Boring History for Sleep - The Forgotten Origins of Medieval Britain 🏰⚔️ | Boring History for Sleep

Episode Date: September 2, 2026

Medieval Britain did not emerge from a single kingdom or a single people. Its origins were shaped by Romans, Britons, Anglo-Saxons, Celts, Vikings, and countless communities whose stories were gradually woven together into the Britain of the Middle Ages.What happened after Roman rule collapsed? Who were the peoples that fought, migrated, settled, and built new kingdoms across the island? And how did centuries of conflict and cultural change create the foundations of medieval Britain?A calm journey through the forgotten beginnings of medieval Britain, exploring ancient kingdoms, migrations, invasions, battles, and the people who shaped the island long before England became the country we recognize today.Boring History for Sleep — Soft stories about the forgotten foundations of history.

Hey there, history lovers. Today we're heading to the cold western edge of Wales, where a petrified forest sleeps under the waves of White Sands Bay. Ancient tree stumps poke out of the sand like the teeth of some long-dead giant, and that's the perfect mascot for what we're about to dig into, because right here sits a 200-year black hole in British history. Rome packs up, the lights go out, and when they flicker back on, somehow there are kings, saints, and a whole new medieval whales. What happened in between? wilder than your textbook ever told you. Burned out Roman cities, Celtic port towns trading Baltic amber, mad emperors, Trojan refugees fighting actual giants, and crumbling forts where

chieftains sipped wine from Mediterranean amphoras, while the rest of Europe was having a rough century. Forget everything you think you know about Dark Age Britain. It was loud, bloody, brilliant, and it built the foundations for King Arthur, Merlin and even Tolkien's elves. Yes, those elves. So before we dive in, smash that like button if you're ready for the deep dive, and drop a comment telling me what city you're watching from. Let's go. There is a particular kind of beach in Pembrokeshire that only exists when the weather has been in a bad mood. Most of the year, White Sands Bay looks like a fairly normal stretch of Welsh coastline, all golden sand and cold green surf and dog walkers in waterproof jackets, pretending it isn't raining. But every few winters, a serious storm rolls in off the

Atlantic, the kind that strips the dunes down to their bones and drags whole metres of sand back out to sea. And when the wind finally drops and the tide retreats, something genuinely strange appears at the waterline, hundreds of black twisted stumps poking up out of the wet beach in clusters and rough lines, looking as more than one visitor has noted exactly like the petrified teeth of a buried giant who fell asleep on the job. These are the remains of an ancient forest that grew here roughly 7,000 years ago, back when the sea was much further out, and the sea was much further out, and Britain was still smugly attached to the European mainland. Mesolithic hunters walked beneath these oaks and pines. Their footprints have even been found in the peat layers underneath,

preserved alongside the prints of red deer and aurochs, the giant wild cattle that once stomped around northern Europe before humans, with characteristic enthusiasm, hunted them into extinction. The forest itself died not in some dramatic catastrophe, but in the slow, patient way that geology tends to prefer. As the great ice sheets up north melted, sea levels rose, and the brackish water crept inland year by year, roots drowned, trunks rotted, eventually a thick blanket of peat formed over the dying woodland, sealing it from oxygen and pickling it for the future to find. The result is a fossil that is technically not a fossil, since the wood has not turned to stone but rather has been preserved as wood, blackened and waterlogged, but still recognisably the same trees that grew there

when the pyramids had not even been imagined. Touching one of these stumps is, in a quiet sort of way, one of the more disorienting things you can do on a British holiday. You're putting your hand on something that watched the end of the last ice age and the beginning of agriculture, which is a lot of historical weight for a piece of wet driftwood. The drowned forest is the deep prologue to the story of this coast, the reminder that, even before there were Romans or Christians or kings, this stretch of land had been changing shape and changing inhabitants for millennia.

But the part of white sands that really matters for our purposes is much younger, and it sits just up the slope from the petrified trees, tucked behind a low ridge of grass-covered dunes. That is where, in the early years of this century, archaeologists started pulling something out of the ground that they had not quite been expecting. Bones. Lots and lots of bones, neatly arranged in long rows of stone-lined graves, all carefully oriented so that the dead would be facing the rising sun on the morning of the resurrection. This was the cemetery of Capel Padrig, the chapel of St Patrick, a small but persistent Christian sight that had sat above the bay

from at least the 6th century until the 11th, after which it was apparently forgotten so thoroughly that for the better part of a thousand years, nobody was entirely sure where it had been. The chapel is named, naturally, for Patrick himself, the same Patrick who would later become the patron saint of Ireland, although in his lifetime he would have been considerably less famous and considerably more concerned with not being kidnapped again.

Local tradition held that Patrick had launched his mission to Ireland from this very beach, gazing west across the water, before stepping onto a boat and disappearing into the spray of evangelical history. Whether he actually stood on this exact patch of sand is one of those questions that will never be settled, but a chapel was built here and it endured, generation after generation, on a tiny strip of land between the dunes' and the cliffs. The structure itself was small, probably no larger than a modest country church, with thick stone walls and a slate roof, but the cemetery surrounding it grew steadily for 500 years until it held something in the region of 200 or more burials, possibly many more. The dead included men, women, and a striking number of children, some of them only days old when they were

placed into their stone boxes and covered with capping slabs. The graves themselves are what archaeologists call cyst burials, which is essentially a fancy word for a coffin made of flat stones rather than wood. You dig a rectangular pit, you line the sides with carefully chosen slabs, you place the body inside, and then you cover the whole thing with another large stone or two. The result is a small underground room for the dead, weatherproof and surprisingly sturdy. Sist burials were extremely common across the Christian Celtic world during the early medieval period, and the ones at Capel Padrig are textbook examples. What makes them remarkable is the sheer number of them and the gentle care with which the bodies were arranged. Adults were laid out flat on their backs with their hands

folded across their chests or by their sides. Children were sometimes buried with small tokens, a single bead or a fragment of polished bone. Infants were placed in tiny cysts barely larger than a shoebox, which is one of those archaeological details that no amount of scholarly distance can quite soften. The fact that we know about any of this is, in a slightly grim way, thanks to climate change. The cemetery had been buried beneath the dunes for centuries, gradually forgotten by everyone except a few local antiquarians who had vague suspicions about the lumpy shape of the ground. Then, in the early 2000s, a series of unusually fierce winter storms started peeling the dunes back like a knife through wet pastry, and bones began to appear on the beach. Beachgoers

stumbled across human skulls in the sand, which is not the standard Pembrokeshire weekend experience, and which prompted, unsurprisingly, a number of frantic phone calls. Welsh archaeologists realised quickly that the entire cemetery was in the process of being eaten by the sea, and a rescue excavation was launched to recover what could still be saved. Volunteers came from all over the country, working in shifts during the brief windows between tides, racing to lift and document burials before the next storm took them out for good. It was archaeology as emergency surgery, and the results have completely transformed our understanding of what was happening on this coast in the centuries after Rome.

For one thing, the people in these graves were not isolated peasants scratching out a meager living at the edge of a forgotten frontier. Bone analysis suggests they were reasonably well fed, comparatively healthy by the standards of the period, and tied into a wider community that took burial seriously enough to invest considerable labour in their final resting place. Carving and hauling stone slabs is not light work, and the fact that hundreds of these graves were dug suggests a community that had time, organisation and a strong cultural conviction that the dead deserved a proper send-off.

The burials also spanned centuries without any major interruption, which means that whoever was running this site, whether a small monastic community or a network of local families, was able to maintain continuity through the entire so-called dark age and well into the medieval period, so much for the idea that everything just collapsed when the Romans went home. What really surprised the excavators, though, was the deeper context of the place. The chapel was not just a remote outpost. It was part of a tightly woven coastal world that looked outward across the water,

rather than inward toward the rest of Britain, to understand why all you have to do is glance at a map. Bruchshire juts out into the Irish sea like a stubborn elbow, and the stretch of water between Whitesons and the Wicklow Coast of Ireland is barely 50 miles wide. With a decent wind, a small wooden boat could cross it in a single day. Seafarers had been making that crossing since prehistoric times, and by the early medieval period it was effectively a busy ferry route, carrying merchants, missionaries, mercenaries, and exiled princeses back and forth in numbers that modern historians are only now starting to fully appreciate. Look at the place next to

names on this coast, and the Irish influence becomes obvious. St. Brinac was an Irish missionary who set up a major church a short distance up the coast. St. Govan, who lived in a tiny chapel built into a cliff face down on the South Pembrokeshire coast, was reportedly Irish as well, although the tales about him have grown so layered over time that distinguishing fact from cheerful invention is essentially impossible. St. Justinian, who was eventually killed by his own monks in one of the more uncomfortable workplace incidents of the 6th century, gave his name to the very headland that closes off White Sands Bay to the south. The whole landscape is studded with reminders that this was not a

Welsh coast facing England. It was a Celtic coast facing Ireland, and the people moving across it spoke at least two languages, often three, and recognised each other as members of a shared sea-going world. That sense of a maritime culture is reinforced by what archaeologists call Ogham Stones, which start appearing along this coast in concert. considerable numbers from the 5th century onward. Ogham is an early Irish writing system that looks to the uninitiated eye, like a series of notches and slashes carved along the corner edge of a stone, somewhere between a tally mark and a fence post.

It was used to record names and short inscriptions, usually marking burials or asserting territorial claims. The fact that Ogham inscriptions are found right across Pembrokeshire, often with bilingual Latin translations carved alongside them, tells us that Irish-speaking communities were not just visiting this coast occasionally. They were settling here, intermarrying with the local population, and producing literate elites who could express their identity in two scripts at once. This was, by the standards of the post-Roman West, a fairly sophisticated cultural arrangement.

It also tells us something rather important about how the post-Roman world actually worked. The conventional story, the one that most of us absorbed from school textbooks somewhere around the age of nine, is that when the Roman legions marched out of Britain in the early 5th century, civilization more or less switched off. The lights went out, the roads cracked, the cities were abandoned, the population fled into the hills, and everybody spent the next several centuries hiding from Saxon Raiders in muddy huts. It is a story with a satisfying narrative arc,

and it has been told so often that it has acquired the polish of received wisdom. The trouble is that, on the western coasts at least, it is almost entirely wrong. While the eastern lowlands of Britain were certainly going through a rough time with Saxon, Anglian and Jutish settlers carving up the old Roman estates and the surviving British population

either retreating, assimilating or fighting a long losing war, the western coasts simply pivoted. They turned their backs on the troubled mainland and faced the sea instead. What emerged was something genuinely new. Christianity, which had been a minority religion in late Roman Britain,

suddenly became the dominant cultural framework along these coasts, but it was a Christianity heavily flavoured by Irish influences and shaped by the realities of small scattered communities. Latin remained the language of learning, but it was a Latin spoken with Welsh and Irish accents, and used to compose poetry rather than imperial decrees. Trade links did not collapse, they shifted. Instead of importing wine from Roman Gaul along well-paved roads, the elites of post-Roman West Britain imported it directly from the Medellings. Mediterranean by sea. Pottery from North Africa and Byzantium turns up in remarkable quantities

at high-status sites in Wales, Cornwall and South West Scotland. Evidence that ships were sailing the long route from the Aegean to the Bristol Channel and finding eager customers at the other end. The economy of this region was not in collapse. It was in transformation, and white sands sat right at one of the busiest junctions of that transformation. There is something almost moving about standing on that beach today, knowing what is buried just behind you. Beneath the dunes, the bones of children who lived and died in a world we struggle to imagine, in front of you the petrified roots of a forest that watch the rise of farming. Across the water the green shadow of Ireland on a clear day,

which is the same view a sixth-century mourner would have looked out at after laying a brother or a daughter into the cold stone of a cyst grave. The chapel itself is gone, or rather it is reduced to a few foundation stones tucked into the grass, but the place still has the quality of a hinge, a point where one age folded into another. Roman Britain, with all its baths and basilicas and bureaucracy, ended somewhere on this coast, and the new world that grew up in its place, the world of saints and seafarers and small Christian kingdoms, started here too, more or less in the same patch of sand. 2,000 years of history concentrated into a windswept beach where, on a quiet day, you can hear nothing but the surf and the occasional outraged seagull.

20 miles or so south down the same coast, that emerging maritime world becomes even more visible in the form of a small, awkward and frankly unlikely island called Gateholm. The name itself is a clue, since it comes from the old Norse for Goat Island, presumably bestowed by Viking sailors who arrived a few centuries later, and noticed, with the brisk practicality of all seafarers, that the place had goats on it. But long before the Norse showed up, Gateholm had already been inhabited by humans for somewhere in the region of 2,000 years, which by the standards of British archaeology is a remarkably long innings.

Today the island is mostly empty, a flat-topped slab of rock about 300 metres long, and only a few dozen wide, separated from the mainland by a treacherous ribbon of seaweed-covered boulders that becomes briefly walkable at low tide. Cross that causeway when the timing is right and you find yourself standing on what was, for many centuries, one of the more curious places in early Britain. The first thing you notice is that the entire surface of the island is covered in archaeology.

Walk along the spine of Gateholm and you can trace the outlines of what were once dozens, possibly more than a hundred small buildings. Most of them are roughly rectangular, with low walls of stacked stone and turf, and they are packed in tightly along narrow lanes that run more or less the length of the island. The effect when you stand among them is uncanny. It is as though someone tried to fit an entire small town onto a rock that was clearly never quite big enough for the job.

Estimates of the population vary, but some archaeologists have suggested that several hundred people may have lived on Gateholm at its peak, which on an island this size would have been distinctly cosy. Privacy, in any modern sense, would have been a theoretical concept, and the smell, particularly in the warmer months, must have been an aggressive presence in its own right. When the first proper archaeological survey of Gateholm was carried out in 1910, the investigators looked at the dense cluster of small buildings, counted the rough outlines, and reached what was at the time the obvious conclusion.

This must, they declared, be an early Christian monastery. The Edwardian habit of identifying any unexplained archaeology in Wales as a monastery was a powerful one, and Gate Home, with its remote setting and its tightly packed cells, fit the template comfortably. The interpretation lasted for most of the 20th century, gently absorbed into guidebooks and local histories, even though there were one or two slightly inconvenient details that did not quite line up. There was, for instance, no obvious chapel, which is the kind of thing you would generally expect a monastery to feature fairly prominently. There was no clear cemetery either, although early medieval cemeteries are notoriously good at hiding from archaeology. who are not specifically looking for them.

But the real problem with the monastery theory only became clear when modern excavations started turning up the kind of finds that monks did not usually accumulate. Mediterranean wine and oil amphury started appearing in the soil, not in ones and twos, but in quantities that suggested regular shipments. Alongside them came North African tableware, eastern Mediterranean glass, Baltic amber beads and Irish bronzework. evidence of metalworking emerged, not just casual repairs but enough slag and crucible fragments

to suggest some kind of organised craft activity. None of this fit the profile of a small ascetic community living a life of prayer and barley bread. It fit the profile of a busy, prosperous, deeply connected trading settlement. The current best understanding of Gate Home, based on decades of revised excavation and reinterpretation, is that it was not a monastery at all. port. More specifically, it was probably a periodically intensified market site, occupied year-round by a core community but expanding dramatically during the trading season when ships put in from across the western seaways. The location is suspiciously perfect for the role. Gaitholm sits at the northern end of St Bride's Bay, with deep sheltered water just offshore and easy access to

the open sea. Sailors approaching from Ireland, Cornwall or further afield could see the island from miles away, and once they had landed they were beyond the reach of any inland authority that might have wanted to tax them. It was in effect an early medieval free port, and the goods that came through it suggests it was doing rather well for itself. What kind of trade was passing through? Imported wine and oil from the Mediterranean almost certainly. The Amfrey do not get to Pembrokeshire by accident. Fine ceramics for the tables of local elites, presumably traded for British goods that the Mediterranean wanted in return. A list of plausible British exports includes tin from Cornwall, lead and silver from Wales,

slaves, hunting dogs, highly prized across the ancient world, and possibly even quality woolen cloth. Amber, coming in from the Baltic, suggests that Gate Home was tied into a trade network that ran clear across northern Europe and probably involved multiple intermediaries. Irish bronze work, with its distinctive interlis patterns and high-quality casting, points to direct contact with the powerful tribal kingdoms across the water. The picture that emerges is of a small, frankly chaotic island settlement that was, in its own peculiar way, a hub of international commerce.

The lifestyle on gate home must have been an interesting one, in the sense that the word interesting has historically done a great deal of polite work. Fresh water was a constant problem, since the island has no springs of its own. Inhabitants would have had to collect rainwater in cisterns or carry barrels across from the mainland at low tide, neither of which is anyone's idea of a relaxing morning. Storms, which roll in across the Atlantic with depressing regularity in the winter months, would have battered the houses mercilessly, and probably made certain parts of the island uninhabitable for weeks at a time. The narrow lanes between buildings would have funneled

the wind in unpredictable directions, and anyone living in the most exposed houses on the western edge would have spent a substantial portion of their lives damp, cold and slightly deafened by the noise of the surf. Whatever appeal Gateholm had as a residential address it was not based on creature comforts, what it offered instead was access, access to the sea, to the trade networks that flowed across it and to the wealth that those networks produced. For a small farming family on the mainland, life in the early medieval period was a grinding cycle of subsistence agriculture, punctuated by the occasional raid or famine. For a family on Gate Home, life was no easier in physical terms, but it offered the chance to handle goods that most

people in Britain would never see in their lifetimes, to learn languages and customs from across the known world, and to occupy a position in society that was, for the period, quite remarkable. There is a reasonable argument that Gate Home functioned almost like a guild city and miniature, a place where the rules were a little different, where outsiders were not just to be. tolerated but actively welcomed because they were the entire point of the enterprise. The continuity of occupation at Gaitholm is what really sets it apart. Most British archaeological sites have a discrete biography. They get founded, they thrive for a period, they decline, and they are abandoned.

Often after some specific event or social shift, Gaitholm just keeps going. The earliest evidence of human occupation goes back to the late Bronze Age, around 1000 BC, when the first round huts were built on the island. There is occupation through the Iron Age, with characteristic roundhouses and evidence of farming on the surrounding mainland. The Roman period sees the site continuing to develop, although it sat well outside the formal Roman administration of the province. And then, when Rome withdraws and the rest of Britain enters its long, complicated transformation, Gate Home not only survives but apparently flourishes, with the densest phase of building dating to the 5th, 6th and 7th centuries. That is a continuous run of around 2,000 years on a single

rock, which suggests that whatever Gateholm offered, it offered consistently, and that successive generations of inhabitants found ways to adapt without ever quite needing to leave. What finally ended the settlement is not entirely clear. By the 8th century, the trading networks that had supported it were starting to shift, with the rise of new political powers and new economic centres. Viking raids, beginning in the late 8th and early 9th centuries, would have made an exposed island settlement an obvious target, and although there is no direct evidence of a Norse attack on Gateholm, the broader pattern of disruption along the western coasts probably eroded its viability as a trading hub. The Norse name suggests they at least visited, even if they did not settle,

and goats on a small island would have made a fine raid target for anyone passing through. Sometime in the early medieval period, the lights at Goulders, Gateholm went out for the last time, and the buildings collapsed slowly back into the turf, leaving only their stubborn stone foundations to puzzle Edwardian archaeologists a thousand years later. Gateholm closed sometime in the early medieval period, its buildings collapsing slowly back into the turf, but the world it belonged to was still very much alive, and that world had a distinctive way of remembering its own beginnings. That is the historical picture. The medieval Welsh themselves, however, had their own story about how the British Isles got started,

and it was a story considerably more colourful than archaeology will ever be able to provide. According to the legend that became firmly established in Welsh and English imagination from the 12th century onward, Britain was not founded by Celtic tribes drifting in from continental Europe. Britain was founded by Trojans, specifically it was founded by a young exile named Brutus, the great-grandson of the Trojan Prince Aeneas, who had himself escaped the burning ruins of Troy and gone on to found eventually, the line that produced Rome. The genealogy is impressively grand if you do not look at it too closely, and looking at it too closely was not really the medieval style. The story begins, like so many of these stories do, with a domestic accident that escalates rapidly. Brutus, growing up in Italy and the generation after Aeneas, was apparently a fine young man with a promising future.

Then he went hunting one day, fired an arrow at what he thought was a stag, and managed to hit his own father instead. This was treated as exactly the kind of awkward incident that Italian noble families could not gracefully ignore, and Brutus was promptly exiled. He gathered up a band of fellow Trojan exiles, men whose families had been wandering since the original disaster of Troy, and who were apparently still up for one more migration, and set off to find somewhere new to live. There followed a long sea voyage with all the standard ingredients, including stops at Greek-controlled islands, the liberation of fellow Trojans who were being held as slaves, naval battles, narrow escapes, and at one notably mythical moment, a vision from the goddess Diana. The vision was helpful in a vague sort of way. Diana told Brutus that there was a great island in the western ocean currently uninhabited

except for a few unspecified problems and that he was destined to settle it. Diana was, as goddesses sometimes are slightly economical with the truth. The island was indeed there, and it was indeed lush and fertile, but it was not in any sense uninhabited. It was, in the original legend, populated by giants. Not metaphorical giants, or particularly tall men, but actual mythological giants of enormous size and unpleasant disposition, supposedly descended either from fallen angels or from the line of Cain,

the same Cain who had murdered his brother in the Book of James. Genesis. Either pedigree placed them firmly on the wrong side of the cosmic moral ledger, which conveniently meant that the Trojans could fight them with a clear conscience. The giants were big, strong and not noted for their tactical sophistication. The Trojans were smaller, fewer, but disciplined, organized, and led by Brutus, who was apparently a talented battlefield commander when he was not accidentally shooting his own father. The decisive encounter of the war was, almost inevitably, a single combat. The giant king of the island was called Gog-Mag.

Sometimes split into the two figures Gog and Magog in later versions of the story, but originally a single, 12-cubit nightmare of a man, give or take a cubit depending on which medieval author you're reading, and how impressive he wanted his giants to be. A cubit was the length from elbow to fingertip, so 12 cubits worked out to somewhere in the region of 5.5 metres-tort. call, which is unreasonable, but by the standards of giant law, not even particularly excessive. Gogmagog issued a challenge, as giants in these stories are obliged to do, and Brutus's

strongest captain, a man named Corinius, accepted. Corinius was famous for his physical strength, his enthusiasm for hand-to-hand combat, and a personality that we might today describe as direct. The wrestling match between Corineas and Gogmog took place on a cliff somewhere in the southwestern part the island, traditionally identified with the coast of what is now Cornwall, although the precise location varies depending on which local guidebook you consult. Both fighters gave it everything they had. Gogmogg reportedly cracked three of Corineus's ribs in the opening exchanges, which was apparently a serious mistake, because it made Corineus furious. Corrinius seized the giant by the waist, hoisted him onto his shoulders despite his injuries, and ran with him to the edge of the cliff.

There he flung Gogmog out into the air and watched him play. lummit down onto the rocks in the sea below, where he was, as the chroniclers tactfully put it, dashed to pieces. The spot was remembered for centuries afterward as Lamb Gomagot, or Gogmog's leap, although the precise cliff in question has been claimed by half a dozen different coastal towns over the years, all of which would naturally like the legend for themselves. With Gogmagog dead and the rest of the giants apparently demoralized by the loss of their king, the war was effectively over. The Trojans took possession. of the island, and Brutus, displaying the modesty typical of legendary founders everywhere,

named it after himself. The land was called Britain, and its people would be the Britons, and they would, according to the medieval imagination, retain a direct genealogical link to the heroes of Troy for the rest of their history. Brutus established a capital somewhere along the River Thames and called it Troa Nova, or New Troy, which was eventually corrupted Trinnovantam, and then, much later, renamed Lundane by some unromantic newcomers, after which it would settle into the name we know today as London. The renaming was apparently an enormous loss of poetic dignity, but at least nobody had to keep explaining where the original Troy was on the map.

Brutus had three sons, and when he died, the island was divided among them in a tidy fashion that medieval audiences appreciated, even if it did slightly oversimplify the actual geography. The eldest son, Lucrinus, received the central and southern portion of the island, the area we would call England, which he ruled from his own. father's capital. The second son, Camber, was given the western part, the mountains and valleys that would become Wales named after him. The third son, Albanactus, was sent north to rule the territory beyond the Fourth and Clyde, which became known as Albany, an old name for Scotland that

survives in places like the Duke of Albany's title and the city of Albany, New York, named eventually by colonists who liked their British history compact and respectable. Three sons, three kingdoms, one tidy origin myth, all of it neatly in place to explain the political geography of medieval Britain to anyone who would sit still long enough to listen. None of this, of course, actually happened. There were no Trojan refugees, there were no demonic giants, and the actual peopling of the British Isles took place over thousands of years, and involved waves of hunter-gatherers, neolithic farmers, Bronze Age herders from the Eurasian steppe, and Iron Age Celtic-speaking populations whose origins are still being argued about by genetic researchers and linguists with

admirable patients. The Trojan story is, by any modern standard, complete fiction. What is genuinely interesting about it is not whether it is true, which it manifestly is not, but how thoroughly it was believed and how powerfully it shaped the political imagination of medieval Britain for the better part of a thousand years. For the medieval reader, especially the educated one, having Trojan ancestors put Britain on the same prestigious footing as Rome. Rome had Aeneas, the Trojan founder of the line that led to Romulus and Remus. France would later claim its own Trojan founder, a fictional figure named Francis, in a bit of obvious imitation. To be Trojan was to be classical, ancient and respectable, with a pedigree

stretching back to the most famous war in literature. For Welsh princes in particular, the Trojan story was politically valuable because it allowed them to claim direct descent from Brutus through the line of Canber, which made them not just Britons but the original inhabitants of the island, the inheritors of the Trojan blood that the Saxon and later Norman newcomers could never quite match. When you are arguing with an English king about whose claim to the island is older, being able to invoke the great-grandson of Aeneas is a useful rhetorical card to have up your sleeve. The myth also did some quiet ideological work in the way it handled the giants. The Trojans were not depicted as conquerors stealing land from rightful owners.

They were, in the medieval framing, righteous heroes cleansing a tainted island of unnatural monsters and making it fit for proper human habitation. Gogmagog and his kind were not fellow people, but demonic survivors of an earlier corrupted world, and their defeat was a kind of cosmic restoration of order. This was a very convenient narrative for a society that was, in the actual present, dealing with the moral complications of conquest, displacement and shifting political boundaries. If your founding myth involves clearing the land of monsters rather than negotiating with neighbours, your descendants get to inherit a much cleaner sense of legitimacy. Modern readers may notice that this rhetorical pattern has not entirely disappeared from political discourse, but that is a rabbit hole for

another day. The legend lasted in something like its full glory for hundreds of years. Medieval English chroniclers cited Brutus alongside Caesar as a foundational figure of the island's history. Welsh poets called their princes the heirs of Brutus and meant it as a serious genealogical claim rather than a poetic flourish. Cities boasted of Trojan connections, with London happily preserving its identification as New Troy in pageantry and civic ceremony well into the early modern period. It was only with the rise of more rigorous historical methods in the 16th and 17th centuries

that scholars began to question, increasingly publicly, whether Brutus had actually existed at all. The Italian humanist Polydor Virgil, writing a history of England in the early 16th century, raised serious doubts about the entire Trojan origin story, which earned him a great deal of furious abuse from English antiquarians, who quite liked having a Trojan founder and resented being told they could not have one.

one anymore. The legend lingered in popular imagination for another century or two, gradually fading into the realm of poetic reference and decorative pageantry, and by the 19th century Brutus had been thoroughly demoted from history into mythology, which is roughly where he sits today. But there is a final point worth making, which is that even though the Trojan story is not historically true, it is culturally true in a way that matters. The medieval Britons did imagine themselves as the heirs of an ancient migration. They did construct an identity that connected them to the wider classical world through legend, even when the actual genealogical lines were impossible to demonstrate. The story of Brutus and Gogma Gog is not a record of real events,

but it is a record of how a society chose to think about itself, what kinds of pasts it found prestigious, and what kinds of cosmic struggles it imagined had shaped its land. In that sense, the myth is its own kind of evidence. Not for what happened in the second millennium BC, but for what was happening in the medieval mind that needed the myth in the first place. When we look at the petrified stumps of white sands or the tightly packed cells of Gate Home,

we see the physical traces of the post-Roman transformation. When we look at the Trojan story, we see how that same world, a few centuries later, chose to remember and reinvent its own beginnings. Both are in their different ways, part of the same ongoing project of becoming Britain. The Trojan story did not survive on its own. Myths of that scale need a champion, somebody willing to write them down with conviction,

and treat them as serious history, and the man who did more than anyone else to fix the British origin legend in stone, was a 12th century cleric whose name has come down to us as Geoffrey of Monmouth. He was not a particularly important figure in his own day, at least not at the start of his career. He was a minor canon attached to a small religious community in Oxford, of Welsh or possibly Breton background, with a Norman patron, an obvious appetite for storytelling, and a name that suggests he came from the small market town of Monmouth on the Welsh border. We know almost nothing about his daily life. We do not know what he looked like, what his voice sounded like, or whether his colleagues found him tolerable company, although the fact that he produced what he produced

suggests he was, at the very least, an interesting dinner guest. What he gave the world was a book called the Historia Regum Britannia, the history of the kings of Britain, finished probably around 1136. It was in its own modest way a publishing event. Within a generation it was being copied in monasteries from Normandy to Scotland, translated and abridged and turned into French-verse romances, cited as authoritative history by writers who really should have known better, and asked, arguing its way into the bedrock of medieval European imagination. It introduced, or popularised, in a single volume, almost every figure that the modern reader would associate with the legendary deep past of Britain. King Arthur, who up to this point had been a shadowy figure in a few

Welsh poems and a brief mention in earlier chronicles, was promoted by Geoffrey into the central character of the entire British story, complete with a brilliant court at Kaelian, a sword called Calaburnus and a continental campaign that took him as far as Rome. Merlin appeared more or less for the first time as a coherent literary character, a prophet and magician with a startling biography and a habit of saying things that nobody quite wanted to hear. The line of British kings stretching back through dozens of generations to Brutus the Trojan was given names, dates and personalities, even though most of those kings had no historical existence whatsoever. Cities were given founders, lakes were given drowned heroes. The whole misty pre-Roman past of the island was

suddenly populated, in great detail, with a cast of memorable characters whose adventures slotted together into a continuous narrative running for nearly 2,000 years. The genius of the book, and the reason it became a runaway success, was that Geoffrey wrote it not as folklore or poetry, but as serious chronicle history, in the same Latin prose style that respectable medieval scholars used for sober ecclesiastical records. He claimed, in his prologue that he was simply translating a very old book in the British language that had been given to him by Walter, Archdeacon of Oxford. This claim was almost certainly complete invention. No such ancient British book has ever been found. No other writer of the period seems to have known about it, and Geoffrey's narrative draws so heavily on identifiable

Latin sources, particularly the much earlier and much shorter chronicles of Gildas, Bede, and Nenius, that the suggestion of a single Welshmaster text falls apart on inspection. But the trick worked. By framing his book as a translation rather than a composition, Geoffrey gave himself a kind of authorial immunity. If anyone questioned the wilder details of his story, he could shrug and say he was only relaying what the ancient book contained, the way a translator handles a difficult Greek tragedy. modern equivalent might be writing a fantasy novel and insisting throughout the publicity tour that you found the manuscript in your grandmother's attic and are just helping to bring it to a wider readership. The contents themselves were to borrow a polite phrase generously imagined. Jeffrey

opens with the now-familiar voyage of Brutus, his battles with the giants, the founding of New Troy, and the division of the island among his three sons. From there he runs through a long list of legendary kings, almost all of them invented, but some of them in striking detail. There is King Lear, for instance, who appears in Geoffrey's pages roughly four centuries before Shakespeare borrowed him for the stage and turned him into one of the great tragic figures of world literature. There is King Bladud, who is said to have founded the city of Bath, dabbled in necromancy, and finally died in a flying experiment that ended badly when he crashed into a temple, which is the kind of detail you cannot make up unless you are Geoffrey

of Monmouth, in which case you can make up almost anything. There is Casavalonus, and who actually did exist as an Iron Age British war leader recorded by Caesar, but who is in Geoffrey's pages given a backstory and a personality that go well beyond anything, Caesar himself bothered to mention. The pattern throughout is the same. Take any thread of half-remembered tradition, weave in some classical literary motifs, sprinkle generously with invented dialogue and convenient genealogies, and present the whole thing in the matter-of-fact tone of a man dictating dates from a parish register. The two figures who came to dominate the legacy of the book, however, were Merlin and Arthur. Merlin appears first, summoned to the court of

King Vortigun, the ruler who has invited Saxon mercenaries into Britain, and is now regretting it considerably. Vortiguan, trying to build a tower that keeps mysteriously collapsing every night, is told by his druids that the foundations need to be sprinkled with the blood of a fatherless boy. A search is launched, the boy is duly produced, and the boy turns out to be Merlin, who informs the assembled adults that the real reason the tower keeps collapsing is that there are two sleeping dragons, one red and one white, fighting beneath the foundations. The dragons are dug up, they fly off into a violent, symbolic battle, and Merlin proceeds to deliver an enormous prophecy about the future of Britain that runs for several pages of densely cryptic verse. The whole sequence is bizarre,

theatrically vivid, and entirely Geoffrey's invention, although it draws on older Welsh traditions of a wild sear figure called Mirdin. The prophecies were, unsurprisingly, the most quoted and reprinted section of the book for the next 400 years since prophecies have the convenient property of being interpretable to fit almost any current political crisis.

Whenever there was a war or a coronation or a dynastic dispute, somebody would dust off Geoffrey's Merlin and find, with great excitement, that the events of last Tuesday had been precisely foretold in a poem written six centuries earlier. Arthur, when he finally arrives, is built up into a figure of genuine grandeur.

Geoffrey gives him a noble lineage, a sword pulled out of a stone in some versions and forged on the magical island of Avalon in others, a queen named Gwynnevere, a court of brilliant warriors at Kirlian on the Welsh border and a series of conquests that take in Ireland, Iceland, Norway, Gaul, and finally a march on Rome itself. The Arthurian campaign against the Roman Emperor is, by any reasonable historical standard, lunatic, since no British king in the immediate post-Roman period had anything remotely resembling the resources to invade the Mediterranean, but it served Geoffrey's purposes magnificently. It put Arthur on the same playing field as the great classical conquerors,

allowed Britain to be imagined as a power capable of standing toe-to-to-to-with Rome, and gave later writers an inexhaustible supply of episodes to expand. upon. The story ends, of course, with Arthur betrayed by his nephew Modrid, recalled from the continent to fight a final battle in his own homeland, and carried away mortally wounded to the island of Avalon, from which he might yet return when his country needs him most. The future kings, whose right to power, leans on this kind of mythology, are, naturally, not entirely sad about the ambiguity of his fate. What was Geoffrey actually trying to achieve? The conventional view, and probably the correct one, is that the book was a careful piece of political theatre, aimed at the Anglo-Norman

elite of his own time. Geoffrey was writing in a Britain that had been violently remade by the Norman conquest of 1066, a conquest that had displaced the old English aristocracy and installed a new French-speaking ruling class with limited cultural connection to the island they now governed. The Normans needed a usable past, a story about Britain that gave their rule, depth and dignity, and ideally a narrative that bypassed the awkward Anglo-Saxon centuries entirely. Geoffrey provided exactly that. By rooting British history and Trojan refugees, by making the Britons rather than the English the original heirs of the island,

and by giving the whole pre-Saxon past a Norman-friendly grandeur, he offered the Norman aristocracy a heritage they could comfortably plug themselves into. The Saxons, in Geoffrey's pages, are the villains of the middle chapters, treacherous mercenaries who turn on their British hosts and force the rightful inhabitants into the Western Mountains. The Normans, by implication, are not a second wave of invaders, but a kind of restoration, returning the island to its proper international, sophisticated, classically descended status after centuries of Saxon misrule. It was, as political myth-making goes, beautifully constructed.

The book also did some quieter work for the Welsh themselves. Geoffrey was almost so. certainly of partly Welsh descent, and although he wrote in Latin for a Norman audience, his sympathies in the long sweep of the narrative are clearly with the Britons, the people whose descendants in his own day were holding out in the kingdoms of Gwynedd, Powis and a Hubert. The Welsh princes of the 12th century, struggling to maintain their independence against an expansionist Anglo-Norman state, suddenly had a published history that traced their ancestry back to Brutus, gave them a heroic golden age under Arthur, and prophesied in the murder of the murdering

in the Merlin sections that the Britons would one day reclaim the entirety of the island. This was useful material. It allowed Welsh poets and chroniclers to present their political struggle not as a regional quarrel, but as the next chapter in a thousand-year cosmic story. The English crown naturally was less enthusiastic about the prophetic elements, although later English kings would themselves claim Arthurian descent when it suited them, which is the kind of intellectual flexibility that politics as always rewarded. Not everyone fell for it. The book had its skeptics from the moment it appeared, and the sharpest of them was a Yorkshire monk named William of Newburgh, who in the late 12th century wrote what is probably the most splendidly furious takedown in medieval historiography.

William, who fancied himself a serious historian, looked at Jeffrey's claims and lost his temper in fluent Latin. He pointed out correctly that no other Roman or Greek author mentions any of Arthur's supposed continental conquests, which is the kind of thing those authors would have noticed. He noted that the entire story of the Trojan migration is geographically and chronologically impossible, that Geoffrey's lists of British kings cannot be reconciled with any other source, and that the prophecies of Merlin are obvious nonsense dressed up in obscure poetry to disguise their lack of content. His verdict, delivered with the controlled outrage of a scholar who has had a long week, was that Geoffrey had simply made it all up, lying with such impudence and such confidence

that ordinary readers had been hopelessly deceived. He called the book a fabrication, a tissue of inventions and a stain on the dignity of historical writing, and he warned future generations not to be taken in by it. History, for what it is worth, has gradually come around to William's view. Modern scholarship is essentially unanimous that the Historia Regan Britannia is mostly fiction, with isolated passages reflecting older traditions, but the bulk of the narrative invented by Geoffrey himself. The Trojan origin is a literary borrowing from Roman foundation myths. Most of the kings between Brutus and the Roman period are entirely fictional, with names sometimes recycled from Welsh genealogies, but stories supplied by Geoffrey's imagination. Arthur, as Geoffrey presents him, is impossible, although there may have been some real military leader of the post-Roman British,

whose name was attached to a body of legends long before Geoffrey got hold of them. The Merlin prophecies are a literary set piece. The book is, in short, a historical novel pretending to be a chronicle, and judged purely by the standards of factual accuracy, it is a disaster. Judged by any other standard, however, it was one of the most influential books written in medieval Europe. Within a generation it had been adapted into French verse by a Norman poet called Wace, who added the detail of the round table and various courtly flourishes that became

standard. Within another generation it was being absorbed into the great cycle of French Arthurian romances, which would in turn give us the Holy Grail, Lancelot, the quest of the knights, and the entire elaborate world that English readers would much later receive in the form of Mallory's Mort D'Arthur. Without Geoffrey's book, none of that literature would exist in anything like its eventual form. Tennyson, T.H. White, the modern Arthurian films and novels, the entire industry of swords and romance fantasy that owes its tropes to the Arthurian tradition, all traced back to a single Latin manuscript produced by an obscure cleric working in the margins of the Norman world. The book also shaped how the British imagined themselves for centuries afterward.

The English kings of the late medieval period took the Arthurian connection seriously enough that Edward I, on conquering Wales in the late 13th century, organised a round-table tournament at Winchester and exhumed the supposed bones of Arthur and Gwynover from Glastonbury Abbey, as though to demonstrate that he, the new ruler, was the legitimate heir of the legendary king. Henry the 7th, founding the Tudor dynasty after a long civil war, leaned heavily on his Welsh ancestry and his supposed descent from Cadwallader, one of Geoffrey's late British kings. He even named his eldest son Arthur, in a transparent attempt to fulfil the old prophecies, although unfortunately for him and for the boy in question,

the young Prince Arthur died before he could inherit the throne, and the crown passed instead to his younger brother, who would be remembered as Henry VIII, and who had distinctly different priorities. Even into the 17th century, English antiquarians were defending Geoffrey's chronology against humanist critics, and Trojan founder figures continued to appear in civic pageantry and royal entries. It took a long time for the educated mind of Britain to fully give up on Brutus. What is perhaps most remarkable about Geoffrey's achievement is that it produced a national myth essentially by literary fiat. He was not recording an oral tradition that already existed

in finished form. He was assembling, embellishing, and inventing a coherent narrative out of fragments, half-remembered names, classical models, and his own imagination, and the narrative he produced was so satisfying, so well-shaped to its audience's needs, that it overrode the more cautious historical traditions that had preceded it. Earlier, Welsh and English Chronicles had been short, fragmentary, and largely uninterested in the Deep Mythical Past. After Geoffrey, the Deep Mythical Past was a fully populated landscape with a thousand-year story arc, and any chronicler who wanted to write about British origins had to either accept his version or argue with it. There was no

neutral ground. He had effectively redefined what the prehistoric past of the island looked like, and although he himself died in obscurity around eleven fifty five having briefly been made bishop of a small welsh diocese he probably never visited the book outlived him by nearly a millennium and counting as acts of historical mischief go It has few equals. If Geoffrey gave medieval Britain its mythologised history, the storytellers of Wales gave the same period something arguably richer, a body of native prose tales that preserved an older, stranger, more authentically Celtic world. These tales survive today under the collective name of the Mabinojion, although the title is a slightly cheerful 19th century invention,

and the stories themselves are scattered across two great Welsh manuscript collections, the White Book of Riddock from around 1350 and the Red Book of Hergest from around 1400. Both manuscripts are working anthologies of the kind that medieval Welsh literary culture produced in considerable quantity, with poems, prose tales, genealogies, religious texts and historical chronicles all bound together in single volumes for the use of educated patrons. The tales themselves are older than the manuscripts. However, they were probably composed in something like their surviving form in the 11th and 12th centuries, drawing on oral traditions that in some cases reach back into the post-Roman period, or even earlier, into the genuinely pagan Celtic past.

They are, in other words, a window into a body of native British literature that had been developing for centuries before Geoffrey of Monmouth ever picked up a pen, and they look very different from anything in his pages. The collection as we now read it consists of 11 tales, of which the four central pieces are known as the four branches. These four interconnected stories form a loose mythological cycle, full of shape-shifting magicians, talking animals, journeys to the other world, magical cauldrons, severed heads that continue to converse over dinner, and battles fought between rival kingdoms whose geography sometimes corresponds to actual whales

and sometimes drifts off into pure mythological space. The hero of the first branch is Pwill, lord of a southwestern Welsh kingdom called Diphid, who in the opening pages encounters a strange figure on horseback while out hunting, and ends up agreeing, somewhat impulsively, to swap places for a year with Arorn, the king of the other world realm of Andman. The arrangement involves Pwill ruling Andwin in Arorn's shape and form, killing Arorn's enemy at the end of the year in a single combat, and then resuming his own life back in Die-Fed. The whole episode is treated with the casual matter of factness that Welsh tales tend to apply to supernatural events, as though crossing into the other world with a

kind of slightly inconvenient travel arrangement that any country gentleman might find himself making in a busy season. Pwill handles the assignment with reasonable competence, refrains from sleeping with Aron's wife despite sharing her bed for a year out of respect for his host, and returns home with a new alliance and a new friendship that will shape his line for generations to come. The second branch shifts north and west, and tells the story of Branwen, the sister of a king named Bran the Blessed, whose kingdom is more or less co-extensive with the island. of Britain itself. Bran is one of those characters who exists somewhere on the threshold between human and divine.

He's so enormous that no house can contain him. He wades the Irish sea on foot like a man crossing a stream, and his name means raven in Welsh. With all the symbolic associations that implies, when his sister Branwen is married to the King of Ireland and then mistreated at his court, Brann leads a British army across the sea to rescue her, and the resulting war is catastrophic for both sides. The Irish are virtually wiped out and only seven Britain survive, including Bran himself, who is mortally wounded, in what is probably the most striking image in the entire collection. The dying Bran instructs his companions to cut off his head and carry it back to Britain for burial,

promising them that during the long return journey, his head will continue to speak to them and host them as a generous companion. The seven survivors duly remove the head and proceed on a slow, dreamlike journey home, during which Brand's head presides over a series of feasts in mysterious enchanted halls, and the listeners forget the passage of time entirely. When at last they bury the head on a hill in London, facing toward the continent, it serves, in the logic of the tale, as a magical defence against future invasion, until a later king digs it up and removes the protection out of misplaced confidence in his own power. The tale is one of the strangest and most beautiful in medieval European literature,

and it has the unmistakable quality of a story that has been polished by oral retelling over many generations before being written down. The third branch picks up with the few survivors of the Irish War and follows them through a series of further adventures, including a long stretch in which two of the central characters are forced into manual labour in England as saddlers, shoemakers and shield makers, and keep being driven out of every town they enter

because they are too good at their crafts and put the local guildsmen out of business. This is one of the small details that suggest the tales were composed for an audience that found such reversals genuinely amusing. The picture of legendary heroes reduced to humble trades and then resented for their excellence. Eventually the surviving heroes return to Wales, find the land mysteriously emptied of all human and animal life by an enchantment, and have to undertake a series of strange tasks involving a magical mouse, a hidden lord of the other world, and a negotiation that resolves the curse and restores the country.

The narrative moves with a logic that is not always rational, but is always somehow satisfying in the way that dreams sometimes are. The fourth and final branch is the most magical of all, and is set in the northern kingdom of Gwynedd, which corresponds roughly to the modern county of Gwynedd in northwest Wales, with its dramatic mountains and rocky coast. The branch centres on the magician's math, the Lord of Gwynedd, and his nephew Gwydion, who is one of the great trickster figures of medieval literature. capable of conjuring whole armies out of seaweed and creating a beautiful young woman,

named Bloed, out of flowers where no woman of human birth is permitted to marry the hero he has decided to support. Blodoved, who is essentially constructed by magic for the convenience of a man, eventually proves to have her own ideas about her marriage, falls in love with a passing huntsman, and conspires with him to kill her husband through an elaborate ritual that exploits a magical loophole in the curse protecting his life. The husband, Liu, who survives in the form of an eagle, and is gradually nursed back to human

shape by Gwydian, eventually returns to take revenge, and Blodoved is punished by being turned into an owl, condemned to fly only at night and to be hated by all other birds. The whole sequence is one of the great mythic narratives of the Western tradition, with its deep ambiguity about who deserves sympathy, who has been wronged, and what it means to be made for someone else's purposes rather than your own. These four tales, taken together, are full of the unmistakable atmosphere of an older religious world. The gods are there in everything but name, in the magicians who shape reality with words, in the other world kingdoms accessible through mounds and rivers,

in the talking animals and the severed heads that continue to host their friends. Yet the tales as we have them are also clearly the work of Christian authors, who handle the supernatural material with a slightly arm's-length attitude, never quite endorsing the old beliefs, but never quite rejecting them either. There are no overt sermons in the four branches. The Christian framework is implicit, present in the moral expectations of the characters and the occasional reference to churches and abbeys, but the material itself is allowed to retain its strange pre-Christian flavour.

This is one of the things that makes the collection so distinctive. It is not a pagan text, but it is also not a thoroughly Christianised one. It is a literary settlement in which an older mythology has been given respectable prose form and allowed to keep its essential weirdness. The other seven tales in the Mabinojean are more variable. Some, like the story of Colhuch and Olwen, are early Arthurian romances with a deeply Welsh flavour, full of long lists of warriors with extraordinary attributes

and impossible quests involving the hunting of a giant magical boar. Others, the so-called three Welsh romances of a wane, Gerrant, and Peradur, parallel French Arthurian stories by Trotienne de Trois, and may represent either Welsh adaptations of his work, or independent Welsh versions of the same underlying material, depending on which scholar you ask. The dream of Maxen Veledig, which is one of the more politically interesting tales, weaves the historical Roman Emperor Magnus Maximus into a Welsh legend

that explains the origins of the San Helen Roman Rhodes, and connects the Imperial Roman past directly to the mountains of St. Snowdonia. Throughout the collection, the prose has a distinctive rhythm, with short declarative sentences, repeated formulas, and occasional eruptions of poetic intensity that reflects centuries of refinement in oral performance. The collection survived through the medieval period in a small number of manuscript copies, read by Welsh aristocrats and scholars, but it was largely unknown to the wider European literary world for centuries. That changed in the 19th century, thanks to one of the more remarkable figures in Welsh literary history,

an English woman named Lady Charlotte Guest, who learned Welsh as an adult, married into the Welsh industrial aristocracy, raised ten children, ran the family ironworks for a period, and somehow found the time and the talent to produce the first complete English translation of the Welsh tales. Her edition appeared between 1838 and 1849, in a series of beautifully produced volumes with extensive notes and decorative illustrations, and it was an immediate sensation. It introduced English-speaking readers to a body of literature they had not realized existed, gave the tales their now familiar collective name of the Mabinogion, through a slightly creative interpretation of a single Welsh word, and made the stories available to a generation

of writers, artists and folklorists who would carry their influence into modern literature. One of the most consequential of those readers, although he came along nearly a century after her translation first appeared was J. R.R. Tolkien was a professional philologist at Oxford, with a deep and lifelong fascination for the languages of medieval Europe, and Welsh was, by his own repeated admission, one of his great loves. He described the sound of Welsh as having a peculiar beauty that the rest of British language somehow lacked, and he spent decades absorbing not just the language itself, but the literary traditions that came with it. When Tolkien set about constructing the elven languages of Middle Earth, he based the principal

one, cindarin, very directly on the sound and structure of Welsh. Anyone who has read the names and phrases in his elvish, and felt that they sound vaguely familiar but cannot place why, this is the answer. They are echoing the cadences of a real language spoken in the mountains of North Wales, and it is one of the more pleasing ironies of literary history that a body of medieval Welsh tales, preserved by a handful of scribes for a small audience of native speakers ends up shaping the most influential fantasy literature of the modern world. The streaming series, the best-selling novels, the role-playing games, all of them, at several removes,

are in conversation with stories first told around firesides in the kingdoms of post-Roman Wales. The Mabinogion also did something subtler for Welsh culture itself. By being recovered, translated and presented to the world in the 19th century, it gave Welsh writers and intellectuals a body of native literature that could stand alongside the great medieval traditions of Ireland, Iceland and France. Welsh poetry had its own long history, but for centuries the tradition had been oriented inward,

addressed to a small audience of bards and patrons. The publication of the prose tales, in both Welsh and English, allowed Welsh literary identity to project itself outward, to claim a place in the wider European story of medieval literature. and to provide modern Welsh writers with a foundation of native material to build upon. Modern Welsh poets, novelists and dramatists have drawn on the four branches in the other tales repeatedly, sometimes retelling the stories directly, sometimes using their characters and settings as a kind of cultural shorthand, sometimes simply absorbing their distinctive rhythm and imagery into their own work.

What runs through both Geoffrey of Monmouth's Latin Chronicle and the Welsh prose tales, despite their very different intentions and atmospheres is a willingness to take the deep path seriously as a place where meaning is generated. Geoffrey, with his political agenda and his inventive genealogies, was building a usable history for a Norman elite. The compilers of the Mabinogian, whoever they were, were preserving a native mythological inheritance that had no obvious utility beyond its own intrinsic value.

But both projects shared the conviction that stories about the long-ago shape who you are now, that the figures of legend are not idle decoration, but active participants in the present, and that a culture without its myths is a culture missing something essential about itself. That conviction was very strong in medieval Britain, on both sides of the language divide, and the body of literature at produced still has the power to fascinate readers who picked up these stories long after the world that produced them had been swept away by other forces. The medieval Welsh and the medieval Anglo-Normans would not have agreed about much. They did agree in their different ways that the past was worth telling stories about,

and that some stories deserve to be told well. Among the eleven tales preserved in the Great Welsh Manuscript collections, there is one that does something none of the others quite manage. It takes a real Roman emperor, a real military campaign, and a real network of stone-paved roads still visible across the Welsh landscape today, and it weaves them all into a single shimmering love story that ought, by any reasonable standard to be impossible to take seriously, and yet has somehow remained one of the most beloved short narratives of the entire medieval Welsh tradition. The tale is called Brudwig Maxen Vleadig, the dream of Maxen Vleadig,

and it is the kind of piece that sits on the boundary between history and legend, with such cheerful disregard for the difference that you almost forgive it for blurring the categories. The Maxon of the title is, in the actual historical record, the Roman general Magnus Maximus, a Spanish-born commander who served in Britain in the late 4th century, and who would eventually be proclaimed emperor by his own legions. But the tale treats him as something much closer to a romantic hero out of an eastern fairy tale, a man whose entire imperial reign is set in motion by a single vision he experiences during an afternoon nap. It is, on its surface, a love story. Underneath the surface, it is one of the more elegant pieces of cultural reclamation produced

in medieval Wales, a story that takes the Roman past and quietly walks off with it. The opening of the tale finds Maxin at his court in Rome, where, the narrator informs us. He is the most handsome and the most powerful man on earth, ruling over 32 crowned kings as their universal sovereign. This is, you may have noticed, not historically accurate, since the actual Roman world of the 4th century was not in fact organised around a single emperor commanding 32 vassal kings, and the Magnus Maximus of the historical record was at best one of several rival emperors who happened to control the western portion of the empire for about five years. The Welsh narrator is not interested in these details. He's interested in establishing Maxen as a

figure of universal authority. The sort of ruler whose attention, when it falls on a particular corner of the world lends that corner its own borrowed glory. So Maxen, in the tale, sits at the very top of the world, and the world arranges itself politely beneath him, and the actual mess of late Roman politics is filed away under the heading of inconvenient detail. One day, says the tale, Maxon rides out from Rome with a vast hunting party. They cross a river, climb a hill, and arrive at a beautiful valley where the emperor decides to rest at midday. His attendants build a shelter for him out of their gilded shields, propped together to keep the sun off, and Maxen lies down on the grass and falls asleep. What follows is a dream of such unusual clarity and detail that the listener

immediately understands it is going to be the engine of the entire story. Maxon sees himself walking up the same valley he has just climbed, but the valley is now twice as steep, and at the top of it he passes through extraordinary forests until he comes to a great river that flows down to the sea. beside the river stands a city and within the city is a fortress with many towers of bright colours and at the harbour beneath the fortress lies a single ship with a plank of walrus ivory leading from the shore to the deck maxon walks across the plank and finds the ship empty except for two young men playing a board game on a board of pure silver with pieces of pure gold while seated on the deck nearby as an old man dressed in expensive robes carving a fresh set of game pieces from a block of bone.

The old man, the young men, and the silver board are described in such precise and slightly hypnotic detail that the narrator clearly wants the listener to feel the strangeness of dream logic, in which everything is simultaneously specific and unaccountable. Maxin does not speak to any of them. He simply walks past, gets onto the ship, and the ship sails on its own, carrying him out across an ocean and to the other side.

When he steps ashore on this far side, he finds himself in a country of the most beautiful land he has ever seen. There are mountains higher than any he has known, valleys greener than any he has imagined, and great rivers running down to the sea on every side. He follows the rivers across the country until he comes to another fortress at the mouth of one of them, a place far larger and more impressive than the one he saw at the beginning of the dream. He enters the gate, he climbs up through the rooms and corridors of the building, and in the topmost chamber where the windows face out across mountains and water, he sees two young men at another board game, this time on a board of solid gold watched by another older nobleman,

this one of regal bearing sitting in a carved ivory chair. Beside the older man, on a chair of red gold, sits a young woman. The description of this woman is one of the most famous passages in medieval Welsh literature, and it is worth pausing on it because it sets the entire emotional centre of the tale. The narrator tells us that her gown is so white that it is harder to look at than the sun rising on a clear morning. Her ornaments are of gold and rubies and pearls gathered in such quantity that no eye can quite take them in.

Her face shines, her hair is black, her brow is white, her cheeks have the colour of fresh apples, and her bearing is so dignified that the dreaming Maxon, who is supposed to be the sovereign of the entire world, immediately drops to his knees in her presence. She lifts him up, takes both of his hands in hers, and seats him in the chair beside her. The two of them sit together in silence, and then, just as Maxen is about to speak to her, he is awakened from the dream by his attendants, who have heard the morning hunting horns

and are unsure why their master is still asleep on the grass. Returning to consciousness, Maxon does not bound up refreshed, the way a normal man would after a midday rest. He lies with his eyes open and his face troubled, and over the following days and weeks his courtiers begin to notice that something has gone seriously wrong with the emperor. He stops eating. He stops attending to the affairs of the empire. He stops speaking to anyone except in monosyllables, and even his closest advisors cannot get him to explain what is the matter. The narrator deploys here a psychological observation that is genuinely clever given the period. Maxon is suffering not from any visible illness, but from a kind of consuming, single-minded, long-endouser.

for a woman who does not, as far as anyone knows, exist outside his own dream. The condition would be diagnosed today as something between obsessive love and depressive withdrawal, and the medieval narrator describes it with the same level of detail he applies to feasts and battles, which is to say with a fair amount of patience and a willingness to let the listener feel the slow weight of the thing. Eventually his advisors take Maxen aside and persuade him, with some difficulty to tell them what is wrong, He recounts the dream, they consider the matter. Their solution is practical to the point of comedy. If the emperor has fallen in love with a woman in a dream, they reason.

The only thing for it is to find her in the waking world, and if necessary, search the entire earth for her until she is found. They organise three groups of messengers, dressing them in the finest clothes and sending them out in three directions, instructing them to find a country that matches the description of the country in the dream and a woman who matches the description of the woman. The messengers ride for a year. They cross many countries.

Two of the three groups return without success, with weary horses and apologetic excuses. But one of the three messengers, After travelling further than any of the others, eventually arrives in a place that fits the description so closely that he is sure he has found it. He is in the mountainous country at the western edge of the world, in what would now be called North Wales,

and the great river he has reached is the Seon, which empties in the sea-en. the Meni Strait at a place where there stands an old Roman fortress called Sigontium, just outside the modern town of Canafon. The messenger climbs the river, reaches the fortress, enters its halls, and finds, in the highest chamber, exactly the scene the emperor described. Two young men playing a board game on a golden board, an older nobleman watching them from an ivory chair, and beside him a young woman of striking beauty seated on a chair of red gold. She's the daughter of the local king, a man named Uda.

and her own name is Ellen, sometimes given in the tales as Ellen Lloyd-Dog, which translates roughly as Ellen of the hosts or Ellen of the many. In reference to her later association with armies and military roads, the messenger, recognising her instantly from the emperor's account, throws himself at her feet and announces that the Emperor of the Romans wishes to make her his wife. Ellen, with admirable composure for a young woman, who has just been informed by an exhausted Roman messenger that the most powerful man in the world has been wasting away for love of her,

declines to leave her home. If the emperor wants her, she says, he can come and find her. So the messenger rides back across the long miles to Rome, breathless with the news, and Maxen, finally hearing that the woman of his dream is real and located in a precise place, gathers an army and sets out on the journey himself. The expedition that follows is,

In the tale, partly a romance and partly a campaign of conquest. Maxon leads his legions across the continent, crosses the sea to Britain, fights his way through unspecified opposition, until he arrives at the western mountains, climbs the same valley he saw in his dream, follows the same river, enters the same fortress, finds the same woman in the same chamber,

and falls again to his knees before her. This time she lifts him up and consents to marry him. The wedding takes place at Sagantium, with feasting that lasts a generous number of days, and Maxon presents his new wife with three rich gifts. The first is the island of Britain, given to her father, Udath, to rule on her behalf. The second is the fortress of Sorgantium itself, granted to her as her personal household. The third, and this is the detail that the tale uses to anchor itself to the actual landscape of North Wales,

is a network of strong roads to be built across the country at her direction, linking her chief fortresses together and allowing her armies to move quickly across the difficult terrain of the mountains. These roads, the tale informs us, were built at her command and are still in use today, and they are called the Sarn Helen, the causeways of Helen in her honour. This is the moment in the story where legend and physical landscape meet, and it is one of the more pleasing meetings of its kind in medieval European literature. The Sarn Helen is a real network of Roman roads. It runs across mid and north Wales for a total length of around 260 kilometres, with sections still clearly visible today as raised earthworks, paved fragments, and fords across rivers that anyone

with a decent pair of walking boots can trace through the upland country between Carmarthen in the south and Kernan in the north. The roads were built by the Romans during their occupation of Britain, primarily for military purposes, allowing legions to move between forts, pacify rebellious tribes, and maintain the imperial supply lines that kept the western frontier provisioned. By the medieval period, when the tale of the dream was being told, the roads were already centuries old, falling into disrepair in places but still very much in use as the principal long-distance routes through the country. They were cut into the mountains, raised on causeways across boggy ground, and built in a kind of obvious foreign engineering style that the medieval Welsh,

looking at them, immediately recognised as not being the work of their own ancestors. They needed an explanation, and the tale provided one. the roads were Helens, she had ordered them built. The local memory wove the Roman past into a Welsh romance and produced a place name that has stuck for well over a thousand years. What is genuinely interesting about the San Helen is that the name itself, in its Welsh form, almost certainly does come from a real association

with the historical Helen of the legend, although the etymological details have been debated by scholars for generations. The personal name Elen, sometimes Latinized as Helena, is reasonably well attested in late Roman and early medieval Welsh sources. There was a real woman, the wife of Magnus Maximus, who was probably of British origin and whose name in some sources is given as Ellen. There was also a famous Helen of Imperial Fame, the mother of the Emperor Constantine, who, according to some traditions, was herself British, and whose cult was widely spread

across Western Christendom. The two figures, the wife of Maximus and the mother of Constantine, may have been confused or merged in Welsh memory, with the result that a single legendary Ellen was credited with the construction of the roads, the founding of churches, and a list of other improvements that no individual woman could have accomplished in a single human lifetime. Whatever the precise process of conflation, by the time the tale of the dream was being committed to writing, the connection between the name of the network and the figure of Ellen Lloyd Dogg was firmly established. Welsh travellers walking across the high passes of Snowdonia, looking down at the long

straight line of the road snaking through the bracken, knew exactly whose causeway they were standing on. The tale has more story to tell after the wedding, and the second half of the narrative concerns Maxon's eventual return to Rome to deal with a usurper who has taken advantage of his long absence in Britain. This part of the tale is shorter, more brisk, and more clearly tied to actual historical events. While Maxon has been busy in the western provinces, somebody back in Rome has noticed that the Emperor's seat is empty for an unreasonably extended period and has decided to install themselves in his place. Maxon gathers his British forces,

sails for the continent and marches on Rome to reclaim his throne. The campaign is supported by his wife's brothers, Sinan and Gadiun, who lead a contingent of Britons across the Channel to fight alongside the Emperor. After a series of battles and a long siege, Maxon retakes Rome and re-establishes himself as the legitimate sovereign of the empire.

He rewards his British. in-laws with grants of land in the area of what is now Brittany in northwestern France, and the tale uses this episode to explain the existence of a Britonic-speaking population in that part of the continent. The Britons of Brittany, in the logic of the tale, are descendants of the troops who followed Maxon to Rome and were rewarded with land for their service. This explanation has a certain charm, and it is not entirely wrong, since there really was a substantial migration of Britons to Brittany, during the late Roman and post-Roman period, although the actual mechanisms involved were considerably less heroic than the tale suggests.

Sinen and his men, the tale continues, settled in their new lands and prospered, but they did so under one rather peculiar condition. According to the narrator, Sinen ordered that the tongues of the local women in the conquered region be cut out, so that the Britain soldiers and their descendants would never have to learn the local language and could maintain their own Britonic speech in purity for generations to come. This is the kind of detail that makes a modern reader sit up and reread the sentence

to make sure they have understood it correctly, and yes, they have. The tale presents this as a sensible, practical measure, rather than as the appalling act of mutilation it actually describes, and the narrator moves on to other matters with no particular comment. It is a useful reminder that medieval narrative sometimes contain casual brutality of a sort that does not survive translation comfortably into modern sensibility, and that the cheerful tone of folk tales can paper over assumptions that would not pass

any decent ethical review today. The detail is also mercifully not historical. The Britonic speech of Brittany was preserved by ordinary processes of migration and settlement, with no need for any of the violence the tale invents. But the story tells us something about how the medieval Welsh imagined the relationship between language, identity and congress. which was a relationship they were thinking about very actively in their own present. Maxon in any case dies somewhere in the course of his European campaigns, although the tale handles his death briefly and with little drama. The actual Magnus Maximus, the historical figure behind the legend,

was executed in 388 AD after losing a battle against the Eastern Emperor Theodosus, who had no patience for upstart Western rivals, and dispatched Maximus with the brisk efficiency that successful late Roman emperors tended to apply to such problems. The tale does not dwell on this. It is more interested in the legacy Maxon leaves behind in the British landscape, the network of fortresses and roads, the Roman heritage that becomes, through this slightly fantastical retelling, a Welsh inheritance. By marrying Ellen, Maxon has effectively married into the British line, and the Great Roman past becomes retroactively a Welsh past as well. The civilising achievements of the empire are reinterpreted as the gifts of a Roman emperor to his

Welsh wife, and the Welsh princes who claim dissent from this marriage, including the kings of Gwynedd, Powis and several other early medieval Welsh kingdoms, were claiming not just ancient prestige, but a direct genealogical link to Imperial Rome itself. This is the political work that the tale is doing underneath its romantic surface. Welsh dynasties of the early medieval period genuinely did claim descent from Maxon, and we will see in a moment how systematically that claim was constructed. Now it is enough that to be a respectable Welsh king meant being able to point to a Roman ancestor, and Maxen was the most useful Roman ancestor available. The fortress at Sagontium, which serves as the setting for so much of the tale, is itself

a real archaeological site that has been excavated in considerable detail. It sits on a low ridge overlooking the Manai Strait just outside the medieval walled town of Kernophon, and its remains include foundations of barracks, a headquarters building, granaries, a temple complex and a defensive wall that was rebuilt and reinforced several times during the centuries of Roman occupation. The fort was established in the late 1st century during the conquest of North Wales and it served as a key garrison post controlling the strategically important many region, the gateway to the island of Anglesey, which was a major source of grain and a centre of native British religious life. The Romans took Sagontium seriously enough to maintain a substantial garrison

there for nearly 300 years, and the archaeology shows continuous occupation and reinvestment well into the late 4th century, which is exactly the period of Magnus Maximus himself. There is, in other words, a real historical association between Sagontium and the period of late Roman activity in Britain, and the medieval Welsh tale is not entirely inventing the connection when it places Ellen and her father at this fortress. They were drawing on a memory of Sorgantium as a centre of late Roman power, even if the romantic details about board games and ivory chairs were their own contribution. When Edward I of England conquered Wales in the late 13th century and built his own enormous castle at Kernephon, just below the ridge where Sorgantium stood,

he was very deliberately drawing on the same set of associations. The towers of his castle were built with deliberately polychrome stone in a striped pattern of light and dark that is sometimes interpreted as a conscious echo of the towers of Constantinople, the great imperial city of the eastern Mediterranean. There is a strong tradition that Edward, who knew his Welsh legends better than most of his English contemporaries, was deliberately positioning his castle as the new imperial capital of the conquered Welsh, casting himself as a kind of maxim figure for the new political order, with Carnaphone as his Sigontium and his castle as the fortress of legend. The visual rhetoric of the building with its bands of striped masonry and its polygonal towers supports this reading,

and it remains one of the more suggestive examples of how a medieval king could deploy local mythology against the local population, taking the Welsh story of imperial connection and using it to reinforce his own claim to the country. The Welsh, of course, had their own opinion of this manoeuvre, and the relationship between Edward and Welsh memory remains one of the more fraught topics in the cultural history of the country. But the fact that Edward thought it worth doing, that he saw the Maxen connection as politically useful, even six or seven centuries after the historical events behind the legend, give some sense of how durable the mythological framework had become.

The tale itself, as a piece of medieval Welsh prose, has a number of features that mark it as a distinctive literary achievement. It is quite short, perhaps 15 to 20 pages in modern editions, and it moves with an unusual economy from scene to scene without the long lists of warriors and kings that clutter, some of the other tales in the collection. Its tone is poetic and slightly ceremonial, full of formal descriptions of clothing, architecture and natural settings,

and it has a distinctly cinematic quality, with successive scenes that read almost like sequential frames in a beautifully illustrated film. The author, whoever he was, had a clear interest in the visual texture of the world he was describing, and he uses repetition of key phrases to give the prose a rhythm that feels close to oral performance. Read aloud in modern wealth,

The tale has the feel of a story that has been polished by generations of telling, with its internal pauses and its careful build-ups, all timed for a listening audience rather than a silent reader. It is one of the more elegant short pieces in the entire medieval European tradition, and it is regrettable that it remains relatively unknown outside Wales itself, where it is taught in schools and remembered with the same affection that English children give to King Arthur or that Irish children give to Cahoulin. What the tale demonstrates more than almost any other piece of medieval Welsh writing is the way in which a culture deals with the fragments of an imperial past it has inherited, but does not in any straightforward sense own. The medieval Welsh of the 11th and 12th centuries lived in a country dotted with the visible remains of Roman power. The roads were still there, used daily. The forts were still standing in places, sometimes reused as the cause of medieval towns

or the foundations of castles. The bathhouses, the temples, the inscribed altar stones turned up in farmers' fields with regularity, and the stories about who had built these things and why had to come from somewhere. The Welsh response was not to simply ignore the Roman past, nor to disown it as the work of foreign occupiers, nor to claim it directly as their own work. The response was to absorb it into a story in which the Roman past becomes, through marriage and legend, a Welsh inheritance.

Maxon comes to Wales, marries a Welsh woman, father's Welsh princes, and leaves behind Welsh roads. The empire in the logic of the tale is not something that happened to Britain. It is something that became Britain, fused through romantic affiliation into the ancestral landscape. This is the kind of cultural work that legends do quietly, and that historians sometimes underestimate. The tale is not history. It is, on the level of factual accuracy, almost completely wrong. Magnus Maximus did not have a vision in his sleep that led him to a fortress on the seant.

The San Helen was not built at the order of a Welsh princess. The settlements in Brittany were not founded by Sinan and his brothers as a thank you gift after the recapture of Rome. None of these things happened in the way the tale describes them, but the tale captures, in poetic form, the way an entire culture chose to remember and reinterpret a period of its past that it did not have direct documentary access to, and the way it chose to make that past meaningful in the present. The Welsh princes who ruled in the centuries after the Roman withdrawal genuinely did inherit a landscape full of imperial infrastructure

and they genuinely did need a narrative that explained where it had come from and how it related to them. The tale of Maxen and Ellen provided that narrative. It has not been historically reliable for very long, but it has been culturally true for over a thousand years, which in some ways is the more impressive achievement. There is also a curious afterlife to the tale that is worth noting, in the way it has been used by later Welsh writers and patriots, to think about national identity. During the long centuries when Wales was politically absorbed into the English state, from the conquests of Edward I in the 13th century onward,

Welsh cultural figures repeatedly return to the story of Maxen and Ellen as a kind of touchstone for the idea that the country had once been the seat of an emperor's love and an empire's roads, and that this dignity, was not something the conquest could erase. The 18th and 19th century revival of Welsh literary culture, which produced a flowering of national poetry, music and historical scholarship, after a long quiet period, leaned heavily on stories like this one. Welsh schoolchildren memorized passages from the tale. Welsh poets wrote new verses inspired by the dream. Local antiquarians went out into the hills to identify the precise roots of the San Helen,

mapping its surviving stretches with the same enthusiasm that English antiquarians of the same period brought to Hadrian's wall. The Roman past of Wales, as filtered through this tale and others like it, became part of the cultural toolkit that allowed modern Welsh identity to define itself in distinction from English identity, even when the political circumstances offered relatively little space for that distinction to be expressed openly. The figure of Ellen herself has had a particularly interesting afterlife. In addition to her association with the network of Roman roads, she became attached to a number of churches and holy wells across Wales, sometimes in the form of St. Helen rather than the Ellen-Louis Dog of the Dreamtale. Whether the medieval Welsh actually

distinguished between Ellen, the wife of Maxon and Helena, the mother of Constantine is unclear, and probably the distinction was not always sharp in practice. Both women were imperial Roman figures with British connections both were associated with church founding and pious works, and both could be invoked as ancestors or patrons by Welsh communities looking for an early Christian pedigree. Some of the wells dedicated to St Helen in North Wales lie close to surviving stretches of the Sarnhellen, which makes a certain practical sense, since wayfaring travellers along an old Roman road would have wanted clean drinking water and would naturally have associated the local saint with the network they were using.

The blending of Roman, Christian and folkloric elements in these dedications is exactly the kind of cultural layering that medieval Welsh religion specialised in and Ellen sits at one of the more visible junctions of all three traditions. There is also lurking just behind the tale an interesting question about what kinds of women the medieval Welsh wanted at the centre of their foundation stories. Ellen is, in some ways, an unusual heroine. She's not passive. She does not simply wait to be rescued.

or to be carried off by the dreaming emperor. When the messenger arrives and tells her that Maxon wants her, she refuses to go to him and insists that he come to her instead. When he arrives, she negotiates her marriage on her own terms, with her father present but clearly not making the decisions for her. She receives the Kingdom of Britain as a wedding gift and uses it to direct the building of roads. Her brothers fight under her direction in continental campaigns.

The tale gives her a degree of agency that is striking for a medieval narrative, and it suggests that whoever composed it had in mind a model of female royal authority that was perhaps inspired by historical figures the author had encountered, whether in chronicles or in the genealogies of his own day. Welsh literature of the medieval period has a number of strong female characters, from the fierce Aronahod of the fourth branch to the mysterious Olwen of the early Arthurian tale, and Ellen belongs in their company. She is not as bizarre as some of them, and not as tragic as others. but she has her own kind of quiet authority, and the tale takes her presence at the centre of the

political action seriously, rather than treating her as decoration. Reading the tale today, one of the things that lingers is the dream itself. Of all the elements of the story, the most memorable is the long detailed description of the journey Maxon takes in his sleep, the way the landscape unfolds before him with the strange logic of dream geography, the silence of the figures he encounters, the empty ship that sails by itself, and the woman who waits at the end of it all in the high chamber of a fortress that he has never seen but that exists in the real world all the same. Whoever wrote this scene understood something about the way that imagination and place can intertwine, the way a vision

can sometimes proceed and experience and structure it in advance. The medieval Welsh listener, hearing the tale and looking out across the same mountains and rivers that the dream describes, have felt the slight prickle of recognition that comes when a story names a place you yourself have walked. The fortress at the river's mouth, the green valleys, the high mountains, all of these were real, all of these could be visited, and the love that Maxon felt for the woman in his dream became, by extension, a love that could be transferred onto the landscape itself. The tale is, in the end, not really about Maxon or Ellen at all. It is about the country and about the way a country can be remembered and longed for are named, and about the way a place can become beautiful through the stories that have been told

about it. In that respect, the tale belongs to the same long tradition that produced everything from Homer's Ithaca to the modern fantasies of imagined geography. The tradition in which writing about a place becomes inseparable from making the place worth caring about. The Welsh, looking at their own mountains and their own coast and their own old Roman roads, decided that this place deserved to have its love story properly told, and they told it well enough that we are still listening to it more than a thousand years later. Behind the dreaming emperor of the Welsh tale stood a perfectly real and rather more pragmatic man,

whose actual life was less concerned with visions of beautiful women in faraway fortresses and rather more concerned with the harsh business of late Roman military politics. The historical Magnus Maximus, who appears in Roman sources sometimes simply as Maximus, was a Spanish-born officer whose career trajectory, while certainly impressive, followed a depressingly familiar pattern of late imperial advancement. A capable soldier proved his loyalty in the provinces, gathered the affection of his troops, was raised on a shield and proclaimed emperor by the men who served under him,

marched on his rivals with a determined army at his back, succeeded for a few years in establishing himself as a legitimate ruler over a portion of the empire, and then lost everything in a single disastrous battle against another general, who happened to be slightly luckier on the day. This was not an unusual life story for a 4th century Roman commander. It was, if anything, the standard career arc, repeated with grim regularity across the western and eastern halves of the empire throughout the troubled decades that historians eventually gathered together

under the polite heading of late antiquity. What makes Maximus distinctive, and what eventually transformed him from one more failed usurper into the central ancestor figure of Welsh royal genealogies, was the particular geographical accident of his rise to power. He was proclaimed emperor in Britain, not in Italy, not in Gaul, not in any of the wealthy old Mediterranean provinces

where most of these little political coups happened, but on the cold, wet, distinctly underappreciated island at the far western edge of the Roman world. His subsequent campaigns took him out of Britain and across the channel and eventually deep into the continent, but the fact that his career began in the British provinces meant that the British took an unusual proprietary interest in him for the rest of recorded history. He was, for a brief moment, our emperor, the one who came from here,

and in the long memory of the Western British, that fact mattered considerably more than the eventual outcome of his Italian campaigns. The man himself was probably born in the early 3 30s in the Roman province of Galicia, in the northwestern corner of what is now Spain, in a region of green hills and rocky coastlines that bore some passing resemblance to Britain itself. his family appears to have been provincial military aristocracy, the kind of moderately well-connected household that produce steady supplies of competent officers for the Imperial Army

without ever quite producing senators or governors of the first rank. He was a relative, possibly a nephew or a younger cousin, of Theodosius the Elder, a senior general who had distinguished himself in a number of campaigns in North Africa and Britain, and the family connection was probably what got Maximus his first significant assignment. Sometime in the later 360s, the Elder Theodosius was sent to Britain to deal with a crisis that had escalated considerably beyond the capacity of the local garrison, and Maximus accompanied him as a junior officer. The crisis in question is known to historians with the dry economy that historians prefer for moments of catastrophic violence as the great conspiracy. In 367, the various tribes who menaced Roman Britain from beyond its borders, the picks in the north, the Scotty from Ireland, the Saxons from the Contextes,

continental coast, and the obscure but apparently effective people called the Atacotti, all decided to attack at the same time, in what may have been a coordinated effort or may have simply been a coincidence of opportunity. The Roman defences cracked in multiple places at once. The wall in the north was overrun. Several of the senior officers in Britain were killed. Bands of barbarian raiders penetrated deep into the southern provinces, looting villas and burning crops. The province descended into a year of chaos. that the Roman administration in Trea and Rome found extremely difficult to address from a distance. The Elder Theodosius was eventually dispatched to put things right,

and by all accounts he did so with the methodical brutality that successful late Roman commanders specialised in. He drove the raiders back, restored the frontier, executed deserters and traitors, re-fortified the towns, and reorganised the provincial administration in ways that would shape British defences for the next half century. Maximus, watching all this unfold at his uncle's side, presumably absorbed a great deal about how to manage a difficult province, how to motivate troops who had been shaken by recent failures, and how to project authority across long distances in a landscape full of enemies. By the late three-70s, Maximus was back in Britain in his own right, no longer as a junior officer attending a senior relative, but as a senior commander in his own command, probably with the title of Comes Britannia Rum, the Count of the Britain, the senior military officer responsible for the field army of the diocese.

The position was important. Britain was a chronically troubled province, with raids continuing along the eastern coast from Saxon shipping, pressure from the unconquered north, and a population that had developed the habit of greeting any new development from the imperial centre with a certain resigned skepticism. To hold the position required a man who could maintain discipline, project authority,

and adapt to local conditions, Maximus appears to have done all three competently. He fought campaigns against the Picts in the north. He probably oversaw the maintenance of the coastal defence system that historians later called the Saxon Shore, a chain of heavy stone fortresses around the eastern and southern coasts. He did all the boring administrative work of a serious military command, the paperwork and the supply requisitions and the disciplinary hearings

and the inspections of equipment, and he apparently did it well enough that the men under him developed a serious affectionate. for him. That affection became politically consequential in 383. In that year, the legitimate emperor of the western half of the empire was a young man named Gratian, who had inherited the throne at 15 after his father Valentinian I first had died of a stroke during a particularly furious shouting match with some barbarian envoys. Gratian was not, by the accounts of his contemporaries, a bad emperor in the sense of being cruel or stupid. He was perfectly intelligent and reasonably well-meaning. But he had the unfortunate habit of preferring the company of his Alan bodyguard,

an elite cavalry unit recruited from the Iranian-speaking step, peoples who had recently arrived in the Roman world over the company of his own Roman officers and senators, and he had taken to wearing the distinctive feathered headdress and ornate clothing of the Allens in public ceremonial, which is Roman subjects found, to put it gently, embarrassing. The traditional Roman officer a corps felt insulted, the senatorial class felt ignored. The pagan aristocracy was furious that Gratian had removed the altar of victory from the Senate House in Rome, and the army in Britain, which had its own complaints about pay and supplies, decided that the situation had gone on long enough. In the spring or early summer of 383, the British legions raised Maximus on a shield

and proclaimed him emperor. The exact location of the proclamation is not recorded in the surviving sources, but it probably happened at York, the senior military headquarters of the diocese, with detachments from the various legions in Britain assembled for the ceremony. Maximus, who was certainly aware that this proclamation was coming and had probably done a fair amount of careful work to ensure it actually happened, accepted the elevation with the kind of orchestrated reluctance that fourth-century usurpers generally affected. He immediately began preparations for a campaign on the continent, drawing troops from across the British provinces and gathering ships at the Channel ports. By the late summer of

383 he had crossed into Gaul with a substantial army, and Gratian, whose own forces in Gaul had been undermined by the same complaints that produced the British proclamation, found himself outflanked, abandoned, and eventually betrayed by his own commanders. Gratian fled south with a small bodyguard, attempting to reach the relative safety of Italy, or the friendly forces of his eastern colleague Theodosius, the son of the elder Theodosius, and therefore Maximus's own cousin. He almost made it. He reached the city of Lyon, where he was received with apparent friendliness by the local commander, a man named Androgathius, who was actually working for Maximus. Androgathius held Gratian at a banquet, distracted him with hospitality, and then had

him killed, possibly on Maximus's direct orders, or possibly on his own initiative as a piece of helpful improvisation. Either way, the legitimate Western Emperor was dead, the throne was effectively vacant, and Maximus was the only candidate within reach of it. He moved his court to the city of Trier on the Moselle, the traditional administrative centre of the Western frontier, and began the work of consolidating his position. Trier in the 380s was, by the standards of the late Roman West, a remarkable place. It had been an imperial capital, on and off, since the late 3rd century, when the The emperors realised that the Italian peninsula was too far from the troubled Rhine and Danube frontiers to be a practical centre of government. Successive emperors had built monumental

palaces, baths, basilicas and city walls in the Moselle Valley, and by the time Maximus established himself there. Troyer had a permanent population of perhaps 80,000 people, with a circumference of city walls nearly six and a half kilometres long, an imperial palace large enough to host substantial state ceremonies, a basilica that still stands today as one of the largest surviving Roman buildings in Europe, and a system of public baths that put most modern leisure centres to shame. The city was sophisticated, prosperous, and bilingual, with a Latin-speaking administrative class and a Gallo-Roman civic culture that had absorbed several centuries of imperial investment. Maximus settled into the imperial palace, surrounded himself with a court that

included a number of distinguished intellectuals and ecclesiastics, and began the standard business of being a Roman emperor. That business included, among other things, minting coins, and the coins of Maximus survive in some quantity in modern collections. They are typical late Roman issues, with his bust on one side, depicted in the standard imperial profile with diadem and military cloak, and various symbolic figures on the reverses, ranging from personifications of Rome and Constantinople to scenes of imperial victory and the Liberum of Christian piety. The coinage was struck in mints across the western provinces, including London, which had its own active imperial mint during the period, and the wide distribution of his coins suggests that

his administration succeeded in maintaining the basic economic functions of empire across the territory he controlled. Soldiers were paid, supplies were procured, tax revenue was collected, the grain ship still sailed from North Africa to feed the cities of Italy, although Italy itself was now in the awkward position of being controlled by a different emperor. That different emperor was the boy Valentinian II, the younger half-brother of Gratian, who had been left in nominal control of Italy after his older brother's murder, and who was, at this point, about 13 years old, real power in Italy lay with his mother, the Empress Justina,

and with whatever generals she could persuade to fight on her son's behalf. Maximus, who had no particular wish to start a full civil war if he could be recognised as a legitimate colleague instead, opened negotiations with the Eastern Emperor Theodosius. The Eastern Court, which was dealing with its own problems on the Danube frontier, and could not afford a major war in the West, eventually agreed to recognise Maximus as a legitimate emperor of the Western provinces, on condition that he leave Valentinian II in possession of Italy. For about three years this arrangement held, Maximus ruled Britain, and then, and he ruled, and Gaul, Spain and the Rhine frontier from Trier.

Valentinian II nominally ruled Italy from Milan. Theodosius ruled the eastern half of the empire from Constantinople. The empire had three legitimate emperors, an unstable but workable equilibrium, and the period might have ended quite differently if Maximus had been content with what he had. He was not, unfortunately, content with what he had. In 387, deciding that Valentinian II was a child whose throne was effectively unguarded. Maximus crossed the Alps with his army and seized Italy, driving Justina and her son into exile in the east. This turned out to be a serious miscalculation. Theodosius, who had been willing to tolerate a co-emperor in Trier, was not willing to tolerate a usurper who had displaced his own protected colleague in Milan.

The eastern army was mobilised. In 388, Theodosius marched west with a force that included substantial barbarian contingents alongside the regular Roman legions, and the campaign that followed was sure that was sure that was short and decisive. Maximus's army was defeated at the Battle of Save in what is now Slovenia. Maximus himself retreated to the city of Aquilea, near the head of the Adriatic where he was besieged, captured and executed in late August 388. His head, in the Roman tradition for defeated usurpers, was sent on tour through the provinces to be displayed as evidence of the failure of his ambitions, presumably with the kind of slow ceremonial parade that nobody particularly enjoyed, but that Roman administrative habit demanded. His body was buried somewhere near Aquilea in an unmarked

grave. His brief experiment in imperial rule was over. So far, this is the standard story of a competent but ultimately unsuccessful late Roman general, no more memorable than a dozen others from the same chaotic century. What lifted Maximus out of the standard category and inter-legendary status was what happened next, both in the actual political life of the Western provinces he had ruled, and in the long cultural memory of the British, the political consequences first. When Maximus crossed into Gaul in 383 to challenge Gratian, he took with him a substantial portion of the British garrison. How substantial is one of the most debated questions in late Roman British studies,

with estimates ranging from a few specialised attachments to a major reduction of the field army. But the consensus is that he certainly removed troops, and that some of those troops never came back. When his administration in Trier began collapsing in 388, the British provinces were left in an awkward intermediate state. The formal Roman administration was still nominally in place. The garrison forts along the wall and the coast were still nominally manned. Tax was still nominally being collected, but the central reinforcement and replacement system that had kept the British army at full strength for centuries had been disrupted and the disruption was never fully repaired.

The British provinces were now, in some measure, on their own. Later British and Welsh writers, particularly those of the 6th century onward, looking back at the eventual collapse of Roman authority in Britain in the early 5th century, identified this moment as the critical turning point. The monk Gildas, writing in the middle of the 6th century, presented Maximus as the man who had stripped Britain of her defenders for a foolish foreign adventure and left the country exposed to the barbarian raids that would eventually overwhelm it. Gildas, who was an angry man writing for an angry purpose, painted Maximus in lurid colours, calling him a tyrant who had drained Britain of its strength and left only weak echoes of the original army to defend the country against the

Saxons, the Picts and the Scots. The chronicler Nenius, writing two or three centuries later in North Wales, picked up the same story and added further details, presenting Maximus as the architect of the British military disaster and the indirect cause of every subsequent calamity that befell the island. This narrative became the standard British account of the end of Roman Britain, and it survived into modern historical scholarship until the 20th century, when archaeologists and historians began to question whether the picture of Maximus as a single bad actor really explained the slow, complicated process by which Roman authority actually unraveled in the Western provinces. Modern scholarship is a great deal more cautious. The departure of troops with

Maximus in 383 was significant, but not by itself fatal. Roman Britain remained a functioning province for another generation after his death, with continuing investment in coastal defences, continuing minting of coins, continuing import of pottery and luxury goods, and continuing operation of the basic administrative machinery. The actual end of Roman authority in Britain came in the early 5th century, in a complicated process that involved further usurpations, the withdrawal of the few remaining troops, a revolt of the British provincials against further Roman taxation, and the eventual famous letter of the Emperor Anorius in 410, instructing the British cities to look to their own defence.

Maximus was not personally responsible for any of this. He was, at most, one factor in a much longer story, and the British monks who blamed him for the whole disaster were doing the kind of simplification that always becomes attractive when the actual chain of events is too long and too dispiriting to confront in full detail. It is easier to identify a single villain than to acknowledge that an entire civilisation may have ended through accumulated administrative neglect, and Gildes and Nenius were too good at their craft to resist the simpler explanation. The Welsh memory of Maximus took a strikingly different turn. While the chroniclers were busy writing him into the role of imperial scoundrel who had ruined the country, the genealogists were busy writing him into the

role of revered ancestor, whose marriage to a British princess had founded the royal lines of Wales. The two traditions sat alongside each other for centuries in the same body of medieval literature, with no obvious sense of contradiction, and Welsh writers seemed to have been able to hold both at once without difficulty. Maximus the betrayer of Britain and Maximus the founder of Welsh kingship were apparently the same man approached from different angles. The genealogies of the major Welsh dynasties almost universally include him as an ancestor figure. The kings of Powis in Mid Wales traced their line back through a sequence of real and legendary figures to Maxen Liedig, treated as the same man as the historical Magnus Maximus. The kings of Diphed in the southwest included him in their

pedigrees as well, although their line had additional Irish components from the heavy Irish settlement of the region, in the post-Roman period. The kings of Gwynedd, the most powerful and most enduring of the medieval Welsh dynasties, also drew Maxinid descent into their genealogies, although the actual founder of the historical Gwynedd dynasty was a 5th-century war leader named Cuneda, whose connection to Maximus was probably constructed retrospectively to give the line additional Roman prestige.

This is one of the more interesting cultural manoeuvres of the post-Roman British world, the Welsh aristocracy, in a country that had been deliberately impoverished and depopulated by repeated waves of conflict, decided that their political legitimacy required Roman ancestry, and they identified the most plausible Roman ancestor available, and built their family trees backward to reach him. Maximus had spent time in Britain. He had probably had relationships with at least one British woman, possibly the historical Ellen of the genealogies, although her existence as a specific named individual is harder to confirm than her appearance in the genealogy suggests. He had ruled,

briefly, as a legitimate Roman emperor, and his coinage, his administration, and his control of the Western provinces gave him a kind of imperial pedigree that no other late 4th century figure in Britain could match. The fact that he had been killed by Theodosius in Italy was inconvenient, but not particularly damaging to his usefulness as an ancestor, since dead Roman emperors were arguably better ancestors than living ones, in that they could not contradict the genealogies being attached to them. The cult of Maximus was strong enough that medieval Welsh princes were still using him in their political self-presentation in the high Middle Ages. The 12th and 13th century rulers of Gwynedd, in their disputes with the English crown, repeatedly invoked their descent from Maxon

as part of their argument for the legitimacy of their independence. The Welsh poets attached to their courts wrote eulogies that traced the royal line back through the imperial ancestor and used that dissent as evidence that the present princes deserve to be treated as the equals of any king in Christendom. This was effective rhetoric, particularly in a period when Anglo-Norman and English claims to authority in Britain were also being constructed through carefully chosen ancestor figures, and it remained part of the rhetorical arsenal of Welsh political identity until the conquest by Edward I effectively ended the independent Welsh dynasties. The archaeology of the period, meanwhile, tells a story that complicates both the chronicler tradition,

tradition and the genealogical tradition in interesting ways. The fortress at Sagontium, where the Welsh tale of the dream set so much of its action, has been excavated repeatedly across the 20th century, and the picture that has emerged is one of continuous and even intensifying activity in the late 4th century, exactly the period when Maximus would have been the senior military authority in Britain. The fort was not abandoned. The garrison was not withdrawn. The buildings were maintained, the defensive walls were repaired, and there is evidence of continuing investment in the site well after 383. This pattern is not unique to Sagantium. Other forts in North Wales show similar evidence of late occupation, sometimes lasting into

the 5th century and beyond. Whatever Maximus took with him to the continent, it was not the entire garrison of Wales. The British provinces remained militarily active for a considerable time after his departure, and the local communities continued to operate within a recognisably Roman framework, even as the central authority of the empire weakened. What does seem to have changed is the pattern of authority. Without a strong central administration in London or York, individual fort commanders, local landholders, and tribal leaders began to take on more decision-making power for themselves. The line between a Roman officer commanding a garrison and a local warlord protecting a territory blurred over the course of the 5th century, and the kingdoms that eventually

emerged in early medieval Wales had their roots in this period of negotiated, fragmented authority. Some of the new rulers may have been actual descendants of late Roman officers. Some may have been native British aristocrats who had absorbed Roman administrative habits during the centuries of occupation. Some may have been military adventurers from elsewhere in the British world, who established themselves through alliance and force. Whatever the precise individual stories, the political world that took shape in Wales in the 5th and 6th centuries was something genuinely new, neither fully Roman nor fully pre-Roman Celtic, and it was a world that found Maximus a useful symbolic founder figure, even when the precise historical connections

were impossible to verify. There is, in this, a particular kind of irony that the historical Maximus would probably have appreciated, given his evident sense of political opportunity. The man whose imperial career ended in defeat and disgrace, whose head was paraded through the provinces as evidence of his failure, whose memory was attacked by the chroniclers of his own former subjects as the architect of their misfortune, ended up becoming the revered ancestor of nearly every royal line in early medieval Wales. The numismatic evidence for his reign is also worth a brief consideration, since coins are one of the few categories of historical evidence that survive in genuinely large quantities from the period, and they tell a story that the literary sources do not.

Maximus struck coins at mints across the western provinces during his five years on the throne, and substantial hordes of his issues have been recovered across Britain, Gaul and Spain. The British finds are particularly revealing. They show that his coinage was reaching the most distant corners of the diocese, including military sites in northern England and Wales, civilian towns in the lowlands, and the prosperous villa estates of the southwest. The pattern of distribution suggests a function monetary economy operating across the whole of Roman Britain in the years of his reign,

with coins moving along the main roads, through the markets, and into the hands of soldiers who spent them in local taverns and shops. This is not the picture of a province in collapse. It is the picture of a province operating more or less normally under the rule of a usurper, who was by all the practical evidence taking his administrative duties seriously. After his death in 388, the supply of new imperial coinage to Britain became increasingly irregular, and by the early 5th century the inflow of fresh coin had effectively stopped, which forced the British provincials into a series of awkward improvisations involving clipped older coins and the occasional locally minted imitation.

It is not too much to say that the British monetary economy died in the generation after Maximus, although whether that death is to be blamed on him personally is a question that gets debated whenever historians of late Roman Britain gather in sufficient numbers. There is also the question of what happened to the soldiers and officers who, followed Maximus to the continent, and who survived the disaster of 388. Some of them were absorbed into the army of Theodosius after the surrender at Aquilea, since the late Roman state had a pragmatic attitude to defeated rebel troops, particularly when there were continuing barbarian pressures that needed addressing on the Danube and elsewhere.

Others were resettled in less sensitive territories, often in northwestern Gaul, where there was a need for military manpower, and where the local population was used to the presence of garrisons. Some of these resettled veterans, particularly those of British origin, formed the nucleus of the British communities that began to appear in Brittany during the late 4th and early 5th centuries, and the migration of Britons to the Armorican Peninsula, which would in time give Brittany its name and its distinctive Britonic culture, may have begun in part with the dispersal of Maximus's defeated army.

The Welsh tale, which presents Sinan and his followers as recipients of grants of land in Brittany after a triumphant campaign, has the direction of the story exactly inverted. The actual British settlement of Brittany happened after a defeat, not after a victory, but the basic fact that British troops ended up in Armourica during the period of Maximus's career is plausibly grounded in real events.

Folk memory has a tendency to convert defeats into victories when enough centuries have passed, and the Welsh tradition, in this case, simply did what folk traditions everywhere have always done. The relationship between Maximus and the Christian church of his own day deserves a paragraph as well, since it is one of the more revealing aspects of his actual administration. Maximus was a baptized Christian, like most senior Roman officers of the period, and he took the religious responsibilities of his office seriously, in his own slightly heavy-handed way.

The most famous incident of his religious career involved a group of heretics led by a man named Prisilian, who was an ascetic teacher in Spain and Gaul, and whose ideas about the proper conduct of the spiritual life had attracted both enthusiastic followers and serious opponents. Prisilian's enemies brought charges against him before Maximus, and Maximus, after consulting with various bishops, allowed Prisilian and several of his followers to be put to death. This made Prisilian the first Christian executed for heresy by a Christian state, a distinction that nobody, including Prisilian himself, would have wanted. The execution provoked horror among many leading bishops of the period, including the Great Martin of Tor,

who refused to have anything more to do with the bishops who had encouraged the executions. The episode shows Maximus operating at the awkward intersection of imperial authority and religious orthodoxy, where late Roman emperors of every rank were finding themselves drawn into doctrinal disputes whether they wanted to be or not, and it does not particularly redound to his credit. It does, however, demonstrate that he was not, as the Welsh genealogies would have him, simply a remote symbolic ancestor figure. He was a real man making real decisions in a complicated political and religious environment, and some of those decisions were unambiguously bad ones.

His brief reign in Trier, which historians treat as a footnote in the long collapse of the Western Empire, became the foundational moment in the political genealogy of an entire successor culture. He was more important in death than he had been in life, and more important in legend than he had been in fact, and the woman he probably did marry in Britain became, through the slow alchemy of folk memory, the namesake of the principal Roman roads of the country. It is the kind of afterlife that very few people achieve,

and it suggests that whatever else the historical Maximus may have lacked, he had a real talent for being remembered in ways he could not possibly have anticipated, whatever else the historical Maximus may have lacked. He had a real talent for being remembered in ways he could not possibly have anticipated, The Welsh genealogies kept him alive long after Theodosius's executioner had done his work, and the legend of his dream proved more durable than the imperial throne he briefly held. If the imperial career of Maximus left Wales with a powerful symbolic ancestor, the Roman occupation that preceded him left the country with something more tangible,

namely a physical infrastructure of forts, roads, towns and engineering works that shaped the landscape long after the legionaries had marched away. The Roman conquest of Wales had been a slow and unusually difficult business, far harder than the conquest of southeastern England, and the network of military installations that the Romans built across the country was correspondingly elaborate. By the height of the occupation in the second century, the Roman army had constructed something in the order of two dozen forts

across what is now Wales, scattered across the river valleys and along the high passes, all of them connected by a system of paved roads and signal stations designed to allow troops and dispatches to move from one fortress to the next within a day or two of marching. This was not the light footprint of an empire that expected the locals to behave themselves. It was the heavy footprint of a power that had learned, through painful experience, that the various Welsh tribes were not in the mood to be casually pacified. The conquest itself had been launched by the Governor Swetonius Paulinus in the early 60s of the first century,

with the campaign that famously ended on the island of Anglesi, where the Roman army cut down the sacred groves of the druids in the smoke and confusion of a battle that ancient sources described in suitably operatic terms. Paulinus did not get to enjoy his victory since news arrived almost immediately that Budica had risen in revolt back in eastern Britain and that several major cities had been burned to the ground in his absence,

requiring him to rush back across the country to restore order. The pattern of his country is. campaign in North Wales, however, set the template for the Roman approach to the country. Heavy commitment of troops, aggressive forward movement, the destruction of religious centers, the construction of permanent fortifications at strategic points, and then the long, slow, patient business of holding the territory through generations of garrison duty, building infrastructure, intermarrying with local populations, and gradually persuading the surrounding tribes that resistance was both expensive and pointless. Two of those permanent fortifications eventually grew

into the most important Roman sites in the country. The first was the Legionary fortress at Karelion, on the River Usk in southeastern Wales, established in the late 70s of the first century to house the Second Legion Augusta and its supporting units. Karelian was not a small frontier garrison, but a major military installation, the permanent base of one of the three legions stationed in Britain throughout most of the Roman occupation, and the scale of the site reflects the importance the Romans placed on holding Southern Wales securely. The fortress covered something like 50 acres within its walls,

large enough to accommodate the full complement of the Legion in barracks, plus headquarters buildings, a hospital, granaries large enough to feed several thousand men through a winter, workshops for the manufacture and repair of military equipment, baths, and a forum for daily administrative business. Outside the walls, a substantial civilian settlement grew up to service the garrison, with shops, taverns, brothels, and the various other amenities that a population of soldiers required and was prepared to spend its pay on.

The most spectacular surviving feature of Karelian is the amphitheatre, which sits just outside the line of the original fortress walls, and which is one of the best-preserved Roman amphitheaters in Britain. The structure was built around 90 AD and could hold around 6,000 spectators, comfortably accommodating the entire legion plus a substantial number of civilians and visitors. The arena floor is still clearly visible today, oval-shaped and sunk below the level of the surrounding seating,

with the eight entrance passages still mostly intact and the foundations of the original timber and stone seating banks rising in a ring around the perimeter. What was performed in this arena is one of those questions that medieval Welsh tradition felt confident answering and modern archaeology approaches more cautiously. The medieval Welsh, looking at the visible ruins, identified the amphitheatre as the roundtable of King Arthur, who had supposedly held his court at Kalyan, according to the chronicle of Geoffrey

of Monmouth. This was, as identifications go, completely wrong, but extremely persistent. Welsh writers across the medieval period treated the amphitheatre as Arthur's seat, with a confidence that no amount of contrary evidence would shake, and visitors well into the 19th century were still being shown around the site as a piece of Arthurian heritage. The actual function of the arena was naturally more pedestrian. It hosted training exercises for the Legion, formal ceremonies, the occasional gladiatorial combat or wild beast show on holidays, and the kind of crowd-pleasing entertainments that allowed soldiers to spend their days off in something approximating relaxation. The fortress at Karelion also contained one of the more interesting institutional structures

of the Roman British world, namely a substantial bathing complex that served both the garrison and the civilian population. The remains of the legionary baths have been excavated extensively and now sit beneath a modern museum building, where visitors can walk along glass walkways above the original Roman flooring and look down at the cold plunge pool, the warm room, the hot room and the underfloor heating system that kept the whole arrangement at a comfortable temperature year round. The scale of the baths is one of the more striking demonstrations of Roman investment in the comforts of military life. A legion stationed at the western edge of the empire in the second century in a damp, mountainous country with limited fuel resources and a hostile native

population was nonetheless provided with a bathing facility that would have been considered respectable in many Italian cities. The thermal sophistication of the system is genuinely impressive. Hot air circulated under raised floors and through hollow wall tiles. Water was channeled through lead pipes from a nearby aqueduct. Bathers moved through a sequence of rooms at progressively higher temperatures, finishing in the hot room where they could sweat out the accumulated grime of frontier service before plunging back into cold water for the closing rinse. By the standards of available alternatives in First Century Britain,

this was frankly Paradise, and one understands why the surrounding civilians, settlement grew so rapidly. Wherever the legions went, hot water followed, and people who had previously made do with rivers and rainwater barrels were not slow to take advantage of the new arrangements. About 30 miles south of Karelyan, on the southern coast of Wales, sat the town of Karewant, which represented an entirely different aspect of the Roman presence in the country. Kerwant was not a military fortress.

It was a civilian town, founded by the Romans, as the administrative centre for the local British tribe, the Siluris, who had given the conquering legions a particularly difficult time during the original campaigns of the first century. The Silurs had been one of the most stubborn opponents of Roman expansion in Britain, fighting the legions to a standstill for several decades and inflicting some serious defeats before being finally subdued. Once they had been brought into the empire, they were granted, as was the standard Roman practice, their own tribal capital, complete with the formal civic apparatus of a Roman town. The place was officially named Venta Silurum, the market of the Silurus,

and it was constructed on a relatively modest scale by the standards of major imperial cities, but with all the essential features of Roman civic life. There was a forum, a basilica for the conduct of legal and administrative business, a public bathing complex, several temples, a network of paved streets laid out in the usual Roman grid pattern, and substantial private houses for the local Romanised aristocracy. The most impressive surviving feature of Kerwent is the town walls, which still stand to a considerable height around large portions of the original perimeter.

The walls were built in the late 3rd or early 4th century during the period of intensified frontier defences across the Roman world, and they were not the kind of decorative civic boundary that some Roman cities maintained. They were serious defensive structures, four metres thick at the base, with bastions projecting at regular intervals to allow flanking fire from artillery against any attacker who might try to approach the gates. The fact that Kerwant invested in fortifications of this scale tells us something about how the Roman administration in Britain assessed the security situation

by the late 3rd century. The town was no longer relying on the Legion at Kerleon for protection. It was, in effect, becoming responsible for its own defence, which was a reasonable response to the increasing irregularity of central imperiality of Central Imperial Authority, but which also marked a significant shift in the way Roman civic life functioned. A walled town in the late empire was a different kind of place from an unwalled town in the high imperial period. It was a place that expected to be threatened, and the texture of daily life inside it was correspondingly more cautious. Carewant's later fate, the slow contraction of its urban

life across the fifth and sixth centuries. We will return to shortly, because it is one of the clearest archaeological pictures we have of how a Roman town actually died. For now, it is enough to say that the walls outlasted almost everything they were built to protect. Up in the north, the third great Roman site in Wales, the fortress of Sagontium near the Mani Strait, has its own distinctive history that deserves a closer look. The fort was established in the late first century, during the campaign of conquest that pushed Roman authority across northwest Wales, and it served as a key garrison post controlling the approaches to the island of Anglesey, which was both economically important as a source of grain and strategically important as a centre of native British religious life.

The site sits on a low ridge overlooking the strait, with a clear view across to the mountains of Anglesey, and the position has the kind of natural strength that Roman military engineers always recognised when they saw it. There is a reliable water supply nearby. The high ground gives commanding views of the surrounding country. The coastline below provides easy action access to seaborne supplies, and the inland approaches can be controlled through the network of mountain passes that the Romans rapidly fortified with subsidiary watchtowers and minor forts. The excavation of Sagontium in the 1920s by Mortimer Wheeler is one of the foundational episodes in modern British archaeology, and it transformed not just the understanding of the site

itself, but the methodology of how Roman military sites should be investigated. Wheeler, who was a young and exceptionally talented field archaeologist at the time, applied a systematic technique of carefully recorded stratigraphic excavation that was relatively new in British practice, mapping the various phases of occupation through the changing layers of soil and the dating evidence of pottery, coins and inscriptions. What he found surprised him. Sorgantium had been occupied continuously from its foundation in the late first century

right through to the very end of the 4th century, with the latest substantial phase of activity dating to the period after about 380. Exactly the years when Magnus Maximus was the senior military commander in Britain, the fort was not, as had previously been assumed, abandoned in the early or mid-fourth century with the rest of the more obscure Welsh garrisons. It was reinforced, rebuilt, and apparently expanded right up to the end of the formal Roman administration in Britain, and the activity at the site lasted long than the activity at almost any other fortress in the country. This was not what anyone had expected. The conventional model of late Roman Britain, as it had stood before Wheeler's excavations, assumed that the Western frontier had been progressively de-garisoned during the 4th century,

with troops being pulled back to defend more strategically important areas. Sargontyum did not fit this model. It looked, instead, like a site that had been actively maintained as a forward position, right up to the moment when central Roman authority finally collapsed, with continuing investment in barracks, defensive walls, and the basic infrastructure of military life. Whatever the high political situation in the rest of the empire, the commanders at Sagontium had been getting their resources, paying their men, and keeping their fort in good order

well into the period when other parts of Roman Britain were already invisible decline. The implications for the historical maximus are interesting, since it suggests that his own administration in the British provinces had not, in fact, stripped this part of the country of its garrisons in the way that later chroniclers would claim. The men at Sagontium were still there, and apparently still well supplied, even after he had crossed to the continent. The three great Roman sites were tied together by the road network we have already met under the name San Helen, the same causeways that the medieval Welsh would later attribute to a Roman emperor's wife.

It was not a single straight line, but a network of connected stretches, taking the shortest practical path through the mountainous terrain and adapting to the engineering possibilities of each particular landscape. In some places the road runs along the floor of a valley. In other places it climbs over high passes, with substantial cuttings into the hillside and embankments built up to maintain a reasonable gradient. The construction quality varies, but is generally impressive, with a layered foundation of large stones, smaller stones, and in the best-preserved sections a paved surface of dressed flagstones laid carefully across the roadbed. The engineering of these roads is one of the more underrated achievements of the Roman occupation of Wales. The country is not friendly terrain for road builders. The mountains are steep, the rivers are unpredictable, the weather turned surfaces into mud at extremely short notice,

and the population of the surrounding hills was, for a substantial part of the occupation, actively hostile and given to ambushing supply convoys. Building a road network across such terrain required not just engineering skill, but a substantial deployment of military labour, with legionaries themselves doing much of the heavy work of clearing routes, laying foundations, and constructing bridges and culverts. The work was done not by specialised civilian contractors, but by soldiers, who in the Roman system were expected to be capable of construction work

as part of their general military duties. This had the side effect of ensuring that the roads were genuinely military in their priorities. They went where the army needed them to go by the most direct practical route, with little concern for the convenience of the local civilian population. Some of the local communities adapted by relocating closer to the road network. Others continued to live where they had always lived, with their daily rhythms operating on a different geography from the Imperial one. The road itself, even in its surviving fragments, is a remarkable thing to walk along today.

The clearer sections run across the upland country of central Wales, where modern agriculture has not entirely covered the original surface and where the line of the road can still be traced for kilometres at a stretch. In some places it runs as a raised causeway across boggy ground, with the original Roman drainage ditches still visible on either side. In other places it climbs the side of a mountain in a long diagonal cut, with the roadway perhaps four to five metres wide and the original stone surface visible in patches where the turf has been worn away by walkers and grazing sheep. Walking along these stretches is one of the more direct experiences of contact

with the Roman past that British landscape can offer. The road does not feel ancient. It feels operational. You can imagine, with very little effort, a column of legionaries moving along it at the standard marching pace, 20 miles a day, with the supply wagons rolling behind them and the scouts ahead, and the rhythm of military movement absorbed.

into the geometry of the route itself. What happened to all this infrastructure after the Roman administration ended in the early 5th century is a question that medieval Welsh tradition answered in one way and modern archaeology answers in another. The medieval tradition, as we have already seen with the tale of the dream, attributed the construction of the road network to a Welsh princess, gave it a romantic provenance and treated the surviving stretches as a continuing visible reminder of a glorious imperial past. The archaeological reality is more prosaic. The roads continued in use, but they were not maintained. The original paved surfaces deteriorated through frost damage, plant growth, and the ordinary wear of human and animal traffic. The drainage ditches silted

up. Bridges decayed and were not replaced. Cullverts collapsed. By the high Middle Ages, what had been a Roman road system was a network of roughly defined routes, sometimes still recognisable as engineered roadways, sometimes reduced to mud tracks following the original line. Travellers continued to use them, because there was nowhere else to go, but they used them in increasingly degraded condition. The forts told a similar story. Kirlian, with its grand legionary fortress, was largely abandoned as a military installation by the early 5th century, although the surrounding civilian settlement persisted and gradually evolved into the modern small town. The amphitheatre survived, becoming over the centuries an object of medieval

curiosity, and eventually the source of the Arthurian legends already mentioned. The bathhouse was buried beneath later occupation layers and was not rediscovered until the 20th century. Cairwant, as we have already seen, contracted dramatically and never really recovered, although its walls survived to ring a much smaller medieval village. Sagontium, despite its remarkable late occupation, was eventually abandoned as a fort sometime in the 5th century, although the site continued to attract activity for centuries afterwards, with what may have been a small early Christian community, establishing itself within, or near, the old walls during the post-Roman period. The Welsh princes of the early medieval period treated all three sites as places of high symbolic importance, and Carnarfon, the modern

town that grew up next to the ruins of Sagantium would eventually become the principal royal centre of the Kingdom of Gwyned, with its later medieval castle constructed deliberately to draw on the Roman associations of the location. What emerges from a large of the land of the land. What emerges from a long look at the Roman heritage of Wales is a landscape in which the Imperial past did not so much vanish as gradually transform into something else. The fortresses and towns and roads of the Roman period were never completely abandoned. They were absorbed into the political and cultural life of the post-Roman world, used in different ways, attached to different stories, fitted into different frameworks of meaning. The medieval Welsh did not look at Roman ruins and see foreign occupation. They looked at

them and saw their own ancestral inheritance, the work of imperial figures who had become, through marriage and legend, members of their own genealogical family. This was a remarkable cultural achievement, and it is one of the reasons that Welsh identity in the early medieval period was able to maintain itself with such tenacity, even under the considerable pressure of Anglo-Saxon expansion to the east. The Welsh had not lost their imperial past. They had absorbed it, and the surviving stones of Kaelian and Kerwent and Sagontium were for their them, not the remains of a foreign empire, but the foundations of their own continuing political identity. The relationship between physical landscape and cultural memory in this period

repays close attention, because it shows how a society can convert the visible remains of an earlier civilisation into the symbolic furniture of its own self-understanding. Stones do not, of themselves, have meanings. They acquire meanings through the stories that are told about them, and the stories that the medieval Welsh told about their Roman inheritance gave those stones a place in the imaginative geography of the country that they would not have had otherwise. Around these three great sites, Kirlian, Kerwant, Sagontium, clustered a constellation of smaller forts and a quiet but lucrative gold mining operation at Dolaukuthi, the whole apparatus tying whales into the imperial supply system for over three centuries. The amphitheatre at Kaelian, in this sense, did not become

Arthur's roundtable because of any factual connection between the two. It became Arthur's roundtable because the medieval Welsh needed a roundtable somewhere in their landscape, and the amphitheatre was conveniently the right shape, the right size, and the right kind of impressive ancient structure to fill the role. The fortress at Sagontium did not become the seat of Maxen-Liedig because Maximus had ever sat in council there in any documented way. It became the seat of Maxen because the medieval Welsh needed an imperial setting somewhere in their north, and Sorgantium was the most obvious candidate available to them. The connections were not invented out of nothing. They were built on real archaeological foundations and on plausible historical associations, but the meanings attached to those

foundations were the work of generations of storytellers, poets, and community elders who had decided what kind of past they wanted to inhabit. There is a final point worth making about the legacy of the Roman occupation in Wales, which concerns the long durability of the basic infrastructure even after the people who built it had been dead for a thousand years and more. The line of the San Helen, where it has survived, is sometimes still used by modern walkers, hikers, and even occasional vehicles in remote upland areas. The walls of Kerwent are intact enough that you can walk along the top of them in places. The amphitheatre at Kerleon is in excellent condition for a structure approaching 2,000 years old, and modern archaeological work

continues to recover material from the site that adds to the picture of its function and its eventual abandonment. The footprint of Sagontium is preserved beneath modern Kernhafen, with the foundations of the original fort visible in a public archaeological park. None of this would survive if the Roman engineers had not built well to begin with, and the long survival of these features is a tribute in its understated way to the quality of the construction. The Roman Empire ended a long time ago in this part of the world. The roads, walls, and amphitheaters it left behind are still here, performing their final and perhaps most enduring function,

which is to remind a much later population that civilizations come and go, but that the things they build, when built well, can outlast almost anything else they leave behind. If the Roman administration in Britain had collapsed in the early 5th century, The question of what happened next was not a question that anyone living through the experience felt confident answering at the time. The contemporary record is frankly awful. There are scattered references in continental sources to events in Britain,

mostly brief and often confused. There are a few inscriptions on stone monuments. There are coin hordes and pottery fragments and the post-holes of timber buildings, all of which can be analysed but none of which can speak. And there is, almost uniquely among the sub-sufficient, surviving evidence for the period, a single book, written in furious Latin by a man who genuinely lived through some portion of the events he describes, and who decided that what his fellow

Britons most needed was a long, biting sermon about how badly they had been behaving. The book is called De Exquidio at Conquestu Britannier on the ruin and conquest of Britain, and its author is known to history as Gildes, sometimes given the additional title of the Wise, although wisdom in his case meant something more like sustained moral indignation than serene philosophical reflection. Gildes wrote, probably in the second quarter of the sixth century, although the precise date is one of those questions that scholars have argued about for generations without ever quite reaching consensus. Some place him as early as the five-tens. Others put him as late as the five-forties. The most plausible window seems to be somewhere in the five-thirties or early five-forties,

after the period of relative stability that followed the British victory at the Battle of Mount Baden, which Gildes himself mentions as having taken place around 44 years before he was writing. Where exactly he was writing from is also unclear. Various traditions have located him in North Wales, in South West England, in Brittany, and at various monastic communities in between. The text itself does not specify a location, although the writer clearly has a strong sense of the geography of Westpherson. Britain, and refers to the rulers of the region, with the kind of familiarity that suggests

he had spent considerable time in their territories. He may have been a monk attached to a monastery somewhere in the Western British world, but the precise institution is unrecoverable, and the man himself remains a slightly obscure figure, even though his book is one of the most important documents of the period. What we do know about Gilders, mostly from his own work, is that he was educated, not just literate, but educated to a very high standard. by the survival levels of the period, with a fluent and often difficult Latin prose style that drew on the entire tradition of biblical, classical and patristic literature available to a serious clergyman of his time. He quotes from the Old Testament, the New Testament, the Apocrypha, the writings of the early church fathers,

and a handful of classical authors, weaving them together in long passages of dense allusion that assume his readers will recognise the references and adjust their understanding accordingly. This level of literary culture in 6th century Britain is genuinely surprising, and it tells us something important about what the immediate post-Roman period actually looked like in terms of intellectual life. The collapse of the Imperial Administration had not destroyed the educational tradition. Somewhere in Western Britain, in the early 6th century, there were schools or monasteries or family households where Latin was still being taught at a high level, where books from the Roman world were still being read, and where young men of promising ability could acquire the kind of literary

training that allowed Gildas to compose his treatise. The lights had not gone out entirely. They had in some places dimmed considerably, but in others they had continued to burn at something like their old brightness. The book itself is divided into three sections of unequal length, and it is worth understanding the structure because it shapes how the contents are usually read. The first section is a historical preface, in which Gildas summarizes the history. history of Britain from the Roman conquest down to his own day. This section is the part that historians have tended to focus on, since it contains the only surviving narrative of the immediate post-Roman period written by anyone close to the events themselves. The second section is a long denunciation

of the British kings of his own time, naming five of them by name and accusing them of various spectacular sins. The third section is an even longer denunciation of the British clergy, accusing them of corruption, ignorance, neglect of duty, and a long list of other failings. The proportions of the three sections are roughly one part historical narrative to three parts moral fury, which gives some sense of where the author's actual interests lay. He was not writing history for its own sake. He was writing a sermon, and the historical material was provided primarily as a kind of moral framework against which to judge the present. The historical framework, when reduced to its bones, run something like this.

Britain was a fertile and well-watered island, blessed by God with natural resources, but inhabited by a stubborn and rebellious people. The Romans came, conquered the island, and gave it the benefits of imperial civilization, including, in time, Christianity. The Britons, who in Gildes' view were always a step behind in moral seriousness, accepted these benefits without quite earning them. When the Roman administration eventually withdrew, the Britons were unable to defend themselves against the raiding peoples of the North, the Picts and the Scots, and they made the catastrophic

decision to invite Saxon mercenaries from across the North Sea to fight on their behalf. The Saxons, having been given a foothold, decided that they liked the country and proceeded to take it over. The Britons fought back under the leadership of a figure Gildes calls Ambrosius Orelianus, who was described as the last of the Romans, of noble, parentage, and who organised a sustained military resistance to the Saxon advance. The Britons won a series of victories, culminating in the famous Battle of Mount Baden, which Gildes treats as a major turning point. After Baden, there was a period of relative peace, but the British leaders of Gildes' own day, instead of using this peace wisely, had fallen into civil war, oppression of their

own people, sexual misconduct, and various other forms of moral degradation. Hence the book, The figure of Ambrosius orelianus is one of the more intriguing in the entire text, and not just because he is described in flattering terms. He is the only late British military leader whom Gildas names, and he has given a Roman name, a Roman pedigree, and a Roman virtue. The implication is that the British resistance to the Saxon invasion was led by a man who consciously identified with the imperial tradition, who saw himself as continuing the work of Roman civilization against barbarian and Cursion,

and who fought under the banner of imperial values rather than under any specifically British or tribal identity. This is interesting in itself, and it has fed into a long and complicated scholarly debate about how the post-Roman British thought of themselves. Were they Britons, in the sense of a distinct ethnic and political identity? Were they Romans, in the sense of inheritors of imperial citizenship? Were they something in between, with a complex layered identity that depended on the context, The answer is probably some version of all three, and Ambrosius is a useful figure precisely because he reminds us that the late Roman cultural framework did not disappear with the withdrawal of the legions.

It continued in the minds and self-understanding of the British elite for at least another century and probably longer. The Battle of Mount Barden, which Gildes mentions but does not describe in any detail, is one of the most famous events in early British history, and it has become inseparable from the legend of King Arthur. Later sources, including the Chronicle and Nenius, writing two or three centuries after Gildas, would attribute the victory at Baden directly to Arthur, presenting him as the British war leader who turned the tide against the Saxon advance. Gildes himself, awkwardly enough, does not mention Arthur at all. There is no Arthur in his book. There is Ambrosius or Elianus,

who organises the resistance, and there is the unnamed victor of Baden, who may or may not be the same person. Whether this silence is significant is one of the great unanswered questions of Arthurian studies. Some scholars argue that Arthur did not exist as a historical figure, and that Gildes' silence is exactly what we should expect from a contemporary writer who knew of no such person. Others argue that Gildes had personal reasons for not naming Arthur, possibly a family quarrel or a doctrinal dispute, and that the silence does not necessarily mean Arthur was not real. Still others argue that Arthur was real but obscure during his own lifetime, and that the legend grew up around his name only later, attaching itself to genuine military victories whose original commanders had been forgotten.

None of these positions can be proved or disproved on the available evidence, and the question continues to generate scholarly literature at a steady rate, with no sign of resolution. What Gildas was actually interested in, however, was not the documentation of military events, but the moral indictment of the rulers of his own day. The five kings he names by name are given the full benefit of his rhetorical skill, which was considerable and not entirely fair. The first is Constantine, ruler of the southwestern kingdom of Damonia, which corresponded roughly to the modern counties of Devon and Cornwall.

Gildes Constantine of a particularly horrible crime, the murder of two royal princes inside a church where they had taken sanctuary. He describes Constantine as having forsworn his oath to God, broken into the holy place, killed the young men in the arms of their abbot, and then continued to live as if nothing had happened. The Latin in this passage is some of the most ferocious in the entire treatise, and it gives a sense of how Gildas approached his subjects when he had real moral material to work with. The second ruler named is Aurelius Caninus, whose territory is unclear but who is accused of civil wars, sexual misconduct, and the persecution of his own family members.

The third is Vortipur, ruler of the southwestern Welsh kingdom of Demisha, who is described as a grey-haired old man whose moral failings include the murder of his wife and an incestuous relationship with his own daughter. The fourth is Cunaglasus, who is accused of, among other things, having abandoned his lawful wife in favour of his sister-in-law, who happened to be a consecrated nun. The fifth and most prominent is Maglacunus, ruler of the northern Welsh kingdom of Gwynedd, who is given the longest and most detailed treatment.

Maglacunus, who is also known to Welsh tradition as Melguin Gwynedd, was apparently the most powerful British ruler of his day. Gildas calls him the dragon of the island, a term that suggests both his political dominance and his spiritual destructiveness. The catalogue of his sins is extensive. He is accused of having begun his reign by murdering his uncle, whose throne he then seized. He is accused of having taken monastic vows in his youth, and then abandoned them to return to royal life. He's accused of having killed his nephew in order to take his nephew's wife, and of having abandoned his own original wife in the process.

He's accused of patronising bards and poets who flattered him in verse, instead of supporting genuinely religious men who would tell him the truth about his conduct. He's accused of being more powerful than any other British ruler and more sinful in proportion. The whole portrait, while clearly polemical and probably exaggerated in the details, has the kind of vivid, specific quality that suggests Gildus knew his subject personally, or at least knew people who did. Maelgwin was a real king, and he is recorded in Welsh genealogies, and in the annals of the Welsh Church, where he is remembered both as a powerful figure and as a complicated one, capable of acts of generosity to the church alongside the various darker actions Gildas catalogs.

Whether Gildes' portraits of these rulers are accurate is another question that has generated considerable scholarly debate. The basic sins he attributes to them, the murders, the broken oaths, the sexual misconduct, the civil wars, are entirely plausible for the period and the social context. Early medieval kingship was often a violent business, with succession disputes routinely settled by assassination and political marriages, routinely complicated by accusations of incest in the technical canon law sense. Gildes may well have been describing real events. On the other hand, his rhetoric is so heated

and is framing so moralistic that it is impossible to take the details entirely at face value. He is not a court chronicler trying to record what happened. He is a preacher trying to terrify his audience into repentance, and the specific accusations may be accurate in outline while being distorted in emphasis or partial in their selection of facts. Modern scholars tend to accept that the King's Gildas names probably did some of the things he accuses them of, but to be cautious about the precise details

and the overall picture of moral decline that he constructs. The third section of the book, on the British clergy, takes a similar approach to the kings. The clergy are accused of various vices, ranging from simony, which is to say the buying and selling of church officers. to ignorance of basic doctrine, to neglect of pastoral duties, to active immorality. The portrait is unflattering, although Gildas is careful to note that there are some good clergymen still in Britain, and that his criticisms are aimed at the corrupt rather than at the church as an

institution. His standards for clerical behaviour are extremely high, and probably impossibly high, and it is difficult to know how much of the picture he paints reflects actual widespread corruption, and how much reflects the ordinary spread of human imperfection in any large clerical body. Either way, the section is the longest in the book, and it leaves no doubt about Gildes's view that the spiritual condition of his country was as compromised as its political condition. The influence of the book in subsequent centuries was enormous, although not always in ways that would have pleased its author. Bede, the great 8th century Anglo-Saxon historian, used Gildas as his principal source for the period before his own time, and Bede's narrative of the British appeal to

the Saxons, the subsequent betrayal, and the long British retreat into the Western Mountains was lifted almost wholesale from Gildes's pages. Bede's perspective, however, was very different from Gildes's. Bede was an Anglo-Saxon writing the history of the Anglo-Saxon people, and he treated the Saxon takeover of Britain, not as the catastrophe Gildas described, but as the beginning of his own people's story, the same events read very differently when filtered through different national perspectives, and Bede's success in presenting the Anglo-Saxon arrival as a kind of providential transition rather than a moral disaster, effectively neutralised Gildas's central argument while preserving his factual outline. Later medieval historians, building on Bede, continued

this pattern, treating Gildas as a useful source of information while quietly setting aside his moral framework. The result is that the the standard medieval English account of the post-Roman period in Britain, owes more to Gildas than to any other writer, but it has been thoroughly de-gilderised in tone, with the fury removed and only the chronology preserved. Welsh tradition, naturally, took the opposite approach. Gildas was, after all, a British writer condemning a barbarian invasion of his country, and Welsh chroniclers and poets continue to use him as a kind of patron saint of British grievance against the Saxon takeover. The Welsh historical tradition,

as it developed across the medieval period, kept Gildes' moral framework intact while expanding the historical material into something far more elaborate. The Britons in this tradition were not just unfortunate victims of Saxon aggression, but the original inhabitants of the island who suffering at the hands of the invaders

deserved both remembrance and eventual revenge. The Welsh prophecies of the medieval period, which we have already encountered through the figure of Merlin, drew heavily on the Gildan model of British Exxion, exile and eventual restoration. The whole imagined arc of Welsh political destiny, from the loss of the island under the Saxons to the eventual reconquest under some future hero, has its deepest roots in the Gilders treaties. The book also became, eventually and indirectly, one of the foundational sources for the legend of King Arthur. Guildess himself, as already noted, did not

mention Arthur. But the British victory at Mount Baden, which Gildas did mention, became attached in later sources to Arthur as the supposed leader of the British forces, and through this attachment Arthur became the central heroic figure of the entire Arthurian cycle. Without Gildes's brief reference to a major British victory of his own time, there would have been less material for Nenius and the later chroniclers to build upon, and Arthur, as a literary figure, might have remained a much shadowier presence. The fact that Gildes' silence about Arthur has, paradoxically helped to fuel, rather than dampen Arthurian speculation, is one of the more curious quirks of historical reception. A writer who said nothing about a possible historical figure

has ended up being treated as one of the most important sources for that figure's existence, which is the kind of thing that happens only when the surviving evidence is thin enough that every gap becomes meaningful. While Gildas was thundering at his contemporaries about their moral failings, the actual physical landscape of post-Roman Britain was undergoing the slow, quiet transformation that we have already glimpsed in the contraction of the Welsh towns. The most striking case is the one we have already touched on briefly, the town of Kerwant in the southeastern corner of Wales, which deserves a closer and more sustained look because it offers one of the clearest archaeological pictures of how a Roman urban

community actually died. The death of Kerwant was not a sudden event. It was a long, drawn-out process, lasting most of two centuries, and it involved the gradual unwinding of every feature of Roman civic life that had given the town its original character. In the late 4th century, when the rest of Roman Britain was still functioning more or less normally, Kerwant had reached the high point of its development. The forum at the centre of the town was a substantial complex of administrative buildings, with the basilica facing onto the open square where the regular markets were held. The bathhouse in the southwestern quarter was operating to the full Roman standard, with hot water, steam rooms, and a steady supply of fuel for the underfloor heating system.

Several temples within the walls served the various religious traditions of the population, including a substantial temple in the town centre that may have been dedicated to a Romano-British divinity combining local Celtic worship with Mediterranean iconography. Private houses range from modest urban dwellings to elaborate towns, residences with their own private bathing suites, mosaics and central courtyards. The streets were paved, the drains worked, and the population, which probably numbered between 2 and 3,000 at its peak, had access to most of the comforts that a small Roman provincial city could provide. By the standards of the period, this was respectable urban living, and the inhabitants

had every reason to expect that their grandchildren would live in similar circumstances. The first signs of trouble, in the form that the archaeology can detect, appear in the late 4th and early 5th centuries. The forum complex begins to show signs of reduced maintenance, with some of the columns and walls being repaired in less expert ways than the original construction. The bathhouse operations become more irregular, with periods of full functioning interrupted by periods when the heating system is clearly not being maintained. The volume of imported pottery arriving at the town drops noticeably, with fine wares being replaced by locally produced ceramics of lower quality. The coinage in circulation also shifts, with new imperial issues becoming rarer and older coins continuing in use long past their normal period of replacement.

None of these changes by themselves would mark a catastrophic decline. They are the kind of slow shifts in standards that any community can experience during a difficult period and recover from when conditions improve. But they were the first warning signs, and the conditions did not improve. By the middle of the 5th century the deterioration had become more severe. The forum had effectively ceased to function as an administrative centre, with parts of the complex being divided up into smaller domestic spaces, occupied by families who had moved in from the surrounding countryside,

or from wealthier districts of the town that were now too expensive to maintain. The bathhouse had been abandoned, with the heating system no longer working and the building partially collapsed. The temples had ceased to be used, partly because of the conversion of the population to Christianity over the preceding century, and partly because the kind of organised civic religious activity they had supported was no longer happening. The walls of the town, which had been the great pride of the late Roman investment in the place, were still standing but no longer being patrolled or maintained,

with vegetation beginning to grow out of the joints between the stones. The population, meanwhile, was contracting. Some of the inhabitants had moved away, presumably to the country. countryside or to other towns that were holding up better. Others had died and not been replaced. The streets that had once been busy with traders and ox-carts were increasingly quiet. The shops along the main roads had closed one by one, as their owners died or moved out and no one took up the businesses. The town that had supported a population of two or three thousand at its peak now supported perhaps a few hundred, and even those few hundred were spread out

across the original urban area in a pattern that no longer reflected the original Roman layout. The grid of streets was still visible, but people were now living in scattered clusters rather than along the main thoroughfares, and substantial portions of the walled enclosure were effectively empty. By the year 500, give or take a generation, the transformation was essentially complete. Kerwant was no longer a town in any meaningful Roman sense. It was a settlement of perhaps 50 or 100 families living among the ruins of a much larger urban complex that they did not have the resources or the social organisation to maintain. The forum had been abandoned, with parts of it being dismantled for building stone and other parts simply left to collapse. The bathhouse was a ruin,

with trees growing up through the broken arches of the heated rooms. The temples had been forgotten, except where parts of their structures had been incorporated into newer Christian buildings. The Roman cemetery outside the walls had been replaced by a Christian cemetery inside the walls, which was a violation of all Roman urban regulations, but which made sense in the new world where the regulations no longer applied. Trade had stopped. Coinage was no longer being accepted as a medium of exchange. The few imported goods that still reached the town came in small quantities, irregularly, and were too expensive for most of the inhabitants to afford. The lights of Roman urban civilization had not gone out in a single dramatic moment. They had flickered and dimmed and finally settled into a much

smaller, much quieter glow. The countryside around Kerwent was undergoing a parallel transformation. The villas that had ringed the town in its prosperous period, the country estates of the local Romanized aristocracy, were being abandoned one by one. Some of them had been actively destroyed by raids or fires, possibly by Saxon raiders pushing up the seven estuary, possibly by internal political disputes. Others had simply been allowed to decay, with their owners moving away or dying out and no one taking over the properties. The fields that had once been cultivated for the urban market were reverting to woodland, with the brambles and ash trees and hawthorn that always come back when the ploughing stops. The drainage ditches that had kept the lowland fields workable

were silting up, and the careful pattern of agricultural production that had supported the town was breaking down into something more local and more subsistence-oriented. By the 6th century, the lowlands of what is now Montmouthshire had effectively been rewilded across substantial areas, with forest cover returning to ground that had been farmed continuously for centuries. What had happened to the people who had once lived in this landscape is one of the more difficult questions to answer. Some of them certainly had died in the various epidemics,

famines and conflicts that mark the period. Others had moved, either to other parts of the British world that were holding up better, or, in some cases, across the channel to Brittany, where the migration of Britons to the continent was continuing through the 5th and 6th centuries. Some had been absorbed into the new political and cultural arrangements of the post-Roman West, becoming subjects of the emerging British kingdoms, or, in the case of the eastern frontier areas, intermarrying with the incoming Saxon populations.

the overall effect was a substantial reduction in the population of the lowland zones, although the exact scale is impossible to estimate with any confidence given the fragmentary nature of the evidence. One of the more catastrophic factors in this demographic collapse may have been the great plague that swept across the late Roman and post-Roman world in the middle of the 6th century, which historians now generally call the Justinianic plague, after the Eastern Roman Emperor under whose reign it first appeared. The disease, which appears from genetic evidence to have been the same Yersinia pestis that would later cause the medieval black death, arrived in the Mediterranean from somewhere in Central Asia

in the early 540s and spread rapidly across the entire Roman world. Whether it reached Britain is one of the more debated questions in late antique studies, but the balance of evidence suggests that it did, probably more than once, with the Western British coasts particularly exposed because of their continuing trade links with the Mediterranean. The settlements at Tintagil and other Western British ports show evidence of disruption in exactly the period when the plague would have been arriving, and the post-Roman population, which was already under pressure from various other factors, would have been particularly vulnerable to a new infectious disease

against which it had no inherited immunity. The combination of Saxon expansion in the east, plague along the western coasts and continuing economic contraction across the country, may have produced a population collapse in the second half of the sixth century, that was far more severe than anything the moralizing chroniclers ever quite captured. Gildes himself, writing before the plague had reached its full intensity, may have been describing only the early stages of a much larger demographic catastrophe that he did not live to see the worst of.

The afterlife of Gildes' text in the wider European tradition is also worth a brief note, since it shows how a book written in obscure circumstances at the western edge of the Christian world could end up shaping the historical imagination of half a continent. The treatise was copied and recopied in monastic scriptoria across Britain and the continent throughout the early medieval period, with surviving manuscripts ranging from the 9th to the 15th centuries. Continental writers used Gildas as a quarry for information about the western edge of the post-Roman world,

and his account of the British appeal to Saxon mercenaries became one of the standard examples cited in medieval discussions of how civilisations destroy themselves through misjudgment. The Carolingian scholars of the late 8th and 9th centuries, working at the courts of Charlemagne and his successors to assemble a comprehensive history of the Christian world, included Gildes in their reading lists and copied his text into the great anthologies that they produced

for the libraries of the major cathedrals and monasteries. From there it passed into the standard medieval curriculum, where it remained as a more or less obscure but persistent presence until the rediscovery of a late antique studies in the 16th and 17th centuries gave it a new prominence. The figure of Gildes himself acquired, over the centuries, a kind of legendary biography that bore only a loose connection to the historical evidence. Various traditions made him a son of the King of the Strathclyde Britons in what is now southwestern Scotland, gave him a series of monastic foundations in different parts of Britain and Brittany, attached him to the great Welsh St. Iltud and to the influential Breton Monastery of Ruiz,

and provided him with a death scene of suitably edifying piety. Whether any of these later traditions reflect genuine memory or whether they are pious inventions of subsequent centuries is impossible to determine. The historical guilders, as opposed to the legendary guilders, was almost certainly a real monk who wrote a real book, but most of the additional details added by later writers should be treated with caution. What is clear is that he became a significant figure in the religious imagination of both the Welsh and the Bretons,

with churches and monasteries claiming connection to him and his name being invoked in genealogies and dedicatory inscriptions for centuries after his death. The handful of families who continued to live within the walls of care during the 6th century occupied a curious position. They were, in a literal sense, the last Romans of southeastern Wales, the surviving remnant of what had once been a Roman urban population.

They were also, by the standards of the wider region, almost invisible. They had no political identity that mattered to the larger world. They had no economic significance beyond their immediate hinterland. They had no surviving institutions of the older romankind. What they had was a continuing presence on the same patch of ground that their ancestors had occupied for centuries, with the great walls of the town still rising around them as a kind of silent monument to what had been lost. They lived among ruins, and they probably did not think of themselves as living among ruins, since this was the only world they had ever known. To them, the broken arches of the bathhouse and the overgrown forum were not symbols of

decline, but simply part of the landscape, the kind of features that had always been there, the sense that something had ended, that a civilization had passed and a darker age had begun, was the sense of the historian and the chronicler. Not the sense of the people who actually lived in the post-Roman lowlands. They were not mourning a lost world. They were just getting on with the next one, whatever. ever that turned out to be. The picture of the post-Roman British world that emerges from the moralising fury of the chroniclers and the slow archaeological contraction of the old towns is not

by itself a complete one. While the great urban centres of the Roman period were quietly slipping into ruin and a small number of literate clergymen were composing furious sermons about the moral failings of their contemporaries, an entirely new political world was being constructed in the same landscape on quite different foundations. The kingdoms of early medieval Wales did not emerge as a continuation of the Roman administrative system, with its civitates and its appointed magistrates and its careful gradations of imperial rank. They emerged as something genuinely new, a network of small chiefdoms based on hilltop strongholds, kin-based political loyalties and a flexible mix of inherited Roman elements with native British, Irish and Christian innovations. Where you went looking for this

new world depended on what you wanted to find. If you wanted to find its political centres, you went to the hilltops. If you wanted to find its written record, you went to the standing stones along the old roads and at the entrances of the new burial grounds. If you wanted to understand how it actually worked, you needed to look at both at once. One of the most revealing examples of the new style of post-Roman political centre is a small fortified site on a hill above the estuary of the River Ely, just south of the modern Welsh capital of Cardiff, called Dynus Powys. The site itself is, by the standards of most major archaeological fortifications,

embarrassingly small. The defended enclosure measures roughly 30 metres across and 60 metres long, an area smaller than many supermarket car parks, and considerably smaller than even the most modest Roman fort. It sits on top of a steep limestone outcrop, with natural cliffs falling away on three sides and a narrow approach from the fourth that was protected in the early medieval period

by a series of ditches and ramparts cut across the access route. Anyone visiting the site today walking up through the modern oak woodland that has grown back over the slopes would be forgiven for assuming that the place could not possibly have been politically important. It looks like a refuge, not a capital. It looks like somewhere a frightened community

might hide for a few days during a raid, not somewhere a ruler might hold court. The man who established that this assumption was completely wrong was the archaeologist Leslie Alcock, who excavated Dynas Pousas across a series of seasons in the late 1950s and early 1960s. Alcock was working with limited resources and with relatively basic technology by modern standards,

but he applied a careful and patient method of stratigraphic excavation that produced one of the most detailed pictures of any post-Roman British site to that date. What he found astonished both himself and the wider archaeological community. The tiny fortification on the hilltop, far from being an obscure refuge, had been the residence of a wealthy and sophisticated elite household, with material remains that pointed to extensive trading connections, considerable conspicuous consumption,

and a settled occupation lasting at least two or three generations. The most striking finds were the imported ceramics. Mediterranean amphorae, the heavy ceramic vessels used to ship wine and olive oil from the eastern Mediterranean, turned up in substantial quantities, with pieces from at least several dozen different vessels recovered from the various deposits across the site. These were not casual finds. They represented a regular, sustained import of Mediterranean liquid commodities into a remote hill fort on the Welsh coast in the late 5th and 6th centuries, a period when, by the conventional model of post-Roman collapse, this kind of long-distance trade should have ceased entirely.

Alongside the amphorae, the excavation produced pieces of fine African red slipware tableware, the kind of luxury dining ceramics that wealthy Roman households had used in the late imperial period, and that continued to be produced in North Africa for export to the Western Mediterranean and beyond. Coptic glass vessels manufactured in Egypt also appeared in the deposits, and beads of similar Egyptian origin were recovered in some quantity. The picture was unmistakable. Whoever was living at Dinus Powis was importing luxury goods from across the Mediterranean basin, paying for them with whatever the British West had to offer in return and consuming them in a way that suggested both wealth and a deliberate cultivation of imperial Roman style.

The metalworking evidence was equally striking. Substantial quantities of bronze and silver scrap, crucibles for smelting, moulds for casting and slag from the various stages of, the production process, all suggested that the site had been a centre for the manufacture of high-status jewellery and personal ornaments. Some of the finished items recovered from the excavations were themselves remarkable. Gold ornaments of local manufacture, designed in styles that combined Roman Celtic, an emerging Christian iconography, showed that the metalworkers at Dynus Powys were producing objects of considerable technical and artistic accomplishment. Glass beads, some of them apparently locally made and others almost certainly imported from continental sources,

formed strings and pendants that would have been worn by the female members of the household, and possibly by guests at formal occasions. The whole picture was one of a small but rich aristocratic establishment, supported by craftsmen of high skill, and connected to an international trading network that brought in the kind of goods that ordinary post-Roman British communities never saw. What was the political identity of the people living at this site? The honest answer is that we do not know with certainty, but the most plausible candidate is that Dynas Powis was the residence of an early ruler of one of the small kingdoms that emerged in southeastern Wales during the post-Roman period.

The kingdom of Glywis Singh, which would later combine with Gwent to form the larger kingdom of Morgendurg, occupied roughly the territory in which Dynas Powers sits, and the dating of the major occupation phase at the hillfoot fits comfortably within the period when Glywis Singh would have been emerging as a distinct political unit. If this identification is correct, and most modern scholars treat it as at least a working hypothesis, then Dina's Powis gives us a remarkable view of what the seat of a small, early medieval British kingdom actually looked like in operation. It was not a town, it was not a palace complex, it was a fortified hilltop with a few timber buildings inside the defences, occupied by a ruling household and its immediate retainers, supplied by tribute from the surrounding countryside, and connected to the wider Mediterranean world through a network of seaborne trade that brought in luxuries far out of proportion to the modest

physical scale of the place itself. The animal bones recovered from the excavation tell their own quiet story about the diet and the economy of the household. Cattle bones predominate, as they generally do at high-status sites of the period, with a pattern of butchery suggesting that prime cuts of beef were consumed at the main residence, while less prestigious portions were processed elsewhere. Sheep and pigs are present in smaller but significant numbers, contributing to a diet that would have been varied by the standards of the period. Deer bones suggest hunting, which was both a practical food source and a marker of aristocratic leisure. Fish bones, including marine species that would have been brought up from the nearby coast,

indicate that the household was drawing on resources from across its territory rather than relying on whatever could be produced within walking distance. The overall picture is of a well-fed elite, eating in ways that reflected both their material wealth and their conscious cultivation of an aristocratic lifestyle that distinguished them from the surrounding peasant population. The contrast with the diet of an ordinary post-Roman British farming household, which would have been considerably more monotonous and considerably more dependent on grain,

is one of the more concrete reminders that the new elites were quite distinct from the wider population they ruled. There is also evidence in the form of post-holes and beam slots cut into the bedrock of timber buildings inside the defended enclosure, These were not large structures by Roman standards, but they were built with care, and they reflect a settled long-term occupation rather than a temporary refuge. The main hall would have been a substantial timber building, probably with a central hearth, walls lined with hangings of woven cloth, and a thatched or shingled roof. Smaller buildings around the hall would have housed the domestic functions of the household, including kitchens, storerooms and accommodation for the retainers and servants. The whole arrangement, while modest by the standards of a Roman villa,

would have been impressive enough to mark out the residence as the seat of a person of importance, and the entire complex was surrounded by the substantial earthworks and revetments that gave the site its defensive character. Visitors approaching the hill from the surrounding countryside would have seen the smoke rising from the household fires, heard the sounds of metalworking and domestic activity from within the walls, and known immediately that this was a place where political authorities, was concentrated, and where careful behaviour was expected of anyone seeking access. This pattern of small fortified hilltop residences with surprising amounts of imported luxury material

turns out not to be unique to Dinah's Powys. Similar sites have been excavated across Western Britain, particularly in Cornwall and South West Wales, where the trading connections with the Mediterranean were apparently most intense, the site of Tintagull on the Cornish coast, which has its own complicated relationship with the legend of King Arthur, produced enormous quantities of Mediterranean amphoree and fine tableware from a period of major occupation in the 5th and 6th centuries. The hill fort of South Cadbury in Somerset sometimes identified speculatively with Camelot also showed evidence of high status post-Roman occupation with imported goods. Various smaller sites along the western coasts have produced similar but smaller assemblages.

The picture that emerges from putting all these sites together is of a network of small Western British political centres. them maintaining trading relationships with the Mediterranean through the Western seaways, all of them supporting elite households that consumed Mediterranean luxuries as part of their political identity, and all of them representing a kind of post-Roman aristocracy, whose existence the moralising chroniclers had no particular interest in describing. The trade itself is one of the more interesting puzzles of the period. How did Mediterranean amphora filled with wine and oil reach the Welsh and Cornish coasts in the 5th and 6th centuries? When the formal Roman shes,

shipping networks had collapsed, and when the conventional wisdom held that all such long-distance commerce had ended. The answer, as far as it can be reconstructed, involves a relatively small number of long-distance voyages by ships that were probably operating out of the eastern Mediterranean, possibly through ports in Byzantine territory, with the cargo broken up into smaller consignments that were then distributed along the Western European coast by intermediary traders. The economics of this trade are not obvious. The British West could not produce gold or silver, or wine in any significant quantity, and yet Mediterranean traders were apparently willing to make the long voyage in significant numbers. What the British West could provide, almost certainly,

was tin, which had been one of the great Cornish exports since the Bronze Age, and which was essential to the production of bronze. There may also have been trade in slaves, in hunting dogs, which were highly prized across the ancient and early medieval Mediterranean for their quality, in fine woolen cloth, and possibly in other commodities that have not left obvious archaeological traces. Whatever the exact mix, the trade was substantial enough to support a network of import sites along the western coasts for somewhere between one and two centuries, after which it tapered off and eventually ceased, leaving the Western British kingdoms to their more local economic relationships. While the wealth was being concentrated in fortified hilltops, the political identity

of the new aristocracy, was being announced in a much more public way through a remarkable explosion of inscribed standing stones across Western Britain. These stones, which are now generally called the early medieval inscribed stones, or, more colloquially, the Ogham stones because so many of them carry inscriptions in the Irish Ogham script alongside Latin texts, mark one of the more distinctive cultural phenomena of the post-Roman West. They were erected, in their hundreds, between roughly the late 4th century and the 7th century, and they cluster particularly densely in southwest Wales, Cornwall and Southwest Ireland,

with smaller numbers turning up in Northern Wales, the Isle of Man, South Western Scotland. And as far afield as Devon, each stone typically carries a short inscription identifying the person it commemorates, often with a patronymic giving the name of the person's father, sometimes with a brief title or epithet, and occasionally with a religious formula that anchors the commemoration to Christian belief. The script in which these inscriptions are written is one of their more distinctive features. Some of them are in Latin only, written in the formal capital letters that had been the standard script of Roman imperial inscriptions. Some of them are in Ogham only, the early Irish writing system that uses a series of short notches and lines, arranged along the corner edge of a stone to represent the sounds of an early form of the Irish language.

And, most interestingly, a substantial number of them are bilingual, with the same name and type. given in both Latin and Ogham, which indicates that the communities erecting these stones were operating in a culturally complex environment, where both languages mattered and both writing systems were taken seriously. The bilingual stones are particularly important because they provide one of the few sources of evidence for the actual sound system of early medieval Welsh and Irish, since the Latin and Ogham versions can be compared to reconstruct how the names were pronounced in each language. A few specific examples give a sense of what these monuments actually look like in practice. The Votporek stone found at Castel Dwyrn in Kamarvanshire

commemorates a man whose name appears in both Latin and Occam forms and who has given the title Protector, a Roman military rank that had been preserved into the post-Roman period and was apparently still meaningful as a marker of elite status. The same name in the form Vortipur appears in Gildas's catalogue of bad kings of his own day, and the stone may well commemorate either that ruler himself or a close relative of his, which would make it one of the few surviving physical traces of the men whose moral failings the angry monk catalogued at such length. The inscription itself is brief, just a name and a title, but the implications are considerable. It tells us that the rulers guildes was attacking really existed,

that they used both Latin and Irish in their official self-presentation, and that they considered Roman military titles worth advertising on their own grave markers more than a century after the formal Roman administration had ended. The Elliseg pillar in North East Wales, although slightly later in date, carries a much longer inscription that traces the descent of a 9th century King of Powis back through Maximus and a series of other figures to a deep British past, demonstrating how durable this style of commemoration would prove across the early medieval period. And dozens of smaller stones scattered across the western landscape from the tips of southwest Wales

to the upland passes of North Wales, carry their brief inscriptions in similar combinations of script and language, each of them marking out the political and cultural territory of the new elites in a way that no Roman administrator would have entirely recognised, but that any Roman administrator would have understood the principle behind. The stones themselves were placed with care and with deliberate intent. They were set up at burial sites, at the boundaries of estates, at crossroads on the major travel routes,

and in proximity to the new churches that were. were being founded across the country. Their placement was meant to be public. Anyone walking the old Roman roads in 5th or 6th century Western Britain would have passed these stones regularly, their inscriptions visible to anyone who could read them, and their physical presence visible to anyone who could not. They functioned as a kind of distributed political announcement, telling the whole landscape who was important, who their fathers had been, and what claims they were making on the territory around them. The whole tradition was, in its quiet, way, a transformation of the Roman habit of monumental inscription into something that fit the new

fragmented political environment, where authority was no longer concentrated in centralised institutions, but had to be asserted repeatedly and visibly by individual families and lineages across the surviving landscape. The dense cluster of Ogham stones in southwest Wales, which we met earlier on the Pembrokeshire coast, points to the same picture. A kingdom of Diefd founded by Irish migrants, ruled by a dynasty tracing itself to a founder called Eokade from across the sea. The Ogham stones are the most visible archaeological trace of this population, marking out the burial places and territorial claims of the Irish-speaking elites, who had taken up residence on the Welsh coast.

The Latin stones, meanwhile, show a continuing tradition of Roman literacy and Roman commemorative practice in the Western British world, long after the formal imperial administration had ended. These are not, however, classical Roman inscriptions in the usual sense. The Latin is often non-standard, with various irregularities of spelling and grammar that reflect the spoken Latin of post-Roman Britain, rather than the polished prose of imperial chanceries. The names commemorated are often Celtic rather than Roman, given in Latinised forms that fit awkwardly into the standard Roman inscriptions formulas. The titles, where they are given, sometimes use Roman terms like Presbyter or Sassados for clergymen,

or magistratus for civil officials, but other times used distinctly post-Roman terms that have no precise Roman parallel. The whole tradition shows a society that had inherited the Roman habit of formal stone inscription, and had adapted it to its own purposes, modifying the conventions to fit a new political and cultural environment without abandoning the underlying practice. The geographical distribution of the inscribed stones gives some sense of the political shape of the post-Roman West. The densest concentrations sit along the old Roman roads and around the surviving administrative centres, suggesting that the literate culture which produced these inscriptions was particularly strong in areas that had been most thoroughly Romanised during the period of formal occupation.

Outside these zones, the practice tapers off, with much smaller numbers of stones in the more remote upland areas, and in the regions where Saxon expansion was beginning to disrupt the older British political arrangements. The pattern is consistent, with what we might expect, namely that the new post-Roman elites were strongest where the inherited Roman framework had been most fully established, and that the cultural innovations of the period built on rather than replace the late Roman foundations. The eastern frontier of the post-Roman British world deserves a brief mention here, because the political developments in Western Britain were taking place in conscious awareness of what was happening to the East. The Anglo-Saxon

expansion, which had begun in the 5th century with the small initial settlements that Gilders had condemned, and which by the late 6th century had absorbed most of what is now eastern and central England, exerted continuous pressure on the British kingdoms throughout the period of their formation. The frontier between British and Saxon territory was not a fixed line but a moving zone of contact, conflict and gradual cultural absorption, with British communities being progressively assimilated, displaced or pushed back westward across the generations. The kingdoms of Wales were taking shape in a sense, as the surviving British political tradition retreated into the Western Highlands and consolidated itself there in defensible form. The construction of fortified

hilltop residences like Dinus Powys, the renewed importance of upland strongholds, and the cultivation of explicitly British political identities through the genealogies and the inscribed stones can all be understood, at least in part, as responses to the eastern pressure. The Welsh kingdoms were not just a continuation of older British political arrangement. They were a deliberate reorganisation of those arrangements into a form that could survive the new geopolitical realities of the post-Roman West. The kingdoms that eventually emerged from this complex political landscape would last, in various forms, for nearly a thousand years. Gwynedd in the north, Powis in the centre, Diffed and its successor de Hobarth in the south-west, Glewising and Gwent in the south-east, with smaller kingdoms appearing and disappearing around their edges,

would form the basic political map of medieval Wales, until the conquest by Edward I in the late 13th century, and even, in some attenuated form beyond it. The dynasties that ruled these kingdoms traced their lines back through genealogies that combined real historical figures, semi-legendary founders, and occasional mythological ancestors

into elaborate constructions of political legitimacy, with figures like Magnus Maximus appearing as imperial Roman ancestors, Kuhneda, as a war leader, who had come south to expel Irish settlers from North Wales, and various biblical and Trojan figures providing the deepest roots of the family trees. The early medieval Welsh aristocracy was, in effect, constructing its political identity by combining whatever materials were available, Roman, British, Irish, biblical and classical, into a synthetic genealogy that could be deployed for whatever political purposes the moment required. What is particularly striking, looking back at this whole period, is how much of it was being negotiated rather than dictated.

There is no single moment when Roman Britain ended and medieval Wales began. There is instead a long process of adjustment, lasting through the 5th and 6th centuries and into the 7th, in which the inhabitants of the Western British provinces gradually constructed new political arrangements out of the materials available to them. Some of these materials were directly Roman, like the inherited literate culture. culture, the inherited Christian framework, the inherited road system, and the inherited concept of formal inscriptional commemoration. Other materials were native British, like the kin-based political loyalties, the Bardic poetic traditions, and the older practices of tribute and clientage that had operated in pre-Roman times, and continued to operate beneath

the surface of the Imperial Administration. Still other materials were imported, like the Irish settlement and its associated language and writing system, like the Mediterranean trade goods, and like the various contacts with the Christian Church in Gaul, Spain, and the eastern Mediterranean. The early medieval Welsh world that emerged from this long process of negotiation was something genuinely new, neither Roman nor pre-Roman Celtic, nor purely Christian, nor purely tribal, but a complex hybrid that drew on all these sources and synthesized them into a distinctive culture. This culture would prove remarkably durable. The political arrangements that took shape in the 6th and 7th centuries provided the basic

framework within which Welsh society operated for the next 800 years. The literary traditions that emerged from this period, including the early Welsh poetry that have survived in fragments and the prose tales that we have already encountered through the Mabnogion, would continue to be copied, recited and built upon throughout the medieval period. The religious institutions that the early medieval Welsh founded, the network of mother churches called Klaus and the various monastic communities scattered across the landscape would persist through the Norman conquest and the subsequent reorganizations

of the Welsh Church. The genealogies, with all their elaborate constructions of Roman and biblical descent, would be invoked by Welsh princes for centuries as evidence of their legitimacy. Even the language itself, the Welsh that emerged from the late-spoken Latin and the older Britonic of the post-Roman West, would continue to be spoken across most of Wales until the great industrial transformations of the 19th and 20th centuries, and it remains spoken today with several hundred thousand speakers maintaining a literary and cultural tradition that runs unbroken from these distant origins.

Looking back across the entire arc of the story, from the petrified forest at White Sands and the trading port at Gate Home through the legends of Brutus and Maxon, the fall of Roman authority and the moral fury of Gildas, the contraction of Kerwent and the rise of Dynas Powis, What emerges is not a story of decline or of disaster, but a story of transformation. The Roman world ended in Britain, certainly. The cities contracted, the trade routes broke, the imperial administration dissolved,

and a way of life that had operated continuously for over three centuries came to an end. But the people who had lived in that world did not disappear. They adjusted, they improvised, they found new ways of organizing their lives, and they built something new on the foundations of what had been. The kingdoms that emerged from this long transformation were smaller than the Roman province had been, less wealthy, less cosmopolitan in some ways, but they were also more responsive to local conditions, more deeply rooted in the immediate landscape, and more capable of sustaining themselves through the long centuries of the medieval period, without requiring the kind of imperial superstructure

that had supported the Roman occupation. There is, in this whole story, a kind of quiet lesson about how cultures actually survived periods of catastrophic change. They survive not by preserving everything from the past in unchanged form, which is impossible, and not by abandoning everything from the past in favour of complete novelty, which is unwise, but by selecting carefully from the inherited materials, combining them with new elements as circumstances require, and constructing arrangements that fit the new conditions while honouring the old ones. The medieval Welsh did this with considerable skill, and the result was a culture that retained genuine continuity with the Roman past,

while becoming something quite different from it. The standing stones along the old roads, with their bilingual inscriptions in Latin and Ogham, are perhaps the most concise visual summary of this whole process. A Roman literate tradition combined with an Irish writing system, both deployed in the service of new British identities, on stones that mark the burial places of figures whose political world was neither fully Roman nor fully tribal, but something genuinely knew. That is the early medieval West in miniature, and it is a more impressive achievement than the moralising chroniclers ever quite acknowledged. The forgotten origins of medieval Britain, in the end, turn out not to be forgotten at all. They are inscribed on stones, buried in hill forts,

scattered across windswept beaches, and preserved in the careful prose of tales that describe dreaming emperors and giants thrown from cliffs. They are present in the names of saints and the roots of old roads, in the genealogies of kings and the foundations of churches, in everything from the petrified roots of vanished forests to the towering walls of empty Roman towns. The story of how this world came into being is not a story of darkness, despite the convenient label that historians attached to it for so many centuries. It is a story of patient improvisation, of cultures adapting to circumstances that none of them had chosen, of a long, quiet conversation between a fading imperial past and an emerging medieval future.

The world that resulted from that conversation would shape Britain for another thousand years, and its echoes, in language and in landscape, are still audible today for anyone who knows where to listen. Thank you for joining me on this long walk through the deep origins of medieval Britain. May your night be peaceful and may your dreams be sweet.