A link between events in Antonine Scotland and Germany has long been envisaged, though the chronology and sequence of those events has been debated.Footnote ¹ Over 20 years ago, Nick Hodgson put forward a well-argued case for dismissing the existence of a second Antonine phase of occupation in Scotland and bringing forward the date of the Roman withdrawal.Footnote ² This allows for another scenario, which is considered here, though it was not envisaged at the time. Hodgson pointed out that hardly any excavated sites had provided unequivocal evidence of a second Antonine occupation following a period of abandonment, and suggested that the minor structural changes that were attested need be no more than piecemeal alterations and repairs to predominantly timber buildings. Where two Antonine phases could be demonstrated, he argued that, rather than indicating any later reoccupation, they reflect changes of garrison arising from a strategic reassessment after the addition of up to 11 forts to the Antonine Wall during its construction.Footnote ³ Accordingly, he dated the beginning of the withdrawal to a.d. 158, some five years earlier than had previously been thought, on the basis of the well-known, but now lost, inscription recording refurbishment of the curtain of Hadrian's Wall or one of its minor structures.Footnote ⁴ At the same time, Hodgson recognised that there was some evidence for occupation of the Antonine Wall after that date, which must indicate either a phased withdrawal or occasional use of certain sites as outposts of Hadrian's Wall.Footnote ⁵ This interpretation of the dating evidence has been generally accepted.Footnote ⁶

It is also widely agreed that, on becoming emperor, Antoninus Pius’ rapid decision to reverse his predecessor's policy and advance the frontier in Britain was to a large extent linked to the potential insecurity of his political position in Rome.Footnote ⁷ Most convincing is the suggestion that, because he lacked military experience and reputation, he needed the prestige that such a military victory would bring. It is particularly worthy of note that this was the only time in his reign that he accepted the imperial acclamation for a victory, as is recorded in his titulature and celebrated on a series of coin issues and in contemporary speeches.Footnote ⁸ Given this close connection between the emperor and the reconquest of Scotland, it is more difficult to understand why that occupation was brought to a close during his reign rather than after his death.Footnote ⁹

Hodgson's solution was to suggest that the withdrawal was primarily a reaction to local opposition, focused in south-west Scotland, perhaps related to unrest and threats of war in Germany and Raetia around this time.Footnote ¹⁰ It is the contention here that the link between problems in Germany and Raetia and the withdrawal in a.d. 158 is a much more direct one,Footnote ¹¹ while the assertion of the precipitatory effect of local rebellion in northern Britain has been exaggerated.

There is, in fact, quite strong evidence that the decision to abandon Antonine Scotland was both ordered and controlled, as the suggested phased withdrawal would itself imply. There are indications that the distance stones on the Antonine Wall were carefully removed and buried, the latter perhaps reflecting some ritual acts.Footnote ¹² Forts along the Wall were deliberately demolished and the sites carefully tidied up, as attested by the infilling of wells at Old Kilpatrick and Bar Hill and the general paucity of the material left behind.Footnote ¹³ At Bearsden, the most recent fort to have been excavated, it is clear that the internal buildings were demolished, as evidenced by the burnt wattle and daub that filled associated gullies and by the throwing down and burning of the putative rampart breastwork.Footnote ¹⁴ The state of the artefacts recovered, however, suggests that material of value, including querns and window panes, had been taken away; this implies a peaceful evacuation. So too at the outpost forts and in the hinterland. Extensive excavations at Strageath indicate systematic demolition, including the graphic recovery of a spade blade broken during the removal of a timber post.Footnote ¹⁵ The end of Antonine occupation at Cramond was sufficiently indistinct for the excavator to argue that the fort may have continued in use as an outpost fort.Footnote ¹⁶ Demolition pits are attested at InvereskFootnote ¹⁷ and Lyne;Footnote ¹⁸ with widespread demolition, including the infilling of ditches and the burning of buildings, attested at Bishopton,Footnote ¹⁹ Crawford,Footnote ²⁰ DrumlanrigFootnote ²¹ and Barburgh Mill.Footnote ²² All this evidence points to an orderly withdrawal, where burning was an integral part of the demolition process.

Only at Birrens does the picture seem to be different. Here the excavator considers the character of the demolition and associated burning prior to rebuilding to be ‘savage and wayward’ rather than ‘Roman and orderly’.Footnote ²³ This contrasts with the evidence from Newstead, which also continued to serve as an outpost fort beyond Hadrian's Wall, where the rampart and most of the primary Antonine phase internal buildings seem to have continued in use.Footnote ²⁴

There are various strands of evidence that suggest that manpower had been stretched throughout the Antonine occupation of Scotland. The auxiliary garrison of Britain at this time is known from diplomas to have been made up of 12 alae and some 44 cohorts.Footnote ²⁵ Lawrence Keppie suggests that this did not provide sufficient troops to occupy all the forts and fortlets simultaneously, though this leads him to argue that those on the Antonine Wall were held only lightly in contrast to forts elsewhere.Footnote ²⁶ There is, however, no unequivocal archaeological evidence that large areas within the forts on the Antonine Wall were devoid of barracks, rather than that early excavators failed to find or seek such buildings.Footnote ²⁷ In fact, a strong case can be made that the Antonine Wall was one of the most intensively garrisoned frontiers in the Roman Empire. The final stage of its development, with forts at intervals of 2.4–5.8 km along a barrier only some 62 km long, involved a closer spacing of associated fortifications than on any other linear frontier in the Empire, with the possible exception of the Valu lui Traian in Moesia Inferior.Footnote ²⁸ This density of installations may best be explained as reflecting a desire to increase local security by dispersing the troops more widely along the linear barrier in response to local resistance. That, in turn, is likely to have been stimulated by the creation of a hard border that would have been highly disruptive to social cohesion and may even have split existing tribal groups.Footnote ²⁹ That such a density of military deployment put considerable pressure on available manpower resources is strongly suggested by two facts. Firstly, those forts in the second phase of building include the five smallest on the Wall,Footnote ³⁰ none of which would have been capable of housing more than a detachment of men. Secondly, it seems likely that several of the Wall forts were garrisoned wholly or in part by legionaries,Footnote ³¹ which is not the most efficient use of elements of this élite infantry battle force. The army of Britain was also weakened by the dispatch of auxiliary units to the Mauretanian wars that appear to have been ongoing through the two middle decades of the second century a.d.Footnote ³² Some of these troops were discharged in a.d. 152/3, including soldiers from cohors I Baetasiorum, which is also attested at Bar Hill on the Antonine Wall.Footnote ³³ The presence of African-style pottery at various sites along the Wall would suggest that at least some troops returned to service there.Footnote ³⁴ The extensive use of small fortlets in south-west Scotland may be a further example of the economic use of manpower, though the network of forts and fortlets in the area is tighter than in any other period, which has also been taken to indicate a potentially troublesome area that required closer control.Footnote ³⁵

This manpower problem seems then to have been further compounded by the transfer of troops from all three of the British legions to Germany. A famous inscription dredged up from the river Tyne in 1903 has traditionally been interpreted as recording a detachment being contributed from the German provinces to the three British legions in the governorship of Julius Verus.Footnote ³⁶ There are, however, linguistic and grammatical problems with this translation, which requires acceptance of either a grammatical mistake (duobus for duabus to agree with Germaniis) or a haplography (a missing ex). Neither of these solutions is entirely satisfactory given the excellent survival of the inscription in its complete state. More telling, however, is the sense of the Latin. How can a detachment from the legions stationed in Britain be deemed to have come from the two Germanies? Accepting John Wilkes’ suggestion that ex should be expanded as exercitibus to resolve the grammatical problem with duobus, Michael Speidel attempted to resolve the problem in a more ingenious way by suggesting that the inscription records troops from the British legions, who had at some unspecified time previously been sent to the two Germanies, now returning to the province.Footnote ³⁷ However, applying Occam's razor and accepting the sense of the Latin text as is stands should reverse the direction of travel and see troops being dispatched from all three British legions to the two Germanies, as is persuasively argued by Wilkes.Footnote ³⁸ This explanation, though accepted as possible in principle, was rejected initially by other commentators primarily because it did not fit historical understanding at the time.Footnote ³⁹ However, that historical understanding has now changed as it no longer includes the continued occupation of Scotland, at least not in full force, into the early a.d. 160s. A withdrawal from Scotland in a.d. 158, as indicated by an inscription recording rebuilding on the line of Hadrian's Wall,Footnote ⁴⁰ would have made the transfer of troops to the two Germanies at this time entirely feasible.

The only precise chronological parameters of Julius Verus’ governorship of Britain come from a diploma and a dedicatory inscription. The former, from Ravenglass in Cumbria, is dated 27 February a.d. 158, and it is suggested that Verus must, therefore, have been in the province by the autumn of a.d. 157 at the latest because of constraints on seaborne travel in the winter months.Footnote ⁴¹ The latter, from Birrens in Dumfriesshire, is dated between 10 December a.d. 157 and 9 December a.d. 158 on the basis of the emperor's titulature.Footnote ⁴² Immediately before his transfer to Britain, Verus had served as governor of Lower Germany, probably from a.d. 154.Footnote ⁴³ If he had served the normal triennium, that would place his transfer to Britain in a.d. 157, which fits perfectly with the dating evidence above. The fort at Birrens was restored to serve as an outpost beyond Hadrian's Wall.Footnote ⁴⁴ It had previously been burnt down, though opinions differ as to whether this was part of the clearing-up process prior to refurbishment or the result of hostile action.Footnote ⁴⁵ Two further inscriptions, not independently dated other than by reference to Verus, attest building activity at forts as far south as Derbyshire.Footnote ⁴⁶ Thus, the pattern of evidence for his governorship is consistent, showing the refurbishment of Hadrian's Wall, along with its outpost forts and hinterland, beginning shortly after his arrival in the province in a.d. 157.

The question remains as to what was happening in the two Germanies that might have required legionary reinforcements in a.d. 158. While there is no literary evidence of major warfare at that particular time, there are clear indications of growing insecurity on the German frontier. During his governorship of Upper Germany (c. a.d. 152–55), Gaius Popilius Carus Pedo was given a special command.Footnote ⁴⁷ Though its precise nature is unspecified, the appointment clearly indicates military activity and implies some level of unrest in the area. Géza Alföldy long argued that this command related to the actual construction of the Outer limes,Footnote ⁴⁸ but this view can no longer be sustained in light of both the epigraphic and dendrochronological dating evidence. This indicates that it was not until almost the end of the decade that the frontier in both Upper Germany and Raetia was advanced to the new line.Footnote ⁴⁹ The extent of the territorial expansion involved was limited, so it remains unclear how much campaigning was necessary to bring the additional area under Roman control. The character of much of the Outer limes, which maintained a straight line across some 81 km of undulating terrain, is usually seen as indicative of administrative convenience and an assertion of Roman power, rather than being linked to any military defensive strategy.Footnote ⁵⁰ Nonetheless, this expansion of the province must have put some considerable stress on the military resources available locally. Even if the actual campaigning involved – assuming that there was any – was limited, simply constructing a new palisade across some 270 km from the Main to the Danube, with all its associated towers, forts and smaller fortifications, was a major undertaking in its own right.Footnote ⁵¹ One might have hoped for some epigraphic record of the involvement of these legions given the magnificence of the distance stones that mark the building of the Antonine Wall, but these records are unique in the Empire and reflect the particular circumstances of the reoccupation of southern Scotland; their equivalent on Hadrian's Wall are no more than small centurial stones.Footnote ⁵²

To summarise and conclude. The year a.d. 158 saw the beginning of the abandonment of the Antonine Wall and other sites in Scotland, and the reoccupation of Hadrian's Wall and its hinterland. It is suggested here, on the basis of the archaeological evidence, that this withdrawal was deliberate, ordered and peaceful, with the possible exception of a small part of the south-west Lowlands. Thus the primary reason for the withdrawal is more likely to have been a shortage of manpower than any localised rebellion. The enhanced density of military installations along the line of the Antonine Wall had put an additional strain on manpower which was further exacerbated by the sending of troops to Mauretania. Finally, the dispatch of troops from all three British legions to Germany in a.d. 158, probably to assist with the reconfiguration of the frontier there, rendered the occupation of Scotland no longer sustainable. Once again, problems closer to the heart of the Empire took precedence, despite Antoninus Pius’ close association with the earlier reconquest of Scotland.