Although labelled as the Antonine Wall, the title is somewhat misleading and it is more accurately described as a frontier. Like Hadrian's Wall in the south, it comprised various components -
The Military Way was a stone and gravel road around 5 metres wide that connected the various forts and fortlets along the Wall and facilitated all weather access across the entire length of the frontier.
Remains of the Military Way at Seabegs (see relevant section)
On Hadrian's Wall a dedicated road had been omitted from the plan with the intent being to use existing infrastructure. It soon became evident this idea was flawed and a Military Road was added to that frontier. That knowledge was transferred to the Antonine Wall with the road being included in the original scheme and was one of the first components to be built.
The rampart was created from turf to form a scarped mound around 3 metres tall built on top of a rough stone base. It may have been topped by a walkway and timber parapet.
Rampart at Croy Hill
Stone Base at New Kilpatrick Cemetery
The rampart was constructed to a height of around 3 metres using blocks of turf dug from the locality or, where this was not available, by similarly sized slabs of earth faced with clay. These were placed on top of a stone base around 4.5 metres wide consisting of finely prepared kerb stones at either edge with unshaped stones in between. The decision for this stone foundation was based on experience gleaned from Hadrian's Wall where the western third had initially been a turf and timber construction. There no stone footing was used requiring the base of the rampart to be significantly wider than the Antonine Wall (6 metres versus 4.5 metre on the later frontier) in order to achieve the same height. This facilitated a significant reduction in the amount of turf required saving thousands of man hours – a factor that should not be under-
Debate rages as to whether the rampart had a Wall walk and parapet. The dimensions of the rampart were sufficient for a wall walk whilst some form of breastwork would have provided an additional barrier. However historians observe that the Roman army was configured and trained to fight in open combat rather than from parapets and also note that other frontiers, particularly the German frontier (Limes Germanicus), did not have any such facility. They also note the forts on the Wall tended to have more defensive ditches and thicker ramparts. On balance however, it can be strongly suspected the frontier did have a walk way; otherwise why invest so many man hours building such an elaborate layered defence when a much cheaper effort would have sufficed?
The berm was a flat space between the rampart and ditch; an essential requirement to stop subsidence.
Remains of rampart and berm at Seabegs (see relevant section).
The average width of the Berm was around 6 metres and there is some evidence it had obstacles to slow down attackers; perhaps spiked pits (Lilias as seen at Rough Castle) or alternatively there could have been stakes projecting from the ground or thorn bushes.
The defensive ditch was a ‘V’ shaped trench, on average 12 metres wide by 4 metres deep, that ran the entire length of the Wall.
The impressive remains of the ditch at Dullatur (see relevant section)
Unlike Hadrian's Wall, the ditch was constructed even where the terrain made it superfluous most notably at Croy Hill where it runs below the tall crags. However, in several places the terrain defeated the Romans; near the Croy Hill fort an untouched neck of rock crosses the ditch.
Another variance from Hadrian’s Wall was the consistency of the width of the ditch. Whilst the average for the Antonine Wall is 12 metres, it narrowed to as little as 6 metres in some areas perhaps reflecting the work of different construction teams.
Upcast Mound (Glacis)
The upcast mound was the spoil from the ditch that was placed on the north side to effectively increase the depth of the ditch.
The upcast mount (left) near Rough.
The ‘pointed’ scarp at Watling Lodge
The shape of the mound varied depending upon the terrain; in most areas it formed a gentle rise but where the ground to the north sloped off steeply, it formed a pointed scarp almost giving the impression of a northern rampart. This steeper mound can be seen near Watling Lodge.
The fortlets were similar in size and configuration to the so-
Once again taking experience learnt from the southern frontier, the Antonine Wall planned for garrison forts to be built on the frontier itself. The original intent placed five forts at intervals averaging 8 miles apart; roughly half a day’s march. Old Kilpatrick guarded the western end whilst Balmuildy, Bar Hill (some authors claim Auchendavy but this is unlikely based on the size of the fort, the multiple garrisons attested to Bar Hill and the naturally strong position of that site), Castlecary and the large cavalry base at Mumrills were founded in the original scheme in some cases re-
Before construction of the Wall had completed, the plan was changed with at least eleven new forts added to onto the line of frontier. Two fortlets were demolished as a result (Duntocher and Croy Hill). The revised plan provided a heavy concentration of troops; the forts were now at intervals of around 2 miles. What prompted this change is unknown. Certainly when the frontier returned to Hadrian’s Wall there was no attempt to increase the concentration of troops so the phenomenon was unique to the Antonine Wall. One explanation could be the unsuitability of the terrain for cavalry. The garrison of Hadrian’s Wall included multiple combined infantry/cavalry detachments whereas evidence for this on the northern frontier is scant; the only known cavalry base was Mumrills (although Castlehill had a mixed infantry/cavalry force and Bearsden is also suspected). Clearly the reduced distances between forts increased reaction time but this theory remains conjectural.
It is interesting to note that the forts were not all equal in size. Whilst the primary forts were all of a similar size – ranging from 3.2 to 4.2 acres plus the larger cavalry base at Mumrills at 6 acres – the additional forts were much more variable. For example Duntocher enclosed just 0.5 acre which wasn’t significantly larger than the fortlet it replaced.
Many of the forts on the Antonine Wall had additional fortified enclosures directly adjacent to them which have been labelled ‘Annexes’.
Aside from providing a secure area for activities, no specific purpose has been successful assigned to Annexes. At Bearsden it hosted the Bath House whilst elsewhere they seemed to support workshops or possibly civilian settlements. Historians have speculated that their purpose was a localised version of the Vallum – the ditch and rampart system that forged a military zone south of Hadrian’s Wall. The majority of the Annexes (all of those east of Balmuildy) were added after the relevant fort’s initial construction.
Expansions were square platforms extending from the rampart and constructed in the same manner, i.e. turf blocks on top of a stone base. The function of these structures are unknown but they are assumed to be for signalling.
One of the expansions at Croy Hill
The expansions measure around 5 metres square and were presumably built to the same height as the rampart. The suggestion that they were signal platforms has not been proven but the topography supports the analysis as they had line of site with forts north or south of the frontier. The expansions exist in pairs further supporting the signalling theory.
Aerial surveys have revealed a number of small enclosures, each measures just 5 metres square, directly abutted to the rampart near Wilderness Plantation fortlet. Seemingly built at the same time as the Wall the purpose of these remains a mystery.
Turrets / Watchtowers
There is no evidence for any turrets / towers along the length of the Antonine Wall (other than those integrated into the forts). It is hard to believe there were none for they are an integral part of all other Roman frontiers. It is likely, therefore, that the evidence of their existence has simply not yet been found in the excavated segments of rampart.
Click for larger map. Each part of Section 3 has a map showing more detail.
The fortlets on the Antonine Wall were similar in size and configuration to the Milecastles on Hadrian’s Wall.
Effectively acting as fortified gateways, the fortlets had timber buildings for the small garrison. The ramparts were the same turf and timber construction as the Wall itself whilst further defensive ditches protected the sides. A causeway was left unexcavated to the south to enable access. Interestingly, at known fortlet locations, no evidence of a causeway has been found over the the Wall’s ditch; perhaps they were removed when the design of the frontier was modified and more forts were constructed along the line.
The known fortlets found on the Wall to date are:
To the west of the Wall, but still key outposts on the frontier, were two additional fortlets:
In the first and second centuries AD, Roman forts were traditionally built to the same design which had evolved from the temporary marching camps which the army used on operations. Constructed in a ‘playing card’ shape the forts had a Headquarters building in the centre, a Commanding Officer’s house adjacent to the HQ, granaries for food storage and barrack blocks along the parameter.
This effective design was simply re-
Although the defences of many of the forts in the south were stone constructions, all but two of the Antonine Wall forts were built with turf ramparts. This was not unusual -
The Headquarters building at Bar Hill
The Bath House at Bar Hill
Reconstructed Gatehouse at the Lunt, Midlands
Reconstructed Granary at the Lunt, Midlands
The Antonine Wall was a layered frontier system consisting of numerous elements. The physical barrier was provided by a ‘V’ shaped ditch and a rampart constructed from turf. Military operations were sustained from a network of forts and fortlets all connected via a Military Way that ran the length of the Wall and beyond.