Ancient History & Civilisation

Chapter Thirteen

The Crises of 306–312

The scheme of succession organized by Diocletian lasted for only one year after his retirement, yet it was sufficiently robust that it broke down only slowly, though this was partly because Diocletian, though retired, was still alive. Within two years of his retirement in 305 three men who had been excluded from his scheme were contending for the control of the Western part of the Empire, two men who he had included were dead, and a civil war – which is what the scheme had been designed above all to avoid – was being fought. It had been, as it proved, essentially a personal creation of Diocletian’s and it was only his continued existence that ensured it lasted that long. When he was dead it was finished.

The first break came in 306, when Constantius I died at York. His eldest son Constantine was acclaimed by the army there and he was assumed to have been designated heir by Constantius. The acclamation was led by the Alamannic contingent and their king Crocus, who had been defeated by Constantius years before and then enlisted into the Roman forces. The designation as heir may or may not have happened; later it was so claimed, but it is not reported at the time.1

However it took place, whoever organized it, whatever construction was later put on the events at York and whether or not Constantius would have approved, it was nevertheless a coup d’état, one of the old sort. Constantine had no official position and his promotion had been explicitly refused in a conference of emperors held only the year before, where Diocletian had organized the succession and then he and Maximian had retired. If Diocletian’s scheme of succession was taken as the law of the land, Constantius had no legal right to designate the successor to his position in the absence of agreement with his brother emperors; the fragment of the army that was with him had no right to acclaim him, and it was certainly not the place of a tribe of barbarian auxiliaries to do so. However, going back a couple of decades, there were precedents for Constantine’s actions aplenty.

It was done, and quickly, immediately it was certain that Constantius was dead. Furthermore it was done in the midst of the whole family, including Constantine’s several half-brothers. These men must have approved and must have been consulted. This is another unusual matter, for Constantine was Constantius’ eldest son, but his half-brothers were the children of Constantius’ later marriage to Theodora, daughter of the Praetorian Prefect Afranius Hannibalianus, adopted by Maximian.

Constantine’s mother was Helena, who may not have been married to his father, and was said, in the usual way of hostile propaganda stories in politics, to have been a tavern-girl and so not of a suitable class to be married to a man of Constantius’ rank (but then all the tetrarch emperors were originally of peasant stock). Constantine’s illegitimacy was not a serious bar to his ascension to the throne. Constantius’ other sons were thus much younger than Constantine, their parents having been married no earlier than 289, and so the eldest was only a teenager in 306. The fate of the whole family at the hands of Constantius’ successor, Severus or Maximin Daia, was unlikely to be pleasant after Constantius’ death, for they could put forward claims to the Empire that many might accept. All in all, it seems most likely that Constantius himself had organized the coup before he died and that it happened like clockwork. It would certainly be assisted by taking place at York, a place as remote from the centres of imperial political power as could be found. No hostile reaction was likely for months, and at the least Constantine would be able to control the province of Britannia, just as had Carausius. Constantius was of the same generation as Diocletian and Maximian, and had waited a long time for his imperial promotion; he could also remember how things had been done in the 270s and 280s. The provinces of Britannia and Gaul, in fact, rapidly signalled their acceptance of Constantine’s enthronement, and the swiftness of all this is surely a good indication of the preliminary planning that had been done.

This was the crucial blow to Diocletian’s scheme. He had deliberately removed heredity as a determinant, preferring the selection of mature men, men of experience, who were then linked into the scheme of rulers by marriages; at least if there were daughters available. Constantine’s coup was a direct challenge to all that, though he did involve himself in the remains of the scheme by marriage later. His claim to power was a claim to succession by inheritance, and a direct refutation of Diocletian’s aims. He got away with it partly because the armies in Britain and Gaul accepted him quickly – and Carausius had shown that Britannia might hold out against a continental army if it had control of the local Roman fleet – and partly because at York he was unreachable by any other ruler, for it would take weeks for the news to reach the other centres of imperial power and further weeks for any condemnation to arrive; by then his local support was firm. In addition he was almost at once copied in seizing power by another disinherited imperial son.

Constantine’s coup in July 306 was followed by that of Maxentius in Italy in October. He had been living on an estate a day’s journey from Rome; his father Maximian, disgruntled at his compulsory retirement, went off to live on another estate in Campania or Lucania. Neither was happy at their demotion. Maxentius’ assumption of power, like Constantine’s, quite deliberately also harked back to the old methods. He was asked by the remnant of the Praetorian Guard that was still in Rome to become emperor, and he was supported by a considerable section of the Roman population, by which we may understand the plebs, and by a considerable proportion of the Senate. Two tribunes and one of the city food officials are said to have supported him; the City Prefect opposed him and was killed.2

Superficially, this looks like a reversion to the old first-century method of proclamation and succession, an impression deliberately reinforced by Maxentius’ use for a time of the outmoded title princeps – just as Diocletian had used similarly old titles – rather than either Caesar or Augustus. Closer examination suggests other reasons and motives. The reasons for the Guard’s action in seeking their own emperor were that their privileges were being reduced, their numbers were declining and they were in the process of being abolished. Similarly the support of the Romans, who are always identified as ‘the people’, was due to the imposition of taxation on the city. Parts of Italy had been subjected by Diocletian to taxation for the first time since the Republic; now the new Caesar Severus, who had Italy in his section of the Empire, was extending this to Rome itself and to southern Italy, both hitherto exempt, and the preliminary census had been taken. So the resort to Maxentius as their emperor was a way of attempting to retain their taxfree privileges, just as the Guardsmen wished to hold on to theirs. (We may note, without surprise, that in the troubles of the next few years, Maxentius was compelled to impose taxation on his tax-shy supporters.) The reasons behind Maxentius’ usurpation were thus very similar to those of many other such rebellions, notably the Egyptian risings under Diocletian only nine years before. Maxentius never had much support outside Rome and southern Italy, even or especially when his father came out of retirement later in 306 at Maxentius’ request and resumed his position as Augustus.3 In this, Maximian was also in a sense rebelling against change, the change of emperors.

Severus, who had moved up from Caesar to Augustus on the death of Constantius and whose military base was in the Balkans, marched on Rome to suppress Maxentius. He came up against the old professional Maximian, who effortlessly induced Severus’ army to change sides and then put Severus himself in prison. Technically, Severus abdicated and was soon afterwards killed, or else committed suicide. In the meantime Maxentius prevented any action being taken by Severus, Galerius or Maximin Daia against Constantine, though this intervention was probably inadvertent, even though the aims of both were aligned for the moment. Constantine was therefore able to consolidate his position in Britain and Gaul, by which he gained command of the two major armies in the West. Galerius perforce recognized him as Caesar, but Constantine, though he accepted this investiture, went on calling himself Augustus. Galerius’ reluctance to accept this was palpable and he seems to have withdrawn his recognition later.

The main problem for Galerius, now the senior emperor, was not Constantine but Maxentius and Maximian. The latter in particular was not acceptable to anyone, and he even fell into dispute with his son. He had attempted to bolster his position by a marriage alliance with Constantine and by awarding him the title Augustus, which was not his to give and which Constantine already used anyway. When Maximian quarrelled with Maxentius, he found that the army preferred the younger, more generous man, and he fled to his new son-in-law, who was not pleased to have him as his guest.4

As a further result of this break, another pretender appeared, this time in Africa. This was a man called L. Domitius Alexander, who had been vicarius in Africa since at least 303. In effect he was protesting at the revived Emperor Maxentius and was able to prevent the despatch of food supplies to Italy. Alexander is universally described as old and lazy, but he cannot have been all that indolent for he managed to develop a navy and to gain control of Sardinia, where his authority is noted on at least one milestone, and so he ruled as emperor there and in Africa for some months at least. He had himself proclaimed as Augustus in 308. However, he did not have much in the way of military forces, and like Maximian he attempted to bolster his position by soliciting an alliance with Constantine, who did not respond. He was eventually defeated by ‘a few cohorts’ sent by Maxentius from Italy in 311, though the size of the invading army was much greater than is implied. This all suggests that Alexander had had substantial local support, and at least the tacit support of the other nearby governors, who had more substantial military forces at their disposal than he did. He had operated as emperor for perhaps a year; he must be accepted as an emperor, just as Maxentius and Carausius were. The victorious army of Maxentius sacked Carthage and Cirta and conducted a brutal purge of Alexander’s supporters. Alexander himself was killed.5

All this activity in Italy and Africa effectively left Constantine some years for consolidation. For Galerius, based in the Balkans, and for Diocletian, living in retirement at Salonae in Dalmatia, events in Italy were more important, particularly since Constantine busied himself with frontier wars. For Maxentius it was the secession of Domitius Alexander in Africa that was important, since the spring of Alexander’s action was resentment at Italian demands on food from Africa and his blockade of food supplies had to be broken. Restoring these supplies was clearly a priority for an emperor dependent in some degree on Roman support.

By 309, therefore, the number of emperors had grown from four to six. In addition, Diocletian intervened occasionally from his retirement, and in a conference at Carnuntum on the Danube in November 308, he replaced the dead Severus with a new Augustus, Valerius Licinianus Licinius, and browbeat Maximian into agreeing to go into retirement again, which turned out to be even briefer than the earlier one.6 In 309, Constantine gained control of Spain and Maxentius suppressed Alexander. This was to be the pattern: jostling for position and territory, suppressing or replacing the weaker rivals. The older generation was dying off: Galerius died in 309 or 310; Diocletian probably in 311. Maximian again returned from retirement, but was at last defeated; he committed suicide in prison.7 By 312 Constantine was ready to take on Maxentius, who had meanwhile forfeited much of the support within his territories.

Maxentius in fact made a good fight of it, demonstrating that it might well have been sensible to have set him up as Caesar in the original scheme. However, Constantine was faster, tougher, more ruthless, a more adroit politician and a better general. He defeated Maxentius at the Battle of the Milvian Bridge in 312, at which Maxentius died.8

Between them, Constantine, Maxentius and Maximian had struck wounding blows at Diocletian’s governing scheme and had destroyed his scheme of organized succession. They had based their claims to power and authority on competing bases: Constantine claimed the Empire by hereditary right, as the son of an emperor; Maxentius based his claim on the older bases of people, Senate and army, though, of course, he was also the son of an emperor. In the past this combination could have been unbeatable, but Diocletian had destroyed the validity of that set of legitimizers. Now the only claim that counted was armed victory. The usurpers had undermined Diocletian’s scheme; Constantine had revealed its fragility, and in so doing he also enforced heredity as the crucial element in a succession. Constantine’s victory was therefore a victory also for his hereditary claim and for that as an organizing principle in the imperial succession.

In fact, the whole Diocletianic scheme, developed piece by piece over twenty years, was unworkable without the master’s touch on the controls. It was only his interventions that kept it going for a few years after his retirement. (When he persuaded Maximian to retire a second time, his appointment of Licinius angered Maximin Daia; the Diocletianic scheme not only ignored hereditary emotion, it ignored human pride as well.) It was not reasonable to expect that emperors with sons would ignore their posterity in favour of choosing other men. Diocletian, of course, had no son.

There was also another aspect to be considered. The emergence of Constantine and Maxentius demonstrated that the selection system devised by Diocletian was very likely to pass over the most capable candidates. The new Augusti as they were promoted were supposed to choose new Caesars, but they would not wish to select men more capable than themselves for they would not wish to be outshone by their deputies. Nor was it sensible that a group of generals – for it was high military command that Diocletian saw as the necessary qualification for imperial office – should be in control of the system of succession since, as Probus had remarked in that unguarded and fatal moment, if the generals were successful, then they and their armies would become unnecessary. That is to say, the new emperors would require different skills and abilities to the military – civilian ones – and generals were not the best-suited men to divine such non-military qualities. As a skilful political operator, Diocletian really should have understood this.

Diocletian’s solution to the succession problem was in fact a personal one, and one that would therefore not outlast its author. Nor was it possible to go back to the old system of army election followed by ratification by the Senate. Maxentius tried that, but rapidly discovered that a conservative reaction was a poor basis for his power, and the maintenance or restoration of old privileges was actually an obstacle to effective rule. Further, election by one army automatically provoked jealousy in others and could all too easily result in multiple claimants. It was in fact inevitable that the emperors would revert to a system based on heredity. Constantius’ (presumed) organization of the proclamation of his son as he lay dying had therefore decisively broken Diocletian’s whole scheme in one blow.

Maxentius’ emergence as princeps – he soon upgraded himself to Augustus – came with his father’s blessing, but he was only a candidate because he was Maximian’s son. Diocletian’s scheme had in fact been a substitute for that process of imperial selection that is called civil war; without Diocletian’s control, or rather his power and personal influence, it quickly reverted to civil war once more. It could not prevent the emergence of competing candidates who were wholly detached from the scheme.

In the process of his tinkering, Diocletian had finally removed the Senate from the system. Only Maxentius made a claim to installation by the Senate, and that was unconvincing to the rest since his support was no less local than that of the Syrian rebels defending their province against the Persians, or the Gallic emperors claiming to protect Gaul from German invaders. Maxentius was claiming to protect Rome and southern Italy against the imperial government’s tax demands. This was no more a defensible position than that of the Alexandrians, and by lending itself to his policy, the Senate was only making itself as local as he was.

This was the final result of the long process of senatorial sidelining that had begun with Septimius, or even Trajan, and whose influence had only revived to a degree with the Gordians in 238 for a single long generation. Carus had not even bothered with senatorial acceptance, nor had Diocletian. Now that the Senate was removed from exercising any influence and Diocletian’s odd and unworkable system had collapsed, all that was left was heredity and military power. Constantine’s claim to be Augustus was based on inheritance, plus the support of his army. The net result of Diocletian’s meddling was to impose an hereditary succession system on what had, in origin, been a military elective system, but it was actually the final working out of the system set in place by Augustus that had combined heredity, the army and the Senate, a combination which had, in the end, proved to contribute to disaster and civil warfare. It was entirely appropriate that Constantine should have begun to style himself Augustus from the moment he was proclaimed.

The Tetrarchy – The Caesars


Galerius (305–311). (Shinjirod via Wikimedia Commons)


Severus (306–307).


Maxentius (306–312). (Jean-Christophe Benoist via Wikimedia Commons)

Constantine and his Enemies


Constantine ‘the Great’ (306–337). (Shakko via Wikimedia Commons)


Maximinus Daia (311–313).


Licinius (308–324). (Adobe Stock)

If you find an error or have any questions, please email us at Thank you!