6

Let One Stone Stand Upon Another

In 1933 Patrick Leigh Fermor disembarked at the Hook of Holland to begin an epic journey. At the age of eighteen, expelled from school, with only a pound a week to live on, he had decided to walk across Europe. Many years later, he wrote down what he had seen, an extended glimpse of a now-vanished Europe, in two volumes of a promised trilogy, A Time of Gifts and Between the Woods and the Water. Following first the Rhine and then the Danube, he kept a diary (later lost) of his travels, and it ends at the Iron Gates, a sequence of spectacular river-gorges. Between the Carpathian Mountains on one side and the Balkans on the other, the Danube races and swirls, its currents deep, treacherous and violent. As Leigh Fermor passed through on board a steamer, he noticed a Latin inscription incised directly into one of the towering cliffs.

The Emperor and Caesar, son of the divine Nerva, Nerva Trajanus Augustus Germanicus, High Priest and for the fourth time Tribune, Father of his Country and Consul for the fourth time, cutting into the mountains and on wooden beams, raised up this road.

When the skipper expertly steered his boat around the whirlpools and underwater rocks, he sometimes sailed very close to the sheer rock face. Leaning over the rail, Leigh Fermor could see something remarkable. No more than a few metres above the boiling river a continuous slot had been quarried, hacked directly out of the cliffs. Two and a half metres high and three metres deep, it was large enough to allow men to walk two abreast. But it was not, as Leigh Fermor believed, a Roman road. In AD 102 Trajan planned to lead a huge army to invade and subdue Dacia, much of modern Romania, and he wanted his soldiers to be supplied by river. Because the currents of the Danube were so powerful as it sped through the Iron Gates, barges would need to be towed upriver. A wooden platform wedged into the slot (replacing an earlier and more fragile structure first built under the Emperor Domitian) would not carry marching legionaries but probably teams of men hauling on hawsers, dragging the barges through the gorge.

Further downstream, at the Great Kazan, where the current is calmer and the channel widens, Trajan ordered a bridge to be built across the Danube. Designed by the Greek architect Apollodorus, it was to become one of the wonders of the ancient world. It reached more than a kilometre in length, was 15 metres wide and raised on twenty stone piers. Completed in AD 105, it was the longest bridge in the world, and would remain so for another thousand years.

Apollodorus faced tremendous structural problems. Thirty metres deep in places and liable to seasonal spates of great ferocity, the Danube could be unrelenting, washing away months of work in moments. But using coffer dams and huge water pumps, and building in bricks, mortar and pozzolana cement (which hardens when in contact with water), all twenty piers were finally sunk and the current diverted around them by sharply pointed cutwaters. A wooden superstructure was prefabricated on shore and lifted into place. And in the spring of 105 Trajan led the legions across, crushed the Dacian army and brought their lands into the Empire.

When his steamer arrived at the same place, here is what Patrick Leigh Fermor saw:

It was the remains of Trajan’s amazing bridge that we had come to see, the greatest in the Roman Empire. Apollodorus of Damascus, who built it, was a Greek from Syria, and two great stumps of his conglomerate masonry still cumbered the Rumanian side; a third stood across the water in a Serbian meadow. Swifts were skimming over the water and red-legged falcons hovered and dived all around these solitary survivors of twenty massive piers. Once they had risen tapering to a great height and supported over a mile of arched timber superstructure: beams over which cavalry had clattered and ox-carts creaked as the Thirteenth tramped north to besiege Decebalus in Sarmizegethusa. On the spot, only these stumps remained, but the scene of the dedication is carved in great detail on Trajan’s Column in Rome, and the Forum pigeons, ascending the shaft in a spiral, can gaze at these very piers in high relief: the balustered bridge soars intact and the cloaked general himself waits beside the sacrificial bull and the flaming altar with his legionaries drawn up helmet-in-hand under their eagle standards.

As Leigh Fermor saw, only the abutments at either end of the great bridge are still visible, but in a year of exceptional drought and low water (1856), all twenty piers were revealed. Trajan’s bridge and towpath were extraordinary structures, two of the greatest feats of engineering undertaken in Europe before the 19th century. Emblematic of the age, the zenith of the Roman Empire, they set Hadrian’s Wall in a clear context. Emperors saw themselves as the Lords of the Earth, not hesitating to alter or overcome geography in pursuit of glory, in the drive for empire.

LUPERCALE

The legend of Rome’s origin centres around the unlikely story of a she-wolf coming to the rescue of the twins Romulus and Remus. Abandoned on the banks of the Tiber, they were taken by the wolf to her cave and she suckled them and saved their lives. No wolf, no Rome. The rest was history (rather than legend). Known as the Lupercale, the cave and the rescue were thought to be little more than myth. But, in November 2007, Italian archaeologists held a press conference. Excavating on the Palatine Hill on the site of the Emperor Augustus’ palace, they had found a cave, a cave which had been a shrine. It was decorated with mosaics, sea-shells, pumice stones and had a representation of a white eagle at its centre. Was this the Lupercale, the sacred place where the wolf had suckled Romulus and Remus? It appears that Augustus, who understood well the power of symbols, had attached himself securely, in architectural terms at least, to the history and destiny of Rome. Professor Giorgio Croci led the excavation, and became very excited: ‘You can imagine our amazement, we almost screamed.’ The Minister of Culture, Francesco Rutelli, was calmer: ‘This could reasonably be the place bearing witness to the myth of Rome, one of the most well known in the world.’ In March 2008 tourists were at last admitted to Augustus’ private apartments. Small and modest, they were decorated with standard frescoes for a man of his status and offer no clues as to his nature.

Hadrian was with Trajan on the Danube and he saw with his own eyes what could be done. With thousands of battle-hardened and highly skilled soldiers at his command, was there anything a determined emperor could not build?

His bridge allowed Trajan to strike fast and deep into the Dacian heartlands. Their king, Decebalus, retreated in front of the legions, hoping, like Calgacus and the Caledonians, to lengthen their supply lines to breaking point at best, and slow and weaken Trajan’s advance at worst. By a remarkable quirk of archaeological survival, what happened next is understood, but through no written source. The spiralling panels of Trajan’s Column show a Roman cavalry trooper riding hard towards a Dacian who holds a curved dagger to his own throat. The scene is set in dense woodland and the trooper leans urgently forward, half dismounting, arm outstretched, in an attempt to prevent this man from taking his own life.

This is the capture of the fugitive Dacian king, the great Decebalus. And the trooper is Tiberius Claudius Maximus. Buried in a field in northern Greece, his headstone was recently recovered and it lists his exploits in a long career with the VII Legion. Maximus’ troop had been tracking Decebalus and his band of loyal die-hards, but when the king realised that he would be caught, that there would be no escape, he took his life. Refusing to be captured, taken in chains to Rome, humiliated, whipped through the streets behind a triumphant Trajan, he slashed his carotid arteries and bled to death. Better to end by his own hand than be strangled in the black depths of the Mamertine Prison. Maximus cut off the king’s head, tied it by the hair to his saddle pommel and rode back to headquarters to give it to his conquering emperor.

Dacia was a vast province, much of it mountainous, all of it difficult to hold down. But Trajan had bridged the mighty Danube, brought the lands to the north into the Empire and sent back to Rome many thousands of slaves and much booty. He used these immense prizes to build a new forum and market in the city, dedicated in 112. The Senate commissioned the great column to take a central place. Rome was marching once more down the roads to glory.

In the east, since the age of Augustus and before, the emperors’ only substantial rivals were the Persians, known as the Parthians in the early second century. Based on the old dominions of the Babylonians, their empire was rich, powerful and very attractive. In 114 Trajan assembled the legions at Antioch in the province of Syria. Hadrian was at the muster, as a senior staff officer, perhaps second-in-command.

The Emperor was sixty by AD 114 but his plans betrayed the vast ambition of a much younger man. Like Alexander the Great, he would stride with his soldiers across the east, and the lands of the Tigris and Euphrates would fall before his unstoppable advance. And that is how it began. Armenia, the mountain kingdom to the north of Mesopotamia, the Empire between the two rivers, was to be the first to be defeated, humbled and brought into the orbit of Rome. On the long march from Antioch, Trajan held what sounds like an imperial durbar from the days of the British Raj: the satraps and princes came to meet him with gifts, one of which was a horse that had been taught to do obeisance, kneeling on its forelegs and placing its head beneath the feet of whoever stood near.

Armenia fell, then Mesopotamia, and by 116 the legions saw the sea once more. Incredibly, Trajan had led them from one stunning victory to another, and when he and Hadrian gazed on the waters of the Persian Gulf, Rome had reached right across the known world. From the seas of Arabia to the chill waters of the River Tyne, the legions had conquered. It was breathtaking – and impossible.

Even Trajan knew it. In 117 a client Parthian king was installed in southern Mesopotamia, and Roman soldiers withdrew from the Gulf. The price of glory could not be paid. The old Emperor had hoped to consolidate and stabilise in Dacia, complete a wide-ranging subjection of the east and hold the line elsewhere. But it was very precarious. Imperial overstretch on this scale needed only two major wars to break out in different parts of the Empire to descend into imperial collapse. As Trajan lay dying in Antioch, and the Empress Plotina moved quickly to manoeuvre Hadrian into the line of succession, it was clear that policy had to change.

Five years later, when Hadrian sailed up the Tyne to the Newcastle quays in 122, he had formally renounced almost all that the great Trajan had achieved. While he still ruled an empire larger than Augustus and all the emperors of the first century, Hadrian was forced to give up what could not be held. While the Parthians had been cowed, the east shrank back to its frontiers before 114. Dacia was retained but no more wars of conquest were planned. As he progressed around the boundaries of his empire, Hadrian set about the work of consolidation. The map began to settle.

HADRIAN’S TRAVELS

Having visited virtually every province in the Empire, Hadrain was almost certainly seen in the flesh by more of his subjects than any of those who reigned before or after him. In 121 he began his first tour by riding from Rome to Lyon and then to Germany and Middle Europe to inspect the defences of the Rhine–Danube line. Perhaps his fleet called in at London before sailing on to the Tyne in 122. After the Wall was begun, the sprawling imperial retinue packed its bags to travel right across France to Tarragona in Spain. Then, in perhaps his most spectacular year, 123, Hadrian crossed the Strait of Gibraltar into Africa, the province of Mauretania. From there he sailed the length of the Mediterranean, following the Atlantic current, to Antioch in Syria, where he was first proclaimed Emperor, thence across eastern Turkey to the Black Sea coast and its Greek cities, where he met Antinous. Back in Rome by 125, Hadrian left for Africa in 128, Greece and southern Turkey in 129, Palestine and Egypt in 130, and finally back to Rome in 133–4. His backside was well used to the saddle and his stomach to the roll of the waves.

Before his arrival in Britain, the new Emperor had ridden through the German forests inspecting his garrisons – and making a more emphatic frontier. In the 1st century a policy of defence in depth had been pursued. There had been frontier zones dotted with forts, depending on good communication, good roads and good intelligence. A limes or frontier path (giving the English word ‘limit’) was more likely to lead directly into enemy territory than lie transverse on a boundary between the Empire and the barbarians beyond. Control, not physical borders, was what mattered.

Nevertheless, in Germany and Middle Europe, handy natural lines of demarcation flowed through the landscape. Until Dacia was taken, the Danube supplied a clear frontier in the east, and in the west the line of the Rhine defended the Empire. To connect what geography had already done for him, Hadrian caused a very long timber palisade to be built in 121–2 between the two great rivers. In some ways the original arrangements resembled the Gask Ridge in Scotland of forty years before. In Germany, Domitian had sanctioned a line of watchtowers half a kilometre apart and linked by a limes, a path. Turf and timber forts were raised along a line which stretched for an immense distance through the forests. It ran for more than 500 kilometres between the Upper Danube and the Rhine, through the Taunus, Wetterau and Odenwald regions. When Hadrian arrived, he made radical alterations, converting the open system of defence into a solid barrier.

In 121–2 the legionaries embarked on building a vast fence. Ditches were dug, and on the upcast a timber palisade was driven in. Oak was the wood of choice and, once a huge number of trees had been felled, perhaps a quarter of a million, they were sawn into lengths. Laid on their sides, the trunks were rived lengthways with wedges and sledgehammers. The flatter, rived surface of white heartwood was turned outwards to present a more wall-like obstacle. Then it was stiffened and tied with rails nailed cross-ways. A later biography of Hadrian in the Historia Augusta recorded that at that time and frequently at other times he marked off the barbarians in many places, where they are separated not by rivers but by limites with great posts driven into the ground and joined together like a wall. Either side of the new barrier the forest was cleared and a military zone created. Clearly the palisade, made from perishable materials available close at hand, was not intended as a solid rampart designed to repel an assault. Rather, it acted like the perimeter of a Roman camp or fort, able to slow down an attack but not stop it in its tracks.

Less dramatically, the new German frontier was also a matter of clear definition, a marking off. South of it was Rome, the Empire, the lands of the citizens, cities and civil order. North lay the trackless forests of Germania Barbarica. The fence also controlled movement in and out of the Empire. Those who wished to pass between could do so peacefully at a crossing-point, and also pay the portaria, a version of customs dues. As at all places of international exit and entry, intelligence could be gathered from the hinterland and, if required, Roman legionaries and auxiliaries could act on it and travel quickly through the barrier if trouble flared.

This was different. A tangible break with the forward policies of the immediate imperial past, the new frontier set an unambiguous limit on the ambitions of Rome. Virgil’s god-given destiny of conquest had been repudiated by Hadrian, and those aristocrats who craved the excitement, the honour and the wealth which came with an expanding empire will have been disquieted as the oak trees crashed to the ground in the German forests.

But a fence? A wooden fence in the depths of the woods? Compared with the blaze of glory trailed by Trajan as he fought his way to the Persian Gulf, and the creation of a mighty bridge to reach into the heart of Dacia, the building of a fence certainly struck a minor key – which did not suit Hadrian’s nature.

A year later, in 122, when the Emperor and his new Governor of Britannia, Platorius Nepos, watched their engineers drive iron-tipped wooden piles into the muddy bed of the Tyne at Newcastle, they must have talked and planned. There had been war in Britain in 117, and probably fighting close by, along the line of the Stanegate. But Hadrian would not lead a punitive expedition north; that was short-termism and in any event looked very like more conquest. Nor would he order the construction of a frontier like the Gask Ridge or the German forests. If Rome was to retrench, then let it be a triumphant retrenchment. Timbers rotted in the ground and collapsed, ditches caved in. Britannia would be different, the scene of a different version of glory. Let one stone stand upon another: let there be a mighty stone Wall.

There was, in any case, no convenient forest to hand. When Hadrian, Nepos, the Praetorians and his surveyors and labourers set out westwards from Newcastle, they rode through a cleared landscape. But stone could be quarried from it, and once it had been established that there were in fact quarries accessible and sufficient to the task – surely the first issue to be dealt with by the military planners – then a stone Wall offered the opportunity not only to build but to build on a spectacular scale.

As the imperial retinue progressed, they moved through the valley of the lower Tyne, a flattish stretch of farmland. The line of the Wall hugs the north bank of the river as the land rises gently towards Benwell. The Celtic name of the Roman fort, Condercum, meant something like Viewpoint Fort and from the crest of the rise it is possible to see some considerable distance both downriver and up as well as north across the Northumberland Plain and south to the high ground at Birtley, where the Angel of the North now spreads his wings.

Decisions were probably made on the hoof. Based on research doubtless done by forward parties of surveyors and scouts, Hadrian and Nepos laid down the line of the Wall as they went. Not only did the Emperor fancy himself as an architect, he was, like most famous Roman generals, reputed to have an eye for good sites for forts and roads. And he had an irresistible urge to interfere, to become involved in detail. Following his directions, the mensores pegged out the line with stakes or marked it with small cairns. No one cared whether or not the Wall cut across good agricultural land. Native farmers were simply removed, lock, stock and barrel. Perhaps they were compensated.

When the frontier had moved up to a new line in Germany in AD 83, the Emperor Domitian specifically ordered that farmers who lost land as a consequence were to be paid for the loss of that year’s harvest. The issue of ownership and any transfer which may have taken place is not clear, but Roman politicians were fussy about property rights (they generally owned a good deal of property themselves) and it is likely that along the line of Hadrian’s Wall some sort of transaction was worked out between the army and local farmers.

At Throckley, a little way west of Benwell, archaeologists have found the marks of native ploughing under the Wall and it looks as though it was fresh. Farmers had probably ploughed in the spring of 122 and sown their fields before the Emperor and his army arrived to obliterate them. It must have been baffling as well as savage. Building a wall? Plough marks have also been found under the forts at Wallsend and Carrawburgh as well as at Wallhouses, 8 kilometres west of Throckley. No doubt many other sites were built over farm fields.

After Rudchester the landscape begins to change, climbing from the coastal plain up to 120 metres above sea level. Further west, at Haltonchesters, there is little trace of the fort but the modern road runs arrow-straight, following the line of the Roman road. Hadrian will have noted Dere Street as the old north road disappeared over the horizon and given thought to how it might penetrate the new Wall. Only a short way to the south lies Corbridge, the Roman fort and settlement at Coria, on the banks of the Tyne. Having clattered down the paved road, it is likely that the Emperor, his Governor and their entourage stayed there, perhaps using it as a base for their reconnaissance of the eastern sector.

At Chesters Fort near Chollerford, the line of the Wall dips down into the beautiful valley of the North Tyne. Clearly a second bridge would be needed and the sites of its abutments were duly fixed on. Once the imperial party had moved on from there, the landscape began to change again. On the long hill up to Walwick, the contour lines crowd together, and upland pasture takes over from arable farmland. At Black Carts a substantial run of surviving wall shows how steep the incline is and, at Limestone Corner, it reaches its most northerly point. This is the beginning of the central section and its geography may have been one of the most persuasive factors in following the line Hadrian chose.

The vantage from Limestone Corner in all directions is breathtaking, especially to the north as the ground falls away to the valley of the North Tyne. And if the view from the Wall was good, then the view of it must have been equally impressive. That was important in choosing sites. As triumphant retrenchment reached across the waist of Britain, it had to make an obvious statement, be something seen by allcomers. Gaze upon my works, ye mighty, and despair. This was to be no fence in a wood.

From Limestone Corner moorland flattens and stretches away to the east but, after less than half an hour in the saddle, Hadrian will have caught his first glimpse of what his surveyors had been talking about, the feature which had drawn them to propose this line: the Whin Sill. It was perfect for what the Romans intended. Sheer cliffs to the north and a fairly gentle slope up over grassland to the south. Geology had been particularly helpful to Hadrian.

The Whin Sill is the result of an ancient collision. An unimaginably long time ago, around 420 million years BC, the crust of the Earth was moving, forming and reforming enormous continents, filling and draining vast oceans. What became southern Scotland lay on the edge of a huge landmass and, separated by a prehistoric sea, northern England lay on the rim of another. When these two continents collided, Scotland’s harder rocks ground and scraped over England’s leading edges, the crust of the Earth corrugated and buckled, the floor of the ancient sea was squeezed upwards, and the foundations of the landscape of Northumberland and Cumbria were laid down.

Like the wavelets of an indrawn tide, the folded ridges of the landscape north of the Hexham Gap supplied the raw materials for Hadrian’s dramatic plans. Not only did the cliffs of the Whin Sill act as the foundation, one of the other results of that ancient collision was a profusion of outcrops, and they too were helpful. Quarries were found near the surface and often in vertical faces, and coal seams also peeped through in various places. These coal-heughs were to prove extremely useful sources of fuel in a treeless landscape.

The Whin Sill is broken by several steep-sided nicks. One of the most famous is at Sycamore Gap, and like the track of a fairground rollercoaster the Wall swoops down from the clifftops, sweeps through the Gap and up the other side. As much as anything these sections of Wall are examples of imperial single-mindedness and unflinching obedience. The legionaries no doubt cursed loudly and shook their heads as they built a stone wall up these near-vertical inclines – but they did it.

The cliffs end abruptly above Greenhead, and the Wall descends to the flatter ground of the Irthing Valley and the Solway Plain beyond it. From the crossing of the river at Willowford to its terminal at Bowness-on-Solway, turf was at first used instead of stone as a basic building material. The Wall must have run over grassland on this stretch. The change was not dictated by a lack of quarries and good stone. Later in the second century the turf was eventually set aside and replaced by stone. The initial difficulty had been a shortage of the limestone needed to make lime mortar.

Hadrian extended the run of the Wall to Bowness in order to cover all of the Solway fords then in regular use. But genuine sea traffic was also seen as requiring close control. Beyond the end of the land-wall, a sea-wall hugged the Cumbrian coastline, possibly reaching as far south as Ravenglass. Five major forts were built at regular intervals and forty-nine turrets and milefortlets have so far been plotted. There was no Wall as such, no need for one, but there was almost certainly a connecting road and possibly some ditching. This appears to have been slight, little more than 1.5 metres wide and 1 metre deep, and perhaps it was the boundary of a military zone. Elsewhere there are the faint shadows of parallel ditches.

Hadrian’s Sea-Wall is often ignored or written off in a footnote, but it was clearly part of the original intention, and the fort at Maryport not only dates earlier, possibly built in AD 72, when Petilius Cerialis was at Carlisle, it was also put under the command of Maenius Agrippa, a friend of the Emperor and presumably a trusted senior officer. In all, from Wallsend to Ravenglass, the Wall ran more than 180 kilometres, an immense, complex and fascinating monument to the power and reach of Rome.

ANOTHER GREAT WALL

Stories of the Far East had filtered to Rome for generations. Trade brought spices, fine silks and tales of exotica. Hadrian himself had been to the Persian Gulf, perhaps a younger Trajan would have emulated Alexander the Great and led armies to India. Antoninus Pius is known to have received ambassadors from China, from the emperors of the Han dynasty. They too were interested in walls, and had been building them for some time. The Great Wall of China, with its broad walkways, crenellations and turrets dates very much later, having been built for the Ming dynasty in the sixteenth century. The wall that Hadrian and Antoninus may have heard about was in some ways more impressive. Built mostly of rammed earth and gravel, the walls raised to defend the northern borders of China had existed since c. 200 BC and the reign of the first emperor of all China, Qin Shi Huangdi. In all, these ran for an astounding 6,400 kilometres and are the largest man-made structures ever built. Perhaps Hadrian was impressed, even inspired.

The symbolism invested by Hadrian in his Wall was not the only reason for its construction. The great labour was undertaken for several practical purposes. As in the German forests, clear demarcation – something unmissable – was important. In the new world of consolidation, being either side of the Wall would have real meaning – in the Empire and outside it. The Wall’s presence had military value (not as a fighting rampart: it was too narrow, not built to allow enfilading fire along its length and not meant to be defended like a medieval castle) in that it could prevent large numbers of enemy warriors from getting through it quickly. The Historia Augusta records a simple strategic rationale. The Wall was built qui divideret Romanos barbarosque. The sense of the verb, it has been persuasively argued, is stronger than merely to divide or separate. It really means ‘to force apart’ the Romans and the barbarians. And this reading becomes even more persuasive when the political landscape is looked at more closely.

The Brigantes, the largest native kingdom in Britain, appears to have been a federation including the Carvetii in the Eden Valley in the east, and the Textoverdi, the Lopocares and the Corionototae along the banks of the Tyne. When King Venutius attempted to defend his huge hillfort at Stanwick, he called on help from outside, that is, from the north. Agricola’s lines of advance in AD 79 encircled the dangerous Selgovae, and the construction of the Sea-Wall down the Cumbrian shore clearly anticipated a threat from the Novantae in Galloway. By driving a concrete, heavily guarded and unmistakable frontier through the Hexham Gap, Hadrian forced apart this powerful alliance of the hill peoples of the north. Of course their kings could communicate, and it seems that sometimes they did succeed in inflicting damage, but the glowering presence of the Wall would prevent them from combining quickly and in force.

Economic imperatives rarely lagged far behind military objectives in the Roman Empire. Just as in Germany, those passing through the Wall probably paid portaria, a tax on movement and goods. Cattle and sheep rustling in the hills and upland valleys of the north Pennines and the Cheviots was not the invention of the Border Reivers of the sixteenth century. Its origins were ancient, and a useful economic effect of the Wall was to prevent cross-border raiding almost entirely (or at least without a considerable outlay in bribery). At the height of the lawlessness of the 1570s and 1580s Elizabeth I of England’s council seriously considered rebuilding Hadrian’s Wall, rejecting the idea only on the grounds of cost.

Much more danger appears to have lurked in the western districts around the Wall. There was thought to be no equivalent need for a Sea-Wall down the North Sea coast, and no outpost forts were built (at first) in the east. The Votadini were not only almost certainly friendly to Rome but also key providers of essential supplies to the large garrison. In the west it was different, and three forts were established in forward positions, at Birrens, Bewcastle and Netherby. The last was known as Castra Exploratorum, the Fort of the Scouts. Patrolling and intelligence gathering were clearly seen as necessary at that end of the Wall. If Hadrian hoped to force apart the barbarians, it seems that he meant to achieve it in the centre and the west. The balance of the garrison was weighted there. In the east, the Tyne was probably already a frontier between the Votadini and the peoples to the south.

Perhaps because it was more settled, building began from the eastern end. And work began everywhere at once. Or so it must have seemed to the native peoples. No doubt exhibiting the impatience characteristic of those in high office, Hadrian will have demanded immediate action, and in the late summer of 122 progress would have been rapid and dramatic.

SOUVENIRS

Not long after it was completed, the Wall appears to have become a tourist attraction in Britannia. The Rudge Cup, the Amiens Skillet and the Staffordshire Skillet all carry the names of forts inscribed on them, usually around the rim of the vessel. All have a sequence of forts at the western end, and not the entire run. Like mugs with Blackpool or Bournemouth painted on them, they seem to have been popular. The survival of three artefacts of such an individual type is surely significant. Discovered recently, in 2003, the Staffordshire Skillet, or patera, has a name inlaid: Aelius Draco. Perhaps he was a soldier who served out his time on the Wall and his messmates had a memento made for him.

Once the line of the Wall had been accurately surveyed and marked (the Roman military made extensive use of maps), the countryside burst into a frenzy of activity. Detachments of soldiers were despatched to quarries, woods, sandpits, rivers and streams to discover local suppliers of all manner of goods, to build limekilns, blacksmiths’ forges and workshops. Smoke from thousands of fires plumed into the Northumberland skies. Not until the Industrial Revolution seventeen centuries later would the landscape of the lower Tyne see such toing and froing.

Three legions were commanded to build the Wall. The II Augusta, the VI Victrix and the XX Valeria Victrix represented a combined workforce of at least 7,000 men. Each legate will not have needed reminding that their men had been set to work in hostile country. Having left a cohort behind at their regional bases at York, Chester and Caerleon, the legions also required protection and their first action will have been to build secure quarters. In the midst of so much activity, it has been difficult to detect where they camped, but while use will have been made of existing forts, the construction of the Emperor’s Wall was a field exercise and his soldiers will have pitched their leather tents behind a palisade. Scouts no doubt rode out on regular patrols to anticipate enemy activity. There had been war in the north only three years before and, as each man worked, his weapons were not far away.

The legions built the Wall because they had all the necessary skills. In addition to their primary role as heavy infantry, many men were also stonemasons, blacksmiths, carpenters and carters. And each legion was used to operating as a self-contained unit. Hadrian and Nepos decided to use this to advantage. Construction of the Wall was split up into legionary lengths of about eight kilometres, what was thought to amount to a season’s work. And there would be productive rivalry between the men of the II, the VI and the XX.

Building was supervised by a Clerk of Works known as the Praefectus Castrorum. His first task was not to send gangs to the line of the Wall but to assemble all the logistical elements needed to make the project happen. It was very complex, and logistics were the key to success. As a general rule of thumb, for every one man building the Wall itself there were a further eight working to support him, supplying materials, digging ditches, watching the horizon for warbands of enemy horsemen.

Gangs worked ahead of each other in a clear sequence. After the turf was cut out and preserved for later use, a shallow trench was dug along a line pegged out with cord by the surveyors. Flags, boulders and other large loose stones were brought forward by oxcart (certainly at the east end of the Wall where the ground was not difficult) and bedded into the clay or earth, making as level a foundation as possible. Where the Tyne was navigable, the river will have been used to bring up materials. Barges and ships were the largest form of bulk transport available to the Romans. Meanwhile at the nearby quarries (the furthest from the Wall appears to have been Black Pasture, 1.3 kilometres from the fort at Chesters) men were clearing vegetation and throwing down bottoming to make hard standing so that they could get at the stone and get it out.

More stonemasons probably worked at the quarries than on the Wall itself. With difficult journeys by oxcart and pack-horse over rough country and often uphill, it was vital to keep carriage weight to a minimum. Once stone had been levered or cut out with wedges, it was roughly sized according to its use. Ordinary Wall stones were manageable and could be loaded onto a cart or a pack-saddle by one man, occasionally two. These blocks of what is known as squared rubble were cut flush at one end (the best end, according to the grain) usually with a scappling hammer rather than a chisel and mallet. The blocks were then tapered away from the cut face into a blunted pear-shape. This was done to make it easier for the less skilled men at the Wall, allowing them to bed stones quickly, only having to present a keyed course on the outside; what we see now. The tapered end was set to the inside so that it bonded better with the rubble and clay core, and also did not touch its neighbours at the sides and need to be cut to fit. This method of working made for rapid progress.

GANGS OF THE NEW WALL

Within the legionary lengths, there were sectors run by each centurion. These were worked by gangs with strictly organised roles. It is estimated that at the site there were 30 men working as a unit, 15 on the north side and 15 keeping pace with them on the south. In each gang, 3 men laid courses of stones and beds of mortar, 4 mixed, kept wet and brought forward mortar, 3 filled the core while 4 provided clay to bond it and there was 1 general labourer, probably the youngest – the lad who would have made the tea if they had had any. Down at the quarry 55 men worked. Of these the vast majority both quarried and roughed out stones, and 10 or so worked on the discarded rubble, breaking it down so that it could be used to fill the core. These numbers say nothing about transport, wood-felling and scaffold-work, sand-quarrying or any of the myriad other tasks. They are the estimates of an experienced modern mason.

The quarries mostly produced sandstone and gritstone, both sedimentary and easy to work. The whinstone from the Sill was usually too hard. Mortar was made from limestone and, happily, the eastern and central sectors appear to have a sufficient supply to hand. In the occasional slack moment, some of the masons at the quarries left inscriptions. At Haltwhistle Burn, the VI Legion made its mark, but nineteenth-century quarrying removed the lettering. North of Housesteads, at Queen’s Crag, there is an undated inscription which is certainly Roman, and at other places men carved a phallus and testicles – a symbol of good luck and no doubt the subject of ribald comparison.

Up on the Wall the turrets and milecastles were built first. Perhaps this was decided because of the volatile nature of the area, perhaps it would also help with plotting the line of the Wall correctly. Archaeologists have been able to demonstrate this order of building because the turrets and milecastles were left with projections designed to bond with the incoming stretch of connecting Wall. Like the irregular edges of an unfinished jigsaw, the courses of stone stick out on either side.

The most complex and difficult elements of each milecastle were the arched gateways. Teams of specialists quarried and assembled them. To cope with the additional structural stress, larger stones were needed for both the piers and the arches themselves. Some survivors have been found to weigh half a tonne. At the quarry, bursting hammers were used to shape the big blocks. With an axehead shape at one end and a flat sledgehammer face at the other, they could be worked quickly in the hands of an experienced man. Known as voussoirs, the wedge-shaped stones for the arch had to fit precisely, and a procedure called setting-out was used. Once all the faces had been cut roughly flush (finer work could be done at the site) the arch was, in essence, built lying on the ground with each stone set in its place and the necessary symmetry achieved. Good masons can judge these proportions by the eye. Once all was ready, the components of the arch were lifted onto a train of oxcarts. Much too heavy to be picked up by muscle-power, they were loaded by block-and-tackle hoists. These were sophisticated and took several forms. Most common was probably a tripod with a pulley at its apex and strong flax ropes threaded through it. It was very dangerous work, and the velocity of a heavy stone block when a rope shears is startling.

For the highly skilled job of building the piers of an archway, the Romans’ lifting technology was simple but ingenious. Once a large stone had been roughed out at the quarry, a mason took a punch and made a hole on each of the opposite sides. They had to be at exactly the same height and width, precisely opposite, otherwise the stone would rise out of balance and could slip. Then pincers, called nippers on British building sites, were held into each hole and the strain taken by men hauling on the pulley rope. The block was slowly hoisted and let gently down onto the cart bed. A centurion will have roared at anyone who slackened his grip too soon and splintered a cart.

Some examples of a different method of moving stones can still be seen by the observant along the Wall. Using a curved-end chisel, concave triangular holes were made in large blocks, like the basic shape of an isosceles triangle with its apex missing. Masons still use this ancient method and call the cuts Lewis holes. These allowed stones, such as voussoirs, and especially the keystones at the top of arches, to be lifted from their centre. Nippers got in the way of any attempt to slot shaped stones into the semi-circle of an arch. Lewis holes were based on a simple notion. Into the triangular hole a metal grip was inserted. Curved outwards so that its teeth gripped the undercut sides when tension was put on the pulley rope, they were immensely strong and immensely useful. Once a keystone was carefully lowered in, the grip was pulled out easily after the rope had slackened. At that moment the arch was formed and, provided it settled correctly after removal of the semi-circular wooden frame (called a centre) that it was assembled on, the gang moved on to the next milecastle.

TOOLKIT

Over 2,000 years, stone has not changed much and neither have the tools used by masons to cut and shape it. The Romans even had saws, although they ran on muscle-power rather than electricity. The ancient iron toolkit fell into two groups. Picks, walling hammers, axes and adzes were used directly on the stone and were preferred by the masons working at the quarries who roughed out what had just been pulled out of the strata. Experienced men take time to look at a stone before they lift up a hammer. Examining the grain, they select a weak point and can often lay open a big boulder with little more than a tap. Several sorts of chisels – claws, bullnoses, gouges, bent chisels, nickers and punches – were in common use for making holes and were all driven by mallets. The wooden version, sometimes called a mel, was used for planing and other less-skilled jobs. Finer judgement needed a metal-headed mallet with most of its weight at the top. These are known as Italian mallets. Working in Greenlaw, the old county town of Berwickshire, Dave Rumbles is a highly skilled mason able to carve lettering in the Roman manner. He works a blank stone with it canted at a 45-degree angle so that the dust and chippings fall away. Loosely gripping an Italian mallet (too tight and tiredness soon sets in with mistakes to follow) and a fine chisel, he can carve at a steady and rhythmic pace, never taking his eyes off the stone.

Transport logistics were the core of the huge operation begun in AD 122. A four-wheeled ox wagon pulled by as many as ten beasts could haul two tonnes of stone, about seventy normal Wall stones, or four or five pier stones or voussoirs. It has been estimated that an astonishing 30,000 vehicles of one sort or another were used in the three years the Wall probably took to complete. To pull them, around 6,000 oxen were yoked to the heaviest loads. Moving at 3 kilometres an hour, they were slow and ponderous, but steady and not excitable. Horses and mules were much quicker and more nimble-footed, important in difficult country, but had less muscle-power. Nevertheless, it is estimated that the Wall builders used more than 14,000.

The legions already had many pack-animals to carry their tents and other kit. Mules were generally preferred to horses because they were hardier and needed less fodder. On an x-frame pack-saddle, mules could carry as much as 200 kilograms and they could reach the less accessible areas of the Wall – as could men. It is important not to underestimate the amount carried by legionaries and other labourers. To a considerable extent, Hadrian’s Wall is a monument to human sweat.

The Romans appear not to have been kind to their pack-animals, driving them beyond what was sensible, as observed in a decree of Constantine Augustus in AD 357: very many persons by means of knotty and stout clubs force the public post animals . . . to use up whatever strength they have. Feed must have been a substantial problem, especially in the difficult central sector of the Wall, and the legionary quartermasters had to have acquired a prodigious supply of hay after the first cut in late June, at the earliest. That implies a great deal of co-operation with native farmers. Twenty thousand draught animals, working hard, will have exhausted the available grazing very quickly.

The core of the Wall was filled by rubble, clay and soil laid in between two outer skins of masonry. Lime mortar was usually used for bonding, and its production was another major industrial operation. Some kilns for burning limestone already existed. The Vindolanda lists mentioned men sent to the kilns and the need to burn stone. But in 122 many more were urgently required. Always sited close to the limestone quarries, they burned for several days, consuming copious amounts of fuel (in itself another constant need), to produce the powdered lime essential for the masons. Sand was also needed. Mortar was made with one part lime, three parts sand and water. At Fallowfield Fell, near Chesters, a very large sandpit was certainly excavated by the Romans, and there must have been many others.

Mortar was almost certainly mixed off-site, probably near the lime kilns, and then taken up to the Wall wet. Logistically very complex – and dangerous – this sort of awkward transfer had to be managed many times, even in very rough country. The problem was that the powdered lime is so acidic that it burns badly to the touch, lifting off skin in a moment. Loading needed tremendous care, especially when it was windy, and if any spilled during carriage then a pack-animal would have been seriously injured. Sand and lime were carefully turned and folded in volcano-like cones as water was added in stages (just as it is now by those without a mechanical mixer). Then the mixture was loaded into panniers slung on pack-saddles, and the mules were carefully led up to the building site. The mortar beds which can still be seen on the Wall are generally not thickly laid, unlike modern compo. This required real skill to keep the beds level and the outer face flush.

SOLDIERS AND SAILORS

Everybody who was at hand helped to build the Wall. Even the sailors of the British fleet were press-ganged into service. Tile-stamps, inscriptions and other scraps of evidence indicate that they were certainly at work at the eastern end. It may be that they organised and piloted barge traffic up the Tyne, much the best way to carry bulky and heavy building materials. But they will not have been Jolly Jack Tars giving their landlubber comrades a hand. The Classis Britannica contained many men who were as much soldiers as sailors, equipped more like modern marines. And they could have possessed many of the building skills needed to work just as effectively on land.

As the milecastles, turrets and the Wall itself rose ever higher, another element of planning came into play. The finished Wall was probably about four metres in height and, to lay the top courses of stone and fill in the core, scaffolding was needed. Even higher scaffolds were required at the milecastles and turrets. At any given time during the building season (this lasted about thirty-five weeks – working in the winter weather involved disabling difficulties) each legion had approximately five work-gangs taking the Wall to its full elevation, four specialist groups working at the milecastles and two completing the turrets. All of them needed scaffolding. Experts have reckoned that 150,000 metres of straight wooden poles were cut to provide it. This meant a massive sourcing and felling operation in AD 122 and 123.

Putlog holes, where scaffolding was wedged into existing courses of stonework for stability, have been found along the Wall, but it is likely that most frameworks were free-standing. This saved on wood, and the frameworks were easier to move on. Straight-sawn timber is a relatively modern invention, and the irregularities of natural tree growth will have made for some rickety and dangerous structures. Lifting up heavy Wall stones off a scaffold in a high wind was not something many men will have volunteered for.

Progress is likely to have been rapid, especially when the Emperor and his Governor were riding back and forth along the line. It is thought that Hadrian stayed in the north-east for three months, probably lodging at Vindolanda for part of the time. Nevertheless the whole area must have looked like a gigantic building site for a long time. At Highshields Crag, just to the north of Vindolanda, there is evidence to show that, after the foundations were dug, no further work took place for a long enough time to allow soil to blow over the site and cover it. Birdoswald Fort was begun and then abandoned. Scrub grew up inside the half-built walls and had to be burned off before work could resume.

Inscriptions allow some secure dating for the construction of Hadrian’s Wall. In addition to that of the Emperor, the name of Aulus Platorius Nepos is found in several places. And since no other Governor is commemorated anywhere, it seems likely that the project was completed, with some omissions, during his tenure of office. Nepos probably left Britannia in 125, 126 at the latest.

That end date was only made possible because the scale of what he and Hadrian had planned was reduced, almost at the outset. The original Wall foundations were nearly 3 metres in width and probably calculated to carry a very high and impressive superstructure. This was quickly modified and a new width of 2.1 metres specified. Along much of the line the masonry is offset at the bottom courses where the narrower wall sits on top of broad foundations. This was almost certainly done to save time and materials.

THE NAMES OF THE FORTS

If Ravenglass is included, and the forts that were off the line of the Wall, like Vindolanda, are not, what follows is a list of the names given to the Wall forts, with their meanings. Most are Celtic; there are only two in Latin:

Wallsend – Segedunum – The Strong Fort

Newcastle – Pons Aelius – Hadrian’s Bridge

Benwell – Condercum – Viewpoint Fort

Rudchester – Vindobala – White Peak Fort

Halton Chesters – Onnum – Waterfort

Chesters – Cilurnum – Riverpoolfort

Carrawburgh – Brocolitia – Heatherfort

Housesteads – Vercovicium – The Fort of the Good Fighters

Great Chesters – Aesica – Fort of the God

Carvoran – Magnis – Rockfort

Birdoswald – Banna – Promontory Fort

Castlesteads – Camboglanna – Fort on the Curved Bank

Stanwix – Uxelodunum – Highfort

Burgh-by-Sands – Aballava – Appletreefort

Drumburgh – Concavata – Hollow Fort (Latin)

Bowness-on-Solway – Maia – Greatfort

Beckfoot – Bibra – Beaver Fort

Maryport – Alauna – Rockyriverfort

Burrow Walls – Magis – Plainsfort

Moresby – Gabrosentum – Goatspathfort

Ravenglass – Glannoventa – Beachmarketfort

There was much more to Hadrian’s Wall than just a wall. After it was more or less complete, work-gangs began to dig a defensive ditch on the north side. It was deep and wide. At a depth of 2.5 to 3 metres with 33 per cent sloping sides, the ditch presented a real barrier and, because the whole excavated area was between 9 and 12 metres wide, it could not easily be bridged. The spoil was piled up on the north side to make the ditch seem even deeper, and it was probably revetted with the turf removed at the beginning of building work. Roman soldiers were used to ditch-digging; they did it routinely while overnighting on campaign. And it was thought to be a vital skill. The great first century AD general Domitius Corbulo once remarked that the pick was the weapon with which to defeat the enemy. The dolabra, an entrenching tool carried by legionaries, no doubt swung hard and often on the northern flank of Hadrian’s Wall, but the work was unusually difficult. When modern-day volunteers at Vindolanda imitated the work done between 122 and 126, they found ditch-digging much harder than wall-building.

The reason for this was not only to be found in differences in fitness and toughness; it also related to the nature of the ground. At Limestone Corner the hard rock immediately in front of the Wall simply could not be removed to make a clean ditch, and the vain efforts of the soldiers can still be seen. One huge boulder still carries the slots cut into it for wedges. These were no doubt hit hard with bursting hammers under the critical eye of a centurion, but the great boulder would not split and had to be left where it was: a monument to Roman frustration.

Limestone Corner was not unusual in presenting difficulties. Much of the ditch had to be dug out of clay deposits and this was what exhausted the modern-day Vindolanda diggers. When dry the clay was almost rock-hard and when wet it turned to putty, sticking to the blades of shovels and very heavy to lift. When it rained, as it will have done often on the work-gangs at the Wall, the ditch simply filled up, did not drain and had to be bailed before any more excavation could be done. Then there was the problem of spoil removal. In such a deep and wide ditch, triple shovelling was necessary as it got near the required depth. One labourer at the bottom hacked out as much spoil as they could lift, brought it up to shoulder height and deposited it about halfway up the slope. Then a second person picked it up and lifted it to the top. There a third mounded the spoil on the northern bank and smoothed it off. If baskets or stretchers were used, the work might have been easier, but much slower.

More satisfying than all that slog must have been the building of the Wall’s four bridges across the Tyne, the North Tyne, the Irthing and the Eden. For it seems that they were things of beauty. All are long gone now, but traces of the bridge at Chesters, where the Wall crossed the North Tyne, are eloquent. Probably one of the earliest structures to be completed, it was carried on eight hexagonal piers which supported small arches of about 4 metres in width. Cutwaters divided the current of the river and, for symmetry and strength, they seem to have been built both upstream and downstream.

DAM DIFFICULT

Like Trajan’s great bridge over the Danube, all of the four Wall bridges were built with the same basic technique. A cofferdam was driven into the river-bed where each of the stone piers was designed to sit. The most difficult of the four would have been the Pons Aelius. The Gateshead Gorge canalised the sprawling Tyne into a strong current between two areas of high ground. In essence a cofferdam was a watertight box constructed from piles banged into the river-bed. Always larger than any of the bridge’s piers, its purpose was to expose the bed so that it could be built on. Iron-tipped oak piles were first hammered in by a pile-driver lashed to a barge anchored at the site. This was a large and heavy stone hoisted by block and tackle on a frame and then dropped. It must often have hit the pile off-centre. Once an oblong had been completed, engineers set to, driving in an inner skin of more piles and then gradually sealing the joins with clay. Using a waterscrew (invented by Archimedes of Syracuse) turned from barges, or from the banks if the river was not wide, water was lifted out. On the North Tyne and certainly on the Irthing, it would have been possible to bail each cofferdam with large buckets and a hoist. When the river-bed was at last exposed, tar-covered piles were hammered into it and building began on top of them. Pre-cut stone was laid with a rubble core and bound together with pozzolana cement. Roman bridges were famously elegant.

Although it made no sense to carry the Wall across a bridge, it seems that the Romans did exactly that. The platform was the same width as the broad wall foundations – and what does make sense is a walkway from one bank to the other. Perhaps there was a boom chain suspended above the water level to inhibit those who fancied rowing under the Emperor’s Wall. In 207 to 208, the Hadrianic bridge was replaced by a grander three-arched version, which had impressive two-storey abutments at either end.

Cawfields milecastle lies near the western edge of the Whin Sill, not far from where the Wall drops down to Greenhead. In a good state of preservation, it is much visited – and marvelled at by some. Like the entire Wall, it displays a certain bloody-mindedness, even an unbending rigidity of thought. Up on the Sill, Cawfields commands long views to the north, but the gate on that side opens onto a sheer drop. Only 30 metres to the west is a much better and flatter site, one of the nicks in the Whin Sill (the route taken by visitors), and it would have allowed access through both of the milecastle’s gates. But the milecastle had to be a mile, exactly a mile, from the next one, so that was where it had to go.

The first phase of building work showed much more general inflexibility. When the Wall was connected to its turrets and milecastles, it restricted Roman as well as native freedom of movement. The garrison was initially very small, on paper only 1,000 to 1,500 men and, if the Tungrian strength report from Vindolanda is any guide, probably many fewer in practice. All that these men could do was observe and patrol. A much larger concentration of troops was stationed at each of the forts along the Stanegate Road, sometimes 2 or 3 kilometres south of the Wall. If an emergency sparked, they had some distance to go to deal with it. And when detachments of soldiers reached a milecastle, they could only funnel through its gates slowly. It was all very unsatisfactory. Roman armies always instinctively sought the open ground and – perversely – their own Wall was preventing them from reaching it quickly.

ROMAN WALL BLUES

Service in the frontier garrison may have had its bleak moments, but it surely cannot have been as bad as it was painted by W.H. Auden in ‘Roman Wall Blues’:

Over the heather the wet wind blows,

I’ve lice in my tunic and a cold in my nose.

The rain comes pattering out of the sky,

I’m a Wall soldier, I don’t know why.

The mist creeps over the hard grey stone,

my girl’s in Tungria; I sleep alone.

When I’m a veteran with only one eye

I shall do nothing but look at the sky.

(from W.H. Auden. Collected Poems, Faber and Faber)

GARRISONS

Many languages were heard along the Wall in the 120s. Forts were garrisoned from all over the Empire, from Spain, France, Belgium, Holland, Switzerland, Syria, Hungary, Greece, Romania and Germany. Troop dispositions are also informative. They show where the Romans expected trouble and what kind. Inscriptions are handy guides to who was where – but there are still some question marks along the line:

Wallsend	Cohors quingenaria equitata [?]
Benwell	Ala quingenaria [?]
Rudchester	Cohors quingenaria equitata
Halton Chesters	Cohors quingenaria equitata
Chesters	Ala Augusta ob virtutem appellata
Carrawburgh	Cohors quingenaria equitata
Housesteads	Cohors milliaria peditata
Great Chesters	Cohors VI Nerviorum quingenaria peditata [?]
Carvoran	Cohors I Hamiorum quingenaria peditata
Birdoswald	Cohors I Tungrorum milliaria peditata
Castlesteads	Cohors quingenaria peditata
Stanwix	Ala Petriana milliaria
Burgh-by-Sands	Cohors quingenaria equitata [?]
Drumburgh	[?]
Bowness-on-Solway	Cohors milliaria equitata
Beckfoot	Cohors quingenaria peditata
Maryport	Cohors I Hispanorum milliaria equitata
Moresby	Cohors quingenaria equitata [?]

Not long after work began in 122, a decision was taken to move forts up to the line of the Wall. And, so that access to the north, the likely source of most trouble, could be much more immediate, gates were sited beyond the Wall line. At Chesters, where a cavalry regiment was based, three of the four fort gates allowed troopers to ride directly out, without having to pass through the Wall. The original plan was a monumental mistake, producing the obvious effect of penning in the garrisons, but the decision to change appears to have been taken quickly.

One of the results of this early blunder and its remedy is spectacular. Housesteads is probably the most beautifully situated of all the Wall’s forts, and because of its location in the central sector, remote until the road-building of the mid eighteenth century, it has survived remarkably. The Border Reivers also helped preserve the old Roman fort during the fourteenth to sixteenth centuries. Using the ready-made cut stone which lay everywhere to hand, they built a longhouse and at least two bastle houses on the site. These were heavily defended, thick-walled structures designed to frustrate rather than repel attackers. The remains of one can clearly be seen outside the southern gate at Housesteads. A ground floor, or pend, was built without windows and it was intended to keep safe a farmer’s beasts during a raid. They were driven in, packed tight, bellowing with fear, and the stout door barred shut from the inside. In the roof of the pend there was sometimes a trapdoor to the first floor. This was more easily and more usually accessed by an outside stair or ladder, and out of its tiny windows defenders hurled abuse and anything else they could find at the horsemen circling below. Bastle houses were primitive but effective, a poor man’s peel tower. To the Armstrongs of Gandy’s Knowe, a branch of one of the most notorious of all the reiving surnames, the irony of rebuilding a much lesser defensive structure on the fringe of one so sophisticated might not have been lost. They were ferocious, ruthless thieves, but not usually ignorant. The perimeter of the fort itself was probably barricaded into use as a cattle corral, often somebody else’s cattle, as the Armstrongs rode the moonlight and raided in the Tyne Valley and to the north.

At the time of the Border Reivers, William Camden compiled his great antiquarian history Britannia (published in 1586). He rode along the line of the Wall, a countryside he thought lean, hungry and waste. Hadrian’s Wall amazed him: Verily I have seen the tract of it over the high pitches and steep descent of hills, wonderfully rising and falling. Camden knew of Housesteads Fort, understood that there was much to see but, to his intense frustration, he could not get near it: I could not safely take the full survey of it for the rank robbers thereabouts. Bandit country.

Considering how the climate had hardened the carcases of Borderers, he observed how the descendants of the horsemen in the Vindolanda letters lived:

In the wastes . . . you may see it as were the ancient nomads, a martial kind of men who, from the month of April into August, lie out scattering and summering with their cattle, in little cottages here and there, which they call shiels and shielings.

It may have been the sort of conservation unknown to English Heritage, but the reivers’ occupation of the old fort and their lack of any architectural ambition kept the circuit of walls intact and the steading full of houses, as they knew it, relatively undisturbed. James VI and I of Great Britain eventually began the process of bringing Housesteads to the notice of a wider world. After his brutal but effective police action of 1603 to 1610 against the Border Reivers, the by-then-redundant frontier began to settle down. The bandits retreated further and further into the hills and, by the end of the seveneenth century, the first visitors were coming.

In 1725, the antiquarian William Stukely made an early drawing of the site. It shows a farmstead built inside the walls, perhaps the successor of the longhouse, and a scattering of altars and inscribed stones over the slopes below the south gate. By 1751, the Military Road had been laid out, much of its length bottomed with Wall stone, and it passed only 500 metres south of the fort. Amongst those who left a record of what they saw was William Hutton, who arrived in 1802 full of enthusiasm:

I retreated next morning over a moss to my favourite pursuit, which brought me to Housesteads, the grandest station on the whole line. In some stations the antiquary feeds upon shells, but here upon kernels. Here lies the ancient splendour in bold characters.

Archaeologists have pieced together the story of Housesteads. It appears always to have been something of a showpiece. As one of the forts integrated into the line from the outset, it occupied a central place in the central sector, perched high and visible on the cliffs of the Whin Sill. It is impressive even when only glimpsed out of a car window from the road below. The site is sloping, although not as extreme as at Cawfields, and it seems to present Housesteads to the south for all to gaze upon it. The outlook on every side is absolutely commanding.

The forts and milecastles of the central sector struggled to find reliable and substantial sources of water, despite the rain. Below Housesteads the Knag Burn runs through a nick lying to the east. It penetrates the Wall through a culvert built at its foot. Flanked by small guardhouses there is also a rare gateway (cut in the early fourth century) directly through the Wall. Inside the fort stand the remains of cisterns probably used to collect rainwater off the roofs. Other garrisons were served by simple aqueducts, occasionally from the north, clay lined and cleverly sited.

THE WORK OF GIANTS

By the eighth century the Roman Wall and the towns of the south of Britain had ceased to function as large communities. Some towns were on sites continuously inhabited by smaller populations whose own buildings eroded and even erased what had stood there before. A ready re-use for quarried, squared-off and handily sized building stone could always be found. Roman altars were incorporated into the structure of Jedburgh Abbey, and in Carlisle Cathedral the facades are speckled with stone robbed from Hadrian’s Wall and nearby Stanwix Fort.

The English and the native British were not contemptuous of Roman architecture. They simply had little use for it as it stood because their social structures were different, smaller in eighth century betrays a sense of awe, and even regret at what time and disuse had done to the old cities:

The Ruin

Splendid this rampart is, though fate destroyed it,

The city buildings fell apart, the works

Of giants crumble. Tumbled are the towers,

Ruined are the roofs, and broken the barred gate,

Frost in the plaster, all the ceilings gape,

Torn and collapsed and eaten up by age.

Despite its elevated position, Housesteads seems to shelter in the lee of the Whin Sill, and it is less buffeted by the winter winds than other forts on the Wall. The free-draining farmland around the fort has historically been cultivated rather than grazed. Aerial photographs show not only the contours of Roman terracing but also the remains of small medieval rig-systems. When John Clayton of Chesters bought the farm so that he could preserve and investigate the fort, he faced stiff competition from bidders anxious to keep tilling such good land.

The Romans called it Vercovicium, a place-name with a shadowy modern survival in nearby Barcombe Hill, the site of an ancient quarry above Vindolanda. Derived from Old Welsh roots, Housesteads’ original name means something clumsier in English, the Fort of the Good Fighters. The earliest garrison was the I Cohort of Tungrians, posted to the fort from Vindolanda around AD 122/123. After their exploits in the front line at Mons Graupius under Agricola, they were certainly reckoned to be good fighters, and the name might simply reflect that reputation. Equally possible is a transfer. Slight traces of a native hillfort have been seen on Barcombe Hill and perhaps the old name travelled 2 kilometres to the north-east to settle on the new fort and its men. That would also explain the similarity between the two place-names. The modern name is certainly a reference to the remains of the Roman buildings: the houses inside what looked like a walled inbye enclosure, a steading.

The layout of Vercovicium is classic. Perhaps because of the very long tradition of success, rarely suffering more than temporary reverses over 500 years or so, the Roman army was very conservative in its thinking. For example, fort design was generally standard throughout the Empire, no matter how different the climate, the geography or the political context and its requirements. From Inchtuthil in the Perthshire woods to the vastness of the Persian desert, the same playing-card shape and internal design was built again and again.

Any innovations were minor. Unlike African or Asian forts, Housesteads had a coal-house. Near the east gate, it was filled with what local people used to call craw coal, the coal that could be quarried and carried away from coal-heughs such as Shawhead Drift, south-west of Vindolanda. There was almost a tonne of it, and at Risingham and Corbridge evidence of more coal storage has been found.

Up until the early nineteenth century two collieries at the foot of Barcombe Hill were still being worked, and over at Ramshawfield coal mining on some scale continued for longer. The ready availability of coal in a landscape with few sources of wood was a boon to the soldiers who shivered in the winter at Housesteads. The bath house, the hypocaust heating system under the commanding officer’s house and wherever the garrison would light a fire or a brazier to keep warm – all of these will have benefited from the supply of Tyne Valley coal.

The Romans’ dogged conservatism is surprising given their unhesitating ruthlessness when making decisions, such as the movement of the forts up to the line of the Wall. At Housesteads and elsewhere, places where permanent garrisons were settling down, a more obviously defensive military architecture might have developed. But there is no evidence for emplacements for artillery, for example, or for projecting gatehouses and towers (like those of medieval castles) which would allow enfilading fire to rake along the length of the fort walls, severely hampering any attempt to break through.

However all that may be, a certain aspect of domestic life is splendidly illustrated amongst the ruined buildings of Housesteads. Tucked into the south-east corner of the fort, the latrines are well preserved. They are surrounded by cisterns, festooned with drains – and communal! No cubicles. This has caught the public’s attention like few other details of Roman army life. Communal, sit-down toilets? The very idea!

Guides to the site delight in explaining that there was seating for about twenty men, a surprisingly small capacity for a garrison of 800 soldiers who regularly consumed a good deal of roughage. A well-drawn coloured reconstruction of men using the latrines, their undergarments around their ankles, adorns the guidebook. It must be one of the more unexpected illustrations in British historiography. But worse is to come. In their hands the soldiers are shown holding sticks with something attached to the end. Much worse. In a small concession to seemliness, the soldier furthest away appears to have used the stick and he leans forward towards a water-filled drain which runs around the middle of the floor.

What, exactly, is going on? The caption for the illustration is vague. It should read: How Roman soldiers wiped their backsides. At the end of each stick, it was said that a small sponge was attached. Having completed the first part of their business, the soldiers were believed to wipe their backsides by scrubbing vigorously back and forth – and then rinsing the sponge in the drain on the floor.

At least two thoughts occur. Did each soldier have his own stick and sponge? And how securely were the sponges attached? Very, it is to be hoped.

The problem with this bizarre scene is the sponges. Well, one of the problems. Although their use in latrines is attested by two Roman writers, Martial and Seneca, it is often forgotten that they were describing an aspect of everyday life around the shores of the Mediterranean, where sponges grow naturally and are harvested by divers. There are none in the North Sea, or the Solway Firth. And it is highly unlikely that sponges were imported in bulk to be used for such an – everyday – purpose by ordinary soldiers. Amidst all the grunting, farting and exhaling, to say nothing of any comments, moss was probably used. And not rinsed.

Perhaps it was sold to soldiers in some quantity in the streets and shops of the vicus, the civil settlement which huddled around the walls of Housesteads. Despite the relatively remote location, the settlement was large with buildings to the south and east of the fort. Two discoveries are particularly vivid.

In the backlands of one of the houses, which sat gable-end on to the main street, a small shrine was found. It housed a well-preserved piece of relief sculpture. Three small figures stand in a line and are more or less identical. Wearing the byrrus Britannicus, a hooded cloak which fastened at the front and was a well-known export to European markets, they stand passive and enigmatic. They represent the Genii Cucullati, the Hooded Gods, and were given offerings as the protectors of the household.

On Chapel Hill, a ridge south of the vicus, altars have been found dedicated to gods native to Frisia in western Holland. They were raised up by units of Germanic warriors hired as irregular troops. The Cuneus Frisiorum, who may have been cavalry troopers, worshipped Mars Thincsus and the Aliagasai. A dedication mentions Numerus Hnaudufridi, another band of soldiers from beyond the Rhine. The translation is ‘Notfried’s Own’ and they sound like a prince and his warband. Whatever their origins and their beliefs, their presence in the north was not decorative. Trouble still broke out on the frontier.

The dates attached to these inscriptions are later, before AD 235. When the Wall was still young, a century before, Rome could garrison the frontier without recourse to barbarian mercenaries. The most prestigious regiment stationed on Hadrian’s Wall also occupied by far the largest fort. Virtually no trace can now be seen of it. At Stanwix, on the eastern banks of the Eden, opposite the centre of Carlisle, a fort of nearly 10 acres was built to house the Ala Petriana. This was the only milliary cavalry cohort in Britannia, a force of around 800 troopers and commanded by a Prefect, the most senior officer anywhere on the Wall.

The suburbs of modern Carlisle have obliterated any remains left by the stone robbers, and the only hint of Uxellodunum, the High Fort, is a pathetic plaque mounted high on a house wall on a street corner and some coloured bricks laid along the line of the fort’s south wall in a car park behind the Cumbria Park Hotel. Some fragments of the original masonry could once be seen under a window of thick bottle glass laid into the tarmac, but the elements have clouded it so much that nothing can now be made out. Compared with the ruined glories of Housesteads, Stanwix’ fate is sad, if inevitable. Perhaps its obliteration is the reason why its pivotal role is sometimes ignored.

RENDER UNTO CEASAR?

The Romans plastered everything, generalised one eminent historian. He was writing about the buildings inside their forts at the time but his assertion is a reminder that the white marble, the naked sandstone and the monochrome appearance of much of antiquity and its remains were not what the Romans – or the Greeks – saw. They had their sculpture painted, the friezes on their buildings were often highly coloured and the interior walls of quite modest structures sometimes had frescos. Floors were tiled with coloured mosaics where they could be afforded. Plaster, or render, on the outside had a practical function in that it kept out the worst of the weather as well as brightening up the environment by making the most of the light. Plastering, like painting kerbstones in a modern army camp, was a good way to keep soldiers busy and out of mischief. Hadrian’s Wall was probably rendered, at least parts of it. At Castle Nick, Denton and elsewhere traces of render or whitewash have been found. One Wall expert believes that not only was the face of the Wall plastered white, but it was also scribed with false joints. What an arresting sight! A white wall snaking through the green and brown countryside, stark, unmissable, dominating.

The crack troopers of the Ala Petriana were stationed near Carlisle because it was seen as the hinge of the Wall, not merely its western terminal. The Cumbrian Sea-Wall stretched far to the south and all indications point to the Roman strategists’ view that the west was the critical sector of the frontier region. In contemporary Celtic society a cavalry force of 800 well-armed and -horsed men represented a small army, and when the Petriana clattered out of the gates at Stanwix, their standards glinting and fluttering, it will have been an impressive sight. What is unclear is the wider role of their Prefect. As the most senior officer on the Wall, did he have jurisdiction over others? Or did orders come directly from York?

Also buried, but less mysterious than Stanwix, the fort at Maryport was a vital link in the military chain which held the north. Sited up on the Sea Brows, high above the Cumbrian shore, the fort has not been smothered in housing or any other modern development. It lies quiet under a grass field, used for pasture and not ploughed, its characteristic shape visible on the ground and especially clearly from the air. Beside it stands a fascinating building, now the Senhouse Roman Museum. Housed in a silent battery used for training gunners in the Royal Naval Volunteer Reserve (guns which did not fire were a condition of the gift of the land by Mrs Senhouse), the collection of objects is startling, better than any on the Wall, with the possible exception of Vindolanda.

In 1870, in the corner of a farm field to the north of the fort site, a cache of seventeen Roman altars was found. Deliberately buried some time in the late second or early third centuries, they tell a clear and continuous story. The altars are in such excellent condition that, not only is the lettering crisp and easily legible, traces of paint can still be seen on them. It was the greatest single find of Roman inscriptions ever made in Britain. In addition to these, there is a great deal of sculpture, some of it native, some of it unique. And yet the fort has never been excavated, nor has the vicus, and the Senhouse Museum is far too little visited. Maryport is not generally seen as an integral part of the story of Hadrian’s Wall. But of course it is and, when the archaeologists finally arrive, treasures are likely to come out of the ground.

Maryport was a sea-fort. From the Sea Brows almost all of the Galloway coast can be seen and, even on dull days, the grey hump of the Isle of Man darkens the western horizon. The fort’s strategic role was primarily defensive from the outset as its lookouts watched for trouble sailing across the Solway Firth. It also turned inland, acting as key link in the encirclement of the Lake District, probably as much a source of opposition to Rome as the North Pennines.

MASTERS AND COMMANDERS

The surviving Roman records of Britannia appear detailed when compared with historical periods at either end of the life of the province. But they are in fact patchy: the sequence of governors, for example, is far from complete. One of the joys of the finds at Maryport is the quality of the inscribed records. Here is a complete list of the first six commanders of the I Cohort of Spaniards based at the fort.

Marcus Maenius Agrippa – Tribune

from Camerinum in Italy

Caius Caballius Priscus – Tribune

from Verona [?] in Italy

Marcus Censorius Cornelianus – Prefect

from Nimes in France

Lucius Cammius Maximus – Prefect

from Solva in Austria

Lucius Anstistius Lupus Verianus – Prefect

from Le Kef in Tunisia

Helstrius Novellus – Prefect

from Italy [?]

The fort prompted the development of a large vicus which straddled the road leading out of the north gate. A detailed geophysical survey has discovered a long ribbon of buildings, stretching for more than 350 metres and including some very large structures. Some are thought to be industrial or used for storage. Associated coal and iron debris has been found. In the lee of the fort, the smoke and fumes from all that work will have blown away from the garrison. The commanding officer in post in 122, Maenius Agrippa, was probably an aristocratic figure with even more high-tone tastes than the Batavians at Vindolanda. Coal smoke is not something he will have wished to sniff as he was entertaining guests.

Some of Maryport’s most fascinating sculpture was found in the area of the vicus. A Roman altar was cut down and reshaped into a large phallus. On one side a serpent wearing a torc slithers up to the head, while on the other a human head with a strange necklace of fish is carved. In the absence of anything comparative or any helpful texts, the iconography is impossible to read, but the use of the phallus as a defence against evil may be what was intended. The sculpture was found amongst several cremation burials and perhaps it was erected to protect them. There is also a wonderfully well-preserved Epona, a Celtic goddess of fertility, who takes the form of a mare. Her name is probably not the derivation of the word pony. She is known from many European sites but the Maryport piece is the only complete representation yet found in Britain. Also recognisable is Cernunnos, a horned god often linked with the Brigantes.

Worship of these deities was not confined to the Celtic population. The Romans often incorporated local cults into their pantheon and sometimes created hybrids, twinning classical gods with Celtic counterparts. Not far away, across the Solway, near Gretna, there was a famous shrine to Apollo Maponus. The name of the latter, a native deity, is still heard in the Dumfriesshire place-name of Lochmaben.

SKY-GODS

The Roman pantheon survives overhead. The planets of our solar system are mostly named after gods venerated in Britannia and some have aspects of their ancient attributes:

Mercury is the smallest planet and is named after the wing-footed messenger of the gods.

Venus is seen as the Earth’s sister planet and is named after the goddess of love and beauty.

Mars is the red, or angry, planet and is named after the god of war, the soldiers’ god.

Jupiter is the largest planet in the solar system and is named after the king of the gods.

Saturn is almost as big and is named after the father of Jupiter and the god of agriculture.

Uranus was a late discovery (1787) and is named after a Graeco-Roman god of the sky.

Neptune has the strongest winds and is named after the god of rivers and, later, of the sea.

After the forts had been brought up to the line of the Wall and the ditch in front of it completed, Roman planners turned their minds to the problem of the south. If part of Hadrian’s intention had been to force apart a tight network of allies, then his dispositions at Stanwix, down the Cumbrian coast, at the western outpost forts, and the general focus of the Wall system appeared to cope with the threats rumbling in the north. But, to its rear, the barrier must have felt exposed. Another momentous decision was taken.

In order to create a clearly defined military zone, a wide ditch would be dug in the area immediately to the south of the Wall. The Romans did not call the ditch the Vallum. This mistaken label was first attached by the normally meticulous Bede of Jarrow, the great eighth-century historian of the English, a man who knew the Wall well. Vallum may be the accepted name for the vast ditch system dug behind Hadrian’s Wall but, in the second century AD, it was the Latin word for a palisaded rampart. And it is the derivation of the English ‘wall’. Fossa is a ditch or trench, not Vallum. But the original misleading label has stuck.

It was a remarkable feat of military engineering. So that it could present a really formidable barrier, the whole entrenchment measures 37 metres across: impossible to bridge and very difficult to traverse on foot. And it ran behind the entire length of the Land-Wall, from the Tyne to Bowness-on-Solway without interruption. In many places where there is no trace of any masonry and even the cleft of the Wall ditch has gone, the Vallum is still visible. Arrow-straight, much larger than any other element of the frontier works, it is perhaps the most enduring mark made by Hadrian on the British landscape.

The line was chosen where the ground allowed a deep ditch to be dug, and occasionally that pulled the Vallum well to the south of the Wall. Between 5.4 to 5.9 metres across the top, it was excavated to a depth of 2.6 to 2.9 metres and, unlike the Wall ditch in the north, which had only a narrow bottom (known as an ankle-breaker), the Vallum was 2.1 metres across at its foot. This made for much steeper sides, a 60 per cent slope – very difficult to climb in and out of. All of the spoil was probably hacked out by gangs of auxiliary troops, rather than legionaries, and they piled it up in mounds on either side. These were separated from the edges of the steep-sided ditch by broad berms 9 to 12 metres in width. This much clearance was needed to avoid the danger of infill or subsidence. The spoil heaps were then revetted with turf to stop them crumbling. Anyone who approached the Vallum had to climb these banks and had no choice but to make themselves very visible to soldiers on the Wall.

The Vallum was a very substantial obstacle. Over time the Romans allowed it to become even more of an obstacle as water pooled in the bottom of the ditch (there is no evidence of any drains) and were probably happy to see thorns and other scrub grow – except on the berms and mounds.

The most telling effects of the creation of the Vallum were twofold. Most dramatically, it cut the number of crossing-points through the Wall from seventy-nine gateways through the milecastles to fourteen. These arched gateways, so painstakingly constructed only a short time before, became secondary as the focus of movement shifted to a series of causeways thrown across the Vallum near the forts. One of the most surprising and quirky sites along the Wall is a Vallum crossing just to the south of Benwell Fort. Half covered by a reservoir and a main road, and drowned by the redbrick suburbs of west Newcastle, the fort has completely disappeared. But at the bottom of a terrace of semi-detached houses behind the local Social Security offices, the causeway over the Vallum has miraculously survived. Protected by iron railings and stout gates, the ditch has been restored to something approaching its original depth. But most impressive is the causeway itself. The bottom stones of one of the piers of its massive gatehouse perch precariously right on the edge. They supported a free-standing massive structure, like a small triumphal arch, which had stout, iron-studded doors and was manned day and night. Behind it rise the cobblestones of several layers of the roadway which led to the vanished fort.

The overall effect of the Vallum was to create a military zone which ribboned through the countryside, with such tightly and clearly controlled access that it was the ultimate authorised persons only area. It protected the forts from damage and theft (a good deal of gear, such as carts, and stock will have been left outside the walls of milecastles, turrets and even forts) and allowed much more secure and unhampered movement between the garrisons. Supply traffic must have been constant and, before the creation of the military zone, quartermasters will have had to secure their valuable goods every night, perhaps behind palisades, certainly under close guard. By excluding the native peoples so completely, this irritant diminished.

What undoubtedly increased was frustration on the far side of the Vallum. One example will suffice. For millennia the stock farmers of the Tyne and Irthing Valleys had driven their flocks and herds up the hill trails on the ancient journey of transhumance. Summering out on the fells, their beasts grazing the upland pasture, shepherds and farm laddies had lived in shielings described by William Camden. North of the Tyne, the Wall and its Vallum stopped this time-honoured traffic in its tracks. The new zone had simply removed a wide swathe of land from the native agricultural economy, and made it extremely difficult for farmers to use land to the north of it. It is unthinkable that commanding officers would allow herdsmen to drive their beasts across one of the fourteen causeways, along the streets of a fort, through the northern gates and out onto the commons beyond. At milecastles it might have been easier, but since the Wall ditch shows no sign of having had solid crossings, it is hard to see how cattle and sheep might have been safely prodded and whacked across narrow wooden bridges – if there were any.

VALLUMS GALORE

Ditches and dykes were used to demarcate bits of Britain long before and long after the construction of Hadrian’s Wall. Offa’s Dyke (or Dike) is the most famous. King of Mercia from 757 to 796, Offa fought fierce campaigns against the invading Welsh princes and, to show where their territory ended and his began, he had a vallum magnum built from sea to sea. Stretching 220 kilometres, much longer than Hadrian’s and Antonine’s walls put together, it runs from the Severn to the Dee estuary. The largest earthwork ever dug in Europe, at any time, it is an extraordinary – and far from well understood – monument. Offa called on his people to give military service and work in gangs to dig a 2-metre ditch and build up a 7-metre rampart to the east of it. The whole layout is more than 20 metres across. Its sheer mass and height made it very difficult for the horse-riding Welsh warbands to cross, and near impossible for them to raid cattle and drive them back. In the north, Wat’s Dyke completed the line and its terminal is at Basingwerk on the Dee. Unlike the Roman walls, Offa’s Dyke appears to have contained the Welsh, and the modern border runs close by.

As the great Wall project neared completion, the standard of mason work began to deteriorate markedly. It is as though there was a scramble to the finish line, perhaps a mad dash to complete before the departure of the Governor of the province, Aulus Platorius Nepos. At Housesteads, Birdoswald and Chesters, the piers and voussoirs at several gateways show poor craftsmanship. There is evidence of more than just hints of great haste. At Housesteads the lower piers of one gate are tidily enough finished but the capstones, where the arch springs, have scarcely been worked at all. In several milecastles some of the massive stonework is not so much badly finished as abandoned in situ. It seems that as long as the arches and gateways stayed up, at least for the moment, that was fine.

Perhaps fighting broke out during the final phases of construction and, in the way that things often happen, the unfinished stonework was simply left in place. With the eventual departure of the legions, most of the really skilled masons in their ranks went too. It may well be that there were too few with sufficient skill to tidy up stonework that had been thrown up.

Once the forts had been more or less completed (the work at Birdoswald seems to have been disrupted more than once), their garrisons moved in. The most common type of auxiliary cohort on the Wall was a mix between infantry and cavalry, the cohors quingenaria equitata. Many of these soldiers had marched north from their postings on the Welsh border and in the South Pennines, where it was hoped that the situation was stable. Not including the Sea-Wall, that added up to a total garrison of 9,090 auxiliaries, on paper.

In the central sector, at Housesteads, Great Chesters, Carvoran and Birdoswald, all units were peditata, or foot soldiers. Why pure infantry should have been preferred in the countryside so well understood and beloved later by the horse-riding Border Reivers is not clear. If their ancestors had similar equestrian skills – and it is highly likely that their very many cavalry did – detachments of foot soldiers chasing them up hill and down dale sounds like a laughable tactical mismatch. The traditional raiding methods of hit and run will have worked every time. But the Romans were usually very astute – and there may well have been a clear but now lost justification for those dispositions up on the Whin Sill.

In any case, the reach of cavalry was long and rapid. Reckoning that the link between the infantry of Housesteads and the troopers stationed at Chesters down in the valley of the North Tyne was not close enough, another fort was built between them. At Carrawburgh, surely the bleakest location on the Wall, Brocolitia was established. Brocolitia means the Heathery Fort, and it still is. Before the masons could begin, the Vallum had to be filled in, and that fact dates the new fort as an afterthought. A cohort of mixed cavalry and infantry appears to have been inserted in the gap.

On the northern bank of the Tyne the original eastern terminal of the Wall may have been the fort at Newcastle, at the Pons Aelius. Under the medieval castle, the site of the Newcastle (and a later Roman fort whose traces have been found under the arches of the Victorian railway viaduct which slashes right through the precinct, separating the keep from the main gatehouse), a modern hotel, the Moot Hall and a pub, the remains of the Hadrianic fort may lie undisturbed for a long time. Little trace of the Wall has been found immediately to the east, but what there is suggests another afterthought. No broad Wall foundations, only narrow Wall building, probably means that the stretch leading out to Wallsend was begun after the drive westwards. It may be that the broad barrier of the lower Tyne was more porous than anticipated.

The course of the Tyne was much wider and meandering in 122. The river had not been dredged and was not as deep as it is now. Perhaps the native peoples were simply ignoring all that impressive stonework upriver at Newcastle, going around the eastern end of the Wall by taking to their boats or wading across at low tide at fording places.

The site chosen for the fort at Wallsend sits on the crown of a long and lazy bend in the Tyne with clear views both upriver and down. The Wall arrived at its north-western corner and then, adopting the fort’s ramparts, turned through 90 degrees to emerge as what is known as the Branch Wall. It ran south down to the riverbank and continued out into the current. And there Hadrian’s Wall terminated (or began) – but not before a final imperial flourish.

Across the Tyne stands the ancient church of St Paul at Jarrow. Once it was part of a monastery (twinned with Monkwearmouth) which was home to Bede, one of Britain’s earliest and very greatest historians. When the first Abbot, Benedict Biscop, founded these communities, he was anxious to build in the Roman manner and brought masons from Europe with the necessary skills. Plenty of authentic Roman building materials were lying about near Jarrow, only a boat trip across the Tyne. Ready-cut stones from both the fort at Wallsend and the Wall itself can still be seen at St Paul’s, especially amongst the monastic ruins to the east of the church. In 1783 stonemasons were beginning to restore the fabric of the old church when they came upon two fragments of a Latin inscription. The lettering was very large and the panel it had been cut into must have been huge. Here is a translation:

Son of all deified emperors, the Emperor Caesar Trajan Hadrian Augustus, after the necessity of keeping the Empire within its limits had been laid on him by divine command . . . once the barbarians had been scattered and the province of Britannia recovered, added a frontier between either shore of Ocean for 80 miles. The army of the province built the wall under the direction of Aulus Platorius Nepos, Pro-Praetorian Legate of Augustus.

Several historians have convincingly conjectured that these words were carved on the plinth of a massive statue, one which stood somewhere near Jarrow. Hundreds of kilometres away, at the southern end of Britannia, is a clue as to what the statue might have been. At Richborough, the massive triumphal arch erected during the reign of the Emperor Domitian was an imposing memorial to earlier conquests and an impressive welcome to the province. It seems that Hadrian planned an equally impressive goodbye. He was later to make a boastful but probably accurate claim: I have achieved more by peace than others [have] by war. More than the hated Domitian! At the end of the Branch Wall, where it had been built out into the midstream of the River Tyne, there probably stood a huge statue of Hadrian, the Emperor who had commanded a Wall to be built across Britain. For all who saw it, and especially those who sailed under its stern gaze, it must have been an awesome sight.

As with every sensible and successful Emperor of Rome, Hadrian’s focus was on the army. Without its loyalty and discipline, he could not survive and nothing could be achieved. If the Wall was principally intended as a symbol, its message to the native peoples was incidental. Whatever they thought, the legions would crush them. Much more important was the army, its morale, the support of its senior officers. By erecting a colossus on the northern frontier of Britannia, subduing the landscape to his will, defeating and containing the barbarians, and having his policy inscribed beneath for all to read, Hadrian was reinforcing the radical strategic changes which he made during his reign. The age of an empire without limit was over, there would be no more conquest for the sake of it. But it would be a triumphant limit. The Roman army would now assume a defensive role, become the keepers of the peace rather than the makers of war.

COLOSSI

Massive statues of gods and emperors were not uncommon in the Roman Empire, but the most famous was probably the Colossus of Rhodes. A huge bronze representation of the sun-god Helios was created by Chares around 280 BC and was said to stand more than 30 metres high. Set up at the harbour, the great statue may have stood astride its entrance with ships passing between its legs. But more likely it did not. The Colossus was toppled in AD 224 in an earthquake, but Hadrian would certainly have seen it on his travels. The other famous Colossus of the ancient world has also gone, but its name survives. A huge statue of the Emperor Nero stood near the Via Sacra in Rome but, after Nero’s suicide, it too crashed to the ground. When Vespasian began to build a great amphitheatre on the site, the name stuck and it is still called the Colosseum.