Towns flourished at Corbridge and Carlisle on the lines of the two main roads. These were substantial settlements by the standards of Roman Britain, if not quite on the same scale as the handful of big cities in the south of the province, let alone the grand cities of provinces like Gaul. In both Corbridge and Carlisle, there was always a military presence and distinct military compounds that housed substantial detachments of legionaries, but this did not alter the essential civil character of the communities. The Vindolanda tablets make it clear that many traders and businessmen were drawn to the northern frontier around AD 100, and the long-term presence of so many soldiers created new markets, as did the people who came to profit from the military community. It is becoming clear that villas and large agricultural estates farming for profit were established farther north in Roman Britain than we used to believe. Hadrian’s Wall was very much part of the wider imperial economy, with goods and people coming in from all over the empire.
Traders followed the legions even on campaign, and almost as soon as any army base was established and occupied in the longer term, civilian settlement swiftly started to cluster around it. As a fort became permanent, so did the community living around it, and it might eventually gain formal status as the vicus (or civil settlement) and some measure of self-government, the community acting together as the vicani. There were sometimes official buildings, such as the waystations, or mansiones, built to accommodate civil and military officials travelling on public business. A grand house outside the Hadrianic fort at Vindolanda was surely constructed for someone of importance, although the suggestion that it was built for the emperor himself during his inspection of the northern frontier is so far no more than an appealing conjecture. There is also evidence for army barracks in the vicus at Housesteads and Birdoswald at some periods, so the area outside a fort was not always purely civilian.1
This is one of the ‘chalet-barracks’ built in the middle of the third century AD at Housesteads. Unlike the earlier barracks, it was not a single range but a row of individual huts or chalets. These narrow structures appear to be equivalent to the pair of rooms that we believe provided accommodation for the men and equipment of an eight-man contubernium, who on campaign would share a tent.
On the whole, most structures within the vicus were smaller and simpler than any official buildings. Access to the roads, especially the main road leading out of the fort, was at a premium, so that buildings tended to be narrow fronted and deep—a style we refer to as strip housing. The front often acted as a place of business, whether a shop or bar. There were also larger taverns and the like, as well as workshops and temples—much the same mix to be found in any normal town or large village, save for the lack of grand public buildings. Most vici were crowded and bustling, and many were large. Recent survey work at Housesteads has shown that the civilian settlement covered an area several times larger than the fort itself, with housing on both sides of the Vallum. Many may have been smaller than this, but even so were substantial. The early third century AD was the time when the vici of forts on Hadrian’s Wall were at their height and showed every sign of prosperity, with plentiful supplies of everyday goods coming from far afield. So far only two, at Wallsend and Housesteads, appear to have had a rampart or ditch, or both, around them, with others simply being enclosed on one side by the Vallum and on the other by the Wall. At Maryport on the Cumbrian coast, the vicus was surrounded by a ditch.
Even when it covered a larger area outside the military compound, the vicus depended on the fort that provided the reason for its existence. Wall forts conform in most respects to the typical design of Roman auxiliary forts in the western empire, shaped like a playing card with rounded corners and four gateways, one in each wall. The main gate, invariably double arched, led onto the most important road in the fort, the via praetoria. This met the via principalis to form a T-junction, the second road running between the two side gates of the fort, and along it lay the key buildings of the base.
The principia (headquarters) stood at the junction of the roads. It was a very large courtyard complex, including an assembly hall, offices, storage space for the great quantity of records generated by the army, a strong room (underground whenever possible) to keep its pay chest and other valuables, and the aedes, the shrine where the unit kept its standards alongside images of the emperor. The life of the unit was regulated and recorded in this building.
Next to the headquarters was the praetorium, the house of the garrison commander (usually a prefect), which was often as large or even larger. Its design followed traditional Italian taste for the home of a rich man and was based on the courtyard house surrounding a small central garden—ideal for offering cool shade in a hot Mediterranean summer. Living quarters, kitchens, and other rooms were provided for the slaves and freedmen who performed much of the work needed to keep an equestrian gentleman in the style expected of someone of his rank. The Vindolanda tablets also make it clear that it was normal for the prefect to be accompanied by his family, so that women and children often lived at the heart of a fort. The tablets tell of birthday parties and other social occasions when senior officers and their families visited each other. All in all, the tablets suggest a very comfortable lifestyle, and even in their timber phases, these houses were large and well provided. When built in stone, there was the prospect of even greater comfort, with a private bath suite and under-floor heating provided by hot air from a furnace blown through a hypocaust system.
More functional, but on an equally grand scale, were the horrea (granaries), which usually also lay on the via principalis. These are very distinctive structures, with huge buttressed walls and the floor raised by stone pillars or small walls. This was intended to deter vermin, but also with the high ceiling, helped regulate the temperature and so preserve the cereals and other foodstuffs stored inside. It was usual for the main entrance of a horreum to be raised, making it easier to load and offload sacks, barrels, and other bulky goods directly from a cart or wagon. Granaries most often occurred as a pair, and it is striking that whenever a fort was wholly or substantially rebuilt on the same site, the new granaries were not in the same place as the old ones. This suggests not only their importance, but also that the old buildings were kept in use for as long as possible until the new ones were ready.
A valetudinarium, or hospital, was provided in each fort, although as yet too little is known about the layout of these to make it easy to identify them—the courtyard building behind the principia at Housesteads may well have been a hospital, but we cannot be sure. The strength report from Vindolanda lists thirty-one men in hospital, including six wounded, ten suffering from inflammation of the eyes, and fifteen others sick. Eye problems may well have been exacerbated by living in dimly lit and smoky barrack blocks, and it is interesting that we hear of an effective eye salve made from a range of ingredients, including mercuric sulphide, which was the work of an oculist in the British Fleet (classis Britannica) named Axius. One of the Vindolanda tablets mentions a seplasiarius, or pharmacist, and others mention medicines or their ingredients in lists of purchases. Army bases, much like ancient cities, were both crowded and periodically visited by individuals and groups from far afield, providing many opportunities for infections to spread, and one estimate suggests that the hospital in each base had the capacity to accommodate 5–10 percent of the entire garrison. Yet on the whole, the army provided better medical facilities than were available to poorer civilians, and soldiers were as well cared for as the medical knowledge of the age permitted.
One of the buttressed granaries at Housesteads gives an idea of the sheer scale of these functional buildings. They are very distinctive archaeologically. Note the rows of pillars used to raise the floor above ground level. Granaries were always built in pairs and formed part of the central range of buildings in a fort.
Much easier to identify from its remains than the military hospital is the balneum, or bathhouse. These were always of stone and always built outside the ramparts bec
ause of the risk of fire created by the furnaces, although this also meant that they were in the vicus, so there were sometimes other buildings nearby. Bathers went through a sequence of warm, very hot, and cold rooms. Olive oil was used as soap and the skin then scraped clean with a strigil. There was also space for other activities, from exercise to board and dice games and simply talking and eating. That the bathhouse was beyond the rampart may have added to the men’s sense that they were off duty.2
The bulk of space within each fort was taken up with accommodation for troops, the number of buildings varying depending on the type of unit. An infantry barrack block was designed to house one century of eighty soldiers. It consisted of a long building, for much of its length divided into pairs of rooms, with each pair allocated to a contubernium of eight men who on campaign would share a tent. As there were ten contubernia in a century, there ought to be ten pairs of rooms per block, and though this was sometimes true, other barrack blocks have slightly more or slightly fewer rooms. Assuming that we are correct to see a pair of rooms as accommodation for a contubernium, we have no idea of how the men slept, but bunk beds, straw paliasses, or cots are all possible. Two or three men may have shared a cot or bed, an arrangement common for much of human history even if it seems deeply alien to those of us born in the age of central heating. Hearths have been discovered on some sites, but they do not appear to have been universal in all periods. At Vindolanda, barrack room floors were covered in layers of rushes, heather, and straw. When dirty, these were not cleared out and replaced but were simply covered with fresh material, leaving a thick mat of dirty and decaying vegetable matter. Finds of lost possessions, including objects as large as shoes, are common, which reinforce the sense of dimly lit and crowded barracks, rife with insects and other vermin, especially in the summer months. At one end, the barrack block widened into a suite of rooms that housed the centurion in greater comfort, probably with plastered and painted walls, and perhaps also rooms for other junior officers such as the signifer (standard bearer).
As we have seen, around the 230s new barrack blocks were constructed that were smaller than the older pattern. These tend to have just five pairs of contubernium rooms but are otherwise very similar in design. This appears to have occurred all along the Wall. One variation seen at some sites are the so-called chalet-barracks, in which instead of a continuous range, each contubernium occupies its own separate little building with side walls and roof, a narrow gap separating the huts on either side. Presumably these were seen as easier to construct and maintain, for they do not seem to reflect any other change.
The stables required for the horses in the cohortes equitatae and alae remained something of a mystery until recently, when excavators at Wallsend identified several examples. Rather than a separate stable block, the Romans used a combined building similar in shape to a normal barrack block. Along one side of the long building was a row of rooms to accommodate the thirty or so troopers of a turma (the cavalry sub-unit commanded by a decurion and equivalent to an infantry century), each room usually equipped with a hearth. Backing onto these was a row of horse boxes, identified as such by the ‘urine pits’—shallow trenches below flooring that allowed liquid to drain into them. This configuration was presumably to make it easier to clean out the stalls and replace the bedding. The rooms for the men and the boxes for the horses on either side of the building do not connect, each one having to be entered from the outside. Each room on one side housed three troopers, with their three mounts in the adjacent (if unconnected) horse box. It is quite possible that there was an attic roof, providing further storage and perhaps even living and sleeping space for the grooms, usually slaves, who assisted many cavalrymen in caring for mounts and equipment. As with infantry barracks, these buildings do not quite conform to what we would expect, having only nine sets of rooms instead of the ten that would have been required if there were thirty men and horses in a turma.
Like infantry barracks, the combined stable-barrack blocks had a suite of rooms at one end for the commander and other officers. However, these do not appear to include any stabling for horses, and it is unclear where officers’ mounts were kept. Living so close together, a pervasive smell of horse sweat, manure, damp leather tack and saddles, straw, and grain must have mixed with all the human smells of clothes, food, and cooking. Any strongly garrisoned Roman fort was a very crowded place, with people and animals in close proximity, but most towns and cities were also very densely packed. It is also worth remembering that recruits to the auxiliary cavalry often came from rural communities in the provinces where people lived close to the land, with animals around them much of the time.
This full-size reconstruction of a third-century barrack block at South Shields gives a good idea of the appearance of soldiers’ accommodation. Barracks in earlier periods were similar in shape but significantly longer. There is no direct evidence for barracks in Britain having more than one storey. Archaeologists traditionally try to reconstruct the theoretical size of a garrison on the number and layout of barrack blocks. However, there is much we simply do not know. Several rooms in this reconstruction are presented to show alternative versions of how soldiers may have lived and slept in them.
Slaves, including many owned by the army itself—known as galearii (helmet wearers) and given a simple uniform and some training—added to the inhabitants of a fort but were not the only noncombatants. From the time of Augustus, Roman soldiers were not permitted to contract a legal marriage, until this ban was lifted by Septimius Severus some two centuries later. In the main, this restriction allowed the state to accept no obligation to care for wives or children during or after a soldier’s lifetime, for it is clear that in practice there was little effort to prevent men from forming long-term liaisons with women and raising families. Soldiers giving their origin as born ‘in the camp’ (in castris) were the products of such unions and readily accepted in the ranks once they were old enough to enlist. As far as the men were concerned, these were proper marriages. Hadrian passed a law making it easier for soldiers to bequeath property to their families, something that was clearly very important to them. On discharge, auxiliaries were granted citizenship not simply for themselves, but also for one wife and the children the man had with her, an acknowledgement that many already were ‘married’.
This raises the question of where soldiers’ wives and children lived, and the answer is not straightforward, perhaps because it varied from unit to unit and place to place. At Vindolanda, there is clear evidence to show that the families were in the barracks with the soldiers, for objects associated with them, including shoes, are frequent finds in the rush and straw material used as floor matting. Such things were clearly dropped and lost by the residents in these poorly lit buildings and were not dumped there as rubbish sometime later. At other sites, such as Housesteads, there is no good evidence that women and children lived inside the fort itself. If a man’s family lived in the vicus, then it is hard to know how much time he would have been able to spend with them—or whether some or all were permitted to spend their nights outside the rampart of the fort. Centurions seem always to have been permitted to contract a marriage, and the rooms allocated to them in a barrack block may well have been a home to their families as well.
Looking at a plan of a Roman fort, it is the living space rather than the fortifications that stand out. When the vicus is included, the whole community looks far more like a garrison town than a purely military installation ready for war. This was where soldiers lived, worked, drilled, trained, ate, slept, relaxed, and raised families. The famous multi-seat latrine block at Housesteads, flushed by a flow of rainwater stored in tanks, gives one example of a mundane activity carried out in a solidly constructed and purpose-built structure. There was a strongly urban feel to life in the army’s bases. There was also a commonality of layout and design in the army’s bases all over the empire, and especially in the western provinces. Soldiers arriving at a garrison would see much that was familiar and
reassuring.
Yet although forts were similar and recognisably of a kind, no two were identical, and the fuller excavations carried out on some sites in recent years reveal successive re-designs that were often drastically different. At Vindolanda in the late second or early third century, there was a phase when an area either in the fort itself or a fortified annex was occupied by people living in traditional Iron Age round houses. It is impossible to tell whether these were friendly local communities, hostages or prisoners, a force of locals used as labour, or soldiers who preferred to live in the traditional way.
Around the time of Septimius Severus’s campaigns in Scotland, the fort at South Shields, on the south bank of the River Tyne (behind the eastern end of Hadrian’s Wall), was converted into a huge depot, with no fewer than twenty-two granaries within its walls. Positioned on the mouth of the Tyne, the fort presumably had a port nearby, but it has not yet been located. This huge supply base was part of the logistical support for the exceptionally large expeditionary force brought by Severus. Although overall troop numbers in the north declined after Severus, there were still substantial forces on and around the Wall, and South Shields probably continued to play an important role in supplying them for several generations.
Seven
LIFE ON THE WALL
IT IS HIGHLY LIKELY THAT outbreaks of serious conflict in northern Britain go unmentioned in our meagre literary sources. As there is good evidence for at least another major campaign under Hadrian apart from the one at the start of his reign, it is a safe bet that other unattested wars were also fought in the second century, let alone during the even more poorly recorded third and fourth centuries. Small-scale military operations are unlikely to appear in our sources even for well-documented periods. In the fourth century, the soldier and historian Ammianus Marcellinus ended his account of some fairly minor operations in Gaul by saying that ‘besides these battles, many others less worthy of mention were fought… which it would be superfluous to describe, both because their results led to nothing worthwhile, and because it is not fitting to spin out a history with insignificant details.’1
Hadrian's Wall Page 7
