There is also evidence to suggest that, like Hadrian’s Wall, the occupation of the Antonine Wall system was subject to changes during construction. In 1975 J. P. Gillam discovered that structural evidence showed several of the larger forts were earlier than the Wall, and most of the small forts were later than the Wall. This indicates that, originally, it was perhaps planned in a similar way to Hadrian’s Wall, with large forts at about 12km intervals and mile forts in between, but the change in plan involved a closer grouping of garrisons in additional small forts. Frere suggests that a shortage of men, evidenced by the numerous fortlets in Southern Scotland, may have compelled economies and simplification in the garrison of the Wall itself. Therefore, posts were maintained by dividing units, but the forts were placed closer together for better supervision of the frontier and so that reinforcement would be easier. Frere further considers it very unlikely that the patrolling and fighting garrisons were organised separately, as they may have been on Hadrian’s Wall. A number of beacon emplacements have been located attached to the back of the Wall, but these are not regularly placed along its length for lateral signalling. Two are known each side of Rough Castle and two on the west slope of Croy Hill, which would indicate that there purpose was for long distance signalling linking the outposts with fighting units to the rear. The presence of the Military Way connecting each fort indicates that a significant garrison needed to be supplied and an inscription from Carriden proves the establishment of at least one regular community with local self-government.
Frere offers evidence to suggest that what we see to-day are the remnants of a once complete system of surveillance and puts forward a view on how the estuaries at each end of the Wall were occupied and protected. He states that in the west a road ran beyond the Wall along the north bank of the Clyde and probably led to a harbour at Dumbarton, above which the river was not then navigable. This would have made naval patrols possible to protect the river from being crossed by boat, with the left shore being protected by a large fort at Whitemoss, with two fortlets further west at Lurg Moor and Outerwards, to watch over the estuary. In the east, forts guarded the southern shore of the Forth at Carriden, Cramond and Inveresk and along Agricola’s road north of the Wall there were outpost garrisons in forts at Camelon, Ardoch, Strageath and Bertha. This garrisoning was not as full and intensive as that carried out by Agricola and suggests a protectorate rather than a full-scale occupation. However, it is evident that no large hostile movements would have gone unnoticed.
We know from the inscriptions on distance slabs that all three legions were involved in the occupation and the building of the Wall. However, all the stones of Legio VI and all but one of Legio XX were set up by legionary vexillations rather than by the legion itself, but the stones of Legio II do not record any vexillations. This suggests that the whole of Legio II was stationed on the Wall, but that the other two legions only supplied detachments. However, if this were the case we would expect the distances built by Legio II to be far greater than those constructed by the other two and this is not so. There is further evidence to suggest that each legion was divided up into two working parties, so we may assume that the greater part of all three legions were present. Duplicate stones were set up at the end of the lengths built by each party, but no cohorts or centuries left records as they did on Hadrian’s Wall. Aerial photography discovered fourteen marching camps near the line of the Wall which were probably the housing for the construction parties.
There seems little doubt that the occupation of the Antonine Wall was developed into a complex frontier system, but the question remains as to why it was built, particularly after so much time, effort and expense had been put into the Hadrianic system. There is very little direct evidence so most theories depend heavily upon assumptions and circumstantial evidence. It is thought that the Greek writer Pausinias refers to these events when he says, ‘Antoninus Pius never willingly made war; but when the Moors took up arms against Rome he drove them out of all their territory…..Also he deprived the Brigantes in Britain of most of their land because they too had begun aggression on the district of Genunia whose inhabitants are subject to Rome.’ However, this suggests that Pausinias thought the Brigantes were outside the province, but they had been within it for more than sixty years, so he may well have his facts mixed up and probably meant a tribe further north. Whatever Pausinias meant, it does seem clear that disturbances in southern Scotland caused the Roman advance and Frere suggests that this was sparked off by an attack on the pro-Roman Votadini by the Selgovae outside Roman territory. If Frere is right, it means that the re-conquest of Lowland Scotland was directed at pushing the Selgovae, and possibly also the Novantae, further north beyond the Forth-Clyde isthmus and then building a frontier wall to keep them out. This would tend to agree with Capitolinus, who stated that ‘after driving back the barbarians, built another wall, of turf.’ This effectively brings the Votadini into Roman territory and also suggests that, maybe, Hadrian’s wall was built in the wrong place. There is further evidence, produced from inscriptions, that by 145 AD there are at least ten units of Numeri Brittonum in Germany, probably recruited from Scotland and would neatly explain what happened to the Selgovae and Novante that were captured or surrendered after the re-conquest. However, Dr. Baatz has shown that British numeri were in Upper Germany from considerably earlier than 145 AD and although not entirely negating this theory, does throw it open to question.
The fort at Newstead is important and would tie in with the Votadini theory. Newstead was rebuilt with a stone wall and a recovered altar stone confirms it was garrisoned with two cohorts of Legio XX and a regiment of auxiliary cavalry, which had been re-deployed from Hadrian’s Wall and some were moved up to the new wall. On Hadrian’s Wall the vallum was filled in every forty-one metres to form causeways and gates were removed from the milecastles to allow free access across the old military zone. Archeological evidence also points to change in habitation patterns in the eastern lowlands. Individual farmsteads now appear on open sites, principally Votadinian as indicated by the number of Roman objects found at their capital, Trapian Law, and hill forts have gone out of use, probably to decrease any chance of resistance. This clearly suggests the impact of Roman suppression of the old warlike way of life.
Inscriptions, coins and troop movements are an indication that there was serious trouble in Britain between 154-158 AD. The building inscriptions from Birrens and Heddon-on-the-Wall recorded the governor Julius Veros and suggest that the Antonine Wall was abandoned at this time. An inscription dredged from the Tyne at Newcastle also indicates the arrival of reinforcements, for all three British legions, from Germany. The inscription from Heddon-on-the-Wall, in particular, even indicates a re-occupation of Hadrian’s Wall. A further inscription found at Brough on Noe, in Derbyshire, led Haverfield to believe that the trouble was not invasion from the north, but a rebellion of Brigantes in the Pennines. Further weight is added to this argument by the re-building and occupation of forts in the Pennines and the fact that the fort at Lancaster was burnt at this time and would also explain the heavy re-occupation of this area. Frere explains that to crush a significant rebellion by the Brigantes, troops would have to be withdrawn from the Antonine Wall because there were no other troops or strategic reserves that could have been called upon. In addition it would explain the destruction of the Antonine Wall and most of the forts in the Scottish lowlands, in as much that they were destroyed to prevent them being used by the enemy. There is no archaeological evidence to suggest that they were destroyed by enemy action. Coins of Pius minted in 154-5 AD show Brittania subdued, which suggests that the war or rebellion had been put down by 155 AD.
Although Roman forces were able to return to Scotland within a year, it is clear that there were renewed outbreaks and that the trouble was acute and prolonged, which is evidenced by the quality of the governors sent to deal with it. Verus, who was a man of outstanding ability, was succeeded in 159 AD by a man whose name is incompletely recorded ‘Long…..’ something, but by 161 he was replaced by Marcus Statius Priscus Licinius Italicus, who was a remarkable general and trusted by the emperor.
The Antonine Wall was rebuilt sometime in the 160’s and Richmond says that it was rebuilt soon after the fort at Newstead was re-occupied. This view is reinforced by the knowledge that Corbridge was being reorganised by Calpurnicus Agricola, by the building of massive granaries and setting up a supply base. This suggests that they were already planning to use it when they went back. Dr. Steer put forward an impressive argument that the second Antonine occupation began before the death of Antoninus, in 161. His evidence was that the original forts were built by legionaries, but close study of two forts, Castlecary and Rough Castle, also have stones from auxiliaries which are of much rougher quality, but the auxiliary building records are dedicated to Antoninus. In addition, at Bar Hill, building inscriptions of Cohors I Baetasiorum were discovered among debris thrown down a well, clearly evidence of the final demolition at the end of the second occupation, otherwise it would have been cleared out. Finally iron arrowheads were found below this debris indicating that they were there first, from a previous occupation. Also B. R. Hartley proved from his study of samian ware that the second Antonine occupation was very brief and that forts were not held concurrently with those in northern England. From this evidence we can deduce that there were rebellions in the Pennines and Scotland during the 150’s and that the Romans pulled back and re-fortified Hadrian’s Wall, but soon they were back on the Antonine Wall and in the outposts beyond it. Excavators found evidence that the fortlet at Lynne was replaced by a full sized fort and that Loudoun Hill and Carzield were abandoned and not rebuilt. The second occupation of the Antonine Wall was strong and nothing much happened on the Wall or north of it, suggesting that the trouble may have been in southern Scotland and the purpose of the Wall was to separate the tribes in Scotland, preventing conflict between them.
Whilst Antoninus Pius lived, Scotland continued to be held, probably because it was the major military achievement of his reign. However after his death in 161, he was succeeded by Marcus Aurelius and policy in the north was re-assessed objectively. It had become obvious that there were just not enough troops to control the vast area between the Pennines and the Forth and in 163 Calpurnius Agricola was sent as governor, the Scriptores Historae Augustae says ‘War was threatening in Britain and Calpurnius was sent to deal with it.’ Calpurnius rebuilt a number of forts in northern England so it was probably he who implemented the decision to, yet again, withdraw Roman troops from Scotland. Certainly small quantities of late Antonine pottery, found by excavators at Glenlochar proves that the area was sparsely populated and that there was no occupation of the Antonine Wall by 165 AD. Under Marcus Aurelius, Scotland was not totally abandoned, but troops were needed to keep a firmer grip on territory further south and he was probably responsible for the reorganisation and re-occupation of Hadrian’s Wall.