The Encircling Sea

Historical Note

THE MUTINY OF the Usipi appears in a single source, Tacitus’ biography of his father-in-law Cnaeus Julius Agricola, legate of Britannia from AD 78–84. His account consists of a single paragraph, and serves a stylistic purpose as a brief interval before he recounts the final year’s campaigning that culminated in Agricola’s victory at Mons Graupius. The text is not well preserved, so that his sense is not altogether clear, but in broad outline the mutineers sailed from somewhere on the west coast of Britain, circumnavigated the island. They raided, but turned to cannibalism, allegedly eating the weakest among them and then others chosen by lot. Eventually they ended up on the coast of north Germany east of the Rhine, where they were killed or enslaved, and some of the slaves were sold into Roman hands and told their story. Depending on the precise dating of Agricola’s operations, the mutiny occurred in AD 82 or possibly 83.

Little is otherwise known about the Usipi. Some of them were part of a group who came into Gaul in 55 BC and were attacked and defeated by Julius Caesar. Their homeland was east of the Rhine and by the end of the first century AD they were not under direct Roman rule, although like most of the peoples beyond the frontier there was presumably some form of treaty relationship. The incident in Tacitus is the only time a unit of Usipi was raised in their own ethnic unit. We do not know the cause of the mutiny and I have embellished the brief story in Tacitus a good deal. He makes no mention of Harii in the unit, but in his discussion of the German tribes, the Germania, he claims that this tribe liked to fight at night, carrying black shields and painting their bodies, relying on the terror caused by their appearance.

Tacitus implies that all the mutineers were killed or enslaved, so it is pure fiction to have one of their stolen ships break away from the others and survive. I have also given them triremes rather than the smaller liburnians in Tacitus’ narrative. This is novel, not a history, so I have felt free to add to the meagre information we have about these years, but have always done my best to set it all within the context of what we do know about Roman Britain, the army and the wider world in this period.

Apart from this short passage from the Agricola, this novel and the others in the series are inspired by the remarkable collection of texts discovered in the excavated forts at modern Chesterholm, once the Roman army base of Vindolanda. These provide fascinating glimpses of life on the frontier of Roman Britain at the very end of the first century and the start of the second century AD. Most deal with the routine and even mundane. There are daily reports made by junior officers of the garrison, requests for leave made by soldiers, accounts of purchases and sales, of goods stored or delivered. Some of the most striking come from the personal archives of the prefect in command of the army unit stationed at the fort, dealing more with their social life than their formal duties. Thus we have a list of food, especially poultry and eggs, consumed in their household to entertain a long succession of guests. We have letters written to and by other commanders, showing the rich social life of these important men. Even more strikingly, we have letters between the wives of these prefects, who accompanied their husbands, whose tour of duty usually lasted for several years. Just as in the more peaceful provinces of the empire, these women supervised the household and the raising of their children in the manner expected of wealthy and well-born wives.

Vindolanda is one of the most remarkable Roman sites in Britain. The first fort was built there in the seventies AD. The fort from our period was the third constructed on the site. The remains visible today are of the later stone fort and the civilian settlement or vicus (a more organised version of the canabae) outside it. A level of laziness in demolishing the earlier forts when the new ones were built, combined with the water-logged nature of much of the site, created unusual conditions that have allowed the preservation of wood, leather, textiles and other material usually lost. Over five and half thousand shoes have already been found at Vindolanda, more than from anywhere else in the Roman Empire. For more information about the site and the Vindolanda Trust visit the website at http://www.vindolanda.com/.

Although less impressive as objects, the wooden writing tablets were an even more surprising and exciting find. Many are tiny fragments or illegible, but hundreds have preserved some text. Papyrus was known and used in Roman Britain, but was expensive, and much everyday correspondence and record keeping was written in ink on thin sheets of wood. Some were covered in thin wax, so that this could be smoothed down again and re-used, but these tend to be impossible to decipher since scratches from numerous different texts overlap. The most useful were the plain wood sheets, which had been rubbed with only a thin layer of beeswax to prevent the ink from spreading and were then used only once. Even so, little of the ink survives, and it requires careful analysis of the scratches made by the nibs of the stylus pens to trace the outlines of letters. Deciphering the texts and then reconstructing and understanding them is a painstaking business. More detail and many of the texts themselves can be found online at http://vindolanda.csad.ox.ac.uk/.

One of the most famous letters is the invitation from Claudia Severa, wife of Aelius Brocchus at Coria (modern Corbridge), asking Sulpicia Lepidina, wife of Flavius Cerialis at Vindolanda, to attend her birthday celebration on 11 September. This formed the basis of the plot for the first novel. All of these people are real, but until the discovery of the tablets none were attested anywhere else. It is highly likely that objects excavated in the area of the praetorium of the third fort at Vindolanda belong to Cerialis and his family, for instance a fine decorated head-piece for his horse, fine shoes belonging to children and the broken and discarded slipper that surely belonged to Sulpicia Lepidina – all of which can be seen in the excellent museum at Vindolanda.

I have used the names of these real people in these novels, but in the main their characters are invented. There are snippets of information in the texts, and I have used as much of this as possible, and tried my best never to invent anything that would jar with what we do know, but even so their appearance and characters, let alone the events that befall them, are fiction. There is certainly no hint in the text that Cerialis’ and Sulpicia Lepidina’s tour at Vindolanda was as eventful as these stories make it. However, there is room for this because we really know very little about them and even less about what was happening in Britain at the time.

Cerialis was the prefect of cohors VIIII Batavorum milliaria equitata (the ninth double-strength Batavian regiment of infantry and cavalry), which at full strength ought to have numbered over a thousand men, a quarter of them cavalry. He was a Roman citizen and a member of the equestrian order, the social class ranking below the Senate. To give an idea of scale, there were tens of thousands of equestrians compared to around six hundred senators. We do not know even the names of the majority of senators in this or most other eras, so it is unsurprising that even fewer equestrians have left any clear trace of their lives. The Vindolanda tablets give us a hint of the vast amount of writing and record keeping in the Roman world, but also remind us that it takes exceptional good luck for such material to survive into the modern era.

An equestrian prefect in command of an infantry cohort or cavalry ala of auxiliary soldiers was a man of considerable importance locally and in his home community. Substantial property was required for a man to be registered as an equestrian in the census. However, by AD 100 there were at least 350 posts as commander of auxiliary units. We think that the average time spent by these officers with a unit was about three years (although there is a lot of guesswork in this), which shows how many of these officers there were and why we should not be surprised that such a small proportion appear in our sources. Cerialis is one of the best documented, and we really know little about him.

His name suggests that he, like the soldiers he commanded, was a Batavian, a Germanic tribe living in what is now the Netherlands. In AD 70 the Batavians revolted as part of the wider civil war and disturbances following the deposition and suicide of Nero. The Civil War was won by the Emperor Vespasian, whose family name was Flavius. The commander he sent to defeat the Batavian rebels was Petilius Cerialis, a task he completed by AD 71. We do not know what happened to Julius Civilis, the rebel leader, for the sources for the end of the rebellion are poor. The combination of Flavius and Cerialis as names strongly suggest that either the prefect at Vindolanda or his father were granted Roman citizenship during or after the rebellion. Presumably the man who was rewarded in this way either remained loyal to Rome (and specifically Vespasian) throughout the revolt, or changed sides early enough to be treated well by the new emperor. Although it is possible that the prefect at Vindolanda was middle-aged, old enough to have attracted attention during the disturbances of AD 70, it seems much more likely that he was the second generation of the family to be a Roman citizen.

Equestrian status came from property, while a command in the army came from influence with those in high places. An old treaty with the Batavians stipulated that they should not pay tax to Rome, but instead provide soldiers for the auxilia, but that these men would serve under their own aristocrats. The appearance of Cerialis at Vindolanda shows that this continued to be the case after AD 70, although there are hints that it ceased during the course of the second century AD. There is a hint that Cerialis came from the royal bloodline of the tribe, like Civilis.

Sulpicia Lepidina is only known through the tablets and it is impossible to say anything definite about her family. The daughter of a senator is attested as the wife of an equestrian prefect on the British frontier later in the second century AD, so such marriages did occur, although they were rare. Neither is anything known about Brocchus, Claudia Severa, Rufinus and other names from the tablets who have become characters in the story. Being a Roman citizen and an equestrian did not require any Italian, let alone Roman, blood and such people came from all over the empire. In language and culture they were primarily, sometimes exclusively ‘Roman’, and they were Roman in law. That did not mean that, like Cerialis, they might also be part of a different ethnic tradition.

Claudius Super is almost a character from the tablets. There is a man who appears to have equestrian status and also to be a centurio regionarius, but he is Clodius Super. When writing the first novel I did not check my notes carefully and it was only after it was finished that I realised I had turned him into a Claudius rather than Clodius. By then it was too late to change, so Claudius he stayed. Some equestrians served as centurions in the legions rather than prefects of auxiliary units, whether because they wanted to spend a longer time in the army, could not afford the lifestyle of a cohort commander, or did not have friends influential enough to secure them the more senior post.

The name Crispinus appears in the tablets and he may be a tribune with a legion, but the character is otherwise an invention. Neratius Marcellus was the governor or legate of Britannia at this time and is also mentioned in the tablets. The poet Martial wrote to his friend Quintus Ovidius when the latter was about to accompany a friend going to govern Britain, and there is a good chance that this was Marcellus. He was a poet and philosopher (possibly a Stoic), and may have been exiled under Vespasian and later recalled.

Ferox and Vindex are wholly invented, although a later tombstone records a Brigantian soldier in the Roman army who was the son of someone named Vindex and in my imagination this is our man. It was common for the Romans to take boys from defeated peoples, educate them and grant them citizenship, sometimes even equestrian status, and make them army officers. This was part of the process of absorbing former enemies. It is not directly attested as happening after the defeat of the Silures (who lived in what is now South Wales) in the seventies AD, but is perfectly plausible.

Legionary centurions are often still depicted as sergeant-major types, tough men who rose through the ranks through sheer talent after long and hard service. The stereotype is a powerful one, but is not based on good evidence. Some centurions were commissioned directly from civilian life without any prior military experience, including a minority of equestrians. A few proudly tell us on monuments, most often their tombstones, that they joined as ordinary soldiers, but these are very few. The vast majority of recorded centurions give no indication of having served in lower ranks. Others served only as junior officers before being elevated to the centurionate. It is important to remember that a centurion had to be well educated, literate and numerate, for the army ran on the written word. These men were professional officers, rather than professional soldiers, and their pay and conditions were substantially higher than those of men in the ranks. It is likely that even junior centurions earned more than ten times the salary of an ordinary soldier, and most were probably drawn from the local gentry and well-off families of Italy and the provinces.

Centurion was not a rank, but a grade of officer, with a considerable range in status and responsibilities. There were some sixty centurions in a legion, and between six and ten in an auxiliary cohort. Many served away from their units on long- or short-term detachment. A cohort at Vindolanda in the nineties AD had only one out of its six centurions at the base. Two more were with the biggest detachment of the cohort at Corbridge, but the others were widely dotted around the province. The post of centurio regionarius or regional centurion was one of the ones that took these officers away. It is attested in Britain and elsewhere, particularly in Egypt, and was probably common. These men acted as the representative of Roman authority in an area, their role a mix of civil and political as well as military. In Egypt there is good evidence for them acting as policemen and investigating crimes.

Before Hadrian’s Wall

The story occurs at the start of the reign of Trajan, whose successor Hadrian came to Britain and ordered construction of Hadrian’s Wall around the year AD 122. Our sources have little to say about major events in Britain under Trajan, although there is talk of major conflict, which may well have prompted the decision to build the Wall. The fort at Vindolanda (modern Chesterholm) lies a few miles south and within sight of the Wall and clearly was incorporated within the network of garrisons serving it.

Although Julius Caesar had landed in Britain in 55 and 54 BC, no permanent Roman presence was maintained, and it was not until AD 43 that the Emperor Claudius sent an invasion force across the Channel. In AD 60 Boudicca’s rebellion devastated southern Britain, but after her defeat there is no trace of any serious resistance in the Lowlands. This is not true of northern Britain, which was garrisoned by substantial numbers of troops for the remaining three and a half centuries of Roman occupation.

In AD 100 few would have guessed that the Romans would stay for so long. Their presence in the north was more recent, for it was mainly in the seventies and eighties AD that this area was overrun. During this time Roman armies marched far into the north of what would become Scotland, while a naval squadron for the first time circumnavigated Britain, confirming that it was an island. An entire legion – one of the four then garrisoning the province and one of twenty-eight in existence – built a base at Inchtuthil in Perthshire, the biggest site in a network of garrisons on the edge of the Highlands. Around the same time, a system of observation towers along a military road was constructed along the Gask ridge.

All of this activity, which to a great extent is known only from the archaeological remains, makes clear the Romans’ intention to occupy this region more or less permanently, but in the late eighties AD priorities changed. The Emperor Domitian, faced with serious trouble on the Danube, withdrew Legio II Adiutrix from Britain and did not replace it. It is probable that substantial numbers of auxiliaries were withdrawn at the same time, so that the provincial garrison was cut by at least a quarter. Inchtuthil and many of the other bases were abandoned, and the same thing happened a little later to the remaining sites and the Gask ridgeline. No Roman base was maintained north of the Forth-Clyde line, and soon the northernmost outpost was at Trimontium or Newstead.

Several forts were maintained or built close to what would one day become the line of Hadrian’s Wall. A couple of years after our story, a proper road running between Carlisle and Corbridge was constructed and more forts and smaller outposts added. Today the road is known by its medieval name, the Stanegate or ‘stone road’, and archaeologists continue to debate its composition and purpose. By about AD 106 Newstead was abandoned in another withdrawal. Our paltry literary sources make no mention of any of this, so it left to us to guess from the archaeology just what was going on.

A novelist has more freedom, and once again I have done my best to reconstruct these years for our purposes in a way that never conflicts with any hard evidence. At the very least I hope that these stories tell of things that could have happened. Something made the Romans station significant numbers of troops in this area at the end of the first century AD, and then made them increase these numbers and develop the deployment along the Stanegate just a few years later. All of the forts mentioned in the story existed and were occupied in AD 98. Syracuse is an invention, but typical of the many small outposts set up by the Roman army as needed. I see it as a predecessor to the excavated sites at Haltwhistle Burn and Throp, which were built alongside the Stanegate, although these were stone structures and larger than the fictional Syracuse. In the late first century AD most of the structures built by the army in Britain were in turf and timber. Some sites were being rebuilt in stone, and in the second century this became ever more common.

One of very many mysteries surrounding Hadrian’s Wall are the defences along the Cumbrian coast. These were part of the initial design of the Wall and consisted of towers and fortlets matching closely to those on the line of the Wall itself and at similar intervals. Many perch on the cliffs just above beaches and all are very close to the sea. All that is missing is a linear barrier like the Wall itself. We have no evidence explaining why this line of outposts and forts was felt necessary. Later in the second century AD many parts of the system were abandoned, but forts – and just possibly some towers and fortlets – remained in use throughout the life of the Roman province. The most obvious explanation of all this effort is that there was a threat from the sea. Later, this either diminished or was dealt with by other means, allowing the coast to be secured by fewer troops.

Alauna (modern Maryport) now boasts an excellent museum (www.senhousemuseum.co.uk) with a particularly fine collection of inscriptions. The fort visible today and most of the collection date to after our story. There are hints of Trajanic activity on the site, but little is really known. The alignment of the later fort with the road suggests that an earlier base was built in a slightly different position. At Aballava (Burgh by Sands) there was a Roman watchtower on high ground overlooking the lowest fords on the River Solway. At some point in Trajan’s reign, this was demolished and a fort for an entire auxiliary unit built on the site. Around the same time, another Roman base was constructed on lower ground to the west. Little is known about it, but it is claimed that timber structures were discovered, suggesting at least semi-permanent occupation and not simply a marching camp used for a few nights or weeks. In my imagination, this is the base used for the legate’s manoeuvres and then kept in occupation for some time afterwards following the raid in our story and the ongoing threat of attacks from the sea.

Hibernia, the Highlands and Islands

Much of the action in this story takes place well outside the province of Britannia, but because of the sources for the history and society of these regions I have been deliberately vague about precise locations. Hibernia/Ireland was never occupied by Rome. Merchants from the Roman Empire were regular visitors there, and there was a fair bit of diplomatic activity. An exiled Hibernian prince came to Agricola asking that the Romans restore him to power by force, but the governor decided not to intervene. There is no clear evidence for Roman troops ever being sent to Ireland. However, a diplomatic escort of the sort imagined in the book would be most unlikely to leave any trace so we need not worry too much about that. Such things did occur elsewhere beyond Rome’s frontiers, so it is not implausible in itself.

In recent decades far more Iron Age sites have been discovered in Ireland, but it is fair to say that we still know a good deal less than we should like about life there during this period. There are echoes of it in later literature, most notably the Ulster Cycle with its chariot-riding heroes, but it is very hard to say how much real history lies in the stories. The ‘Place of Kings’ in the story is inspired by Tara in County Meath, which, like Navan Fort in Armagh, figures in the later poetry. It is a huge complex of monuments, some very ancient even in the Iron Age, and how it was used and by whom is hard to reconstruct. I have taken tribal names from the Greek Geographer Ptolemy, but it is hard to say how accurate his information was.

The same is true of all our literary sources for the Highlands and islands of Scotland. The tower that Ferox and the others take and hold is one of those remarkable dry-stone buildings often known as brochs, of which some of the most splendid are on Shetland. The name came from Norse and goes back to a time when it was thought that these were forts built by the Vikings. Now it is clear that they were much earlier, and part of a wider style of building that appeared in islands and on the western coast of Scotland and occasionally further afield. In some cases it is hard to tell from the existing remains whether the structure was originally lower, forming what archaeologists would call a complex roundhouse. In either case these were buildings for more than a single family, and were strong statements of power. Yet they do not seem designed primarily for defence, and there is a good deal about them that we simply do not understand.

The island occupied by the pirates is fictional, although there are examples of broch-towers built out on islands in lakes. Similarly the idea of a woman warrior teaching young heroes how to fight comes from the Irish poems, which seems to place it all somewhere off the coast of Scotland. The historian in me considers it unlikely; we might remember Greek heroes supposedly going off to be instructed by a centaur, so such romantic invention is a feature of other heroic myths. The novelist is quite happy to take a good story and use it, at least as long as it cannot be proven to be nonsense.

The Roman Army

This is a vast subject, but it is worth making a few points for those new to the topic. In AD 100 the Roman army consisted of twenty-eight legions – two more would soon be added by Trajan – each with a paper strength of some 5,000 men. Each one was divided into ten cohorts of heavy infantry and had a small contingent of some 120 horsemen. Legionaries were Roman citizens. This was a legal status without any ethnic basis and by this time there were over four million Roman citizens scattered throughout the empire. We may think of St Paul, a Jew from Tarsus in Asia Minor, but a Roman citizen and entitled to all the legal advantages that brought.

Supporting the legions were the auxiliaries who were not citizens, but received citizenship at the end of their military service. These were organised as independent cohorts of infantry and similarly sized cohorts of cavalry. There were also the mixed cohorts (cohortes equitatae) like the Batavians, which included both infantry and cavalry in a 4 to 1 ratio. Legionaries and auxiliaries alike served for twenty-five years. Most were volunteers, although conscription did occur and was probably especially common with some auxiliary units.

We know a good deal about the Roman army, about its equipment, organisation, command structure, tactics, ranks and routine, although it must be emphasised that there are also many gaps in our knowledge. As a historian, it is my duty to stress what we do not know, but a novelist cannot do this and must invent in order to fill in these gaps. Some aspects of the depiction of the Roman army in these books may surprise some readers, but often this will be because some of the evidence for it is not well known outside academic circles. I have invented as little as possible, and always done my best to base it on what we do know. As an introduction to the army, I am vain enough to recommend my own The Complete Roman Army, published by Thames and Hudson. I would also say that anything by the late Peter Connolly is also well worth a look. Once again for more specific recommendations, I refer readers to my website – adriangoldsworthy.com.