7

Abandonment, Invasion and Desertion

So that it would be instantaneous, or at worst very rapid, Hadrian drew a coloured line on the left-hand side of his chest, exactly in the place his doctor had shown him. Then he commanded Mastor, his huntsman, to stab him, to drive the blade deep and hard between his ribs, straight into his heart and put him out of what seemed like an eternity of misery. With promises of great riches, of piles of gold and silver coins, and threats of dire consequences if he did not do what his Emperor instructed, Hadrian had brought his huntsman armed to his bedside. He begged for death. There would be no retribution, no prosecution, Mastor would be immune – if only he would do this last, desperate service for Hadrian. He bitterly lamented the state to which his illness and his helplessness had reduced him – notwithstanding that it was still within his power, even when on the brink of death, to destroy anyone else.

No doubt terrified, Mastor lost what nerve he had and refused to kill the Emperor. He was not an aristocrat, not a Roman, but a tribesman from the Iazyges, from the Danube, a nobody who would probably be killed in the first frantic hours after Hadrian’s death. Wise to refuse his master’s pleadings, Mastor reported what had taken place.

The pain must have been very hard to bear. It seems that by the early summer of AD 138 the Emperor was passing into the terminal stages of heart disease, suffering from increasingly severe haemorrhages, his arteries hardening and preventing proper circulation. It had begun as nose-bleeds and was quickly developing into something very much worse. Hadrian also had what was known as dropsy, a painful buildup of fluid in his body.

After the failed attempt to persuade and coerce Mastor, Hadrian’s adopted successor came to see the dying Emperor. Here is the passage from the Historia Augusta:

Antoninus and the Prefects went in to see Hadrian and begged him to endure the necessity of the disease with equanimity. Antoninus told him that he would be a parricide if, after being adopted, he allowed him to be killed. Hadrian was angry and ordered the person who had informed them to be killed – he was, however, saved by Antoninus. He at once wrote his will. But he did not lay aside the business of state . . . He did in fact try to kill himself again; when the dagger was taken from him he became more violent. He asked his doctor for poison, who killed himself rather than comply.

Hadrian was sixty-two. Worn out after a reign of more than twenty years, eleven of them spent travelling around his vast domain, he had become increasingly unwell. And just as insidious, he began to attack, dismiss and even dispose of those who had been friends and companions, who had given solace and support through the loneliness of absolute power. Aulus Platorius Nepos, former Governor of Britainnia and builder of the Wall, had been very close and at one time a candidate for the succession. So confident was he in Hadrian’s favour that he dared to send the Emperor away when he had had the decency to come to Nepos’ house and enquire after his health. No one, anywhere, was ever too sick to see an Emperor – but there appear to have been no immediate repercussions. Led on by suspicions and listening to whatever was whispered about his friends, Hadrian cast Nepos out from his inner circle, calling his old comrade an enemy.

It could have been worse. Two senators, Polyaenus and Neratius Marcellus, were compelled to suicide and two men of equestrian rank who had been imperial staff members and advisors were dismissed; Valerius Eudaemon was reduced to poverty and Avidius Heliodorus’ services were no longer required.

Hadrian continued to be obsessed by the death of Antinous: the boy was deified and his worship formalised into a cult. Twinned with the goddess of the hunt, Diana, Antinous was to have a temple dedicated to him in a town near Rome. And in the gardens of the imperial palace on the Palatine Hill, there was a further public commemoration, a series of sculptured reliefs of Hadrian and his young lover hunting. In his temple Antinous was shown as the god Silvanus, the patron of huntsmen. Perhaps that connection supplied the tortured reason for Hadrian’s selection of Mastor as the agent of his own death. In any case it seems like a rather grandiose celebration of nothing more than happy times, of the Emperor doing what he loved best with the person he loved most alongside him.

After his return to Rome in the mid 130s, Hadrian indulged his other passion: architecture. Building work on several imperial projects was coming to an end. In 121 the Emperor had sanctioned a temple to Venus and Rome, as usual taking a meddling hand in the design himself. There was a basic flaw. Inside, the massive colossi of the goddess and Rome were seated on thrones of some sort, but the roof over their heads was much too low. Trajan’s architect, Apollodorus, made the great mistake of pointing this out: for now, if the goddesses want to stand up and go out, they will be unable to do so. Hadrian was embarrassed at something so basic – and enraged. One historian recorded Apollodorus’ murder and, given the Emperor’s unhesitating ruthlessness with others, it may be an accurate report.

Perhaps the grandest, most emblematic architectural project was Hadrian’s own villa. Built in the countryside around Rome, near the town of Tivoli, from which it took its name, it was both vast in scale and in conception. More than a kilometre from one sprawling end to another, Tivoli compassed the Empire, with Greek, Egyptian and Italian motifs. There was an amphitheatre, a canal, sculpture galleries, baths, an outlook tower and at its centre, at the Teatro Marittimo, a private retreat for the Emperor. Surrounded by a moat, this selfconscious symphony of curves and innovative design was supposedly where Hadrian could spend time in contemplation.

The impression is of an epic emptiness, a show of power and patronage for its own sake, a monument rather than a pleasure palace or an idiosyncratic expression of character. As Hadrian rattled around the cavernous halls of Tivoli, becoming more and more unwell, growing increasingly embittered, loathed and lonely, he trailed a gathering cloud of misery and frustration behind him.

Little had worked out as the Emperor had planned it. His first choice as heir, Lucius Verus, had tuberculosis. He frequently coughed up blood, but Hadrian had had the Senate confirm his adoption anyway. It seems that the Emperor’s real preference was for a boy just a shade too young to succeed to the purple. Marcus Annius Verus was only fifteen, but if Lucius Verus could stagger on for a few more years, then he would at least have had this bright young man beside him looking and learning as the throne was kept warm. But Hadrian’s extended family, not suprisingly, grew resentful and in 137 there was an attempted coup. Led by Pedianus Fuscus, the Emperor’s grandnephew, it probably did not progress beyond a plot, and it failed completely. Retribution followed swiftly as Fuscus was executed, and his grandfather, Servianus, thought to be implicated or at least a potential danger, was driven to suicide. Even though he was close to ninety years old, he was seen by Hadrian as someone who might outlive him and take the throne. Servianus made his preparations, burning incense and praying to the gods: That I am guilty of no wrong, you gods are well aware. As for Hadrian, this is my only prayer: may he long for death but be unable to die.

Servianus’ curse came to pass. As Hadrian lay in agony, his groans echoing around the marble corridors of Tivoli, the plans for the succession unravelled. On the first day of 138, the consumptive Verus died and the future of the Empire was clouded with uncertainty. But, while he still breathed, Hadrian politicked. Having chosen a replacement, a respected and even well-liked senator in the shape of Antoninus Pius, he insisted that the boy he attached to Verus, the bright boy who would become Marcus Aurelius, be confirmed at the same time. In this way the Empire would be in safe hands for two reigns. Here are Hadrian’s thoughts, expressed in a speech to the Senate, probably embellished by Dio Cassius:

But since Heaven has taken [Lucius Verus] from me, I have found as emperor for you in his place the man I now give to you, one who is noble, mild, passionate and prudent. He is neither young enough to do anything rash nor old enough to be neglectful. He has exercised authority in accordance with our ancestral customs, so that he is not ignorant of any matters which concern the imperial power, but can deal with them all. I am speaking of Aurelius Antoninus here. I know that he is not in the least inclined to be involved in affairs and is far from desiring such power, but still, I do not think that he will deliberately disregard either me or you, but will accept the rule, even against his will.

Antoninus appears to have been less than enthusiastic. The wealthy owner of large estates in Italy, he was fifty-one when Hadrian adopted him – and he could see at first hand what the pressures of high office had done. His calm and common sense seems to have had an effect. After the episode with Mastor and the suicide of his doctor, Hadrian wrote to Antoninus with what, from an emperor, amounted to an apology:

Above all I want you to know that I am being released from life neither prematurely nor unreasonably; I am not full of self-pity, nor am I surprised and my faculties are unimpaired – even though I may almost appear, as I have realised, to do injury to you when you are at my side, whenever I am in need of attendance, consoling me and encouraging me to rest. This why I am impelled to write to you, not – by Zeus – as one who subtly devises a tedious account contrary to the truth, but rather making a simple and accurate record of the facts themselves.

The day finally came. On 10 July 138 Hadrian was released, it seemed, from Servianus’ curse. At the desperate last, he had aggressively disregarded medical advice and, eating and drinking whatever he pleased, the Emperor appears to have hastened his end. Without fuss or political difficulty, Antoninus became Emperor. According to the Historia Augusta, Hadrian was first buried at Cicero’s old villa at Puteoli and, in a withering final comment, the author added invisus omnium or ‘hated by all’.

BLACK AND WHITE

We live in a world lit by bright colours. Two thousand years ago the world was less obviously vivid – and yet the sophistication of the ancient language of colour was much greater. Perhaps this was because of the wide spectrum of subtle hues seen in nature, and the fact that more precision is required to describe it. In Latin, for example, black and white were not just black and white. Candidus meant gleaming or shiny white, and St Ninian’s Candida Casa for his very early church at Whithorn attached a more spiritual quality than a mere ‘White House’. Albus meant matt white or chalky white, as in the White Cliffs of Dover, and the adjective must be related to Albion. By contrast ater was matt black or dead black, like a moonless night, and it had overtones of gloom or even malice. Niger was glossy black, like the fur of a sable or lustrous like the skin of a black person. Celtic languages were equally nuanced. The Gaelic adjectives odhar and lachdann describe points in the spectrum somewhere between a parchmenty sort of beige and porridge – but have no satisfactory equivalent in English.

Whatever judgement contemporaries offered, or subsequent historians made, there can be no doubt that Hadrian possessed a towering intelligence and tireless energy. Perhaps he saw something similar in Marcus Aurelius. Despite Hadrian’s cruelties and obstinacies, there was a cultured sensitivity in his nature. While he lay dying, enduring the long hours of pain, Hadrian must have thought on all that he had seen in his vast empire. From the Persian Gulf to the Strait of Gibraltar, from the Danube to the German forests and the ancient grandeur of Athens and Egypt, he had travelled more widely than any who had held his office. Perhaps in the heat of July 138, in his last days, he thought of the windy hills of northern Britannia, the place where he had planned and built the great Wall. At least that would endure, outlasting him and all who followed. His travels seemed to be much on Hadrian’s mind when he composed a remarkable and brief epitaph in a few moments when the pain had receded and he was calm. It seemed at last that he was ready to die.

Little soul, little wanderer, little charmer,

body’s guest and companion,

to what places will you set out for now?

To darkling, cold and gloomy ones –

and you won’t make your usual jokes.

The Senate was not amused. So powerful was their hatred for Hadrian that they at first refused to vote for his deification, an honour which had previously been automatic. When they attempted also to annul his acts, Antoninus intervened. If the Senate persisted with this, then it should realise that one of those acts was Antoninus’ own adoption as heir. Was that what the Senate wanted to see annulled? The argument was as much about respect for the authority of emperors, the integrity of the office itself, and Antoninus was sufficiently astute to suffer no challenge to that, even for someone as loathed as Hadrian.

Where he did play to patrician sentiment was in the matter of Britain. Antoninus planned to repudiate the policy of triumphant retrenchment where it had been most emphatically made manifest. The great Wall in Britannia would become irrelevant. Exposed for what it was, a folie de grandeur like the absurd villa at Tivoli, it would be bypassed as Rome moved beyond the constraints placed on her by Hadrian. There would be conquest once more! The legions would march and the old glories (and, of course, opportunities for ambitious aristocrats) would unfurl and the eagle standards gleam again.

As Claudius did in 43, and other emperors since, Antoninus also needed an opening blaze of prestige. Having no experience himself as a soldier, he would nevertheless bathe his reign in the glow of immediate military success. Britain was a low-risk option. Literally insulated from the European Empire and the Mediterranean, any failure there could be contained, and if there was success then remoteness could only enhance it.

All of this was well received in Rome and, early in his principate, Antoninus acquired the honorific Pius. It means ‘faithful’ or ‘loyal’ and could have had several applications. By reviving the tradition of conquest, Antoninus Pius was being faithful to the spirit and history of Rome. A new and more harmonious relationship with the Senate, the Conscript Fathers, showed loyalty to old Republican institutions. Some might have seen Antoninus’ defence of Hadrian, his adoptive father, as filial piety of an attractively old-fashioned – if misguided – sort.

MAGICAL MYSTERIES

The otherwise sober-sided Antoninus Pius and his empress were both members of the eastern mystery cult of Cybele. It involved self-flagellation, frenzied dancing – and self-castration. Here is the poet Juvenal’s description of their rites: ‘And now comes in procession / Devotees of the frenzied Bellona, and Cybele, Mother of Gods / Led by a giant eunuch, the idol of his lesser / Companions in obscenity. Long ago with a sherd / He sliced off his genitals: now neither the howling rabble / Nor the kettledrums outshriek him.’ Evidence of the worship of Cybele has been found in northern Britannia. Digging at Catterick in 2005, archaeologists came across the grave of a transvestite priest. He or she castrated himself in a gruesome religious ceremony. Using a clamp, similar to one found in the Thames, he had cut off his testicles in imitation of Cybele’s lover, Attis, who had made himself a eunuch as a punishment for an extramarital affair. Mystical eastern religions were popular in the Roman north, and temples to Mithras, an eastern god much favoured by soldiers, can be found along the line of the Wall.

Whatever the calculation and the cold reality behind all that spin, Antoninus seemed a canny political operator. While repudiating Hadrian’s wrongheaded imperial legacy, he appeared to act like a faithful son. And, while ordering the legions to march against the barbarians, he seemed to be leading Rome into a fresh, new era. In fact history turned out differently. Far from harking back to the triumphs of Trajan, Antoninus never left Italy during a long reign of twenty-three years, and never once set eyes on an army or a frontier. Unlike Hadrian, freezing in the forests of Scythia or tramping around Britain, he stayed in Rome, kept the Senate and the mob happy and never went to war unless he absolutely had to.

Contemporary commentators had the delicate task of reconciling the competing images of a powerful war leader at the head of his legions, in spirit if not quite in body, with that of an Emperor who had decided to stay in Rome. In 142, the year of his consulship and the successful conclusion of the war in Britain, Cornelius Fronto managed this awkwardness very adroitly. Here is part of his speech to the Senate:

Although he [Antoninus] had committed the conduct of the campaign to others, while sitting at home himself in the Imperial Palace in Rome, yet like the helmsman at the tiller of a ship of war, the glory of the whole navigation and voyage belonged to him.

Amongst Antoninus Pius’ earliest acts was the appointment of Lollius Urbicus to the governorship of Britannia. In 139 Urbicus was at Corbridge, rebuilding the military depot, no doubt inspecting the Wall garrison, gathering intelligence, making his plans. Dere Street, the old invasion road, crossed the Tyne nearby and the legions would soon be marching north. It seems that there had been war in Britain. The Greek writer Pausanias reported it in this enigmatic passage:

Also he [Antoninus Pius] deprived the Brigantes in Britannia of most of their territory because they had taken up arms and invaded the Genounian district of which the people are subject to the Romans.

Genounia is a mystery (it may be a garbled rendering of an Old Welsh place-name cognate to gwyn or even Guotodin, but it seems unlikely) and its inclusion in a notice of what was happening in Britain may simply have been a blunder. There was a district known as Genaunia in the province of Raetia, modern Switzerland. The use of Brigantes looks like a catch-all for northern British barbarians much in the way that Siberia came to stand for a vast tract of the north-east of the old Soviet Union even though the name originally referred only to the lands immediately to the east of the Urals. Brigantes probably included the Selgovae of the Southern Uplands and the Ettrick Forest, the Anavionenses of Annandale and the Novantae of Galloway. They were, in any case, allies of the Brigantes, part of the great federation of northern hill peoples. The later dispositions of forts in southern Scotland also strongly suggest that the Damnonii of the Clyde basin had sent warbands to attack the Romans.

In the second century AD the kingdoms of lowland Scotland were vigorous. Archaeology has revealed an expanding population and a growing density of settlement. It may be that the kings in the west and the south had turned on the Votadini of the Lothians and the Tweed basin. They had been Roman allies and suppliers, if not exactly subjects, but probably bound to them by treaty. One interpretation of Genounian district holds that it refers to the territory of the Votadini and their cousins across the Forth in Fife, the Venicones. Not part of the Empire in 139, they had briefly been subjects of Rome when Agricola was provincial Governor. Perhaps Antoninus Pius and Lollius Urbicus planned to bring them back into the fold after Hadrian’s badly sited Wall had excluded them, leaving them to the mercy of their covetous neighbours.

In the summer of 140, the new Governor at last rode at the head of an invading army. Their route or routes to the north are not known for certain but they may have followed Agricola’s lines of advance up Dere Street in the east and the modern A74 in the west. Forts were built and rebuilt along these approaches. Three legions were stationed in Britannia in 140, and later inscriptions suggest that the VI Victrix based in York and the XX Valeria Victrix at Chester sent only detachments. Only the II Augusta came in its entirety from Caerleon on the Welsh border, probably leaving behind only a skeleton garrison. This at first glance appears perverse – most men travelling the longest distance, but it probably reflects the pattern of unrest in Britannia at the time. The south was sufficiently secure and quiet to allow all of the II to march north, while the Pennines and the territories either side of them needed watching. Enough legionaries to deal with an emergency had to be left at Chester and York.

Of the auxiliary regiments, the Batavians were serving in Europe, but the feared Tungrians almost certainly formed part of Urbicus’ strike force. Others, such as the Hamian Archers from Carvoran and the cavalry, the Ala Augusta, from Chesters had raised their banners and were going to war.

In two years of campaigning, perhaps less, that war had been won. No details survive but in AD 142 coins were minted to commemorate a great victory. And in Rome Antoninus Pius was acclaimed Imperator, the first and only time this happened. Job done.

Or it seemed to be done. The frontier of the Empire had been successfully extended by 150 kilometres, a victory gained and new territory subdued. But some in the imperial administration sniffed. Here are the comments of a Greek civil servant, Appianus:

The Romans have aimed to preserve their empire by the exercise of prudence rather than to extend their sway indefinitely over poverty-stricken and profitless barbarian peoples. I have seen embassies from some of these in Rome offering themselves as subjects, and the Emperor refusing them, on the grounds that they would be of no use to him. For other peoples, limitless in number, the Emperors appoint the kings, not requiring them for the Empire . . . They surround the Empire with a circle of great camps and guard so great an area of the land and sea like an estate.

And later, Appianus offered his own interpretation of imperial attitudes to the province first conquered by the armies of Claudius:

They have occupied the better and greater part of it [Britannia] but they do not care for the rest. For even the part they do occupy is not very profitable to them.

Others saw it differently and, paradoxically, were more pragmatic about the political wisdom of spending money and resources on invading bits of territory of little economic value. Here Hadrian’s friend, the poet, Annaeus Florus, is talking about two remote provinces, Britannia and mountainous Armenia:

It was fine and glorious to have acquired them, not for any value, but for the great reputation they brought to the magnificence of the Empire.

In Britain the job was done efficiently. Over a wide swathe of southern Scotland Urbicus set his victorious soldiers to the tedious task of consolidation. Roads were refurbished and new stretches constructed. Forts were dug. From Glenlochar, near Castle Douglas, and from what is now Gatehouse of Fleet to Castledykes and Bothwellhaugh in Clydesdale, garrisons showed the flag and began to patrol and gather intelligence. Roads struck directly through the heart of dangerous hill country, nowhere more so than the dramatic Dalveen Pass between Upper Nithsdale and the upper reaches of the Clyde. There was an unmistakable sense in the 140s of the beginnings of provincial government.

The hub of the occupation of the south was at Newstead on the banks of the Tweed, in the lee of the Eildon Hills, the place called Trimontium. Close to Dere Street, which probably bridged the Tweed nearby, at Leaderfoot, Trimontium was the largest fort north of Hadrian’s Wall. With several substantial annexes (fenced and guarded enclosures) around it, Newstead was a supply depot as well as important tactically. Brilliantly excavated by a Melrose solicitor, James Curle, before the First World War, it has given up one grisly memory of warfare in southern Scotland. Buried with a good deal of military rubbish, several skulls were found in a series of pits. Roman burials always took place well away from human habitation, often on roadsides, and in any case the skulls had no associated skeletons. They had not been interred. On Trajan’s Column soldiers are shown raising up the severed heads of Dacians and, at the gates of a camp, two have been impaled on poles and stuck into the ground. At Vindolanda Andrew Birley’s team have recently found a skull which had been stuck on a pole in exactly the same way. DNA analysis has demonstrated that the dead man was probably a native warrior from the kingdom of the Anavionenses. It seems certain that the Newstead skulls were those of captured or killed Selgovan soldiers whose severed heads were spitted on poles and set up at the fort gates in a gruesome display of Roman ruthlessness. We are inclined to see the legionaries and auxiliaries are clinical and efficient killers. This shows an atavistic savagery – something surely worthy of a barbarian.

After the native kings had been defeated and their lands brought under control, something extraordinary then took place. Here is the entry in the Historia Augusta:

[Antoninus Pius] conquered the Britons through the agency of the Governor, Lollius Urbicus, and having driven off the barbarians built another wall in turf.

Another wall! Twenty years after Hadrian commanded his vast project to commence and perhaps fifteen years after its completion, the Roman army built another wall across Britain. The same legions were involved and many of the same soldiers are likely to have started work on a second wall. While they are often stoic about the insanities of army life, soldiers are not immune automata. Many of the older men must have shaken their heads when the orders were given. What was wrong with the first wall?

Nothing at all. Except that it had been built on the orders of Hadrian. His name was inextricably attached to it, and if Antoninus wished to be seen actively repudiating his policies then one of their greatest monuments needed to be abandoned. Immediately. But this extraordinary decision has another fascinating aspect. It was not extraordinary to the Romans. We see these massive works differently, as expressions of vast expense, resources, will and effort. To abandon something like Hadrian’s Wall as soon as it was completed would have been a profligate waste.

Antoninus and his generals did not see the decision in this light. At their command was a large, versatile and highly skilled army which would need to be paid and fed whatever happened. The abandonment of Hadrian’s Wall was not a disaster or a waste, just a matter of strategy. Because they had the army, as much a huge and relatively well-disciplined labour force as a fighting machine, the Romans thought on a different scale. They were masters of the world and it was for them to order it as they saw fit.

These were pervasive attitudes. If their officers thought like this, so did the men and, while the prospect of building another wall will have raised the eyebrows of more than a few centurions, there was no hesitation in setting about the task. In any case, this wall would be easier: half the length and built out of turf, not stone: no problem.

As with Hadrian’s Wall, the western kings appear to have given more cause for concern and, before any soldier began work on the Wall, it seems that forts were built to hold down the lower Clyde basin. The line chosen was indeed much shorter at 59 kilometres compared with 120 kilometres. It was to run from Bridgeness on the Firth of Forth in the east to Old Kilpatrick on the bank of the Clyde in the west, across the narrow waist of Scotland. Crucially, there was to be no Vallum, a tremendous saving of labour and time, and, while a short sequence of forts was planned at either end, nothing on the scale of the Cumbrian Sea-Wall would be needed.

TURF WARRIORS

The garrison of the Antonine Wall is less well attested than that of Hadrian’s Wall and only those forts with named units are listed below.

Mumrills	– Ala I Tungrorum
	– Cohors II Thracum
Rough Castle	– part of Cohors VI Nerviorum quingenaria peditata
Castlecary	– part of Cohors Tungrorum milliaria peditata
	– part of Cohors I Vardullorum milliaria equitata
	– vexilations from II Legion and VI Legion
Bar Hill	– Cohors Baetasiorum quingenaria peditata
	– Cohors Hamiorum quingenaria
Cadder	– Cohors quingenaria peditata
Balmuildy	– Cohors quingenaria peditata
Castlehill	– Cohors IV Gallorum quingenaria peditata
Old Kilpatrick	– Cohors I Baetasiorum quingenaria peditata

The Antonine Wall straddled more than the shortest distance between the North Sea and the Atlantic sea-lochs of the west; it was also a man-made recognition of one of Scotland’s most profound geographical divisions. To the south lay Clydesdale and Ayrshire, the fertile Lothians, the gentle slopes of the Southern Uplands and Galloway, and the rich farmland of the Tweed basin. In the north were the mountains and the islands, the formidable rampart of the Highland Line visible for many miles. There was Fife to the north-east, and beyond it the coastlands of Angus and Tayside, and while these places had powerful links with the south they were nevertheless culturally distinct. What is now Scotland used to be split into two, and Tacitus thought that the north was like adifferent island.

And it was. Motorways and modern drainage have erased this ancient frontier. Two thousand years ago, landward communications were much more difficult. At the western end of the line of the Antonine Wall, the Highlands rise up abruptly. In the central section, overland travel to the north was made circuitous and even dangerous by the Flanders Moss. A wide tract of treacherous marshland, it made much of the valley of the meandering River Forth impassable in winter and very awkward in summer. Drainage had to wait until the eighteenth and nineteenth centuries. Only in the east, through the Stirling Gap, was there firm passage. Funnelling under the glowering crag of the great castle, armies intent on the conquest of all of Scotland were forced to march that route. For this now-lost reason of geography, Stirling held the key to the whole kingdom for many centuries.

With no emperor to interfere, Lollius Urbicus and his surveyors rode from the Forth to the Clyde pegging out the line of the new Wall. They made good decisions. Even though the geography is much less dramatic than the Whin Sill, the Antonine Wall nevertheless commands the ground emphatically. Rising up from the coastal plain at Bo’ness, it quickly finds the high country to the south of the valley of the River Carron. From there long vistas stretch out over west Fife and as far as the Ochil Hills. In contrast with the bleak prospects from the high parts of Hadrian’s Wall, these are views of heavily populated areas, fertile flatlands bordering the Forth and the better-drained terrain towards Alva and Dollar. In the central section, the Wall hops from one vantage point to the next, is by no means straight, and at Croy Hill and Bar Hill it climbs to its highest point above sea-level. The valley of the little River Kelvin provides a handy southern slope where the garrison could look out over the Campsie Fells and the threatening mass of the Drumalban Mountains behind them. At Balmuildy the line turns sharply north-west for a few hundred metres and then strikes through what is now the well-set Glasgow suburb of Bearsden before reaching its terminus at Old Kilpatrick on the northern bank of the Clyde. Only the ford at Bowling lay beyond the end of the Wall; all of the others were upstream.

Outlier forts and fortlets were built across the river at Bishopton and even further west at Lurg Moor and Outerwards. More substantial were the forts beyond the eastern terminal. Cramond stood on the Forth shore, at the outfall of the River Almond, and, 18 kilometres further east, Inveresk was built near another river-mouth. From these, soldiers could watch the Fife shoreline, probably not fearing attack from that quarter but, rather, attacks on it. The corn-growing Venicones were likely Roman clients, well within the protective compass of the new Wall, and the lookouts at Cramond and Inveresk will have watched for suspicious movement on the northern shore. Further help for native farmers lay beyond. Lollius Urbicus reactivated part of the old Gask Ridge system and there were outpost forts once more at Ardoch, Strageath and north-west of Perth. This was a comprehensive pattern of occupation and an entirely unmissable show of strength. The Empire had come back north.

Work began on the new Wall. Probably in the summer of 142 the three legions dug the first of their temporary labour camps. The outlines of eighteen of these have been detected, mostly from the air. As at Hadrian’s Wall, construction was organised in legionary lengths and also appears to have begun in the east. Following the line laid out by the surveyors, the gangs first dug a shallow foundation trench. Stone footings were tapped in, made as level as possible and the edges set out with large kerbstones finished only on the outside face. Even though it was to be a turf wall, a great deal of stone was quarried. The building principle was simple: to build a flat and stable foundation, and border it with kerbstones which would keep the width of the Wall consistent and help prevent spreading.

Culverts were let into the foundations at frequent intervals so that, in winter, or after heavy rain, there would be no ponding behind the rampart. There had been too few culverts on Hadrian’s Wall, and it seemed that the legionary builders were learning.

Since the basic building material, turf, lay around the line of the Wall and did not need to be quarried, the logistical support for the men working at the site was much less extended and much less complex. Once the stone founds had been laid, virtually every soldier could be involved in construction because they undoubtedly possessed the skills required. On the march, most nights, the Roman army dug temporary camps which used exactly the same methods. Despite the fact that some learned opinion now believes that the Antonine Wall took several years to build, in fact it rose quickly.

Each legionary was issued with a crescent-shaped turf cutter, something like the tool now used to straighten up the edges of perfect suburban lawns. With it, soldiers expertly sliced out large rectangular pieces of turf. And, as with almost everything they built, the Romans specified a standard size, in this case 45 centimetres by 30 centimetres by 15 centimetres. That is large, and if taken from damp ground, a single turf might have weighed 30 kilograms. One of the advantages of using them was that one man could carry a single piece but, at that weight and bulk, it might easily have disintegrated. If the ground was dry, it would be even more difficult to keep a large piece of turf together. On Trajan’s Column there is an illustration of turf cutting and it shows one soldier loading a large piece onto another man’s back with the help of a rope sling. That seems a lot of handling for a simple job, but it may well have been standard procedure.

Perhaps ground conditions varied more than they do now. The point of turf as a building material is the grass on top of it. Its roots knit the earth together sufficiently for it to keep its basic shape when lifted out of the ground. These were not the tidy squares of turf delivered by today’s garden centres. Natural pasture in AD 122 had much longer and stronger roots than the wispy stuff grown by farmers and horticulturalists today. If the work on the Antonine Wall was done in the spring and summer when the grass and all the other herbage and the weeds that grew in it were tall, then not only would cutting be awkward, the roots in the ground would have been more vigorous and binding – and in some cases very much longer and tougher.

Whatever their size, the turfs were laid like stone with each course’s joints bonding as neatly as possible with the middle of the one below. On a base of 4.2 metres, the Wall was able to rise safely to 2.5 metres. For stability the sides sloped, narrowing to a 1.8 metre top. This was not like stone in that sense, and there could be no vertical faces. In any case the profile and possibly the height of the Wall would have altered quite quickly as the turf settled and moisture drained downwards.

As with Hadrian’s Wall, there is much debate about what the top of the Antonine Wall looked like. Was there a walkway? Was there a wooden breastwork with crenellations? The answer is probably yes to all of these. A sloping turf surface is much easier and quicker to climb than a vertical stone wall, especially one which had been rendered. It seems inconceivable that the soldiers on watch would not have had access to the top of their Wall, or indeed a breastwork to shield them – from the weather if nothing else. And one with crenellations will have made the rampart look higher and more formidable.

What also made the Wall look taller was the forward ditch. North of the Wall, work-gangs dug down to 3.5 metres and piled the upcast on the northern flank to make the downslope even deeper (the Antonine Wall used much less turf than the original western end of Hadrian’s Wall, between the Irthing and Carlisle, but the ditch was significantly deeper, again giving an impression of greater height). The whole earthwork was 12 metres across. The ditch is the most visible relic of the Wall, there being very little to see of the turf rampart, and surprisingly at its most impressive in Falkirk, at Watling Lodge. Between it and the Wall a substantial berm was left. And behind both, the legionaries laid down a military road which ran the entire length, from Bridgeness to Old Kilpatrick.

One particular element of the building programme is eloquent. On Hadrian’s Wall the soldiers dug pits known as lilia on the north side. Remains of them have been found at Byker and Throckley in Newcastle. Filled with sharpened stakes and then covered over with leaves and twigs, they were the Roman equivalent of barbed wire, intended to break up an assault. On the Antonine Wall the builders appear to have expected attacks and there are examples of extensive fields of lilia. At Rough Castle, near Falkirk, there are no less than ten rows of pits offset like the black squares of a chessboard, and in each row there are twenty pits. This sort of added defence took a good deal of effort and some maintenance – but it was obviously thought necessary.

The organisation of the garrison on the Antonine Wall was different. And, as in the south, there were changes of plan. At first, forts were to be arranged in a similar order with one every 12 kilometres or so. There were to be seven in all, with fortlets interspersed and, at Camelon, one fort forward of the rampart. And then the plan suddenly changed. As the work-gangs reached the stretch between Bearsden and Duntocher, not far from the Clyde and the end of the line, it was decided to add no less than ten new forts to the Wall. Now there was to be a garrison (of differing sizes) every 3.5 kilometres. What prompted this radical change in plan is not recorded – but it must surely have been something dramatic. Perhaps there was hostile action against the Wall and its soldiers during construction. It lay close to the Highlands, difficult country to patrol, and perhaps warbands erupted out of their glens without any warning and launched themselves at this extraordinary structure and the men who were daring to make it.

BLOCKED IN BEARSDEN

Property prices in the genteel Glasgow suburb of Bearsden might be thought to benefit from the presence of Roman remains. History on a doorstep adds a certain cachet. Less appealing to estate agents might be the fact that one of the most interesting finds at the fort was the effluent from a latrine. Like the famous example at Housesteads, it was communal. The sewage drained into the ditches beyond the rampart where it was covered with water – and therefore probably did not smell. Analysis has revealed a largely vegetarian diet for the average soldier: cereals supplemented by wild fruits and nuts, such as raspberries, brambles and hazelnuts. Food was not as fresh or as hygienically prepared as it is now and some of the men suffered from roundworm and whipworm. Too much information? Less worryingly, bits of ancient moss were found in the ditches and absolutely no trace whatsoever of sponges and sticks.

The native kings knew of the great stone Wall in the south and it may be that some of their warriors had joined their allies, the Brigantes, and fought in the countryside around it. When a second Wall began to slice through their own territory, it is not likely that they were content to sit quietly and watch. And as Roman soldiers worked, they were more vulnerable to surprise.

Meanwhile Hadrian’s Wall appears to have been all but deserted. To allow easier access, the gates were removed from the milecastles, and every 40 or 50 metres some of the upcast was backfilled into the Vallum to make causeways. Garrisons fell back to the old Stanegate forts of Corbridge, Vindolanda and Carlisle. The focus was now firmly in the north as the turf Wall took shape.

The density of the Antonine Wall defences is very striking. The garrisons were stationed very close to one another. Admittedly they varied in size from the small detachment which fitted into the tiny fort at Duntocher to the much larger units based at Balmuildy, for example. Most of the soldiers were infantrymen and there was cavalry only at Mumrills in the east (the Tungrians) and probably at Bearsden in the west. Given the boggy and difficult nature of the ground immediately to the north of the Wall, infantry may have been a better option. The Roman army always preferred to fight in formation, using well-rehearsed tactics. By contrast, native horsemen were raiders and skirmishers, like the Border Reivers long after them, with sure-footed ponies able to move quickly through uncertain ground. This seems like a mis-match but, in reality, it was probably not sensible for the Romans to pursue but rather to police effectively those places they could reach on foot. As ever they were determined to fight only on their own terms.

ANTONINE NAMES

Perhaps because the Antonine Wall was in operation for such a comparatively short time, we do not have a complete list of well-understood names for the forts along it. Some are entirely unknown, others wildly speculative. However, place-names sprinkled on either side of the Wall are interesting. ‘Medionemeton’ means Middle Shrine and it might have two references. The less likely location is a Roman monument known as Arthur’s O’en. Looking a bit like a huge oven or a giant beehive, it stood 3 kilometres north of the Antonine Wall, near the River Carron. It was probably dedicated to the goddess Victory, perhaps in thanks and celebration for the defeat of the northern kings in the campaign of AD 139–142. A large, circular and domed structure, it was demolished in 1743 and its dressed stones used to build a mill dam for the Carron Ironworks. The other, more likely, location of the Middle Shrine is Cairnpapple Hill near Linlithgow. An elaborate prehistoric temple and burial site for millennia, it seems to stand on a Scottish meridian with clear and long views in every direction. The Highland Line, the Forth, the Southern Uplands and the Firth of Clyde can all be seen from its summit, and the name means the Stone of the Priests in early Gaelic.

At Newstead, the conditions were different. The well-drained rolling farmland of the central Borders was good cavalry country. In addition to two cohorts of the XX Legion, a full regiment of troopers was stationed on the banks of the Tweed. The Ala Augusta Vocontiorum had the hitting power and the speed to confront the fast-moving warbands of the Selgovae. Horses for courses.

At Newstead, and outside the walls of most of the forts on the Antonine Wall, the soldiers built annexes. In part these enclosures were a quicker and more economical solution to the problem solved by the Vallum dug so laboriously to the south of Hadrian’s Wall. Instead of all that sweat spilt to create a secure military zone behind the entire length of the Wall, Lollius Urbicus’ planners decided to localise it. At the flanks of the Antonine forts, annexes could be used to house additional troops, enclose stock or provide a well-guarded cordon around goods, which might otherwise have disappeared during the night.

The date of the completion of the Antonine Wall is nowhere recorded, but the finishing of its legionary lengths is. A set of unique stone plaques, known as distance slabs, has been found, which carry the names of the legions who did the work, commemorating who did what. Set up on both the northern and southern faces of the Wall at the end of each length, some are very ornate – and informative. Not surprisingly the mascot animals of the legions appear: a charging boar for the XX, and a goat and Pegasus, the winged horse, for the II. Various representations of the goddess of Victory are also carved.

The story of the War for the Wall is told in two of the most interesting slabs. Set up by the II Legion, one of these carries a central inscription, which is mostly taken up with the many names of the Emperor Antoninus Pius. Reading from the left, there is a panel showing a Roman cavalryman riding over four naked barbarians, clearly vanquishing them. Or him. The four figures may represent one warrior being attacked, knocked to the ground and wounded, and perhaps begging for mercy, his weapons cast away. No mercy was given, for at the bottom of the panel he has been decapitated. The sculptor was no great artist, but he has tried hard to impart a sense of movement to the cavalryman and his pony. Its tail is up, its forelegs leaping into the gallop, and its rider’s cloak flies out behind as he thrusts downwards with his spear.

On the right of the inscription there is a religious scene. Under the banner of the II Legion, a priest, probably the commander, Aulus Claudius Charax, pours out a libation onto an altar. Wearing what is probably a senatorial toga, he is surrounded by his fellow officers. Some are bearded, like both Hadrian and Antoninus, while others are clean-shaven, and it may be that these are rough portraits. A figure plays the pipes and another kneels beside three animals: a bull, a sheep and a pig. Soon these poor creatures will have their throats cut in sacrifice, and the strong impression is that this ceremony is a dedication to celebrate the completion of the Wall, or at any rate the section at Bridgeness. It may also be a thanksgiving for the victorious outcome of the war being fought on the other side of the panel.

A distance slab was found at the western end of the Wall, at Hutcheson Hill, which shows another legion celebrating. In a central panel this time, the goddess of Victory awards a wreath to the XX Legion. As she places it on the beak of one of their eagle standards, two kneeling, naked and bound captives watch from the side panels. There can be no doubt about who is who and what is what. Rome has not just won but triumphed over the miserable, naked barbarians of the north.

Nakedness on the distance slabs was clearly the condition of defeat and enslavement but, as noted earlier, when one of the Vindolanda letters observes that the natives were naked, it meant that they are unprotected by armour. This was something the Romans not only sneered at but found difficult to understand. But in reality the sort of well-made and effective protection worn by legionaries and auxiliaries was unusual in the ancient world. Because the Roman army was professional and the deposit of a great deal of investment in training and pay, it made sense for soldiers to be well protected and for them to survive as long as possible. Dead soldiers were simply a waste. In the main, native warbands were not professional and did not wear armour. Their cavalry did not resemble the helmeted trooper riding across the Bridgeness slab.

The Antonine reoccupation of Scotland and discoveries at Corbridge have contributed a great deal to an understanding of this crucial Roman advantage. At the site of Newstead Fort the remains of a lorica segmentata were found by James Curle in 1906. This was the most common type of body armour worn in the western Empire from the late first century AD on into the middle of the second. Designed like a set of wide, overlapping metal ribs, it fitted around the abdomen and was flexible, allowing a soldier to move without much constraint. In combat an ability to move freely and quickly could be a matter of life or death. The Newstead type, and also a damaged lorica found at Corbridge, had heavy protection on the shoulders. Sword strokes over the top of a shield were obviously common and the overlapping shoulder plates are also fitted together like scales, both to protect and make it possible for a fighting soldier to raise his arm unhindered.

These cuirasses must have been very expensive, but it seems that soldiers bought them, either second-hand or direct from the manufacturer. Factories must have existed all over the western Empire, for, over the period of their use, hundreds of thousands of loricas were worn. The process of manufacture was all manual. Plates were hammered into shape, not produced by rolling mills. Roman armour lasted a long time and was passed on by fathers to sons; some examples are known to have survived in use long after the Empire in the west had disappeared.

Chain mail was more expensive, but more flexible. It took around 180 hours to make the most rudimentary shirt because it had at least 22,000 rings. Officers wore them, and often scales were latched onto the rings to make an even tougher piece of armour. The surprise is that mail was worn by auxiliaries. How did they afford it?

Parts of helmets have been found along Hadrian’s Wall and they all appear to have followed the same basic design. A skull cap or pot had three elements added. At the back there was a wide neck-guard, to the sides hinged cheek-pieces, and a thick brow-band was fitted at the front to deflect downward blows. Most distinctive, at least in Hollywood epics, were the red horsehair cockades worn like a brush on the top. An example was preserved in the anaerobic mud of Vindolanda and in the main such decorations were the prerogative of senior officers. It should perhaps be expected that cavalry helmets had a little more dash. Often heavily decorated, some had a plume-tube fitted so that when the horses galloped whatever they had attached would stream out behind. The trooper on the Bridgeness slab wears a fine example, shaped like a pony tail. At Newstead a splendid metal cavalry mask was found. Not a piece of armour, it was used only for display on the parade ground. The comments of the infantry can easily be imagined.

The Antonine Wall showed every sign of permanence. A large garrison had been concentrated in a small, well-placed and heavily defended area – in fact it was only slightly smaller than that of Hadrian’s Wall, which was of course twice the length. There were between 6,000 and 7,000 soldiers in seventeen forts, compared with a total of around 8,000 on the stone wall in the south. Such large numbers suggest a large native population. Most of their duties were more like modern police work: the pursuit, trying and punishment of criminals (with commanding officers as magistrates), the escort of important people and a general effort to keep order. During the reign of Marcus Aurelius’ son, Commodus (AD 180–192), forts and towers were built along the banks of the Danubeto prevent the secret crossings of petty raiders.

ROMAN SPORRANS

The lorica segmentata, a mail shirt, could only reach down so far, and Roman soldiers were unable to protect what many men reckon to be the most vital part of their anatomy. The Roman solution was a kind of sporran. A few leather straps were sewn onto a belt and had metal studs hammered into them. This was flexible enough to allow free movement and it did offer some protection – although not much. Like a heavy sporran, it could probably have been uncomfortable while running or even walking quickly. Perhaps most Roman soldiers relied on their long shields and a sporting respect from the similarly vulnerable warriors they fought against. But then again, perhaps not.

The Antonine system seems also to have been much more clearly thought out, more integrated. All the forts faced north and formed part of the rampart, while the road behind them linked them all very closely. The Antonine Wall was perhaps the most advanced example of a Roman linear frontier.

There was no sense that it would be a short-lived expedient. Vici, the familiar civil settlements, quickly grew up and, beside the walls of the fort at Carriden, the inhabitants seem to have been particularly independent-minded. Having organised themselves into some sort of self-governing entity, they set up an altar, dedicated to Jupiter Best and Greatest. It also named Aelius Mansuetus on it, possibly a civil official or leader.

A pottery workshop was established at Bearsden by a man called Sarrius. It was part of a chain. He already had factories near Leicester and Doncaster, and, as the north opened up, Sarrius could clearly recognise a business opportunity when he saw one. It appears to have been successful: several sherds have been found with the firm’s name on them.

Despite all this activity and all the effort expended to build a new northern frontier, it was abandoned after only fifteen years’ occupation. It was Antoninus Pius who made the final decision. Probably in 157, he and his council in Rome decided to pull the frontier back to Hadrian’s Wall. Lollius Urbicus, the former Governor of Britannia and builder of the new Wall, was instrumental. By 157 he was almost certainly Prefect of the City of Rome and a key member of the imperial council. Knowing the situation in Britain better than anyone else, he may have advocated the pull-back himself. The invasion and holding of the north had long since served its purpose and after twenty years on the throne Antoninus no longer needed to be associated with any fading glory. And, with legionaries certainly included in the Antonine garrisons (usually this sort of frontier posting was given to auxiliaries), there is a hint of overstretch.

The Wall was not demolished nor the ditch backfilled. It stood high enough to be clearly visible to the surveyor, William Roy, in the middle of the eighteenth century, and the ditch has survived well in places even now. Forts and their buildings were slighted or burned, and the distance slabs carefully buried. Having military honours for the legions inscribed on them, they were semi-sacred objects, and no barbarian, naked or otherwise, would be allowed to deface them.

TAGINES

Archaeologists digging at Bearsden Fort came across pottery made for cooking in an African style. Large pots were produced to sit directly on top of a brazier. How did they get there? There are no records of units from North Africa posted on the Antonine Wall – but there was a war in the province of Mauretania (part of modern Algeria) during the reign of Antoninus Pius. It began in 145 and ended five years later, and it may be that units were sent from the frontier in central Scotland on an immense journey to the frontier in North Africa to fight against more barbarians. There they appeared to have enjoyed the local cuisine and, when they arrived back in Bearsden, they had the right sort of pot fired in a kiln and made a brazier to hold it. The pots sound very much like tagines, heavy earthenware pots used by North African nomads to cook on charcoal braziers. They have conical lids which ensure that none of the condensation caused by cooking escapes, and this enrichs the stew or whatever else is being cooked. Cosmopolitan indeed.

As the Antonine Wall passed into history, misconceptions and myths began to gather around it. The normally scrupulous Bede of Jarrow got it wrong when he reported that it had been built in the fifth century when a Roman army returned to southern Scotland. Having helped the locals repel an invasion of Picts and Scots (some whispers of genuine history here), the Romans advised them to build a wall to keep marauders at bay. By the eighth century, when Bede was writing at Jarrow, building in the Roman mannermeant stonework. Turf had to be the work of the primitive British:

The islanders built this wall as they had been instructed, but having no engineers capable of so great an undertaking, they built it of turf and not stone, so that it was of small value. However they built it for many miles between the two estuaries, hoping that where the sea provided no protection, they might use the rampart to preserve their borders from hostile attack. Clear traces of this wide and lofty earthwork can be seen to this day. It begins about two miles west of the monastery of Abercorn at a place which the Picts call Peanfahel and the English Penneltun, and runs westward to the city of Dumbarton.

By the fourteenth century bad history had turned into myth-history. The Scottish chronicler John of Fordun reckoned that the Wall had been cast down (perhaps it was in a ruinous state by his time) by Gryme, the son of King Eugenius. Writing in the sixteenth century, George Buchanan thought that Graeme, a leader of the Picts, had broken through it. King Graeme? In any event the local name for the Antonine Wall was the Grimsdyke and it lives on in modern streetnames. In Bo’ness there is a Grahamsdyke Road and a Grahamsdyke Lane, while in Laurieston there is a Grahamsdyke Street. If Gryme and the very unlikely Graeme are set aside, what did the name mean? There is an intriguing old Scots expression, a Grime’s Dyke, which means a ditch made by magic. The Old English grim originally meant fierce or aggressive. Either interpretation could work.

There are powerful archaeological arguments that the withdrawal from the Antonine Wall was gradual, managed in stages over four or five years. Dating after 158, signs of rebuilding have been detected at Chesters, Corbridge and Vindolanda, and redeployment for some units may have had to wait until the old Wall had been repaired and made habitable. The turf section in the west was replaced with stone during this period.

Most important, there seems to have been trouble in Britain. Around 155 the Brigantes may have risen in rebellion and the depleted legionary garrisons at Chester and York may have been unable to contain them. Coins were issued with the image of Britannia subdued, usually a sign that a war had been won, and troop movements are also suggestive of trouble in the Pennines. An inscription pulled out of the River Tyne records the arrival of reinforcements for all three legions in Britannia. But, in the nature of fragmentary evidence, it is possible to interpret it in the opposite direction. Soldiers may have been sailing down the Tyne to reinforce comrades in Germany. What is certain is the arrival of another experienced and talented general as Governor. Julius Verus is recorded ordering building work at Birrens Fort, just to the north of Hadrian’s Wall and at Brough, at the southern end of the Pennines, in Derbyshire. Despite the ambiguities around the question of reinforcements, it looks as though the Brigantes once again forced a change in Roman policy. After almost a century of occupation, their kings were still powerful. The simple cause and effect may be that Lollius Urbicus advised his Emperor to withdraw elements of the Antonine garrison to suppress revolt in the Pennines.

After a long reign, from 138 to 161, longer than any emperor since the first two, Augustus and Tiberius, Antoninus Pius died. He was seventy, and the succession followed on untroubled. Marcus Aurelius had been nominated by Hadrian and in the last years of Antoninus he was closely involved in government. In turn, his reign was to bring to an end what the great historian of Rome, Edward Gibbon, called a Golden Age:

If a man were called to fix the period in the history of the world, during which the condition of the human race was most happy and prosperous, he would, without hesitation, name that which elapsed from the death of Domitian to the accession of Commodus.

It is doubtful if the kings of the Brigantes would have seen their world in quite the same roseate glow. Gibbon’s estimate of the reign of Antoninus is in the same vein:

His reign is marked by the rare advantage of furnishing very few materials for history, which is, indeed, little more than the register of the crimes, follies, and misfortunes of mankind.

Despite this sense of Britannia dozing contentedly in the high summer of imperial Rome, the south was threatened again in 161, and another good soldier, Calpurnius Agricola, was despatched to deal with it. His governorship is well recorded in a series of inscriptions set up on Hadrian’s Wall as it was being recommissioned, and also elsewhere in the north. Newstead continued to be held, well forward of the Wall, and other outpost forts were maintained in the meantime.

In the Senate, Marcus Aurelius had insisted that Lucius Verus, long seen as the spare rather than the heir, be confirmed as co-emperor, the first time this had happened. It appeared that Marcus was the senior partner and Verus was given rein to campaign abroad in the eastern provinces. After an expedition to Parthia, in the Middle East, the Historia Augusta was not impressed:

He had brought with him [on his return] both minstrels and pipers, actors, pantomime jesters and jugglers, and all kinds of slaves in which Syria and Alexandria take pleasure, to such an extent that he seemed to have finished not a Parthian war but an actors’ war.

Verus died of apoplexy in 168, probably a suffering a stroke (he lived on speechless for three days) before expiring and Marcus was left to govern unhindered. Hadrian was enormously fond of him, and his judgement of Marcus Aurelius’ abilities seems to have been sound. While forced to campaign almost ceaselessly on Rome’s frontiers, especially the Danube, and endure the same sort of rigours Hadrian relished, he showed himself a true philosopher-king. What impressed Gibbon (and indeed anyone who reads them) was a series of writings, To Myself, which survived as Marcus Aurelius’ Meditations. They are the deposit of a tremendous intelligence:

What peculiar distinction remains for a wise and good man, but to be easy and contented under every event of human life . . . ? Not to offend the divine Principle that resides in his soul, nor to disturb the tranquility of his mind by a variety of fantastical pursuits . . . To observe a strict regard to truth in his words and justice in his actions; and though all mankind should conspire to question his integrity and modesty . . . he is not offended at their incredulity, nor yet deviates from the path which leads him to the true end of life, at which everyone should endeavour to arrive with a clear conscience, undaunted and prepared for his dissolution, resigned to his fate without murmuring or reluctance.

Marcus’ equanimity was tested in 168 when German barbarians burst through the Danube frontier and reached northern Italy before being caught and defeated. This scare must have had a profound impact in Rome. The Emperor was on the Danube in the late 170s, fighting, amongst others, the Sarmatians and their heavy cavalry. At the same time, Avidius Cassius rose in rebellion in the east, claiming the imperial throne. To buy a hasty peace, Marcus accepted 8,000 Sarmatian cavalry into the Roman army and, in AD175, 5,500 of them arrived in Britannia. The culture shock on both sides must have been considerable. Marcus’ rationale was probably very simple: a continental European posting for these horse-riding soldiers might see them slip away and ride home to the Danube shore, but escape from the island of Britannia might prove a little more tricky. Some time later, Sarmatian veterans settled at Ribchester in Lancashire, and there are traces of them at Chesters Fort on Hadrian’s Wall. It may be that they were used in the war in the Pennines against the Brigantes.

Gibbon’s Golden Age ended abruptly in 180 when Commodus Antoninus succeeded his father. Assassination attempts were made almost immediately, and it must be a testament to his ingenuity, strength of will and good fortune that he survived for twelve years. He was probably the most dissolute and chaotic emperor since Nero.

During the 180s war erupted again in Britain, and this time it threatened the whole province. Dio Cassius reported that native armies crossed the Wall that separated them from the Roman garrison, had killed a general (probably the Governor) and massacred his army. Evidence of destruction has been found along the eastern section, at the neighbouring forts of Halton Chesters, Rudchester and Corbridge. This suggests an invasion route down the Northumberland Plain and through the lower Tweed, sidestepping the garrisons at Newstead and the outpost forts on Dere Street.

A DRAGON ON A STICK

The Sarmatian heavy cavalry drafted into the Roman army by Marcus Aurelius wore armour and protection for their horses, but they held another innovation in their hands. They used a long cavalry lance, called a contus in Latin, perhaps 6 feet long and a deadly weapon with the momentum of a charging pony and rider behind it. The Romans adopted the contus quickly. Up to and including the Light and Heavy Brigades of Crimean fame, cavalry squadrons began their charge with the lance held upright and only levelled it at an enemy at the last moment. The Sarmatians attached flags and often a mythical beast to the end of them. The draco was a hollow, open-mouthed dragon’s head with a long tube of red or white material flowing behind so that, when the horseman kicked his pony into a gallop, it would fill with air like a windsock. For extra effect, reeds were inserted into the dragon’s mouth so that when air passed through, it seemed to scream. How this struck those hoping to repel a Sarmatian charge can only be imagined. Perhaps the dragons first hissed along Hadrian’s Wall. By the fourth century many Roman cavalry regiments had adopted the draco. Britain’s oldest national emblem, the red dragon of Wales, may well have come from the Sarmatians.

Ulpius Marcellus was sent to replace the Governor and it took at least three campaigning seasons to restore some sort of order. But it seems that the forts to the north of the Wall were abandoned once more, presumably because their garrisons were needed to strengthen the south, and because Scotland could no longer be held. Birrens and Newstead were never again reoccupied, and on Dere Street, over the Cheviot watershed and down into Northumberland, High Rochester and Risingham were also given up. In spite of a coin issue in 184 to celebrate a victory, and the adoption of the title of Britannicus by Commodus, it sounds more like a stand-off than any clear-cut result. Native kings had seen a Roman retreat in the late 150s, and when a weak Emperor succeeded in 180 they hit hard and cleared his army out of all the territory occupied in 139 to 142. In the circle of firelight their bards would have sung of victory.

The invasion had a very visible impact in the south. Walls began to rise around towns. Before 180 only a few had protected themselves in this way but the shock of events in the north, the massacre of the army and the killing of the Governor, persuaded local authorities to spend a good deal of cash and effort to make their communities safer. Imperial permission was required for the building of town walls and it was not easy to obtain. Always suspicious of allowing independent strength of any kind, emperors were traditionally reluctant – but it seems that there was a general recognition that Britannia was now vulnerable.

The central difficulty for Rome was that the province was not very Roman. In Spain and France a thoroughly Romanised society had been created. Large cities had grown up and Latin widely adopted. In Britain it was very different. Out of a population of 2 or 3 million, only 10 per cent lived in the hundred or so small towns, and perhaps 50,000 on villa estates in the countryside. When the army and its dependants, and the villagers in the vici, are added, the total of those who might reasonably be called Romano-British makes up a fifth, or at best a quarter, of the whole population. The Celtic speech community was overwhelmingly dominant, and Latin remained the langauge of authority, the towns and the army – the apparatus of colonisation. In Britain Rome simply did not catch on.

Against a background of that sort of cultural arithmetic, the continuing struggles of native kings appear in a different light. Ulpius Marcellus and the other governors who dealt with regular British rebellions may well have believed that a densely populated and largely Celtic countryside would rise to support insurgents if they looked like succeeding.

Meanwhile Commodus became crazier and crazier. In 182 he appointed the Praetorian Prefect, Perennis, as, in effect, his prime minister, handing over government almost entirely so that he could concentrate on ever more exotic and cruel forms of debauchery. In an ill-judged attempt to widen political power beyond the Senate, Perennis decreed that legionary commanders would now be drawn from the equestrian order, one rank below. At a stroke he removed a key senatorial prerogative – the command of the army. There was uproar, especially in Britain. Led by young aristocrats who saw their careers thwarted, the army mutinied. The British legions wanted to protest directly against the changes and they sent a deputation of no less than 1,500 soldiers to Rome to put their case directly to Commodus. Astonishingly, they proved persuasive and Perennis was removed from office.

To try to settle Britain down, Commodus sent Helvetius Pertinax, a seasoned soldier who knew the province well, having served two tours of duty there. After some initial success in defusing more mutiny and squashing a plot to promote himself as a rival to Commodus, Pertinax found that the situation began to deteriorate. A legion rebelled and attacked Pertinax and his amici, his bodyguard. Only the Governor survived, having been left for dead. His revenge was swift and severe but it only served to foment further disaffection, and Pertinax was forced to ask Rome to relieve him of his command. The discipline considered so important by Hadrian was breaking down.

Commodus’ behaviour did nothing to settle the gathering chaos. His madness, random cruelties and perversions were beginning to overwhelm the government of the Empire. Believing that he was the reincarnation of the god Hercules, he announced that he would appear publicly in the arena to demonstrate his divinity. As Consul and Hercules, at the same time, Commodus planned to fight as a gladiator and, of course, emerge entirely invincible. His assassination was arranged immediately.

Already Prefect of the City of Rome, Pertinax was proclaimed Emperor. But more chaos engulfed the Empire. When he attempted to bring the Praetorian Guard under closer and more direct control, Pertinax was murdered on the orders of the Prefect. The throne was then auctioned to the highest bidder and the Praetorians sold it to a fabulously wealthy and very foolish senator, Didius Julianus. No more durable than his predecessor, he was quickly removed as the frontier legions revolted, proclaiming no less than three candidates for the purple. Clodius Albinus, Governor of Britannia, claimed the throne at the same time as Septimius Severus in Pannonia (modern Hungary) and Niger Pescennius in Syria. A deadly game of deception and double-cross followed. Severus offered Albinus the title of Caesar, something reminiscent of Verus’ role under Marcus Aurelius, while he marched east to confront Niger. After an emphatic victory in Syria (ending with the pursuit and murder of Niger), Severus then reneged on his promises to Albinus. A decisive battle became inevitable. With the British legions behind him and others from Spain and Gaul, Albinus met Severus’ army at Lyons in 197. He lost.

How much of the British garrison also fought on the losing side that day is difficult to work out. Widespread changes of units in the auxiliary forts and on Hadrian’s Wall have been taken to mean that the likes of the Tungrians followed Albinus on his European adventures and were removed after his defeat. But there is no certain information. What was definite was Severus’ mastery of the Roman Empire, a mastery he would soon extend to the mutinous province of Britain.

He was the first African to sit on the imperial throne. Born in Lepcis Magna, now in modern Libya and one of the most substantial and impressive Roman sites anywhere, Severus was one of a growing number of powerful African senators. Clodius Albinus, from Tunisia, was another. Fertile, reliably productive and easily accessible by sea, the north African coastline had become wealthy and, by the end of the second century AD, its leading citizens were increasingly politically active. Severus’ rise had been steady, unspectacular. By the time he defeated Albinus at Lyons, he was fifty-one and beginning to become what the Historia Augusta described as crippled in the feet. Towards the end of his life, his gout forced the Emperor into the indignity of being carried on a litter. It is said that the pain from gout is especially acute, and it cannot have made the already exhausting business of establishing his authority any easier for Severus. Thought harsh by the Historia Augusta, he was probably moved to irritation more often than most by the growing pain in his feet.

From the very outset of his reign the new Emperor carefully cultivated the loyalty of the army. He awarded soldiers their first pay rise since the days of Domitian – a long wait, almost exactly a century. Three new legions were recruited in the east, the I, II and III Parthicae. And, perhaps most popular, Severus at last permitted soldiers to marry legally.

Retaining a healthy suspicion of factionalism in Rome, he took immediate radical action. After removing the uncontrollable Praetorians entirely and replacing them with a much larger and much more loyal bodyguard drawn from the Danube legions, Severus also augmented the Vigiles, the urban cohorts who policed Rome. He stationed one of the new legions in Italy, treating it like any other province, and he further insisted that the II Parthica be based at Albinum, only 30 kilometres from Rome. When he lay on his death bed in York in 212, the old emperor gave his sons some simple advice. Having risen and prospered with the support of the army, he told them to give money to the soldiers and ignore the rest.

Conquest also bound the legions to Severus and as soon as he had established himself, he led them on campaign in the east. The Parthian Empire was believed to be vulnerable, and by the end of 198 two new provinces had been created in Mesopotamia and Osrhoene, the first significant additions to the Empire since the time of Trajan. In North Africa an extensive frontier system of roads and forts pushed the barbarians further south and another new province was set up in Numidia, what is now eastern Algeria.

PARTHIA

Rome’s greatest rivals in the east are usually treated much in the same way as the barbarians of the north. Given very little historical personality, they seem like a buffer, a monolithic enemy. The Parthians took over Iran and Mesopotamia in the first century BC, and in 53 BC delivered a massive blow by defeating and killing Crassus at Carrhae, massacring many legionaries. Parthia appears to have been a federation of vassal kingdoms governed by a dynasty which originated amongst the Parni, semi-nomads from the north. Their capital place was at Ctesiphon on the lower Tigris and their armies boasted a heavy, armoured cavalry and squadrons of deadly horse-archers. Zoroastrianism seems to have been the state religion, although there was widespread toleration of other cults. Modern Zoroastrians are known as Parsees. The frontier with Rome ebbed and flowed between the two rivers, the Euphrates and the Tigris, until a new dynasty, the Sassanids, pushed further west in the third century, humiliating the Empire with the capture of the Emperor Valerian in 260.

The focus on the Parthian campaigns called for holding action at the farthest end of the Empire, in Britannia. In 197 Virius Lupus, the Governor installed by Severus, was forced to buy peace in the north. A very large bribe was handed over to the kings of the Maeatae and the Caledonii, and in return some sort of treaty was agreed and Roman prisoners returned. These last were probably captured during raids into the province. If Severus wished to prosecute his war against the Parthians in the east, then he had little option but to buy time in the north-west of the Empire.

Dio Cassius had heard of these powerful native kingdoms:

the two most important tribes of the Britons [in the north] are the Caledonians and the Maeatae; the names of all the tribes have been practically absorbed in these. The Maeatae dwell close to the Wall which divides the island into two parts and the Caledonians next to them. Each of the two inhabit rugged hills with swamps between, possessing neither walled places nor towns, but living by pastoral pursuits and by hunting.

The Antonine Wall is the more likely of the two and the Maeatae have left gossamer traces of their ancient name in the hills around it. Five kilometres to the north of Stirling rises the steep rampart of the Ochil Hills, and the most prominent part of the western ridge is known as Dumyat. It means Fort of the Maeatae. Near Denny, looking south at the central sector of the Antonine Wall, is Myot Hill, another stronghold. The name outlasted the Empire and in his Life of Columba, written in the 7th century, St Adomnan describedthe war with the Maeatae. It was fought by the Gaelic-speaking King of Argyll, Aedan macGabrain, and Adomnan remembered that Columba prayed hard for victory for his fellow Gael. His prayers may not have been answered because, in the event, two of Aedan’s sons, Eochaid Find and Arthur, were killed. The same battle is recorded in the Irish Annals of Tighernach, but it is called the battle of Circenn. And Circenn was the name of the later Pictish province of Angus and the Mearns. All of which places the Maeatae where the place-names hint, just to the north of the Antonine Wall and in the Angus glens and coastal plain. The name itself is hard to parse – Maeatae may mean something routine like ‘the Great Ones’. Certainly great enough to extract cash from the coffers of the Roman Empire.

Behind them or next to them was the territory of the Caledonii. The likelihood is that their kings ruled the lands to the north and west of their confederates. Unlike them, their name has endured, and expanded its meaning to include all Scots.

The treaty concluded by Virius Lupus did not last long. By 207 the warbands of the Maeatae and the Caledonii were mustering once more for raids to the south. The new Governor of Britannia, Alfenus Senecio, wrote an urgent despatch:

. . . the barbarians had risen and were overrunning the country, carrying off booty and causing great destruction . . . for effective defence more troops or the presence of the Emperor was necessary.

This request was not made lightly. No governor wanted the massive burden of an imperial visit and all the disruption – and scrutiny – involved. The Maeatae and the Caledonii had probably made serious inroads; their hardy little ponies would have been capable of long distances and would have thought nothing of making 50 miles a day over difficult country. The warbands had probably penetrated deep into the richest part of the province, bypassing or overrunning Hadrian’s Wall, staying clear of the main roads and the legionary fortresses.

Severus made immediate preparations for a campaign in Britain. Some historians have suggested that his main motive was the usual desire for prestige, but it was far too late in Severus’ reign for him to bother about that. Britannia was probably in uproar, and Alfenus Senecio’s request desperate.

The imperial expedition into Scotland and the heartlands of the insurgents was to be primarily focused on the east coast and heavily supported by sea. The fort at South Shields, at the mouth of the Tyne, was converted into a massive supply dump with no less than twenty-three granaries built to store food. It has been calculated that there was enough for an army of 40,000 men for three months in the field. Corbridge also saw new building, and at Cramond, west of Edinburgh, on the Forth shore, the old Antonine fort was refurbished.

The intended target was Tayside, the territory of the Maeatae. It seems likely that the German fleet, and perhaps even the Danube river flotilla, were brigaded with the Classis Britannica as transports and supply ships. And in 208 Septimius Severus, the imperial family, a senatorial council, most of the enlarged Praetorian Guard, detachments from several legions and the whole administrative apparatus of the Roman Empire arrived in Britain. And the dark and shifting shadows of palace intrigue came with it.

Based at the legionary fortress at York, Severus planned his campaign in the north. Contemporary historians recorded that part of the reason he came to fight in Britannia was to toughen up his sons, Caracalla and Geta, and to remove them from the fevered and unhealthy atmosphere of Rome. Probably because he did not trust him, the old Emperor decided to take Caracalla with him on campaign – where he could keep a wary eye on him. Severus’ judgement turned out to be sound. Geta was set in charge of the province (what happened to Alfenus Senecio is not recorded) so that he could gain experience of government. The Empress, Julia Domna, was also at the fortress in York. Having regained her influence at court, she was to continue to play a motive role in imperial politics for the following two reigns.

When the huge Roman army rumbled northward, most of them marching up Dere Street, some on transports, all of them supplied and shadowed by the fleet, the kings of the Maeatae and the Caledonii will have shuddered. Severus’ sprawling strike force was strung out on the metalled road for at least 8 kilometres, well screened by cavalry, the eagles flashing in the summer sun, the menacing thud of marching feet and rattling carts audible for miles. The army was so huge that it took four days to cover the 60 kilometres between Newstead and Inveresk on the Forth. When the mensores were pegging out the ramparts of the next camp, soldiers had only just passed through the gates of the one before. As they halted each afternoon, enormous temporary camps were dug, the largest at 165 acres. Thousands of cooking fires will have lit the night sky. Scouts no doubt reported back and the native kings sent envoys to offer peace. The Emperor brushed them aside, and the might of Rome swept on into Scotland.

At Carpow, on the southern shore of the upper Tay estuary, probably in the territory of the friendly Venicones, Severus’ legions built a base. It could be supplied and reinforced by sea, and it looked west into the glens of the Maeatae; Strathearn, Strathtay and Glenalmond all lay within marching distance. Across the estuary the fertile fields of the Carse of Gowrie and Strathmore to the north may also have been readily reached. A coin minted in 208 shows troops crossing a bridge, and the historian Herodian had heard of plans made by Severus’ commanders which anticipated water barriers to the advance. Perhaps the Tay was crossed by a bridge of pontoons from Carpow.

In any event no pitched battles or decisive victories were recorded for the great army. Archaeologists believe that they penetrated far to the north, to the Moray coastlands, maybe as far as Agricola, but it seems that the Maeatae and the Caledonii would not be drawn into a set-piece. They probably scorched the earth and forced Severus and Caracalla to rely completely on their fleet. In turn that shortened any campaigning season. Instead of glory in battle, the emphasis may have to have been on great engineering projects – like a bridge across the Tay – how the Roman army could tame the landscape, and its inhabitants, with technology. A treaty appears to have been made and territory ceded. In 210 more coins were struck and victory in Britain was celebrated.

PALATINE PALACE

The English word ‘palace’ derives from the name of the hill in Rome where the emperors had theirs. Originally the location of Augustus’ house, it developed quickly into a large imperial compound. Domitian greatly enlarged the Palatine and by the time Septimius Severus became emperor, he was forced to have a platform built on the flanks of the hill and his residence plonked on top. To hide it, Septimius’ architects built an ornamental screen which came to be known as the Septizonium. On another platform, this time on the north-western side, Elagabalus had his temple to the Syrian sun-god erected. The Palatine was not defended by walls but closely patrolled by the fearsome Praetorians. And it was handy for the Circus Maximus, the arena used for chariot racing. There was a passageway from the palace leading directly to the imperial box.

The treaties held for a year. In the summer of 211 the kings of the Maeatae and the Caledonii mustered their warriors once more and attacked the Roman garrisons. Severus was too ill to lead the legions north, and Caracalla went in his place. This time the strategy was brutally single-minded, nothing less than the annihilation of the warbands and the society which sustained them. But bloodshed seems to have been prevented by a single death. In the fortress at York, Septimius Severus died and Caracalla seems to have broken off his campaign to turn his immediate attention to the succession. In addition to paying the army, the old Emperor had exhorted his sons to act together. But his judgement of Caracalla was good. Seeing his chance with the army under his command on Tayside, he tried to persuade his senior staff to use their men to acclaim him as sole emperor. They would not, and Caracalla was forced to hurry south to York, or possibly London, to confer with his brother, Geta, and his mother, Julia Domna. There might be other claimants from outside the imperial family and it was best to present a united front, for the moment. Severus had intended his sons to succeed him as joint emperors and that was the legal position. At Carpow, detachments of the II Augusta were left, probably a significant presence. They were certainly still on the Tay in 212. An undignified retreat, a complete abandonment of the war in the north, would have made Caracalla look weak.

By 212 he looked strong. Having disposed of his brother, Caracalla reigned as sole emperor and he quickly set about reorganising Britannia. In order to reduce the risk of governors making an attempt on the throne with the backing of the entire garrison of three legions, the province was divided in two. Because it lay further from Rome, the north was renamed Britannia Inferior, or Lower Britain. It included the VI Legion at York and probably the garrison of the Wall. Britannia Superior, or Upper Britain, comprised the south and the legions based at Chester and Caerleon.

Caracalla strengthened the frontier. The fabric of the Wall itself was altered slightly: some turrets in the central sector were abandoned, the curtain repaired, and new, grander bridges crossed the North Tyne at Chesters and the Irthing at Willowford. At milecastles the double gateways to the north had probably been little used by carts and horsemen and they were blocked up and replaced by small postern gates. Elsewehere many forts saw refurbishment and even some new building.

The most striking aspect of Caracalla’s frontier strategy could be seen forward of the Wall. In the west, two outpost forts were reinforced. Netherby, in the Esk Valley north of Carlisle, was known as Castra Exploratorum, the Fort of the Scouts. And it seems that the new emphasis was on patrolling and intelligence gathering. At Bewcastle, north of the Wall at Birdoswald, the fort was essentially better adapted to suit its site. Now it can only be reached by a single-track road and feels as though it stands in the middle of nowhere, with grey and dun-coloured moorland stretching away to the Bewcastle Fells. But in fact the fort lay astride a well-ridden route taken by raiders for millennia. In the sixteenth century Elizabeth I’s government used Roman stone to rebuild what amounted to a small, squat castle to hold a troop of light cavalry. Their purpose was to police the hill trails followed by reivers from Liddesdale down into the Irthing Valley. The Romans probably built Bewcastle with a similar purpose in mind.

In the east, the fort at Bremenium, now High Rochester, commanded Dere Street as it climbed up towards the watershed at the Carter Bar. The reconstructed walls of the fort and three of the original gateways can be seen from considerable distances by those travelling along the line of the Roman road. Between it and the Wall, the fort at Habitancum, now Risingham, was really formidable. A milliary cohort of 1,000 mixed cavalry and infantry was based there along with a unit of scouts and a detachment of spearmen from Raetia, modern Switzerland. The fort is too small to accommodate all of these soldiers, even allowing for below-strength numbers, and it seems likely that some were either outposted elsewhere or that patrolling was constant and undertaken in force, or both. Traces of third-century Roman activity have been found at Jedburgh and at Tweedmouth, south of Berwick.

Caracalla’s strategy worked. The evacuation of Scotland and the thickening of the frontier zone discouraged incursion, and for seventy years the Wall held firm. But successful Roman retrenchment would have been seen differently by the bards of the Maeatae, the Caledonii and the Selgovae. The great Empire had been driven back, the turf wall overrun, and even though they had the effrontery to patrol, the Roman scouts rode through a hostile land. Most of them cowered behind their Wall!

In the reign of Caracalla it became an even more meaningful divide. In 212 he proclaimed that all free men who lived within the bounds of the Empire would become citizens of Rome. Cives had real legal standing, rights which could offer protection and redress. If the miserable Brittunculi of Vindolanda lived on the right side of the Wall, they were now as Roman as the senators who strolled around the Forum. What this meant in practice is of course another matter, and social class and economic clout will have continued to outweigh all other considerations.

ROMAN AFRICA

Money brings power and, by the end of the second century, the economy of Roman North Africa was booming. Septimius Severus was the first – but by no means the last – African to come to great prominence in the Empire. The old Carthaginian empire had drawn its wealth from the fertile fields of what is now modern Tunisia and the green coastal strips to both east and west. Rainfall was regular and irrigation opened up new areas for the cultivation of olives and corn. Only Egypt produced more of the latter. The cities of North Africa prospered, and at El Djem stands a huge amphitheatre. Throughout the entire Empire only the Colosseum in Rome was larger. All of this abundance was protected by geography. In the Atlas Mountains and the Sahara Desert to the south, relatively few people lived and therefore the cost of defending the provincial frontier was nothing like as heavy as in Britain.

Caracalla was murdered in 217 by one of his bodyguards, evidently a man with a personal grievance to settle, and was succeeded by another African emperor, Macrinus, but he managed only a year. The army intervened once more and, loyal to the family of Septimius Severus, the Syrian legions backed Elagabalus. He was only fourteen years old and the hereditary high priest of a local sun-god cult at Emesa. His unlikely promotion to the purple is explained by the shadowy machinations of the old Empress, Julia Domna. Before her marriage, she had been a Syrian princess, and Elagabalus was the son of her great-niece, and almost certainly her puppet.

The cult of Emesa, complete with its sacred black stone, was uplifted and relocated to a temple on the Palatine Hill in Rome. While Elagabalus celebrated strange and sexually exotic rituals in the worship of the sun-god, the Senate were at first puzzled, then embarrassed at the antics of the young Emperor, and finally hostile. He was insisting that his Syrian god be installed as the supreme god of the Empire. And then everyone would have to cavort around like Elagabalus. Symbolic marriages, divorces, remarriages and adoptions followed in rapid sequence, and the young Emperor was finally murdered in 222 and replaced by Severus Alexander. Although he lasted a good deal longer, until his assassination in 235, both his and succeeding emperors’ total dependence on the support of the army began a damaging cycle. Between the death of Alexander and the end of the third century, there were fifteen emperors and many more candidates supported by various legions. The lowest point came in 260 when the Emperor Valerian was defeated and captured by the Persian king, Shapur I. Miserable and humiliated, he died a prisoner.

Insulated and remote, Britannia appeared to be little affected by the seething convulsions of imperial intrigue and infighting. The Wall garrison had been increased by a third, from just under 9,000 men in the reign of Hadrian to 12,000 under Caracalla. And the units began to settle, with few changes and almost certainly more and more local recruitment. After Septimius Severus permitted soldiers to marry, there seems to have been an attempt to house them differently inside the forts. Instead of the old barrack blocks with communal sleeping quarters for each platoon, or contubernium, new chalet-style rooms were built. At Vindolanda, Wallsend, Great Chesters, Risingham and High Rochester, this sort of accommodation has been detected, while at Housesteads the earth rampart backing onto the fort’s walls was removed and new buildings erected.

The vici, the civil settlements, probably also expanded as a result of Severus’ edict, and archaeologists see the third century as a period of some vigour along the Wall. But there appears to have been very little fighting. The continuing turmoil in Rome may have persuaded governors of Lower Britain to buy off the northern kings with more bribery.

Against a background of some security, Carlisle began to develop as an urban centre. Luguvalium, a name incorporating the Celtic god, Lugh, was little more than a fort built by Petilius Cerialis during his campaign of 71 to 74 in the north. After the establishment of the frontier along the line of the Wall, and the fort across the Eden at Stanwix was built, Carlisle slowly began to develop. Its site was attractive. Cerialis had had the fort built on the promontory where the castle now stands, and it was bounded on three sides by rivers and marshy ground but accessible from the south, up a gentle slope. That was where the vicus first developed, and keyhole archaeology has found early wooden buildings around the area of Blackfriars Street.

By the middle of the third century Carlisle had grown sufficiently to merit promotion. Named as the Civitas Carvetiorum, it became the principal urban centre in the lands of the Carvetii, the Deer People, what appears to have been the Eden Valley, North Cumbria and the lower Irthing Valley. Literally meaning a city-state, a civitas was run by the ordo, a council of decuriones, men of property drawn from the region as well as the town: more a county council than a town council. At full complement, the ordo of acivitas had 100 members with property over a certain value but in a relatively less wealthy part of Britannia; not all owned grand houses in Carlisle. Many probably had estates in the landward area.

Based on the ancestral lands of the Carvetii, the civitas was probably run by decurions who were mostly native aristocrats that had become partly Romanised, or at least wished to participate in Roman-style local government. They were expected to endow civic projects personally, just as Roman aristocrats were in the habit of doing, and also they probably paid a fee for the privilege of being a decurion. Two senior and two junior magistrates were elected to administer justice, oversee local tax-gathering and manage civic amenities such as the water supply, planning and road-building.

Most towns in Roman Britain had an engineered water supply, and Carlisle’s must have been solidly built. As late as 685, it was still working. After the ambitious Northumbrian kings had brought the western end of the Wall and Galloway into their power in the seventh century, they appointed reeves to administer their growing royal estates. Carlisle was theirs, and when St Cuthbert came to visit, with the Northumbrian queen, in 685, Bede recorded remarkable Roman survival:

Cuthbert, leaning on his staff, was listening to Wagga the Reeve of Carlisle explaining to the Queen the Roman wall of the city . . . the citizens conducted him around the city walls to see a remarkable Roman fountain that was built into them.

Wagga and his people (still called citizens by Bede) must have maintained whatever aqueduct or pumping or piping systems continued to bring water to this fountain. The walls were intact, and they enclosed a Roman street grid which was still being adhered to. Some time after 698 when a church dedicated to St Cuthbert was built in the city, the east–west alignment was altered so that it fronted onto a Roman street. Much later, a large arched stone building was still standing when the twelfth-century historian William of Malmesbury described its inscription to Mars and Venus.

Archaeologists have found grand relics under modern Carlisle. Stone columns and capitals are kept on display at Tullie House Museum, and hypocaust flooring used under the houses of wealthy citizens has been uncovered. They also displayed their wealth in death.

Since it was illegal to bury the dead inside the walls of Roman towns, the habit was to set up tombs by the roadside. Not only do these allow the line of the roads out of Carlisle to be plotted, but two in particular offer valuable information. A Greek merchant, probably with a specialised trade and who followed the Roman army as a supplier, died in the town. His wife, Septima, left a valedictory verse:

To the spirits of the departed

Flavius Antigonus Papias

a citizen of Greece, lived

60 years more or less, and

gave back to the Fates his

soul lent for that time,

Septima Do . . . set this up.

Later an elaborate tombstone with a sculpture of the deceased was found by the west road passing through Denton Holme. Wearing rich and expensive drapery, the lady sits on a high-backed armchair holding a circular fan and with a pet bird on her lap. A child stands beside her. It is a confident, stylish memorial to a luxurious way of life not seen again in Carlisle until the twentieth century.

It has been argued – convincingly – that Christianity was the only substantial historical legacy of the Roman occupation of southern Britain. One of our greatest saints almost certainly had his origins in the countryside around Hadrian’s Wall and the civitas of Carlisle and its early Christian community. St Patrick was born in Britannia, not Ireland, and in his writings he left several tantalising snippets of information about himself and his early life. These have been brilliantly analysed by Professor Charles Thomas.

All of the episodes in Patrick’s young life point to origins in the north-west of Britannia. He was abducted by Irish slavers and sold into captivity in Northern Ireland, becaming a shepherd for six years. When Patrick escaped, he eventually returned to Britannia to train as a priest before finally returning to Ireland to begin his great mission of conversion and thereby leave an indelible mark on history.

Patrick wrote that his father was Calpurnius, a deacon in the church and a decurion in the ordo of a civitas. He was wealthy enough to have both male and female servants and to own a villula, a small estate in the countryside. According to Patrick, it lay near Vicus Bannavem Taberniae, a place-name which appears slightly corrupted. Vicus is simple enough, a civil settlement outside the walls of a fort. But Calpurnius was also the decurion of a city – and where in Britannia was there a city close to forts with vici? If the text is taken to read Vicus Bannaventa Berniae, then the location comes slowly into focus. Banna is the distinctive name for the Wall fort at Birdoswald, and venta is a market, the market held at the vicus. So – was Calpurnius’ small estate near Birdoswald? The last element in the puzzle of the place-name is Berniae, and that appears to be a transliteration of an Old Welsh word bern for a defile or a narrow pass. Near Birdoswald, there are several candidates. Perhaps the river-cliff at Greenhead, perhaps the steep banks of the burn at Poltross? In any event, the fort, and by extension the villula, are near Carlisle, the city where Calpurnius may have been a decurion.

Patrick also wrote that his grandfather, Potitus, was an ordained priest. This implies an organised church with a hierarchy (able to conduct ordinations) around 360, and this in turn strongly suggests links with a developed town. All of the elements seem to fit. Charles Thomas is certain that Patrick was born and raised in Carlisle and the countryside through which Hadrian’s Wall runs. Perhaps the conversion of Ireland ought to be seen as part of the Wall’s legacy.

Carlisle’s new status as a civitas was first mentioned during the reign of the breakaway emperor, Postumus. After the shocking affront of Valerian’s capture in the east, rebellions at the opposite end of the Empire saw the provinces of Spain, Gaul, Germany and Britain form what was known as the Gallic Empire. Under Postumus, it lasted little more than a decade before the Empire was put back together again by the energetic Aurelian. Britain appears to have been unaffected by these continental convulsions. But reform was in the air. After his accession in 285, Diocletian began the process of dividing the Empire in four. There were to be two senior emperors, the Augusti, including himself, and two junior figures, the Caesares. As Roman Europe settled down into this new pattern of power, Britain rebelled.

ZENOBIA

When the antiquary William Stukely first gazed upon Housesteads Fort in 1775, he proclaimed it the Tadmor of Britain: a reference which escapes most modern readers. Stukely used it as a comparison with the recent discovery of the magnificent city of Palmyra in the Syrian desert. By 271 the Empire of this remarkable place had all but eclipsed Rome in the east. Under their warrior-empress, Zenobia, the Palmyrenes had conquered Egypt, Arabia, Palestine, Syria, Mesopotamia and much of Asia Minor. The Historia Augusta was dazzled: . . . in the manner of a Roman emperor, she came forth to public assemblies, wearing a helmet and girt with a purple fillet . . . Her face was dark . . . her eyes were black and powerful . . . her spirit divinely great, her beauty incredible. The oasis city of Palmyra had ingathered fabulous wealth as a consequence of its position astride several long-distance trade-routes from the east to the cities of the Syrian coast. Zenobia’s amazing empire lasted only a year. The Emperor Aurelian defeated the Palmyrenes in three bloody battles and successfully besieged the city. Caught while fleeing to Persia, Zenobia was brought to Rome and rode a camel in Aurelian’s triumph. Instead of the dark horrors of the Mamertine Prison, she retired to a villa near the city – and no doubt thought often on her year as Empress of the East.

The English Channel had become badly infested by pirates and barbarian raiders. Having been successful against them, the admiral of the Classis Britannica, Carausius, was thought to have done rather too well out of the proceeds of captured booty. Diocletian’s co-emperor, Maximian, summarily condemned Carausius to death. With nothing to lose, the admiral declared himself Emperor of Britain. Not as overblown as it sounds, for the south of the main island had been redivided into four provinces (from the Wall southwards, these were Britannia Secunda, Flavia Caesariensis, Maxima Caesariensis in the south-east and Britannia Prima in the south-west); Carausius also controlled parts of northern Gaul. In 293 Maximian’s junior imperial partner, Constantius Chlorus, drove Carausius out of Gaul and, crucially, retook the base of the Classis Britannica at Boulogne. A secondary coup d’état then occurred when Carausius was assassinated by the unlikely figure of his Financial Secretary, Allectus. Archaeologists have discovered the remains of an imperial palace he had time to build in London before a Roman army invaded Britain. At a battle near Silchester, Allectus was killed, and the short-lived British Empire died with him.

Discord and weakness in Rome once again stirred predatory instincts in the north. After four generations of peace, of generous subsidies, of basking in ancient glories – what their kings probably saw as a victory against the vast army of Severus and his son – the Maeatae and the Caledonii once more talked of war and of great raids in the south. In 296 warbands crossed the Wall, or sailed around it, and they were recorded riding far to the south, even attacking the legionary fortress at Chester. They must have been numerous, well organised and confident. And they were given a new name. Writing of the victories of Constantius Chlorus, a historian noted a fearsome people he called the Picts.

Picti, meaning ‘the painted or tattooed people’, was probably a soldiers’ nickname, perhaps coined in 296 by those garrisons on Hadrian’s Wall who had not seen the northern warbands before. It stands in a tradition of noms de guerre which called the Saxonsafter a short-bladed knife they carried, the Franks, whose name means ‘the wreckers’, or the Vikings who did exactly that, dodged in and out of creeks, or viks.

The nickname stuck and almost certainly applied to a federation of Maeatae, Caledonii, a people called the Verturiones (probably from Strathearn and Menteith) and other groups. These came to include the Scots from the Argyll coastlands and a people called the Atecotti. This name translates simply as ‘the Old Peoples’ and they may have originated from the Hebrides. St Jerome believed the Atecotti to be aboriginal savages and claims to have witnessed them practicing cannibalism.

When the first raiders broke through into Britannia, they encountered much less resistance than in former times. By the beginning of the fourth century the Roman army had reorganised. Gone were the old distinctions between legions and auxiliaries. Now the frontier garrison was known as the limitanei, and under imperial command was a mobile field army called the comitatenses. Control over the men on the Wall was removed from the Governor of Britannia and given to a soldier known as the Dux Britanniarum, the Duke of the Britains (meaning all four provinces).

The accent was firmly on defence in Britain, and while strong archaeological evidence found at Cramond and Carpow suggests that Constantius led an expedition to Scotland to quieten the Picts in 306, that was the exception. Numbers seem also to have declined markedly; units with the names of the old legions attached only had 1,000 or so, compared with 4,800 or 5,000, and cavalry troops were 150 rather than 460 or 500. And soldiers were conscripts, not volunteers, with the sons of veterans now being compelled to join the army. The decline was reflected in the occasional fraud, when commanders continued to claim dead men’s pay by failing to report casualties in their units. On paper the Roman army sounded a great deal more formidable than it was in the field.

One solution was the co-option of mercenaries, and by the fourth century these were mainly German warriors. Some estimates put the proportion at a quarter of the total strength in the western Empire. When Constantius returned to York in 306, like Severus a century before, he died and was immediately succeeded by his son. Diocletian’s power-sharing arrangement remained in force, but when Constantine was proclaimed at the legionary fortress, he needed solid support from his father’s army, and prominent was a German mercenary king, Crocus. He commanded a band of Alemanni, men who had presumably fought against the Picts in Scotland.

Along the Wall repairs and refurbishment took place at the beginning of the fourth century. Work is recorded at both Birdoswald and Housesteads forts, and a new garrison arrived at South Shields. The Numerus Barcariorum Tigrisensium, or the Tigris Bargemen, originated in modern Iraq, but by the time they reached the mouth of the Tyne, they may have lost that specific function. But they probably still recruited in the Middle East because they gave the fort a new name. It was called Arbeia, the fort of the Arabs.

It took Constantine the Great until 324 to establish himself as sole emperor and his long reign produced two significant shifts of policy: the transfer of the focus of the Empire from Italy and Rome to the east and Constantinople, and the official adoption of Christianity as the state religion. The new religion had probably been imported into Britain by soldiers, and in the third century martyrs were killed at St Albans and at Caerleon. The beautiful church plate found at Water Newton was made in the early fourth century and Christian-inspired mosaics at the villas at Hinton St Mary and Frampton a little later. These finds were made in the south, but on the Wall the signs of the new faith are sparse and of uncertain date. Chapel-like buildings have been identified at Vindolanda and Housesteads, but they were probably built after the end of Roman Britain.

FORGERIES

Clay moulds for making counterfeit Roman coins have been found both north and south of the Wall. This shady trade was made possible by the continuing debasement of the currency of the Empire. By the middle of the third century the imperial budget was running at 225 million denarii per annum and hundreds of millions of coins were being struck to feed it. There was not enough silver in the Empire to make what was needed and consequently coins were primarily minted from base metals. The Emperor Diocletian attempted to check the runaway inflation by issuing an Edict on Prices in 301. It listed cereals, beer, meat and other commodities and attached standard measures and prices to each, as well as the rates of pay for different sorts of worker. Like all attempts at a prices-and-incomes policy, it failed immediately. The market corrected the situation with characteristic crudeness. Exchange was based on bullion, silver or gold, no matter what form it came in. The Roman Empire ultimately fell because it ceased to produce sound money and became less and less able to pay for itself.

For the first half of the fourth century, the province appears to have been calm and prosperous. No raiding from the north is recorded for almost forty years after Constantine’s accession at York. But in 342 the long period of peace was broken. The Emperor Constans arrived with detachments of the field army and immediately moved up to the Wall. It seems that Pictish warbands, perhaps in concert with their Scots allies, had attacked the outpost forts at High Rochester, Risingham and Bewcastle. All three were burned. Constans encouraged the areani, a new name for the scouts formerly called exploratores, to take a more actively defensive role, and the most northerly fort, at High Rochester, appears to have been abandoned. A treaty, and probably subsidies, were accepted by the northern kings.

By 360 these arrangements no longer held. The Pictish federation mustered and sent its warbands to raid an area north of the Wall, and perhaps behind it. Four regiments from the field army (including a unit of Batavians) in Europe were sent by the Emperor Julian to deal with the emergency. No details of the campaign have survived, but it was certainly not decisive. More attacks from the Picts, the Scots, the Atecotti – and a new group of barbarian raiders, the Saxons, came in in 365. Serious though they seem to have been, these assaults on Britain were only a prelude.

In 367 the province suffered as it had never done before – even during the rebellion of Boudicca. In a concerted series of incursions, coming from several directions, the Picts, the Scots, the Atecotti, the Franks and the Saxons descended on Britannia. Known as the Conspiratio Barbarica, the Barbarian Conspiracy, it saw the Pictish army pour through the Wall, the Scots and probably the Atecotti invade from the west, across the Irish Sea and the Solway, and the Franks and the Saxons attack the coasts of Gaul and perhaps the south of England. A pre-arranged plan was being put into action. Having communicated well in advance, raised the necessary forces, been in possession of good military intelligence, and worked out a timetable for invasion, the barbarian kings swept Britannia’s defences aside and tore into the province. For two years they burned, looted and killed without check and over a wide area. It was a catastrophe, and a hammer-blow to an ever-weakening western Empire.

In the north there had been betrayal and collusion. The Pictish kings had bribed the Roman areani, the scouts operating forward of the Wall, to report nothing of the preparations for war, and probably to supply crucial intelligence on the state of the defences and troop dispositions. It seems likely that the great conspiracy was co-ordinated and planned in Pictland, an operation of considerable sophistication which modifies the usual image of screaming hordes, for once allowing that savagery could be accompanied by brains.

Once the Pictish army had broken through the Wall (they would not necessarily need the co-operation of the areani to sail around it), they sought out units probably deployed on the east coast, what was known as the Saxon Shore, and defeated them and killed their commander, Nectaridus. Then they turned on the Dux Britanniarum, Fullofaudes, and either killed or neutralised him. It was a triumph, a sweeping and comprehensive victory – complete mastery of the province. All the riches of Britannia, the villas and the prosperous towns, lay at the mercy of the fearsome barbarians.

Roman reaction was at first hesitant. The Frankish and Saxon attacks on the coast of Gaul were probably partly designed as a screen to prevent the continental field army from reaching Britain, and the two-year delay might simply be explained by difficulties in reaching the Channel. But by 369 a capable and experienced soldier arrived at last. Theodosius brought four regiments of the field army and his first action was to proclaim an amnesty for deserters (there had been many) from the limitanei and to swell his ranks with their numbers. Never intent on invasion, only interested in raiding, the barbarian army had broken down into small warbands. Theodosius was able to mop them up or chase them out of the province and restore some sort of order. Longer-term security depended on shoring up the northern frontier and, as soon as the south seemed more settled, Theodosius and his men set out for the Wall.

Because of the treachery of the areani, a new approach was needed. Two years before, in the North African provinces, Theodosius had dealt with the incursions of Berber tribesmen by creating alliances with buffer kingdoms between them and Roman territory. It looks as though he did something similar beyond the Wall.

The Picts, the Scots and the Atecotti all originated beyond the Firth of Forth and the Clyde Valley. Culturally distinct from the kingdoms between the two Walls, the Picts spoke a language which survives only in tiny scraps and elements of place-names. No one can now utter a sentence in Pictish, but philologists have deciphered enough to show that, while it formed part of the P-Celtic family of languages, it was different from the Old Welsh or Brittonic dialects spoken in the south of Scotland. The Scots talked to each other in Q-Celtic, the ancestor of modern Gaelic, and the Atecotti may have lived to the north of them, perhaps in the Hebrides. It is therefore likely that the Damnonii of the Clyde and the Votadini of the Lothians and the Tweed basin had more in common with the peoples who lived south of Hadrian’s Wall than those beyond the Antonine.

The garrison of the Wall had probably relied on native recruitment for many generations. As early as c. 150 the tombstone of a Brigantian who fought in the Roman army was set up on the Antonine Wall. With all of this background in mind, it seems that Theodosius did indeed set up buffer kingdoms beyond Hadrian’s Wall. To replace the duplicitous areani and the dangerously exposed outpost forts, the Britons would have to resist the ferocious Picts, and prevent them from overrunning their own territory in so doing.

The evidence for this change of policy is slight but very intriguing. The earliest genealogies, the kinglists, for the Votadini show strikingly exceptional names around the end of the fourth century. Amongst all of the early and clearly Celtic kings, Aetern, Tacit and Patern Pesrut stand out. All they require is the terminal -us. Pesrut is particularly telling: an Old Welsh epithet which means ‘the man with the red cloak’. Was this a native Roman officer, perhaps from the Wall garrison, put in place by Theodosius? Or a native king given a Roman rank? Elsewhere there are more Celticised Roman names. Early rulers of the Damnonii are listed as Cluim, or Clemens, and Cinhil, or Quintilius, and in Galloway a powerful king was known as Annwn Donadd, or Antonius Donatus.

There was nothing unsual in this enlistment of barbarians in the defence of the Empire, it was happening all along the Rhine and Danube frontier. In return for settlement and a grant of land inside the Empire, and no doubt the payment of subsidies, new peoples and their kings pledged loyalty to the Emperor and helped turn back others who threatened from outside. Two further scraps of evidence suggest that this sort of transaction had been concluded by Theodosius for the territories beyond the Wall. On the summit of the Votadinian hillfort on Traprain Law, in East Lothian, a large hoard of late-fourth-century Roman silver was discovered. Much of it had been cut up from cups and bowls and folded over for easier handling and transport. Probably collected from wealthy Romano-British aristocrats, some of it carried Christian images and symbols. In an age when currency had become debased, and with a miniscule content of precious metals, the Traprain hoard looks like bullion. It may represent the fruits of a successful raid, but a much more likely explanation is that it was a payment, a Roman subsidy for an ally prepared to fight in defence of the Empire.

Theodosius may have been the creator of a fifth province in Britain. Named after the Emperor Valentinian, its location is uncertain – but what would have been the purpose of carving out yet another province from the four which already existed? Valentia was probably the new Roman name for southern Scotland, the territory between the two Walls controlled by the kingdoms of the Damnonii and the Votadini, and ruled by the likes of Cluim/Clemens and Paternus Pesrut. Its elevation was a way of bringing the northern allies inside the Empire, conferring romanitas (and perhaps their names) on them, and also incidentally celebrating a notable victory over the barbarians.

KING’S RANSOM

In a field near Kelso in the Scottish Borders hundreds of Roman coins have been found. Often ploughed, the King’s Haugh reveals more each winter through careful metal detection. The coins seem to be scattered, not part of a hoard. But they do not amount to a King’s ransom. Most are from the fourth century and are not silver-based, or gold, but small radiates struck from bronze. Far from being treasure, they are the small change of a money economy, only worth anything as a means of exchange. But Kelso is more than 70 kilometres north of Hadrian’s Wall, in the centre of the ancient territory of the Votadini. Was the Tweed Valley briefly inside the bounds of the Empire, in the province of Valentia? Many of the bronze radiates come from the period after AD 369 and Theodosius’ reorganisation of the north. The heaviest concentration is from the House of Valentinian (364–378) and there are more from the reigns of Arcadius and Honorius, shading into the fifth century. The find spot is very suggestive. At the confluence of the Teviot and the Tweed, the King’s Haugh is commanded by the remarkable, and unexcavated, fortress of Roxburgh Castle. Famous in medieval times, it certainly had a longer history before then.

If the Picts and the Scots were discouraged by these new alliances, they had only to take to their ships to reach the tempting plunder of the south. But, despite their best efforts, they did not bring about the end of Roman Britain. Under pressure from the kingdoms to the east, especially the Huns, the Goths were pressing hard on the European frontier along the Danube. Valens, the eastern Emperor, allowed one group to settle in the province of Moesia, modern Bulgaria. But in 378 they rebelled after some harsh treatment and, at Adrianople, the Gothic army inflicted a crushing defeat and Valens was killed in the fighting. It was a turning moment. After Adrianople the Empire began to shudder and, in the west, to shake itself to pieces.

Meanwhile Britannia appeared to have rallied once Theodosius’ measures had been put in place. The Pictish threat persisted, but it was contained by a dynamic new Duke of the Britains. A Spanish officer, known as Magnus Maximus, assembled an army and led it to victory in the north in 382. So successful was the campaign and so warm the glow of its prestige that Maximus was encouraged to make a bid for the Empire. Crossing into Europe, he took Spain and Gaul under his control and he ruled in the west until 388. Brought to battle at Aquileia in northern Italy by the legitimate emperor, Theodosius I (son of the saviour of Britain in 369), Maximus was defeated, captured and executed.

There were many usurpers at the end of the fourth century and their struggles for power weakened the Roman Empire just at the point when external pressures threatened as never before. But Magnus Maximus was remembered in Britain, and indeed his memory still lives. A sixth-century poem in Old Welsh, ‘The Dream of Macsen Wledig’, recalls his triumphs and it is likely to have been composed by the bards of the kingdoms of southern Scotland, known as Yr Hen Ogledd, the Old North; the courts of these kingdoms were the fount of much of the earliest Welsh literature. The survival of ‘The Dream of Maximus the General’ is perhaps understood by the success against the Picts of a native army raised in the Old North by a competent and ambitious Roman commander. Perhaps some of the warriors of the Damnonii and the Votadini went to Europe with Macsen on his quest for empire, to fulfil his dream.

THE NOTITIA DIGNITATUM

Compiled during the fourth century, this fascinating document, loosely translated as ‘The Ascertaining of Ranks’, had as its primary purpose the setting down of the chain of military command in North Britain (and elsewhere) by listing officers, staffs and the units under their control. At face value it is impressive. Eleven regiments of cavalry and infantry appear to be based in Durham and Yorkshire and three others west of the Pennines, and then the entire garrison of Hadrian’s Wall is set down with their forts in the correct geographical order. The overall impression is of continuity; some of the same units appear to have been soldiering on the Wall for almost three centuries. But what appears on paper almost certainly had no more than a nominal existence on the ground – no more than a shadow of ancient power.

Theodosius’ young son and heir, Honorius, was nominally Emperor after 395, but his armies in the west in reality lay under the command of Stilicho the Vandal. The son of a barbarian soldier, he proved adroit in maintaining the balancing act which imperial government had become. Stilicho withdrew troops from the British garrison to plug gaps and shore up weaknesses in Europe, and it must be significant that the large-scale import of coins into Britain ceased in 402. This was likely cash to pay the army and the imperial administration, and after that date there may have been little left of either.

In the winter of 406 nature took a decisive hand. The Rhine froze over and many thousands of barbarians flooded into the Empire. Vandals, Alans and Suebi rampaged into the province of Gaul, looting, killing and causing chaos. Three years later their warbands crossed the Pyrenees into Spain and eventually the Strait of Gibraltar. After 429 the remarkable Vandal kingdom of North Africa came into being.

Britain found itself ever more embroiled in the ferment of imperial ambition. In 406 no fewer than three usurper emperors attempted coups with support from the province. In 407 more troops were withdrawn, and the following year, as if prompted, a series of serious barbarian attacks caused great damage. This time the Romano-British themselves, probably with the help of the kingdoms in the north, rebelled and expelled the representatives of the most recent usurper, Constantine III. Here is a concise account from the historian Zosimus:

The barbarians across the Rhine attacked everywhere with all their power, and brought the inhabitants of Britain and some of the nations of Gaul to the point of revolting from Roman rule and living on their own, no longer obedient to Roman laws. The Britons took up arms and, braving danger for their own independence, freed their cities from the barbarians threatening them; and all Armorica and the other provinces of Gaul copied the British example and freed themselves in the same way, expelling their Roman governors and establishing their own administration as best they could.

Led by Romano-British aristocrats, backed by the warriors of the Old North and augmented by the Wall garrison (such as it was), the rebellion against the Empire appears to mark an end-point. The conventional signal for the passing of Britannia is usually seen as around 410 when a letter written by the chancery of Honorius advised the cities of the province to look to their own defences. It seems as though they were doing it anyway.