The trail of the historical Arthur leads not to medieval England, but to Dark Age Britain, recently cast off by the dying Roman Empire, and struggling to maintain its authority over an island beset on all sides by barbarian invaders.

Centuries have passed since the English adopted Arthur as a national hero. Their attachment to him is somewhat ironic. When first sighted in the mists of antiquity, Arthur was the mortal enemy of their ancestors. The Britain of his saga was Celtic, its people, including himself, being forebears of the Welsh. He defended it against Anglo-Saxon invaders and opposed the advance that turned most of it into Angle-land . . . England.

Medieval authors transformed Arthur from a Celtic warlord into a monarch whom English kings could claim as a predecessor. He acquired an official biography, included in History of the Kings of Britain, written by a cleric known as Geoffrey of Monmouth in the late 1130s. Wildly imaginative, this account can never be relied on for facts, yet it preserves some basic truths about Arthur. It introduces familiar details that romancers develop--his origin at Tintagel, his magic sword, his order of knighthood, his betrayal by Modred. But in Geoffrey's narrative Arthur first rises to general popularity by routing Saxon aggressors, and the northern Picts in alliance with them, who have been ravaging the country. His glory is founded on that triumph, restoring peace to a troubled Britain lately separated from Rome. And here Geoffrey knows what he is about; he grounds his story on historical bedrock. Despite all his fantasy, he shows us where to look for Arthur and in what role.

Roman Britain--comprising, roughly, England and Wales as modern geography defines them--endured for more than 300 years. But in the 4th century AD, as the Empire lost its grip, this Celtic land was increasingly harassed by barbarian raiders: Irish from the west, Picts from the unsubdued north, Saxons from the Continent. Imperial rescues had little lasting effect. About 410, with no Roman forces left on the island, the Emperor authorized the Britons to look after themselves, virtually granting them independence.

Beset by further Pictish attacks, a ruler called Vortigern tried to solve the problem by following an imperial precedent. He allowed barbarians to settle in Britain on the condition that they defended it against other barbarians, in this case, the Picts. His imported auxiliaries were known collectively as Saxons; they included Angles and others. At first the policy seemed to work, but more settlers poured in from the Continent, and Vortigern lacked the resources to support so many. Probably in the 440s, the Saxons revolted. They allied themselves with the Picts whom they were treaty-bound to keep out, and began to raid across Britain.

In the resulting period of turmoil and devastation, the native Britons abandoned many of their Roman-built towns. Then, surprisingly, a British recovery began. They managed to contain fresh incursions. After decades of to-and-fro warfare they won a decisive victory at a place called Badon Hill, sometime about the year 500. For a time the position stabilized, and the Saxons were contained within their authorized enclaves near the east coast. But over the ensuing decades the Saxons increased in relative number, and presently they were able to gain ground again. The conquest was slow, and Wales never succumbed at all. Its bards still held on to the traditions of their forebears' achievements when England was irreversibly Anglo-Saxondom.

The essential point in all this is that despite the ultimate outcome, something unique had happened during the 5th century. Throughout the rest of the crumbling western Empire, barbarians moved in and the provincial subjects didn't care. In Britain the subjects had won their independence, and did care. When barbarians encroached, an appreciable number of Britons fought back--for a while, successfully.

This is where Arthur belongs, an unparalleled leader in an unparalleled war effort. Opinions differ as to what his status actually was. Opinions also differ as to whether he existed at all. However, attempts to interpret him as a warrior or demigod of Celtic myth, wrongly made out to be historical, have failed to convince. His name is a Welsh form of the Roman name Artorius, a perfectly human one. In the next century or so, we find several Arthurs up and down Britain, all presumably named after a great original.

Can we reconstruct him? We have many items of information, or semi-information, that have gone into the making of his story, and the best of them are earlier than Geoffrey of Monmouth, not inventions of his. Continental records mention a 'King of the Britons' who may well be the man himself. Welsh poems and tales extol Arthur as a peerless fighter with a formidable band of companions. A work ascribed to Nennius, a 9th-century Welsh monk, lists twelve victories won by him as the Britons' war leader, giving him the credit for their crowning triumph at Badon. A Welsh chronicle reaffirms his Badonic role and adds a cryptic entry about a later battle at 'Camlann' where 'Arthur and Medraut fell', the first known hint of the fatal conflict with Modred. Arthurian place-names and local legends, more than 100 of them, extend in a vast sweep from Cornwall to Scotland.

Taken together, these items may look impressive, yet they need cautious handling, even apart from the fact that most were not written down until long after the time they refer to. Chronologically, they don't cohere. They spread out too far, giving Arthur a lifespan of at least 100 years, from the mid-5th century to the mid-6th. To pin him down we must decide where in the time-range to place him, and then find explanations for the dates that don't work.

There are ways of doing this. It appears, for instance, that a force called Arthur's Men may have remained active after his death. Bards not sure of their facts might have attributed exploits of Arthur's Men to Arthur himself as their patron or inspiration, prolonging his imagined life. That is a guess. Other interpretations are possible.

One of the most intriguing pieces to the puzzle is Nennius' account of Arthur's battles. His list is full enough to be interesting, and whether or not it is all authentic, it at least tells us how his people remembered him and helps answer the question, 'When?' The battles themselves are hard to identify. Nennius gives the battle sites pre-English names and several have vanished from the map. He speaks of Bassas, Guinnion, Tribruit, and Agned. No one is certain where these were. But he takes us on to safer ground with a river called Glein; with another river called Dubglas, in the region Linnuis; with the Forest of Celidon; and with the City of the Legion. These battles took Arthur to Lincolnshire, on the east side of England; to southern Scotland; and to Chester. As for Badon, at the end of the list, its whereabouts can at least be narrowed down to southern England. Geoffrey of Monmouth makes it a hill near Bath, and some historians agree. Other candidates are Badbury Rings in Dorset and Liddington Castle, near Swindon, which has a village of Badbury below it. On this occasion, says Nennius, Arthur single-handedly slew 960 of the enemy. A growth of legend has weakened his credibility. Still, legends can grow without disproving facts.

Two inferences follow. First, the list portrays Arthur as ranging over a large territory: as a national commander or paramount chief or both, not merely a local resistance leader. Second, in view of the times when enemies could have been active in the right area, the battles make good sense during an earlier rather than a later phase. They put Arthur's warfare somewhere around the 450s and 60s. That fits in with the Continental texts mentioning a 'King of the Britons', which some now believe provide the best clues to Arthur. It is also the time that Geoffrey apparently has in mind. A numerical date that he gives, 542, is incompatible with his other indications and may be a mistake for a date actually in the same early phase.

Badon remains puzzling. It seems too late to square with the rest, yet Arthur was surely associated with it somehow. If another commander had won this crucial victory, why did somebody so important drop out of the record? Maybe we should suspect Arthur's Men here, military heirs fighting under his standard after his death, and winning the battle. Bards counted the victory as his, but grew vague as to whether he was present himself, so that over the course of time the event became blurred. Something like this is documented in other cases. But again, it's only a guess.

Yet to admit guesswork is a step towards insight rather than a confession of failure. That, after all, is the way things are. Some might wish that we could replace Arthur's pseudo-biography with a real one. Would that be better? He is a strange, haunting, betwixt-and-between figure, not a figment to be dismissed, yet not a clear-cut individual like, say, Alfred the Great. If we could confront him plainly as we confront Alfred, who knows how we would react? It is right that the mists around him should not be allowed to blot him out. But perhaps it is also right that they should not disperse entirely.