons was exempt (Arist. Pol. 2.9, 13). Those beyond that age were, however, sometimes employed in the less arduous kinds of service--as at Mantineia, where they had charge of the baggage (Thuc. 5.72). On the occasion of any military expedition, the kings at first, and afterwards the ephors, made proclamation what class, according to age, were to go on the expedition (τὰ ἔτη εἰς ἃ δεῖ στρατεύεσθαι, Xen. Rep. Lac. 11.2): as, for example, all citizens between twenty and [p. 1.769]thirty, or between twenty and thirty-five, &c. (τὰ δέκα ἀφ᾽ ἥβης, τὰ πεντεκαίδεκα ἀφ᾽ ἥβης, &c.). On one occasion (B.C. 418), on a sudden emergency, when probably there was not time to collect the Perioeci, all the citizens of the military age were called forth (Thuc. 5.64).

The political and military divisions of the Spartans were mixed up together in some way which it is not easy to unravel. The whole life of a Spartan was passed in the discipline of a kind of camp. The citizens messed together in companies, and slept in a sort of barracks. It appears from Xenophon (Rep. Lac. 11.4 ff.) that the whole body of citizens of military age was divided into six divisions called μόραι (πολιτικαὶ μόραι he terms them: we are not to read ὁπλιτικῶν: cf. Xen. Hell. 5.3, 25), under the command or superintendence of a polemarch, each mora being subdivided into two λόχοι (commanded by λοχαγοί), each λόχος into two πεντηκοστύες (headed by πεντηκοστῆρες), each πεντηκοστὺς into two ἐνωμοτίαι (headed by enomotarchs). Xenophon indeed speaks of four λόχοι, but compare Xen. Hell. 7.4, 20, τῶν δώδεκα λόχων: 5, 20; and E. Müller, Jahrb. für Phil. 75.99. The ἐνωμοτίαι were so called from the men composing them being bound together by a common oath (τάξις τις διὰ σφαγίων ἐνώμοτος, Hesych. sub voce). It was composed of men of the same age, as is implied in the fact that the members of the enomoty were trained to act together, and that men under thirty or thirty-five years were often detached in a battle to pursue the light troops of the enemy (Xen. Hell. 4.5, 15, 16; Grote, 2.459), It is to be further noticed that fathers, brothers, and sons appear in different μόραι (Xen. Hell. 4.5, 10); and inhabitants of the same locality did not serve together, for we find the Amycleans scattered through the whole army (ib. § 11). These were not merely divisions of troops engaged in actual military expeditions. The whole body of citizens at all times formed an army, whether they were congregated at head-quarters in Sparta, or a portion of them were detached on foreign service. Herodotus (1.65) speaks of enomoties, triacades (which are not mentioned elsewhere), and syssitia as military divisions, and we learn that the polemarchs presided over the public tables (Plut. Lyc. 12). But these were not military divisions, but civil societies framed in Sparta to increase the feeling of comradeship so useful in war (Plat. Legg. 1.625 E). It was a principal feature of Spartan discipline that among the youths the elder should teach the younger, and so the wellknown relation of the εἰσπνήλας and ἀϊτας was established. (For details, see Gilbert, Staats-alterthümer, 1.70.) When a portion of the citizens was sent out on foreign service, the army that they formed was arranged in divisions corresponding to, and bearing the same names as, the divisions of the entire military force of Sparta, i. e. of the entire body of citizens of military age. As has already been remarked, an army sent on foreign service consisted of citizens between certain ages, determined according to the number of soldiers wanted. So that, as it would seem, every enomotia of the general body sent out a certain proportion of its numbers for the expedition in question, who (with some Perioeci) formed an enomotia of the army so sent; and the detachment of those enomotiae which formed a mora of the whole body of citizens, formed (apparently) a mora of the army on service. All the accounts that we have of Spartan military operations indicate that the Perioeci who served as heavy-armed soldiers, formed integral members of the different divisions to which they were attached; so that an enomotia, pentecostys, &c., in the field, would contain a number of soldiers who did not belong to the corresponding larger divisions of the whole body of citizens of military age (Thuc. 4.8, 38; Xen. Hell. 6.4, 15-17, 7.4, 27). However, Gilbert (op. cit. 1.74 ff.) thinks that up to 425 B.C. the Perioeci served apart, and he refers for the time of the Persian Wars to Hdt. 9.10, 11, 28. Thirlwall (Hist. of Greece, vol. i. app. ii.) talks of thirty families being represented in the army by thirty soldiers; an idea totally at variance with all the accounts that we have. Supposing a family to consist of a father and three sons, if the latter were above twenty, and the father not above sixty years of age, all would be soldiers, liable to be called out for active service at any time; and according to the limits of the age proclaimed by the ephors, one, two, three, or all of them might be called out at once. The strength of a mora on actual service, of course, varied according to circumstances. To judge by the name pentecostys, the normal number of a mora would have been 400; but 500, 600, and 900 are mentioned as the number of men in a mora on different occasions (Plut. Pel. 17; Xen. Hell. 4.5, 12; Schol. ad Thuc. 5.66; Diod. 15.32, &c.; Müller, Dorians, 3.12.2, note t). That these variations arose from variations in the number of Spartan citizens (Haase in Ersch and Gruber's Encyclopädie, art. Phalanx), is an assumption which leaves out of sight the proportion of citizens called out, and the number of Perioeci in the army. (Of the 292 heavy-armed soldiers who surrendered at Sphacteria, 120 were Spartans, Thuc. 4.38. At the battle of Plataeae, one-half of the heavy-armed soldiers of the Lacedaemonians were Spartans.) When in the field, each mora of infantry was attended by a mora of cavalry (Xen. Rep. Lac. 11.4), consisting at the most of 100 men, and commanded by an hipparmost (ἱππαρμοστής, Xen. Hell. 4.4, § 10; 5.12). Plutarch (Plut. Lyc. 23) mentions squadrons (οὐλαμοί) of fifty, which may possibly be the same divisions. It is not easy, however, to see in what manner the cavalry could have been thus apportioned, or how each mora of cavalry could have “belonged to a mora of infantry without being in close connexion with it” (as Müller says). The cavalry seems merely to have been employed to protect the flanks, and but little regard was paid to it. The feeblest and least ambitious served in the cavalry (Xen. Hell. 6.4, 10). The Spartans generally used mercenary cavalry (Thuc. 2.9; Xen. Hipp. 9.4). The corps of 300 ἱππεῖς (Hdt. 8.124) formed a sort of body-guard for the king, and consisted of the flower of the young soldiers. Though called horsemen, they fought on foot (Strabo, p. 482). Their commanders were called ἱππαγρέται (Xen. Rep. Lac. 4.3; Hesych. sub voce).

Thucydides in his account of the battle of Mantineia (5.68) describes the Lacedaemonian [p. 1.770]army as divided into seven lochi, each containing four pentecostyes, and each pentecostys four enomotiae, with thirty-two men in each; so that the lochus here is a body of 512 men, and is commanded by a polemarch. It is clear, therefore, that the lochus of Thucydides, in this instance, answers to the mora of Xenophon. It would be absurd to suppose such a large number to be the ordinary lochus or tactical unit; the largest such tactical lochus known is the Sacred Lochus at Thebes, which consisted of 300. As on this occasion the pentecostys contained four instead of two enomotiae, and as four pentecostyes were thrown together into one division, Thucydides may have been led to call this division a lochus, as being next above the pentecostys, though it was, in fact, a mora commanded by a polemarch (Thirlwall, l.c. p. 445; comp. Arnold on Thuc. 5.68). Aristotle appears to use the terms lochus and mora indiscriminately (Λακώνων πολιτ. Fr. 5 and 6; Photius s. v. λόχοι). The suggestion of Arnold (l.c.) that one of the seven lochi spoken of consisted of the Brasidean soldiers and Neodamodes, who would not be taken account of in the ordinary divisions of the Spartan forces, is not unlikely, and would explain the discrepancy between the number of lochi (or morae) here, and the ordinary number of six morae; but even independently of it, no difficulty need be felt with respect to that particular point, as the whole arrangement of the troops on that occasion was a departure from the ordinary divisions. The Neodamodes were not usually incorporated in the morae (Xen. Hell. 4.3.15). The mora was used for the arrangement in the camp (Xen. Rep. Lac. 12.5), and formed the highest unit for the levy, but tactically it had no importance (Bauer, op. cit. 249, note).

It seems a probable opinion that the number of morae in the Spartan military force had reference to the districts into which Laconia was divided. These, including Sparta and the districts immediately around it, were six in number. Perhaps, as Thirlwall suggests, the division of the army may have been founded on the fiction that one mora was assigned for the protection of each district. The same writer also suggests a very probable explanation of the λόχος Ριτανάτης which Herodotus (9.53) speaks of, and of which Thucydides (1.20), though doubtless erroneously, denies the existence. Thirlwall suggests that as each mora consisted of four lochi, the four lochi of the mora belonging to the district of Sparta may have been distributed on the same principle among the four κῶμαι--Limnae, Cynosura, Mesoa, and Pitana--of which Sparta was composed.¹ Gilbert (op. cit. 1.74) supposes that the λόχος was composed of the inhabitants of Pitana, but that its official title was not derived from that district.

A Spartan army, divided as above described, was drawn up in the dense array of the phalanx, the depth of which depended upon circumstances. An ἐνωμοτία sometimes made but a single file (reading εἰς ἕνα), sometimes was drawn up in three or six files (στίχοι, Xen. Rep. Lac. 11.4; or λόχοι, as they are called by Polybius). At the battle of Mantineia the phalanx was eight deep, so that each enomotia made four files (Thuc. 5.68; comp. Xen. Hell. 3.2.16, 6.2.21). At the battle of Leuctra it was twelve deep (Xen. Hell. 6.4, § 12). The enomotarch stood at the head of his file (πρωτοστάτης), or at the head of the right-hand file, if the enomotia was broken up into more than one. The last man was called οῦραγός. The divisions of the rows of twelve men each were πεμπάδες (Xen. Hell. 7.3, 6; cf. Cyr. 2.1, 22). It was a matter of great importance that he, like the enomotarch, should be a man of strength and skill, as in certain evolutions he would have to lead the movements (Xen. Mem. 3.1, 8, Cyr. 7.5, 5; cf. 3.4, 41; Hom. Il. 4.297). The commander-in-chief, who was usually the king (after the affair of Demaratus and Cleomenes it was the practice not to send out both kings together, Hdt. 5.75; but comp. 6.73), had his station sometimes in the centre (as at Mantineia, Thuc. 5.72), but more commonly (as at Leuctra) on the right wing. The deployments by which the arrangements of the phalanx were altered took place under the direction of the enomotarch. When the troops were drawn up in a line in the ordinary battle array, they were said to be ἐπὶ φάλαγγος. Supposing an enomotia to consist of twenty-five men, including its leader, and to be drawn up eight deep, the front line of the army would consist of 288. In an ordinary march the army advanced ἐπὶ κέρως (or κατὰ κέρας, Xen. Hell. 7.4, § 23), the first enomotia of the right wing filing off, and the rest in succession following it; so that if the enomotia was drawn up in three or two files, the whole army would march in three or two files. The most usual arrangement was in two files, εἰς δύα (Xen. Hell. 7.4, § 23, 3.1.22; Polyaen. 2.1.10). According as there were one or two or four columns, the march was called μονοφαλαγγία, διφαλαγγία, τετραφαλαγγία, of which there are the most various species given in the tactical writers (Asclep. 11; Ael. Tact. 37). If an army in marching order had to form in phalanx, the movement began with the hindmost enomotia of the column, which placed itself on the left of (παρ᾽ ἀσπίδας) and on a line with (εἰς μέτωπον) the enomotia before it. These two then performed the same evolution with respect to the last but two, and so on, till all were in a line with the first enomotia, which now, with the commander-in-chief at its head, occupied the extremity of the right wing. This evolution was called παραγωγή (Xen. Rep. Lac. 11.6), a name also given to the reverse movement, when a phalanx had to fall into marching order, and to subordinate movements of the same kind for changing the depth of the phalanx. In the latter the evolutions were conducted on much [p. 1.771]the same principle. Thus, if the depth of the phalanx was to be diminished by half, the hinder portion of each enomotia marched forwards and placed itself on the left of the half in front of it. Similarly, if the depth had to be increased, the left-hand portion of each enomotia faced about towards the right, took up its station in the rear, and then, facing to the left again, assumed its proper position. (Xen. Rep. Lac. 11.8.) The facing to the right was always the usage, because if the evolution were performed in the face of an enemy, the shielded side could be presented towards him. Wheeling to the right was called ἐπὶ δόρυ: wheeling to the left ἐπ᾽ ἀσπίδα. With the cavalry the former was also called ἐπὶ δόρυ, the latter ἐφ᾽ ἡνίαν. Modifications of this evolution, conducted on the same principle, were employed if the depth had to be increased or diminished in any other proportion. It is very likely that at those points of the files where in such evolutions they would have to separate, there were placed men suitable for taking their station in the front rank, where it was always an object to get the best men. These would answer to the δεκάδαρχοι and πεμπάδαρχοι of Xenophon. (Cyrop. 2.1, 23; comp. Hipparch. 2.6, 4.9.) If an enemy appeared in the rear, it was not enough that the soldiers should face about towards the enemy. The Spartan tactics required that the stoutest soldier should be opposed to the enemy. This was accomplished by the manoeuvre termed ἐξελιγμός. Of this there were three varieties:--1. The Macedonian. In this the leader of each file kept his place, only turning towards the enemy; the man behind him (ἐπιστάτης) retreating and again taking up his station behind him, and so on. In this way the army retreated from the enemy by a distance equal to its depth. 2. The Laconian (the one usually adopted by the Macedonian phalanx of Philip and Alexander). This was the reverse of the preceding, the rear man remaining stationary and the others advancing successively one before the other. In this way of course the army advanced against the enemy by a distance equal to its depth. 3. The Cretan. In this the leader and rearman, the second and last but two, and so on, changed places, so that the whole army remained at the same distance from the enemy. This species was also called χορεῖος (Haase ad Xen. Rep. Lac. 11.9; Müller, 3.12.8; Aelianus, Tact. 26, 27, 33). These evolutions would of course leave the general on the left wing. If it was deemed expedient that he should be upon the right, it was not enough that he should simply remove from the left to the right, the whole army had to reverse its position, so that what was the left wing should become the right. This was effected by an exeligmus, termed (at least by the later tacticians) ἐξελιγμὸς κατὰ ζυγά, as contrasted with the ἐξελιγμὸς κατὰ στίχους. Further evolutions were the different kinds of wheelings (ἐπιστροφαί). The quarter wheel was called ἀναστροφή, the half περισπασμός, the three quarters ἐκπερισπασμός, in which one quarter wheel more brought them back to their original position, ἐπικατάστασις. The pivot was the πρωτοστάτης (Hermann-Droysen, 1.42). If the army changed its front by wheeling round through a half circle, round one corner as a pivot, the movement seems to have been expressed by περιπτύσσειν or ἀναπτύσσειν (Xen. Anab. 1.1. 0, 9; cf. Hermann-Droysen, 1.47, note 2). One more evolution remains to be noticed. Suppose an enemy appeared on the right, while the army was marching in column, two abreast. The different lochi wheeled round through a quadrant of a circle, round their leader, as on a pivot, so that the army presented twenty-four columns to the enemy, consisting of two files each, and separated by a considerable interval from each other. The depth of the whole body was then lessened, and these intervals filled up by the ordinary paragoge, and by the different lochi siding up nearer to each other in case the intervals still remained too great. If it was necessary for the general to take his station on the right, this would be effected, as in other cases, by an ἐξελιγμός (cf. Grote, 2.457). Similar manoeuvres took place if the enemy appeared on the left, though, as this was the shielded side of the soldiers, and the danger was consequently less, it was frequently thought sufficient to keep the enemy in check by means of the cavalry and light troops. (Xen. Rep. Lac. 11.10.) Bauer (op. cit. 257), however, represents this manoeuvre as being effected according to a different order. He supposes (1) the ordinary παραγωγὴ ἐς μέτωπον παρ᾽ ἀσπίδας to form a phalanx; then (2) a wheel round through a quadrant of a circle to the right or left, according to the quarter from which the enemy appeared. One point that a general had to be on his guard against was the tendency of an army, when advancing ἐπὶ φάλαγγος, to sheer off towards the right, each man pressing closer to his righthand neighbour in order to protect his unshielded side, so that the right wing frequently got beyond the left wing of the enemy. (See especially the account of the battle of Mantineia, Thuc. 5.71.) A slight consideration will show that the analogy traced between the evolutions of an army and those of a chorus is by no means fanciful. One kind of ἐξελιγμὸς was, we have seen, called χορεῖος. The importance attached to the war dances among the Spartans as a means of military training was consequently very great. [CHORUS]

When an army was led to attack a height, it was usually drawn up in what were termed λόχοι ὄρθιοι, a term which merely implies that the lochi had greater depth than breadth (παράμηκες μὲν λέγεται πᾶν τάγμα ὃ ἂν τὸ μῆκος ἔχῃ πλεῖον τοῦ βάθους, ὄρθιον δὲ ὃ ἂν τὸ βάθος τοῦ μήκους, Aelian. Tact. 100.29). The breadth of the lochi would, of course, vary according to circumstances. They were drawn up with considerable intervals between them. In this way the army presented a considerable front to the enemy, and was less liable to be thrown into confusion than if drawn up in close phalanx, while at the same time the intervals between the lochi were not left so great that the enemy could safely press in between them. (Xen. Anab. 4.2, § 11; 8, § § 10-19; 5.4.22; Cyrop. 3.2.6; Anab. 4.3.17; Polyaen. Strat. 5.16.1.) There is no ground for affirming that a λόχος ὄρθιος was drawn up in two files, or even one, as Sturz (Lex. Xen.) and Grote (8.403) say. Such an arrangement would be perfectly useless for attack. This [p. 1.772]system of arrangements, which in breaking up the rigid phalanx formed some approximation to the Roman manipular tactics, was not, however, employed except in the particular case mentioned. This was an innovation due to the genius of Xenophon, and it was frequently employed in later Greek tactics.

In special circumstances, such as those of the retreating Greeks in the Anabasis, the arrangement in a hollow square was adopted (cf. the τετράγωνος τάξις in Thuc. 4.125), and in the retreat from Syracuse (Thuc. 7.78); the troops being so placed that by simply facing about they presented a front for battle on whichever side it was necessary. But if the enemy were following, it proved a very bad arrangement (Xen. Anab. 3.4, 19). The term πλαίσιον was generally applied to an army so arranged, whether square or oblong. Afterwards the term πλαίσιον was restricted to the square, the oblong being called πλίνθιον.

Though at first sight the arrangement and manoeuvres of a Lacedaemonian army seem exceedingly complex, they were in reality quite the reverse, owing to the carefully graduated system of subordination which prevailed (σχεδὸν γάρ τοι πᾶν τὸ στρατόπεδον τῶν Λακεδαιμονίων ἄρχοντες ἀρχόντων εἰσἰ, Thuc. 5.66). The commands of the general were issued in the first place to the polemarchs, by these to the lochagi, by these again to the pentecosteres, by the latter to the enomotarchs, and by these last to their respective divisions. From the orderly manner in which this was done, commands were transmitted with great rapidity: every soldier, in fact, regulating the movements of the man behind him, every two being connected together as πρωτοστάτης and ἐπιστάτης.

In later times the king was usually accompanied by two ephors, as controllers and advisers (Xen. Rep. Lac. 13.5). These, with the polemarchs, the four Pythii, three peers (ὅμοιοι), who had to provide for the necessities of the king in war,--among whom was the κρεοδαίτης, who was an important person, the post being held by Lysander (Plut. Lys. 23),--the λαφυροπῶλαι, and some other officers, such as the ταμίαι, treasurers, and the Ἑλλανοδίκαι, judges of disputes (Xen. Rep. Lac. 13.1. 1), constituted the οἱ περὶ δαμοσίαν of the king [DAMOSIA]. The polemarchs also had some sort of suite or staff with them, called συμφορεῖς (Plut. Pel. 17; Xen. Hell. 6.4, § 14), just as the king had the 300 so-called ίππεῖς. The polemarchs took the place of the king in cases of emergency. With the exception of the enomotarchs, the superior officers and those immediately about them are not to be reckoned with the division which they led. They stood distinct, forming what was called the ἄγημα (Xen. Rep. Lac. 11.9

The Spartan and Perioecan hoplites were accompanied in the field by helots, partly in the capacity of attendants, partly to serve as lightarmed troops. The number attached to an army was probably not uniform. At Plataeae each Spartan was accompanied by seven helots; but that was probably an extraordinary case. One helot in particular of those attached to each Spartan was called his θεράπων, and performed the functions of an armourer or shieldbearer (Eustath. ad Dionys. Per. 533). Xenophon (Xenoph. Hell. 4.5.14; 8.39) calls them ὑπασπισταί (cf. Hdt. 5.111; Müller, Dor. 3.3.2), or δορυφόροι (Xen. Hell. 4.5, § 8). In extraordinary cases, helots served as hoplites, and in that case it was usual to give them their liberty (Thuc. 7.19, 4.80, 5.34). Distinct corps were sometimes composed entirely of these Neodamodes. A separate troop in the Lacedaemonian army was formed by the Sciritae (Σκιρῖται), originally, no doubt, inhabitants of the district Sciritis. In battle, they occupied the extreme left of the line. On a march, they formed the vanguard, and were usually employed on the most dangerous kinds of service. (Thuc. 5.67, with Arnold's note; Xen. Cyrop. 4.2, § 1.) Similarly the Tegeans laid claim to a position on the right wing (Hdt. 9.26 ff.).

Light-armed troops (ψιλοί) are mentioned in the fourth year of the Peloponnesian War (Thuc. 3.1), and in 424 B.C. a corps of bowmen was established by the Spartans (Thuc. 4.55). Brasidas used great numbers of such light-armed troops (ib. 111, 123, 125); and they appear later in the armies of Agesilaus (Xen. Hell. 3.4, 23) and Archidamus (7.4, 22). The peltasts from Orchomenus which Xenophon mentions (Hell. 6.5, 15) were certainly mercenaries; so that we may infer that to a large degree the Spartans used mercenaries and allies for light-armed.

Though the tyrants (e. g. Pisistratus, Hdt. 1.61) and some Peloponnesian states (e. g. Corinth) had previously a large mercenary force (Thuc. 1.60), the first appearance of a μισθοφόρος ὄχλος ξενικὸς of the Spartans was in Acarnania in 426 B.C. (Thuc. 3.109). Brasidas used mercenaries frequently (Thuc. 4.80; 5.6); also Lysander. (Xen. Hell. 2.4, 30), Thibron (who wanted 300 mercenaries from Athens, ib. 3.1, 4), and Agesilaus (ib. 6.5, 15). The expedition of Cyrus gave a great impulse to the development of a mercenary army. In fact, about the time of Agesilaus they were so numerous that Xenophon says the citizens were only used for garrison duty, while mercenaries were employed for war (ib. 4.4, 14); and further (7.1, 23) that Arcadian mercenaries (they were the great people for supplying this kind of force: cf. Thuc. 7.57; Hesych. sub voce Ἀρκάδος μιμούμενος) were always used for an invasion of Attica. The Carians, too, appear frequently as mercenaries. The Spartans generally contented themselves with sending out thirty of their own body to take the chief posts of command in the army (cf. Bauer, op. cit. p. 261). The state used to look for a general of known ability, and it was he who used to agree with the soldiers, it was to him that the latter used to look for pay and provisions, and so practically they became his soldiers--they were no longer Spartans or Arcadians or Achaeans, but Brasideans or Cyreans. In later times there was a regular market of mercenaries at Taenarum (Diod. 17.108). The pay was generally about a drachma a day (Thuc. 7.27), to the λοχαγὸς twice as much, and to the στρατηγὸς four times as much (cf. Xen. Anab. 4.6, 1; and Gilbert, op. cit. 2.354). The divisions of the mercenary force were not those of the φρουρά. The divisions were much smaller, the τάξις being the highest unit for both cavalry and infantry. The τάξις of cavalry was commanded by the hipparch, and that of the infantry by the taxiarch. The τάξις fell [p. 1.773]into λόχοι, commanded by λοχαγυί, with substitutes ὑπολοχαγοί (Xen. Hell. iv. i, 26; 5.2, 13). This breaking up of the phalanx into smaller tactical units which were capable of being used independently is the chief military feature of the mercenary forces.

When the φρουρὰ was declared, the Spartans generally required, beside the native citizen force, a definite proportion of allies, καὶ τῶν συμμάχων τὸ μέρος (Xen. Hell. 4.6, 3), which was very frequently two-thirds of the available force (Thuc. 2.10). The number of allies compared with that of the Spartans was very great: in one case the proportion was 10,000 to 1500 (Thuc. 1.107). Later, when the system of mercenaries was fully in vogue, the allied forces could be commuted for a money payment. A hoplite was considered equal to two lightarmed, and a horse-soldier to four hoplites--commutation for a hoplite being three Aeginetan oboli (Xen. Hell. 5.2, 21; Diod. 15.31). The forces of the allies, besides their own commanders (ἀπὸ τῶν πόλεων στρατηγοί, Thuc. 2.10; Xen. Rep. Lac. 13.4), had also assigned to them special commanders, who were Spartans. These commanders were called ξεναγοί (Thuc. 2.75; Xen. Hell. 3.5, 7); and they were sometimes used in times of peace as governors in the allied states (ib. 5.2, 7).

The arms of the phalanx consisted of the long spear and a short sword (ξυήλη). The chief part of the defensive armour was the large brazen shield,--generally marked with the letter A, signifying Λακεδαιμονίων (Eupolis, fr. 359, Kock; Photius, s. v. λάμβδα), just as the Sicyonians had Σ on their shields (Xen. Hell. 4.4, 10),--which covered the body from the shoulder to the knee (Tyrtaeus, fr. 2.23), suspended, as in ancient times, by a thong round the neck, and managed by a simple handle or ring (πόρπαξ). The improved Carian handle (ὀχάνη) was not introduced till the time of Cleomenes III. Besides this, they had the ordinary armour of the hoplite [ARMA]. The helmets were made of felt and called πῖλοι (Thuc. 4.34, and Arnold's note: though Poppo takes these to be cuirasses). The heavy-armed soldiers wore a scarlet uniform (Xen. Rep. Lac. 11.3; Ages. 2.7). A day's march in the Anabasis was about 15 miles. Droysen (op. cit. p. 83) has collected a variety of statistics of the time taken to effect certain marches in Greek history ; but we cannot be sure of the distance, owing to our ignorance for the most part of the courses of the Greek roads between the places. In marching through a friendly country a market was usually granted to the soldiers by the inhabitants (Thuc. 6.44, 8.95; Xen. Hell. 3.4, 11): in an enemy's country they took what they wanted, getting vegetables and fuel are frequently mentioned, λαχανισμὸς καὶ φρυγανισμός (Thuc. 3.111, 7.4), or else made formal requisitions, προνομαί (Xen. An. 5.2, 24). In transmarine expeditions the Greek armies often took provisions with them (Thuc. 6.22, 44): though we hear of ἔμποροι and an ἀγοραῖος ὄχλος following the army (Thuc. 1. c.; Xen. Hell. 1.6, 37; 6.2, 23), so that no doubt a large part of the provisions were purchased; so, too, we may be certain that the ἑξαμήνου σῖτος mentioned in Xen. Hell. 3.4, 2, and the τροφὴ in Thuc. 6.93, 94, were money. The baggage and camp-followers came after the line of march, except in the rare occasions of night-marches, when they, as being the slowest, went in front (Xen. An. 7.3, 37; Cyr. 5.3, 37). We saw that the citizen-soldiers had each a servant, but the mercenaries did not keep such. Besides these servants, there was a large mass of impedimenta, such as the traders, booty, cattle, soldiers' mistresses, &c. (Xen. An. 3.5, 9; 4.1, 14), which Xenophon tells us were a serious shackle on the army (ib. 3.2, 27). Among the Spartans the baggage was all looked after by the ἄρχοντες τῶν σκευοφόρων (Xen. Hell. 3.4, 22; Rep. Lac. 13.4). The Spartan encampments were circular. Only the heavy-armed were stationed within it, the cavalry being placed to look out, and the helots being kept as much as possible outside. As another precaution against the latter, every soldier was obliged always to carry his spear about with him (Xen. Rep. Lac. 12.4). Though strict discipline was, of course, kept up in the camp, it was less rigorous than in the city itself (Plut. Lyc. 22; comp. Hdt. 7.208). Preparatory to a battle the Spartan soldier dressed his hair and crowned himself as others would do for a feast. The Spartans at Corinth, when but a stadium from the enemy, offered the customary sacrifice to Artemis Agrotera (Xen. Hell. 4.2, 20). The signal for attack in ancient times was given by priests of Ares (πυρφόροι), who threw lighted torches into the interval between the two armies (Schol. ad Eur. Phoen. 1186). Afterwards it was given not by the trumpet, but by the music of flutes, and sometimes also of the lyre and cithara, to which the men sang the battle song (παιὰν ἐμβατήριος), such as Tyrtaeus wrote (Paus. 3.17.5; Plut. l.c.), or raised a wild shout (ἀλαλά, ἐλελεῦ). The object of the music was not so much to inspirit the men as simply to regulate the march of the phalanx (Thuc. 5.70). This rhythmical regularity of movement was a point to which the Spartans attached great importance. The Greeks had not standards like the Roman signa. The σημεῖα which we hear of being raised and lowered (Thuc. 1.63) were signals rather than standards. Signals were given by fire-beacons also (φρυκτοί, Thuc. 3.22, and Arnold's note). Trumpets were little used except to sound the attack (τὸ πολεμικόν) or retreat (τὸ ἀνακλητικόν). (Cf. Hermann-Droysen, 1.54.) To prevent the ranks being broken the soldiers were forbidden to stop in order to strip a slain enemy while the fight lasted, or to pursue a routed enemy. The younger hoplites or the cavalry or lightarmed troops were despatched for this purpose (Xen. Hell. 4.4, § 16; 5.14.16). All the booty collected had to be handed over to the laphyropolae and ephors, by whom it was sold. The loss in battle was always much greater on the side of the vanquished than on that of the victors: for the wounded of the former could not be carried away, the beaten side not having any start as they have in modern battles; and so either were killed by the victors or bled to death. As average instances may be taken the battles of Delium and Corinth: the defeated party at each of these lost 1 in 7 men, the victors at the former 1 in 26 and at the latter 1 in 23. The loss of the Spartans at Leuctra (1 in 2) was the greatest we know of [p. 1.774]in an important battle. (Cf. Hermann-Droysen, 1.101.)

The rigid inflexibility of the Spartan tactics rendered them indisposed to the attack of fortified places (Thuc. 1.102). At the battle of Plataeae, they even assigned to the Athenians the task of storming the palisade formed by the γέρρα of the Persians (Hdt. 9.70).

In Athens, the military system was in its leading principles the same as among the Spartans, though differing in detail, and carried out with less exactness; inasmuch as, when Athens became powerful, greater attention was paid to the navy. Of the times before Solon, we have but little information. We learn that there were twelve phratriae, and in each of these four naucrariae, each of which was bound to furnish two horsemen and one ship (Photius, Suidas, s. v. ναυκραρία: Poll. 8.108). Of the four classes into which the citizens were arranged by the constitution of Solon, the citizens of the first and second served as cavalry, or as commanders of the infantry (still it need not be assumed that the ἱππεῖς never served as heavyarmed infantry), those of the third class (ζευγῖται) formed the heavy-armed infantry. The Thetes served either as light-armed troops on land, or on board the ships. The same general principles remained when the constitution was remodelled by Cleisthenes. All citizens qualified to serve either as horsemen, or in the ranks of the heavy-armed infantry, were enrolled in a list [CATALOGUS]. The case of Thetes serving as heavy-armed soldiers is spoken of as an exception to the general rule; and even when it was the case, they were not enrolled in the catalogus (Thuc. 6.43). Every citizen was liable to service from his eighteenth to his sixtieth year. On reaching their eighteenth year, the young citizens were formally enrolled (εἰς τὸ ληξιαρχικὸν γραμματεῖον), and received a shield and spear in a public assembly of the people, binding themselves by oath to perform rightly the duties of a citizen and a soldier (Aristot. ap. Harpocr. p. 241). The actual words of the oath are given in Poll. 8.105, 106 (cf. Hermann-Droysen, op. cit. 1.57). During the first two years, they were only liable to service in Attica itself, chiefly as garrison soldiers in the different fortresses in the country. During this period, they were called περίπολοι, and their commander περιπόλαρχος. (Harpocr. s. v. περίπολος: Pollux, 8.105; Lycurg. Leocr. § 76; Thuc. 8.92.) [EPHEBUS] The levies were made under the direction of the generals [STRATEGI]. The soldiers were selected either according to age, as among the Spartans (Aristot. ap. Harpocr. s. v. στρατεία and Phot. s. v. στρατία: ὅταν ἡλικίαν ἐκπέμπωσι, προσγράφουσι ἀπὸ τίνος ἄρχοντος ἐπωνύμου μέχρι τίνος δεῖ στρατεύεσθαι; the archons being, of course, those in whose year of office they had entered the military service), the expeditions being then called ἔξοδοι ἐν τοῖς ἐπωνύμοις: or else according to a certain rotation (Aeschin. Fals. Leg. p. 330.168, τὰς ἐκ διαδοχῆς ἐξόδους). Another kind of levy was that called ἐν τοῖς μέρεσι. It appears to have followed the former; each being used alternately (ἐκ διαδοχῆς, Aeschin. Fals. Leg. § 168). This expression ἐν τοῖς μέρεσι is variously interpreted. Some (e. g. Droysen, Gilbert, &c.) consider that the levy ἐν τοῖς ἐπωνύμοις called out all those of the specified years, while that ἐν τοῖς μέρεσι only called out a portion of these in some specified order of rotation (cf. Dem. Ol. 2.31., πάντας ἐξιέναι κατὰ μέρος ἕως ἂν ἅπαντες στρατεύσησθε). Others again (Bauer, Rustow and Köchly) suppose that ἐν τοῖς μέρεσι refers, not to the levy by years but by tribes or portions of tribes. (There is a full discussion on the subject by Lange, Leipziger Studien, 1.160; cf. Gilbert, op. cit. 1.302.) Both these kinds of levies were, however, only partial ones: when the levy was universal, it was said to be πανδημεί. The levying-place was generally the Lyceum (Schol. on Aristoph. Peace 356), though occasionally in other places (Andoc. de Myst. § 45). Each soldier was expected to be provided with provisions for three days (Aristoph. Ach. 197). The services of those below or above the ordinary military age, were only called for on emergencies, or for guarding the walls. (Cf. Thuc. 1.105, 2.13.) Members of the senate during the period of their office, farmers of the revenue, choreutae at the Dionysia during the festival; in later times, traders by sea also, were exempted from military service (Lycurg. Leocr. § 34; Demosth. c. Neaer. p. 1353.27; c. Mid. p. 516.15; Aristoph. Eccl. 1027, with the Schol.). Any one bound to serve who attempted to avoid doing so, could be accused on a γραφὴ ἀστρατείας. This, with the γραφὴ λιποταξίου and the γραφὴ δειλίας, were the chief indictments to which the soldier was liable [ASTRATEIAS GRAPHÉ]. The resident aliens commonly served as heavy-armed soldiers, especially for the purpose of garrisoning the city, but only in case of an universal levy (πανδημεί). They were prohibited from serving as cavalry (Xen. de Vect. 2.5; Hipparch. 9.6). Slaves were only employed as soldiers in cases of great necessity, as at Marathon (according to Paus. 1.32.3） and Arginusae (Xen. Hell. 1.6, § 17). The levy of the cleruchs and allies appears to have been made ἐκ καταλόγου (Thuc. 6.26); and recruiting was conducted by Athenian officers when the order was given (στρατιὰν ἐπαγγέλλειν ἐς τοῦς συμμάχους, Thuc. 7.17). The Athenians often used the allies solely for their own special wars (cf. Bauer, op. cit. p. 290; Grote, 5.199); though when we hear of Lemnians and Imbrians and other peoples who had large numbers of cleruchs serving, such are generally to be understood as referring to the cleruchs.

The dress and armour of the hoplites consisted of a white jerkin (χλανίς) which reached down to the hips. Over this was the θώραξ [LORICA]; and over this again they wore a cloak which in the case of officers was purple (Aristoph. Peace 1175). Of course in battle they took this off. On their legs were greaves [OCREAE], and on their head a plumed helmet [GALEA]. Their arms were a round or oval shield [CLIPEUS], short sword [GLADIUS], and lance [HASTA]. These arms the soldier appears to have provided himself, but the orphans of those who fell in battle, on arriving at man's estate, received a πανοπλία from the state (Aeschin. Ctes. § 154).

Of the details of the Athenian military organisation, we have no distinct accounts as we have of those of Sparta. There was less exactness than among the Spartans. Especially the [p. 1.775]Athenians were not so particular that the same individuals should always hold the same positions in the phalanx or its subdivisions (Xen. Rep. Lac. 11.6; Thuc. 5.66); and more frequent exemption from service was allowed than among the Spartans (cf. Bauer, op. cit. p. 270). The heavy-armed troops, as was the universal practice in Greece, nearly always fought in phalanx order, very rarely in a square (Thuc. 6.67). They were arranged in bodies in a manner dependent on the political divisions of the citizens. The soldiers of each tribe formed a separate body in the army, also called a tribe, and these bodies stood in some preconcerted order (Hdt. 6.11 ; Plut. Arist. 5; Xen. Hell. 4.2, § 19, with Schneider's notes). Each φυλὴ appears to have formed a τάξις (Lys. Agorat. § 82, cf. § 79), and the members of each dome probably stood together (Isaeus, de Pyrrhi hered. § 42; Lys. Polystr. § 23). A further subdivision was that into λόχοι under λοχαγοί (Xen. Hell. 1.2, 3; Mem. 3.4, 1; Isocr. 15.117). The strength of the λόχος, however, cannot be determined; and it must be remembered that λόχος is the most ordinary term for the largest tactical unit, which varied, though it was generally about 100 men (Bauer, op. cit. p. 248). The συσσίτια or συσκήνια (Lys. Agorat. § 79; Isaeus, Nicostr. § 18) are voluntary unions of friends, and we cannot infer from them that the members belonged to the same deme or even to the same branch of the service. Before Potidaea the hoplite Socrates of Alopece was a σύσσιτος of the knight Alcibiades of Scambonidae (Plat. Symp. 219 E). Every hoplite was accompanied by an attendant (ὑπηρέτης, Thuc. 3.17), like the Helots among the Spartans, to take charge of his baggage, and carry his shield on .a march. Each horseman also had a servant, called ἱπποκόμος, to attend to his horse (Xen. Hell. 2.4, 6).

It would appear that in 490 B.C. there was no cavalry (Hdt. 6.112). In the time of Cimon it was 300, soon after 600 (Andoc. de Pace, § 5; Schol. on Aristoph. Kn. 627), and at the beginning of the Peloponnesian war 1200, of whom 200 were hired Scythian bowmen (Thuc. 2.13; Aristoph. Kn. 225). The number 1000 remained down to the time of Demosthenes (Xen. Hipp. 9.3; Dem. de Symm. p. 181.13). It is probable that the organisation of cavalry was due to the experience the Athenians obtained in the Persian wars with regard to that service. The cavalry was commanded by two Hipparchs, part of whose duty it was to see that the number was kept up to the 1000 (Xen. Hipp. 1.2). The ἱπποτροφία was made one of the liturgies of the rich (Xen. Oec. 2.6; Lycurg. Leocr. § 139), the performance of which could be enforced by law (Xen. Hipp. 1.9). It could only attach to the rich, for an ordinary horse cost as much as three minae (Isaeus, de Dicaeog. hered. § 43), and a war-horse would cost a great deal more. Of course those only were required to serve who were physically capable of doing so efficiently; but such were required to serve as long as they were able (Xen. Hipp. 1.2): so we find the corps divided into νέοι and πρεσβύτεροι (ib. 1.17; 2.3). It belonged to the duties of the hipparch to train the horsemen (Xen. Mem. 3.3, 5), especially in being able to leap up on horseback (Xen. Hipp. 1.1. 7)--for they had no saddles or stirrups--in throwing the javelin (ib. 1.6, 21), and in kinds of sham fights (ἀνθιππασία), in which each hipparch commanded five tribes (1.20, 3.11; Aristot. fr. 56 M.); for the cavalry were divided by tribes (Xen. Hell. 2.4, 31), each governed by a phylarch (Xen. Hipp. 1.8). The hipparch appears to have had to examine the horses (Ken. Mem. 3.3, 4), which must have been very necessary when one remembers the various requirements of a warhorse as given by Xenophon (de re Equestri, 3.7, 8.1). The hipparch and the phylarch were elected by the people (Aristoph. Birds 798). At the beginning of the year the βουλὴ held a review of the knights. They went through exercises (ἀκοντισμός, ἀνθιππασία, &c.) which are described at length by Xenophon (Hipp. iii.); and if this examination (δοκιμασία） proved satisfactory and the horses seemed suitable (1.13), a fee for equipment (κατάστασις) was paid to each horseman (Suid. s. v.). Besides this, pay (μισθός) was given to the cavalry to the amount of 40 talents a year, i. e. 240 drachmas for each horseman (cf. Dem. Phil. 1.28, p. 48 R.; Boeckh, Staatshaushaltung, i.3 317 ff., 340 ff.). A knight regularly appointed could not serve as a hoplite ; and if he did not pass the examination and yet attempted to serve, he was disgraced (Lys. Alc. 2.7; 1.8, 9). The arms of the cavalry were two spears of cornelwood (Xen. de re Eq. 12.1. 2, 13) and a small sword like a cleaver. Full account of the armour is to be found in Xen. de re Eq. 12.1 ff. The cavalry appear to have been an imposing corps, and so were much used in processions (Aristoph. Frogs 652; Dem. Phil. i. p. 47.26; Xen. Hipp. 3.2). Besides the light-armed soldiers drawn from the ranks of the poorer citizens, there was at Athens a regiment of Thracian slaves, armed with bows. The number of these increased from 300, who were purchased after the battle of Salamis, to 1000 or 1200 (AESCHIN. FALS. LEG. § § 173, 174; Boeckh, Staatshaushaltung, book 2.100.11). They, however, were generally employed as a sort of police or city guard. Besides these, however, the Athenians had a troop of bowmen of their own citizens, amounting, at the beginning of the Peloponnesian war, to 1600 (Thuc. 2.13). They were commanded by τόξαρχοι (C. I. A. 1.79). The two classes of ἀστικοὶ and ξενικοὶ τοξόται are mentioned in C. I. A. 1.433. Besides the 1000 knights, there were at the beginning of the Peloponnesian war 200 ἱπποτοξόται, mostly taken from the Scythians (Thuc. 2.13, compared with Aristoph. Kn. 225). It was considered degrading for a citizen to serve in this corps (Lys. Alc. 2.6). They used to ride before the hipparchs (Xen. Mem. 3.3, 1). There is an elaborate account of the military forces of Athens in Boeckh (op. cit. i.3 321-339).

For the command of the army, there were chosen every year ten generals [STRATEGI] and ten taxiarchs [TAXIARCHI]. Respecting the military functions of the ἄρχων πολέμαρχος, see the article ARCHON The number of strategi sent with an army was not uniform. Three was a common number. Sometimes one was invested with the supreme command; at other times, they either took the command in turn (as at Marathon), or conducted their operations by common consent (as in the Sicilian expedition). [p. 1.776]

The practice of paying the troops when upon service was introduced by Pericles (Ulpian. ad Demosth. περὶ συντάξ. p. 50 a). The pay consisted partly of wages (μισθός), partly of provisions, or, more commonly, provision-money (σιτηρέσιον). The ordinary μισθὸς of a hoplite was two obols a day. The σιτηρέσιον amounted to two obols more (Dem. Phil. 1.28, p. 48). Hence the life of a soldier was called, proverbially, τετρωβόλου βίος (Eustath. ad Od. p. 1405; ad Il. p. 951). Higher pay, however, was sometimes given, as at the siege of Potidaea the soldiers received two drachmas apiece, one for themselves, the other for their attendants. This, doubtless, included the provisionmoney (Thuc. 3.17). Officers received twice as much; horsemen, three times; generals, four times as much (comp. Xen. Anab. 7.6, § 1; 3.9).

As regards the military strength of the Athenians, we find 10,000 heavy-armed soldiers at Marathon (Just. 2.9); 8,000 heavy-armed, and as many light-armed, at Plataeae (Hdt. 9.29); and at the beginning of the Peloponnesian war there were 13,000 heavy-armed ready for foreign service, and 16,000 consisting of those beyond the limits of the ordinary military age and of the metoeci, for garrison service (Thuc. 2.13).

It was the natural result of the national character of the Athenians and their democratical constitution, that military discipline was much less stringent among them than among the Spartans (χαλεπαὶ γὰρ αἱ ὑμέτεραι φύσεις ἄρξαι, Thuc. 7.14), and after defeat especially it was often found extremely difficult to maintain it. The generals had some power of punishing military offences on the spot, such as treasonable correspondence with the enemy (Lys. Agorat. § 67); but the punishment for insubordination was trifling (Lys. Sim. § 43). For the greater number of offences, the trial only took place after the return of the army home before the generals (Lys. Alc. 1.21), or the taxiarchs as their substitutes (Dem. in Booet. de nom. § 17), and a jury of persons who had served in the army (Lys. Alc. 1.15); the punishments were various kinds of ἀτιμία (Andoc. de Myst. § 74). Various rewards also were held out for those who especially distinguished themselves for their courage or conduct, in the shape of chaplets, statues, &c. (cf. Aeschin. Fals. Leg. § 169). In connexion with these, the λόγος ἐπιτάφιος, spoken over those who had fallen in war, and the τάφος δημόσιος, must not be omitted. Respecting the provision made for those who were disabled in war, see the article ADYNATI

The Peltastae (πελτασταί), so called from the shield which they bore [PELTA], were a kind of troops of which we hear very little before the end of the Peloponnesian war. The first time we have any mention of them is in Thuc. 4.111, where they are spoken of as being in the army of Brasidas. With the more frequent employment of mercenary troops a greater degree of attention was bestowed upon the peltastae; and the Athenian general Iphicrates introduced some important improvements in the mode of arming them, combining as far as possible the peculiar advantages of heavy-(ὁπλῖται) and light-armed (ψιλοί) troops. Thus distant as well as close combat was possible for them (Schol. on Eur. Rhes. 311; cf. Plut. Legg. 7.813 E). The great advantage of this light infantry is well sketched by Xenophon (Xenoph. Hell. 2.4, 12). Iphicrates substituted a linen corslet for the coat of mail worn by the hoplites, and lessened the shield, while he doubled the length of the spear and sword. He even took the pains to introduce for them an improved sort of shoe, called after him Ἰφικρατίδες (Pollux, 7.89). This equipment was commonly adopted by mercenary troops, and proved very effective. The almost total destruction of a mora of Lacedaemonian heavy-armed troops by a body of peltastae under the command of Iphicrates was an exploit that became very famous (Xen. Hell. 4.5, § 11). The peltast style of arming was general among the Achaeans until Philopoemen again introduced heavy armour (Pluto. Philop. 9) with the Macedonian phalanx.

When the use of mercenary troops became general, Athenian citizens seldom served except as volunteers, and then in but small numbers. Thus we find 10,000 mercenaries sent to Olynthus with only 400 Athenians (Demosth. Fals, Leg. p. 425.263). With 15,000 mercenaries sent against Philip to Chaeroneia, there were 2,000 citizens (Demosth. de Cor. p. 306.237). It became not uncommon also for those bound to serve in the cavalry to commute their services for those of horsemen hired in their stead, and the duties of the ἱπποτροφία were ill executed. The employment of mercenaries also led in other respects to considerable alterations in the military system of Greece. War came to be studied as an art, and Greek generals, rising above the old simple rules of warfare, became tacticians. The old method of arranging the troops, a method still retained by Agesilaus at the battle of Coronea, was to draw up the opposing armies in two parallel lines of greater or less depth, according to the strength of the forces, the engagement commencing usually very nearly at the same moment in all parts of the line. The tactics employed by Epaminondas were a further development of the Boeotian practice of deepening one point in the attacking line. Thus, while the Greek army was generally drawn up eight deep, at Delium the right Boeotian wing was twenty-five deep (Thuc. 4.93). Epaminondas at Leuctra not only deepened the left wing to fifty, placing the Sacred Band in front; but he further (and this was a distinct innovation) moved it forward obliquely, λοξὴ φάλαγξ (Xen. Hell. 6.4, 12; Diod. 15.55; Plut. Pel. 23), so as to attack with this tremendous force the centre, and not as usual the wings, of the enemy.

Thessaly under Jason of Pherae had a vast army: 8,000 cavalry and 20,000 hoplites, a large force of peltasts, and 6,000 mercenaries (Xen. Hell. 6.1, 5). This Jason has the distinction of having invented an arrangement of cavalry in the shape of a rhomboid (Asclep. 7.2;. Ael. 16.3), but a still greater one in having been the first to take measures for tending soldiers when sick (Xen. Hell. vi. 1, 6). The tyrants in Sicily had very large armies, though probably the actual numbers are exaggerations of their conquerors. Gelon is said to have had 10,000 mercenaries; and Dionysius I. an army of 120,000 foot and 12,000 horse, together with every kind [p. 1.777]of improvement in engines, arms, &c. He paid great attention to cavalry, and had mercenaries from all quarters, Iberians and Gauls (Diod. 15.70). For a full account of this great military organiser, see Diod. 14.42, 43; Grote, 10.240; Hermann-Droysen, op. cit. 76; Bauer, op. cit. § § 53, 54. The chief features in the Sicilians were their great enterprise and their readiness to adopt all improvements as well in arms as in tactics: thus, for example, Dion divided his army into λόχοι ὄρθιοι in accordance with the important improvement made by Xenophon (Plut. Dio 45).

Under Philip and the subsequent Macedonian kings, the army was much more systematically attended to, and we have much fuller accounts of it than we have of the armies of free Hellas. Till Philip's time the Macedonians had a very indifferent army. Archelaus indeed did something to improve it; but the exact nature of his improvements we cannot be very certain about (Thuc. 2.100). Philip was the first to organise the foot-soldiers in a phalanx, and to develop still further the resources of the Macedonian cavalry, which had always been excellent. In times past it was the cavalry who used to protect the country, while the inhabitants fled into the fortified towns (Thuc. 1. c.); but it is very rarely that we read of foot-soldiers, such as the hoplites of the Lyncestae and the infantry which Perdiccas supplied to Brasidas (Thuc. 4.124, 83).

During his stay at Thebes, Philip learned the tactics of Epaminondas, though the Macedonians must have learned already much of the Greek method of warfare of that day from the troops of the Athenians and Spartans which were certainly carrying on war in Chalcidice, and from the number of Greek mercenaries which were now everywhere abroad. With special knowledge derived from these sources, and with a population accustomed to monarchical rule, whom he could call out for service and use according to his own judgment for as long as he pleased (cf. Dem. de Cor. 305.235), Philip organised the Macedonian army. The cavalry were the principal element; they were called his Companions or ἕταιροι, a term applied in Macedonia to persons of distinction (Plut. Pel. 27). Those summoned for service in the infantry were called Foot-Companions (πεζέταιροι), a term which shows that they were a kind of subsequent appendage to the cavalry. These two terms, ἕταιροι and πεζέταιροι, appear to have been originally applied to a picked body-guard (Dem. Ol. ii. p. 23.17), and gradually to have been extended to the cavalry and infantry generally (cf. Grote, 11.386). The ἕταιροι, 800 in number (Theopomp. fr. 249), were mostly Macedonians, but there were some Greeks and Thessalians amongst them. It has been supposed that there were six ἴλαι of Companions and six τάξεις of Foot-Companions, one raised from each of the six recruiting districts of Macedonia (Grote, l.c.). But though the divisions of the cavalry and infantry were called ἴλαι and τάξεις respectively, and there were recruiting districts (cf. the analogous division of Thessaly into four recruiting districts after its conquest, Dem. Phil. iii. p. 117.26), we cannot be quite certain that there were only six (see Hermann-Droysen, op. cit. 108). Thirteen ἴλαι are mentioned in Plut. Alex. 16, in such a way as to imply that there were more. A picked body of the cavalry was called the ἄγημα or βασιλικὴ ἴλη. Further divisions of the τάξεις were λόχοι and perhaps σκηναί, the latter consisting of ten men (Anaximenes ap. Harpocr. s. v. πεζέταιροι: Arrian. Anab. 4.21, 10). If Anaximenes in this passage refers the establishment of the πεζέταιροι eo nomine to Alexander, this does not refer, as is thought, to Alexander son of Amyntas, but is probably only due to the practice of ascribing all innovations to and beginning all things from the great Alexander, though Abel (Makedonien vor König Philipp, p. 131, note 1) supposes him to refer to the enrolment of Persians in the army by Alexander the Great. The infantry was arranged in a phalanx (Diod. 16.3). Their armour consisted of a helmet, metal-plated leather jerkin, greaves, and small circular shield, about two feet in diameter (Asclep. Tact. 5), fastened on the arm, not grasped by the hand: for both hands had to be used to hold the sarisa [SARISA], twelve cubits long (Theophr. Hist. Pl. 3.12, 2). They had also a short sword used for thrusting; and accordingly the long sword of the Roman horsemen caused the greater terror (Liv. 31.34, 4). Over the armour they wore a cloak except in battle (Polyaen. 4.3, 3). Their arrangement was that of a serried mass, a feature due to the influence of Epaminondas; for it was doubtless with a view of facing his method of warfare that Philip first organised the phalanx. Like the Greek hoplites, the phalanx was arranged in ζυγὰ and στίχοι (cf. p. 770b). The armour of the cavalry did not at all differ from the armour of the ordinary heavy-armed horsemen in the Grecian armies. The lance we was the ξυστόν, made of cornel-wood (Arr. An. 1.15, 5). The riders had no stirrups or saddles (cf. Hermann-Droysen, op. cit. 109). A very large portion besides of Philip's army consisted of light-armed troops and mercenaries (Dem. Phil. 3.123.49). Philip was most particular about drill (Polyaen. 4.2, 10); he made the soldiers carry their own provisions, thus getting rid of a great number of waggons; and he allowed no more than one servant for each horse-soldier and one for each ten foot-soldiers to carry such articles as the hand-mills, &c. (Frontin. 4.1, 6). He supplied the arms to the soldiers (Diod. 16.3; Liv. 42.52), arranged the different corps and appointed all their commanders himself (cf. Arr. An. 3.5, 5). His especial body-guard were called σωματοφύλακες, and were composed of his closest friends (Diod. 16.93; cf. Arr. An. 6.28, 4). A further body closely connected with the king were the Royal Pages (βασιλικοὶ παῖδες), sons of the most distinguished Macedonians, who received a military training at the court. On the cavalry Philip bestowed great care: with the united Thessalians he had 3,000 in his war against the Phocians (Diod. 16.35). He took the greatest care in improving the breed of horses (Just. 9.2, 16). Even the foot-soldiers were trained to fight on horseback (cf. Arr. An. 4.23, 2); and a peculiar wedge-shaped arrangement of cavalry (ἐμβολοειδὴς τάξις) is attributed to Philip (Arr. Tact. 16.6). For the Macedonian ἐξελιγμός, see p. 771 a. Philip, too, made the greatest improvement in [p. 1.778]military engines [TORMENTA], such as catapults and besieging-towers: from the sieges of Perinthus and Byzantium a new era dates in the art of besieging towns (Athen. de Mech. p. 10; Grote, 11.262); though we must remember that Dionysius of Syracuse had considerable artillery before Philip's time (Diod. 14.42, 48, 50).

One of the chief introductions of Alexander was that of the ὑπασπισταί. They held an intermediate position between the heavy-armed Foot-Companions and the quite light-armed. They carried a round shield, a short thrustinglance, and were clad with a chiton [CHITON], and wore on their head a causia [CAUSIA]. Some time during the campaigns of Alexander, probably after the battle of Arbela, the ὑπασπισταὶ were divided into chiliarchies (cf. Arr. An. 3.29, 7), according to Persian custom (cf. Xen. Cyr. 4.1, 4; Diod. 18.48). Alexander also divided the cavalry, no longer into ἴλαι in the first instance, but into ἱππαρχίαι (at first eight, later four, in addition to the ἄγημα), which were then divided each probably into two ἴλαι and several ἑκατοστύες (Arr. An. 6.27, 6): a previous alteration tending towards this arrangement being the dividing of the ἴλη into two λόχοι, this latter title not having been applied to a division of cavalry previously (Arr. An. 3.16, 3). Prior to the appointment of these ἱππαρχίαι, the chief commander of the cavalry was called ἱππάρχης (ib. 3.27, 4). Besides the heavy cavalry, Alexander also introduced a lighter kind of cavalry, called Sarisophori, used, as Grote (11.389) says, like Cossacks for advanced posts or scouring the country. Their sarisa was shorter than that of the phalangites, but much longer than the ξυστὸν of the Companions. The Paeonian cavalry were something similar. We hear of a body of reconnoitring πρόδρομοι (i. e. Paeonians) and Sarisophori in the march to the Granicus (Arr. An. 1.12, 7). The Greeks seem to have used but little such reconnoitring parties. The phalanx of Alexander was much less rigid than is usually supposed: according to circumstances, it could be divided into various independently acting corps. We find the infantry sometimes running to the attack (Arr. An. 1.26, 3),: and a glance at the plan of the battle of Arbela in either Rustow and Köchly's Griechisches Kriegswesen, or in Bauer, op. cit. Taf. xi., will show the phalanx broken up. However, the normal condition of the phalanx was in close array (πύκνωσις κατ᾽ ἐπιστάτην καὶ παραστάτην, Pol. 18.30). The breaking--up into separate divisions is mentioned as an exception; it appears to have been often done to allow the chariots of the enemy to rush through (Arr. An. 1.1, 8; 3.13, 6). The phalanx was usually eight deep, though at the Granicus Alexander made it sixteen (Arr. An. 1.13, 1). The front of the individual λόχοι was greater than the depth, because λόχοι ὄρθιοι are early mentioned as quite unusual (ib. 4.25, 2). When standing in close array, each phalangite stood three feet from the soldier at his side and two feet from the soldier behind him. Given the sarisa to be even eighteen feet long, and six of these to be used in grasping and balancing it (Plb. 18.12), the sarisa of the fifth rank would project two feet beyond the first rank. [It is very questionable, however, whether Polybius is in error in this passage in using the word πήχεις instead of πόδας all through, as Rüstow and Köchly (op. cit. 239), followed by Droysen (op. cit. 173), think. Hultsch, in his edition of Polybius, adheres to πήχεις, and see Grote, chap. xcii. Appendix, and 11.384. That the longest sarisa was eighteen feet in Alexander's time, and was lengthened afterwards to twenty-one feet, twenty-four feet being the ideal length aimed at, seems the most satisfactory way to explain the various passages.] The other ranks held the sarisa slanting upwards to ward off missiles, but took no other part in the fight except so far as their pressure urged the front ranks forward. On the march the phalangites carried their sarisas on their shoulders ; and bringing them down into position for action was called καταβάλλειν τὰς σαρίσας (Pol. 18.24, 9; Weiss. on Liv. 33.8, 13). The first man of each file (στίχος) was the λοχαγός, a soldier of tried bravery and great strength. This method of fighting in phalanx, which was characteristic of the Hellenistic empires down to the end of their existences, is severely criticised by Polybius in various passages, chiefly on the score of its incapability of manoeuvring easily--a disadvantage attaching more to the phalanx of the successors of Alexander than to that of Alexander himself. The phalanx, we are told, required even ground (Pol. 18.31; Liv. 31.39, 10), was powerless against an attack from flank or rear (Pol. 18.26, 4; Liv. 44.41, 6), and in attempting to wheel round was generally thrown into confusion (Liv. 32.18. 16). All this is no doubt true; and the truth was all the more apparent when the phalanx had to face the Roman manipular tactics (Pol. 18.29, 30): besides, the phalanx of the successors of Alexander was used generally in the offensive. But we must remember that the phalanx was quite invincible as long as it remained together, and its attack in front irresistible (Pol. l.c.), and that it was not subject to any inconveniences greater than those experienced by the ordinary Greek hoplite army (cf. Grote, 11.385). The strength of the phalanx, according to the tacticians, was 16,384 men--16 men to one Lochus, 16 Lochi to one Syntagma, 2 Syntagmata to one Pentacosiarchy, 2 Pentacosiarchies to 1 Chiliarchy, 16 Chiliarchies to the phalanx (cf. Grote, 11.386). But this is merely theoretical. The actual numbers of the phalanx varied. It was 10,000 at Sellasia, 16,000 at Cynoscephalae, and 20,000 at Pydna (Droysen, pp. 158, 172). The phalanx of Antiochus at Magnesia consisted of 16,000 men, and was formed in 10 divisions (μέρη) of 1600 men each, arranged 50 broad and 32 deep (App. Syr. 32; Liv. 37.40, 1).

We cannot be very certain about the strength of Alexander's army at different times; but even at the outset the Macedonians were distinctly in the minority--12,000 out of 30,000 infantry, 1500 out of 4,500 cavalry (see Grote, 11.397). The Grecian allied forces stood under Macedonian commanders: the infantry is rarely mentioned, but we hear of the cavalry of the Thessalians and an ἴλη from Orchomenus in Boeotia. It is noticeable that mercenary cavalry appear in Egypt (Arr. An. 3.5, 1). We do not hear much of the light-armed troops in Alexander's army. The Agrianes, a people from [p. 1.779]Northern Macedonia, were valuable archers. There were, of course, slingers and javelin-throwers, later all kinds of native Asiatic troops, amongst them ἱππακοντίσται and ἱπποτοξόται (Arr. An. 3.24, 11; 4.24, 1). The introduction of these native troops after the conquest of the Persian empire led to changes in the army. The phalanx was now made 16 deep; the first three ranks being Macedonians, armed in Macedonian fashion with the sarisa, as was also the last, but the twelve intermediate ranks were Persians, armed with bows and javelins (Arr. An. 7.23, 3; cf. Grote, 12.73). This phalanx, says Bauer (op. cit. 315), is the first examaple of the union of differently-armed troops, light and heavy, into one tactical body. Droysen (Gesch. Alex. ii.2 232) thinks it was intended to be used against the Italians. The Macedonian cavalry was posted on the wings, and was generally used to bear the chief part in the battle after it had been opened by the light troops, the heavy infantry being seldom required in Alexander's battles. Alexander himself usually headed the cavalry on the right wing. It was arranged in ἴλαι, separated by intervals (Arr. An. 5.15, 2), which advanced to the attack one a little in advance of the other, so that on a plan the different ἴλαι would look like a flight of steps. (This is probably the meaning of ἐξελιγμὸς τῶν ἵππων in Arr. An. 3.15, 2.) It was quite exceptional if the cavalry were placed in a solid mass (ib. 5.17, 4). The extreme wings were generally formed by the Agrianes and other bowmen: next to the cavalry on the right wing, adjoining the heavy infantry, were the Hypaspistae (see Bauer, op. cit. p. 316). At Arbela Alexander made use of a considerable reserve (Arr. An. 3.12, 1). The excellence of his cavalry brought about an important alteration in the conclusion of battles; for he used to pursue the beaten enemy (cf. Arr. An. 3.15, 5), a practice not at all usual in Grecian warfare. Another feature in which the excellently-drilled Macedonian army differed from the Grecian was in being able to form in battle array a long distance from the enemy (7 miles at Arbela, Arr. An. 3.9, 3). The use of field artillery is a striking feature of Alexander's tactics. He used it specially to protect the army in crossing a river (Arr. An. 1.6, 8), but also in other cases (Arr. An. 4.29, 7). Heavy artillery was of course used in sieges. All the engines were administered by μηχανοποιοί (Arr. An. 2.26, 2). Alexander did not use elephants; though he appears to have intended to do so, as he had 200 brought from India (Arr. An. 6.2, 2). There was a very large mass of camp-followers in the army of Alexander, women and children being in it (Arr. An. 6.25, 5), besides the number of soldiers servants, merchants, and drivers of the waggons, which carried the tents, engines, and other baggage. The Macedonians do not appear to have fortified their camp; but each division of troops had definite positions in the encampment. We hear of Thracians used as pioneers (Arr. An. 1.26, 1). We know little about the way the commissariat was administered: and as regards pay, all we know is that the phalangite, who held a position midway between the common soldier and those who were said to get double pay (διμοιρῖται, got 10 staters a month (Arr. An. 7.23, 3).

The wars of the Hellenistic period show in many respects a continuation of the later system of Alexander. The armies become much larger, and as a necessary consequence the material much worse, and a still larger proportion than before were mercenaries. Demetrius is said to have had an army of 110,000 men taken from every kind of class and nation (Plut. Dem. 43). In Hermann-Droysen (op. cit. p. 133 ff.) there are extensive statistics of the strength of the armies of this period. But many of the features of the army of Alexander continue. The large introduction of native troops beside a few Macedonians, the use of cavalry for the first attack, the position of the cavalry and light-armed on the wings, and the phalanx of heavy infantry in the centre, the carrying on of campaigns during the whole year and the extensive use and improvement of artillery, are features we still find exhibited. Also special corps retain their names, such as the παῖδες βασιλικοί (Pol. 5.82; Liv. 45.6, 7), σωματοφύλακες (Diod. 19.12), ὄγημα (Pol. 5.25), ἴλη βασιλική (Pol. 5.84), ἕταιροι for the cavalry (Pol. 5.53). But, on the other hand, there are differences. The phalanx comes to be used in the attack, no longer in the defence, and it becomes far more rigid and immovable than it was with Alexander. The sarisa was lengthened to 21 feet. Philopoemen (Pol. 11.11.6) alone appears to have broken up the phalanx into smaller divisions. Pyrrhus (Pol. 18.28) placed Italians and phalangites alternately (ἐναλλάξ), but apparently without the whole mass being appreciably less serried. This rigid kind of phalanx remained the cardinal feature of the Hellenistic armies of the time: the best soldiers for it were considered to be the Macedonians and Achaeans (Pol. 4.8); on the wings were generally placed special forces (ἐπιτάγματα) to prevent flank attacks (Pol. 5.53). The various Grecian states who aspired to any distinction in war had to adopt it--Aetolians (Liv. 38.7), Achaeans and Megalopolitans (Pol. 2.65; Plut. Phil. 8), Spartans (Plut. Cleom. 11). The smallest division of the Hellenistic phalanx was the σημαία, which often appears in Egyptian papyri for bodies of fourteen men (Bauer, op. cit. p. 320). The σπεῖρα, which is the general term Polybius gives to the Roman cohort, was a larger division. There were no standards. Livy (33.7, 2） wrongly applies to the Macedonians Roman customs. The purple cloth φοινικίς) raised on a sarisa in Pol. 2.66.11, was only for a signal. Alexander had after the Indian expedition assigned to certain veterans of the Foot-Companions the privilege of carrying silver shields, and these veterans were called ἀργυράσπιδες (Just. 12.7, 5) [ARGYRASPIDES]. These are mentioned subsequently in the wars of Eumenes and Antigonus (Diod. 18.58), and in the Syrian kingdom (Pol. 5.79, 4). We also hear of phalangites who carried white shields (λευκάσπιδες, Plut. Cleom. 23) and brazen shields (χαλκάσπιδες, Pol. 2.66). This looks like the introduction of a kind of uniform (cf. the στρατιωτικαὶ χλαμύδες and φοινικοῖ ὑροδύται in Plut. Phil. 11: cf. Plut. Aemil. 18; Pol. 31.3). The breastplate of the phalangites appears to have been of great weight and power of resistance (Plut. Demetr. 21). The term ὑπασπισταὶ appears to be now used no longer for the troops [p. 1.780]midway between the heavy and light style of armour, but for the body-guard (Diod. 18.45; Pol. 5.27; Liv. 42.51), as indeed it was occasionally in Alexander's time (Arr. An. 3.8, 3). The 300 with Eumenes were the remains of Alexander's corps (Diod. 19.28). We also hear of a special body-guard called Nicatores (Liv. 43.19.11 ; Hesych. sub voce Νικατῆρες).

In the Hellenistic armies the light-armed and cavalry stood on the wings. Peltasts are found in the Greek armies, and the Greek mercenaries were mostly armed in this style. They are regularly distinguished from the archers, slingers, javelin-throwers, &c. (Pol. 8.15.6). They were divided into σπεῖραι and σημαῖαι (Pol. 5.4). These only very rarely acted in serried ranks: they generally fought in a scattered manner in small detachments (Liv. 35.29). The best archers were Cretans, who were excellent too at ambushes, raids, and such like irregular warfare (Pol. 4.8, 68), or Agrianes (Pol. 2.65, 79): the best slingers were Achaeans (Pol. 4.61); yet it is noticeable that in the armies of the Achaean league no Achaean slingers appear (Hermann-Droysen, 170, note 3). The cavalry was largely supplied by the Thessalians (Liv. 33.4; Pol. 18.22). It was armed with the heavy breastplate and thrusting-pike (ξυστόν), and carried a wooden shield covered with brass (Pol. 6.25). The Thessalian cavalry was unconquerable when fighting in line and phalanx, but difficult to be used for single combat out of line; the case being just the reverse with the Aetolian cavalry (Pol. 4.8, 18.22). The Aetolian cavalry was divided into οὐλαμοί (which was the smallest division, Pol. 18.19), ἴλαι, and ἱππαρχίαι. Cavalry was usually posted eight deep, with intervals between the individual ἴλαι (Pol. 12.18). The statement that in the battle of Gabiene the two ἴλαι were posted fifty deep (Diod. 19.27) is hardly to be believed; rather they were fifty strong (Rustow and Köchly, op. cit. 370, note 16). Special kinds of cavalry also used were : (1) the Median cavalry, with long lances (λογχοφόροι, Diod. 19.39), probably like the Sarisophori of Alexander; (2) Tarentini, lightarmed cavalry, using javelins and each having two horses (Pol. 4.77, 11.12, 16.18; Ael. Tact. 45; Liv. 35.28: cf. Pollux, 1.131, ἅμιπποι), such being originally mercenaries from Tarentum; (3) Cataphracti (Pol. 31.3; Liv. 37.40) [CATAPHRACTI]. We occasionally hear of ἱπποτοξόται (Diod. 20.113) and scythed chariots (Plut. Demetr. 28; Liv. 37.40). But the most special feature of the Hellenistic armies, especially that of the Seleucids, is the use of elephants. Alexander had no doubt intended to use them, but never did. Seleucus got no fewer than 500 from Sandracottus (Strab. p. 724). The elephants were generally placed before the line in the centre, though sometimes on the flanks, ἐν ἐπικαμπίῳ (Diod. 19.27-29), i. e. at an angle with the main line, stretching forward or backward according as they were to be used for attack or defence. Sometimes we find them in a square (Diod. 19.39). Between the elephants were numerous light-armed, and both used generally to open the battle. Pyrrhus, however, was accustomed to keep them for the final effort--they were the ultima ratio regis. The elephants were mostly Indian ones, and were driven by men called Indians (Diod. 19.84; Liv. 38.14), except in Egypt, where the African elephant was used (Pol. 5.79, 82). The latter was much inferior to the Indian elephant (ib. 84). In the wooden tower (θωράκιον) on each beast there were, besides the driver, three or four combatants armed with sarisas or bows (Pol. 5.84; Liv. 37.40; Strab. p. 709). Sometimes there were coats of mail put on the animals (Liv. l.c.), and on their backs were purple trappings (Plut. Eum. 14), in front plumes and frontlets (Liv. l.c.). There were special commanders for the corps of elephants (Plut. Eum. 16; App. Syr. 33). One elephant generally used to lead the way (προηγούμενος, Diod. 19.42). When the elephants violently assailed the enemy with tusks and trunk, they were very formidable (Pol. 5.84); but if they did not strike terror at once into their opponents, their influence was frustrated with comparative ease: simple devices, like chains or boards with nails, sufficed to stop their advance (Diod. 18.71; 19.83).

The miscellaneous character of the Hellenistic armies can be very well seen by an examination of the lists of the armies of Eumenes and Antigonus at the battle of Paraetakene (317 B.C.) and of Demetrius at Gaza in 312 B.C. (Diod. 19.27-29 and 82). Mercenaries were very cheap (Diod. 20.113). Antigonus promises each Celt no more than a χρυσοῦς Μακεδονικός, i. e. a stater (Polyaen. 4.4, 17). The frequent use of Celts is very noticeable (Pol. 2.65, 5.53): their characteristic is carrying the great oblong shield (θυρεός), which is so often portrayed by the Pergamene artists. Owing to the long marches, considerable care was bestowed on the commissariat, as is evidenced by the preparations of Antigonus in his Egyptian expeditions (Diod. 19.58; 20.73): for it is to be noticed that military operations, ambuscades, forced marches, surprises, and such κλοπαὶ τοῦ πολέμου, are more frequent in this period than previously, and these required waiting for opportunities (cf. Rustow and Köchly, op. cit. 358 ff.). There was a still larger and more indiscriminate camp-following than in Alexander's time, wives and children of the soldiers being found there in great numbers (Diod. 19.43; Plut. Aemil. 16). The army, says Diodorus (20.41), was actually like a colony. Owing to its size, different divisions had sometimes to take up winter-quarters six days' journey from one another (Diod. 19.37). Great arsenals and military centres were established in the different kingdoms, e. g. at Alexandria and Apamea (Strab. p. 752), where the soldiers were drilled during the winter (Pol. 5.66, cf. 63). Field-artillery was occasionally used (Pol. 11.11, 12). For the camp, which was protected by a rampart and a ditch (Diod. 19.39), but by a very indifferent palisade (Pol. 6.42, 18.18), see CASTRA The generals and distinguished officers in close attendance on the monarch wore purple mantles (Pol. 11.18); hence are called purpurati by Livy (31.35). The bestowing of a purple mantle and a causia was considered a mark of distinction (Plut. Eum. 8). The officers sometimes fought with sarisas, and had, like the phalangites, small shields (Diod. 18.34), but usually they fought with a ξυστὸν on horseback. In Macedonia there were not nearly so many mercenaries [p. 1.781]as in the other Hellenistic armies; moreover, the armies in that kingdom used generally to be disbanded in winter (Pol. 2.54.14; 4.67.3). In Egypt the Ptolemies enrolled a great number of native Egyptians (Pol. 5.65); but apparently only in consequence of a revolution by the Egyptians (Mahaffy, Greek Life and Thought, p. 486): as reserves they had the colonists (κάτοικοι) settled in the military colonies (κατοικίαι) throughout the country. These colonists were of the most varied nationalities--Greeks, Macedonians, Thracians, Celts, Jews, Cretans. They and their sons (ἐπίγονοι) were divided into σημαῖαι, and in time of war could be called out for service as infantry or cavalry, but in time of peace they might engage in any business like ordinary civilians. (For full details on these κάτοικοι in Egypt, see Lumbroso, L‘économie politique de l‘Égypte, p. 224 ff.; and Hermann-Droysen, op. cit. 163.) In Greece at this period, with the exception of the efforts of Philopoemen in the Achaean league and Cleomenes at Sparta, military affairs were in a deplorable state of inefficiency. Mere raids seem to have been most that were attempted; we come across such phrases now as τὸ λάφυρον ἐπικηρύττειν.

(The chief work on the Grecian military systems is that of Rustow and Köchly, Geschichte des griechischen Kriegswesens, 1852. This is the main foundation of all subsequent treatises, the best of which are Die griechischen Kriegsalterthümer by Adolf Bauer, in Iwan Müller's Handbuch der klassischen Alterthumswissenschaft, 1887, 4.226-231, where the vast bibliography on the subject is to be found; and H. Droysen's edition of Die griechischen Kriegsalterthümer, in vol. ii. of K. F. Hermann's Lehrbuch der griechischen Antiquitäten, 1888. Much information is also to be got from Gilbert, Handbuch der griechischen Staatsalterthümer, 1881, and H. Droysen, Untersuchungen über Alexander des Grossens Heerwesen und Kriegführung, 1885.

2. Roman

(down to Septimius Severus). The most generally adopted periods in the development of the army during the Republic is into the citizen army of the first six centuries, and what is called the mercenary army of the last century dating from Marius. The latter formed a transition to the standing army of the Empire. Adopting this division, we shall say something about the legionary and auxiliary force of each period.

I. From Romulus to Marius: the Citizen Army.

1. The early Kings.

Under the early kings the legion or “gathering” (from legere, Varro, L. L. 5.87)--called in Greek generally στρατόπεδον, often τάγμα, τέλος, στράτευμα (D. C. 38.47), τεῖχος (ib. 79.7)--appears to have consisted of three “thousands” (milites) commanded by three tribuni militum and three hundred horse (celeres) commanded by three tribuni celerum. Tradition says (Varro, L. L. 5.89, 91) that the Ramnes, Titles, and Luceres each contributed a third to the legion of Romulus, though the Luceres were not in existence at that time. The tribuni were so called because they presided over a third part of the whole force (Serv. on Aen. 5.560). Besides these regular troops, the old terms velites and arquites point to a subsidiary force of light-armed, especially archers (Mommsen, R. H. 1.79, Eng. trans.). The patricians were the heavy-armed and their clients the light-armed. The method of fighting, if we may judge from the early stories of Livy (e. g. 1.10, 2.6), appears to have been single combats of horsemen in Homeric style. Each warrior, we are told (Festus, s. v. paribus equis, p. 121 M.), brought two horses into battle. Instances of single combat appear long after the Republic was established: e. g. Cossus (Liv. 4.19), Q. Fabius (5.36), Corvinus (7.26), Asellus (23.46, 12), Scipio Aemilianus (V. Max. 3.2, 6), but it was no longer more than an exceptional incident of the battle. Still for a long period the cavalry formed among the Romans a highly important department of the army, used either to charge the enemy at the outset and throw them into disorder, which was completed by the advance of the infantry (Liv. 1.30, 9; 4.47, 2), or to act as a reserve in order to charge at the critical moment (Liv. 3.62, 8; 9.39, 8). A still greater superiority in cavalry lasted among the other Italians till much more recent periods. In the Second Punic War it was the most important part of the Capuan army (Liv. 23.46, 11), and the Romans during the flourishing period of the Republic always got the main portion of their cavalry from the Italian allies.

2. Servius Tullius.

During the monarchy of Servius Tullius we must suppose the increased army due to the increased population. Under him we find four legions, two of juniores from 17 to 46 years (Mommsen, Staatsrecht, i.2 487) and two of seniores from 47 to 60, with 1800 cavalry. Later the calling out of the seniores only took place in great emergencies (Liv. 10.21, 4; cf. 6.6, 14). The infantry now became the chief department of the army, and was arranged in a phalanx like the Doric one, except that the soldiers had missile weapons borrowed from the Etruscans or Samnites (Sal. Cat. 51, 38), especially the hasta, called in Sabellian language quiris (hence Quirites). It is probable too that the Romans also got from the Etruscans the clipeus, and from the Samnites the scutum, though Mommsen (R. H. 1.456, note) in the face of tradition (Diod. Vat. Fr. 54; Verg. A. 7.665; Festus, s. v. Samnites, p. 327 M.), thinks the clipeus and scutum derived from the Greeks. The men were taken from those who had a stake in the country, the ASSIDUI or LOCUPLETES who were formed into five classes, paid tribute, and served at their own expense (Festus, s. v. assiduus, p. 9 M.); while the proletarii, who were outside the classes, were only called out in cases of emergency, and were equipped by the state (Gel. 16.10, 11). The phalanx consisted of six rows of 500 men each and 1200 light-armed. Scholars interpret the accounts of the phalanx as given by Livy (1.43) and Dionysius (4.16, 17) differently. Marquardt (Staatsverwaltung, ii.2 326) supposes the first two ranks only to have been of the first class and to have had the full armour, viz. helmet [CASSIS, GALEA], breastplate [LORICA], round iron shield [CLIPEUS], greaves [OCREAE]; the third and fourth ranks to have been of the second class, and to have had an inferior suit of armour, viz. breastplate, greaves, oblong leather shield [SCUTUM]; and the fifth and sixth ranks to have been of the third class, and to have had only breastplate and scutum. Mommsen, on the other hand (Röm. Tribus, [p. 1.782]p. 138; R. H. 1.99), holds that the first four ranks were taken from the first class, and had the full armour; the fifth from the second class, with the inferior armour; and the sixth from the third class, wearing the third class kind of armour. The fourth and fifth classes, who either formed the rear of the phalanx or more likely fought alongside it, contributed the light-armed or rorarii, so called because they opened the battle with a “shower” of missiles ( “Tractum quod ante maximas pluvias caelum rorare incipiat,” Non. 552, 31), and then retired behind the phalanx (cf. Lucil. x., “Pone paludatus stabat rorarius velox” ). But Dionysius says the fourth class formed the seventh row of the phalanx, and carried a scutum, sword, and spear; and in this he is followed by Huschke (Serv. Tull. 425),> while Livy declares that they had “nihil praeter hastam et verutum,” i. e. fought sometimes in the phalanx with the hasta, sometimes as light-armed with the verutum. The fifth class, according to Dionysius, was outside the phalanx, but had spears (veruta, σαύνια) and slings; Livy says they had only slings and stones. The ferentarii, who fought with missiles ( “quae ferrentur non quae tenerentur” ), were the same as the rorarii (Varro, ap. Non. 520, 12 M.; Festus, p. 369 M.). It is to be carefully remembered that there were twenty centuries of the first class; five of the second, third, and fourth; and seven of the fifth class in each legion. Besides, there were two centuries of artisans (fabri, aerarii et tignarii) to look after the engines, two of trumpeters (cornicines, liticines), and one of accensi velati (unarmed supernumeraries), who were most likely brought to fill up the places of those who fell (Festus, p. 369), and it is barely possible that they may have acted as a sort of pioneers to clear the roads in advance: for in imperial times a body with this name was a college (Orelli, 2182; C. I. L. 6.1607), and had something to say to road-making (Orelli, 111; and Mommsen in Annali dell' Institulo, 1849, pp. 209 ff.). Adscriptitii (from the point of view of the levy) or accensi (from the point of view of the census) when used alone appear to be generic terms, embracing all those not in the phalanx (Festus, s. v. ad scriptitti, p. 14 L.), though Livy (8.8, 8） and Plautus (ap. Varro, L. L. 7.58: “Ubi rorarii estis? En sunt. Ubi accensi? Ecce” ) distinguish them. Lange (Röm. Alt. i.3 535) supposes that the rorarii were the fifth class, and the accensi the proletarii. Subjoined is a table taken from Mommsen (Die römischen Tribus, p. 141), which shows the theoretic arrangement, and how the numbers were derived from the four local tribes. We say “theoretic” because, for example, one can hardly suppose that the seniores can have been as numerous as the juniores. There cannot be a shadow of doubt but that the arrangement was previously a military one (Lange, R. A. i.3 456, 464, 522, &c.); for if it were not so, why should a man of sixty years of age be excluded from the centuries (Mommsen, R. H. 1.100)?

Order of Voting.	Legio I. Junior.	Legio II. Juniorum.	Legio I. Seniorum.	Legio II. Seniorum.	Order of Battle.
Centuria I.
Suburbana.	Palatina.	Esquilina.	Collina.	Total of Century.
Classis I.	25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	First rank	The Phalanx.
25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	Second rank
25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	Third rank
25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	Fourth rank
II. Centuriae fabrum.
Classis II.	25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	Fifth rank
Classis III.	25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	Sixth rank
Classis IV.	25	25	25	25	100 [multi] 5	100 [multi] 5	100 [multi] 5	100 [multi] 5	Rorarii.
II. Centuriae litic. cornic.
Classis V.	25	25	25	25	100 [multi] 7	100 [multi] 7	100 [multi] 7	100 [multi] 7
I. Cent. accens. velat.									Accensi.
Total of Legion.					100 [multi] 42	100 [multi] 42	100 [multi] 42	100 [multi] 42

The cavalry was increased by twelve new centuries by Servius Tullius; and though they still were personally the principal classes in the army, yet in a military point of view they lost their pre-eminence, being now placed on the wings of the phalanx, and in a social point of view they lost their exclusively patrician character. The knights are now the richest of the citizens (Cic. Rep. 2.22, 39), those whose property surpassed that required for the first class, and there was perhaps a fixed sum necessary to enable one to become a knight, as we hear of a census equester in 403 B.C. (Liv. 5.7, 5). They no longer had two horses for themselves, but one for themselves and the other for their attendant; and for procuring and keeping these horses they got from the treasury an aes equestre and aes hordearium, the latter a sum defrayed [p. 1.783]by a tax on widows and orphans (Marquardt, op. cit. ii.2 172). At a triumph they got a threefold share of the booty (Liv. 45.43, 7), when planted in a colony a share of land larger than the others (Liv. 35.40, 5; 37.57, 8), and threefold pay when pay was introduced (Plb. 6.39, 12). The number of active knights who received the aes equestre remained 1800; but those who acquired the property necessary to become knights, without obtaining the equus publicus (i.e. getting the aes equestre), increased greatly, and they were the origin of the ordo equestris (Liv. 2.1-10; Lange, R. A. ii.2 20 ff.). [EQUITES] Thus we hear that at the siege of Veii there were many who served without having an equus publicus (Liv. 5.7, 5); a passage, by the way, which shows that those who did not get the aes equestre probably were not required to serve in the infantry.

3. Camillus.

Important changes were introduced into the army by Camillus: (1) payment of the soldiers from the time of the siege of Veii (406 B.C.), in consequence of the necessary continuance of war operations, from summer into winter; (2) certain alterations of arms, plating of helmets and shields with brass, and teaching the soldiers to use long pila to ward off the stroke of the Gallic sword (Plut. Camill. 40); (3) probable beginning of the breaking up of the phalanx into maniples. This breaking up was most likely introduced in order to resist the first and most dangerous charge of the Celtic sword-phalanx, and was further developed in the Samnite wars, reaching its full perfection in the war with Pyrrhus. The result of the breaking up of the phalanx was that the system of census divisions in the army disappeared, and the arrangement was made to depend on the age and experience of the soldiers: the soldier now advanced from rorarius to triarius (see below). This was quite necessary, as a better military training was required for steadfastness and efficiency in the small manipular unit than in the solid phalanx, where the inexperienced could not leave their place so easily (Mommsen, R. H. 1.454).

But we do not know all the steps in the development of the army to the arrangements it exhibits in the most flourishing period of the Republic, viz. the 6th century of the city. From Camillus to Polybius is over two hundred years, and many changes were effected in that time which we can only guess at. As to the manipular arrangement described by Livy (8.8) in his account of the Latin war of 340 B.C., which he appears to have taken without understanding it from some older authority, it is all too confused and uncertain to allow any definite result to be obtained from it (see Weissenborn ad loc., and Madvig, Verfassung, 2.483, 489). For the immense literature which has gathered round the passage, but reached no definite explanations, see Marquardt, op. cit. ii.2 360, note 1. Passing by this perplexing description, let us proceed to sketch the state of the army as described by Polybius.

4. Polybius.

A certain property was still required for service in the army, but it was now only 4,000 asses instead of the 11,000 of Servius Tullius (Plb. 6.19, 2). The normal number of men in the legion continued as in the time of Servius Tullius, viz. 4,200, roughly 4,000, infantry (cf. for 494 B.C. Dionys. A. R. 6.42; for 381, Liv. 6.22, 8; for 349, Liv. 7.25, 8; for 225, Plb. 2.24, 13; for 146, Plb. 6.21, 9). In especial cases the numbers were 5,200 or 5,000 (Plb. 2.24, 3; 3.107, 10; 6.20, 8;--Liv. 26.28, 7; 40.1, 5). It was first in the war against Perseus that the legion was raised to 6,000 (Liv. 42.31, 2), and the regular number from Marius's time was 6,200 (Festus, s.v. sex millium, p. 336 M.), a number which appears occasionally earlier (Liv. 29.24, 14; 35.2, 4). The old quota of cavalry for the legion was 200 (Pol. 3.107, 10; Liv. 42.31, 2), but the usual one for the legion, whether of 4,200, 5,000, 5,200, 6,000, or 6,200, was 300 (see Polyb. and Liv. ll. cc.).

Taking then the normal legionary force at 4,200 infantry, it was divided in Polybius's time (6.21, 7-9; cf. Gel. 16.4, 6) into 1200 hastati, 1200 principes, 600 triarii--these three classes arranged behind one another in the form of a quincunx, and 1200 velites (γροσφόμαχοι). The latter are the “youngest and poorest,” the hastati “those next them” (i. e. in these two respects), the principes “the most vigorous in years” (τοὺς ἀκμαιοτάτους ταῖς ἡλικίαις), and the triarii “the oldest” (cf. Liv. 8.8, 8). All three classes were armed with a metal helmet [CASSIS], having a red or black plume 1 1/2 feet high (Pol. 6.23, 12), but no visor, a leathern shield [SCUTUM], greaves [OCREAE], a leathern breastplate [LORICA], a short Spanish two-edged sword [GLADIUS], which was worn at the right side while they carried a dagger at the left. The spear of the hastati and principes was the light pilum [PILUM], used for throwing, while the triarii carried the hasta for thrusting [HASTA]. Besides this armour, the soldier of course had such articles of attire as the sagum [SAGUM

The hastati, principes, and triarii were broken up into thirty “bundles” (manipuli), each “bundle” forming the smallest tactical unit which was under one standard (Varro, L. L. 5.88; cf. Tac. Ann. 1.34, “discedere in manipulos” ). The manipulus derived its name, if not as indicated from itself being a bundle of men (cf. Madvig, Verfassung, ii. p. 486, note), from the bundle or handful of hay (Plut. Rom. 8; Ov. Fast. 3.115 ff.) fixed on a pole which served as the standard. Later signum was the term for the standard of the maniples (Liv. 27.14, 8), and Polybius very frequently uses σημαία for the maniple (though the usual word is σπεῖρα, τάγμα sometimes being found). Occasionally the maniple had more than one signum (Pol. 6.24, 6). The maniple continued as a recognised division of the army even after the introduction of the arrangement by cohorts (Caes. B.C. 2.28, 1; Tac. Ann. 1.34) down to very late times (Amm. Marc. 21.13, 9), though we sometimes find it used in a vague sense for any detachment (Tac. Ann. 14.58, 59). The maniple was further divided into two centuries, each commanded by a centurion, the centurion which commanded the right century being elected before him who commanded the left (Pol. 6.24).

The hastati then formed 10 manipuli of 120 men each or 20 centuries of 60 men each. So did the principes. The triarii, who were always 600, and never varied with the varying numbers [p. 1.784]of the legion (Pol. 6.21, 10), formed 10 maniples of 60 men each or 20 centuries of 30 men. To each century 20 velites were added (ib. 24, 4). The maniple of the first two ranks with its velites was usually arranged with 20 men in front and 8 in depth, according to Nast (Röm. Kriegsalterthümer, p. 51) and Marquardt (Staatsverw. ii.2 346), on the not very strong grounds that such was the usual depth among the Greeks and among the Romans in the time of Trajan. Mommsen (R. H. 1.453) appears to think that the depth was not more than four files, the maniples of the three ranks having a front accordingly of 40, 40, and 20 men; and this seems the preferable view, as the main aim of the manipular arrangement was to encourage the individual mode of fighting. This latter was the cause of the considerable space, it may have been 6 feet, which was allowed in what was called loose array (laxatis ordinibus), between each soldier and his fellow-soldier beside or behind him; in close array (confertis ordinibus） the space was only 3 feet (cf. Pol. 18.13, 6; Liv. 22.47, 5; Caes. Gal. 2.25, 1; Veg. 3.14, though it must be confessed these passages only make for and are not quite conclusive as to such an accurate distinction). The general form of opening the battle was for the hastati to hurl their pila at a distance of about 10 or 20 paces from the enemy, and then proceed to the attack with the sword, where single combats prevailed. If this did not finish the battle, the principes advanced. The triarii always acted as a reserve, sometimes at the camp (Liv. 5.19, 7; 7.23, 7), as was their duty originally (Dionys. A. R. 5.15; 8.86). For further details, see the account of the Battle Array below, p. 807 b.

The great advantages of this divided light manipular arrangement as against the serried unwieldy phalanx, are that the former was ready for any emergency, that it was not disarranged by uneven ground, that a small disorder did not break up the whole array as in the phalanx, that the individual soldier was better able to vary his method of fighting according as it is necessary that the maniple, century or individual should engage ; and many more which are set forth by Polybius, 18.31, 32. Mommsen (R. H. 1.455) notices how the Roman military arrangements of this time attained the three great military principles of keeping a reserve (the triarii), of combining the distant and close methods of fighting (in the discharge of pila preparatory to the sword attack), and in combining the offensive and defensive (the latter especially in the formation of camps).

The 300 cavalry of the legion fell into 10 turmae (ἴλαι） of 30 men. Each turma was commanded by three decurions (ἴλαρχαι) and three optiones (οὐραγοί), and had a vexillum (Veg. 2.14). The turma stood in three rows, each row having a decurio and an optio in the first and last place, the first elected decurio leading the whole troop (Pol. 6.25, 1, 2). They charged in close order (Sal. Jug. 101, 4), often taking the reins from the horses (Liv. 8.30, 6; 40.40, 5) ; but in standing fight they doubtless extended their ranks if they did not dismount and fight on foot (Liv. 31.35, 5). Originally the cavalry had no armour, only a tunic so as to enable them to leap on and from their horses easily, ox-hide shields incapable of resisting severe thrusts and easily damaged by wet, very light lances with a point only at one end; but in the time of Polybius they had adopted the Greek cavalry equipment, which consisted of breastplate, covering for the loins, greaves, metal helmet, round shield [PARMA], lance, and long sword, “which as soon as they saw they adopted speedily; for the Romans are pre-eminently good at adopting new practices and striving after improvements” (Pol. 6.25, 3-11). Possibly, as among the Greeks, the horses wore protections for the head, breast, and sides (προμετωπίδια, προστερνίδια, παραπλευρίδια). Saddles appear to have been used in Caesar's time (B. G. 4.2, 5), but they had no stirrups, as may be seen from illustrations, and from the fact that recruits were taught to vault into the saddle (Veg. 1.18).

During the Republic the Roman cavalry was always weak; and its inferiority was so marked as against the fine Campanian cavalry that, in order to be able to contend with it at all, the Romans in 211 B.C. (Liv. 26.4) selected from the legions the most active youths, who carried each a parma and seven iron-tipped javelins (hastae velitares), 4 feet long each. These youths, whom Livy calls velites, used to ride behind the horsemen, and leaping down at a given signal hurl their darts in rapid volleys. This practice of foot-soldiers fighting amongst the cavalry existed also amongst the Germans (Caes. Gal. 1.48, 5; Tac. Germ. 6), and Caesar formed such a troop of Germans (B. G. 7.65, 4; 8.13, 2). This practice appears to have been afterwards regarded as a regular specific for strengthening weak cavalry (Veg. 3.16). Of course velites, meaning light-armed, existed long before, as can be seen even in Livy (21.55, 11); but it appears that from this time (viz. 211 B.C.) the rorarii disappear, and velites is the name for the lightarmed of the legion (Liv. 30.33, 3). They combined both distant and close conflict (Liv. 31.35, 5), whereas the rorarii only fought from a distance; and further the velites now formed an integral part of the maniples or turmae, whereas the rorarii were troops apart.

A word must be said in conclusion on the terms hastati, principes, and triarii, terms which were not clear (minus illustria) in the manipular arrangement, as Varro says (L. L. 5.89). The front rank is called hastati, but their spear was the pilum. How is it that the second rank is called principes, i.e. “the first men” ? And the triarii, who are also called pilani, do not carry the pilum at all, but a hasta (δόρυ, Pol. 6.23, 16). We must take refuge in conjecture to explain these difficulties. The terms probably came from the old phalanx, the principes or proci being the front rank (Festus, p. 249 M.), as principia means the front rank often (Liv. 2.65, 2; Sall. Jug. 50, 2), and so being members of the first class. The term hastati was one applied to all the members of the phalanx who were all armed with the hasta, but it got narrowed down to a portion of what it originally implied when the name of principes was appropriated by the first class, just as, for example, the term dies fasti became confined to a portion of the Dies Fasti (see Lange, R. A. i.3 360, 532), or the term centuria to half a century when the maniple was introduced. The triarii, the third rank (we should more naturally expect tertiarii), [p. 1.785]used originally, as we saw, to guard the camp. Doubtless in this duty they used the long heavy pilum [PILUM]; hence were called pilani (Varro, L. L. 5.89), a name which was in a measure retained not only in the titles of the centurions, but also in the term antepilani as applied to the hastati and principes (Liv. 8.8, 7). Now, when the class-distinctions began to be superseded and efficiency made the ground of the arrangement in the legion, the chief and most able soldiers, who socially too were the most important and who had been in the front rank originally and as such still retained the name of principes, were not used for the first assault, but kept in case the battle proved a serious one; the inferior classes, the hastati, being placed in front. The triarii continued to be the reserve. They had been probably the older men left to guard the camp; but now that age meant generally experience, as being used for the reserve these veterans formed the most efficient and tried portion of the army.

II. From Marius to Augustus--the Mercenary Army.

The terrible defeat of the Romans at Arausio in 105 B.C. rendered it necessary to make every effort possible to raise sufficient and adequate forces to meet the Northern invaders. This serious pressure hastened and consummated the introduction of changes in the army which the altered political and social condition of the times rendered inevitable sooner or later. The gradual shrinking of the rich upper classes from service (Sal. Jug. 85, 3) and the disappearance of the middle class left nothing open but the admission of all free-born citizens to the legions; that is, if an adequate army was to be raised. The census indeed in Polybius's time was only 4,000 asses (6.19, 2), but now nothing further was required but free birth,--a recruit need possess nothing but his caput (capite census). This was the most vital change possible. Men with no stake in the country and nothing to go home to after service, who looked on service as a means for enriching themselves and not as a temporary burden, were certain to become, as they did, faithful followers of their leader if he led them to plunder (Sal. Jug. 86, 3) and to care little about their country (Appian, App. BC 5.17)--in fact, practically to become mercenaries. It is not without reason, then, that Marquardt (Staatsv. ii.2 321) calls the period from Marius to Augustus the period of the Mercenary Army (Söldnerheer). The essence of such an army is the supremacy of the general and the equality of the soldiers. All the traditional distinctions of velites, hastati, principes, and triarii disappear; their definite and traditional place in the order of battle, their military rank, armour, standards, all were superseded; the recruits were now all on an equality and uniformly trained, a new and severe method of drill and training like that of gladiators having been devised by P. Rutilius Rufus, consul in 105 B.C. (V. Max. 2.3, 2). In place of the manipular division, the last mention of which is its use in the Jugurthine War by Metellus (Sal. Jug. 49, 6), the unit now becomes that of the cohort, viz. a combination of three maniples. Whether it was the maniples beside one another, or the three behind one another, which were massed together, is uncertain, probably the latter. The number of the legion at this time was, roughly speaking, 6,000 (Plut. Sull. 9, compared with Mar. 35; Appian, Mithr, 72; Cic. Att. 5.1. 5, 1, compared with Plut. Cic. 36); though Caesar had generally less than this number, sometimes not more than about 3,500 men in his legion: e. g. the 13th Legion, which he had at Ariminum at the outbreak of the Civil War (Caes. B.C. 1.7, 7), consisted apparently of 5,000 (Plut. Caes. 32, Pomp. 60), but Caesar himself reckons two legions (B. G. 5.49, 7) and a few cavalry at 7,000 men. Still, whatever was the strength of the legions, the number of cohorts was always 10. These had no traditional arrangement, and were disposed of in battle just as the general thought fit: perhaps the triplex acies (e.g. B. G. 1.24, 2), viz. 4 cohorts in front and 3 in each of the other ranks, was the most usual; though the single (Bell. Afr. 13, 2), double (Caes. B.C. 3.67, 3), and quadruple lines of battle (ib. 89, 3) are also found. The special application of the word cohors to a definite portion of a legion must be noticed, beside the technical use of the word to express the divisions of the auxiliary force, the vague use that Livy makes of it by applying it to any division of troops at all (parmata cohors, 4.38, 3), and the misuse by so translating (σπεῖρα (30.33, 1, compared with Plb. 15.9, 7). The pilum was now made the common weapon of the whole legion, the haste of the triarii being done away with. The old standards--the eagle, wolf. Minotaur, horse. and boar (Plin. Nat. 10.16)--give place to the silver eagle (Cic. Cat. 1.9, 24), which now becomes the chief standard of the legion.

The Roman cavalry had disappeared before Marius. Its last mention is in the Spanish campaign of 140 B.C., where it behaved with great pretension and insubordination (Mommsen, R. H. 3.200); and after the Jugurthine War it vanishes entirely. Even the Italian cavalry had been for long unable to cope with the enemies of Rome: it had been defeated by Hannibal in Italy, and Scipio only won Zama by the cavalry of Masinissa. It vanished completely in the Social War; and after that foreign troops--Gauls, Spaniards, Thracians, and Africans (Plut. Ant. 37)--were taken into the service in larger numbers than heretofore. In Caesar's army they were about a fourth or a fifth of the infantry: with 6 legions (=about 21,000 men) he had 4,000 (Caes. Gal. 1.15, 1), with 5 legions 5,000 (B.C. 1.39, 1); in the army of Brutus at Philippi there were more than 1,000 to the legion (App. BC 4.108). Some minor changes and improvements in practical details are attributed to Marius, such as the muli Mariani, a forked contrivance fastened on the shoulder of the legionary to enable him to carry his baggage (see below, p. 807a), altering the pilum [PILUM], taking away the parma from the auxiliary troops (Festus, p. 238, s. v. parmulis), &c. The numbering of the legions probably began also in this period; certain it is that we find it quite regular in Caesar's time (C. I. L. 1.624; Cic. B. G. 8.8, 2). Originally the legions had been numbered according as they were raised (D. C. 38.47); but then the numbers had no durable significance, owing to the speedy disbanding of the legions. [p. 1.786]

The Auxiliary Forces under the Republic.

1. The Socii.

In the time of Tarquinius Superbus the Latins were incorporated in the Roman army (Liv. 1.52, 6): but on their regaining their independence after the establishment of the Republic, they furnished their contingent to the united army, and held the chief command alternately with the Romans. But after the great Latin War, which ended in 338 B.C., the Latins became cives sine suffragio. The members of those municipia who lost their constitution were enrolled in the legions, those who retained their municipal constitution formed self-subsisting divisions, such as the legio Campana we hear of in Liv. Ep. xii. (cf. Plb. 2.24, 14). After the Second Punic War, when the differences of the old municipia ceased by the attainment of the full franchise, such legions disappeared. The other kind of Italian states besides the municipia--viz. the civitates foederatae and the Latin colonies which retained self-government on the basis of a treaty guaranteeing a definite amount of troops--it was these communities which furnished the socii to the Roman armies (cf. Marquardt, Staatsv. i.2 25-58, ii.2 390-1).

They contributed nothing to the legions, but had to supply each year a force of auxiliaries for the army, and ships and sailors for the fleet. The allied state raised, swore in and paid (Liv. 27.9, 13) this force, but it was fed during service by the Romans (Plb. 6.39, 14). The consul proclaimed by an edict the amount of the force each state was to supply, and where and when it was to assemble; and it appeared there with its own leader and paymaster (ἄρχοντα καὶ μισθοδότην, Plb. 6.21, 4, 5).

As to the actual force contributed, Polybius varies in his account. Sometimes (3.107, 2; 6.26, 7) he says the force of foot equalled that of the Romans, the horse being three times that of the Roman horse--in both passages including the extraordinarii, who were one-third of the horse and one-fifth of the foot (see below). At another time he says (6.30, 2) that the force of foot equalled that of the Romans, the horse being twice that of the Roman horse--this time excluding the extraordinarii. He agrees then as regards the horse in all three passages. But the discrepancy is patent about the infantry. All we can say is that Polybius, if he did not make a mistake, was speaking only with a loose approximation in the first two passages, as the expressions πάρισον ( “about equal” ) and πάρισον ὡς ἐπὶ τὸ πόλυ may lead us to infer. The last passage indeed probably gives the correct view for Polybius, in his account of the battle on the Trebia (3.72, 11), gives 16,000 Romans and 20,000 allies as the complete force of infantry of two consular armies. Deducting one-fifth, we get exactly the same number of allied and Roman infantry; but the regular legionary force for two consular armies ought to be 16,800, though Polybius himself, as we saw, speaks of the legion as roughly consisting of 4,000 men (3.107, 10). But the force of allies is generally greater than those given in these passages by Polybius. The elaborate schedule of forces available by Rome in 225 B.C. given by Polybius (2.24) from Fabius the annalist, and, as it is reproduced in some of its numbers by other historians (Liv. Ep. xx.; Diod. xxv. fr. 13 Dindorf; Eutrop. 3.5; Oros. 4.13; Plin. Nat. 3.138), doubtless derived from official sources, is too defective to decide details with certainty; but it shows that the active force of the Romans in that year was only one-fourth that of the allies (see Mommsen in Hermes, 11.49-60). Generally it was about one-half (Vell. 2.15; Liv. 36.2, 8). According to Livy (40.36, 6), the normal number (quantus semper numerus) of allies directly attached to two legions was 15,000 infantry and 800 cavalry, and as a matter of fact we do find these numbers in Liv. 33.43, 3, 40.26, 7. Other numbers of allies for two legions are 15,000 and 600 (37.2, 4), 15,000 and 500 (34.56, 6), 10,000 and 600 (41.21, 4), 12,000 and 600 (41.9, 2), 20,000 and 800 (35.20, 4). For further details, see Marquardt, op. cit. ii.2 395, note 1; Weissenborn on Livy, 40.36, 6; and Mr. Strachan-Davidson's Selections from Polybius, pp. 22-32.

As to the divisions of the allies, they were never independent divisions, but always portions of the combined army (Weissenborn on Liv. 2.53, 5). In battle the allies were placed on the wings, the two sections being called dextra and sinistra ala or κέρας (Liv. 31.21, 7; Plb. 6.26, 9). After the Latin War in 338 till the end of the Punic Wars, the allied infantry regularly formed cohortes, called after the separate nations; thus we have cohortes Samnitium (Liv. 10.40, 6), cohortes Marrucina et Peligna . . . Firmana Vertina Cremonensis duae turmae equitum Placentina et Aesernina (Liv. 44.40, 5, 6). The cohort was the usual contingent for each community, and was at this time the unit of the allied infantry, just as the unit of the Roman troops was the maniple: cohorts and maniples are often opposed (Liv. 10.33, 1; 41.1). Like the maniple, the cohort had a standard (Liv. 25.14, 4), and was regarded as a regular division in the camp [CASTRA]. Each ala had 10 cohorts (Liv. 10.43, 3); sometimes we find 15 (Liv. 30.41, 5), though perhaps the legion on this occasion was stronger than usual (Weiss. ad loc.). The ordinary cohorts were called cohortes alares or alariae, to distinguish them from the cohortes extraordinariae (Liv. 10.40, 8), and in Caesar (Caes. Gal. 1.51, 1) from the legionarii.

The commanders of these allies were, for the contingent assigned to 4 legions, 12 praefecti sociorum, appointed by the consuls (Plb. 6.26, 5). These praefecti were Roman citizens (Liv. 23.7, 3; 25.1, 3). For a contingent of 2 legions (there were 6 praefecti, three for each ala. Besides these Roman officers, there were the native officers, called ἄρχοντες by Polybius (6.21, 5), and praetores or even praefecti by Livy (23.19, 7; 25.14, 4). The former term was, however, the most common one used by that historian for the generals of foreign states; and the latter native praefecti who commanded a cohort must be carefully distinguished from the Roman praefecti, who commanded the whole ala (Weissenborn on Liv. 21.3, 4; 25.14, 4).

The number of the cavalry of the allies, Polybius says, was three times that of the Romans; accordingly 1800 for two legions. It was divided into 4 alae (in the strict sense) of ordinary cavalry (called equites alarii, often opposed to legionarii equites, Liv. 35.5, 8; 40.40, 9; Tac. Ann. 3.39; Veg. 2.1) and two alae of [p. 1.787]extraordinarii. The ala in this strict sense of squadron of cavalry consisted of 300 men, divided into 5 double turmae of 60 men each [CASTRA]. Each double turma had a signum (cf. Liv. 27.12, 17). This number of 30 always remained constant for the single turma, and in case of increase of numbers it was the number of turmae which was increased (Hyg. §16). Ala is a general term for a squadron of allied cavalry, just as cohors is for a battalion of allied infantry: thus we find during the Republic alae of 400 and 500 men (Caes. Bell. Afr. 78, 7; Liv. 10.29, 12), just as there were cohortes of 460, 500, and 600 men (Liv. 23.17, 8, 11; 28.45, 20). But the ala differed essentially from the cohort in this, that while in the latter the soldiers were of the one nation or people, the ala was composed of different nationalities. Marquardt (op. cit. ii.2 400), however, thinks that as in the camp there was a double turma of equites put beside a cohort of infantry, these latter may have been of the same nationality as the cohort; cf. Liv. 29.15, 6--a passage which shows that a double turma of cavalry appears to have been a normal contingent of cavalry from a Latin colony, as a cohort was the normal contingent of infantry (cf. Weissenborn on Liv. 28.45, 20).

The extraordinarii (ἐπίλεκτοι) were a picked body of horse and foot from the allies (Fr. Fröhlich, Die Gardetruppen der röm. Republik, 1882, pp. 4 ff.). The extraordinarii pedites were one-fifth of the infantry of the allies (Plb. 6.26, 8). If we take this exactly, we have 8,000 ordinarii and 2,000 extraordinarii for two legions. But we saw Polybius avoided exact numbers, and it suits the arrangement of the camp and the order of battle (e. g. Liv. 37.39, 7) better if we suppose the ordinarii equal to the legionaries, i. e. normally 8,400, and place the extraordinarii at 1600. These extraordinarii were chosen by the praefecti sociorum as being most suitable for real service (πρὸς τὴν ἀληθινὴν χρείαν, Pol. 6.26, 6). They formed four cohorts (Liv. 40.27, 3; cf. 27.12, 14) of 400 men. The extraordinarii equites were one-third of the whole contingent (which normally was 1800, three times that of the legionary cavalry), and accordingly numbered 600. These formed two alae (in the strict sense of squadrons of cavalry) of 300 men each (Liv. 40.31, 3), which were divided into 5 double turmae of 60 men each [CASTRA]. The extraordinarii formed a class intermediate between the heavy and the light forces. They were used for reconnaissances, flank movements, and generally on difficult services which required vigour and celerity. They appear to have been free from the duties of constructing the camp and keeping watch (see Fröhlich, l.c.).

2. Auxilia

The Auxilia properly so called did not come into any prominence till the foreign wars of Rome. Though usually they were taken into the service as allies (Festus, p. 17: “Auxiliares dicuntur in bello socii Romanorum exterarum nationum” ), yet there are some traces of mercenary forces employed by the Romans in the Punic Wars, e. g. Gauls (Zonar. 8.16) and Celtiberians (Liv. 24.49, 8); but the Cretan archers taken at Trasimene (Liv. 24.30, 13) were sent by Hiero (Plb. 3.75, 7). During the wars in the East their numbers increased greatly. After the Social War the class of socii disappeared entirely, and there were again but two classes of soldiers: (1) legionaries from Italy, (2) auxiliaries from the provinces, from allied states and princes, and even from independent nations. The number of auxiliary infantry varied according to necessity. We know that they were divided into cohorts, out no more. Some of the cohorts were not armed or disciplined in the Roman fashion, but used their national arms, sagittarii, funditores (Caes. B.C. 3.4, 3), cohortes cetratae et scutatae (Caes. B.C. 1.39, 1). The cavalry of the Roman army, however, in Caesar's time consisted entirely of auxiliary troops (Caes. Gal. 1.15, 1; 1.42, 5; cf. 2.24, 4). Italians are found among the cavalry (App. BC 2.70), but they are volunteers. The cavalry consisted of Gauls, Spaniards (Plut. Ant. 37), Thracians, Numidians (Sal. Jug. 38, 6; 68, 2), even Germans (Caes. Gal. 7.13, 1), They were formed into alae of about 400 men each, which were subdivided into turmae and decuriae (Bell. Afr. 29, 4), commanded by praefecti equitum (Caes. Gal. 3.26, 1), who were doubtless, in the case of allied and independent states, natives (cf. Hirt. B. G. 8.12, 4).

III. The Imperial Period--the Standing Army.

The Empire established by the sword had to be retained by the sword, and the army now becomes a standing one,--στρατιῶται ἀθάνατοι, as Dio Cassius calls them (56.40). The emperor is the supreme commander (imperator, αὐτοκράτωπ, Dio Cassius, 57.8); and to him, as represented by his images carried on a pole, the legions and auxiliaries swore allegiance twice a year, on the date of his accession (Plin. Epp. 10.52 (60)) and on the Kal. Jan. (Tac. Hist. 1.55), besides of course the oath of allegiance at accession (Dio Cassius, 57.3). The oath was administered by the provincial governors in the provinces and by the three praefecti at Rome.

1. The Legions.

Under the early Empire the legion contained 10 cohorts and 60 centuries--altogether between 5,000 and 6,000 men. Now that the Pax Romana was established, they were mostly used on garrison duty, and accordingly got again a small quota of cavalry, each legion getting 120 in 4 turmae of 30 each. These are the equites legionis of Tac. Ann. 4.73; Hist. 1.57. From the time of Hadrian the first cohort had 10 centuries (Eph. Epigr. 4.227), as was the case in the latter half of the third century also, according to the arrangement described by Vegetius (2.6, 13), which was probably established by Aurelian. This arrangement varied slightly from that of the early Empire. The first cohort had 1105 men in 10 centuries, and 5 ordinarii (cf. p. 799 b); the other nine cohorts, 555 men in 5 centuries each. Each century consisted of 10 contubernia (also called manipuli), and each contubernium was presided over by a decanus: so that the infantry of the legion consisted of 6,100 men and 5 ordinarii. The cavalry were 726 soldiers, viz. 132 to the first cohort and 66 to each of the others; arranged in 22 turmae, each of 30 men and 3 officers (decurio, duplicarius, sesquiplicarius: cf. Veg. 2.6, 14; Hyg. § 16; and see below, p. 809 a). [p. 1.788]

Octavian, when in 36 B.C. he took over the 22 legions of Lepidus (Appian, App. BC 5.123), amongst which were 8 of Sextus Pompeius, had the command of 45 legions (ib. 127, cf. 53). He then disbanded 20,000 veterans (ib. 129), all the Pompeian legions (Dio Cassius, 49.12), and most of those of Lepidus. At the battle of Actium, Antonius had at most 30 and Octarian 25 (Plut. Ant. 61), so that when the victory was won Octavian had a little over 50 legions to dispose of. He retained the first 12 of his own and 6 of those of Lepidus and Antonius. This accounts for two legions being found at times with the same number. He retained these numbers in order not to confuse the soldiers by unnecessary alteration, a distinction being sufficiently marked by the name of the legion, partly perhaps to delude the people that he had only 12 legions, or possibly with a hope that they might be speedily disbanded. Augustus enrolled legions XIII. to XX. in order to face the great German war waged in 3 A.D. against Maroboduus, and afterwards against the Pannonians and Dalmatians (cf. Vell. 2.104; Suet. Tib. 16). In the subjoined list it will be seen that none of these were planted in colonies, and none are found to have duplicates. Legions XXI. and XXII. were added after the defeat of Varus, in which three legions fell (XVII., XVIII., XIX.). As to Legio I., which was probably disbanded after the defeat of Varus (for we hear of a First legion getting its standards from Tiberius, Tac. Ann. 1.42), Mommsen supposes its existence from the consideration that a first legion can hardly have been wanting in the army of Augustus. We now add the list of Augustus's legions, which were 25 at his death, 23 after the defeat of Varus (this reconciles Tac. Ann. 4.5 with Dio Cassius, 4.23, 2), with the localities in which they were stationed, as given by Mommsen (Res gestae d. Aug., pp. 68-9), adding in square brackets here and there a word as to the reason why such and such a legion was disbanded; with a continuation of the legions which were enrolled up to the time of Septimius Severus.

LEGIONS.	QUARTERS.	SPECIAL ACTIONS.	DURATION OF EXISTENCE.
I.		Lower Germany		Disbanded after defeat of Varus(?).
I.	Germanica	Lower Germany		Not mentioned after 70 A.D. [Probably disbanded on account of its conduct in the rebellion of Civilis （Tac. Hist. 4.25, &c.).]
II.	Augusta	Upper Germany		Existed in Dio Cassius's time.
III.	Augusta	Africa		do.	do.
III.	Gallica	Syria (Marq. staatsv. ii.2 447, note 7)	Took part in the Parthian war of M. Antonius (Ac. Hist. 3.24)>	do.	do.
III.	Cyrenaica	Egypt		do	do.
IV.	Macedonica	Spain	At battle of Philippi: hence probably its name	Not mentioned after 70 A.D. [Perhaps disbanded on account of its vigorous support of Vitellius (Tac. H. 3.22).]
IV.	Scythica	Moesia		Existed in Dio Cassius's time.
V.	Macedonica	Moesia	Took part in African war, 46 B.C. (App. BC 2.96), also in battle of Philippi: hence its name. Veterans of this legion planted in a colony at Berytus in 16 B.C.	do.	do.
V.	Alauda	Lower Germany	Enrolled by Caesar (Suet. Jul. 24): in 16 B.C. lost its eagle in Lower Germany (Vell. 2.97)	Not mentioned after 70 A.D. [probably destroyed by the Sarmatae (Suet. Dom. 6): yet Mommsen (Eph. Epigr. 5.214) and Schiller (Kaiserzeit. 511) say it was disbanded by Vespasian.]
VI.	Victrix	Spain		Existed in Dio Cassius's time.
VI.	Ferrata	Syria		do.	do.
VII.	(Subsequently Claudia)	Illyricum	At battle of Philippi	do.	do.
VIII.	Augusta	Illyricum	Veterans planted at Berytus, 16 B.C.	do.	do.
IX.	Hispana	Illyricum (Tac. Ann. 3.9; 4.23）	At battle of Philippi	Not mentioned after Trajan. [Probably annihilated by the Brigantes in Britain (Schiller, l.c. 607, note 6).]
X.	Fretensis	Syria	Fought in Sicilian war at the Straits against Sext. Pompeius: hence its name. Veterans of both tenth legions planted at Patrae, 16 B.C.	Existed in Dio Cassius's time.
X.	Gemina	Spain		do.	do.
XI.	(Subsequently Claudia)	Illyricum		do.	do.
XII.	Fulminata	Syria	Veterans planted at Patrae 16 B.C.	do.	do.
XIII.	Gemina	Upper Germany		do.	do.
XIV.	Gemina	Upper Germany		do.	do.
XV.	Apollinaris	Illyricum		do.	do.
XVI.	Gallica	Upper Germany		Not mentioned after 70 A.D. [Probably, like I., disbanded on account of its mutinous conduct in the rising of Civilis (Tac. Hist. iv.).]
XVII.		Lower Germany		Fell in defeat of Varus, A.D. 9.
XVIII.		Lower Germany	G. I. Rh. 209	do.	do.
XIX.		Lower Germany	Tac. Ann. 1.60	do.	do.
XX.	Valeria Victrix	Illyricum (6 A.D.), afterwards in Lower Germany		Existed in Dio Cassius's time.
XXI.	Rapax	Lower Germany		Not mentioned after Domitian. [Disbanded for taking side of Antonius Saturninus.]
XXII.	Deiotariana	Egypt		Not mentioned after Trajan. [Last mention in C. I. L. 3.36, of year 84 A.D.: if we are not to suppose with Borghesi (4.254) that this was the legion annihilated by the parthians in 162 A.D. (D. C. 71.2).]
Added by Claudius by division of XV. and XXII.
XV.	Primigenia	Lower German		Disbanded by Vespasian as having revolted to Civilis (Pfitzner, Gesch. der Kaiserlegionen, p. 262).
XXII.	Primigenia	Upper Germany (Tac. Hist. 1.18, 55）		Existed in time of Victorinus (Cohen, 40, p. 16). For its omission in Dio Cassius, see Marq. Staatsv. ii.2, 452, note 1.
Added by Nero.
I.	Italica	In Gallia Lugdunensis (Tac. Hist. 1.59, 64）
Added by Galba.
I.	Adjutrix	Spain (Tac. Hist. 2.67; 3.44）	Enrolled from the fleet (C. I. L. iii. p. 907)	Existed in Dio Cassius's time.
VII.	Gemina	Spain (Tac. Hist. 2.11）		do.	do.
Added by Vespasian (Dio Cassius, 4.24, 3).
II.	Adjutrix	Pannonia		do.	do.
IV.	Flavia	Upper Moesia		do.	do.
XVI.	Flavia Firma	Syria		do.	do.
Added by Domitian.
I.	Minervia	Lower Germany		do.	do.
Added by Trajan.
II.	Trajana, Fortis Germanica	Egypt	Enrolled 109 (C. I. L. 3.79).	do.	do.
XXX.	Ulpia Victrix	Lower Germany	Was located for some time in Upper Pannonia (C. I. L. iii. p. 482)>	do.	do.
Added by M. Aurelius.
II.	Italica	Noricum		do.	do.
III.	Italica	Rhaetia		do.	do.
Added by Sept. Severus.
I.	Parthica	Mesopotamia		do.	do.
II.	Parthica	Alba, near Rome	Henzen, Annali dell' Inst. 1867, p. 73.	do.	do.
III.	Parthica	Mesopotamia		do.	do.

[p. 1.789]

An interesting list of the quarters of the different legions in the middle of the first century, derived from the Histories of Tacitus, is to be found in Borghesi, iv. p. 240, reproduced by Marquardt, Staatsv. ii.2 p. 449, note 4; also of their quarters in the middle of the second century in C. I. L. 6.3492 a, b (= Orelli, 3368, 3369). Detailed histories of the separate legions are numerous. See Marquardt's notes; an admirable history of all the legions individually is given by Grotefend in Pauly, 4.868 ff.

As to the names of the legions, they arose from various causes: (1) in the later Republic, sometimes from the place of levy, e. g. V. Urbana (C. I. L. 5.2514), VIII. Mutinensis, IV. Sorana (see Mommsen on C. I. L. 10.5713); (2) from the people against whom they successfully fought, e. g. IV. Scythica, or from the place, X. Fretensis; (3) from divinities, e. g. I. Minervia; (4) from division of a single legion, e. g. Primigenia (this part was probably held the superior, and retained the eagle), or uniting of two, e. g. Gemina (Caes. B.C. 3.4, 1), yet cf. also Mommsen, Res gestae d. Aug. p. 73, note 1 fin.; (5) from certain insignia, e. g. probably the V. Alauda, certainly XII. Fulminata (κεραυνοφόρον); (6) various epithets expressive of [p. 1.790]courage and devotion, e.g. pia, fidelis, constans, victrix, aeterna, rapax; (7) up till Caracalla's time, when the emperor's name was added to the legions, except in the case of a few, as VII. and XI. Claudia pia fidelis, it marked the founder, as II. Trajana, XXX. Ulpia Victrix. After Caracalla all legions bore the name of the reigning emperor (cf. Schiller, Kaiserzeit. 742, note 1).

2. The Auxilia.

Concerning the auxiliary troops under the Empire, Tacitus says (Ann. 4.5): “At apud idonea provinciarum sociae triremes alaeque et auxilia cohortium, neque multo secus in iis virium: sed persequi incertum fuit cum ex usu temporis huc illuc mearent, gliscerent numero et aliquando minuerentur.” They consisted of those forces, beside the legions, which were raised in the provinces. Perhaps we may assume that in numbers they were about equal to the legionaries (Suet. Tib. 16). They supplied the whole of the cavalry and a large contingent of infantry. The latter (cohortes auxiliariae, sociae, Tac. Ann. 1.49, 51) partly were equipped in Roman fashion, partly retained their native weapons, and as such are styled sagittarii, funditores, scutati, contarii, catafracti (see Index to C. I. L. iii. p. 1148 ff.), or collectively leves cohortes or ferentarius miles (Tac. Ann. 1.51; 12.35). The latter passage contrasts the swords and pila of the legionaries with the spathae (broad two-edged swords without a point) and hastae of the auxiliaries.

The regular auxiliary cohorts were divided into cohortes quingenariae of 480 men, each in 6 centuries (Hyg. § 28); or cohortes milliariae of 1,000 men, each in 10 centuries. Some of these cohorts were exclusively infantry, and such were called peditatae: but from the time of Vespasian at all events (Tac. Hist. 4.19) it was found necessary, in order to enable the cohort to be an independent body capable of acting efficiently in garrison duty on the frontiers, to join a certain number of cavalry to

Roman Soldier. (From Iwan Müller's

Handbuch,

vol. iv. p. 745.)

some cohorts. These were called cohortes equitatae. a cohors quingenaria equitata would consist of 360 infantry in 6 centuries of 60 each, and 120 cavalry in 6 turmae of 20 each (Hyg. § 27; C. I. G. 5053): a cohors milliaria equitata of 760 infantry in 10 centuries of 76 each, and 240 cavalry in 10 turmae of 24 each (cf. Mommsen in Eph. Epigr. v. p. 31).

The cavalry were further divided into alae milliariae of 960 men each, in 24 turmae of 40 each, and alae quingenariae of 480 men each, in 16 turmae of 30 men each. Josephus (B. J. 3.5, 5) tells us that under Vespasian the cavalry wore helmets and cuirasses; carried a long sword, a long pole (κόντος ἐπιμήκης), and a buckler; and were furnished with a quiver, containing three or more javelins with large points and as broad as spears. Under Hadrian, as we learn from Arrian (Tact. 34), they had an iron visored gold-plaited helmet, with a plume of red horse-hair, a light shield, and instead of a cuirass a red Cimmerian tunic (i. e. made of Crimean leather). In the preceding cut we see many of the above accoutrements. The cuirass seems to be of leather. Note also the trousers (bracae) and boots.

Besides the names we have mentioned, which distinguished the cohorts according to their character and numbers (equitatae or peditatae, quingenariae or milliariae) and the alae according to their numbers, some, especially the alae, were also called after the provincial governor who first organised them: e. g. Auriana (Tac. Hist. 3.5), Petriana, Siliana (ib. 1.70), cohors Lepidiana (Diplomata xi. xii. in C. I. L. iii. pp. 854, 855), &c. (A full list in Marquardt, Staatsv. ii.2 473, note 5: also in Mommsen, Eph. Epigr. 5.246 f.) It is quite a mistake to suppose that they got their names from their commanders, as Borghesi has shown (4.192) in the case of the ala Frontoniana, which appears with many different commanders (cf. C. I. L. 3.788, 789, 793, 5331; Eph. Epigr. v. p. 175). Very commonly they are called after the emperor who organised them--Claudia, Flavia, Ulpia, Aelia (C. I. L. iii. Index, p. 1148 ff.). Sometimes, again, they were called after the country in which they were stationed or had distinguished themselves, e.g. Ala L. Flavia Augusta Britannica (Dipl. xxvi.), Cohors I. Lusitanorum Cyrenaica (Dipl. xx., xxii.). Also they got such honorary titles as Augusta, Victrix, Veterana, Pia Fidelis (C. I. L. iii. Index, l.c.). The cohorts were numbered progressively only so far as to distinguish cohorts otherwise undistinguished, e.g. Cohortes I., II., III., IV., Tungrorum milliaria equitata.

They were raised in the imperial provinces only (Mommsen in Hermes, 19.45), though we hear of a levy in Cyrenaica (Tac.> Ann. 14.18). They were seldom (Josephus, B. J. 2.13, 7) stationed in the same province as they were raised in, nor indeed ever left very long in any one province: see, e. g. the transferences of Cohors I. Thracum in the Diplomata (ix., xiv., xxxvi.). But as time went on it was found impossible to keep each cohort of one nationality; and very frequently vacancies were filled up by men of other nations, especially from natives of the province in which the division happened to be stationed, e. g. among the Equites Pannonii in Britain there was a decurio from Spain (Dipl. xxi.); in the cohortes Alpinorum and Lusitanorum in Pannonia, [p. 1.791]Pannonians are found (Dipl. xxxix., xlii., xii.).

There appear to have been special corps of auxiliary troops called equites and pedites singulares, the oldest mention of which is in the time of Vitellius (Tac. Hist. 4.70). They consisted of picked individuals from different nations, and as such differed from the ordinary cohorts and alae. One special division of cavalry set apart for the service of the emperor was called equites singulares Augusti (see below, p. 795 b). Sometimes the cohorts or alae had the additional title of Civium Romanorum (Tac. Hist. 3.47), which signifies either that the people from whom the corps was raised had citizenship previously, or that it was given to the whole corps--most probably the latter (Mommsen in Hermes, 19.60).

The commanders of the auxiliary cohorts and alae were for the most part Romans, and were called praefecti (Tac. Hist. 2.59; Dig. 3, 2, 2). The praefectura cohortis was the second and the praefectura alae the fourth step in the equestrian career of honours, the primipilatus being the first and the tribunatus legionis being the third (e. g. Wilmanns, 1249 b; Stat. Silv. 5.1, 94 if.). Some of the auxiliary cohorts, however, e. g. those equitum singularium (C. I. L. 6.224, 226, 228, Dipl. li.), and perhaps the cohortes milliariae, like the cohortes praetorianorum, vigilum, urbanae, voluntariorum (see below), were commanded by tribuni instead of praefecti, who stood on a level with the tribuni legionis: e.g. a tribunus cohortis becomes a praefectus alae without being tribunus legionis often (C. I. L. 3.1193, 9.5357, 10.3847). In later times the title tribunus came to be applied to the chief officers of all the cohorts (Boecking, Not. Dig. 2.536).

After the grant of Roman citizenship to all the provinces by Caracalla, the auxilia were almost entirely taken from Roman citizens. All that remained outside this category were a few barbarian mercenary troops. These were called nationes or numeri, and were arranged neither in cohorts nor in alae. They were commanded by praepositi. The introduction of them may perhaps go back to Trajan (Dipl. xxv.), and they formed the precursors of the Foederati. For the divisions of barbarian cavalry in the third century called cunei, see below, p. 808 a.

Special Extra-Legionary Troops.

1. The Cohors praetoria (σπεῖρα στρατηγίς) of the Republic²

This must have been an old institution, not so much owing to Livy's statement (2.21, 5) of its existence at the battle of Lake Regillus as from the title praetoria pointing to the time when the praetor was general. But it was not in accordance with the nature of the citizen army, and does not really appear in historical times till the Numantine war in the army of Scipio the Younger. He formed 500 of his friends and clients into a troop which Appian (App. Hisp. 84) calls an ἴλη φίλων and Festus (s. v. p. 223 M.) a cohors praetoria. They received once and a half the pay of the legionary. But the term ἴλη (turma) seems to point to their being cavalry, yet the regular cavalry received three times the pay of the legionary. Mommsen (l.c. p. 27, note 3) supposes that it was a privilege of the general to give certain members of this body-guard a. horse; that he did so mostly to his amici, who were men of good birth; and that it is of these Appian is chiefly thinking; while the notice about increased pay has reference to the mass of the troop who were his clients and served on foot. The term cohors can of course be applied to a mixed body of horse and foot. Caesar had only one praetorian cohort (B. G. 1.40, 15). Towards the end of the Republic each commander had a praetorian cohort (Sal. Cat. 60, 5; Cic. Fam. 15.4, 7, cf. 10.30; App. BC 4.7). After the battle of Philippi the various praetorian cohorts had in all 8,000 men (ib. 5.3). Later each of the triumvirs had several (ib. 5.24; Plut. Ant. 39, 53): at the battle of Actium, Octavian had at least five praetorian cohorts (Oros. 6.19). These cohortes praetoriae were soldiers picked from veterans and from the equites extraordinarii; they were the successors of the clients of Scipio, and had got more and more separated from the body of friends, who gradually cease to be soldiers. The two classes are strikingly opposed under Caligula (Suet. Cal. 19).

2. Cohortes civium Romanorum.

When Italians came to be practically excluded from legionary service and confined to serve in the city troops (see below, p. 806 a), such of them as wished to make a profession of arms entered the auxiliary cohorts as volunteers, the more readily as service in the cohorts was lighter than in the legions and rewards quicker (Veg. 2.3). Hence the origin of these troops which appear under many names: e.g. Cohors I. Italica civium Romanorum voluntariorum (Henzen, 6709); Cohors II. Civium Romanorum (C. I. L. 2.4114); Cohors I. Civium Romanorum ingenuorum (ib. 5.3936), and a vast number of titles with the word voluntariorum occurring in them in Mommsen's index of cohorts in Eph. Epigr. 5.248-9. That scholar supposes (Res gestae d. Aug. p. 72, note 1) that the cohortes voluntariorum originated from the freedmen Augustus enrolled (Vell. 2.113; D. C. 55.31; Suet. Aug. 25; Macr. 1.11, 33, “Caesar Augustus in Germania et Illyrico cohortes libertinorum complures legit quas voluntarias appellavit” ); but that they were not recruited afterwards from freedmen, for we find recruits of foreign extraction admitted to them, just as to the auxiliary cohorts, to whom they are further assimilated in having to serve 25 years (C. I. L. iii. p. 907). An inscription (C. I. L. 9.5835) shows us that there was a thirty-second cohort of these voluntarii. The centurion Cornelius in the Acts of the Apostles (10.1) belonged to one of these cohorts (ἐκ σπείρης τῆς καλουμένης Ἰταλικῆς). Whether the Ala I. Civium Romanorum which recurs so often (Dipl. xi., xii., xxv.) was a similar institution is uncertain. Some of the cohortes civium Romanorum at any rate, e. g. I., II., had cavalry attached to them (C. I. L. 6.3520; cf. Arrian, ἔκταξις, 100.9, 13).

3. The Evocati (ἀνάκλητοι）

(Schmidt in Hermes, 14.321-353).--The ancients (especially Serv. on Aen. 8.1) distinguished three kinds of service: (1) legitima militia or sacramentum, (2) tumultus, (3) evocatio (see p. 809 b); but logically, as Mommsen has pointed out (Eph. [p. 1.792]Epigr. 5.142), the division should be state-ordered service, comprising sacramentum and tumuitus, and voluntary service. Those who served in the latter were the evocati. It was originally only in periods of great crisis (Serv. on Aen. 7.614) that such invitation was made, not by a magistrate, but by any man of spirit and influence, who called on those who wished their country's safety to follow him. According to strict legal right, these volunteers were not soldiers, but pro militibus (Serv. on Aen. 2.157); they did not serve in a legion or a cohort, they had no definite leaders, nor had they any right to demand pay, though probably they always received rewards in larger measure than the ordinary soldiers; certainly they did in later times (Caes. B.C. 1.3, 2; cf. Mommsen, l.c. 143, note 1). But afterwards it became the practice for generals, no longer in the name of the state but in their own name, to invite, as a rule specially (nominatim, Caes. Gal. 3.20, 2), veterans (B.C. 1.3, 2) to renew their service. These evocati stood in rank above the ordinary soldiers, probably on a level with the centurions (Caes. B.C. 1.17, 4; 3.53, 1 ; 91, 1; Vell. 2.70, 3, compared with D. C. 47.46, 4); were perhaps only used in battle, and freed from all ordinary duties (Marquardt, Staatsv. ii.2 387, note 6); they had horses on the march (Caes. Gal. 7.65, 5). We find such troops as these in the time of Flamininus (Plut. Flam. 3; Liv. 32.3, 3), Marius, Catiline (Sal. Jug. 84, 2; Cat. 59, 3), Cicero (Cic. Fam. 15.4, 3), Caesar (ll. cc.; cf. C. I. L. 10.3886, 6011), Octavian (D. C. 45.12, 3). They played a considerable part in the civil wars, but seldom appear under the Empire, it not being consonant with the order of the standing army to have forces which could not be formed into definite troops (Mommsen, op. cit. 144). When they do appear, the invitation was of course made no longer by a private individual, but by the emperor (Tac. Hist. 2.82). Those who responded to the invitation were sometimes called revocati (Orelli, 3580). But if they practically disappear under the Empire, a new body was formed, viz. the evocati Augusti.

4. The Evocati Augusti³

These were a special corps (σύστημα ἴδιον, D. C. 55.24, 8) established by Augustus which continued till Christian times (C. I. L. 6.2870). The evocati which he called out after the death of Caesar (Dio Cass. l.c.) gave the name; and the features of voluntary service and of not being formed into regular troops, i. e. cohorts or legions (App. BC 3.40), which these evocati possessed, in some measure formed the transition to the evocati Augusti properly so called. But the latter differed essentially from the former, who were rather like the regular evocati described in the last section, in the following respects:--(1) The evocati Augusti had no fixed time of service; the regular evocati were released after the crisis which called them out was passed. (2) The evocati Augusti were an ordinary and continuous branch of the service: the regular evocati extraordinary and temporary. (3) The evocati Augusti had civil rather than military functions, e.g. ab actis fori (C. I. L. 9.5839), a quaestionibus praef. praet. (ib. 6.2755), architectus (ib. 6.2755), librator, “a surveyor” (ib. 6.2545), agrimensor (Grom. Vet. p. 121, ed. Lachm.). Such functions as these they had already performed while in the army; and the main cause of the establishment by Augustus of this special body of evocati was to get men who had been specially trained in definite branches of state service which were usually administered by soldiers to remain, after they were entitled to discharge, in the performance of their duty outside the regular army. This he effected by giving them increased rank (for they stood next the centurions above the principales, C. I. L. 6.1009), increased pay (probably nearly that of a centurion), and the right of carrying a vitis (Dio Cass. l.c.). They were often employed on special duties; thus we find one appointed to guard Vonones (Tac. Ann. 2.68). Further proof that they were in a measure civilians is that their pay was called salarium, not stipendium; that militia in caliga is always (e.g. C. I. L. 9.5840) opposed to evocatio, not to centurionatus; and that they never receive the military distinctions of torques, armillae, phalerae, but only a crown (Mommsen, l.c. p. 152). Nor did the evocatus in legione (less correctly legionis) belong to the legion in the same sense as the soldier or centurion. He acts for the legion, e. g. caters for it (C. I. L. 6.2893), but does not belong to it. The privilege of invitation to this service, which was coveted so as to avoid becoming a veteranus and being discharged, appears to have belonged practically, if not of right, to the city soldiery, and especially to the praetorians: for of all the evocati mentioned in inscriptions only three belong to the urban cohorts, one to the fleet of Misenum, the rest to the praetorians. They were probably subject to the praefectus praetorio (C. I. L. 6.3445). The body-guard of equestrian youths formed by Galba, whom he called evocati, was quite special and temporary (Suet. Galb. 10).

5. Vexillarii⁴

This word has two meanings: (a) one who bears a vexillum (=vexillifer), and this is its general use when it is added to a name to mark a particular branch of duty, e.g. C. I. L. 3.4834; (β) one who serves under a vexillum, and such is its meaning when used generally and in the plural. Now a vexillum and a signum differ in that the former is temporary and extraordinary, the latter is fixed and regular. Hence different classes of those who serve under vexilla. (1) Those veterans who had served out their time of twenty years, and who for one reason or another, financial or military, were not yet provided for. These were nominally dismissed (exauctorati, Tac. Ann. 1.36), but remained under a vexillum (ib. 26). They were treated as a select troop, used only in battle, and were free from all other duties (ib. 36). These troops of veterans are usually called vexilla veteranorum (Tac. Ann. 3.21; C. I. L. 5.4903). (2) Any troop separated from the main body under a special commander had its own vexillum (Caes. Gal. 6.36, 3; 40, 4), was called vexillum or vexillatio (the latter term only in Inscriptions), and its members vexillarii. Of this class, besides the quite general expressions vexilla equitum, tironum (Tac. Hist. 2.11; Ann. 2.78), we find vexillationes of the legions used for making roads, bridges, fortifications, &c. (Tac. Ann. 1.20; C. I. L. 3.1979, 1980, 3200), as outposts [p. 1.793]through the provinces--such are doubtless the multae et diversae stationes Hadrian refers to in his speech to his soldiers at Lambaesis (C. I. L. 8.2532 A, b; cf. Eph. Epigr. 4.528)--or as detachments dispatched to the theatre of war; e.g. C. I. L. 10.5829: T. Pontius Sabinus . . . praepositus vexillationibus milliariis tribus expeditione Britannica (i.e. Hadrian's expedition), Leg. VII. Gem. (stationed in Spain), VIII. Aug., XXII. Primigen. (both in Germany). This, as well as the inscription in Bull. dell' Instit. 1868, p. 60, praef. vexillation. eq. Moesiae infer. et Daciae, shows that a combined band of vexillarii need not necessarily be taken from the one province (cf. C. I. L. 2.3272, yet cf. Mommsen ad loc.), though usually such was the case (Tac. Hist. 2.100, 3.22; Wilmanns, 1429). Vexillarii of the auxiliary cohorts appear also (Tac. Hist. 3.6). The numbers of a vexillatio, though often 1000 (C. I. L. 8.2482, 10.5829), varied: hence the different rank of the commander: for he is sometimes a praepositus of equestrian rank, sometimes a centurion (Eph. Epigr. 4.524), or a tribunus legionis (Henzen, 6453), or a legatus Augusti (C. I. L. 8.7050), sometimes a dux of senatorial rank. (This term dux generally signifies an active military command of an unusual nature, as when a subaltern officer holds command of a legion: cf. C. I. L. 2.4114, and Mommsen in Eph. Epigr. 1.135, note 2, and in the appendix to Sallet's Die Fürsten von Palmyra, pp. 72-75.) From the time of the Gordians, at any rate (C. I. L. 8.2716), vexillatio takes an entirely new signification, viz. that of a troop of cavalry (ib. 3.405, 8.9045, 255 A.D.), and in this sense it is used in the Code (7.64, 9; 10.54, 3): cf. Veg. 2.1; 3.4, 10; Kuhn, Die Verfassung und Verwaltung des röm. Reichs, 1.133-5.

6. Frumentarii⁵

There was probably a numerus (a very vague term for “troop” under a single commander) of frumentarii in each legion (C. I. L. 6.3341), if that inscription does not refer to the special numerus at Rome. The numerus was commanded by a centurion, who is called in inscriptions centurio frumentarius (e.g. C. I. L. 8.2825). They appear in all the provinces which had legions, and even in the inermes provinciae (Marquardt, Staatsv. ii.2 492-3). Their functions were partly connected with the corn-supply (C. I. L. 6.3340), and as such their inscriptions are found all along the Appian Way to Puteoli (ib. 10.1171, 6095, 6575), and nowhere else in South Italy. Their chief station was in Rome, but there was probably a division at Ostia also (Marquardt, l.c.). Partly they acted as letter-carriers (Capitol. Max. et Balb. 10, 3), partly too as spies (Spart. Hadr. 11, 4), and in later times as police (Euseb. Hist. Eccl. 6.40), a function perhaps confined to the provinces. They appear to have been instituted by Hadrian: for we first hear of them under that emperor, and no frumentarii are found to belong to a legion which ceased to exist before Hadrian's time.

With regard to the numerus frumentariorum in the city, they nearly always state the legion to which they belonged (C. I. L. 6.3331-3361), a fact which shows that, though practically free from the legion, yet they were technically recognised as belonging to one. Confirmatory evidence on this point is that inscriptions of frumentarii have been found in places where the legion to which they are stated to have belonged never lay, e. g. in Lambaesis of a frumentarius of V. Macedonica, which belonged to Upper Moesia (ib. 8.2867). In the post-Severian times these frumentarii had a place in the castra peregrina in the Second Region, near Mount Caelius. For further particulars see Henzen, op. cit.

The Garrison of the City under the Empire.

1. The Cohortes Praetoriae (οἱ δορυφόροι).

These were what the cohors praetoria of the Republic developed into under the military monarchy. The praetorium was now wherever the emperor was. We have seen above that there were several praetorian cohorts at the end of the Republic (p. 791 a). Augustus established nine praetorian cohorts (Tac. Ann. 4.5): three of these he quartered in different parts of Rome, keeping one to act as guard (Tac. Ann. 1.7), and the rest he scattered about Italy in places where he was wont to stay himself (Suet. Aug. 49; Tib. 37). It was a great stroke of Sejanus concentrating the praetorian cohorts into one camp before the Viminal gate; and the vast power this concentration threw into their hands is duly insisted on by historians (Tac. Ann. 4.2; Suet. Tib. 37; D. C. 57.19, 6). The supreme commander of these was of course the emperor (in Tac. Ann. 1.7 he gives the watchword); but from the year 2 B.C. (D. C. 4.10, 10) the praetorians were commanded in the name of the emperor by two praefecti praetorio, sometimes by one, after the time of Commodus by three (Mommsen, Staatsrecht, ii.2 828). For a list of these praefecti praetorio up to Diocletian, see Hirschfeld, Verwaltungsgeschichte, p. 219 ff.

The number of praetorian cohorts was raised probably by Claudius to 12, for we find the xi. and xii. praetorian cohort mentioned in several inscriptions (C. I. L. 6.2762-2768). Mommsen (Hermes, 16.644) attributes this increase of the number of cohorts to Claudius, because it certainly is found in existence in Nero's time (C. I. L. 5.7003), is not mentioned by Tacitus, and Claudius certainly raised the number of the cohortes urbanae (Suet. Cl. 25). Vitellius raised the number to 16 (Tac. Hist. 2.93); Vespasian reduced it to 9 (Diploma x.). But before 130 A.D. (C. I. L. 6.208) a tenth must have been added, and this number certainly subsisted till the third century (Diploma lvii., 298 A.D.).

Each cohort was a cohors milliaria (D. C. 55.10, 10; Tac. Hist. 2.93) equitata, each century having a turma of cavalry (Tac. Ann. 1.24, 12.56; C. I. L. 6.100). We may possibly conjecture that, like the cohorts of the legionaries and auxiliaries, each praetorian cohort was divided into 10 centuries and 10 turmae; but such an analogy is not very reliable, for we know that each cohort of the Vigiles, though a cohors milliaria, had only seven centuries (cf. Mommsen in Eph. Epigr. 4.241). Each cohort was commanded by a tribune (e. g. Wilm. 1639). The praetorians were recruited from volunteers coming principally from Italy or the civilised provinces of Spain, Macedonia, or Noricum [p. 1.794]（Tac. Ann. 4.5; D. C. 74.2, 4); and this is fully attested by inscriptions--viz. C. I. L. 6.3885 (147 A.D.); 2381 a, b, c (153-156 A.D.); 2382 a, b (175-178 A.D.): cf. Mommsen in Hermes, 4.118. From the time of Septimius Severus, however, this was changed: the praetorians were recruited from the most tried of the barbarian troops (Dio Cass. l.c.), as is attested by the names found in inscriptions, e.g. C. I. L. 6.2385. On the nationalities and native places of the praetorians, see the indices of Bormann and Henzen in Eph. Epigr. 4.322-326; Oscar Bohn, Ueber die Heimath der Praetorianer, 1883, and in Eph. Epigr. 5.250-258; Mommsen in Hermes, 19.52. The length of service in the praetorian cohorts was 16 years (D. C. 4.23, 1; Tac. Ann. 1.17); the pay from the time of Tiberius, 720 denarii yearly (see below, p. 809 a).

A word must here be said on the officer called trecenarius, who often appears (e.g. C. I. L. 2.4461, 3.454. Complete list in Eph. Epigr. iv. pp. 242-3). In early times the chief centurion of the first cohort of the praetorians was called primus ordo (C. I. L. 10.4872); but, owing to the excessive danger to the state of such a special position among a body like the praetorians, it was arranged that the chief centurion of the first cohort should not hold superior rank to the others. The chief centurions of the different cohorts appear to have been called trecenarii, probably because they commanded 300 men, certainly not because the pay was HS. 300,000, which would be three times that of a tribunus militum; and the trecenarius ranked on a level with the second centurion of the legion primus praetorii (cf. Henzen, 6767, 6877), inferior to the primus pilus (ib. 6771 and passim). Just as the primus pilus was the highest position a centurion could gain in the legionary force, so the trecenarius was the highest he could gain in the praetorian; and, if promoted, he had to pass to another branch of the army, viz. the legions. Hence the title ex trecenario or ex CCC (Orelli, 3444, 3457; C. I. L. 3.3427). The principes castrorum among the praetorians (Orelli, 3457; C. I. L. 6.216, 10.5064) seem to have been two centurions selected from the trecenarii, who, like the principes in the legion (Veg. 2.8), had the general oversight of the camp.

2. The Cohortes Urbanae, οἱ ἀστικοί

(O. Eichhorst, De cohortibus urbanis imperatorum Romanorum, 1865: cf. also C. I. L. 6.2861-2948).--The praetorians and the urban cohorts, both in the lists of veterans and in the Diplomata (e. g. x., xlvii.)--see Bormann in Eph. Epigr. 4.317, 318--appear to form one closely connected body, though the urban cohorts were inferior in rank, only getting half the pay of the praetorians (see below, p. 809 a), and having to serve twenty years (Dig. 27, 1, 8, 9). Soldiers are advanced from the legions to the urban cohorts, and thence to the praetorians (e.g. C. I. L. 2.4461). They were established by Augustus in custodiam urbis (Suet. Aug. 49; D. C. 55.24, 6), put under the authority of the praefectus urbis (Tac. Hist. 3.64), and perhaps appointed at the same time as that officer, viz. 24 A.D. (Eichhorst, p. 3). Each cohort had originally 1500 men (Dio Cass. l.c.), in Vitellius's time 1000 (Tac. Hist. 2.93), and was commanded by a tribunus (Tac. Ann. 6.9; Eichhorst, Nos. 5-25). Each cohort consisted probably of 10 centuries commanded by centurions (Eichhorst, Nos. 26-76 and p. 7). The number of the praetorian cohorts seems to have been four in the time of Tiberius, three for duty in Rome (Tac. Ann. 4.5) and one for service at Lyons (ib. 3.41). After the model of this latter Claudius (Mommsen in Hermes, 16.645) added two at Puteoli and Ostia to guard against fires (Suet. Cl. 25); Vitellius reduced them to four (Tac. Hist. 2.93), while Vespasian increased them to five by creating Cohors I. Urbana for service abroad. In Caracalla's time they were four (Dipl. xlix.), and were still in existence at the beginning of the fourth century (C. I. L. 6.1156). This inscription shows that at this time they were located in the Forum Suarium in the Seventh Region (Preller, Die Regionen, p. 98; Mommsen, Staatsrecht, ii.2 1021, note 2).

The history of the urban cohorts is perplexing. It must be remembered that they formed a closely connected body with the praetorians, the numbers of the urban cohorts continuing those of the praetorians. Thus under Tiberius the urban cohorts are numbered X., XI., XII., XIII., the last, viz. XIII., being used for service at Lugdunum (Boissieu, Inscr. antiques de Lyon, pp. 353-361). When Claudius raised the praetorian cohorts to 12 and the urban to 6, these latter were probably numbered XIII.-XVIII. We do find about this time XV. (C. I. L. 10.1765), XVI. (Henzen, 6767, though this is not strong evidence), XVII. at Ostia (Tac. Hist. 1.83), at Lyons (Inscript. of Moulins in Hirschfeld, Lyon in der Römerzeit, p. 27), XVIII. (Tac. Hist. 1.64). For the possible transferences of these cohorts, see Mommsen in Hermes, 16.646. Vespasian created the Cohors I. Urbana and sent it to Lugdunum, sending XIII. to Carthage. In the second and third centuries XIII. appears at. Lugdunum and I. in Africa (Mommsen in Eph. Epiqr. 5.119, 120). The sending of these select troops to Africa was probably for fiscal purposes, as Africa from Nero's time became highly important in the matter of taxes. Under Antoninus Pius there appear in Rome cohortes urbanae X., XII., XIIII. (C. I. L. 6.1009); and note that this inscription proves the simultaneous existence of Coh. X. Urb. and Coh. X. Praet. In Caracalla's. time (Dipl. xlix.), X., XI., XII., XIIII. are the cohortes urbanae. After 216 Cohort XIII. disappears; perhaps it was abolished by Severus when he conquered Albinus at Lugdunum (Renier, Rév. Arch. 1864, p. 214).

3. The Statores (
C. I. L.
6.2949-2958).

These appear in the Republic among the apparitors of the provincial governor, on a level with the lictors (Cic. Fam. 2.1. 9, 2; 17, 1; 10.21, 2). Under the Empire they seem to have been entirely confined to the special service of the emperor, and so often called Statores Augusti (C. I. L. 6.2949). They took rank between the cohortes vigilum and urbanae, as one can see from the lists of promotion (e. g. Wilmann's, 1598, 1617), and stood in close connexion with the praetorians (Hyg. § 19; cf. C. I. L. 6.1009). They formed a numerus statorum praetorianorum (ib. 10.1766), consisting of several centuries (ll. cc.). Their commander appears to have been called praefectus (Eph. Epigr. ii. pp. 291, No. 339) or Curator (Herzog, Gall. Narb. No. 876).

[p. 1.795]

4. The Cohortes Vigilum, οἱ νυκτοφύλακες

（C. I. L. 6.1056-1058, 2959-3091).--These cohorts were established by Augustus in 6 A.D. as a night-watch, both to prevent fires and to act as police (Mommsen, Staatsrecht, ii.2 1008-1012, where a full account of the night-watch of the Republic is given). They were part of the regular army (Dig. 37, 13, 1), as promotion was made from them to the urban and praetorian cohorts and the legions (Orelli, 3444; C. I. L. 6.2780), but still we find them of a somewhat lower rank than the other corps of soldiers; for their functions required a variety of un-military appliances for rendering aid in case of fire--viz. centones, siphones, scalae, perticae--and on this account they appear to have been nicknamed Sparteoli (Du Cange, s.v. Tert. Apol. 39: cf. Mayor on Juv. 14.305); and we find they had assistants with duties essentially those of police, e.g. carcerarius and quaestionarius (see below, pp. 802 a, 803 b). The sebaciarii, of whom so frequent mention is made in the inscriptions on the walls of the excubitorium of Coh. VII. (C. I. L. 6.2998-3091), were men who supplied by the month (ib. 3062) tallow lights (faces sebacias) to the vigiles for use in their patrols. Their standards were the extraordinary vexilla, not the ordinary signa (Eph. Epigr. 4.356, 370), probably owing to their duties not being altogether strictly military. Accordingly Tacitus passes them over in his enumeration (Ann. 4.5).

They consisted of freedmen (D. C. 55.26, 5; Strabo v. p.235; Suet. Aug. 25). According to the Lex Visellia (24 A.D.), a freedman got citizenship if he served six years in the vigiles; a later senatusconsultum shortening the period to three years (Ulpian, fr. 3, 5; Gaius, 1.32a, ed. Studemund), and qualifying them for further advance in the army. After Sept. Severus there were many freemen in the vigiles (Dio Cass. l.c.). The whole corps consisted of seven cohorts of about 1000 men each (in C. I. L. 6.1057 the total in the cohort is 924: ib. 1058 it is 960). Each cohort had two regions of the city to look after (Dig. 1, 15, 3, pr.), and in each region a watchhouse (excubitorium). Regions 1 and 2 fell to Cohort V., 3 and 5 to II., 4 and 6 to III., 7 and 8 to I., 9 and 14 to VII., 10 and 11 to VI., 12 and 13 to IV. (Jordan, Topographie der Stadt Rom, 1.1, 307; cf. ii. p. 573). There were probably equites among the vigiles, though Mommsen on C. I. L. 6.3045 thinks not. Each cohort fell into 7 centuries, averaging about 135 men each (ib. 6.1056-1059), under centurions (Mommsen, Staatsrecht, ii.2 1009, note 2), and was commanded by a tribunus (C. I. L. 6.1599). The whole corps was under the praefectus vigilum ὁ ἔπαρχος ὁ νυκτοφυλακῶν (D. C. 52.33), who was of equestrian rank, not a magistrate, but only appointed extra ordinem utilitatis causa (Dig. 1, 2, 33). The only higher equestrian positions were the praefecti Aegypti and praetorio. In early times his jurisdiction was very restricted, as all serious cases came before the praefectus urbi. But under Sept. Severus the jurisdiction of the equestrian praefectus vigilum was extended at the expense of that of the senatorial praefectus urbi. For further details as to his police jurisdiction, see Dig. 1, 15, 3, 1; 47, 2, 57, 1; 47, 2, 18, 2; 12, 14, 5; 1, 15, 4: and Mommsen, Staatsrecht, ii.2 1011. For a list of praefecti vigilum, see Hirschfeld, Verwaltungsgeschichte, 1.145 if. Modestinus the jurist was a praefectus vigilum (C. I. L. 6.266). The praefectus had a deputy, a subpraefectus (ib. 414 b, 5.8660). These facts have been mostly elucidated from (1) two lists of Cohort V. treated by O. Kellermann, Vigilum Romanorum latercula duo Caelimontana, 1835; (2) discoveries made in 1858, which fixed the locality of four stations of the vigiles (De Rossi, Annali dell' Instituto, 265-297, 391-2); (3) in 1866 the discovery of the excubitorium of Coh. VII. and all its inscriptions (C. I. L. 6.2998 f.).

Non-Roman Troops in the Garrison.

1. Germani or Batavi.

This troop was instituted by Augustus. Previous to the battle of Actium he had a corps of Calagurritani as a body-guard, but afterwards a manus of Germans. These he disbanded after the defeat of Varus (Suet. Aug. 49), but they appear again under Tiberius (Tac. Ann. 1.24), Caligula (Suet. Cal. 43, 45, 55), and Nero (Tac. Ann. 13.18). They were taken from the German tribes belonging to the Empire, and their duty was the guarding of the imperial family. Thus we find them in the body-guard of Tiberius (C. I. L. 6.4339), Germanicus (ib. 4337), sons of Germanicus, Nero and Drusus (ib. 4337, 4342), Agrippina (Tac. Ann. 13.18; Suet. Nero 34), &c. We cannot argue from C. I. L. 6.4334 that they had cavalry among them. They were not divided into centuries and turmae, but formed a collegium. (ib. 8802-8809), which fell into decuriae just like the collegia and decuriae of slaves. For these Germans were legally slaves, though practically soldiers; and their being slaves is the mark which distinguishes these servi milites from the other foreign troops (cf. Mommsen, Schweizer Nachstudien, in Hermes, 16.459, note 1; also Hermes, 19.30, 31). They were disbanded by Galba (Suet. Galb. 12), and do not appear in Tacitus's time, as may be proved from his use of tum in Ann. 1.24. Nor do they appear in later times either; for the Γερμανοὶ and Βατάουοι mentioned in Herodian (4.13, 6) and Dio Cass. (55.24, 7; cf. Capitol. Maxim. et Balb. 14) are not this German body-guard at all--for none of the inscriptions referring to the German body-guard under that name are later than the Claudian era (Mommsen, l.c.), nor do they appear in the Hyginian camp--but are the equites singulares, who took their place.

2. The Equites Singulares Augusti

(Henzen, Sull' Equites Singolari, 1850): cf. C. I. L. 6.3173-3323.--This troop was instituted by either Trajan or Hadrian (C. I. L. 6.3309). Henzen places their origin in Domitian's time, but Tacitus (Tac. Ann. 1.24) does not mention them, and they are exactly the sort of institution we should expect from Hadrian (cf. H. Schiller, Kaiserzeit, 616, note 4). The inscription in Orelli (3525) which refers them to Augustus is utterly spurious. They are sometimes indeed styled equites singulares without Augusti or imperatoris added (Henzen, 26 ff.), but this is exceptional; and they must be carefully distinguished from the ordinary singulares of whom something has been said on p. 791 a. They appear originally to have been taken from the provinces on the Rhine or the Danube; [p. 1.796]none come from Italy or the Gallic or Spanish provinces. They formed a force complementary to the Praetorians, who were mostly raised from these more civilised provinces (Mommsen in Hermes, 19.54). They were connected with the auxiliary cavalry, but stood in a higher grade: thus we find frequent promotions from the auxiliary cavalry to the Equites singulares Augusti (C. I. L. 6.3199, 3234, 3238, &c.). If a soldier is promoted from the latter to the former, he becomes a decurio in the auxiliary horse (ib. 6.228, 1. 17). They ranked close to the Praetorians, as may be inferred from their position in the Hyginian camp ( § 23); but we cannot infer from C. I. L. 2.4147 that the Equites singulares Augusti were promoted to be centurions; for the omission of Eq. in that inscription renders its validity very questionable (Mommsen in Hermes, 16.462, note 2). They had probably two divisions, as they had two camps in the city, viz. the castra priora (C. I. L. 6.3183, &c.) and the castra nova (ib. 3217) or Severiana, which was built by Septimius Severus (cf. Herodian, 3.13, 4; cf. Dipl. i.). They left these camps only when the emperor took the field (Hyg. § 23). They seem to have been under the general command of the praefectus praetorio, each division having a tribune (C. I. L. 6.224, 226-8). We find, too, perhaps a praefectus (ib. 3261) and a praepositus (Wilm. 1639), but these were probably temporary officers. The Equites singulares Aug. wore a helmet without any plume; carried an oval shield, sword, and lance (ib. 3214, 3290; Henzen, op. cit. p. 50). The figures on these Inscriptions appear to show that they had slaves, and were therefore free themselves,--a fact to be noticed, as this is the main distinction between them and their predecessors, the Germani. They probably possessed Latin rights: for they certainly were not Roman citizens (Dipl. li.); nor again are their names in the inscriptions at all of the usual type of nomina peregrina, i.e. cognomen of son with cognomen of father (e. g. Adiatullus Vepotali f.), but nearly always they have the regular Italian three names or they omit the praenomen (C. I. L. 6.3309). It is to be noted, too, that the tribe is never added. For further particulars, see Mommsen in Hermes, 16.460 if.

The Provincial Militia.

Tacitus often (Hist. 1.11; 2.81, 83; 3.5) speaks of the inermes provinciae, by which he means those provinces in which no legions were stationed. These were (1) Senatorial provinces (except Africa), in which we learn that the proconsuls had some soldiers (Dig. 1, 16, 4, 1; 1, 16, 7, 1) separated from the main corps; but that they were very few Pliny's correspondence with Trajan shows. Pliny had one cohort, but from that he had to give 10 beneficiarii, 2 equites, and a centurion to the praefect of the coast, and more were asked for (Plin. Ep. 10.21 (32)). He asked for a century of legionaries for Juliopolis, but was refused (77 [81], 78 [82]). In fact he had so few that servi publici were guards of the prison (19 [30], 20 [31]), not as usual in other provinces, statores, who were military men. Trajan's reason seems to have been that as few soldiers as possible should be separated from the main corps (ne milites a signis absint). At certain points we hear of small military posts in senatorial provinces, e. g. at Eumenia in Greater Phrygia, where several high roads met (C. I. G. 3902 c). (2) Imperial provinces in which there was no legion stationed. But generally in such a province there was some small force: e. g. in Lyons we have seen there was a cohort of urban soldiery; in Dalmatia we find two cohorts at all events (C. I. L. iii. p. 282); in Pontus one cohort (Tac. Hist. 3.47). (3) Provinces governed by Procurators, e. g. the Cottian Alps (Suet. Tib. 37); Rhaetia, in which in 108 A.D. there were 4 alae and 11 cohorts=over 8,000 men (Dipl. xxiv.); Noricum, which had 4 alae and 14 cohorts (Dipl. lxx.; Eph. Epigr. 4.503); Thrace, which had 2,000 men (Joseph. Bell. Jud. 2.164); Judaea, which had 2 alae and 5 cohorts (ib. Ant. 19.2, 9); Sardinia (Dipl. xviii.). But they were so few in comparison to the extent of the country and so scattered that Aristides (i. p. 349) rightly says that the nations do not know where the forces are which keep them in subjection.

To make up for this deficiency of soldiers, there was a threefold kind of militia in the provinces:--

1. Municipal.

By the Lex Coloniae Genetivae (Urso in Baetica) we find it incumbent on the municipal magistrates to keep the fortifications in order, and for protection of their boundaries to appoint commanders to lead out armed forces (armatos educere), such commanders to have the same authority as the tribuni militum in the Roman army (Lex Col. Genet. xcviii., ciii.; Eph. Epigr. 2.110, 112); and we know also that colonies had often to defend themselves (Tac. Hist. 4.65; Amm. Marc. 25.9, 2).

2. Special Provincial.

E.g. (a) the Praefectus orae maritimae in Tarraconensis, who was a commander of coast-guards (Dig. 47, 9, 7） and had two cohorts under him (C. I. L. 2.4138),--cf. the tribunus militum cohortis maritimae in Baetica (ib. 2224); though the title cohort points to these being regular imperial troops; (b) in Noviodunum there was a praefectus arcendis latrociniis, which implies a force for the same purpose (Inscr. Helvet. 119); (c) various kinds of police officers and their men, e. g. εἰρήναρχοι (Dig. 48, 3, 6, pr.), διωγμῖται (Capit. M. Aurel. 21), κορυνηφόροι (Marq. Staatsv. 1.213). For more about these, see Kuhn, Verfassung, 1.43, 44.

3. General Provincial.

In times of danger or confusion the main body of youths in a province appears to have been sometimes utilised for military service. This was often the case against the Pirates in Sicily (Cic. Ver. 5.17, 43; 5.24, 60). In imperial times we hear of such troops of native youths in not completely civilised provinces, such as Cappadocia, Rhaetia, the Maritime Alps, and Noricum (Tac. Ann. 12.49; Hist. 1.68, 2.12, 3.5). We hear, too, of a people supporting at its own expense fortresses which probably commanded military roads, e. g. the Helvetii (Tac. Hist. 1.67). Prior to Hadrian's time such forces were only used locally; but later similar corps often appear among the imperial forces, still retaining their nationality (e.g. C. I. L. 7.1002). Tacitus (Ann. l.c.) calls this militia auxilia provincialium, Arrian (ἐκτ. chap. 7) τὸ συμμαχικόν, and [p. 1.797]Hyginus (chap. 43) symmacharii (according to Mommsen's emendation for the corrupt sumactares: see Hermes, 19.224). They ranked below even the classiarii (C. I. L. 8.2728) and were usually commanded by praepositi not of equestrian rank, appointed probably by the provincial governor: sometimes, however, the commander is styled praefectus or tribunus. Cf. Mommsen in Hermes, 19.228. For further details, see R. Cagnat, De municipalibus et provincialibus militiis in imperio Romano, 1880, and Mommsen in Hermes, 22.547-558.

The higher Officers of the Legions

(see especially Madvig, Die Befehlshaber und das Avancement in dem römischen Heere, in his Kleine philologischen Schriften, No. 10, pp. 477-560).

1. The Tribuni (χιλίαρχοι).

They were six for each legion, so twenty-four for the four legions, and were chosen originally by the consul. In 362 B.C. the people laid claim to elect six out of the twenty-four (Liv. 7.5, 9） in the comitia tributa (hence called by Asconius, p. 142, comitiati, but see Mommsen, Staatsrecht, ii.2 562, notes 1, 2); in 311 B.C. they elected sixteen (ib. 9.30, 3), and in 207 B.C. all twenty-four (ib. 27.36, 4). They were regular magistrates of the Roman people (see the Lex Acilia, C. I. L. 1.198, passim; Cic. Clu. 54, 148); they canvassed for the post (Sal. Jug. 63) and held office for a year (Cic. Att. 13.3. 3, 3), beginning on Jan. 1st (Cic. Ver. 1.10, 30). But the people never claimed to elect more than the twenty-four; so when the number of legions increased, the consuls had the appointment of all the rest. Those tribunes chosen by the consuls were, as opposed to those elected by the people, from the first called Rufuli (Liv. 7.5, 9), because, says Festus (p. 260 M.), Rutilius Rufus brought forward a law fixing their rights. Under Augustus, but apparently not later (Mommsen, Staatsr. ii.2 564; C. I. L. 10.788), we find the same distinction maintained in the tribuni militum a populo (often in Pompeian inscriptions, e.g. C. I. L. 10.820, 822, 830, 837, &c.) and the tribuni militum Augusti (ib. 2.3852). It was customary in Polybius's time to elect fourteen tribunes who had served five campaigns and ten who had served ten (Plb. 6.19, 1); and in the Punic and Macedonian wars tribunes of the soldiers appear who had held the consulship, praetorship, and aedileship; but as a general rule the tribunes were young men of rank: Scipio was one at the age of twenty (Liv. 22.53, 1), and Flamininus at that of eighteen (Plut. Flam. 1; cf. Madvig, p. 544, note 1). These young men began their duties either by serving in the cavalry or by being contubernales of the commander. After getting some little experience in this way, they either entered the civil service by becoming a vigintivir; or if they continued in the military service were appointed to the praefecture of a cohort of auxiliaries or to the tribunate of a legion--both of them equestrian positions (Caes. Gal. 3.7, 3, compared with 3.10, 2), the latter slightly but not much the more distinguished--and later to the praefecture of an ala of cavalry (C. I. L. 8.95, 2394). The arrangement of Claudius whereby the praefecture of the ala came before the tribunate of the legion (Suet. Cl. 25) was quite transitory. The tribunes wore the gold ring [ANULUS] of the equites (App. Pun. 104): if they were of senatorial birth, they were called laticlavii (Suet. Aug. 38, and the Inscriptions passim, e.g. C. I. L. 10.3722, C. I. G. 3990, χειλίαρχος πλατύσημος), those of equestrian birth angusticlavii (Suet. Otho 10). The three positions--viz. praefectura cohortis, tribunatus legionis, and praefectura alae--were the equestres militiae often mentioned (equestribus militiis functus, Plin. Ep. 7.25, 2; Veil. 2.111; 104, 3). In later times one who had held these positions was called a militiis, a tribus militiis, a iiii. militiis, ἀπὸ στρατειῶν ἱππικῶν, ἀπὸ τριῶν χιλιαρχῶν, &c. (Evidence in Marquardt, Staatsv. ii.2 367, note 8.) In order to get the rank of eques the tribunate was sometimes given for six months or a little over it (semestris tribunatus, Plin. Ep. 4.4, 2: cf. Juv. 7.88; C. I. L. 3.101, 9.4886). Such tribunes appear to have got a full year's salary, viz. 25,000 sesterces; see the Inscription of Torigny.

Under the Empire the tribunes still held a certain command in the legion (Hor. Sat. 1.6, 48; Tac. Hist. 3.9), but to all intents and purposes they were subject to the legatus, who, as we shall see, commanded both legions and auxiliary troops. Macer in the Digest (49, 16, 12, 2) enumerates several of the duties of the tribunes: viz. milites in castris continere, ad exercitationes producere (cf. Veg. 2.12), claves portarum suscipere, vigilias interdum circumire, frumentationibus commilitonum interesse, frumentum probare, mensorum fraudem coercere, delicta secundum suae auctoritatis modum castigare, principiis frequenter interesse, querellas commilitonum audire (cf. Tac. Ann. 1.44), valetudinarios inspicere. Besides these duties, some of which devolved on the praefectus castrorum during the early Empire, the tribunes commanded on the march (Lampr. Alex. Sev. 50, 2) and in battle (Plin. Nat. 22.11; Vopisc. Aur. 7, 1), took part in the council of war (Caes. Gal. 5.28, 3), acted as spokesmen for the soldiers (ib. 1.41, 8), kept the lists of the dead and living [by putting the latter Θ (θανὼν: cf. Pers. 4, 13; Mart. 7.37, 1) and Τ (τηρούμενος) or V (vivit) before the names on the roll (cf. Isidor. Orig. 1.24, 1; Marini, Atti de' Fratelli Arvali, pp. 165, 609)], granted the discharge to soldiers who had served out their time (Tac. Ann. 1.37), also granted leave of absence (Cod. Just. 12.35 (36), 13, 1), was answerable for the exercise and discipline of the soldiers in the camp (Capit. Max. 5, 5; Spart. Hadr. 10, 3). From the time of Claudius there were nominal tribunes, who did no duty, but they had only the title (Suet. Cl. 25); not like the tribuni vacantivi (Lampr. Alex. Sev. 15, 3), who appear to have got pay for doing nothing. In Ammianus (15.3, 10; 18.2, 2) tribuni vacantes are tribunes who had served out their time, but in case of emergency were invested with an extraordinary command (Godefroi ad C. Th. 6.18, 1).

2. The Legatus legionis.

On days of battle! (Caes. Gal. 1.52, 1), and when he used to go towards Italy to look after Roman politics (ib. 5.1, 1; 25, 5), Caesar was in the habit of appointing special commanders (legati) for the legions, one for each. This commander was [p. 1.798]retained by Augustus; and in imperial times, to distinguish him from other legati, was styled legatus legionis, or more strictly legatus Augusti legionis (see examples in Indices to C. I. L. passim, e. g. 8.7050, 7). He was of senatorial rank, usually an ex-praetor (Tac. Hist. 1.48; C. I. L. 10.6006; and the exception proving the rule in Tac. Ann. 2.36, 14.28). Previous to being praetors, they could only hold the title of pro legato (Tac. Ann. 15.28). The important distinction between the legati and the tribuni was that the former had command of the legion and the annexed auxilia (Tac. Hist. 1.57; cf. C. I. L. 8.2582, 2637), while the tribuni were strictly legionary officers. The office of legatus disappeared under Gallienus, its functions being undertaken by the praefecti legionum, who were no longer senators (Aur. Vict. Caes. 33, 34).

3. The Praefectus Castrorum (Wilmanns in
Eph. Epigr.
1.81-95).

Passing over the reference to στρατοπεδαρχία put into the mouth of Siccius Dentatus by Dionysius, 10.36 (where it may possibly be a word common people would use for a commander generally: cf. Lucian, de conscr. Hist. 21, 30), we only find the praefectus castrorum in the Empire, from Augustus to Severus. The locating of the soldiers in fixed camps caused its institution. These officers were created by the emperor (C. I. L. 10.4868), chiefly from among primipili (Orelli, 3509; Henz. 6758; C. I. L. 3.454, 2028), as being purely military men of non-senatorial rank; and as such they differed from the legati, who were senators, and from the more cultured youths who aspired to the militiae equestres. These praefecti were subject to the legatus of the province (Tac. Ann. 14.37), seem to have held the chief post after the legatus legionis, though they were not subject to him (Tac. Ann. 13.39; Hist. 3.14), and probably took his place if it was left vacant from any cause (Tac. Hist. 2.29). We find praefecti castrorum advanced to be praefecti fabrum and praefecti classis (e.g. C. I. L. 10.4868). The name of the legion was not added to his title during the first century; for each legion had not a camp, a camp being often formed of several legions (Tac. Ann. 1.20; Hist. 4.59), or even of vexillationes of separate legions (D. C. 55.8, who uses φρούραρχος to express praefectus castrorum). If several legions were drawn together for a campaign, the united force had only one praefectus castrorum (Vell. 2.112); whereas if it was divided, it had two (Tac. Hist. 2.26, 29; cf. 89). But from the time of Domitian (Suet. Dom. 7) each camp had only one legion, and so each legion had one praefectus castrorum: accordingly for this time it becomes customary for the praefectus castrorum to add the name of his legion to his title--e.g. C. I. L. 3.451, praef. Kastror. leg. xiii gem.: gradually the shortened title praefectus legionis becomes the usual one, and finally in the post-Severian time, when the senate was being deprived of all its privileges, the senatorial legati legionum give place to these praefecti. In Egypt, where no senators were permitted to enter without special permission (Tac. Ann. 2.59), the praefectus castrorum, even in Augustus's time, took the post held by the legatus legionis elsewhere (Joseph. B. J. 6.4, 3); and in the other provinces, just as Augustus sent his non-senatorial procurators to perform certain functions beside the legati of the provinces, so in the army he appointed non-senatorial praefecti castrorum, well-tried soldiers devoted to his interest, to be a counterbalance to the legati of the legions (Wilmanns, op. cit. 1.104). The duties of the praefectus castrorum are partly given by Vegetius, 2.10: “Totius positio valli et fossae aestimatio pertinebat. Tabernacula vel casae militum cum impedimentis omnibus nutu ipsius curabantur. Praeterea aegri contubernales et medici, a quibus curabantur, expensae etiam ad eius industriam pertinebant. Vehicula, sagmarii (‘pack-horses’) necnon etiam ferramenta a quibus materies secatur vel caeditur, quibusque aperiuntur fossae, contexitur vallum, aquaeductus, item ligna vel stramina, arietes onagri ballistae ceteraque genera tormentorum ne deessent aliquando procurabat.” He kept discipline in the camp (Tac. Ann. 1.20), and so had a certain jurisdiction, but could not punish capitally (ib. 1.38). He was often sent with detachments to make roads, bridges, or fortifications (ib. 12.38). During the battle he generally remained in the camp with the reserve (Tac. Hist. 2.26), though he sometimes leads out the reserve forces to battle (ib. Hist. 2.43; Ann. 13.39). Who performed his duty when he left the camp is not certain; perhaps one of the tribunes (cf. Dig. 49, 16, 12, 2).

4. The Centuriones, ἑκατοντάρχαι, ταξίαρχοι⁶

They first become important from the offensive order of battle instituted by Camillus. They were 60 in number, and were appointed by the consul, through the military tribunes, on grounds of merit (Liv. 42.53, 34, 35; Plb. 6.24). The qualifications required for a centurion are set forth in a weighty sentence of Polybius (6.24, 9): they are to be not so much daring and courters of danger as men fit to command and steadfast, the courage of whose souls is deep rather than superficial (βαθεῖς μᾶλλον ταῖς ψυχαῖς); nor should they engage before the battle is joined and commence the fray, but when they are being conquered and hard-pressed they should stand firm and be ready to die in defence of their post. They were arranged partly according as they belonged to the triarii principes or hastati; partly according to the maniple they were in within these ranks (thus secundus hastatus, Liv. 26.5); partly again according as they commanded the first or the second century of the maniple (prior or posterior); but there does not seem to have been any difference of rank between these latter. It is to be noticed that the triarii were called pilani (Varro, L. L. 5.89), and every ordo, whether in the sense of rank or centurion (see below), was called pilus: thus primipili centurio (Caes. Gal. 1.46: cf. primi ordinis centurio, Tac. Ann. 1.29), primum pilum ducere (ib. 5.33), primos pilos ademit (Suet. Cal. 44); and again in the sense of centurion, octavus pilus prior (C. I. L. 5.7004) and the common primus pilus. Thus the lowest centurion was called decimus ordo hastatus (Liv. 42.34), in which ordo = centurion (cf. primi ordines below), and notice that it was not thought necessary to add prior or posterior. There cannot have been any regular advancement till the institution of the standing army; but this [p. 1.799]advancement (successio, Orelli, 3454) developed itself during the Empire. On the introduction of the arrangement by cohorts, the meaning of the three divisions--hastati, principes, and triarii--disappeared, but the names were retained to signify the rank of the centurions, with the number of the cohort usually added. What exactly was the nature of that rank?

There are two main theories: (1) either Marquardt's view, that the first thirty centurions of the legions are priores (the first ten pill, the second principes, the third triarii) and the second thirty posteriores; or (at least for the period of the manipular arrangement) A. Müller's view (Die Rangordnung und das Avancement der Centurionen der römischen Legion, in Philologus, 38.136-148), that the first ten centurions are the pili priores, the next ten pili posteriores, the next principes priores, &c. (2) The theory of Mommsen (Eph. Epigr. 4.229, note 1) and that of A. Müller for the post-Marian organisation, that the rank of the centurion was fixed by the number of the cohort: thus the primus princeps was second in the legion, the primus hastatus third, the primus posterior fourth, and so on; the centurion of the first cohort ranking before those of the second, those of the second before those of the third, and so on. But first let us see who the primi ordines were, who so often appear (Caes. Gal. 6.7, 8; 5.30, 1); for it is in discussing these that Marquardt argues for his view of the rank of the centurions.

Marquardt says the primi ordines are the ten priores pili; for (1) the cohort surely must have had a commander, viz. the first of its six centurions, and these leaders of the cohorts must have held a position of distinction--an à priori argument which begs the question; and besides there were no commanders of the cohorts as such recognised in the legions, the only commanding officers being tribunes and centurions. (2) It is impossible to suppose that a centurion who had commanded a maniple or a cohort should have to sink to command a simple century, as must have been the case in (say) the step from primus hastatus prior to decimus princeps posterior. To which further à priori argument we reply that it is just as impossible that the second centurion of the first legion--which had, at least after Hadrian's time, a double number of men, had the charge of the eagle and a definite position of honour in the camp--should rank only twenty-first in the legion. (3) Vegetius (2.21) says that soldiers are promoted “ita ut ex prima cohorte ad gradum quempiam promotus vadat ad decimam cohortem et rursus ab ea crescentibus stipendiis cum maiore gradu per alias recurrat ad primam.” But this only means that a common soldier of the first cohort, if advanced to be a centurion, begins at the bottom of the centurions of the tenth cohort and works his way up. The other view, that of Mommsen (op. cit. 240), sees in the primi ordines the three chief centurions of the first cohort, viz. the primus pilus, primus prineeps prior, and primus hastatus prior--at least for the time subsequent to Hadrian, and it is certain that they were a definite class then, as may be seen from Hadrian's speech to the army at Lambaesis (C. I. L. 8.2532), “primi ordines et centuriones agiles et fortes more suo fuerunt,” and Hyginus (Grom. p. 176), “legiones cum signis et aquila et primis ordinibus et tribunis deducebantur.” They appear to have been also called ordinarii (C. I. L. 5.92, 8275, 7.404; Capitol. Alb. 11). But in classical times Mommsen considers that the primi ordines were no definite class, but simply the most distinguished centurions ( “qui loco et virtute prae ceteris eminerent” ), just as the superiores and inferiores ordines are opposed in Caesar (Caes. Gal. 6.40, 7; B.C. 1.46, 4). A similar view of the primi ordines is held by H. Bruncke (Die Rangordnung der Centurionen, 1884), who considers that the only difference in rank among the centurions was between these primi or superiores ordines and the rest the inferiores ordines (Caes. ll. cc.).

Now the reason for this view of Mommsen's will be further seen by reverting to the original question of the rank of the centurions. Suppose we hold with Marquardt, how are we to explain Caes. B.C. 1.46, 4: “Q. Fulginius ex primo hastato legionis xiv qui propter eximiam virtutem ex inferioribus ordinibus in eum locum pervenerat” ? He would only have been 21st in the legion. Again, what could Caesar (B.C. 3.53, 5) mean by “Scaeva centurio quem Caesar ab octavis ordinibus ad primipilum se traducere pronuntiavit” ? On Marquardt's theory those ordines would have been of too varied rank (8th, 18th, 28th) to be explicit; whereas it doubtless means from the eighth cohort to the first. Further, Mommsen's view will explain Vegetius, 2.8: “Vetus consuetudo tenuit ut ex primo principe legionis promoveretur centurio primipili;” cf. Henzen, 6747, “P. Aelius Marcellus . . . astatus et princeps et primipilus legionis vii geminae piae felicis.” Lastly, the great number of inscriptions marking the primi principes and primi hastati are additional proofs of the distinction of these positions. It is to be noticed that whereas in these three first centurions the cohort is seldom (Caes. B.C. 3.64, 4) added, it is universal in the centurions from the fourth on (see Mommsen's list, op. cit. p. 231 ff.). Mommsen considers a similar arrangement subsisted even before the introduction of the arrangement by cohorts: for the centurions who are mostly mentioned in this age besides the primus pilus are the primus princeps (Liv. 42.34, 8; 25.14, 7） and primus hastatus (ib. 27.14, 8), and why should these latter, the 11th and 21st in the legion on Marquardt's theory, be mentioned?--an argument which tells still more against A. Müiller's view that, during the time of the manipular arrangement, the first ten centurions are the pili priores, the next ten the pili posteriores, the next principes priores, and so on; for, according to this view, the primus princeps would be 21st, and the primus hastatus 41st, in the legion.

The following are a sufficient number taken from Mommsen's list (Eph. Epigr. 4.231-238) to show his arrangement of the order of the centurions.

1. Primus pilus. The term triarius prior is an invention of Vegetius, 2.8. The abbreviation P. P. is often found in Inscriptions, e.g. P.P. bis iterum (C. I. L. 5.867, 6.1637). This means that they held the primipilatus twice: so Tullius in Liv. 7.13, 1, who septimum primum pilum ducebat, was seven times primipilus (cf. Liv. xlii. [p. 1.800]34, 12). Occasionally the second primipilatus was taken after one of the militiae equestres had been held, so that on discharge he might get the honours of a primipilaris (Mommsen on (C. I. L. 5.867).

2. Primae cohortis princeps prior (Caes. B.C. 3.64, 4). Also called princeps pretorii in C. I. L. 3.5293, 2917; Henzen, 6767 (see Mommsen on 3.830), or simply princeps (ib. 8.2841, and passim). As he had considerable duties in seeing after the camp (Veg. 2.8), he had an adjutor (Wilm. 1556), librarius and optio (Henz. 6778). The latter is sometimes called optio pretorii (ib. 6940).
3. Hastatus primus (Caes. B.C. 1.46, 4; C. I. L. 2.4147) or simply hastatus (ib. 8.2825).
4. Cohorte i princeps posterior (ib. 2883).
5. Cohorte i hastatus posterior (ib.).
7. Cohorte ii princeps prior (Eph. Epigr. ii. p. 290).
8. Cohorte ii hastatus prior (ib. p. 287).
10. Cohorte ii princeps posterior (C. I. L. 6.3584).
12. Cohorte iii pilus prior (Eph. Epigr. v. p. 2).
13. Princeps tertius = cohorte iii princeps prior (C. I. L. 9.2770).
17. Cohorte iii hastatus posterior (ib. 3.1480).
21. Cohorte iv pilus posterior (ib. 6.3404).
28. Cohorte v princeps posterior (ib. 3.102).
42. Octavus pilus prior=Cohorte viii pilus prior (ib. 5.7004).

43. Octavus princeps (Cic. ad Brut. 1.8, 2).

50. Cohorte viiii hastatus prior (C. I. L. 8.2938).
53. Cohorte viiii hastatus posterior (ib. 2877).

All centurions carried the vitis, with which to inflict punishment on refractory soldiers (Plin. Nat. 14.19), and which was a synonym for the centurionship (Juv. 8.247). In the legion the primipilus was held in especial honour. He took part in the council of war (Plb. 6.24, 2). In republican times, and for the most part during the Empire, this was the last position in an ordinary soldier's career. When they had finished (consummaverunt, Suet. Cal. 44; διανύσαντες, Dig. 27, 1, 8, 12） their service, they retired into private life, and were then called primipilares. They frequently with large property repaired to municipal towns, where they became great local magnates (Cic. Fin. 1.3, 8; Hor. Sat. 1.6, 72). For inscriptions of such centurions, see Karbe, De centurionibus Romanis quaestiones epigraphicae, 1880, p. 338. They often made sufficient money to enable them to buy equestrian rank for themselves (Orelli, 3049) or for their sons (C. I. G. 2793): cf. Mart. 5.58, 10. If they wished to continue in the service, they were employed in important military duties (Tac. Ann. 2.11, 4.72, 13.36; Hist. 1.87, 4.15), sometimes as praefecti castrorum (see above, p. 798 a) or praefecti cohortium (Tac. Ann. 1.20): sometimes, too, they got positions in the civil service (Suet. Cal. 21 ; C. I. L. 5.698). But in imperial times the primipilatus became the first step in the equestrian career, leading to the tribunate of the cohorts of the city soldiery, i.e. vigiles, urbanae, praetoriae (ib. 867), occasionally to the tribunate of the legions (ib. 10.4868), and then to procuratorships (ib. 5.867), and even to the praefecture of the praetorians, e. g. Justus Catonius (Tac. Ann. 1.29; D. C. 60.18); cf. Hirschfeld, Verwaltungsgeschichte, 1.219-239, 247. Often the honorary title of centurion was given to young men of position by grant of the emperor (Juv. 14.193; Suet. Gramm. 24; D. C. 52.25, 6), who, after filling the three posts of the equestrian career--viz. praefectura cohortis, tribunatus legionis, praefectura alae (Wilm. 1259 b; Stat. Silv. v. 1, 94)--retired from the service with the title a mititiis (see above, p. 797 b). Both classes of aspirants, the ordinary centurions and the honorary ones, were called petitores militiae sc. equestris (C. I. L. 6.2485; Wilm. 1602). It is to be noticed that sometimes a young man of rank resigned that rank (intermissus eques, Stat. l.c.), and became a centurion in hopes of arriving at the lucrative and honoured positions, tribuneships and procuratorships, of the militia equestris--which hopes were not always fulfilled. (C. I. L. 3.1480).

In republican times and in the early Empire the number of centurions in the legion was 60 (Gel. 16.4, 6; Tac. Ann. 1.32). But after Hadrian's time it was 59 though the first cohort had double the number of men, yet it had only five centurions (Veg. 2.8, confirmed by C. I. L. 8.2555), while all the other cohorts had six. And in the list given above it will be seen that there is no pilus posterior in the first cohort. If 64 optiones occur in C. I. L. 8.2554, we must suppose that the adjutores, who as well as the optiones belonged to the centurions of the first cohort, are reckoned along with the optiones. In the first cohort, according to Vegetius 2.8 properly corrected, the primipilus commanded 400, the princeps 200, the hastatus primus and princeps posterior 150 each, and the hastatus posterior 100.

The Inferior Officers (principales).

For the title principales, see Veg. 2.7; Dig. 49, 16, 13,4; C. I. L. 6.221; Eph. Fpigr. 4.524. A most exhaustive treatise on these officers is that of P. Caver, De muneribus militaribus centurionatu inferioribus, in Eph. Epigr. 4.355 sqq.

Mommsen (Eph. Epigr. 4.532) divides the principales into two classes: (1) those belonging to a corps, and the commander under which these serve is never mentioned; (2) those belonging to a higher officer, and their title was not fully specified unless they had the name of the officer attached. This was partly because they were special attendants on certain officers (e. g. the stratores), partly because their rank was determined by the officer whose name they added, this especially in the case of the beneficiarii. Adopting this classification, we shall treat of the different principales of each class.

I. The Principales belonging to a corps.

1. The Standard-bearers.

a. vexillarius, Among the vigiles each century had a vexillarius ( = vexillifer), Orelli, 3480.
b. Among the praetorians each century had a signifer (C. I. L. 2.2610)--so that we may assume that the signifer was a higher functionary than the vexillarius. As each cohort had its signa [SIGNA], in cases [p. 1.801]where a signifer of a cohort is mentioned, e. g. (ib. 5.4371), it cannot be decided whether they performed this function in a century or a cohort.

c. Among the urbanae cohortes, just as among the praetorians, there were two kinds of signiferi, one belonging to the century, the other to the cohort. Mommsen thinks that the signifer of the first century was the signifer of the cohort (cf. Caes. Gal. 2.25). That Marquardt is quite wrong in supposing That the standard-bearer of the first cohort among the praetorians and the first cohort of the vigiles was called aquilifer, while the standard bearer of the century was signifer, is shown by the correct interpretation of “AQ praet. urbis” (ib. 6.2880) being a quaestionibus. Nor had the first cohort of the vigiles an aquilifer, as Marini and Marquardt (l.c.) assume on the strength of 6.1056: “AQ. Julius Julianus centuria Ingenui, AQ. Sittius Chryseros centuria Juvenis;” for (1) there is no sure evidence of an eagle outside the legions, (2) there would be two or more aquiliferi in one cohort, which is quite incredible.

d. Among the auxiliary cohorts a few vexillarii are found (ib. 3.3261), but it cannot be proved that they were vexilliferi. Also a few signiferi (Eph. Epigr. 3.106); and with regard to them it cannot be decided whether or not there were signiferi of other divisions than the cohort, though it is probable that each century had its signifer.
e. Among the equites singulares each turma had a signifer (ib. 6.3304).
f. Among the equites, each turma had a signifer (ib. 8.2094), also each ala (ib. 3.6274), just as the centuries and the cohorts. And as in their case, we cannot decide whether the signiferi of the alae were the signiferi of the first cohorts or not. Vexillarii alarum appear (ib. 3.4384), perhaps by an inaccuracy of language, the cause of which possibly may have been that the standard-bearers of the equites legionarii were called vexillarii to distinguish them from the signiferi of the cohorts and centuries--as may be inferred from 3.4061.
g. The numeri had signiferi (3.1396).

h. As to the legions there is nothing in the inscriptions to decide definitely whether the separate cohorts and centuries had signiferi or not. The signiferi are said to belong to the legions. Wherever vexillarii legionum occur, we can never be sure that they are vexilliferi. Most probably, as was the case with the praetorians and the auxiliary cohorts, the standard-bearers of the legions were not called vexillarii, either under the Empire or indeed at any time, for σημαιοφόρους in Plb. 6.24, 6=signiferos. From Veg. 2.20 we may infer that, besides the ten signiferi cohortiium, there were no other in the legions.

As to Aquiliferi, there was one for each legion (C. I. L. 2.266). We have shown that there were no aquiliferi among the praetorians and the vigiles. And as to Imaginiferi, “qui imperatorum imagines ferunt” (Veg. 2.7), there was probably one in each legion (8.2935) and in each auxiliary cohort (3.3256).

2. The Trumpeters

[see BUCINA, CORNU, TUBA]--Touching these, Vegetius 2.22 says: “Ergo quoties ad aliquod opus exituri sunt soli milites tubicines canunt: quoties movenda sunt signa, cornicines canunt: quoties autem pugnatur, et tubicines et cornicines pariter canunt. Classicum ita appellatur quod bucinatores per cornu dicunt. Hoc insigne videtur imperii, quia classicum canitur imperatore praesente vel cum in militem capitaliter animadvertitur.” The watches (vigiliae) were signalled by the cornicen. The term aeneator is probably a generic one, as Suet. Jul. 32 says that his tuba was taken from him, though Festus (p. 20) identifies him with the cornicen. The legionaries and auxiliaries, of course, had trumpeters. Of the urban soldiery, the praetorians had tubicines, cornicines, and bucinatores (6.2570, 2379, 2545), the vigiles bucinatores only (6.1057); in the cohortes urbanae the only trumpeter which appears is one tubicen (6.2404). The lituus, we are told (Ascon. in Hor. Carm. 1.1.25), was the trumpet of the cavalry and the tuba of the infantry, yet we find the equites singulares had a bucinator and a tubicen (6.3176, 3179). The inscription of Orelli, 3519, which mentions a liticen of a legion, is Ligorian.

3. Tesserarii

“qui tesseras per contubernia militum nuntiarent” (Veg. 2.7), were one in each century, as is proved principally by the latercula of the vigiles (6.1056). They also appear among the praetorians (6.2379 b), cohortes urbanae (9.1617), legions (6.2672), auxiliary cohorts (3.5046).

4. Curatores.

As the curatores fisci only appear in the praetorian (6.2375, 6, 27) and urban cohorts (8.4874), we may perhaps infer that the rest of the parts of the Roman army had not fisci. Accordingly the curatores turmarum which we find among the equites singulares (6.225), auxiliary cavalry (8.4510), auxiliary infantry (3.6025), are not to be explained as administrators of the military money. They are rather extraordinary commanders (cf. Mommsen in Archäol. Zeitung, 1869, p. 126), just as the commanders of veterans were in extraordinary cases called curatores (5.7005).

5. Custodes armorum.

They appear very often in the legions (e. g. 3.4275); also frequently in the auxiliary cavalry (3.5655), but still more frequently among the equites singulares (6.3177, 225, 228). The mysterious armatura which sometimes occurs (e. g. 8.2569 a, 17) is not to be explained with Borghesi as simply=miles, as miles armatura occurs in 6.2699. Mommsen in a letter to Hübner, published in the Jahrbuch des Vereins für Alterthumsfreunde im Rheinland, 68, p. 53, explains it as a thoroughly-drilled soldier ( “qui a campidoctore omni arte bellandi imbutus fuerit” ). The riding school (basilica equestris) had also a custos (Henz. 6811).

6. The Drill-masters: doctores and (a higner title) campidoctores or exercitatores.

They seem to have belonged to the cohort (6.533), and are found in all kinds of corps, the frumentarii (8.1322), equites singulares, where he is a centurion (6.228), speculatores (Wilm. 1617), equites praetoriani (6.2464).

7. Optiones

so called because “chosen” by the decurion or centurion as his assistant in private matters (rerum privatarum ministrum, Festus, p. 184; Varro, L. L. 5.81), also as his substitute in case of sickness or accident (Veg. 2.7). With regard to the optiones Mommsen's view appears right (Eph. Epigr. 4.449), that there were two classes of them, one belonging to the centuries and superior in rank, the other in a manner supernumerary and outside the centuries, [p. 1.802]taking care of the navalia. (Brambach, 1301), carcer, acta (C. I. L. 9.1617), equites (6.100), valetudinarii (6.175). Thus in C. I. L. 9.1617 optio alone is distinguished from optio of certain functions. We find optiones belonging to the centuries in the legions (3.3530), praetorians (6.2447), cohortes urbanae (8.4874), vigiles (10.3880). This view is better than that of Caver, who supposes the optiones of the special departments were optiones of the centuries, who were freed from all functions as such. In the latercula of the vigiles (C. I. L. 6.1056-1058), Kellermann, Henzen (6791), Wilmanns (1499), Marquardt (2.540) supply “optionem a(rcarii), optionem ca(rceris), optionem ba(lnearii), optionem c(ohortis), optionem co[h(ortis)] v.” But these functions are not elsewhere mentioned; optio carceris is expressed in other parts of the latercula by karc. or car.; and it is not at all plain what an optio cohortis can be.

In C. I. L. 10.135 the optio tribunorum quinque was that tribune of the soldiers who was in command of five vexillationes, which were sent on some expedition with their own tribunes, these latter probably electing the tribune in question to the chief command: cf. optio signiferorum in Brambach, 1048.

8. Speculatores

appear first in the Civil War (Caes. Bell. Hisp. 13; Bell. Afr. 37). During the Empire there were ten in each legion (C. I. L. 2.4122; 3.4452). They were often marked as belonging to cohorts (e. g. 6.2453), though in 5.2832, 5071 we find speculatores of centuries. They used to carry despatches (Tac. Hist. 2.73; cf. Liv. 31.24, 4), and sometimes act as executioners (Seneca, de Ira, 1.18, 4; St. Mark 6.27). They were numerous in the praetorian cohorts, and we find them in Tacitus (Tac. Hist. 2.11, 23) forming a separate corps, with separate officers tesserarius, optio (ib. 1.25), &c., and having the special function of acting as body-guard to the emperor (Suet. Cl. 25). The ninth Diploma (of Vespasian) separates (C. I. L. iii. p. 853) the praetorian speculatores from the other praetorian cohorts: and such was often the case, but not so after the time of Vespasian, when they were ingrafted among the praetorians. None of the inscriptions which show that the speculatores were allocated to the centuries and cohorts can be proved as old as Vespasian. The speculatores Augusti (ib. 3.5223) probably belonged to the praetorians.

9. Carcerarii

（harc. or car. in the latercula of the Vigiles) were prison-warders, who generally held also the functions of some other kind of principalis, e.g. beneficiarius (3.3412), optio (9.1617). There was also a clerk attached to the prison a commentariis custodiarum (Orelli, 3206).

10. Candidati

(strangely omitted by both Marquardt and Caver) are mentioned with other principales in C. I. L. 8.2568, 2569, 2618, 2866; 3.1190. What their function was we cannot say for certain. Mommsen (Eph. Epigr. 4.532, note 2) supposes that they were so called from a white robe they were in certain ceremonies (compare the white-robed priests of Jupiter Dolichenus, 6.406, 409). They appear to have been selected for their strength and stature (Chron. Pasch. ad ann. 243, 251).

11. Sacrificial Officers (haruspices and victimarii）

haruspex> (C. I. L. 8.2586, 57) and victimarius (Grut. 545, 4).

12. Artisans (architectus and librator）

Under this head may come the architectus (Orelli, 3489; C. I. L. 8.2850), and the librator who made aqueducts, such as Nonius Datus at Saldae, who tells us (ib. 8.2728 = Wilmanns, 785) how with the help of the classici milites and provincial soldiers (gaesates), perhaps. Raeti, he cut a tunnel through a mountain, from both ends meeting exactly in the middle (cf. Mommsen in Arch. Zeitung, 1871, p. 5. Very similar duties to those of the librator were performed by the canicularius (Henzen, 6785) and the aquilex (Dig. 50, 6, 7). This same section of the Digest mentions as attached to the army coppersmiths, iron-smiths, stone-cutters, roof-builders (scandularii), lime-burners, wood-cutters, charcoal-burners; makers of catapults, helmets, bows, arrows, javelins, trumpets, horns, chariots:, workers in leather and in lead. There were also the metatores, who measured out the camp (Cic. Phil. 11.5, 12), different from the mensores, who distributed corn in the legions (C. I. L. 8.2857) and praetorian cohorts (Henz. 6820; cf. Lyd. de Magistr. 4.46). If mensor does appear as a land-measurer in C. I. L. 3.586, he does not belong to the legions.

13. Medici.

The Army-doctors (Dig. 4, 6, 33, 2), who ranked among the principales (Eph. Epigr. 4.524), were an institution of the Empire. Under the Republic the chief officers. used to bring their private physicians with them on campaign (Plut. Cat. Mi. 70; Suet. Aug. 11),; and these used probably to attend what few wounded soldiers were lucky or unlucky enough to obtain their treatment (cf. Cic. Tusc. 2.16, 38); but we hear that often more soldiers died. of their wounds than fell in battle (Liv. 9.32, 12). But as early in the Empire as Augustus and the regular organisation of the standing army (Vell. 2.14; Tac. Ann. 1.65） attention began to be paid to the medical department. Every division had its medicus. There was the medicus legionis, perhaps several bearing that. title, in each legion, e. g. in Legion iii. Aug. (C. I. L. 8.2872, 2874, 2951); each praetorian cohort (Grut. 66, 1) had them, as had the urban. cohorts and those of the vigiles, perhaps four in each cohort (C. I. L. 6.1058, p. 201), the equites singulares Augsti (Orelli, 1576), cohorts of the auxiliary troops (ib. 3507; C. I. L. 7.690), vexillationes on garrison duty (Eph. Epigr. 4.530), &c. The doctors were immunes (Dig. 50, 6, 7), and sometimes duplicarii (i.e. getting double pay), especially in the fleet (C. I. L. 10.3443). During the first six centuries of the city there were no professional doctors at Rome (Plin. Nat. 29.12); but after Caesar gave citizenship to all professional physicians (Suet. Jul. 42), the medical art came to be held in respect by the Romans; and the recognised physicians of the legions, praetorians, and city cohorts were citizens. Doubtless, however, many, of their assistants were freedmen and slaves, and in the cohorts of the vigiles and auxiliaries freedmen and foreigners may have been the authorised physicians. Marquardt says that the assistants of the regular physicians are those designated as medici ordinarii (3.4279, 5959); but they are rather, as Mommsen (Eph. Epigr. 4.530, note 3) shows, those who served in the, numeri in contradistinction to those who served in the legions.

The hospital of the garrison was called valetudinarium: [p. 1.803]for three legions there was one; for five or six, several (Hyg. § § 35 and 4). It was technically under the superintendence of the praefectus castrorum (Veg. 2.10) or a tribunus militum (Dig. 49, 16, 12, 2); but the real officer was the optio valetudinarii, who often occurs (C. I. L. 8.2563, 9.1617), even in Rome (6.175). He had a number of male nurses under him (Dig. 50, 6, 7). There was also an infirmary for horses and mules (veterinarium, Hyg. § 4) and veterinary surgeons to attend to them (C. I. L. 5.2183, a freedman; C. I. G. 5117, ἱππωΐατρος; Dig. l.c.). On the whole of this interesting subject, see René Brian, L'assistance médicale chez les Romains, among the Mémoires présentés par divers Savants à l'Académie des Inscriptions, 1ère série, t. viii., and Droysen in the Deutsche militärärtzliche Zeitschrift, 1874.

II. The Principales belonging to an officer.

1. Clerks.

Though not all clerks (e.g. actarius） appear as belonging to an officer, yet as the majority do, it will be convenient to treat them all under this head. Among the various kinds of clerks are:

a, b. Librarii (in Inscriptions, Lib. or Libr.), who kept the accounts (rationes pecuniae, frumenti, hordei, &c.: cf. Veg. 2.7), differing herein from (b） Commentarienses, who entered the daily proceedings in journals. Several inscriptions of librarii have IM prefixed to the name (C. I. L. 6.3395). This means immunis (see below, p. 804 b); and such were the librarii qui docere possint, those horreorum, depositorum (of money deposited by the soldiers), caducorum (of the property of intestate and heirless soldier): cf. Dig. 50, 6, 7. The librarii do not mark the department of business to which they are assigned, but the officer under which they serve or the division of the army to which they belong. Thus we have librarii of the provincial governor (C. I. L. 3.5435), of the legate of the legion (ib. 3538), of the tribune of the praetorian cohorts (6.2632), of the centurions (Henzen, 6778); also of the legions (6.3395, and often), vigiles (ib. 221), equites (3.804), &c. We find commentarienses belonging to the provincial governor (comm. cos. 3.2015) and to the legatus and tribunus legionis (8.2586). The comm. legionis is the same as the comm. legati (2.4156).

c. A notarius, who was perhaps a shorthand writer, appears in 8.2755.
d. Exceptores, “qui verba dictata celeriter accipiebant,” belonged to the praef. praet. (Murat. 864, 3), to the provincial preses (6.2997), to the praef. and trib. of the vigiles (6.1058).
e. Codicillarii, account keepers, only appear in the vigiles (6.1056, &c.).
f. The functions of the tabularii are uncertain, though it is certain that the tabularium castrense contained the books and accounts of the camp (Marini, Atti, p. 499).
g. Nor is it known who the capsarii were, who are mentioned with the librarii in 8.2553, and like them were immunes (Dig. l.c.).
h. The actarii not only made up accounts (Veg. 2.19), but also looked after the procuring and distribution of the corn: cf. Aur. Vict. Caes. 33, who in this capacity gives them very bad characters. They must be held to have had the same functions as the actores bonorum privatorum et publioorum, harsh though it is to etymologically connect actarii with actor rather than acta. We find actarii legionis (C. I. L. 3.4232), cohortis (3392), cohortis urbanae (9.1617).
i. The exacti had also to look after the acta militaria, and did not differ essentially from the commentarienses. He was ex actis, and the word is formed like aborigines. We find exacti consularium (3.5812), legionis (8.2956), procuratoris (Boissieu, Inscr. de Lyon, p. 253).
j. The arcarius (8.3289) and the dispensator (3288) were slaves, appointed to guard and dispense actual cash. There were several slaves among the clerks of the camp (cf. familia rationis castrensis, 8.2702).

The functionary called ab indicibus (Orelli, 3464; C. I. L. 6.3414) was not a soldier, as the duty only attached to him after discharge from the army.

2.
Quaestionarii
for a quaestionibus.

These are torturers (Cod. Theod. vi. p. 348). Marquardt indeed urges (ii.2 552, note 1) that though such might belong to the vigiles, such cannot belong to the legions, the praetorians (C. I. L. 6.2755), and the cohortes urbanae (ib. 2880); for Roman citizens could not be put to the torture. He supposes them to be investigators concerning crimes. But Mommsen replies (Eph. Epigr. 4.421) that Quaestionarii ex legionibus are only found with those legati who held a province, and so cannot be considered to have held investigations on soldiers only, not to mention the fact that there were many soldiers not Roman citizens. There were at least five quaestionarii in the legion (C. I. L. 8.2751; 2586, 45).

3. Cornicularii,

so called from the adornment (corniculum, Liv. 10.44, 5) on their helmets . They were adjutants belonging to each large division of troops. As the praef. urbi had no military authority (Mommsen, Staatsrecht, ii.2 1020), neither had he a cornicularius. During the Republic one belonged to the tribunus militum (V. Max. 6.1, 11). Under the Empire we find one belonging to the legatus legionis (C. I. L. 3.887) and the tribunus militum (ib. 5974). If it is signified that there is one cornicularius in each legion (ib. 4452), the cornicularius legati is meant; if two are mentioned, that of the tribune is included. The arrangement of the commentarienses was very similar (cf. 8.2587). Provincial governors probably had two cornicularii (3.252); hence their office is called officium corniculariorum (ib. 3543). Similarly they belonged to praefecti praetorianorum (6.2776), vigilum (1057, 4, 2), annonae (11.20)--for the officia annonaria were among the first civil departments to assume a military form--castrorum (3.6023 a), to the subpraefecti vigilum (6.2997), to the tribunes of the praetorians. (2440), urban cohorts (2869), vigiles (2984);, and they also appear in the auxiliary cohorts. (Orelli, 4979). We find an adjutor corniculariorum (3.2052; C. I. G. 4453): probably he is the same as the subcornicularius (6.3596).

4. Secutor tribuni,

an adjutant who appears only among the soldiery of the city, praetorians (C. I. L. 6.2659), urban cohorts (ib. 2931), and. vigiles (ib. 1056-1058). The inscriptions which mention secutores in the legions are uncertain (Eph. Epigr. 4.406). The adjutor tribuni only appears among the vigiles (C. I. L. 6.220).

5. Singulares

so called, says Lydus (de Mag. [p. 1.804]3.7), because they rode one horse. In inscriptions the abbreviation for their title is sometimes (e.g.) Sing. Cons., sometimes S. C. (C. I. L. 3.5938; 7.723). They were orderlies belonging to different officers, to the praef. praet. (6.2634), trib. praet. (9.1617), legati legionum (6.3339, 3614), praef. vigilum (1056, 3, 1), provincial governor (3.5938). They were inferior to the beneficiarii (10.410). There were equites singulares as well as pedites singulares (8.2911, 9393). The equites and pedites singulares appear to have performed for the legati of the provinces and for the praefects and tribunes of the praetorians the same functions which the praetorians did for their leader; accordingly pedites and equites singulares formed a corps separate from the main body. The secutores probably did not differ in function from the singulares, but were inferior in rank; hence there are no secutores of the praesides, or of the praefecti praet., nor singulares of the tribuni vigilum; while both secutores and singulares are found belonging to the tribuni praetorianorum, who held a middle position (Mommsen in Eph. Epigr. iv. p. 404). Cf. Lange, Hist. Mutat. p. 50.

6. Stratores,

to be carefully distinguished from statores (see above, p. 794 b), belonged only to very high officials, whom they assisted to mount on horseback (Spart. Carac. 7): e. g. to the emperor (ib.), the governors of imperial provinces (C. I. L. 2.4114; 3.4317, 4836), procurators of provinces (Wilm. 1283), the praef. praet. (6.3408); but not to the proconsuls of senatorial provinces (Dig. 1, 16, 4, 1). Marquardt (ii.2 548) says they were generally either centurions or decurions: but they are often simply called milites, nor is it probable that the purchase of horses (Ammian. 29.3, 5) would be assigned to centurions. Again, in Wilm. 1293, we have a strator advanced to the position of a decurio. The fact is the strator retained the name and office, which he thought an honour, even when advanced to higher positions in the army. That the function of strator was one held in estimation may be proved from the fact that it is found alone, and not united to any other military position (e. g. 8.2792). He seems to have been generally a legionary soldier, though we find one from the auxiliaries (3.2067).

7. Beneficiarii

were nominees of the higher officers (Liv. 9.30, 3; Tac. Hist. 1.25), who varied in number according to the rank of the officer; e. g. the imperial legate of Numidia had thirty, a tribunis semestris had five. They were used for most varied services, frequently on police duty (C. I. L. 3.3412; Tert. de Fuga, 12). The legati Aug. pro praetore used to send their beneficiarii into outlying stations (3.3949): cf. Eph. Epigr. 4.529. We find beneficiarii belonging to the provincial governors, the legatus legionis, the tribunes of the legions and of the three departments of the city soldiery, the praefects of the praetorians and of the vigiles, the imperial procurators, but not of the praefectus urbi, who had no military authority (Mommsen, Staatsr. ii.2 1020; so pr. ur.=praetoris urbani in C. I. L. 9.1617). Evidence for this list in Eph. Epigr. 4.379 ff. Also there were beneficiarii of the tribunes of the equites singulares (6.3285) and of the auxiliary cohorts (5.898). Mommsen holds (Eph. Epigr. 4.394, note 1) that it was only the tribuni laticlavii or semestres who had beneficiarii, because they were engaged in civil administration, and not merely in military command.

8. Immunes.

The soldier who had fulfilled all duties in the construction of the camp, the watch, and the field was called munifex (Festus, p. 33). He who was freed from these duties was called immunis. A list of such privileged soldiers is given in Dig. 50, 6, 7, which comprise nearly all the principales. But some of the principales are called immunes simply. They got their privileges from their commander, but it is hard to see how they differed from the beneficiarii. We find thus immunis turmae, immunis legionis (Eph. Epigr. 4.409).

III. On the Promotion of the Principales.

The full arguing out of this difficult question would require too much space to be attempted here. However, we have attempted to draw up a list in order of rank, beginning from the lowest; and in doing so have consulted most of the inscriptions bearing on the question, the most important of which are Wilmanns, 1596, 1596 a, 1598, 1617; Orelli, 3176, 3480, 3464; C. I. L. 9.1617, 2.2610.

1.	Secutor.
2.	Singularis praef. pr.
3.	Beneficiarius trib. (C. I. L. 10.410).
4.	Tesserarius.
5.	Optio [legati].
6.	Vexillarius (in vigiles).
7.	Signifer (in praetorians).
8.	Curator fisci.
9.	Cornicularius tribunis.
10.	Beneficiarius praef. pr.
11.	Cornicularius praefecti (C. I. L. 11.20).
12.	Evocatus Augusti (Wilm. 1584).
13.	Centurio.

On this we have to remark (1) that in many of the above inscriptions one or more steps are omitted: e. g. in Wilm. 1596 a there is a sudden transition from 6 to 10. Such gaps may possibly be due to omissions on the part of the stonecutter, but, as is more likely, are evidence of special advancement owing to distinguished merit. Thus in C. I. L. 5.898 a beneficiarius tribuni is made a signifer, obviously from merit, as he is at the same time transferred from a cohort to a legion. Compare optiones becoming evocati (11.19, 5.7160) or centurions (5.7004, 6.215), a signifer becoming a centurion (8.217), &c. (2.) For the omissions of optio valetudinarii, optio carceris, optio ab actis (8.4874), which appear in 9.1617, see the section on optiones above (p. 801 b). (3.) Where exactly the librarius (cf. 8.217) and the librator (=mensor; cf. 6.2545) come is uncertain, but they certainly come before the tesserarius: as did also the commentariensis (cf. 11.19, 5.7004) before the optio, and the custos armorum before the signifer (8.2094). (4.) A signifer sometimes becomes aquilifer, and then curator veteranorum (5.3375). (5.) The codicillarius tribuni mentioned by Marquardt in his list (ii.2 557) occurs only in the vigiles, so is not given in this list. It should take the lowest rank below the secutor; for in 6.1057 (A.D. 205) three names of codicillarii appear--Dignus, Gaetulicus, Festus, who in 6.1058 (A.D. 210) are secutores tribuni. [p. 1.805]

The Levy (delectus, dilectus, κατάλογος）

Cf. W. Soltau, Ueber Entstehung und Zusammensetzung der altrömischen Volksversammlung, 1880, p. 335 ff.

In the time of the citizen army the levy took place yearly thus. The consuls (Plb. 6.19, 5-9) appointed by edict a definite day on which the citizens liable to service were to assemble at the Capitol (Liv. 26.31, 11), in later times at the Villa Publica in the Campus Martius (Varro, R. R. 3.2), if the latter was not the place of the regular arrangement in the ranks as opposed to the levy on the Capitol. Sometimes recruiting officers (conquisitores) were used throughout the districts of Italy (Liv. 23.32, 19; 25.5, 6), for the most part but not necessarily in times of great danger. After the Lex Plautia Papiria (89 B.C.), when all Italy was liable to legionary service, recruiting by conquisitores became constant. They were often most dishonest, and used to sell vacationes (Plut. Crass. 17; Auct. Bell. Alex. 56, 4). Whether the levy was in Rome or by conquisitores did not make any difference in the legal position of the soldiers--both kinds were legitimi milites.

The tribunes were first assigned as a nucleus to form the four legions in the order of their election (Plb. 6.19, 8 ff.): the four junior tribunes first elected and two senior tribunes first elected being assigned to the first legion; the three juniors and three seniors next elected to the second; the next four juniors and two seniors to the third, and the last three juniors and three seniors to the fourth. A tribe was then chosen by lot; the names of four of its members, generally those with lucky names (Festus, p. 137, 15; Cic. de Div. 1.4. 5, 102), called out, sometimes at random, sometimes by lot (V. Max. 6.3, 4; App. Hisp. 49). These, answering to their names (citati ad nomina respondebant, Liv. 7.4, 2), were chosen by the tribunes, one for each of the four legions; then four more (the tribunes of the second legion now having the first choice), and so on. A lot was then cast for the second tribe, and the same procedure followed; and so on till the legion was completed. Most probably all the tribes did not contribute an equal quota, for if they did it is hard to see the point of the lot; yet each tribe did contribute a quota, as may be proved from the exception in Livy, 4.46, 1. Volunteers (qui voluntate nomina dabant, Liv. 42.32, 6) were frequent. Their legal position in the service differed in no respect from that of those who served under compulsion. The cavalry, in early times, were chosen after the infantry, as the census of the centuriae equitum was held after that of the tribes (Liv. 29.37, 8); but in Polybius's time (6.20, 9) the cavalry were picked first from the equestrian list of the last census and 300 appointed to each legion.

If those bound to serve did not answer to their names, they sometimes had their goods confiscated (V. Max. 6.3, 4), or were scourged or imprisoned (Liv. 7.4, 2), or even in old times sold into slavery (Dig. 49, 16, 4, 10). Certain excuses were allowed: e. g. holding a magistracy or priesthood (Plut. Camill. 41), absence on state-service (Liv. 23.49, 2), specially-granted remission of service for distinguished bravery (ib. 20, 2), physical incapacity (these were called causarii, Liv. 6.6, 14). The inquiry into the validity of the excuses was conducted by the consul (Liv. 3.69, 7).

The levy being completed, the consuls administered the oath (sacramento milites rogare or adigere). One soldier stood out and repeated the oath (sacramentum or sacramento dicere) that he would obey orders and execute the commands of the officer to the best of his ability (Plb. 6.21, 2); after that each soldier was called out separately and said idem in me (Festus, p. 224 M.). The obligation of the oath lasted till the next campaign, a new oath being required for the new general (Liv. 3.20, 3; Cic. Off. 1.1. 1, 36). Without an oath it was unlawful to fight with the enemy (Cic. Off. 1.1. 1, 37), and those who did so were considered bandits (Liv. 8.34, 10). Desertion was a nefas (Senec. Epist. 95, 35) and deserved death (Dionys. A. R. 11.43). From the time of Marius the oath was taken once for all for the whole period of service (App. BC 5.128, 129), and was to the effect “pro republica se esse facturos nec recessuros nisi praecepto consulis post completa stipendia” (Serv. on Aen. 8.1; 7.614).

Besides this deliberate sacramentum, there was frequently in times of great danger (in tumultu) a swearing--in of the soldiers en masse (conjuratio), which first appears in 216 B.C. (Liv. 22.38). Something of the kind doubtless also took place when what was technically called tumultus, i.e. an Italian or Gallic war, suddenly arose (cf. Serv. on Aen. 8.1). The conjuratio, however, differed from the sacramentum in that the soldier bound by the latter was not released from service till he received his discharge (missio), while he who was sworn--in by a conjuratio was legally free from service at the end of the crisis (Mommsen, Eph. Epigr. 5.143).

The qualifications for service were after the time of Servius Tullius, besides physical capacity (cf. Plb. 6.20, 3) and citizenship, chiefly having a considerable stake in the country, i. e. a rated property. The proletarii and capite censi were enrolled only in times of great danger (Gel. 16.10, 11). Plutarch (Plut. Mar. 9) states that slaves also were deliberately enrolled; and though such did happen occasionally in the confusion of the Civil War (Plut. Mar. 41 ; Sulla, 9 Brut. 45; Flor. 2.9; Caes. B.C. 1.24, 2; Bell. Afr. 19, 3; App. BC 2.103, 3.49), still it was a practice which did not continue in settled times. Even in the age of Theodosius a slave was not allowed to serve in the army (Cod. Theod. 7.13, 8). The lowest census which admitted a man to the army was in the time of Servius Tullius 11,000 asses, but by the time of Polybius (6.19, 2) it was only 4,000. In the Social War libertini were enrolled (Liv. Epit. lxxiv.); and already in the time of Marius the capite censi had been taken into the service, the revolutionary nature of which alteration is duly emphasised by ancient writers (Plut. Mar. 9; V. Max. 2.3, 1; Gel. 16.10, 4), and its disastrous results noted by Appian (App. BC 5.16), as we have seen above (p. 785 a).

Caesar and Pompeius assumed the right of enrolling whole legions of native troops in the provinces (legiones vernaculae), who were not citizens (Caes. Bell. Hisp. 7; B.C. 2.20). But this practice was restricted by Augustus to the [p. 1.806]Eastern provinces. For it is to be especially noted that Augustus made Italy and the West supply the western legions and the East the eastern ones, and so paved the way for the final division of the Empire into the partes Orientis and partes Occidentis of the Theodosian era (Mommsen, Die Conscriptionsordnung der röm. Kaiserzeit, in Hermes, xix. pp. 11, 23--a profound and spirited article). But in the West, although the legions were partly raised from communities of Latins and peregrini, those who were enrolled were made citizens (ποιησάμενοι δὲ πολίτας, οὕτω καὶ στρατιώτας ἐποιήσατε, Aristides, i. p. 352, Dind.: cf. Tac. Ann. 3.40, 11.24), and a large number of the soldiers were by birth Roman citizens (Mommsen, op. cit. pp. 13, 63). Still we must remember that freedmen served in the vigiles (D. C. 4.26, 5; Suet. Aug. 25).

The qualifications for the legionary service under the Empire, besides such physical ones as health, strength, and a certain height, viz. 5 ft. 10 in.--which regulation height was called ἔγκομμα or incomma (Veg. 1.5)--and not ever having been convicted of any serious crime, such as adultery (Dig. 49, 16, 4, 7), were: (1) to belong to an urban community, which was considered to be a guarantee of being civilised; (2) free-birth--qualifications which were very often got over. For though formally the legionary in the roll always had his native community attached to his name, still Pannonians and Thracians figure very largely among them; and freedmen were easily made free-born by the fiction of natalium restitutio (Mommsen, pp. 16-18): cf. Plin. Ep. 10.29, 30.

During the Empire there were seldom regular levies (Vell. 2.130, 2). We find such indeed after the defeat of Varus (Suet. Aug. 24), under Nero (Suet. Nero 44), Vitellius (Tac. Hist. 3.58), Hadrian, M. Aurelius, Maximinus (Mommsen in Hermes, 4.119); but as a general rule vacancies were filled up by volunteers (Dig. 49, 16, 4, 10); and even in the levies getting a substitute was allowed (Plin. Ep. 10.30 (39), 1). Gradually the Italians ceased to take service in the legions; and as in the legions enrolled in the Vespasian era (e. g. I. Adjutrix) they do not appear, while they do appear in those of Nero's time (e. g. Leg. xi. at Vindonissa), we may put the consummation of this important result in the time of Vespasian (Mommsen in Hermes, 19.19), though we must remember that this was a result arising from natural causes, and not from any regular legal enactment. In the Digest (l.c.) the levy is still in theory universal and exceptionless. The Latin provinces, then, from Vespasian's time contributed more largely to the legions; and as a counter-balance the African forces were from that time on assigned to the Eastern division of the Empire (Mommsen, pp. 9, 19). The second important feature of the imperial conscription was, for the most part, the local conscription for each legion. Under Hadrian the African legion was almost exclusively enrolled from Africa and Numidia, and to Hadrian we may attribute the extension of this regulation to the whole Empire (ib. 10, 21).

The city-guards were raised by Augustus from the citizens of the old Latin communities; under Tiberius Cisalpine Gaul is added: but up to Severus Italians predominate, though in the second century many soldiers appear who were natives of Spain, Macedonia, and Noricum (D. C. 74.2). After Severus Italians rarely appear (Mommsen, pp. 52, 53).

From the time of the Social War and the extension of the citizenship by the Lex Plotia Papiria, the Socii had vanished entirely, and the legions and auxilia form the two main bodies of the Roman army. The latter were the non-Roman troops raised partly in the provinces, partly from allied lings and nations. In imperial times they generally stand opposed to the legions (Vell. 2.112, 4). These auxiliary troops were for the most part raised from peregrini--in the first instance in definite districts from which they got their names, but the vacancies which arose came to be filled up with natives of other districts, especially from natives of the place where the troop happened to be located. Thus in 60 A.D. (see Diploma ii.) an Illyrian gets his discharge from the Cohors II. Hispanorum located in Illyricum (Mommsen, p. 41). These non-citizen troops are designated by the name of the nation to which they belonged; for if they belonged to a town, they could have served in the legions (ib. 25). Local conscription for each division probably began in the case of the auxiliaries before it did in that of the legions (ib. 42). By Augustus the auxiliaries were only taken from the imperial provinces and Africa (ib. 44), the senatorial provinces supplying in larger measure than the imperial ones the legions and the guard. But as the senatorial provinces had no garrisons, when local conscription was introduced by Hadrian, the legionaries ceased to be taken from them; while those who belonged to urban communities in these provinces, and wished to join the service, served only in the guard, and after the time of Severus not even there (ib. 51).

The officers who effected the recruiting under the Empire, which they did by instituting an inquisitio (Plin. Ep. 10.30 (39), 2), were called dilectatores (Wilm. 1256, 1257), officials of senatorial rank in Italy and the senatorial provinces, of equestrian rank in the imperial (Mommsen, Staatsrecht, ii.2 819; Hermes, 19.56, 59). Sometimes the levy was held by the proconsul in a senatorial province (Tac. Ann. 14.18).

The Line of March (Agmen).

The usual order of march in the second century B.C. is described by Polybius (6.40, 4 ff.). It was in a single column. First came the extra-ordinarii--that is, if the enemy were in front; for if they were in the rear, the extraordinarii were placed in the rear. Next came the ala dextra of the allies, followed by its baggage and the baggage of the extraordinarii. Then one of the legions, followed by its baggage; then the other legion, followed by its own baggage and that of the ala sinistra of the allies, which latter brought up the rear. The cavalry usually followed the divisions of infantry to which they were attached; sometimes, however, they were placed on the flanks. The legions shifted places each day, as did also the alae, so that each should get turns in the advantage of first arrival at fresh food and water. In open places where an attack was imminent, the line of march was [p. 1.807]almost that of battle, into which it could most readily be changed (Tac. Ann. 2.16). Caesar calls it triplex acies (B. G. 1.49, 1, and passim). According to it, the hastati, principes, and triarii marched in three parallel columns beside one another, each maniple having its baggage before it. On any danger approaching, by a movement at one time to the right, at another to the left, the maniple got clear of the baggage, and, having got it in the rear, formed the triple battle array. But Caesar (e.g. B. G. 2.19, 1; cf. 8.8, 3) often varied this, arranging the whole main body in front, then the whole of the baggage, and behind it a rear-guard. The triplex acies is usually called by Livy agmen quadratum, which by no means implies a square with four sides (Liv. 44.9, 6), but simply a rectangular arrangement (cf. Cato in Non. 204, fronte longa, quadrato exercitu): see Liv. 10.14, 7; 21.57, 7, &c. But sometimes the Romans formed a hollow square or rectangle, with the baggage in the middle, when a sudden attack was at all expected. We first hear of this arrangement, itself also called agmen quadratum, in the Spanish wars (e. g. in 151 B.C.; App. Hisp. 55), and it was continued in the Jugurthine and Parthian wars (Sal. Jug. 46, 100; Plut. Crass. 23), and frequently elsewhere (Tac. Ann. 1.51, 64; 13.40)--from which passages we see also the difference of arrangement on different occasions, sometimes the front and rear being legions and the sides the allies, sometimes a legion being on each side of the square, with the auxiliaries as a front and rear guard, &c.

Nothing is more insisted on by the Roman writers than the great weight of baggage the Roman soldier used to carry: “injusto sub fasce viam cum carpit,” says Virgil (Georg. 3.346). Cicero (Tusc. 2.16, 37) says that, besides his shield, sword, helmet, &c., “which he no more reckoned than he did his shoulders, arms, or hands,” he had to carry food for half a month or more (Caes. B.C. 1.78, 1), sometimes for a whole month (Liv. 44.2, 4), one or more stakes, and various odds and ends: of the latter Josephus (B. J. 3.5, 5) mentions a very miscellaneous collection--saw, basket, spade (ἄμην), axe, strap, sickle, rope. The whole weight of the baggage Vegetius (1.19) estimates at 60 Roman pounds. Marius made an arrangement for carrying all this on a board fastened on to a forked support which was strapped across the shoulders. They were called muli Mariani (Frontin. 4.1, 7: cf. Festus, 24, 148, M.). By this arrangement they could be both more easily carried, and more readily cast aside and resumed when circumstances required it.

The heavy baggage, such as tents, hand-mills, &c., were carried on horses or mules (jumenta) driven by calones. Every ten men had one tent, thereby forming a contubernium (Hyg. § 1; Joseph. B. J. 3.6, 2). [CASTRA] Each centurion and higher officer had at least one tent. If we suppose each jumentum carried one tent, we should have over 2,000 jumenta in an ordinary army of 20,000 men, so that the impedimenta, even before artillery was much used, was very considerable.

In the time after Hadrian the order of march appears to have been that the legion was divided into two divisions, the first consisting of the first five cohorts arranged from right to left, the second the sixth to the tenth cohorts similarly arranged (Veg. 2.15).

The Battle Array (Acies).

According to Livy (8.8, 9 ff.), the ten maniples of the hastati stood first, next immediately behind the intervals (intervalla, διαστήματα) between the maniples of the hastati stood the principes, and behind the intervals of these the ten maniples of the triarii. The arrangement was that of a quincunx (Plb. 15.9, 6). Should the hastati be defeated, they retired through the intervals of the principes (which shows by the way that the intervals between the maniples must have been nearly if not quite as great as the front of the maniple). Should the principes likewise be defeated, “the contest came to the triarii” (res ad triarios redit), which was a proverb for a most serious state of things. This arrangement was the ordinary one; but sometimes there was a variation, as for example, in the battle of Zama, Scipio, in order to allow a free passage for the elephants through the maniples, drew up the maniples of the three lines straight behind each other (Plb. 15.9, 6).

The standards (signa, i.e. the standards of the maniples; for there were neither standards of the cohorts nor was the eagle the legionary standard till the time of Marius [SIGNA]) usually stood behind the last row of the maniple. The signa, especially alluded to in accounts of battles, are those, of the hastati (Liv. 37.39, 8). Hence we see clearly what Livy means when he says (8.11, 7), “Caesos hastatos principesque, stragem et ante signa et post signa factam, triarios postremo rem restituisse.” The antesignani, then, are the prima acies (cf. Liv. 47, 2; 8.39, 4; 10.27, 9); and in a similar sense postsignani (a word, however, of late date) is used for the second line, the principes (Frontin. 2.3, 17). If the enemy are repulsed, the hastati advance, signa promota (Liv. 8.38, 10); and if the hastati have to retire, the secunda acies ad prima signa successit (Liv. 9.32, 8). But sometimes the signa certainly appear to have been in the front of the maniples (Bell. Afr. 15; Tac. Hist. 2.43), and this was no doubt the usual position not only in the march, but also in a charge in battle ( “et signae prae se ferri plenoque gradu ad castra hostium oppugnanda succedere jubet,” Liv. 34.15, 3). Signum is, however, often used in the very general sense of the main body (Liv. 21.55, 2; 38.21, 2), and similarly we find it in miles subsignanus, used for a legionary soldier in opposition to an auxiliary (Tac. Hist. 1.70), and to desert was relinquere signa (Liv. 5.6, 4). Somewhat in this sense, too, we must understand the antesignani in Caesar, who applies that term apparently to a select body different from the common soldiers (cf. Cic. Phil. 5.5, 12), formed inside the legion (Caes. B.C. 1.57, 1) from the most active and brave men, three (Bell. Afr. 78, 5) or four hundred (B.C. 3.75, 5), who were free from all baggage (expediti), and who could be used for sudden exigencies, such as seizing a position (B.C. 1.43, 3), reinforcing a hard-pressed division (Bell. Afr. l.c.) or fighting among the cavalry (B.C. 3.84, 3). They would be so called because they were independent of the [p. 1.808]regular organisation of the legion, and mostly used for rapid movements in front of it. H. Schiller (in I. Müller's Handbuch, iv. p. 737) says that these corps of antesignani probably formed the training school of able subaltern officers.

But something more must be said of the order of battle of the whole combined army, legionaries and allies. The usual order appears to have been the two legions in the centre, flanked on the right by the ala dextra of the allies, and on the left by the ala sinistra. On the extreme right was the legionary cavalry and the equites extraordinarii (Liv. 22.45, 7; 27.12, 4); on the extreme left the allied cavalry. But there were frequent variations--sometimes the cavalry was behind the legions (Liv. 10.5, 6), or the allies were in front, the legions in reserve (35.5, 8), or the legions were on the wings, e.q. in the battle of Scipio against Hasdrubal at Ilipa (Plb. 11.20, 5). Vegetius (3.20)--who is plainly copying from Cato, de re militari (cf. Cato ap. Nonium, 204), whom Vegetius reproduces word for word--gives a whole series of recognised kinds of attack: (1) fronte longa quadro exercitu--the normal kind described above; (2, 3) obliqua acies, i.e. right (as at Pharsalia, Caes. B.C. 3.91) or left wing attacking; (4, 5) sinuata acies, both wings attacking, legionaries exposed or guarded by light-armed; (6) whole army attack left wing of enemy (usual when a battle ensues on the march); (7) a wing rests on a fixed point (e. g. a river or a mountain), so that it cannot be surrounded; (8) cuneus, called by the soldiers caput porcinum (Veg. 3.19; Amm. Marc. 17.13, 9), viz. the centre projecting like a wedge, a kind of attack to be received by a V-shaped arrangement, called forfex. For an example of a cuneus and a forfex, see Liv. 39.31.

(By way of parenthesis we notice here other meanings of cuneus: (1) any division (Augustin. de Ordin. 2.18, 48); (2) any serried band--thus the maniples in battle are so called (Liv. 7.24, 7） and the Macedonian phalanx (Liv. 32.17, 11); (3) in the later Empire a definite division of cavalry (Cod. Theod. 5.4, 1; Not. Dign. Occ. 40, 54; cf. C. I. L. 7.218), which appear among German troops (Tac. Germ. 6), e. g. the Frisii in Britain in the third century (cf. Mommsen in Hermes, 19.232-3).)

In particular emergencies other combinations of soldiers were formed, with special names: e.g. (a) orbis, a serried, not a hollow, square or mass, which soldiers hard-pressed by the enemy formed (volvere, facere), to protect themselves on all sides and to prevent themselves being scattered (Sall. Jug. 97, 5; Caes. Gal. 4.37, 2); (b) testudo, used either by the whole army against a violent discharge of missiles, or by a small division in attacking a wall. The front rank held their long half-cylindrical shaped shields before them, and the other ranks, each stooping a little more than the one before it, held their shields closely locked together above their heads so as to form a sloping roof. Full description in D. C. 49.30; Liv. 34.39, 6 [TESTUDO]. (c） globus, a general word for a closely-formed attacking division (Liv. 4.29, 1; Tac. Ann. 14.61).

The labours of Hadrian in his organisation and improvement in the discipline of the army are often mentioned (Spart. Hadr. 10; D. C. 69.9, 4; Aurel. Vict. 14, 10). Under him we may place the definite return to the arrangement by phalanx, which was the earliest arrangement of all in Roman history. Not but that the phalanx was used earlier. A fight of man against man, which in a broad way of speaking was the style of fighting used in the Republic, had to give way when barbarian hosts were met; for barbarians are far less dangerous to masses than to individuals. In a battle Suetonius Paullinus fought against the Britons he marshalled his forces in three phalanxes, so that they could not be broken (D. C. 62.8). But the work of Arrian, ἔκταξις κατ̓ Ἀλβάνων (not Ἀλανῶν: see D. C. 69.15), legatus of Cappadocia in 136 A.D., which describes officially the arrangement in a phalanx, causes us to fix Hadrian's reign as the period of its definite adoption. The phalanx is eight deep, in close array, the first four lines armed with pila (κοντοφόροι), the last four with lanceae (λογχοφόροι) [LANCEA]: cf. Tac. Hist. 1.79. The pila [PILUM] were to be used by the front rank for piercing, by the other three lines as missiles (Arrian, op. cit. 15). The cavalry and artillery stood both on the wings and in the rear of the phalanx (ib. 19), and further back still a reserve of picked men to make a charge at the right moment (ib. 22, 23). We hear that Caracalla had a phalanx of 16,000 Macedonians (D. C. 77.7, 1), and Alexander Severus one of 30,000 (Lampr. Alex. Sev. 50, 5).

Conditions of the Service

Augustus, we are told in Suetonius (Suet. Aug. 49), “quicquid utique militum esset, ad certam stipendiorum praemiorumque formulam astrinxit; definitis pro gradu cuiusque et temporibus militiae et commodis missionum, ne aut aetate aut inopia post missionem sollicitari ad res novas possent:” cf. also D. C. 54.25, 5, ἐπὶ ῥητοῖς . . . καταλεγόμενος, and Herodian, 2.11, 5, where, as in Suetonius, the conditions alluded to have reference to pay while in the service, the number of years' service which was required, and some consideration or reward on discharge.

1. Pay.

Prior to the siege of Veii, in 406 B.C., the expenses of the Roman soldier were defrayed by his tribe (Dionys. A. R. 4.19, 5.47), except the knights, who got from the treasury an aes equestre to purchase and aes hordearium to feed his horse. And even after 406 the pay was rather for maintenance than for remuneration, ὀψωνιασμὸς rather than μισθός. It was paid at the end of service--for service up to six months, which was a semestre stipendium (Liv. 40.41, 11), six months' pay (Dionys. A. R. 9.59); for service beyond six months, which was an annuum stipendium (Liv. 42.34, 5), a year's pay (Diod. 14.16): cf. Varro ap. Non. p. 532, s. v. acre dirutus; Liv. 24.11, 8. (The six months were from March to August.) The payment was laid on the conquered (Liv. 5.27, 15; Dionys. A. R. 9.17, 36, 59). Polybius (6.39, 15） says that the provisions, clothes, and accoutrements were deducted from the pay; and, though we hear that C. Gracchus brought in a law providing clothes for the soldiers at state expense (Plut. C. Gracch. 5), we know that in the [p. 1.809]time of the Emperor Tiberius they (Tac. Ann. 1.17) complain that they had to supply out of their pay vestem arma tentoria--a passage which, by the omission of frumenta, proves that at this time their food was supplied by the state, as it certainly was later (Veg. 2.19, 20; 3.3). But if the soldier had these items deducted from his pay, it was incumbent on the state to see that there were proper supplies of them (Liv. 27.10, 13; 44.16, 2; 9.29, 4; 29.22, 3; 34.6, 13; 42.27, 8. Caes. Gal. 1.16, 1; 6.33, 4).

The amount of the pay in the time of Polybius was two obols a day for the legionary, four for the centurion, and six for the horseman. Now a drachma to Polybius was equal to a denarius, which originally was 10 asses: therefore the annual pay of the legionary for the year of 360 days was 1200 asses, of the centurion 2,400, and of the horseman 3,600. In 217 B.C. the denarius was made equal to 16 asses, but the old reckoning of the denarius (=10 asses) remained for the soldiers (Plin. Nat. 33.45). Caesar is said to have doubled the military pay (Suet. Jul. 26). What he actually did was, he trebled their nominal pay (i. e. 1200 asses), but paid them in the new coinage (1 denarius=16 asses), so that they got 225 denarii a year. There is, however, a certain confusion. In Tac. Ann. 1.17 the legionaries say they got 10 asses a day and claim to get a denarius, while Pliny (l.c.) says that in military pay the denarius was always given for 10 asses nominal; so that the claims of the legionaries appear to be for what they already got. Perhaps the claim was that the 225 denarii, which was now their pay, be paid at 16 asses the denarius, and not at 10. Other troops were paid certainly in denarii, e. g. the praetorians (qui binos denarios acceperint, Tac. l.c.) = 720 denarii a year. Sometime between 27 B.C. and the time referred to in this passage of Tacitus the pay of the praetorians was raised from 1 1/4 of a denarius to 2 denarii, for in 27 they got double the pay of the legionary (D. C. 53.11). Domitian increased the pay of the legionaries by one-third, so that it was now 300 denarii. From the time of Caesar the payments were made every four months, each payment being called a stipendium, so that the soldiers got three stipendia each year (Suet. Dom. 7; Zonar. 11.19). The pay of the cohortes urbanae is to be inferred from the legacies left to them by Augustus and Tiberius (D. C. 56.32, 2; 59.2, 1, 3), which were doubtless, as in the case of the praetorians and legionaries, one stipendium, i.e. 120 denarii; therefore their yearly pay would be 360 denarii. The only superior officer's pay we know of is that of the tribunus legionis in the third century, i. e. 25,000 sesterces (see Inscription of Torigny).

Military men sometimes, for distinguished bravery or for acting as emeriti or evocati after their legal period of service had expired, got increased pay; and as such, if the pay was doubled, were styled duplarii (C. I. L. 8.2564), duplicarii (Varro, L. L. 5.90), dupliciarii (Bramb. 475); if the pay was increased by one-half, sesquiplicarii (C. I. L. 3.3164; Veg. 2.7). The term salariarii points to officers who were employed in civil rather than in strictly military duties; and their years of service are called salaria (C. I. L. 6.2495, 2589). It is applied to adsessores, professores, medici, and evocati: see Mommsen in Eph. Epigr. 5.152.

2. Length of service and discharge.

During the Republic the citizens were bound to serve between the completed 17th year and the completed 46th (Gel. 10.28; Mommsen, Staatsr. i.2 487), the legionary being bound to serve sixteen or at most twenty campaigns, the horseman ten (Plb. 6.19, 2; reading δεκαὲξ for ἕξ οὐ). Under Augustus the legionary had to serve sixteen years (D. C. 54.25, 6), but afterwards it was fixed at twenty, that of the praetorians being sixteen (ib. lv, 23, 1; Tac. Ann. 1.17, 78), that of the urban cohorts being twenty (Dig. 27, 1, 8, 9), and that of the auxiliaries being twenty-five: see the Diplomata passim, where the formulae are for the legionaries qui vicena stipendia aut plura meruerant (e.g. Dipl. vi.), for the auxiliary forces qui quina et vicena stipendia aut plura meruerant (e.g. Dipl. ix.). Note aut plura; for as a matter of fact we find that soldiers were very commonly retained in the service as evocati long after their legal term of service had expired, e. g. we have evocati of thirty-eight (C. I. L. 3.2818) and forty-six years' service (Mommsen on C. I. L. iii. p. 282). After the year 108 A.D., in the Diplomata we always find the privileges of veterans accorded only to those who are discharged (dimissi honesta missione), but yet these often appear to be still serving in such and such an army under such and such a leader (e. g. Dipl. xxxiv.). After Hadrian's time it appears (see Mommsen in Eph. Epigr. v. p. 4, No. 10) that a soldier did not get the honesta missio till twenty-five years' service, but during the last five years he was released from the harder duties, or in a later age chosen among the protectores (Mommsen on C. I. L. 3.6194). Hence we see how it is that the period of service is said sometimes to be twenty years (Dig. l.c.; Cod. Just. 7.64, 9), sometimes twenty-five (Serv. ad Aen. 2.157). Discharge, though properly a function of the emperor, yet is often said to have been performed by such and such a legatus (C. I. L. 3.1078, 1172; Eph. Epigr. iv. p. 502). In the case of honesta missio, if the veterans did not get land, they got a lump sum of money, the praetorians 5,000 denarii, the legionaries 3,000 (D. C. 55.23, 1), which Caligula (Suet. Cal. 44) cut down to half and Caracalla raised to 5,000 (D. C. 77.24). In case of physical incapacity or sickness, a soldier got what was called a causaria missio. As to drumming out (ignominiosa missio), see below (p. 811 b), and compare Dig. 49, 16, 13, 3.

Miscellaneous Features of the Service.

1. Marriage.

All the Diplomata give certain privileges with regard to marriage to the person or persons named in each document. The only Roman citizens these privileges are granted to are the praetorians and urban cohorts, and in their case grant them the right of forming full connubium with the first woman of foreign birth they marry after discharge, and the consequent legitimacy of the children arising from such a marriage (cf. Gaius, 1.57). Dio Cassius (60.24, 3) says that a soldier (meaning doubtless one who was a Roman citizen) could not legally marry while in the service and prior to his [p. 1.810]discharge, or at any rate till performance of his full term of service; but Claudius allowed them to do so (τοῖς στρατευομένοις ἐπειδὴ γυναῖκας οὐκ ἐδύναντο ἔκ γε τῶν νόμων ἔχειν τὰ τῶν γεγαμηκότων δικαιώματα ἔδωκε). If he was married before joining the army (Dig. 24, 1, 32, 8; 60-62; 49, 17, 6, 8), he could retain his wife, but could not keep her in the camp, and so had to part from her; but as men entered the army about their 18th year, there cannot have been many such cases. The only women in the camp were either meretrices (App. Hisp. 85) or non-Romans (Liv. 43.3, 1): in early times it was a great disgrace for a Roman woman to be in the camp (Serv. on Aen. 8.688; Prop. 5.3, 45). Hence the auxiliaries could easily get wives of their own class; and as they served for a somewhat longer period than the legionaries, they were allowed to marry while still soldiers; and in the Diplomata, which are given in great numbers to the auxiliaries, they get the right of connubium with the wives they already have when the latter have got citizenship, or if unmarried the right of connubium with whatever wife they chose to marry subsequently to the issue of the diploma. The Roman soldier, then, could not legally marry; and this was the rule before the time of Severus. But by him all soldiers were allowed to have a concubine (technically called focaria, Cod. Just. 5.16, 3); and auxiliaries and legionaries were put on the same level. For that is the meaning of γυναιξί τε συνοικεῖν in Herodian, 3.8, 5 ; not that the soldiers could contract regular connubium, as may be seen from the Diplomata subsequent to this date: e. g. liii., which runs, “ipsis filiisque eorum quos susceperint ex mulieribus quas secum concessa consuetudine vixisse probaverint, civitatem Romanam dederunt et connubium cum iisdem quas tunc secure habuissent cum est civitas his data, aut, si qui tune non habuissent, cum iis quas postea uxores duxissent” (cf. Eph. Epigr. v. p. 100). The children sprung from such a focaria bore the name of the mother (cf. the four children born to one man from four different focariae in Dipl. liii.), were said to have been born in the camp, and all assigned to the tribe Pollia, so as to enable them to take service in the legions (Wilmanns in C. I. L. viii. p. 284; Mommsen in Hermes, 19.11, note). In the fourth century all soldiers could marry (Cod. Theod. 7.13, 6), and permission could be granted by the commander to a soldier to keep his wife in the camp (Cod. Theod. 7.1, 3, and Godefroi's notes; Cod. Just. 2.52, 2).

This is Mommsen's view (C. I. L. iii. p. 905 if.), and as it is generally adopted has been set forth fully. But Mispoulet (Le Mariage des Soldats romains in the Revue de Philologie, 1884, pp. 113-126) has raised some strong arguments against it. The exigencies of a military life, he says, always preclude the majority of soldiers from marriage; but that is different from a legal prohibition of it. It is à priori improbable that Augustus, who encouraged marriage so much, would have condemned 200,000 men in the prime of life to celibacy. Again, the inscriptions constantly speak of the conjux of a soldier and call him maritus: e.g. C. I. L. 3.102, 5949; 7.23, 25, 121, 245; 8.3084, &c.; but we do not know the dates of these inscriptions, and we need not suppose that they necessarily speak with strict accuracy, nor that what was legally cohabitation may not have been expressed by the more honourable term. But the passages adduced from the Digest are strong: 49, 17, 16, pr., “Dotem filiofamilias datam vel promissam in peculio castrensi non esse respondi (Papinianus). . . Dos matrimonio cohaerens;” 23, 2 (de ritu nuptiarum), 45, 2, “Plane si filius-familias miles esse proponatur non dubitamus, si castrensis peculii ancillam manumiserit, competere ei hoc ius,” &c.; 23, 2, 35, “Filiusfamilias miles matrimonium sine patris voluntate non contrahit.” These passages cannot be got over by saying that they refer to marriage contracted before entering the service. Mispoulet's explanation of the passage in Dio Cassius is that Claudius allowed the soldiers who could not live with their wives according to military rules, and so did not marry, relief from all the penalties which fell on unmarried civilians--an exceedingly reasonable interpretation of the words. He notices also that Severus's enactment was that wives might live with their husbands; and just about that time, as Wilmanns has shown (l, c.), the camp at Lambaesis ceased to be lived in by the soldiers: they go to the town and only return to the camp at stated hours. The connubium granted by the Diplomata, he says, was the right of full marriage, which would put the children in the position of those born from two Roman citizens, even if one or both of the parents were peregrinae condicionis.

2. Business of the Soldier in peace. Exercises and the Construction of Public Works.

It was always in accordance with Roman discipline that the soldiers should not get demoralised by inaction, nor should they, as the Emperor Probus put it, eat the bread of idleness ( “annonam gratuitam milites comedere non debere,” Vopisc. Prob. 29): cf. Frontin. 4.1, 15. In the first place, then, they had to go through the drill (see especially Vegetius, 2.23, and i. passim) necessary to keep them in efficient training, which gave the name to the army (exercitus, Veg. 2.23). Hadrian laid great stress on this drill (Spart. Hadr. 10, 3), which was held twice a day for recruits, once for full soldiers. They were further trained in wood-cutting, riding, jumping, running, swimming, shooting javelins and arrows, slinging stones, fencing and sparring at a stake with an unpointed wooden stick (vectis, Sall. Hist. 2.11; clava, Veg. 1.11; sudes, Juv. 6.248; κυιντανὸν κόντακα, Cod. Just. 3.43, 1, i.e. ludus quintanus, perhaps so called because often practised in the via quintana), &c. Besides this, there was occasionally the decursio, an evolution whereby the army divided into two parts, and engaged in a sham fight; this especially after the review (lustratio): see Liv. 40.6, 5; cf. Eckhel, 6.271, 503. Also once a month the ambulatio, i.e. fully armed and equipped to march ten miles forward and ten miles back at the regulation military pace (four miles an hour), sometimes at an accelerated pace (pleno gradu=nearly five miles an hour): cf. Veg. 1.9, 27.

Soldiers often executed certain public works. They were never allowed to work for private people (Dig. 49, 16, 12, 1; cf. Liv. Epit. xi.); but sometimes they were allowed to build temples and public edifices for the municipalities in the provinces (Dig. 50, 16, 7, 1). In republican [p. 1.811]times the soldiers were occasionally employed in executing large public works. For example, Flamininus had the road from Bononia to Arretium made by his soldiers (Liv. 39.2, 6); Marius dug the Fossa Mariana from Exagium (St. Gabriel), about 5 Roman miles N. of Arles, to the village of Fos (Fossae Marianae portus), about 28 Roman miles N.W. of Marseilles. The length of the canal was nearly 30 Roman miles. (Full discussion in Desjardins, Géographie de la Gaule romaine, i. p. 199 ff.) Sulla diverted the river Cephissus from its channel (Plut. Sull. 16): cf. also Caes. Gal. 1.8, 1; Bell. Afr. 32, 2.

But the great works of the Roman soldiery appear under the Empire.

(1.) The Limites, or frontier fortifications; such as, the Vallum Hadriani or Picts' Wall in the N. of England, from Wallsend (Segedunum), E. of Newcastle-on-Tyne (Pons Aelius), to Bowness (Glannibanta), W. of Carlisle (Luguvallium). This work is even now in its ruins one of the most impressive monuments of Roman military greatness. It is well known to all English scholars, with its stations, trenches, mile-castles, &c., thanks to Dr. Bruce's learned and eloquent work (The Roman Wall, 3rd edit. 1867). Again, the Limes Transrhenanus, from the mouth of the Lahn near Kelheim to the Danube; the Limes Rhaetiae, the Limes Orientalis, the Limes Africanus, &c.

(2.) The Military Roads, which were always the main high roads (Quint. 2.13, 16). From the duties of road-making even veterans were not free (Dig. 49, 18, 4). These roads were everywhere. To take two examples: in Dalmatia, the road from Salonae to Andretium was built by the Seventh Legion (C. I. L. 3.1627); and in Africa the road from Carthage to Thevesta was built by the Third Legion (C. I. L. 8.10048). For more, see the chapters on Viae Publicae in the various volumes of C. I. L. Both legionaries and auxiliaries worked at these roads: see the important inscription from Coptos in Eph. Epigr. v. pp. 5-17.

(3.) Keeping up and extending the fortifications already in existence. In all the provinces there are vast numbers of bricks found with the stamp of the legions which executed works there. For British ones, see C. I. L. vii. p. 255 sqq.

(4.) Miscellaneous works (cf. generally Vopisc. Prob. 9, 2), such as making bridges (Tac. Ann. 1.20), cutting canals (ib. 11.20), building ampnitheatres (ib. Hist. 2.67), temples (e. g. that of Aesculapius and Salus at Lambaesis, C. I. L. 8.2579), dredging rivers and harbours (Suet. Aug. 18; Liban. i. p. 324, Reiske), planting vineyards (Vopisc. Prob. 18, 8), trying to make head against locusts (Plin. Nat. 11.106), even working in mines (Tac. Ann. 11.20), though this was generally one of the severest punishments (Dig. 48, 19, 28, pr.), and as a punishment was not inflicted on soldiers (ib. 49, 16, 3, 1).

3. Rewards and Punishments.

With regard to rewards, it will be sufficient to refer to the Index, s. v. Military Rewards, and the various articles mentioned there, such as CORONA, OVATIO, TRIUMPHUS. But a word here may be said on the punishments, which Modestinus summarises in the Digest, 49, 16, 3, 1: “Poenae militum hujuscemodi sunt, castigatio, pecuniaria multa, munerum indictio, militiae mutatio, gradus deiectio, ignominiosa missio.” (Compare, for many examples, V. Max. 2.7, De disciplina militari.）

(1.) Castigatio could be inflicted even on officers (Val. Max. l.c.).

(2.) Pecuniaria multa might be deprivation of whole (ib. § 5) or part of pay (Liv. 40.41, 11), the campaign not being reckoned in the year of service. One who thus lost his pay was called aere dirutus (Fest. s.v. Non. p. 532).

(3.) Munerum indictio could be inflicted only on those who were in some degree immunes, by making them munifices (V. Max. 2.7, 4), which may be regarded as gradus dejectio.

(4.) Militiae mutatio, degradation, e. g. from the cavalry to the infantry (Frontin. 4.1, 18), or to the funditores (V. Max. 2.7, 15). This was also called ordinis ademptio (cf. C. I. L. 1.206, 50.121).

(5.) Ignominiosa missio--public drumming-out before the whole army. See the striking scene in Bell. Afr. 54, 4. The soldier thus cashiered could not appear at Rome or at the emperor's court (neque in sacro comitatu, Dig. 49, 16, 13, 3).

To these add--

(6.) When whole troops were in fault, they were often given barley instead of wheat for food (a punishment inflicted on individuals also, Veg. 1.13), and made to bivouac outside the camp (Plb. 6.38, 3): cf. Liv. 10.4, 4; Suet. Aug. 24.

(7.) Decimation [DECIMATIO].

(8.) Capital punishment, which could only be inflicted by the consul during the Republic (Dionys. A. R. 11.43), or by the emperor or legatus legionis during the Empire (D. C. 52.22, 3). It was the penalty for desertion (desertio), excessive insubordination, secret correspondence with the enemy, or loss of standards (Dionys. l.c.; Joseph. B. J. 3.5, 7; Tac. Ann. 13.36 ; Dig. 49, 16, passim). It was frequently inflicted in the terrible manner called fustuarium [FUSTUARIUM]. (With regaro to desertion, it may be parenthetically noted that the desertor differed from the emansor, in that the former intended to stay away and was brought back, while the emansor either came back of himself or at any rate intended to come back--the same distinction as subsisted among slaves between fugitivus and erro.）

4. Commissariat.

Not the least important function of the general was to see that the army had sufficient provisions. When the commissariat had been properly seen to, the campaign or expedition used to begin (cf. Caes. Gal. 1.37, 5; 2.2, 6, &c.). Each foot-soldier in the time of Polybius (6.39, 13) used to get per month two-thirds of an Attic medimnus of wheat; each Roman horse-soldier, 7 medimni of barley and 2 medimni of wheat; each horse-soldier of the allies, 5 medimni of barley and 1 1/3 medimni of wheat. The extra supply for the horse-soldiers was for the support of two servants for each Roman and one for each ally. The food of the Greeks was mostly barley (cf. Xen. An. 7.1, 37); but the Romans used more wheat, as that is more nourishing for those who have to go through a great deal of physical exertion (Galen, 6.507, ed. Kuhn). Barley was sometimes, as we have seen, given as a punishment. The soldier used to carry rations for seventeen days (Amm. Marc. 17.9, 2; Lampr. Alex. Sev. [p. 1.812]47), or perhaps in strictness for only sixteen days (see A. Langen, Die Heeresverpflegung der Römer im letzten Jahrhundert der Republik, p. 5 ff.). The corn when measured out (metiri) to the soldiers was neither ground nor baked. For the former process hand-mills (molae manuales or manuariae） [MOLA] were used, which were carried by the beasts of burden. Once ground, the meal was made into a kind of porridge (puls, Plin. Nat. 18.83); hence the Romans are jestingly called pultiphagi, or pultiphagonides (Plaut. Most. 3.2, 143; Poenul. prol. 54). Often it was baked into bread (Plut. Mar. 7; Plin. Nat. 18.12). Cocta cibaria are frequently mentioned in the times of the early Republic (Liv. 3.23, 3; 27, 3); and often reference is made to supplies for the fleet (ib. 21.49, 7; 24.11, 9; 44.35, 13). We do not hear of biscuits (buccellatum) till the Empire (Spart. Pesc. 10; Ammian. 17.8, 2). The soldiers baked their own bread (Plut. Mar. 13); and it was only in times of lax discipline that they sold their corn and bought bread day by day (Sal. Jug. 44, 5). Meat was quite a secondary article of food in the time of the Republic and early Empire (Caes. Gal. 1.48, 6): and it was considered a case of real hardship when life had to be supported on meat alone (Caes. Gal. 7.17, 3; Tac. Ann. 14.24). It is only in one case, and that when sufficient corn could not be supplied, that Caesar made a requisition on the allies for meat (B.C. 1.52, 4). Vegetables (legumina) were also occasionally used in case of an insufficient supply of corn (ib. 3.47, 7); and of course salt was always given (cf. Plut. Crass. 19). For drink we hear of vinegar being supplied to make the draught called posca (Spart. Pesc. 10; Hadr. 10). Wine was often served out (Veg. 3.3; cf. Plut. Ant. 41), though not by strict disciplinarians (Spart. Pesc. 7), and it was often obtained in great quantities when discipline was relaxed (Sal. Jug. 44). We know that the price of the corn doled out to the Roman soldier was deducted from his pay (which was called ὀψώνιον by Polybius, 6.39, 12). Langen (op. cit. p. 13) calculates that it amounted to about 40 denarii a year. Anything in the shape of luxuries was supplied by the sutlers (lixae), who used to follow the army. They were so subversive of discipline that real soldierly generals used frequently to banish them from following the camp (App. Hisp. 85; Sall. Jug. 45). The army used often to be supported by quartering it in towns, and the extravagant abuses of this practice in the last century of the Republic are insisted on by Cicero (pro Imp. Pomp. 13, 38; cf. Plut. Sull. 25; Caes. B.C. 3.5, 2; 31, 4). The frequent practice of Caesar in his wars was to lay a requisition for corn on tributary states, making the magistrates responsible (B. G. 1.16, 6; 48, 2; 2.2, 5, &c.). Indeed, the great Roman skill in organising is shown by the way Caesar was generally so well supplied with provisions, and also by the arrangements by which the large armies of Pompeius (Caes. B.C. 3.5) and Brutus and Cassius (App. BC 4.100) were kept in supplies. These supplies were of the nature of a forcible requisition, and there is no evidence of their having been paid for (Langen, op. cit. p. 27). The war, as Cato used to say (Liv. 34.9, 12), should support itself. Magazines (horrea) were erected for the supplies to be transported to (Caes. B.C. 3.42, 4). Vesontio, for example, was one (B. G. 1.39, 1). In the provinces of the Empire, too, there were large magazines (Tac. Agr. 19), whither the provincials brought the corn supply (annona militaris, Veg. 3.3). One of the accounts of the supplies in the imperial magazines tells us that they consisted of vinegar, corn, suet (larido), barley, and straw (Capit. Gord. 28): another that they consisted of wine, vinegar, biscuits, suet, meat (Cod. Theod. 7.4, 6), probably pork. The duty of transport (vectura) lay on the magistrates of the tributary towns (Caes. B.C. 3.32, 2). There is only one case in the later Republic of the supply of the army being dependent on merchants (App. BC 4.108), though contracts for military supplies are mentioned in the early history (e. g. Liv. 23.49, 10). Foraging was also a frequent means of getting provisions. For details on this and other points connected with the provisions of the army, see Langen, op. cit.; for foraging especially, pp. 21-23.

Among the principales we have reckoned (p. 803 a) the clerks of the stores (horrei). We may add the pecuarius (C. I. L. 8.2553) and venatores (Eph. Epigr. 4.526), who looked after the procuring of the meat, and the lanii, who saw to the killing of it (Dig. 50, 6, 7 (6)). For further, cf. Arnold Langen, Die Heeresverpflegung der Römer, 1878-1882.

5. Associations in the Army.

Belonging to every cohort of the legion, under charge of the signifer and subordinate clerks (librarii depositorum), was a follis, into which one-half of all extraordinary grants of money, such as donativa, were paid by the soldiers, and probably also what additional sum each soldier might wish to deposit. There was also a burial fund in each legion, to which every soldier contributed a trifle (Veg. 2.20). We also find kinds of friendly societies formed by certain classes of the principales. Such a society, as well as the building in which it met (Wilmanns in C. I. L. 8.2552 b; Comm. in honorem Mommseni, p. 200, note 62) was called scola (Eph. Epigr. iv. p. 146). None is more interesting than in C. I. L. 8.2557, the scola of 36 cornicines of Legion iii. Aug. at Lambaesis. Each member had to pay an entrance fee (scamnarium--his “sitting” apparently, not his “footing” ) cf 720 denarii. If he is ordered across the sea, he gets 200 denarii; when he becomes a veteran, he gets an anularium (whatever that means) of 500, if promoted 500, if degraded (quod abominamur) 250, if he dies his heirs or executors are to get 500--all apparently on condition that his full liability has been paid up. The scola of optiones in C. I. L. 8.2554 has a quaestor.

For military armour and weapons, see CALIGA, GALEA, GLADIUS, HASTA, LORICA, OCREA, PILUM, SCUTUM. For Dress, Military Ensigns, Military Engines, see the articles cited in the Index.

(Besides various articles on special points connected with the army which are scattered throughout German periodicals, especially those by Mommsen in the Ephemeris Epigraphica and Hermes, there are systematic treatises on the Roman army in Lange, Historia mutationum rei militaris Romanorum, 1846; Madvig, Die Verfassung und Verwaltung des röm. Staates, ii. [p. 1.813]465-579; Marquardt, Röm. Staatsverwaltung, ii.2 319-612; H. Schiller, in Iwan Müller's Handbuch der klassischen Alterthumswissenschaft, vol. ii. In this latter work there is a most extensive and well-nigh complete bibliography of books, monographs, and articles bearing on Roman military affairs.)

[L.C.P]

1 Müller (Dorians, book 3.100.3, $ 7) talks of a πόλις distinct from these κῶμαι. But the latter were certainly not mere suburbs, but component parts of Sparta itself (comp. Paus. 3.16, $ 9). Haase (l.c.) speaks of five divisions of the city besides Pitana, so that the six morae or lochi in the sense of Thucydides corresponded to these six divisions. For the arrangement, there seems no authority, except the statement of the scholiast on Aristoph. (Lys. 453), that there were six lochi at Sparta, five of which he names, one of the names being corrected conjecturally by Müller to Μεσοάτης. But there seems here little more than a confused version of the division into six morae.

2 Mommsen in Hermes, 14.25-30

3 Mommsen in Eph. Epigr. 5.142-154).

4 Mommsen in Eph. Epigr. 4.370-1

5 Henzen in Bull. dell' Instit., 1884, pp. 21-29

6 marked in Inscriptions 〉 for Ↄ = 100: cf. P as the symbol for the ἑκατοντάρχης in Greek.

A Dictionary of Greek and Roman Antiquities. William Smith, LLD. William Wayte. G. E. Marindin. Albemarle Street, London. John Murray. 1890.

Greek Display:
Arabic Display:
View by Default:
Browse Bar:

A Dictionary of Greek and Roman Antiquities (1890) William Smith, LLD, William Wayte, G. E. Marindin, Ed.