The Military Reforms of the Emperor Diocletian

Throughout the course of the third century A.D., the Roman Empire was subjected to repeated attack, both by Germanic tribes and by the armies of a resurgent Persia under the militaristic Sassanid dynasty. Compounded by an ever-deteriorating internal state of affairs, the result of frequent civil conflict, Roman military efficiency declined so appreciably that the once impregnable fortified frontier, the limites, were frequently pierced, exposing the interior of the empire to attack and plunder. By the time of Diocletian’s accession as emperor, four great problems had arisen in the military sphere: (1) deteriorating discipline within an army committed to self-aggrandizement, (2) deficient military capability of politically appointed commanders, (3) inadequate tactical mobility of army units, and (4) divestment of the limites to provide contingents for emergency field armies.

Diocletian undertook to resolve these problems by a bold reorganization of the military—the first significant revision of the army since the time of Augustus three centuries before and the greatest alteration of the theoretical basis of Roman military practice ever undertaken. The reforms were effective in curtailing both external aggression and internal disintegration. Moreover, the later military revisions of Constantine were not original but rather continued in logical fashion the reforms of Diocletian, laying the foundation of a military system and methodology that extended beyond the end of antiquity and throughout the Middle Ages.

The extensive military reforms and accompanying frontier reorganization accomplished by Diocletian originated from a desire to maintain a strong and internally secure empire through a defensive system that would be effective against external aggression. The basic principle of Roman military strategy since the time of Augustus, the defense of the frontiers, was retained. However, in order to administer the greatly increased number of troops, the organization of the structure was so transformed that it may be considered a new system.

The concept of frontier defense had for four hundred years been a basic and essential part of the Roman military system.1 The purpose of Diocletian’s frontier reform was twofold: first, to inspire enemies outside the empire with the fear of defeat;2 second, “to create a more effective system of frontier defence by a more detailed distribution of the available garrison troops.”3 In an effort to satisfy these designs, basic reforms of fortification and troop placement were instituted, but additional reforms were necessary for the maintenance of those basic reforms.4

The fortifications along the limites required extensive reconstruction. In former times frontier fortifications were designed in concert with the principle of “the overwhelming striking-power of the Roman army, and fortifications had been confined to purely military structures such as forts and signal-towers (only a few towns were fortified), and these fortifications themselves were relatively slight—earthen ramparts, fronted with timber or later with stone.”5 Even so, the frontier of the Antonines and Severans had been pierced many times in the course of the third century by barbarian invaders. The limites had failed to serve their purpose. The frontier was to be reconstructed on a greater scale by Diocletian, even if the frontier would not constitute the cardinal principle of Diocletian’s defensive strategy.

There were far more fortifications—not only those occupied by the army (praesidia, castra or castella proper), but civil sites as well, and not merely along or close to the Rhine, but deep into the interior of Gaul. . . . These late fortifications differed materially from those of the earlier period. Earth ramparts must virtually have disappeared, their place being taken by massive stone walls, at least ten feet thick and of greater height than ever before; and the great thickness of their walls shows that signal-towers and blockhouses (burgi) too, along the Rhine and far into its hinterland, must have been loftier than in earlier times.6

The Rhine and Danubian fortifications, which had existed since the time of Hadrian, were not the only sections of limites to which Diocletian turned his attention, but the entire perimeter of the empire was refortified in various ways.7 The empire was beset by enemies on all its frontiers. Those areas where no system of fortification existed required the immediate erection of defensive structures. Although the frontiers varied greatly in their terrain and particular needs, Diocletian utilized the same principle of defense for all.8

In brief, Diocletian’s strategy was to maintain the limites through colonies of limitanei (troops manning frontier forts) charged with the first line of defense of the empire. This thin cordon of militia was supported by cavalry units stationed at key strategic points behind the front lines. The regulars, or ripenses, were organized in legionary concentrations at major centers, some distance behind the cavalry network, thereby constituting the major striking force of the system.9 With minor modification, such an organization could be and was applied to areas devoid of natural boundaries such as the eastern frontier, or was used to strengthen the natural boundaries of the Sahara in Africa or the Rhine and Danube in Europe. The plan and the structures were the same.10

Diocletian took great personal interest in the development of the frontier fortifications. Lactantius characterizes him as possessed with a compulsion for constant construction: “to this there was added a certain limitless desire of building, and for supplying all the workers, craftsmen, carts, and whatever was necessary for constructing the works.”11

William Seston cites Theodor Mommsen’s observation that Diocletian personally supervised the reoccupation of several sections of the Danubian limes (frontier fortification) that had been lost to the barbarian tribes. He was present at the conquest of the Rhaetian limes in 288, at the reoccupation of the “ripa samarticae [north bank of the Danube occupied by Samartic barbarians]” on the far bank of the Danube in 294. In 289 and again in 300 he traveled to Antioch for the reconstruction of the Syrian limes.12 Diocletian’s involvement in this process was clearly more extensive than that of any of his predecessors.13

The empire featured five primary defensive areas: the African, Syrian, Danubian, Rhine, and the British limes. Diocletian began the reorganization of the Rhine limes immediately after his accession. Maximian was sent to Gaul to reestablish order and begin the reoccupation of this limes. He established a Frankish buffer state on the far bank of the Rhine and constructed a line of forts in advance of the Rhine from Mayence to the sources of the Danube.14 In 294 Constantius extensively augmented the Rhine fortifications to provide increased security for Gaul so that he might undertake the invasion of Britain without fear of barbarian incursions. He constructed a second line of fortifications on the other bank of the Rhine, behind the first, so that a double line of fortification existed between Basel and Constance.15

Whereas the defense of the Rhine was essentially a process of refortification along a natural boundary, the fortification of the Syrian limes resulted in the original development of the threefold, in-depth defense system described above, which was later adapted to other frontiers. Because of the ever-present Persian menace, Diocletian instituted new eastern frontier defenses immediately after the beginning of his reign.16 The greater part of the system was constructed in 287 and 293, adding to a small northern section erected in 297 after the victory of Gallienus over the Persians.17 In the eastern desert there was no natural boundary such as the Rhine or Danube that would encourage the construction of a continuous vallum (wood and earthen rampart), which could, if built, be easily pierced by the mobile Persian army. Relying on what natural fortifications existed on hills, plateaus, and other strategic points, Diocletian constructed a series of forts across the desert. Manned by limitanei, these constituted his first line of defense in the east. The towns behind the line were occupied by cavalry and legionaries who served as the second and third lines of defense, protecting strategic roads and areas.18

The African defense system, prior to the rule of Diocletian, had consisted of a series of castella (fortified guard posts), which served as desert watch stations, and of a structure called the fossatum (ditch and earthen ramparts). During the course of Diocletian’s reign, this system was strengthened and modified in such a way that it came to resemble the Syrian limes. The fossatum became a continuous rampart employed as the boundary of the Roman frontier as well as a part of the defensive system. Denis Van Berchem describes the fossatum as four to ten meters high with an earthen rampart topped by a wooden wall.19 It was not large enough to serve the same defensive purpose as the great walls of Europe and was far too extensive to be completely garrisoned. Indeed, because of the presence of fortification works and other structures in front of the fossatum, it must be concluded that in this system the rampart constituted not the first defensive position, but rather a last line of defense.20

Diocletian’s treatment of the Danubian defenses was much like that of the Rhine limes. It was his policy to reoccupy and refortify the abandoned limes while strengthening the system with secondary defensive positions of cavalry and legions. Moreover, on the right bank of the Danube new castella were constructed in the same way that they had been placed in front of the African fossatum.21 But what is most significant, all took the form characteristic of the Diocletianic defenses in other areas.22

The improved and expanded frontier defense system occasioned the introduction of other reforms. The greatly extended defenses needed to be properly garrisoned, thus necessitating a tremendous troop increase, which in turn was administratively effected through a complete provincial reform. It must, however, be remembered that even though these other reforms came as a result of the original reform, together they gradually and simultaneously developed and were dependent on each other.

H. M. D. Parker asserts that “on his accession [Diocletian] found some forty legions in existence and, although certainty is not attainable, it seems probable that he raised their total to about sixty.”23 Seston generally agrees that during the twenty-five years between 280 and 305 the number of legions throughout the empire rose from 39 to 59 or 60.24 A. H. M. Jones adds a further corroboration of this remarkable troop increase: “The evidence suggests that the army was approximately doubled between the Severan period and the reign of Diocletian, and that the greater part of the increase was due to Diocletian himself.”25 The increase was gradually realized through the application of the provincial administrative reform:

These new units were not raised simultaneously, but at intervals, to meet the requirements of the new provincial organization, which Diocletian gradually established throughout the Empire. As the smaller frontier provinces were created out of the old larger administrative units, each was given its own garrison troops. The principle of distribution was in accordance with the practice of the third century, namely, the assignment of a pair of legions to each province.26

This, of course, provided for the assignation of a greatly increased number of legions.

Little precise factual information is available regarding Diocletian’s recruiting procedures to increase the number of troops. J. B. Bury concludes that the manpower was recruited from four sources: the sons of soldiers, serfs, barbarian settlers, and adventurers.27 Seston believes that recruiting was accomplished through the application of a quota system to the regions of the empire in much the same way that taxes were collected.28

E. C. Nischer has proposed the theory that the legions raised by Diocletian were composed of a greatly reduced complement of only one thousand men. If this supposition is correct, the view expressed above that part of the motive behind provincial reform was to provide additional administrative supervision for an increased number of legions, two to a province, should be reexamined.29 Nischer’s opinion is opposed by that of several other historians.30 Van Berchem regards this estimate of legion strength as merely a hypothesis. In a question where no clear and definite evidence is available, Jones offers a sound analysis that

The legion of the principate numbered about 6,000, and there is good reason for believing that the new legions which Diocletian raised were of the same strength. They, like the old legions, later contributed detachments to the comitatus, and in those areas, the Danubian provinces and Egypt, where the frontier legions were later broken up into a number of detachments, the Diocletianic legions were divided in the same way as were the older legions.31

No primary evidence exists to illuminate the problem of legion strength in the Diocletianic era, either pro or con. However, without any reputable indication that the size of the legion was altered by Diocletian (whose increase of the size of the army is definitely asserted by Lactantius,32 seeming at least to imply his interest in the maintenance of a large army, and thus explaining the increased number of legions of substantial troop strength), we must assume that the legion is composed of its customary contingent of six thousand soldiers. Accordingly, for the present, no argument can be offered against the interpretation of troop increase presented above.

The military detachments of the empire were of differing classification and role. Van Berchem explains the function of various kinds of troops employed in Diocletian’s frontier system, with limitanei garrisoning the frontier and cavalry and infantry legions stationed behind the frontier at key strategic points, under the command of the military governor, or dux.33

The first line of defense was provided by the increasingly militialike limitanei who, according to the Codex Iustinianus, not only defended the forts and settlements of the frontier but also engaged in agricultural cultivation of land to encourage settlement of the frontier regions by civilians.34 These were soldiers who functioned in a fashion similar to the ancient auxilia (auxiliary troops of non-Romans) with the dual responsibilities of defense and settlement; their military function is described by Van Berchem as ensuring Rome’s access to the frontier through holding strategic routes and points.35

The limitanei were supported in their function by the equites. These cavalry units were organized in independent tactical units called vexillationes, each vexillatio most likely being composed of five hundred men.36 It was their function to provide accessible routes of communication from their garrisons strategically positioned behind the frontier and to check any incursions from across the frontier. The primary fighting force of a province consisted of the two legions assigned to it. Their role was strictly one of support, except in times of conflict when it was their responsibility to engage and destroy the enemy. These first-class troops are sometimes designated by the term ripenses. By A.D. 325 the term was applied to the regulars stationed in the provinces as opposed to the comitatenses or regulars serving with the comitatus or field army. The term ripenses (riparienses) derives from the Latin ripa and, as its name denotes, relates to the defense of a riverbank37 and was used in reference to certain troops in the Danube area as early as the reign of Aurelian.38 Indeed, it was in that area, the boundary of which is the Rhine and the Danube, that these troops were most commonly located.39

The command of the provincial garrisons had been exercised throughout the history of the empire by the provincial governor. It should, however, be remembered that from the time of the reign of Gallienus a tendency had existed toward the separation of military and civil powers and, moreover, that in those provinces still governed by a senatorial governor, a praepositus, rather than the governor or praeses, often exercised the command over troops. Under Diocletian, however, governors who were of equestrian rank still retained the function of command. Nevertheless, inscriptions dating from the early years of Diocletian’s reign attest to the fact that military operations were still occasionally performed by the praeses.40

The separation of the military and civil powers was carried to its conclusion under Diocletian. J. G. C. Anderson informs us that

    In the later years of Diocletian’s joint reign the principle of divided authority was applied almost universally. By A.D. 304—5 the separation of military and civil functions was normal: only in such provinces as Mauretania and Isauria, where conditions were disturbed, were the two functions still combined, and there the combination was permanent.41

The military command of the governors was superseded by the creation of a new officer—the dux. “The appointment of an equestrian military officer as general commander of the troops of a province or of more than one province, the dux provinciae or dux limitis provinciae,” was “a natural sequel”42 after the separation of military from civil powers. Jones asserts that “In the system of command Diocletian introduced one innovation, establishing in certain frontier areas zone commanders (duces) distinct from the provincial governors, who retained civil functions only. This change however was far from universal: in many areas the provincial governor continued to command the local forces as heretofore.”43

The origin of the dux is rightly attributed to Diocletian. The initial existence of the office in the time of his reign is certain. The first literary mention of Diocletian’s new military commanders occurs in the early part of his reign when the author of the second panegyric refers to those occupying the new office of dux as preserving the glory of Diocletian. Regarding this account, Norman H. Baynes explains that “there is no undoubted reference to the later provincial military commander—the dux—until the year 289.”44 The existence of the dux is further attested by several inscriptions dating to the period of the tetrarchy.45

Nevertheless, the question may be raised whether the dux originated before the time of Diocletian. Duces are mentioned in the passage from Aurelius Victor describing the accession of Diocletian.46 It may, however, be argued that Aurelius Victor, writing in the late fourth century, applies the terminology of his day to a situation that had existed almost a century before. A further argument for the early origin of the dux cannot be so easily dismissed, however. An inscription found at Verona, dating to the year 265, records that a certain construction project was undertaken at the direction of one “Aur. Marcellino, v.p., duc. duc.”47 Mommsen expanded “duc. duc.” into “duce ducenario.”48 If his interpretation is correct, the office of dux must have existed, at least in some form, before Diocletian. Baynes attempts to account for the use of the term dux by citing Homo’s argument that Aurelius Marcellinus was not a dux assigned to the command of the troops of a province, but simply a director of a particular project.49 Clinton W. Keyes, on the other hand, disagrees with Mommsen’s rendering of “duc. duc.”; as he explains, “the stone-cutter may have repeated these letters by mistake [i.e., duc. duc.], and the man’s title may have been simply v. p. ducenarius, a title which is found in another inscription of a little later date (CIL, III, 1805).”50 Of course, there is no good reason to expect a dux limitis to have served at Verona. “It seems much more likely that Marcellinus was a financial officer.”51

The command of the regular troops of the African provinces was not entrusted to a dux but to an equestrian officer with the title praepositus/limitis. It is possible that this officer had originally been placed in command of troops in the old province of Africa, a consular province and thus governed by a senator whose title was later retained by officers who exercised the same functions as the dux in other provinces. Seston attributes the widespread employment of the praepositus within the African provinces to a general reorganization of the area effected by Maximian at the conclusion of his African expedition and campaign of 198.52

It was the function of both the dux and the praepositus to command troops and maintain defensive structures.53 However, regarding military command in the provinces, Jones explains that “in most of the provinces, which were ungarrisoned, the governor had civil functions only. In some [provinces] which had garrisons Diocletian separated the military command from the civil government, but this was by no means a universal rule.”54 Furthermore, as we are informed by Van Berchem, it was the provincial governor who exercised the command of the limitanei of the province, while the dux might command the legions and cavalry.55

The lack of uniformity in the system and the occasional overlapping and reduplication of the duties of various commanders, whether duces, praepositi, and praeses, may be accounted for by the fact that “the re-organisation of the frontier provinces was not carried out simultaneously,”56 but that “the policy was applied at different times during the reign, as occasion dictated, as the circumstances of particular territories pressed themselves on the attention of the Emperors.”57 Consequently, the reform of frontier defense, the military, and the administrative system occurred both simultaneously within the individual provinces of the empire, and gradually from province to province. As Parker writes, “The division of the old provinces into smaller units, which further weakened the power of their governors, was carried out in gradual stages and the completed scheme is reproduced in the provincial list of Verona.”58

Some may have held that the sole motive of the provincial reforms was to continue the reduction of the power of the governors and, by so doing, decrease the likelihood of civil strife. Although this was certainly an effect of the reform, the purpose for which it was undertaken was to provide effective leadership for the armies.59 Jones supports the military interpretation, arguing that “some at any rate of the new duces commanded far larger concentrations of troops than any provincial governor had had at his disposal for generations.”60 Parker concurs with Seston’s analysis of the motive, declaring that

although the primary purpose of this new territorial delimitation was doubtless to facilitate the work of administration, it is worthy of note that in each case the re-organisation appears to have followed a period of extensive military operations. This suggests that at least a secondary motive for the change may have been the creation of a more effective system of frontier defence. For the subdivision of provinces situated in a danger zone would, if each province had its own garrison legions, provide a stronger and more intensive resistance to hostile attacks from beyond the frontier.61

The civil (and in a few cases military) administration of the province was accomplished by its governor. In the appointment of governors, Diocletian continued Gallienus’s policy of assigning provinces to equestrian governors, to such an extent that under his rule senatorial governors were almost eliminated. Jones affirms that there “were still a few proconsuls and legates, . . . and only the two consular proconsulships of Asia and Africa, both greatly reduced in territory, still survived, and were still filled by senators.”62 However, these two governors “were responsible directly to the Emperor and not to the praetorian prefect.”63 Jones writes that “all the other provinces . . . were governed by equestrian praesides.”64 Although the empire was divided into four administrative sections under the tetrarchy, the continual division of provinces and the formation of new provinces undoubtedly represented a great strain on the administrative machinery of the imperial government. Parker suggests that “in order to facilitate the control of the provincial governors by the central bureaucracy, Diocletian introduced a new and more comprehensive scheme of internal organization. The Empire was divided into twelve dioceses.”65 Provinces were grouped together to form the various dioceses, “each of which was directed by a deputy of the praetorian prefects, vices agens praefectorum praetorio, or vicarius for short.”66 The establishment of the system of vicars weakened the praetorian prefects,

whose representatives they were, and appeals from their decisions were heard not by the prefects, but by the Emperor himself. By this check upon the monopoly of power in the hands of his chief ministers, Diocletian sought to protect himself from a repetition of the military pronunciamentos [declaration of troops to elevate general to imperial rank], which in the past had so vitally sapped the strength of the Empire.67

The vicar not only served as a check on the prefects but also as a restraining force on provincial governors.68

The provincial and diocese divisions of the empire are given in the “Verona List.” Jones’s research and observations on the value of the list demonstrate its accuracy. Based on the known dates of creation for provinces appearing in the list, Jones dates the document several years after the abdication of Diocletian, writing, it “cannot be earlier than 312, the earliest possible date for the creation of Aegyptus Herculia, nor later than 320, the latest possible date for the amalgamation of Numidia Cirtensis and Militiana.”69

The administrative organization of the regular troops, cavalry, and infantry, under the command of the duces corresponded in part to the new provincial organization. The empire was divided into military districts, each administered by a dux. Several provinces were grouped together to form a military district. The Eastern Prefecture was divided into six military districts that, with the inclusion of the Egyptian district, constituted the organization of the empire directly under Diocletian’s supervision.70 In Galerius’s administrative portion of the empire, seven duces commanded troops along the Danubian limes.71 The western half of the empire was placed under the military supervision of thirteen duces.72 With the exception of the field armies maintained by the emperors near their persons, the entire military strength of the empire was stationed along the frontiers and commanded by duces, as they were placed in the military districts.

A different interpretation of the military administration of Diocletian has been proposed by Nischer. In addition to the frontier reorganization, he contends that the reforms “consisted in the institution of a number of independent bodies of divisional and main reserves which were not tied to any particular point of the frontier, but which possessed a greater freedom of action and could even be employed at need in more distant provinces without any substantial weakening of the frontier-defence proper.”73

Nischer seems to have conjured up twenty additional legions whose existence at the time of Diocletian is unverified. He assumes that all legions of Flavian Constantinian nomenclature were originated by Constantius rather than Constantine, and that a number of other obscure legions also functioned under Diocletian. These legions are deployed in the Nischer system to fill the posts of regional and main reserves. A “Rhine reserve” is supposed to have been organized, composed of the legions I Flavia Constantiniana, II Flavia Constantiniana; a “West Danube reserve” of the legions III Herculia and IV Iovia; an “East Danube reserve” of the I and II Flavia Gemina; a “Pontus reserve” of the I and II Armeniaca; an “Eastern reserve” of the V and VI Parthica; an “Egyptian reserve” of the I Maximiana, II Flavia Constantia, and III Diocletiana; an “African reserve” of the I Flavia Pacis, II Flavia Virtutis, and III Flavia Salutis; a “Julian Alps Main reserve” of the legions I, II, and III Iulia Alpima; and an “Isaurian Main reserve” of the legions I, II, and III Isauria.74

Nischer’s theory is refuted by both Baynes and Parker. Baynes believes Nischer’s system of divisional reserves to be “unsupported”75 and Parker, after examining the origin and positioning of each legion mentioned by Nischer, concludes that his theory is “arbitrary and unsatisfactory.”76 He continues and summarizes his position in support of the Diocletianic provincial defense system as follows:

    Now since the time of Septimius Severus it had been the practice to assign not more than two legions to each frontier province, and, although in some provinces one legion was sometimes deemed sufficient, the upper limit was not exceeded. . . . This policy appears to have been continued during the third century A.D., and we find Aurelian raising the garrisons of Phoenice and Arabia to the normal strength of two legions apiece. It is surely then not improbable that Diocletian adhered to what had become a recognised principle. Certainty is unattainable; but, if we adopt the theory that Diocletian garrisoned his frontier-provinces on the basis of not more than two legions to a province with a preference for a system of pairs, then we can both account for the increase in the number of the legions and also discover the play of their distribution without having recourse to Nischer’s theory of “Divisional and Main Reserves,” for which not a shred of evidence exists.77

Thus, through programs of refortification, troop increase, reformation of the military command system, and provincial reorganization, Diocletian endeavored to provide the empire with secure boundaries.

The defense of the empire against its enemies could not be accomplished merely through the improved system of frontier fortification and administration that Diocletian had instituted. A further element of defense was required for Diocletian to realize his goal of a secure empire—an effective offensive striking force.

The central component of Diocletian’s offensive strategy was a permanent field army. Through its creation he provided the foundation for the enlarged and expanded field army whose existence during the time of Constantine is revealed in the document Notitia Dignitatum. Mommsen calls this later force “the joint creation of Diocletian and Constantine.”78

The contention that Diocletian was the architect of the field army and that Constantine was the builder who constructed the edifice on Diocletian’s foundation is challenged on the basis of the familiar passage in which Aurelius Victor describes the labor of Constantine as originating a new military order (novando militiae ordine).79 Those who oppose the theory of a Diocletianic field army hold that Diocletian was only the augmenter of the number of troops that were still administered and distributed solely in accordance with the concept of positioning them along the frontier. It is Constantine alone whom they consider worthy to be credited with the innovation of the field army.80

Nischer attempts to refute the existence of a Diocletianic field army through three arguments. He first contends that Constantine is surely the great reformer of the army and the originator of the field army in accordance with the statement of Aurelius Victor. This is essentially the line of reasoning expressed above. His second argument is that the troops that comprised the field army of Constantine are listed in the Notitia Dignitatum under titles unknown to the pre-Constantinian army. Nischer finally argues that it seems illogical to believe that if Diocletian had been the founder of the field army, he would have begun his process of reform with a tremendous increase in troop strength and then have continued to form entirely new units out of units whose formation was barely complete. Furthermore, the execution of such measures would have required a longer period of time than the reign of Diocletian.81

Nischer’s first argument must be called into question because the passage from Aurelius Victor is by no means clear in its meaning. Perhaps it may be concluded that some kind of a military reform was initiated by Constantine. However, the nature of that reform is not indicated by the historian. It may or may not have been associated with the field army. Seston goes so far as to question the reality of any sort of reform on the part of Constantine.82

Parker answers Nischer’s second objection by presenting strong evidence that the legions of Constantine’s field army were, in part, Diocletianic in origin.

These statements, I take it [i.e., Nischer], mean that in the field-armies very few legions are found which are designated by a number only, or by another title in addition to the number, and the exceptions can be explained by holding that the legions so styled date back to frontier-legions of Diocletian. Now if we take the list of legiones palatinae et comitatenses in the Notitia we find in the eastern half of the Empire two Palatine legions (Primani, Undecimani) designated by a number, nine legiones comitatenses (V Macedonica, VII Gemina, X Gemina, I Flavia Constantia, II Flavia Constantia Thebaeorum, I Maximiana Thebaeorum, III Diocletiana Thebaeorum, I Flavia Gemina, II Flavia Gemina) by a number and name, and three legiones comitatenses (Tertiodecimani, Quartodecimani, Secundani) by a number; while in the western half of the Empire we meet with one Palatine legion designated by a number (Octavani) ten legiones comitatenses (Secundani Italiciani, III Italica, III Herculia, II Britannica, I Flavia Pacis, II Flavia Virtutis, III Flavia Salutis, II Flavia Constantiniana, Tertio Augustani, III Julia Alpina) by a number and name, and one legio comitatensis (Undecimani) by a number. If we subtract from this list the eight “Flavian” legions and III Julia Alpina, whose origin we have shown to be doubtful, that leaves us with no fewer than seventeen “easily explained exceptions.”83

This information casts doubt on Nischer’s reasoning. As Parker asserts, the existence in Constantine’s field army of at least seventeen legions during the tetrarchy cannot be lightly dismissed as an “exception.” Nischer’s disregard of such weighty evidence calls his entire methodology into question.

Parker also responds to Nischer’s third query regarding Diocletian’s logic. Because small legionary vexillationes drawn from the frontier legions were used in the composition of the field army, Parker is able to explain that “the increase in the number of frontier-legions . . . was not merely a defensive measure, it was also an attempt to provide for a mobile army raised on the principle enunciated by previous emperors, . . . each in all probability 1,000 strong, drawn from the larger frontier-legions, and these, whether they had received by then the title of ‘comitatenses’ or not, are the foundation of Constantine’s field-army.”84 Nischer’s error is that he supposes Diocletian’s reforms were accomplished in accordance with an established plan. This is contrary to the indication that the reorganization of the frontier provinces was not carried out simultaneously, but as necessitated by particular military circumstances. Thus frontier legions were raised at varying times.85 We may suppose that contingents were added to the field army on a similarly gradual basis. Diocletian did not raise a huge number of troops for service on the frontier and then transform them into a field army. Here Nischer is correct. Rather, a gradual process of formation occurred both in raising frontier legions and in organizing the field army. This was not an impossible feat, nor one that could not be completed in twenty years.

It seems that the concept of the mobile field army originated with the cavalry corps that Gallienus stationed at Milan to retaliate against enemy incursions. This aggregation of cavalry cannot be considered a true field army, but was certainly its forerunner. Seston expresses a similar view.86

The field army of Diocletian consisted of a nucleus of permanent troops, augmented in times of emergency by vexillationes from the provinces. At the center of this force were the emperor’s most elite troops, organized in the sacer comitatus (sacred body of companions). The existence of this body is attested by both literary and epigraphic sources. A papyrus that treats the composition of the expeditionary force Diocletian led into Egypt to quell the rebellion of the rebel Achilleus in 29787 mentions the comitatus in several places.88 An inscription dated to the reign of Diocletian further attests the existence of the comitatus under the tetrarchy.89

Jones proposes that the “mobile forces under the immediate command of the emperor, which, since they accompanied him on his movements, were called the comitatus.”90 The comitatus was apparently composed of several different bodies of troops and comites, the companions, or court, of the emperor. A select military guard, the lanciarii, also held a place in the comitatus. Several inscriptions provide evidence of the existence of lanciarii.91 Nischer refuses to consider the lanciarii as part of the comitatus but rather considers them to be “apparently a detachment of the praetorian guard.”92 On the other hand, Parker maintains that it is virtually impossible to consider these troops as part of the praetorians, since an inscription implies that the lanciarii were especially selected for this honor (lectus in sacro comitatu lanciarius),93 and suggests that the lanciarii were troops selected to serve as a bodyguard to the Augusti and Caesars on their campaigns. They may have begun to fill this role in place of the praetorian guard when that body was limited by Diocletian to garrison duty in Rome.94

The protectores comprised another of the units included within the comitatus. Jones explains their function as that of an imperial honor guard, which status “seems to have been invented by Gallienus, who bestowed it on high-ranking officers, prefects of legions and praetorian tribunes.”95 Jones further asserts that “by Diocletian’s time there was certainly a corps of protectores which accompanied the emperor.” It is very possible that these are the same troops that are less formally called domestici and who Diocletian himself commanded before his accession.96 This detachment is mentioned in an Oxyrhynchus papyrus as being present with Diocletian on his Egyptian expedition.97 An inscription from the tombstone of a quartermaster of this unit, dating to Diocletian’s reign and found at his official residence at Nicomedia, shows that the protectores did not constitute an official unit; since they were stationed at the emperor’s residence, they were likely members of the comitatus.98

The imperial bodyguard proper was called the scholae. The existence of this unit in the comitatus of Diocletian is clearly attested by certain literary sources. The Acta Sergi et Bacchi in the Analecta Ballandiana indicate that two senior members of the Scholae Gentilium were martyred during the persecution of Galerius.99 Lactantius speaks of the rapidity with which Maximinus was promoted through the ranks to the office of Caesar, having once served as a scutarius or a member of the Scholae Scutariorum.100 It may then be postulated that at least two units of scholae, the Scholae Gentilium, recruited from barbarians and the Scholae Scutariorum, were raised among the Romans.101

The inclusion of other privileged units among the crack troops of the comitatus is very possible. Two such elite corps are the Ioviani and the Herculiani. Literary sources reveal their frequent presence with the emperor, and they are listed in the position of highest honor among the legiones comitatenses in the Notitia Dignitatum. They were probably attached to the comitatus as units originally drawn from the legions of the province of Scythia, the I Iovia and the II Herculia, to render temporary service that was, apparently, later made permanent.102

To be able to provide additional legions to augment the comitatus in times of danger and thus form an expanded field army, Diocletian increased the size of the Danubian troops. The increase in troop numbers along the Danube enabled Diocletian to draw from the frontier legions without jeopardizing the safety of any one section of the frontier by depleting legionary garrisons below a safe level. It is certain that Diocletian used this method of adding vexillationes from the frontier legions to organize the field army. This is illustrated in the Oxyrhynchus Papyrus account of Diocletian’s Egyptian expedition. In it, nine praepositi are said to be in command of the troops. Each commanded vexillationes from a pair of legions, probably sister legions assigned to the same province.103 Thus the only real change in the policy of Constantine from that of Diocletian is “the permanent concentration by Constantine of those vexillationes in an army separate from the frontier-army and their subdivision into two sections which differed from each other only in rank and distinction.”104

The infantry that formed the bulk of the expeditionary forces was raised in the method described above. Although the vexillationes were transferred from duty with their larger parent legions stationed on the frontiers, they did constitute what must be termed a field army. Like so many other innovations of Diocletian, the field army was probably formed in response to a particular military necessity occasioned by a particular problem situation. The Egyptian expedition of Diocletian is an example of such a situation.

Of the troops comprising the army that Diocletian led into Egypt, it appears that the detachments drawn from the two legions of a province acted in unison and were under the command of the same officer. At times units from sister provinces were organized in the same fashion. The Oxyrhynchus Papyrus specifically mentions troops from Upper Moesia, consisting of units of the IV Flavia and the VII Claudia and a detachment of the legion XI Claudia, which was stationed in Lower Moesia, as commanded by the same officer, one Iulianus.105 After the campaign was successfully concluded, several vexillationes remained in Egypt. They were detachments from legions originally stationed in the same province. The Notitia Dignitatum identifies the location of units from the legions V Macedonica and the XIII Gemina, originally assigned to Dacia, in northern Egypt.106

Just as Diocletian led his expedition to quell the rebellion in Egypt in 297, Maximian, in the same year, led a force of troops into Africa to end disturbances by desert tribes. Maximian’s field army seems to have been formed in the same manner as that of Diocletian, with the greater part of the force consisting of troops drawn from frontier legions. A number of inscriptions provide us with this information. The epitaph of a praetorian guard evidences the presence of detachments of this unit with the comitatus of Maximian.107

The legion XI Claudia, as indicated by another inscription, was stationed at Aquilea, which may have represented a reserve base at which the vexillationes used in the field army could await the next campaign without rejoining their parent legion.108 Because a number of the units of Diocletian’s field army remained in Egypt rather than returning to their parent legion, and because the XI Claudia vexillatio likewise did not rejoin its legion, it may be questioned whether other units also remained detached from their parent legion and continued as permanent mobile independent units.109

Cavalry was an important and integral part of the field army. According to Parker, the creation of the mobile cavalry unit of the field army, detached from service to a particular legion, was not the accomplishment of Constantine.110 As I have shown in a longer version of this paper, the formation of independent cavalry units must be attributed to Gallienus, and the separation of the legionary cavalry from their legions, with the reception of the new name promoti, to Aurelian. During the reign of Constantine, the promoti were designated by the unit title of vexillatio, which at this time began to refer only to units of cavalry and was no longer applied to infantry. However, it seems that perhaps even this new terminology originates with Diocletian instead of Constantine. Certain passages from the Codex Iustinianus indicate that such terminology was followed before 293.111 Independent cavalry units, or promoti, were utilized in the armies of Diocletian. Some were certainly incorporated in the field army and others served in frontier areas. The evidence that cavalry was used in both the comitatus as well as on the frontier is strong. An inscription from Noricum mentions a cavalry unit that had been part of the comitatus, but on completion of its term of service had returned to its station on the frontier.112

On the basis, then, of numerous literary and epigraphical evidences, it must be concluded that the mobile field army was the creation of Diocletian, gradually formed to meet the military necessities of the empire. It consisted of several special units permanently assigned to duty in the comitatus, of both infantry and cavalry vexillationes assigned on either a temporary or permanent basis. The mobile field army was a vital and important component in Diocletian’s comprehensive strategy for the defense of his empire.

The reforms of Diocletian proved successful in giving new life to an ailing empire. The organization of the tetrarchy, the refortification of the limites, the augmentation of the army’s size, the reformation of the system of military command, the reorganization of the provinces, and the inauguration of a permanent mobile field army constituted a comprehensive program of reform advanced by Diocletian. These changes occurred in response to definite and specific threats to the security of the empire and were implemented gradually in response to individual situations and events. Their effectiveness, which developed as a result of their establishment, is perhaps best described by a historian of the world.

Thanks to Diocletian’s foresight all the frontiers of the Roman Empire had been fortified in the manner already described with towns and citadels and towers where the entire soldiery lived. Thus the barbarians could not effect passage anywhere as forces would encounter them and repel invasions.113


The preceding is a chapter from a senior honors thesis written in 1975. Richard Lloyd Anderson was a member of the committee that accepted the thesis for “university scholar” designation.

1.    See William Seston, Dioclétien et la Tétrarchie: Guerres et Réformes (284—300) (Paris: De Boccard, 1946), 177; see also Graham Webster, The Roman Imperial Army of the First and Second Centuries A.D. (London: Black, 1969), 47, 88—90.

2.    See Seston, Dioclétien, 296.

3.    H. M. D. Parker, A History of the Roman World from A.D. 138 to 337, 2nd ed., revised by B. H. Warmington with additional notes (London: Bradford and Dickens, 1958), 271.

4.    This idea is supported by Seston, Dioclétien, 296:

De l´ des décisions techniques qui eurent pour but de fortifier les frontières, d’accroître les effectifs ainsi que de régler le dispositif des troupes et leur ordre de bataille. En même temps, pour assurer avec régularité l’entrentien des armées qu’on maintenait sur le qui-vive mais, autant que possible, loin des combats, il devint nécessaire de créer des cadres administratifs qui fussent assez forts pour réunir les ressources indispensables et en contrôler l’emploi, sans qu’ils pussent devenir un danger pour l’autorite impériale. Toutes ces réformes sont liées entre elles et on ne doit les séparer qu’autant que la clarté de la description l’exige.

. . . from that the technical decisions that have as a goal the fortification of the frontiers, in order to improve the effectiveness as well as regulate the disposition of the troops and their order of battle. At the same time, in order to assure with regularity the upkeep of the armies that are being kept in readiness, but, as far as possible, away from combat, it became necessary to create an administrational cadre which was strong enough to organize essential resources and control their use without becoming a danger to the imperial authority. All these reforms are interlinked and cannot be separated beyond the limitations of their job description.

5.    Franz Oelmann, “The Rhine limes in Late Roman Times,” in Congress of Roman Frontier Studies 1949, ed. Eric Birley (Durham, England: Durham University Press, 1952), 81.

6.    Ibid.

7.    See Seston, Dioclétien, 296.

8.    See ibid.

9.    See ibid., 178.

10.    See ibid., 297—98.

11.    Lactantius, De mortibus persecutorum 7.8.

12.    Seston, Dioclétien, 296—97.

13.    See ibid., 298.

14.    See ibid., 105. Mamertinus, Panegyrici Latini 7.1—7.

15.    See Seston, Dioclétien, 106. Mamertinus, Panegyrici Latini 2.6—7, 5.1—3. R. G. Collingwood, “Reviews and Notices of Recent Publications,” Journal of Roman Studies 18 (1928): 241.

16.    See Seston, Dioclétien, 167.

17.    See ibid., 163. Sextus Aurelius Victor, 39.33—37. Mamertinus, Panegyrici Latini 2.7.

18.    See Seston, Dioclétien, 178.

19.    See Denis Van Berchem, L’Armée de Dioclétien et la réforme constantinienne (Paris: Geuthner, 1952), 43.

20.    So also concludes Van Berchem, L’Armée de Dioclétien, 44. Van Bercham’s view is well corroborated by a description of the fossatum cited by him from Codex Theodosianus, 7.15. Van Bercham, L’Armée de Dioclétien, 43 n. 2.

21.    See Seston, Dioclétien, 131.

22.    See Van Berchem, L’Armée de Dioclétien, 98.

23.    Parker, History of the Roman World, 271.

24.    See Seston, Dioclétien, 298.

25.    A. H. M. Jones, The Later Roman Empire 284—602 (Baltimore: Johns Hopkins University Press, 1986), 1:60.

26.    Parker, History of the Roman World, 271.

27.    See J. B. Bury, The Invasion of Europe by the Barbarians (New York: Russell and Russell, 1963), 39—40.

28.    See Seston, Dioclétien, 302; cf. Jones, Later Roman Empire, 1:615; Lactantius, De mortibus persecutorum 7.2—5.

29.    See E. C. Nischer, “The Army Reforms of Diocletian and Constantine and Their Modifications up to the Time of the Notitia Dignitatum,” Journal of Roman Studies 13 (1923): 8—11.

30.    See Van Berchem, L’Armée de Dioclétien, 111.

31.    Jones, Later Roman Empire, 1:56—57.

32.    See Lactantius, De mortibus persecutorum 7.5.

33.    See Van Berchem, L’Armée de Dioclétien, 22.

34. See Codex Iustinianus, 1.27.2, cited in Van Berchem, L’Armée de Dioclétien, 20.

35.    See Van Berchem, L’Armée de Dioclétien, 21.

36.    See Jones, Later Roman Empire, 1:56.

37.    See Van Berchem, L’Armée de Dioclétien, 89.

38.    See Flavius Vospicius Syracusius, Divus Aurelianus 38.4, Scriptores Historiae Augustae, 3.270; J. Frank Gillian, “The Dux Ripae at Dura,” Transactions and Proceedings of the American Philological Association 72 (1941): 157—75.

39.    See Van Berchem, L’Armée de Dioclétien, 89.

40.    See Hermann Dessau, Inscriptiones Latinae Selectae (hereafter ILS) (Berlin: Weidmann, 1892), inscriptions 627—28.

41.    J. G. C. Anderson, “The Genesis of Diocletian’s Provincial Reorganization,” Journal of Roman Studies 22 (1932): 30.

42.    Ibid.

43.    Jones, Later Roman Empire, 1:608.

44.    Norman H. Baynes, “Three Notes on the Reforms of Diocletian and Constantine,” Journal of Roman Studies 15 (1925): 200.

45.    See Dessau, ILS, 664, 4103.

46.    See Sextus Aurelius Victor, 39.1.

47.    Corpus Inscriptionem Latinarum (hereafter CIL), V, 3329.

48.    Cited in Baynes, “Three Notes,” 200.

49.    See Homo, as cited in Baynes, “Three Notes,” 200.

50.    Clinton W. Keyes, The Rise of the Equites (Ph.D. diss., Princeton University, 1915), 51 n. 2, cited in Baynes, “Three Notes,” 200.

51.    Ibid.

52.    See Seston, Dioclétien, 120; cf. Eutropius, Breviariaum ab urbe condita 9.23; Sextus Aurelius Victor 39.22; CIL, VI, 1130.

53.    See Seston, Dioclétien, 308; also see Mamertinus, Panegyrici Latini 2.3—4.

54.    Jones, Later Roman Empire, 1:43.

55.    See Van Berchem, L’Armée de Dioclétien, 23.

56.    H. M. D. Parker, “The Legions of Diocletian and Constantine,” Journal of Roman Studies 23 (1933): 184.

57.    J. B. Bury, “The Provincial List of Verona,” Journal of Roman Studies 13 (1923): 128.

58.    Parker, History of the Roman World, 263.

59.    See Seston, Dioclétien, 243.

60.    Jones, Later Roman Empire, 1:45.

61.    Parker, “The Legions of Diocletian and Constantine,” 177.

62.    Jones, Later Roman Empire, 1:45.

63.    Parker, History of the Roman World, 264; compare Dessau, ILS, 1213, 1217, 1220, 1227.

64.    Jones, Later Roman Empire, 1:45.

65.    Parker, History of the Roman World, 264; Lactantius, De mortibus persecutorum 7.

66.    Jones, Later Roman Empire, 1:47.

67.    Parker, History of the Roman World, 265.

68.    See Seston, Dioclétien, 340.

69.    A. H. M. Jones, “The Date and Value of the Verona List,” Journal of Roman Studies 44 (1954): 22.

70.    See ibid., 23—24; Notitia Dignitatum (Register of Dignataries), in Translations and Reprints from the Original Sources of European History 6/4 (n.d.): 17.

71.    See Notitia Dignitatum (Or. 39—43).

72.    See Notitia Dignitatum (Occ. 1).

73.    Nischer, “Army Reforms,” 6.

74.    See ibid., 8.

75.    Baynes, “Three Notes,” 203.

76.    Parker, “The Legions of Diocletian and Constantine,” 175.

77.    Ibid., 177—78.

78.    Cited in Baynes, “Three Notes,” 204.

79.    See Sextus Aurelius Victor, 41.12.

80.    See Nischer, “Army Reforms,” 4.

81.    See ibid., 4, 12—13; Parker, “The Legions of Diocletian and Constantine,” 182.

82.    See Seston, Dioclétien, 304.

83.    Parker, “The Legions of Diocletian and Constantine,” 182—83. Cf. Notitia Dignitatum, 17.

84.    Parker, “The Legions of Diocletian and Constantine,” 184.

85.    See ibid.; Baynes, “Three Notes,” 203.

86.    See Seston, Dioclétien, 305; Sextus Aurelius Victor, 39.22.

87.    See Eutropius, Breviariaum ab urbe condita 9.23; Sextus Aurelius Victor, 39.23.

88.    See The Oxyrhynchus Papyri, part 1, ed. Bernard P. Grenfell and Arthur S. Hunt (Boston: Kegan Paul, Trench, Trübner, 1898), 90 (col. II, line 17).

89.    See Dessau, ILS, 2781.

90.    Jones, Later Roman Empire, 1:52.

91.    See Dessau, ILS, 2045, 2782.

92.    Nischer, “Army Reforms,” 55; see Parker, “The Legions of Diocletian and Constantine,” 185; Josephus, Bellum Iudiacum 3.6.2.

93.    See Dessau, ILS, 2781.

94.    See Parker, “The Legions of Diocletian and Constantine,” 185.

95.    Jones, Later Roman Empire, 1:53; Sextus Aurelius Victor, 39.1.

96.    See Jones, Later Roman Empire, 1:53.

97.    See Oxyrhynchus Papyri, 90 (col. II, line 7).

98.    See Dessau, ILS, 2779.

99.    See Jones, Later Roman Empire, 1:54.

100.   See Lactantius, De mortibus persecutorum 19.6.

101.   See Codex Theodosianus, 17.9.389, in Jones, Later Roman Empire, 2:1076 n. 33.

102.   See Zosimus, Historia Nova 2.42, 3.30; Notitia Dignitatum, 9.

103.   See Parker, History of the Roman World, 271—72.

104.   Parker, “The Legions of Diocletian and Constantine,” 183.

105.   See Oxyrhynchus Papyri, 93—94 (col. V).

106.   See Notitia Dignitatum, 17.

107.   See Dessau, ILS, 2038.

108.   See Dessau, ILS, 2332.

109.   See Parker “The Legions of Diocletian and Constantine,” 186; CIL, V, 895.

110.   See Parker, “The Legions of Diocletian and Constantine,” 187—89; Dessau, ILS, 569.

111.   See Codex Iustinianus, 10.55.3; 6.64.9, in Parker, History of the Roman World, 272.

112.   See CIL, III, 5565.

113.   Zosimus, Historia Nova 2.34, in Historia Nova: The Decline of Rome, trans. James J. Buchanan and Harold T. Davis (San Antonio: Trinity University Press, 1967), 76.