The Saxon Shore. Frontiers of the Roman Empire project/RGZM/David John Breeze
Barely one year into the reign of Diocletian, another usurpation shook the northwest of the Roman Empire. Carausius had been in command of ‘the fleet that formerly protected the Gauls’ (Panegyrics Lat. Vet. 8(5) cited after Nixon and Rodgers, 1995). This may well have been the classis Britannica, but the scarcity of records might hide other commands that may have been created since then. According to Eutropius (9.21) he was based at Boulogne and charged with pacifying the sea from the Franks and Saxons.
Carausius is recorded to have been Menapian by birth (i.e. from the area of Belgian and French Vlanders) and to have been a pilot or helmsman in his youth (the Latin term can mean both). Immediately before his command in Boulogne he had campaigned with Diocletian’s later Caesar Maximian against the Bagaudae in Gaul.
His command in Boulogne was successful, but according to Aurelius Victor and other sources he failed to hand his booty over to the central government and was thus ordered to be executed. Instead he revolted and declared himself Emperor in Britain and the north of Gaul. This scenario is a topos (a prejudice on the basis that usurpers should be considered greedy), and Carausius is not the first Emperor to be tarred with it, as the same story is also told about Postumus twenty years earlier.
Apart from the short entries in the historians (Eutropius and Aurelius Victor), and the references in the Panegyrics, our main understanding of the size and history of the Carausian Empire derives from the study of his coinage, both with regard to the images and legends used, as well as the distribution of the coinage. The conclusions drawn differ between different historians and numismatists, but the general sequence appears clear.
It is frequently assumed that the main power base of the Carausian revolt was the fleet (British or otherwise) and its base at Boulogne, although the appearance of his early coins in northern France could point to a second base, the assumption being that because Constantius laid siege to Boulogne, this must always have been the main base. In addition to any fleet he may have controlled, he seemed to have claimed at the height of his power the support of nine legions or parts thereof, as he commemorated them on his coinage.
It seems that around 289/290 Maximian tried to invade the breakaway Empire and remove Carausius, but failed, a fact that is beautifully circumlocuted in the Panegyrics and commemorated on Carausian coinage by issues with images of ships and ‘felicitas’ (good luck) and ‘salus imperatoris’ (the wellbeing of the Emperor) as their legend. Furthermore Carausius from this point onwards portrayed himself with Diocletian and Maximian as the third Emperor. The arrangement was clearly one-sided, as neither of the other Emperors returned the gesture, but it is possible that the stand-off in the following year resulted because Maximian did not have the time or manpower to deal with Carausius and problems on the Rhine frontier at the same time.
The situation changed with the appointment of Constantius Chlorus (293/306) as the Caesar and thus as junior Emperor in the West in March 293 (at the same day Galerius (293/311) was made junior Emperor in the East), creating for the first time a college of four Emperors, the Tetrarchy). This meant that Diocletian and Maximian had no intention of accepting Carausius as their colleague and unsurprisingly, soon after the appointment, the attacks on the Carausian Empire were renewed. As a consequence Carausius lost the north of Gaul for good. Constantius proceeded to lay siege against Boulogne and managed to block the harbour exit through the construction of a dam of timber and stone at the mouth of the Liane, leading to the surrender of the base. It seems shortly after this disaster Carausius himself was replaced with Allectus, one of his co-conspirators and officers, but it is symptomatic of our understanding of this period, that we do not know Allectus’ rank under Carausius. Aurelius Victor describes him as ‘cum ei permissu summae rei praeesset’ (he was in charge of everything with his (i.e. Carausius’) permission). This has frequently been interpreted that Allectus may have been the ‘rationalis summae rei’ – a very high finance officer in the late Empire, frequently with reference to the abbreviation RSR found in a number of Carausian coins. These have now been shown to refer to a Virgilian quote from the Fourth Eclogue (de la Bèdoyére 1998). However, Aurelius does not give the correct title ‘rationalis’ and his apparent circumlocution makes too many other interpretations possible.
Having regained control of northern Gaul, and limited Carausius, or rather his successor Allectus, to Britain, it took Constantius another three years to invade Britain itself. The most detailed account of the Constantian campaign both against Boulogne and the Rhine Delta, and the final attack on Britain is provided by the Panegyric of 297 honouring Constantius. While this text provides us with the chance of constructing a narrative for the events leading up to the fall of London, it is worthwhile remembering that Panegyrics are the ultimate praise poems or eulogies, constructed to let both the speaker feel erudite and the Emperor feel flattered. These speeches are clearly not designed to be, in any shape or form, balanced or interested in providing a full picture of what happened. The positive highlights of the campaign are what matters, as chosen by the court of the winning side. Without the chance of verifying the details, we are once again dependent on the account as presented. Critical assessment on what might be covered up may be helpful, but is still just one of several possible scenarios.
The Panegyric stresses the personal involvement of Constantius. It relates the fact that the fleet aimed for the Solent and slipped past the Allectan fleet in thick fog. It also suggests that Allectus, during the final encounter, did not draw up a battleline or make use of any of his considerable preparations. This failure as well as his moral shortcomings consequently (in the mind of the unknown orator) led to his death in battle, where in clear imitation of Tacitus’ statement two hundred years earlier, no Roman was killed. Meanwhile a part of the fleet became lost in the mist and found itself in front of London, just in time to kill all the surviving Frankish mercenaries fleeing from the battle, thus saving the city and ‘providing a spectacle’. Then ‘the necessity of obedience was imposed on many other people who had been accomplices in that criminal conspiracy’ (Panegyric VIII, 14, 1ff in Nixon 1995).
Eutropius (IX, 22, 2) and Aurelius Victor (39, 22) suggest that Constantius’ praetorian prefect Asclepiodotus may have been in charge of some of the troops, or at least of a detachment of the navy and legions. Asclepiodotus is not mentioned in the Panegyric, as in the context of an Imperial eulogy, it would have been considered a faux pas.
Throughout the account it is never quite made clear where Constantius is. The impression is one of ubiquity, but more than one scholar has suggested that Constantius may have followed after the army and thus it is generally assumed that Asclepiodotus fought the battle against Allectus (Nixon and Rodgers 1995, 136–138).
Constantius is celebrated as an avenger and liberator; Britain’s liberation is perceived ‘as its restoration to the whole world’. (Panegyric IX, 18,4) Finally, the city of Autun thanked the Emperor for receiving many artisans, with which Britain was amply supplied, for its own reconstruction.
The account is further underlined by the Trier Gold medallion, which celebrates Constantius as REDDITOR LUCIS AETERNAE, the restorer of eternal light. The reverse shows Constantius on horseback approaching a city, whose personification kneels to receive him. LON identifies her as London, and beneath Constantius a galley signifies the accompanying fleet.
Allectus clearly met Constantius’ army in battle, but where this was and whether Allectus really was not willing to draw up a battle line should be doubted. The Panegyrist goes to great lengths to present Allectus and his army as barbarians, and barbarians are too wild to do anything but charge wildly against the Roman army. The description that no Roman died is clearly an exaggeration, as the Panegyrist himself admitted. The people on the battlefield were barbarians, or imitating barbarians with their reddish hair and the clothes they wore. That can just mean that the soldiers who fell wore late Roman uniforms, including trousers. But the praise of a victory in a civil war is always problematic, and stressing the foreign character of the opposing side meant the Panegyrist could create a clear ‘us’ vs ‘them’ scenario, which is not uncommon in late Roman historiography, and thus offered enough of a distance to remove the bad feeling arising from gloating over your dead fellow citizens.
The actions in London are clearly equally problematic. The Panegyrist claimed that his army killed the barbarian survivors in the town, before they could loot and move out. This sounds harmless and like a welcome rescue, but the barbarian army is the army of Allectus, and therefore the defenders of London, and street fighting in the city of London itself should, therefore, not be ruled out. The latter would also provide an explanation for the phrase ‘providing a spectacle’, with the invading army portrayed as providing gladiatorial style games at the same time as ‘liberating the city’. The latter, a novel way of portraying the situation, was clearly designed to distract from its seriousness by making it appear to the audience like an entertaining game. It can probably be assumed that the residents of London felt very differently about the event.
We do not know what stood behind the phrase ‘necessity of obedience’. In other cases of usurpations the winning side was only too willing to execute the leaders of the opposing factions and/or to confiscate their property. That the arrival of Constantius was not all the happiness suggested by the coin issue is made abundantly clear by the reference in the same Panegyric to British artisans arriving in Autun ‘to rebuild the city’. If we use the events associated with other usurpations as a parallel, then we have to consider the possibility of mass deportations or even enslavement of the British supporters of Allectus. It is certainly unlikely that Constantius would have found a large volunteer force just waiting to be taken across to Autun in southern France for construction work. The reference to Autun also provides interesting evidence to the extent of Carausius’ Empire. Autun in central France clearly thought of itself as loyal to the central government and the Panegyrist clearly saw the provision of craftsmen as a well-deserved boon.
That the opposition were unlikely to have been quite so barbaric as described by the Panegyrics is clear from the Carausian coinage (which ceased to be used shortly after 297). In addition to the standard repertoire of deities lending support to the Emperor and the personification of the core virtues of his reign, Carausian coinage contains frequent allusions to the poetry of Virgil and mention of other Roman institutions such as the Secular Games, thus showing that Carausius was promoting a considerable familiarity with Roman culture including literature and urban Roman traditions.
In addition to the legionary support celebrated on the coins, the armies of Carausius and Allectus were likely to contain foreign (including Frankish) mercenaries, but the same would be true of the army of the Tetrarchs, especially Constantius. The political need for distance between the Roman Emperor Constantius and the ‘foreign’ usurper Carausius, may have suggested to the Panegyrist that he focus on the foreign mercenaries, as well as Carausius’ origin in the borders of the Empire to discredit his claim as a ‘Roman’ ruler.
In 1961 White proposed that the Saxon Shore was in its origin Carausian. Numerous excavations on the coastal sites in the south and west of Britain have since then proved that the history of the coastal defences is more complicated than this statement suggested.
In its origin the modern concept of the Saxon Shore goes back to a command of the Litus Saxonicum listed in the pages of the Notitia Dignitatum (which is highly problematic as a source: see appendix), to which nine forts were assigned:
Branodunum / Brancaster, Gariannonum / Burgh Castle, Othona / Bradwell, Regulbium / Reculver, Rutupiae / Richborough, Dubris / Dover, Lemanis / Lympne, Anderita / Pevensey, and Portus Adurni / Portchester.
Published excavation reports have shown that these forts were not built as a unified scheme, but piecemeal over a considerable period of time. Of those listed Brancaster and Reculver can now be shown archaeologically to have been started before or at the end of the second century; thus at the time of the Carausian revolt they would have been in existence for nearly 100 years.
Richborough, at the opposite end of the Wantsum Channel from Reculver, appears to have started in 275; a similar date is now assigned to Pevensey (Fulford 1995). Lympne has produced early third century pottery, but most of the pottery and coins date to the period 270/348, suggesting that this may have been the heyday of the fort. Therefore all four may actually predate the Carausian usurpation.
By contrast extensive metal-detecting surveys at Burgh Castle produced a coin list whose earliest significant coin losses do not begin before the 320s, and which would thus suggest that its construction post dated Constantius’ re-conquest and indeed the accession of his son Constantine in 306 (Davies 2009, 219). The southernmost fort in the sequence, Portchester is the only one where coin finds suggest a start date in the late third century, perhaps even in the course of the reign of Carausius.
On the other hand, even in the area covered by these forts, there are other structures, which are clearly designed to be coastal forts and are designed to the same construction techniques with thick walls and the (subsequent?) protruding towers. These additional forts include Walton Castle (now eroded), Caister-on-Sea, as well as Reedham in Norfolk (Davies 2009, 221) and possibly Carisbrooke Castle on the Isle of Wight (underlying the medieval castle), a candidate proposed by Brian Philp (2005) on the basis of the overall similarity of shape to Reculver and Brancaster. Of these Caister-on-Sea, across the bay from Burgh Castle, was constructed in the early third century, and remained in use until c. 370.
Burrow Walls, Workington. David John Woolliscroft
Even beyond the Saxon Shore area, shore defences in the form of large forts close to coastal inlets, and natural harbours and beaching sites, can be found by the early fourth century in other parts of Britain, including Burrow Walls, as well as Cardiff, Holyhead, Lancaster and Moresby in the west (White 2007, 55–59), suggesting a growing interest in protecting the harbours and possible shipping routes around the island. Distribution maps of Late Roman pottery and other finds clearly document the importance of maritime shipping for the supply of the army, but also the civilians of the third century and later. Similar coastal defences are also found along the opposing side of the Channel, documenting that the protection these forts afforded was seen as a universal need in the north-western provinces, and was not specific to Britain or the southeast of the island.
So what did Carausius do along the coast? His ability to hold his British powerbase for nearly ten years against the rest of the Roman Empire suggests that he was very lucky (in picking a period where the Roman Empire was facing pressure from numerous other enemies), although he is usually also credited with being a great strategist and tactician, although admittedly in adulatory biographies. It is thus hardly surprising to find him credited with the creation of the Saxon Shore command by modern scholars, despite the lack of any supporting literary evidence. It is also clear from Casey’s coin diagrams, that most of the shore forts were in use in the Carausian period. Unfortunately, archaeology is not able to provide evidence for the interpretative jump from ‘in use’ to ‘used as a single strategic unit’. We thus are simply unable to answer the basic and most important question of when this command was created, what size it was or what it was meant to achieve. Its creator may have been Carausius, if we assume that he regrouped existing forts together for a new purpose, or it may not have been him. What we do know, however, is that by the time he came to power, Britain was provided with several forts close to harbours and coasts. Whether they were used as coastal defences, which would suggest that they were envisaging their role as a defence of the hinterland from sea-borne raiders, or whether they served as naval bases, in which case they might have seen their role primarily as protecting the shipping lanes around Britain, or as a combination of both, needs to be further studied. Preliminary studies at Pevensey and Richborough suggest that the answers might vary from fort to fort.
After the takeover, Constantius Chlorus remained in Britain. For how long is not known. Tony Birley (2005, 397–8) points out that it was likely to have taken some time to replace Allectus’ administrative structure and implement the changes necessary to integrate the provinces within the reformed structures of the Tetrarchy, but this could have been done by delegating, in a similar manner to Claudius’ appointment of Aulus Plautius.
In the aftermath of the reintegration of the Empire, Britain became exposed to the reforms of Diocletian and Constantine. These reforms, which in many areas created or at least codified the changes that made the difference between the early Roman Empire and the Late Empire, affected most areas of daily life. Instead of two provinces under direct rule from Rome, Britain was now faced with a two-tier administration, the higher, the diocese of Britain was further subdivided into four provinces: Prima, Secunda, Flavia Caesariensis and Maxima Caesariensis (the latter two clearly named after Constantius’ victory titles and family name). These four were administered by praeses or later consulares, who while in charge of the day-to-day running of the provinces, only had very limited control of the military. That came under the sole administration of military officers, comites (sg. comes) and duces (sg. dux), whose commands did not have to coincide with the boundaries of the civilian provinces. The boundaries of these military commands or the provincial boundaries within Britain are subject to numerous reconstructions, all equally based on very slim evidence. Nor is it likely that these boundaries remained static over the next 120 years. New military commands may have been created or others amalgamated, without leaving any traces in the archaeological record. Only at the very end of Roman Britain do we have a potential source for the disposition of some of the army and the administration in the form of the Notitia Dignitatum, although this has recently been questioned (see appendix).
The reforms also affected the structure of the army units. We have already seen that there is evidence for a decrease in unit size during the course of the third century and the creation of smaller vexillationes that operated for long stretches away from their parent unit, often in more mobile parts of the army, which would become the field army of the later Emperors (the comitatenses, literally: the companion army). A large number of these changes had gradually evolved, but were now made permanent. Many of the surviving units remained at their bases with more defensive duties and were eventually referred to as limitanei (literally: the border troops).
The sources often imply that this field army may have been of higher quality than the stationary troops. At least in some cases this can probably be put down to inter-corps rivalry, similar to the cordial comments passed today about members of the army by navy and air force personnel, and vice versa. But service as an officer in the comitatensian units appears to have provided better chances of promotion, as it was easier to attract the attention of the Emperor and his commanders; while service on the frontiers may have gone unnoticed by anybody but the neighbourhood and the local commander. Curiously, while there is a fairly decent understanding of the accommodation of the limitanei in their late Roman forts, so far it has proved difficult to identify the barracks or winter quarters of the field army anywhere in Britain.
Typically changes in accommodation can be seen in the whole province; the most striking (if only due to the scale involved) is the fortress of Chester in the early fourth century. There is continued evidence for the maintenance and substantial rebuilding of many of its main buildings, such as the baths and the principia. The barracks provision, however, appears to be reduced, with the barrack buildings in Deanery Field going out of use with the dismantlement of the structures, while in the neighbouring Abbey Green site, the accommodation survived well into the fourth century (Hoffmann 2002). The results are large open areas inside the Roman fortress. Similarly reduced accommodation areas can be identified in many forts and fortresses in active use, which survive into the fourth century. Michael Gechter (2001) has recently argued, on the basis of similar evidence in Bonn fortress, that these open spaces may have been used as protected campsites for the field army. While this argument is tempting, it is hard to prove. Elsewhere the finds of military equipment in walled civilian towns are associated with billeting of the field army, but other scenarios, such as the rise of the buccellarii, the armed body guards of the wealthy, cannot be ruled out.
In the auxiliary forts, new smaller types of barracks, first seen in the third century, continued to be built. Outside the civilian vici, which up to now were considered as constant companions to Roman forts, seem to vanish from the late third century onwards. By the beginning of the fourth century, vici were a rarity in the north of England, with Malton the striking exception. It used to be thought that the civilians used the open spaces inside the forts as their new residences, but with the recognition of ‘chalets’ as regular barracks instead of family accommodation, Bidwell and Hodgson (2009, 34) have recently argued that the reduced garrisons and the disappearance of the vici may be linked, suggesting that the reduced numbers inside the fort removed the economic viability of the civilian settlements outside it. However, this economic argument is problematic: historic evidence for the abandonment of villages in other periods suggests that economic viability is only ever one reason why a site is abandoned; frequently single houses survived, when the rest of a village may have gone elsewhere. Also, the abandonment of vici appears to be independent of the size of fort. If the reduced numbers were to blame, we would expect the larger forts to be able to continue to maintain smaller vici, while the smaller forts lose theirs. However, the loss of the civilian settlements can affect forts of all sizes, including the large fort of Pevensey, which was originally built next to a civilian settlement of second and third century date of considerable size.
In 305, after the resignation of Diocletian and Maximian, and his promotion to senior Augustus, Constantius Chlorus returned to Britain. The main source material for his presence in the island is the Panegyric for his son Constantine in 310 (Panegyrics Lat. vet 7(6)). The text’s flowery language reduced Constantius’ presence in Britain to a need to move closer to the stars at the end of his life, and to provide York with the opportunity to be the place of Constantine’s inauguration in 306, and thus the first step to Constantine’s route to becoming the first sole ruler of the Roman Empire in a generation.
The contrived suggestion of the Panegyrist that dying and making room for his son was Constantius’ only achievement in Britain is contradicted by Constantius’ coinage which proclaims him in 305 ‘Britannicus Maximus’. The title is shared with his son Constantine, who joined his father in Britain in 305, early enough to get involved in the campaigning. This suggests that in a similar manner to other Emperors (especially Septimius Severus), Constantius may have come to Britain to provide his son with the chance of gaining experience in warfare at Constantius’ side, and thus establish him in the eyes of the army as the heir designate. If so, then the ploy worked. The army appointed Constantine Emperor immediately after the death of his father and remained his staunchest supporter in the years to come (Barnes 1981, 71).
If we know little about the events of 305/306, there is even less to say about further possible visits of Constantine between 307 and 311/312, and in 313/314 or 314/315. The evidence for all of these are a series of coins struck in London proclaiming the ADVENTUS AUG and two passages in Eusebius (VC 1.8.2 and 1.25.2), which describe Constantine as returning to Britain to wage war and possibly to gather troops for his campaign against Maxentius. The visits in 313/314 or 314/315 appear to have resulted in Constantine being called Britannicus Maximus. Four Imperial visits in 10 years would suggest serious problems in the province, especially in view of the precarious relationships with Constantine’s colleagues Severus, Maxentius, Severus II, Galerius and later Licinius in Italy and a volatile situation on the Rhine frontier. A campaign in Britain always carried the risk of the inability to return if more serious trouble erupted along the eastern and southern borders, but again there is little beyond the coins and the fleeting references in Eusebius to tell us about the situation.
Archaeologically, we see the abandonment of the outpost forts of Hadrian’s Wall by 312 (Casey and Savage 1980). The continued mentioning of the Picti in the Roman sources, as well as finds of Roman silver, including the late Roman Traprain Treasure suggest that the abandonment of these advanced posts did not mean that Rome had lost all interest in the lands north of the Wall; but it may signify a change of policy, with the importance in how the northern neighbours were ‘engaged with Rome’ shifting from military presence to payment of money (one assumes for continued good behaviour). Elsewhere, the late third and early fourth century generally appear to have seen another phase of rebuilding in the North of England, covering both Hadrian’s Wall, as well as some hinterland forts such as Chester-le-Street and Old Penrith, and a major refurbishment of Chester fortress (Mason 2005, 69–81).
In c. 325 Constantine closed the mint in London. This probably happened as part of the general reorganization of Constantine’s part of the Empire, and would be of little significance for a military history of Britain, were it not for the fact that we now lose the ADVENTUS coinage as an indicator of Imperial visits to Britain. We are thus even more dependent on the slim literary sources for our understanding of events in the province.
The next Imperial visit recorded was by Constantine’s son, Constans in 343. Our sources are Ammianus Marcellinus who refers back to a part of the Histories that is now lost, as well as passages in the speeches of the late Roman orators Libanius and Firmicus Maternus: the two agree that the most noteworthy item was that Constans crossed the Channel in mid-winter, and Libanius (who lived in the eastern part of the Empire) added that it was not caused by any military emergency. However, Ammianus’ corresponding reference suggested that Constans was active at the northern frontier and that he may have come to the aid of people in Britain. In addition coin issues claimed a victory around this time.
Seven years later Britain became part of the Imperium of Magnentius, a usurper, who had been declared Emperor in Autun in January 350; his direct opponent, Constantine’s son Constans, was killed shortly after, leaving Magnentius in charge of the Western Empire. Constantius II, brother of Constans and Emperor in the eastern part of the Empire managed to defeat Magnentius at Mursa a year later, but Magnentius held on to power in the West until his eventual defeat in 353.
Battles between Emperors in the fourth century tended to be costly in manpower, even more so, if they involved the comitatenses of both the Western and Eastern Empire. In the past it has become a stock-in-trade theory amongst historians to link the two facts, and claim that usurpations resulted in the army of Britain being depleted in preparation for these battles, and – if the battle was lost – that the soldiers never returned. There is little evidence to support these claims, but there is as little evidence that Britain remained exempt from the repercussions of these recurring bouts of Roman self-destruction.
Ammianus Marcellinus is, however, explicit that Britain did suffer in the aftermath of the fall of Magnentius in 353. The winning Emperor had sent Paulus, nicknamed Catena (the Chain), to ‘clean up’ in Britain. The episode is reported only by Ammianus Marcellinus (14.5.6), and remains thus unconfirmed. According to Ammianus’ account, Paulus, a notarius (a high civilian official) at the court of Constantius II was sent to Britain to apprehend some military persons associated with Magnentius. Upon arrival, he appears to have started a witch-hunt. When the highest civilian official in Britain, the vicarius Martinus could not stop Paulus, he attempted to kill him, but in the event failed and subsequently Martinus committed suicide.
The next historical event known is equally difficult to understand: Constantinian coins overstruck with the legend DOMNO CARAUSIUS CES may refer to another usurpation by another Carausius in the late 350s. This so-called Carausius II is not mentioned in the admittedly scanty historical record, and thus it is impossible to decide whether the coins are forgeries (especially in view of the misspellings) or perhaps the only witness to another problem in the province in the middle of the fourth century (Birley 2005, 420).
In the winter of 360, Julian, the cousin and brother-in-law of Constantius II, revolted against his cousin in Paris. Shortly before these events, Ammianus Marcellinus (20.1.1–3) reported that either the Scotti or Attacotti (the text is corrupted at this point, but both would refer to originally Irish tribes) and the Picts harassed the area close to the frontier. Instead of coming to Britain himself, as apparently his uncle Constans had done, Julian sent his magister militum Lupicinus to deal with the problem. Again we hear about a winter crossing, but the problem was apparently quickly resolved as Lupicinus was back in Gaul in the spring after the declaration. There is more than a hint in the surviving account that he may have been sent to Britain mainly to have him out of the way during the usurpation, rather than to solve a military problem. Ammianus mentioned previous disaster befalling Britain, but again the account is too short to make any sense of what may have happened.
After Constantius II in 361, Julian united the Empire under his sole rule (something last seen in the 330s), but on 26 June 363 Julian died during fighting in Parthia: with him and his successor and distant cousin Jovian (363/364), the Constantinian dynasty came to an end: Valentinian (364/375), commander of one of the Imperial bodyguards, became Emperor, and quickly divided the Empire with his brother Valens (364/378), who took over the Eastern half. This led to a militarily successful period in the West for the next 11 years, frequently associated with large-scale defensive building and vigorous campaigning on the Rhine and Danube.
The period is also a watershed for Britain. The early fourth century has been described by some as the ‘Golden Age of Britain’, with many of the villas seeing their largest expansion, and the development of the large distribution networks for pottery, which can be taken as an indicator of comparative safety for large-scale transport. This is also, for many towns, the period where their buildings see their final expansion phase, although as noted above, in the North, this goes hand in hand with the loss of the civilian vici around many forts.