Dinner on the Stone Road

Claudia Severa to her Lepidina greetings. On the third day before the Ides of September, sister, for the day of the celebration of my birthday, I give you a warm invitation to make sure that you come to us, to make the day more enjoyable for me by your arrival . . . Give my greetings to your Cerialis. My Aelius and my little son send him their greetings.

I shall expect you, sister. Farewell, sister, my dearest soul, as I hope to prosper, and hail. To Sulpicia Lepidina, (wife) of Cerialis, from Severa.

Only the names and the word hail give away the dates of this birthday invitation. It could have been written at almost any period in the last two thousand years of our history. The formalities, the gushing sign-off, the evident warmth and the slightly superior tone might as easily have come from the memsahibs of the British Raj as from a woman living on Rome’s northern frontier some time around the year 100.

Sulpicia Lepidina was the wife of Flavius Cerialis, the commander of the IX Cohort of Batavians stationed at the fort of Vindolanda. It lay on the Stanegate Road, the Stone Road, midway between Newcastle and Carlisle, a mile or so south of the line of what was to become Hadrian’s Wall. The invitation from Claudia Severa was miraculously preserved in the anaerobic mud of Vindolanda and, along with many hundreds of other wooden writing tablets, it was discovered in a series of brilliant excavations directed by Robin Birley, which began in the 1970s and are ongoing. The tablets form what is considered the greatest archaeological treasure ever discovered in Britain. Not gold, not silver, nor precious gemstones, the Vindolanda letters and lists are priceless because they allow us directly into the thoughts of people from the long past. In the cloying, black mud of soaking trenches, Robin Birley found Roman voices.

They whispered stories of remarkable immediacy. From the small change of everyday life – shopping lists, parcels of socks and underpants, schoolroom exercises – to the great affairs of state – a visit from the Governor of Britannia, a petition to the Emperor Hadrian, an assessment of regimental strength – the Vindolanda letters and lists take us into the Roman world of northern England, 2,000 years ago.

Their discovery was the result of a problem. When the Vindolanda Trust began life in 1970, its sole source of income was from entrance fees to the site. Excavation and its unpredictable excitements drew visitors and it was first undertaken outside the western gate of the fort, in the civilian settlement. To get at the Roman layers, Robin Birley and his team had to remove modern field drains. Water then collected in the hollow between the settlement and the long mound covering the western wall of the fort. Sometimes it was so deep that the precious visitors could not get into the fort. Something had to be done, and instead of Roman remains, a drainage ditch was dug.

As the excavators moved south towards the slope and the stream running below, they were forced to dig deeper to achieve sufficient fall for the drainage pipe. And they ran straight into what was first believed to be a Roman rubbish dump. Then the stumps of two timber posts were uncovered and Robin Birley realised that they had come upon the buildings of the early timber fort. It being too late in the year to begin a comprehensive excavation, the trench was quickly backfilled for the winter.

In March 1973, Birley moved his team back into the drainage ditch. He quickly saw that, before each phase of rebuilding, soldiers had flattened the site, covering over demolition with clay or thick turf. This process had effectively sealed each layer, trapping material in an anaerobic, or oxygen-free, environment. All sorts of organic remains came up. The carpets of bracken which had lain on the floors of rooms were at first brightly coloured, brown, green and yellow, just as they had been 2,000 years before. But as they came into contact with the oxygen in the atmosphere, they quickly turned black and also began to smell and rot.

Wood was well preserved, as was leather and other normally rare sorts of finds. In a room of what appeared to be a timber building, some thin and oily shavings were noticed amongst much larger objects, like barrel staves. But because there was such a volume ofmaterial, much of which needed to be conserved urgently, they were ignored. Then one of the excavators picked up two of these shavings which seemed to be stuck together. Separating them, he noticed that the inside surfaces were covered with illegible markings. I had another look, wrote Birley, and thought I must have been dreaming, for the marks appeared to be ink writing. He later commented:

If I have to spend the rest of my life working in dirty, wet trenches, I doubt whether I shall ever again experience the shock and excitement I felt at my first glimpse of ink hieroglyphics on tiny scraps of wood.

Hastily collecting as many of these scraps as they could find, Birley and his assistant raced down to Durham to find a leading expert on ancient handwriting, Richard Wright. When they unpacked the shavings from their moss-filled box, they had turned black. Birley wrote:

We were shattered. But Richard Wright had faith in us, and suggested we contact Miss Alison Rutherford, at the Newcastle University Medical School, to see whether ultra-violet or infra-red photography could reveal the texts again. It took time, but the infra-red photography produced the images, and when we saw them, we realised we were no further forward, because it was impossible to read the scripts. Even Richard Wright had to admit defeat, but one of our Trustees, Professor Barri Jones, put us in touch with Dr Alan Bowman at Manchester University, who was an expert in dealing with papyrus records from Egypt, and with the assistance of Dr David Thomas at DurhamUniversity, the long process of reading the texts began. At the time, no-one realised the magnitude of the task.

But it turned out to be immensely rewarding. As scores of finds grew into hundreds and then thousands, a unique record of Roman life began to form. Life on the northern frontier only decades before the Wall was built.

The writers of all this extraordinary material, it should be made clear, were not in fact Romans, or at least not Italians. Flavius Cerialis was a Batavian, an aristocrat from the peoples who lived on what Tacitus described as the Rhine Island. This was the land lying between the old course of the Rhine and the River Waal, the area around the modern city of Nijmegen in the Netherlands. Batavi means ‘the Better Ones’ and the legacy of the name, a certain superior attitude, often leaps off the page at Vindolanda.

Considered exceptional cavalry warriors, Batavians formed the Imperial Horseguards and, from the reign of Augustus to Nero, a detachment was stationed at Rome. Very unusually, the eight cohorts and one purely cavalry regiment drawn from these people were commanded by their own aristocracy, and Flavius Cerialis may even have been a member of the royal family. Like most commanding officers in the Roman army, the Batavian officers were permitted to bring their wives and families to live with them wherever they were posted. It was not approved of by everyone. The stern Republican virtues of discipline and gritty sacrifice come through in this speech by the historian Tacitus to an old-fashioned senator, Caecina Severus:

An entourage of women involves delays through luxury in peacetime and through panic in war. It turns the Roman army into the likeness of a procession of barbarians. Not only is the female sex weak and unable to bear hardship but, when it has the freedom, it is spiteful, ambitious and greedy for power. They disport themselves amongst the soldiers and have the centurions eating out of their hands.

Out of a total population of only 35,000 or so on the Rhine Island, almost 5,000 Batavians had enlisted in the Roman army. Few able-bodied young men can have been left at home. The reason for this was another unusual arrangement. Along with their southern neighbours, the Tungrians, the Batavians were exempt from taxation in cash or in kind. Instead their ruling families supplied recruits to swell the ranks of Roman auxiliary regiments. Their military prowess must have been impressive for the normally grasping imperial procurators to prefer soldiers to cash or goods.

And they were undoubtedly crack troops. Batavians served with distinction in Britain. As part of the Claudian invasion force, it was almost certainly they who swam the Medway in full armour with their ponies, the decisive tactic in winning the first, and critical, battle. They were also at Mons Graupius with Agricola in 83, fighting alongside the Tungrians in the first rank. Having become embroiled in the ebb and flow of the Year of the Four Emperors, they showed great loyalty – in several directions. Under their commander, an ex-gladiator called Tiberius Claudius Spiculus, the Imperial Horseguards fought for the doomed Nero to the last. Revolt erupted on the Rhine Island in AD 70 and Petilius Cerialis hurried north to suppress it. He negotiated with the alarming Batavian leader, Julius Civilis, who had lost an eye in battle, grown his hair very long and dyed it red. Proclaiming his loyalty to Vespasian, claiming to be a friend of the new emperor – and he probably was, having fought at his side on the Medway and at Maiden Castle – Civilis appears to have extracted a pardon for his people from Cerialis. Some Batavians had not joined the revolt, and it was because of this that the family of Flavius Cerialis were probably granted citizenship in 70. Taking the name of Flavius for the family of Vespasian and Cerialis from Petilius, who had agreed to the rehabilitation of the Batavians, it was almost certainly Flavius Cerialis’ father who first became a Roman citizen.

Family wealth appeared to be untouched by the rebellion. Commands in the Roman army demanded property qualifications, and the rank of prefect of an auxiliary regiment was second only to the senatorial class who led the legions. These equestrians, or knights, had to be worth 400,000 sestertii to qualify as commanders, or prefects. There was real wealth washing around the northern frontier in AD 100. More than twenty men of equestrian rank are mentioned in the Vindolanda letters and lists and, in Britain as a whole, sixty prefects were appointed to lead auxiliary regiments. By contrast, the average legionary earned 1,200 sestertii a year – before deductions.

Flavius Cerialis appears not to have been a career soldier, and Vindolanda and the IX Batavians may have been only his second command, possibly his last. Unlike his friend mentioned in the invitation, Aelius Brocchus, his name disappears from the historical record after his tour of duty on the frontier in Britain. Brocchus’ name is found in Pannonia, modern Hungary, some time after 105.

The surviving letters and lists suggest that Cerialis left the daily grind of military life to his senior officers. Centurions seem to have organised training, duty rosters and manoeuvres. These men were the backbone of the Roman army, many dedicating their entire adult lives to its service. They not only stiffened discipline, usually able to back their commands with an implicit threat of violence, they also added a deep reservoir of experience. Most centurions did not retire after twenty-five years of soldiering. Several appear on inscriptions on both Hadrian’s Wall and the Antonine Wall, and there were cases of men still in active service at the age of 80. Like many young officers who relied heavily on their sergeants in the modern British army, Cerialis would have been wise to listen to the advice of these gnarled old veterans.


The Romans were very interested in money. Although coinage was invented by the Mesopotamians and developed by the Egyptians and Greeks, the Romans were the first to create an economy based purely on money. Around 300 BC their first coins were closely related to livestock; the Latin word for money is pecunia and it is derived from pecus for cattle, while denarius meant sixteen asses’ worth. But once coins acquired an intrinsic value and began to be mass-produced in many centres, money drove imperial expansion as hard as political ambition. In 61 BC, before his conquest of Gaul, Julius Caesar owed his bankers 25 million denarii, but after he had subdued and taxed what is now modern France, he made a fortune many times larger than his debts. Caesar’s nephew, the first Emperor Augustus, used the expansion of the Empire to make himself by far the richest man the world had yet known, or at least the first whose immense wealth was quantifiable. Equivalent values for currency are very difficult to measure but relative values are useful to know. A silver coin known as the denarius was worth four sestertii and there were four asses to each sertertius. Measures were also standard and the most common in the Vindolanda lists is the modius. It was 8.62 litres.

If most military matters were of little interest to him, some administration did engage the fort commander, and twelve applications made to him for leave were found in the black mud: I ask you, Lord Cerialis, that you hold me worthy for you to grant me leave.Another soldier supplies more information. For his period of leave, he wants to make for the bright lights of downtown Corbridge, only a few miles to the east. Corbridge was growing into a substantial Roman town, and the stumps of the massive columns on its main street suggest an urban substance not yet seen elsewhere along the Stanegate. Yet another man requests time off so that I can buy something. Perhaps it was a gift for family or friends, perhaps a night on the tiles and a prostitute. Before these applications were found, it was assumed by historians that the Roman army granted leave, but at Vindolanda (and almost certainly elsewhere), it seems to have been normal practice.

Like almost all aristocrats in almost all periods, Flavius Cerialis loved hunting. In a letter to Aelius Brocchus, he pleaded: if you love me, brother, send me hunting nets. They were used to catch everything from a charging wild boar to the sort of small songbirds whose consumption (especially in modern Italy where uccellini are a delicacy) appalls the tender-hearted British. Like several local ducal families until very recently, Cerialis also kept hounds to hunt in north Northumberland. Perhaps like the men and women in pink and black jackets, he rode to them. Batavian culture was certainly equestrian and there can be no doubt that Cerialis would have been schooled into an accomplished horseman. Ancient hunting methods lasted a long time, well into the nineteenth century. Hounds were of two sorts. Sight hounds (vertragi in Latin) were bred and trained for speed, able to run down a quarry such as deer over long distances. Greyhounds and wolfhounds are modern versions – although mastiffs went after wolves, their spiked and studded collars not a matter of show but designed to protect their necks from vulpine fangs. Scent hounds (segosi) picked up a trail, like bloodhounds, and led huntsmen to their prey. Sometimes an animal might have been wounded by spears or arrows, but not brought down, and scent hounds would follow the trail of gore until the weakening creature slowed and was at last found. Hunting dogs were a famous British export to the rest of the Roman Empire.


Insults rarely stand the test of time. Most of the best were spoken, spontaneous and probably shouted, and others contain contemporary references which have lost their force. But Cicero, the great Roman politician, lawyer and orator, was a past master. Because some of his best insults were launched in the law courts, they were written down – and he did not hold back when attacking his opponents: You were such a moron that throughout your speech you were at war with yourself, firing out statements which were not just inconsistent, but which were utterly devoid of any coherence or logic, to the point where your adversary in battle stopped being me and became yourself.

Aristocrats playing at being soldiers were not spared: These weakling softy-boys . . . how in the hell do they expect to survive the frosts and snows of the Apennines? . . . Unless they imagine that they are better armed against winter on account of their expertise in dancing naked at banquets.

Nor was a political opponent: No-one can say whether he spent more time drinking, vomiting or relieving himself.

These young aristocrats were also intensely competitive in the hunting field and, after a few cups of wine before dinner, they may have become boisterous in their boastfulness. In one amusing record, it is possible to hear more than a whisper of triumph. Gaius Tetius Veturius Micianus, who commanded a cavalry regiment at Binchester, near Bishop Auckland, was so mightily chuffed with a kill that he set up an altar to give thanks. It was dedicated to the god of the woods and the fields, Silvanus the unconquered, for the capture of a boar of exceptionally fine appearance, which many of my predecessors have been unable to bag. It sounds like the boar had evaded capture for some time and, like many big salmon or pike, was known to those who hunted it.

The Romans may well have also hunted in the sett. Beaters would drive through cover, as they do now to put up grouse and pheasants in front of guns, and, in a semi-circular sweep, push wild boar, deer and smaller game towards a defile, or a geographical feature otherwise suitable, where hunstmen waited with spears or arrows, or even nets. The hunting horn is ancient, used as a means of communication when a hunting field became extended or confused, or lost, and it may be that Flavius Cerialis’ venatores, or hunters, used a version of the cornu. This was a parade-ground and battlefield horn which could blow at least seventeen notes and was as useful as a modern bugle to give commands. The yelping of excited dogs, the pluming breath of panting horses in the frosty air, the piping notes of the horn will have echoed across many a winter morning.

The land around Vindolanda is prime hunting country. High moorland along the line of the Wall descends gradually to the South Tyne Valley with virtually every variety of cover in between. Cerialis and his men celebrated it. Near Crow Hall Farm, almost 3 kilometres south of Vindolanda, a relief of Diana, the goddess of hunting, was found. Like Micianus’ altar, it may have been set up at the scene of a particulary exhilarating kill. There was another shrine to the goddess at the fort and, when he was stationed in Pannonia, Aelius Brocchus set up yet another.

Without doubt these young men hunted for the thrill of the chase, but also to enrich their supper tables. There were few better introductions to a dish at a smart dinner party than I caught this myself. And few better follow-ups than the story of how and when. The rivers and lochs around the fort were fished, swans and geese were snared, ducks, thrushes and other birds netted.

All of this food was prepared at the commander’s house at Vindolanda by slaves in the kitchen. Astonishingly, buried in the bracken and rushes on the kitchen floor, an incomplete inventory of the dinner service it was served on was found. Here is the list:

Shallow dishes – 2

Side plates – 5

Vinegar bowls – 3

Egg plates – 3

A platter, or shallow dish

A container

A bronze lamp

Bread baskets – 4

Cups – 2

Bowls or ladles – 2

The Romans dined very differently from us and, to make sense of what seems to be an odd collection of plates and vessels (no cutlery), some explanation is needed. For the people who lived at Vindolanda, and all over the Empire, breakfast and lunch (if available) were no more than snatched snacks: pieces of fruit or cheese, a hunk of bread, leftovers, a cup of water. Cena, or dinner, was the main meal of the day, as it still is in Mediterranean cultures, and in the commanding officer’s house at Vindolanda, it will have been an elaborate affair, especially if guests were staying.

Diners reclined on couches around a central, low table. Before food arrived, there was always a drink. Mulsum is on at least one shopping list found at the fort. It was a mixture of wine and honey, a precursor of the upper-class habit of a glass of sherry before dinner. At banquets, food was often carried in by slaves on a pre-set table and laid down where the reclining diners could reach it.

Otherwise food was plated in the kitchen on large ashets or in round bowls, and put in front of the diners. They took small cuts and quantities and, from their side plates, ate the food with their fingers. In her dinner service, Sulpicia Lepidina lists five such small side plates. Eggs were a favourite for the gustatio, the first course, as were oysters and fish. The ovaria, or egg plates, in the inventory would have been used often. Salad, olives and bread also appeared, and isicia, meatballs, were also often on the menu at dinner parties. Throughout the meal, slaves will have brought in bowls of dipping sauces; two recipes are mentioned in the Vindolanda lists: a garlic paste and a spicy dip simply called conditum.

The mensae primae, literally the first tables, were more substantial. Roast haunch of venison, joints of all sorts, baked fish and game pie (probably Hadrian’s favourite) were all carved and sliced by slaves. Diners took small pieces for their side plates and dipped them in the sauce bowls. The pace of a Roman dinner was slow and steady, different from the modern habit of piling up a single plate with a meal and devouring it. Evidently dinners could go on for many hours.


Neither wine, olive oil nor tomatoes were first made or grown by the Romans, or the Greeks, or in the Mediterranean. Olive oil production began in Mesopotamia, roughly modern Iraq. It was introduced into Italy by the Greeks in the eighth century BC. Homer called it ‘liquid gold’, and winning athletes at the original Olympic Games were presented with it instead of medals. It is a very healthy food and one of the reasons why the peoples of the Mediterranean suffer much less from heart disease. Wine was first made in the area between the Black Sea and the Caspian before its manufacture spread westwards. Some of that eastern heritage is preserved in the occasional word. ‘Shiraz’, for example, is a Persian name for a popular style of red wine. The best vines in ancient Italy were not grown in free-standing rows, as now, or on trellises, but trained to climb up the trunks of trees. Tomatoes are seen as the quintessence of Italian cooking, but the Romans did not eat them. The Spanish first saw tomatoes when they conquered the Aztecs of Mexico, and they brought them to Europe. The early varieties were yellow and that is why they are called pomodoro in Italian: golden apples.

Wine featured in the Vindolanda shopping lists, but it must have been expensive, having been brought so far north. River traffic could reach as far up the Tyne as Corbridge, but amphorae will have been loaded onto carts and shoogled the last few miles to Vindolanda – at some cost, both financial and to its taste and condition on arrival. Most wine came from Gaul and southern Spain, but the lists do mention Massic, an expensive brand from Campania in Italy. Unlike modern purists, the Romans did not hesitate to add things to their wine, and in winter they used an even further-travelled additive. Pepper is recorded at Vindolanda, and along with other exotic additives such as dates and saffron, the Romans used these to made a spiced wine to warm themselves when the bitter west wind blew through the Hexham Gap.

In late summer Sulpicia Lepidina’s kitchen slaves probably gathered wild berries and apples to make a sort of fruit cocktail. Wine was poured over the bilberries, raspberries and brambles and the resulting mixture left to infuse and ferment. The process produced either a sweet drink with bits in it or a runny pudding but, however it was described, it was very alcoholic.

An amphora found at Newstead, near Melrose, has the word vinum scratched on the handle. This meant vintage wine and, from records elsewhere in the Empire, fort commanders appear to have ordered it from vineyards in southern Italy and Sicily. Ordinary soldiers, by contrast, had to be content with posca, a drink made from acetum, sour wine mixed with water. At Vindolanda and no doubt elsewhere, wine lees, the bits left at the bottom, had water added to them to make what seems to have been an acceptable beverage.

Beer appears regularly in the lists and it must have been a staple for the soldiers, a drink they knew well from home in Batavia. Known as cervesa, or Celtic beer, it was brewed from local barley, or bracis. No hops were used and, as a result, Celtic beer did not keep for long. It probably did not need to. One of Vindolanda’s many claims to fame is the name of the first brewer recorded in the north-east. Atrectus began a long and honourable tradition still carried on with distinction along the banks of the Tyne. When the beer ran out, it could be a problem. Cerialis received a plaintive request from one of his cavalry officers: the comrades have no beer, which I ask that you order to be sent. It should be pointed out that beer was drunk as much for its calorific content as its alcoholic.

Flavius Cerialis, Sulpicia Lepidina and their guests dressed for dinner. Another list details the sorts of clothes they wore and, appropriately for the family of a wealthy equestrian, they seem to have been fashionable, of high quality and made by a specialist. What is surprising is the gender of the sender and the receiver of this list of garments. Aelius Brocchus sent clothes for formal wear, some synthesi (matching items, probably in different combinations), scarves, capes or cloaks, a plain tunic, half-belted tunics more appropriate for dining and one or two more utilitarian items. No doubt Flavius Cerialis was glad to have a new wardrobe. Were these two wealthy young aristocrats dandies as well as keen huntsmen? It seems so.

Lepidina was not to be outdone, and a remarkable correlation between the Vindolanda lists and archaeology at the fort demonstrates her awareness of fashion in an eerily vivid way. In the anaerobic mud one of her sandals has been found. The Jimmy Choo or Gucci of its day, the acme of contemporary Roman footwear for the wealthy, and something which would not look out of place in the streets of Milan or Paris today, it is something almost certainly worn at a very smart dinner party. Preserved by the black mud, it is almost complete, showing only a break in a small strap – probably while Lepidina was wearing it. Perhaps she threw it out of the window in fury? Stamped on the sole is the maker’s name, L Aebutius Thales, a shoemaker with a factory in Gaul – serving the luxury market. On the other side, there is enough wear to show that Lepidina wore the sandal for some time before the strap broke. In the Vindolanda lists, her expensive shoes are of course itemised.

Footwear probably belonging to Cerialis has also been found. One pair of shoes in particular seems to reinforce the image of him as a dandy. The design for the uppers is a latticework pattern, and it would only have shown up if Cerialis had worn brightly coloured socks. Can there be any doubt that he did – and Brocchus too?

If the talk at these smart dinner parties was of the chase, of fashions in Rome and of the politics of the wider world, there was sometimes someone whose decorum slipped, lowering the high tone. In a letter which may have been sent to Cerialis, a correspondent warns him about a difficult guest: you know that he is [often?] drunk . . . and kindnesses are ruined by envy. That last reference smacks of long-lost gossip. Other guests were more considerate. In a letter sent to Flavius Genialis, Cerialis’ predecessor at Vindolanda, an invitation is politely declined: . . . that I could not come, for a headache is affecting me very painfully.


The principal garment of a free-born Roman, the toga, was huge; it must have been very warm in a hot Mediterranean summer. Usually made from undyed wool, it was shaped like a semi-circle and measured up to 5.5 metres along the straight edge and 2.75 metres at its widest point. Unlike the kilt, which in some ways it otherwise resembled, it was worn without a clasp or fastening – anywhere. As many Roman statues suggest, toga wearers had to keep their left arm crooked to cope with the volume of cloth. To put it on, one corner was laid at the feet while the straight edge was pulled up the back and over the left shoulder, then across the back, under the right arm, or over it, across the chest and then over the left shoulder again, leaving the other corner to hang down behind the knees. How it caught on is a mystery.

Sulpicia Lepidina was almost certainly a Batavian aristocrat, possibly a princess. Her cognomen suggest that her family were granted Roman citizenship during the brief reign of the Emperor Sulpicius Galba (AD 68–9). Like most Roman brides, she would have been much younger than her husband. Analysts of the letters have calculated that Cerialis was about thirty years old in AD 100, but Lepidina is likely to have been little more than a teenager. In the ancient world, girls as young as 12 were encouraged to marry and begin bearing children. Inasmuch as the literary conventions of the letters allow an interpretation, the relationship between Lepidina and her friend Claudia Severa has the atmosphere of breathless intimacy and intensity associated with teenage girls. Here is another letter sent to Vindolanda:

I, sister, just as I had spoken with you, and promised that I would ask Brocchus, and that I would come to you – I did ask him, and he replied that it is always, wholeheartedly, permitted to me, together with . . . to come to you in whatsoever way I can. There are truly, certain intimate matters which [I long to discuss with you (?). As soon as I know for sure(?)] you will receive my letter, from which you will know what I am going to do . . . I was . . . and I will remain at Briga. Farewell my dearest sister and my most longed-for soul. To Sulpicia Lepidina from Severa, wife of Brocchus.

What is going on? What is this matter needing such urgent and private discussion? Such is the power of the Vindolanda material that the reader, at a distance of 2,000 years, wants to know. Perhaps these young women were very lonely, marooned in a sea of soldiers with little or no female company of the same age. Another letter, only a fragment, from an unnamed woman to the wife of the later Tungrian prefect Priscinus, whispers at a sense of isolation:

[as?] my Lady has done, whereby you console me eloquently, just as a mother would do. For my soul . . . this state of mind . . . [during these(?)] days . . . and I was able to convalesce comfortably. As for you . . . what will you do with your Priscinus?

There yawned an unbridgable social gulf between regimental commanders, their wives and families, and the men they commanded. It would have been unthinkable for Lepidina or Severa to have any social commerce with the men in the camp or any of the women associated with them. That made these letters and the meetings they facilitated all the more important.

Letters were important to ordinary soldiers too. News was hungrily devoured in the forts and contacts kept up as good wishes were passed on the acquaintances in other units. Here is Sollemnis beginning a letter to his errant friend Paris:

So that you may know that I am in good health, which I hope you are in turn, you most irreligious fellow, who haven’t even sent me a single letter – but I think I am behaving in a more civilised way by writing to you!

The substantive part of the letter has been lost, but the sign-off survived:

. . . you are to greet my messmates Diligens from me and Cogitatus and Corinthus, and I ask you to send me the names . . . Farewell dearest brother.

Paris was another Batavian, stationed with the III Cohort at the unidentified fort of Ulucium. It is mentioned several times in the lists and letters and was likely to have been near Vindolanda. Perhaps it was at Newbrough, a few miles to the east.

When Robin Birley unearthed the first letters in 1973, Alan Bowman and David Thomas deciphered the text. Because of what the first letter to be published said, public reaction immediately warmed to these discoveries and a flush of recognition lifted its contents and context out of the academic or abstruse.

. . . I have sent you . . . pairs of socks from Sattua, two pairs of sandals and two pairs of underpants, two pairs of sandals . . . Greet . . . ndes, Elpis, Iu . . . enus, Tetricius and all of your messmates, with whom I pray that you live in the greatest good fortune.

Like the mothers and sweethearts of modern soldiers, someone had sent a welcome parcel to a loved one on a remote and chilly northern posting far from home. The letter conjured up images of knitting socks and scarves by the fireside for our brave boys. In fact it is much more likely that the letter accompanied goods ordered and was sent to Vindolanda by a fellow former soldier. But it was fortunate that the very first letter carried such a homely message and it allowed an interested public to relate immediately to the garrison of the Roman frontier two thousand years ago. It conveyed how different, and how important these amazing finds were.

The socks and underpants in themselves also contributed new information – as often with the contents of the letters. Hitherto there had been no evidence that the Romans wore either, although bare feet in leather sandals in a Northumberland January would have demonstrated unimaginable hardihood. And underpants no doubt supplied much needed manly support.

The reference to messmates, contubernales, suggested a male sender who understood something of army organisation. Perhaps he was a Batavian veteran who had returned home. In any case he knew that each century in an auxiliary regiment, and a legion, was divided into ten platoons of eight men. On the march they shared a tent, and at Vindolanda a barrack room. Bunk beds probably made these more spacious and congenial than the archaeological remains imply. Each platoon cooked its own meals, the men probably taking turns. There was no communal dining (although a club, the scola, existed for noncommissioned officers) and the army did not supply rations. The fort appears to have sold barley to its soldiers as well as chickens and eggs, perhaps through some accounting mechanism which reverted the cash to the commander or the army. Not surprisingly, large volumes of beer were sold; one individual order was for the equivalent of 50 pints and it seems to have been very cheap.


Slaves usually ran the kitchens of the well-off in ancient times, and they also invented many recipes. The best-known collection is thought to have been compiled by Marcus Gavius Apicius, a gourmet who sailed the Mediterranean in search of giant prawns. In fact it is much more likely to have been written by a series of scribes who were given recipes by slave-cooks. Sally Grainger and her husband have translated many of these and, in an excellent new edition of Cooking Apicius, have made them very accessible to modern methods; blenders are unblushingly used. The recipes are superb and unnusual: favourites might include a starter of soft-boiled eggs in pine kernel sauce, a side dish of spicy mushy peas. The Romans made good vinaigrettes for salads and called themoeno-garum. In addition to several dipping sauces for meats, two of the best mensae primae are the roast lamb and the belly pork known as ofellae. Almond and semolina pudding is surprisingly light. As with modern Mediterranean cooking, the emphasis is always on the excellence of the ingredients.

Far to the north of Vindolanda and built forty years later, Bearsden Fort in Glasgow revealed much about the diet of ordinary soldiers. Through painstaking analysis of the contents of a blocked latrine drain, strong-minded experts have been able to show that porridge was a staple. In season, wild fruits, berries and nuts could be gathered by soldiers and mixed in to add some variety and flavour. Hazelnuts, brambles, bilberries, crab apples and raspberries would have been almost farmed and protected by each platoon. This wild harvest also had the attraction of costing nothing, no deductions from pay. The men were probably allowed to fish, and to trap and snare small animals, but just as medieval kings guarded their hunting reserves, it is likely that commanders like Cerialis and Brocchus forbad the pursuit of bigger game by anyone else.

The arrival of so many men in northern England at the end of the first century AD transformed the local economy. Although their numbers were small compared with the 30,000 or so who came to build Hadrian’s Wall in AD 122 and its subsequent garrison of 12,000 auxiliaries, the soldiers stationed at Vindolanda and the other early forts strung out along the Stanegate Road provided a substantial new market for goods and services. There were perhaps 5,000 of them – and they all needed something.

Once the countryside had recovered from the initial shock of conquest – and this is not to be underestimated – trade began. But it must be remembered that the forts were built on good land which was worked by the native peoples. The Romans did not enter an empty landscape. Summarily driven off wide swathes of pasture and arable ground, the farmers will have nursed their fury for at least a generation. It is estimated that a cohort of 500 men needed the yield of 700 acres of cornfields each year. Add to this at least an equivalent amount of pasture for 200 oxen (used to pull the regiment’s carts), 200 or so ponies, and woodland for the fort’s pigs and chickens to root around in and the result is a very large appropriation of land, a dispossession and hardship on a huge scale. Vindolanda’s territorium may have run to 1,500 acres. This sort of deprivation took place all along the Stanegate, and on a renewed and greater scale when the Wall builders came. It will have taken some time before local producers began to trade significantly with the invaders who had evicted so many.

But trade they did. Markets developed as goods and food were brought for sale to the gates of the forts. And while the occasional complaint about prices and supply surface in the Vindolanda letters and lists, it was in everyone’s interests for these markets to benefit both buyer and seller. Local goods were almost always cheaper than anything shipped from the south.

Perhaps the most radical economic shift is nowhere explicitly recorded, or at least no record has survived. The Romans operated a money economy, and the natives seem to have used barter. In order for trade to function at all, cash must have been universally adopted. How else could transactions have been completed? The soldiers had little to offer in the way of goods to barter, and cash was what they were paid. The only hint of rapid adoption is in the Vindolanda lists where people with native-sounding names do write of the cash costs of items. But it must at first have been bewildering.

In order for trade to transact smoothly – there was after all money and livelihoods involved – it is inevitable that the native peoples will have had to learn to speak Latin. Many loan-words made their way into Old Welsh but the linguistic traffic did travel in the opposite direction in one significant way.

The names used by the Romans for their forts were almost all Celtic in derivation. And this process of adoption took place early. Within less than twenty years of the building of the first timber fort, letter writers marked their address clearly as Vindolanda. The meaning of the name is a rare insight into native reaction to the Roman invasion. From uindo (gwyn in modern Welsh), the first element means white or bright. This is almost certainly a native reference to the Roman habit of plastering their buildings with render to help keep out the weather. A deposit of wall plaster was found outside the northern rampart, and it more than suggests that the fort’s gatehouses were rendered as the IX Batavians rebuilt Vindolanda around AD 100. There are several other Vindo names in the Celtic regions of the Empire and at least four in Britain. They probably echo a similar reaction to the rapid appearance of bright, white buildings in the landscape; Vindobala is nearby at Rudchester in Northumberland, Vindomora is in County Durham, Vindogara in Ayrshire and Vindocladia in Dorset. These foreign, new and threatening structures stood in startling contrast with the earth colours of the native roundhouses built out of natural materials which blended into the soft greens and browns of the Northumberland landscape.

The second element of the place-name is simpler. It comes from llan, a modern Welsh term for a church or, more correctly, a churchyard. Originally it described a fenced or walled enclosure. The White Fort is probably the native name for what its soldiers knew as Vindolanda.

The Tungrians were first on the scene, perhaps around AD 85, and the earliest letters and lists say a good deal about their tenure of the White Fort. But some of the most detailed archaeological and documentary information is confusing. Excavation indicated a standard layout for the early fort, probably built for a cohort of 480 men. Significantly the ditching defending the western wall was deep and elaborate (the other three walls overlooked downslopes), a reminder that the Tungrians were digging in hostile country. So far, so predictable. But a fascinating list came to light early in the excavations. It is a strength report of the I Tungrian Cohort, and it describes the state of a regiment of 752 men, far more than could be accommodated in the new fort:

18th May, net number of the I Cohort of Tungrians, of which the commander is Julius Verecundus the prefect, 752, including 6 centurions.

Of whom there are absent:


Guards of the governor


at the office of Ferox


at Coria


including [?] 2 centurions


at London [?] a centurion


including 1 centurion


including 1 centurion


at [?]




total absentees


including 5 centurions,


remainder, present


including 1 centurion


from these:






suffering from inflammation of the eyes


total of these


remainder, fit for active service


including 1 centurion.


Most striking is how few soldiers there were at Vindolanda fit and able to deal with any emergencies. Only a little more than half a standard cohort held down a wide swathe of, at best, unsettled countryside. The neighbouring forts of Carvoran in the west and Newbrough in the east lay only 8 or so miles away, but too far to be helpful in the event of surprise or night attack. Morale amongst the native kings must have been low, to say nothing of their military intelligence.

The dangerously sparse garrison at Vindolanda was probably in the process of being upgraded. The large detachment at Corbridge, 337 soldiers, may have been new recruits undergoing some sort of induction. Their origins are uncertain. They cannot have been local – too much bad blood had been too recently shed – but they might have been drafted from southern Britannia. More likely they were tax-equivalent Tungrians fresh off the boat.

The Vindolanda letters and lists often provide a disconcertingly everyday background to the grim grind of the Roman military apparatus. While his soldiers were sweating at the bottom of deeper ditches dug below the western wall to protect their fort from a threatening hinterland, their commander, Julius Verecundus, was being very particular about the sort of apples he liked. Another aristocrat from the Netherlands, he owned at least one slave; here is the shopping list he gave him for what sounds like a trip to the more sophisticated market at Corbridge:

Two modii of bruised beans, 20 chickens, 100 apples – if you can find nice-looking ones – 100 or 200 eggs, if they are on sale there at a fair price . . . mulsum [honey-flavoured wine] . . . eight sextarii [about four litres] of fish sauce [muria] . . . a modius of olives . . .

Vindolanda was enlarged to accommodate a double-strength cohort, but Army Command North, in the legionary fortress at York, sent orders for the infantry of the I Tungrians to move out and be replaced by the part-mounted Batavians. From this time onwards the volume of letters and lists increases markedly, and they paint a vivid picture of what garrison life was like in the White Fort.

The day began with a morning muster when a roll call was taken and duty rosters read out. At Vindolanda both a list of these duties and the names of the men who worked at specialist tasks has been found. One morning, two thousand years ago, a centurion or his second-in-command, the optio, barked out that detachments were to be assigned to building work at the new bath house (recently discovered outside the south wall), making shoes, collecting lead and rubble, working in the lime and clay pits, and carrying out general plastering and building work. If it was raining and cold, the immunes, the men with inside clerking jobs, might have smiled a sly smile.

Another list detailed the regiment’s specialists. Because it was a part-mounted outfit, the IX Batavians had two veterinarii, who looked after their ponies, the large herd of oxen and all the other livestock. Horses were particularly valuable and, with mounts for 240 cavalry troopers and a reserve of remounts to cope with, the vets will have been busy.


Julius Caesar and other Roman commentators noted the British habit of wearing tattoos and decorating their bodies with paint. But the Romans did it too, albeit more discreetly. It seems that soldiers were allowed to wear a legionary tattoo once they had completed some sort of test, perhaps an initiation. Some auxiliary regiments did the same thing. In his fourth-century treatise ‘The Epitome of Military Science’, Vegetius wrote that the tattoo was earned by a physical test, perhaps one concerned with endurance. The tattoo may have been an eagle and, after discharge, it was a handy proof of military service.

Horsemanship was clearly premiated amongst the Batavians, and some Vindolanda correspondence leads to insight on this. Letter-writing appears also to have been an important business, and much valued. Here is Chrauttius complaining that Veldeius has not been keeping up:

And I ask you, brother Veldeius – I am surprised that you have not written anything back to me for such a long time – whether you have heard anything from our kinsmen, or about Quotus – in which unit he is – and you are to greet him in my own words – and Virilis the vet. You are to ask him whether you may send through one of our people the shears which he promised me for a price. And I ask you, brother Virilis, that you greet me from sister Thuttena and write back to us about Velbuteius, how he is. I wish you may be very happy. Farewell.

The address on the back of the letter is London, and Veldeius appears to have been a groom on the Governor of Britannia’s general staff. The Tungrian strength report showed 46 men on secondment in London; the practice of drawing men from the northern frontier to serve in the Governor’s bodyguard was not uncommon.

When the Governor visited Vindolanda in 105, it seems that Veldeius was in his retinue, and that he brought Chrauttius’ complaining letter with him. Once again written sources link with archaeology to produce a rich picture of life in the fort. Preserved in the black mud, a piece of leather was found with Veldeius’ name stamped on it. Perhaps it was an offcut from a tanned hide he had bought and brought to a saddler to have some piece of tack made. A chamfron, a beautifully worked mask worn to protect the heads of cavalry ponies (but in reality used mainly in parade-ground displays) was also discovered. It belonged to Veldeius, and if it was made for his own horse, then he owned a beauty. The chamfron best fits a pure-bred Arab mare, a big horse for its type, standing more than fifteen hands high.

The vets at Vindolanda will have seen many fancy horses, but their main concern was with the shaggy little ponies grazing in the regimental paddocks, the cavalry mounts. Most would have been smaller than Veldeius’ Arab, and much more strong-boned and chunky. Their brood mares must have interbred with local native stallions to refresh their bloodlines. Native hill-ponies were hardy and well used to the uncertain ground and the wet and windy conditions which often obtained. The Vindolanda lists and letters have maddeningly little to say about the people who rode them, the native British, and when one Roman observer sniffs at their poor military prowess, he focuses as much on their ponies:

the Brittones, rather many of them cavalrymen, are naked [perhaps meaning ‘without body-armour’]. The cavalrymen do not use swords, nor do the Brittunculi mount to throw javelins.

The most revealing aspect of this fascinating passage is perhaps the least suprising. As has been noted, Brittunculi is a dismissive diminutive meaning something like ‘pathetic little Brits’. Invaders and colonists always look down on the natives they have conquered, and Brittunculi could not have been worse than many names conferred by the British Empire – wogs, niggers, kaffirs, fuzzie-wuzzies are even more offensive.

The rather many cavalry of native hosts is observed elsewhere on the frontier and the society appears also to have been a horse-based one. But it sounds as though the pathetic Brits not only fought on horseback but, as seems to have been the case in the north and at Mons Graupius, also sometimes used their ponies for speed and mobility, dismounting to throw javelins, fire arrows or use sabres and shields. The memorandum mentioning the Brittunculi may have been a disparaging comment on raw British recruits who needed licking into shape. But the date is early, and training local warriors in advanced military methods seems a little risky, even if they were to be posted elsewhere. British regiments turned up on the Rhine frontier in the late first century AD and into the second century.

The morning roll call and duty roster will have held no surprises for some of the men. In peace-time they did the same things every day. Candidus looked after the piggery, making sure farrowing sows did not squash their litters (or indeed eat them). And then once they had been weaned, he found them safe access to the delights of the woodland surrounding the fort. The swineherd was an important man. Archaeologists have found enough bones to know that the garrison was very fond of pork. Bacon and ham was cured and kept well through the hungry winter months.

Oxherds are also noted on the rosters. These men, who looked after the feeding and safety of 200 oxen, probably lived in shielings outside the fort so that they could be near the 100 or so acres needed to keep the beasts fit and strong. At one point, a supply of food is sent to the wood to feed the men and keep them in good condition.

Condition, or well-being, could also be found at the bath house, which had been built by a work-party. Archaeology has discovered something which may have caused the workmen to shout and shoo. The caldarium and laconicum, the hot rooms, were heated by a hypocaust, a system which worked on the principle of warm air circulating under the floor. This needed a series of supports made from tiles. They could stand the heat and also allowed more precision in holding up a flat floor. When the workmen made the tiles from puddled clay, they laid them out on the grass to dry off before being fired in a kiln. It was at that stage the shouting and shooing took place. Several tiles have been found with clear animal paw and hoofprints on them. While they dried on the grass, dogs, cats, cows, pigs and either sheep or goats, or both, stepped on them. Perhaps no one was around to do any shooing.

Other skilled workers went about their tasks on what must have been a daily basis when they were not soldiering. Only one cartwright, Tullio, is mentioned, but he probably had charge of a workshop of several men. Oxcarts were vital to the running of the fort, and soldiers were careful with them:

The hides which you write are at Cataractonium [Catterick] – write that they be given to me – and the wagon you write about – and write to me what is with that wagon. I would have collected them already – except that I did not care to wear out the baggage animals while the roads are bad.

Bad Roman roads! Another revelation. Despite the above writer’s consideration, carts often broke down, and Vindolanda’s cartwrights will have been busy. Evidence of trade with a man who sounds like a local supplier suggests a large number of wagons were used by the fort. Metto wrote the following to Advectus:

I have sent you goatskins . . . sent through Saco . . . 34 wheel-hubs; 38 cart axles, one of them turned on a lathe; 300 spokes.

Other traders also supplied Vindolanda with hides. The regimental cobblers, the sutores, made hob-nailed boots, not sandals, for the soldiers, and there are several mentions of small numbers of nails being bought to make do-it-yourself repairs. Holes were deliberately punched through the uppers to allow water to squelch out. Boots were designed to protect the feet from sharp stones and worse, but not to keep them dry. Waterproof footwear for soldiers is a recent invention. The Highland army which crossed Hadrian’s Wall in 1746 wore very similar shoes. The Gaels called them brogan. Changed only a little into brogues, the principal design feature of these modern shoes is the tooling on the uppers which resembles half-cut holes.

The duty roster included other roles such as the scutarius, or shield-maker, the venetus, the butcher, and of course, most important, the cervesarius, the brewer. The roster itself had probably been compiled by the men with the cushiest number of all, and their boss was known as the cornicularius, the chief pen-pusher. Records were kept in the principia, the headquarters building in the centre of the fort, and files were copied in duplicate and sometimes triplicate before being stored in capsae, document boxes. It is astonishing how few files have survived. From the beginning of the reign of Augustus in 31 BC to the beginning of Diocletian’s in AD 284, one historian has calculated that the pen-pushers generated about 225 million records of Roman army pay, but only three have been found in reasonable condition. Clerking may have been dull, but it was also dry and warm, a lot better than road-building and ditch-digging, and a lot better paid.


Carbo marinus, or sea-coal, was first picked up on Northumberland and Durham beaches a very long time ago. Prehistoric peoples probably burned it. The Romans certainly did. Coal holes have been found at both Housesteads and Vindolanda, but not for sea-coal; this sort came straight out of the ground. Along the banks of the Tyne and in isolated pockets inland, there were outcrops known as coal-heughs and, until the late Middle Ages, miners could get at them easily. Water transport for such a bulk item was handy, and barges may have brought it up the Tyne as far as Corbridge. Coal was probably used for heating up the bath houses. Until the nineteenth century there was a resistance to using coal for cooking. The fumes and gases which sometimes hissed out of big lumps persuaded people to use wood to cook on.

Training was considered essential to maintain the physical fitness of all soldiers, and at Vindolanda the centurions organised it. It must have been difficult to dodge. Every ten days a route march in full kit was undertaken and the required rate was at least three miles an hour. More elaborate manoeuvres were planned and executed, temporary marching camps dug in remote locations and mock attacks mounted. Hadrian approved heartily of these as means of maintaining discipline and avoiding the ever-present danger of mutiny. But Army Command North at York had to approve manoeuvres in advance in case native kings misread them as genuine acts of warfare.

Vindolanda was kept busy – perhaps precisely because, in all the lists and letters, there is not one solitary reference to war or fighting. All along the frontier, the Roman garrison spent 99 per cent of its time doing something else. Idle soldiers can become undisciplined and dangerous, and their commanders kept them almost neurotically at work.

Building seems to have been a near-constant activity. Until 122 and the construction of Hadrian’s Wall, Vindolanda and the other Stanegate forts were made from timber. Posts rammed and chocked directly into the ground rotted quickly and that meant that the life of such buildings was no more than eight to ten years. Even the stockaded rampart could weaken, and soldiers packed a bank of earth and rubble against the inside to add strength and also allow rapid access to the rampart top in the event of a surprise attack. Later, ovens and kilns were dug into the back of this internal mound. Needing fiercely hot fires to be efficient, both could also be dangerous if let out of control, and their location at the rampart was safer than in the body of the fort. Ovens and kilns are also found in gatehouses. There is a particularly well-preserved example at Birdoswald Fort. It will have had the welcome effect of keeping sentries warm on long watches through winter nights. At Housesteads their famous latrines were also dug into the rampart bank, well away from the barracks block, but not for reasons of safety.


Under the floor of Hexham Abbey, surely one of the most beautiful and atmospheric churches in Britain, lies a remarkable structure. The crypt is all that remains of St Wilfrid’s seventh-century foundation. Before work began, a huge hole was dug and a small chapel built in it. Every stone came from the ruined Roman town at Corbridge. Still legible is an inscription dating from the visit of the Emperor Septimius Severus in 208. Unfortunately this stone has been removed now and put in the nave. In situ is a fragment of a pagan altar to Maponus Apollo holding up a passageway. Much of the Roman masonry has been broached so that render would stick to it. A leaf and berry design from the frieze around the walls of a fine house at Corbridge is visible in several places. It probably still had its roof on when Wilfrid’s masons robbed out the stone. The tiny chapel housed a shrine to St Andrew. Wilfrid had been to Rome, bought a relic of one of the disciples and brought it back to Northumberland. A direct link with Christ, it would have been visited by many pilgrims, and its planting under the church sanctified the ground it was built on. The seventh-century mason work is crude, some of the robbed stones are upside down, but it is, after a fashion, a complete Roman building.

In some ways the timber forts were like more solid versions of marching camps, not designed to be defended like medieval castles, but to slow down an attack. The Romans’ instinct to get out into the open to fight is well illustrated at the cavalry fort at Chesters, near Chollerford. It abuts Hadrian’s Wall but has three of its four gateways to the north of it – so that squadrons of troopers could gallop out and get at an attacking force quickly. At Ambleside in the Lake District, an exception was recorded. The tombstone of Flavius Romanus, a regimental clerk, notes that he had been killed by the enemy inside the fort.

As the frontier settled on the line of what was to become the Wall, and the province of Britannia quietened, the Romans began to rebuild their fortresses in stone. The garrison consolidated into a permanent army of occupation. By AD 122 many auxiliary regiments had been in Britain for eighty years. As soldiers completed twenty-five years of service and were discharged as Roman citizens, some will have returned to Batavia and Tungria with their savings, their bronze diplomas and their stories. Others undoubtedly stayed. As Kipling judged in ‘The Roman Centurion’s Song’, many had known no other place as home. Some will have formed lasting relationships with local women, and although army regulations forbad marriage while in service, they will have settled into a common law arrangement. A blind eye was usually turned. What happened to these men? Where did they go after all those years in the army?

Not far, is the likely answer. When Tacitus wrote of smart dinner parties and the subtle process of Romanisation, he was describing the subversion of the native aristocracy. With ordinary soldiers who decided to stay in Britannia, the process took a different route. Having spent most of their adult lives in the army, they were the living agents of Romanisation, at least in the north. Those who married local women and had children by them created a slim but significant stratum of society with a direct stake in developing a more Roman culture. Enfranchised by marriage to a veteran, citizen-families settled in Britannia, and it is likely that those who left their posts in the frontier forts moved only a short distance, going to live outside the gates rather than inside them.


British admiration for the Roman Empire springs partly from a recent memory of our own imperial centuries. One of the great bards of the British Empire may have been recalling stories of old India hands who had sailed home to an uncomfortable retirement in the Home Counties, far from the country they had come to love. In this moving poem, Rudyard Kipling cleverly reverses the compass:

Legate, I had the news last night – my cohort ordered home,

By ship to Portus Itius and thence by road to Rome.

I’ve marched the companies aboard, the arms are stowed below:

Now let another take my sword. Command me not to go!

I’ve served in Britain forty years, from Vectis to the Wall,

I have none other home than this, not any life at all.

Last night I did not understand, but, now the hour draws near,

That calls me to my native land, I feel that land is here.

Here where men say my name was made, here where my work was done,

Here where my dearest dead are laid – my wife – my wife and son;

Here where time, custom, grief and toil, age, memory, service, love,

Have rooted me in British soil. Ah, how can I remove?

. . .

Legate, I come to you in tears – My cohort ordered home!

I’ve served in Britain forty years. What should I do in Rome?

Here is my heart, my soul, my mind – the only life I know.

I cannot leave it all behind. Command me not to go!

Vici, or civilian settlements, appeared early around forts and most, like the one outside the west gate of Vindolanda, were built very close to the walls. This is surprising because the presence of these houses removed a vital defensive advantage. Fort builders usually cleared wide areas on all sides, so that an attacking enemy was forced to approach over open ground, and thereby be seen and be vulnerable to fire from the ramparts, or at least be quickly seen. The houses of the vici huddle so close that they must have been the homes of absolutely trustworthy people. And who could be more so than regimental veterans? They would also understand if their homes had to be fired or levelled if there was war and the fort was threatened. In any case the vici were built on the territorium belonging to the army – portions of which could be granted to veterans on their discharge.

These small villages set up an important link between the native and military communities. In addition to holding regular markets, they supplied services. The Emperor Hadrian disapproved of vici and, in listing his reasons, he outlined exactly what was attractive to soldiers: drinking booths, gambling halls and prostitutes. In most there were at least two buildings which appear to have been standard; a bath house and an inn, called a mansio. The latter is usually primly interpreted as a place where travellers or visitors to the fort might lodge. It was also almost certainly a whorehouse. With either 500 or 1,000 young and unmarried soldiers living mostly bachelor lives inside Roman forts, the notion that they were not serviced by a large number of whores is entirely absurd. Of course no trace of them remains, no archaeological or literary evidence has yet been found at Vindolanda, but its absence does not mean that the White Fort was populated by whiter-than-white, clean-living young men with a wide variety of hobbies to occupy them in the evenings. They were soldiers and they behaved in the way soldiers have always behaved.

Gambling, by constrast, has left a mark, and it certainly went on in the bath houses. Many games revolved around dice and a throw-board. A favourite was ludus latrunculorum, or robber-soldiers. It was a battlegame like chess. At the fort of Segedunum at Wallsend a board and a set of counters were found, and it seems that pieces were moved according to throws of the dice. All of them moved in straight lines like the rook in chess, and could be captured and removed if both their advance and retreat were cut off. According to contemporaries, the wedge formation was most effective – just as on the battlefield. No doubt substantial sums changed hands as players bet against each other and others bet on them.

The Romans seemed fascinated by gambling. A good deal of literary evidence survives – the Emperor Claudius even wrote a history of dice-games and was apparently devoted to them. Another game, called duodecim scripta, or twelve points, and resembling backgammon, inspired the poet Ovid to describe it: A sort of game confined by subtle method into as many lines as the slippery year has months.

Played without a board, tesserae, or dice, were often marked with different values, and players would bet on how they landed. At Birdoswald four dice were found as a complete set and each face had I, III, IV or VI inscribed on it. The principle was the same as for modern poker dice. Venus was the best throw since it showed all of the different values at once, and Dogs was the worst with four Is. Poor throws were penalised by adding more stakes to the pot, which was scooped by the first player to throw a Venus. Sums could spiral. Gambling chips came into use: there were three denominations, I, V and X. As the stakes rose, a crowd must have pressed hard to watch the high rollers in the warmth of the caldarium. Perhaps it was not just the hypocaust that made men sweat.

Snacks were served in the bath houses. Oysters, mussels and other titbits were on sale, and Hadrian’s third vice, drinking, must have washed down many a plateful. There is no doubt that overindulgence, sexually transmitted diseases and indebtedness through gambling were all a danger, but camp commanders were more than likely glad of the diversions of the vici. They allowed soldiers to let off steam, kept up morale and broke up the dull, quotidian routines of guard duty, route marches and cleaning out the latrines.

Over time the fort-villages grew. Outside Chesters, which had a garrison of a thousand cavalry troopers, there were four streets arranged around a crossroads. And at Old Carlisle in North Cumbria surveys have hinted at a sizable settlement of several hundred souls. To the army one of the most attractive aspects of these villages was probably as a reservoir of recruits. Many sons of veterans will have followed their father’s footsteps onto the parade grounds. Batavian and Tungrian military traditions might even have been sustained by a second generation, perhaps a third as the army renewed itself. An inscription found in Egypt and dated to AD 194 records the retirement of forty-six veterans from the Legio II Traiana Fortis. Unusually it lists the origins of the soldiers, and more than half (twenty-four) said that they had been born at a fort. In order to enlist in a legion, they will all have had to be citizens, and they were probably enfranchised on the discharge of their fathers and their formal marriage to their mothers. As in modern times the names of the regiments never changed to reflect different recruiting grounds, and it would be a mistake to assume that the men of the Batavians or the Thracians later came from either of these places.

Later altar dedications at four forts show that the villages were assuming a measure of self-government. At Vindolanda, Old Carlisle and Housesteads cash seems to have been contributed, perhaps through taxation, perhaps donation, to pay for altars. Two are dedicated to Vulcan, the god of blacksmithing, and the army may have been happy to see fire-hazardous smiddies set up outwith the walls of its forts. At Old Carlisle the inscription was specific: dedicated by the village elders with money contributed by the villagers.


To our sensitive noses the ancient world-would have stunk to high heaven. Worst, by far, were the leather-tanning pits. These used a disgusting soup of urine and dog turds to produce the tough, treated leather that the Roman army badly needed. Next on the scale was probably animal dung. Oxen splatted the streets of Vindolanda regularly, ponies piled out their muck, chickens shat their eye-watering guano, and no one bothered – unless it landed on them. Almost all the soldiers had been raised on farms and they were used to it. In fact, up until the early twentieth century, most people were – from the farm workers who shared their cottages with their cow in the winter (imagine steaming piles of dung dropped only a few feet from where people slept) to the dainty ladies of the cities dodging street manure of all kinds. There were good smells at Vindolanda too: cooking, baking, woodsmoke and someone who had just visited the bath house.

A macabre discovery by Eric Birley in his 1930 excavation of the vicus at Housesteads suggests that civilian life on the Roman frontier was occasionally as wild as the equivalent in nineteenth-century America. Under a false floor he found two skeletons. One had a knife blade embedded in its ribs. The site became known as the Murder House, for no one doubted that was what had taken place. All Roman burials had to be located outside settlements and forts, and the obvious concealment of two bodies amounted, at the very least, to what the police are fond of calling suspicious circumstances. More resembling an incident from Tombstone or Dodge City, it vindicated Hadrian’s stern disapproval. The vici could be lawless, dangerous places. When Eric Birley reported his find to the coroner at Hexham, as he was bound to do, a verdict came back of murder by person or persons unknown shortly beforeAD 367.

A hundred years before the dark deeds at Housesteads, the vicus at Vindolanda seems to have been abandoned. Around 270, it may be that the substantial reduction in the garrison allowed villagers to move inside the fort’s sheltering walls. There was plumbing, the buildings were better constructed, and there was running water. By 400, all of these settlements had gone. They existed only to service the forts and clearly made no economic sense without them. Only Carlisle sustained itself. When St Cuthbert visited in 685, the water supply was still working, and medieval chroniclers reported that Roman streets were still paved and buildings still standing as late as the 1200s. But Tacitus’ cynical predictions did not come to pass in the north. Despite the four centuries of a large garrison along the line of the Wall, Romanisation simply did not take. Celtic culture was too powerful.

All of this lay far in the future. Around AD 100 the IX Batavians were as busy as ever. Part of the reason why they had replaced the I Tungrians was their part-mounted capability. Their 240 cavalry troopers were also excellent intelligence gatherers and watchers, patrolling the countryside, asking questions, looking out for suspicious movement. Tagged onto their title, the regiment had the additional description of exploratorum, the scouts.

The immediate vicinity of Vindolanda appears to have formed part of the territory of the Textoverdi. Only 3 kilometres from the fort an altar dedicated to the goddess Saitada has been found. It is unique. Like hundreds of other Celtic deities whose names appear only once in the historical record, Saitada sounds like the focus of a local cult. Perhaps she was the genius, the patron goddess of the Textoverdi, since the inscription informs that the altar was set up by curia Textoverdorum. This may be a mis-spelling of coria, the Old Welsh word found in the place-name of Corbridge and meaning the host or the hosting-place. Or it may mean something like council, perhaps in the sense of elders.

The Textoverdi of the South Tyne Valley may have allied themselves with the Brigantes, been part of their federation. Some of the personal names of the people who appear in the Vindolanda letters and lists were almost certainly Textoverdi and will have spoken their dialect of Old Welsh. The name itself is obscure, but it may be related to the Celtic root-word teach, which meant something like fleet or fast. Maybe they and their ponies were famously adept at the gallop across the rough country they knew so well.

Like all Celtic peoples the Textoverdi left no written record, no means of confirming such supposition. By contrast the ability of the Vindolanda garrison to use written communication was a distinct military advantage because of the way in which it spread precise information, and made the Roman army more efficient, better able to punch above the weight of its often depleted numbers. Messages could be written quickly and carried at speed along the Stanegate to neighbouring forts where they might be read and acted upon. A sense of this urgency and immediacy was captured in a remarkable record of a mistake. When Flavius Cerialis was dictating to a scribe, he did not make himself clear. Perhaps he stumbled over the words of what was after all his second language. The scribe first wrote tempestates et hiem, then scrubbed out the last two words (not well enough – probably because he was being hurried) and replaced them with etiam, what Cerialis had really meant to say.

When the Governor of Britannia came to Vindolanda in 105, he may have brought orders from the Emperor, Trajan, who was fighting a ferocious war on a vast scale in Dacia (modern Romania) and against a formidable enemy, King Decebalus. Troop deployments would have to be reorganised to allow reinforcements to be sent to the Danube frontier. Flavius Cerialis and Sulpicia Lepidina packed their many belongings onto oxcarts and, after a long posting and very distinguished service, the IX Batavians left Britain in the summer of 105.

The astonishing detail which careful archaeology can produce shows that much of Vindolanda lay empty for some months before a new garrison rumbled in through the gates in the late autumn. Blown off the trees fringing the fort, dead leaves had rustled into many of the rooms, piling up in drifts in some of the corners. Birds had flown in through the gaping windows and left some feathers and droppings on the floor. Squirrels had hopped into Cerialis and Lepidina’s apartments and buried their hazelnuts beneath the carpet of leaves and bracken. Seven caches were found.

The contrast must have been baffling to the native peoples. At one moment the White Fort buzzed with the racket and clangour of soldiers, blacksmiths, wagon trains and parade-ground commands. The next, it was silent as the wind sighed through the buildings and the leaves swept into its rooms and blew down the via principalis.

It was the I Tungrians who came back in late 105. Their commander was Priscinus and it is likely that his men occupied the fort until 122. Having no cavalry, the Tungrians were reinforced by troopers from the I Cohort of Vardulli, originally from northern Spain. There seems to have been serious trouble in Britain. The forts to the north, at Newstead near Melrose, Dalswinton and Glenlochar in the southwest, and Cappuck and Oakwood in hills above the Tweed basin, were all destroyed by fire, most of them deliberately. The army was retreating from the lands of the Selgovae and the Novantae. But it seems that the departure of the Romans was encouraged. Archaeologists believe that Newstead shows signs of having been attacked. Human bones were found charred amongst the wreckage as well as a great deal of valuable kit and some damaged armour. Perhaps the Selgovan kings led their warbands down the Tweed, roared their war-cries and launched themselves at the great army depot at the foot of the sacred Eildon Hills.

High Rochester, impressively sited on the line of Dere Street and commanding much of Upper Redesdale, was also burned, but the flames did not stop there. Corbridge also went up in smoke. It seems certain that these two forts were attacked, possibly also by the Selgovae. Lying too far south to be part of any deliberate withdrawal from Scotland, Corbridge and its growing town presented too tempting a target. It looks very much as though a large and powerful warband rode down Dere Street, from Newstead to Corbridge, leaving fire and destruction in its deadly wake. As the Tungrian strength report showed, forts could on occasion be desperately short of manpower.

After the Selgovan kings and their warriors had ridden back home over the hill trails through the Cheviots, no doubt lugging their plunder after them, the northern frontier grew quiet again. And its line seemed to settle along the Stanegate Road.

Security at Vindolanda probably tightened in the wake of 105, its huge iron-studded wooden gates closed and barred at nightfall, the guards on the platform above scanning the darkening horizon. As a matter of routine each morning a daily password was agreed, written on a wooden tablet and given by the duty officer, usually a centurion, to an orderly. He then toured the gates, ramparts and the principia making sure all the soldiers on guard read and remembered it. This laborious business was designed to preserve the password’s secrecy, but it also suggests that ordinary soldiers had some ability to read. Simple words like courage or victory were commonly used.

The most routine problem encountered by sentries was probably theft. The fort was normally full of supplies and valuable items. Hungry natives and greedy or needy soldiers may have raided granaries and other stores. The Vindolanda letters and lists occasionally complain of things going missing. After nightfall it is likely that four-man pickets patrolled outside the walls of the fort to keep an eye on stock and other movables, like carts. They could also have made out the shapes of anyone attempting to climb the rampart. Perhaps they had dogs on a leash. Watches rotated frequently throughout a 24-hour period with around a fifth of the garrison involved on any given day.

Arguably Rome’s most dynamic and successful emperor, Trajan had succeeded Nerva in AD 98. His interests lay in the east, both on the Danube and in what is now modern Iraq. Unlike his predecessors, he showed little interest in Britain, certainly none in advancing up into Scotland. Glory lay elsewhere. During his reign the frontier stayed on the Stanegate Road and the forts along its length. It was a well-made, two-way paved road able to carry heavy military traffic. A clear outline of its camber and the drainage ditches on either side can be seen from the western approach road to Vindolanda. The number of Stanegate forts was increased, and the frontier seems to have been extended beyond Carlisle to Kirkbride and Bowness-on-Solway. Seaborne raids across the firth may have been troublesome. Lookout towers were built on the high ridges to the north of the forts in the central section, and one, at Walltown Crags, was certainly incorporated into Hadrian’s Wall. But sources for the period between 105 and 117, the death of Trajan and accession of Hadrian, are sparse. The name of the Governor who took over Britannia after L. Neratius Marcellus is not even known.

In the south of the province the legionary fortresses at York, Chester and Caerleon were all rebuilt in stone, and new colonies of veterans had been founded at Lincoln and Gloucester. More troops arrived in Britain around 105, perhaps as reinforcements after the attacks of the Selgovae. The prefect of the II Cohort of Asturians was buried in Alexandria, in Egypt. The inscription recalled his service in Britain with the Asturians when it noted that C. Julius Karus had been decorated for bravery, probably in the war of 105.


Three generations of Birleys have found themselves at the bottom of Vindolanda’s wet and muddy trenches and, without their decades of hard work, the state of our knowledge of Roman Britain would be immeasurably poorer. It all began in 1929. The estates of John Clayton were put up for sale. The owner of Chesters, Housesteads, Carvoran, Vindolanda and Carrawburgh, he had done a great deal to rescue Hadrian’s Wall from nineteenth-century stone robbers and destroyers. A young Eric Birley had become fascinated by the Wall, and on being told that Vindolanda was a site of huge archaeological potential, he persuaded his father to help him buy the farm it stood on. The first excavations began. Then war intervened, and then Eric Birley’s academic career at Durham University took up more and more of his time. By 1950 he had sold Vindolanda. But family interest continued. Robin Birley started to do research in 1956 and became increasingly certain that the site would ultimately reveal a great deal, if only he could get at it. The farm came up for sale in 1970. Mrs Daphne Archibald, the mother of one of Robin Birley’s volunteer excavators, bought it and gave what was known as ‘the Camp Field’ to a newly formed Vindolanda Trust. Work began in earnest. Sustaining itself solely by receipts from visitors and the revenue from excavation courses, the site slowly fulfilled all that potential, and more. Robin Birley’s colleague, Patricia Burnham, became Mrs Birley and by 1974 the Trust began to expand. The nearby cottage at Chesterholm was bought and converted into a museum. The discovery of the first letters and lists built momentum, and the publicity they generated helped to bring in more than 100,000 visitors each year. A third generation now runs Vindolanda’s excavations. Andrew Birley has taken over from his father and, each summer, the distinguished Roman scholar Professor Anthony Birley finds his wellies and overalls and helps with excavation. Vindolanda is a remarkable place, dynamic and constantly adding to the store of knowledge about the Wall and Roman Britain. Without the Birleys, it might, as with several other important sites, have remained a green field with a few strange humps and bumps.

At Vindolanda another tombstone commemorated Titus Annius, a centurion of the I Tungrian Cohort, possibly their acting commander, who was killed in the war, fought in 117 in Britain. All around the Empire native peoples were well aware of the shifts of imperial politics and knew that the death of an emperor and the shaky accession of another could herald a period of instability – and opportunity. In a list of the problems facing Hadrian in 117, a third-century historian noted that the Britons could not be kept under control. The kings in the north appear to have made a sustained attempt to take advantage of a moment of Roman weakness.

Annius’ death suggests that there might have been conflict on the Stanegate. If the Roman army had indeed been attempting to conscript young native men, resentment may have flared into open revolt. When the attacks came, they seem to have been savage – and effective. Both the II Augusta and the XX Valeria Victrix took heavy casualties and, in the initial period of his reign, Hadrian was forced to send 3,000 reinforcements to Britain. The historian Cornelius Fronto, writing not long afterwards, in 157, lamented the killing of many soldiers at the hands of the Britons. Even after thirty years of occupation, the hill peoples of the north still hated the Roman conquest and pitched themselves against a mighty empire to be rid of it.

By 119 it was reported that the barbarians had been scattered and the province of Britannia recovered, and coins were issued which showed the figure of Britannia (looking not unlike those which used to appear on old pennies) on the obverse in a submissive pose. The war, it seems, had been won – for the moment.

Hadrian knew that Britannia required sustained attention and he determined to give it personally. When news of his impending arrival spread in 122, the commander at Vindolanda put his soldiers to work. The fort would be an excellent base for any survey of the frontier and it is likely that the imperial household had sent word ahead. Well-appointed quarters were quickly built. Spacious, with plastered and painted walls, the house appears to have been a large courtyard building. The Emperor’s interest in architecture was well known and the work done will have been finished with great attention to detail.

Once the imperial visit was confirmed, another man went to work. It was well known that Hadrian liked to make himself accessible and was happy to receive petitions from his subjects in person. At Vindolanda the writer of the following probably believed that he had a good chance of getting his complaint into the Emperor’s hands. Over twenty centuries the anger is still palpable:

. . . he beat [?] me all the more . . . goods . . . or pour them down the drain[?] . . . As befits an honest man [?] I implore your majesty not to allow me, an innocent man, to have been beaten with rods and, my lord, inasmuch as [?] I was unable to complain to the prefect because he was detained by ill-health I have complained in vain [?] to the beneficiarius and the rest [?] of the centurions of [?] his unit. I accordingly implore your mercifulness not to allow me, a man from overseas and an innocent one, about whose good faith you may enquire, to have been bloodied by rods as if I had committed some crime.

The complainant appears to have been a civilian merchant from overseas. Whatever he had supplied had been thrown away, presumably because it had spoiled or been of poor quality. But the merchant insists that, whatever the merits of the dispute, he should not have been beaten – because he was from overseas. The clear corollary is that it was appropriate to beat Britons.

It is highly unlikely that the petition was ever put into the hands of His Majesty. The scribe asked by the merchant to write out a fair copy had probably gone to see the centurions. They then sought out the poor man, took the draft from him (it was found in the centurions’ quarters) and probably gave him another thrashing for his impudence.

The petition and the incidents around it are a fascinating, very human perspective on an event of immense significance. Hadrian had arrived on the northern frontier and, as he rode through the gates of Vindolanda, history was rumbling into place. Within ten years his plans and wishes would transform the north forever, and the greatest monument to Roman power would ribbon through the Northumbrian and Cumbrian countryside.

If you find an error please notify us in the comments. Thank you!