The soldiers at Banna — the fort the Romans called Birdoswald, on the western reach of Hadrian's Wall — did not receive an official order to leave. There was no formal ceremony, no lowering of standards, no dispatch rider arriving with instructions from Rome. What happened was slower and stranger than that. Around 408 AD, the coin supply dried up. Wages stopped. The bureaucratic machinery that had kept three hundred miles of northern frontier supplied, staffed, and operational for nearly three hundred years simply ceased to function, and the men stationed along the Wall were left to work out what that meant for themselves.
The western Roman Empire was collapsing in stages, and Britain had been becoming peripheral for decades. Magnus Maximus had stripped the island of troops in 383 AD when he marched on Rome to claim the purple. Constantine III did the same in 407. What remained in Britain by 408 was a depleted, underpaid, increasingly local force — soldiers who had children in the nearby vicus settlements, who had put down roots, who were as British as they were Roman. When Emperor Honorius sent his famous rescript in 410 telling British cities to look to their own defence, he was confirming a reality that had already arrived on the ground. The legions were not so much withdrawn as they simply dissolved.
What happened next is one of the great historical puzzles of the British Isles. The period immediately after 410 — roughly 400 to 450 AD — is sometimes called Sub-Roman Britain, and the archaeological record for it is extraordinarily thin. Not thin in the sense of gaps, but thin in the sense of a society that stopped generating the kind of material evidence archaeologists can read. The coin evidence stops almost completely. Wheel-thrown pottery, a Roman staple, disappears. The sophisticated road maintenance system breaks down. Towns that had been functioning urban centres for three hundred years show signs of radical contraction.
At Wroxeter, the second largest city in Roman Britain, excavations in the 1980s led by Philip Barker revealed something astonishing. Where Roman-style masonry buildings had stood, there emerged evidence of large timber halls built on the same footprint in the early fifth century. The city didn't simply die. Someone — almost certainly a local leader of considerable authority — had deliberately maintained the urban centre and rebuilt it in a different vernacular. Wroxeter's sub-Roman phase may have lasted into the sixth century. It is a ghost of Rome, recognisably shaped by Roman urbanism but no longer Roman in any meaningful sense.
The Wall itself tells a different story. The fort at Birdoswald, excavated by Tony Wilmott in the 1990s, produced evidence of a granary being converted into a large timber hall in the fifth century. The interpretation is compelling: someone with power, someone who controlled a garrison community and needed a feasting hall, had taken over the fort after the Roman military apparatus evaporated. This is not barbarian collapse. It is local continuity — a British chieftain or a senior soldier's descendant running a diminished but functioning estate from inside the shell of a Roman fort. The Wall, in other words, did not go dark. It changed hands.
Further east, at Vindolanda near Bardon Mill in Northumberland, the picture is different again. The famous writing tablets that had captured so much of daily Roman life — letters about birthday parties, requests for more beer, a child's homework — stop in the third century. The site's sequence in the fifth and sixth centuries is poorly understood, though more recent excavations have continued to refine it. What seems clear is that the transition was not uniform across the Wall zone. Different forts experienced different fates depending on the local communities that surrounded them and the individuals who took control.
In the towns of the south, the breakdown was also uneven. London — Londinium — contracts sharply after 410. The forum, once the commercial and administrative heart of the city, falls out of use. The basilica, one of the largest buildings in Roman Britain, is stripped for material by the early fifth century. But people remain. Cemeteries from the mid-fifth century suggest a residual population maintaining some kind of community life in the old Roman city, now reverting to farmland and ruins in its former civic core. The mosaic floors crumble. The underfloor heating systems fill with silt. The baths cool.
What fills the political vacuum is uncertain, but the literary sources — fragmentary and often written decades or centuries after the events they describe — give us names and outlines. Gildas, a British monk writing around 540 AD in his polemical text De Excidio et Conquestu Britanniae, describes a period of civil war, plague, and Saxon raids that follows the Roman withdrawal. He is furious with the British leaders of his day, and his account is full of biblical rage rather than historical precision, but the broad strokes — factional conflict, vulnerability to external attack, a society unable to coordinate defence — seem plausible.
The Saxons are the great complication. Germanic settlers and raiders had been present in Britain during the late Roman period, often serving as federate soldiers. After 410, that relationship changes. The eastern and southern coastal areas of Britain experience increasing Saxon settlement through the fifth century, and by 450 the balance of power in much of what would become England has shifted fundamentally. Vortigern — a British leader mentioned by Gildas and later by Bede — allegedly invites Saxon mercenaries to defend the island against Pictish raids. According to the tradition, those mercenaries mutiny, and the Saxon takeover begins in earnest.
What is striking, when you stand at the Wall today and look north toward the Whin Sill's escarpment dropping away into Northumberland, is how completely the landscape absorbed what came before. There is almost nothing at human eye level that announces itself as Roman. The Wall runs, yes — its stones rebuilt many times, consolidated in the Victorian era by John Clayton, who bought the farmland and excavated sections almost as a private hobby. But the Roman world that the Wall once enclosed, the empire that built roads from Deva Victrix in Chester to Eboracum in York to the gates of Banna and Segedunum at Wallsend — that world has to be imagined into existence.
The archaeology rewards that imagination richly. The Vindolanda tablets, now in the British Museum, are among the most intimate documents to survive from Roman Britain. Claudia Severa's invitation to her friend Sulpicia Lepidina to attend her birthday party — written around 100 AD — is the oldest surviving example of writing in Latin by a woman's hand. These are real people, and their world was warm and specific, a frontier life full of mundane concerns. The contrast with the fifth century silence is stark. One era generates thousands of words on wooden leaves. The next generates almost nothing.
LocationsLive's Hadrian's Wall walking experience crosses ground where that transition happened — where Roman soldiers became something else, where the empire's edge became its own world. Standing at Birdoswald, looking at the outline of the granary that became a chieftain's hall, you are standing at the exact pivot point between two eras. It is not a ruin. It is a hinge.
There is a persistent myth that the Romans left and the lights went out. The reality is more interesting and more melancholy. The lights did not go out — they changed colour. The Roman system of government, law, taxation, and military organisation dissolved over decades rather than overnight. What replaced it was not chaos but a patchwork of smaller authorities, some descended from Roman administrators, some military men who stayed when the money stopped, some genuinely new arrivals reshaping a changed landscape. Britain in 450 was still recognisably related to Britain in 400. By 500, the relationship had grown much harder to trace.
For related reading, the post on life in Roman York at its peak fills in what urban Roman Britain actually looked like before this dissolution, and the post on the women of Hadrian's Wall recovers the human texture of the frontier community.
The Wall is still there. On clear mornings, when the light is low and long across the ridge and the sheep are scattered across the pasture below the crags, it looks permanent. Ancient, unargued, inevitable. But permanence was exactly what the Romans believed they had built, and they were wrong. Empires are not eternal. Their roads last longer than their armies. Their stones outlast their emperors. The morning light on the Wall illuminates something that was always already gone — and something that never quite left.
