The first written account comes from Hector Boece, a Scots historian writing in 1527 - more than five centuries after the supposed battle. In his Historia Gentis Scotorum, Boece tells how Sueno, king of Denmark and England, sent a naval force to Scotland under a commander named Camus, half sailing from Denmark and half from the Thames. The Scots, led by Malcolm, ambushed them. After Camus fell, Malcolm dipped his fingers in the dying man's blood and ran them along the top of a shield carried by a man named Robert Keith, creating - so the story claims - the red and gold stripes still found on the Keith clan coat of arms today. It is a wonderful piece of foundational mythology. Boece was a native of the area; he likely drew on local folklore. The earlier 14th-century chronicler John of Fordun, on whom Boece relied for much else, makes no mention of any such battle.

The bones at the Lochty Burn were real, but the interpretation was wrong. When Robert Dickson examined the burials carefully in 1878, he noticed two awkward problems. There were no weapons in the graves - unusual for warrior dead. And many of the skeletons appeared to be female. Stranger still, the burials were laid out in supine, east-west orientation, the Christian position. These were not pagan Viking warriors hastily interred where they fell. They were ordinary local people, buried in the manner of their Christian faith. The Camuston Cross, long held to mark Camus's death, turned out on closer study to be a late Pictish monument from the 10th century, with carvings that have nothing to do with Vikings at all.

The name "Camus" itself dissolves under scrutiny. It does not appear in any record outside the works that followed Boece. Boece almost certainly derived it backwards from a place name, Camuston, which earlier documents from 1425 and 1426 record as "Cambistown" - a Celtic origin, not a Scandinavian one. The Roman marching camp at nearby Kirkbuddo, once also attributed to the Danes, has likewise been reassigned to its actual builders: Roman legionaries on the march. Even the famous Aberlemno Stone with its battle scene, sometimes cited as evidence, has been dated by stylistic analysis to the mid-8th century or later, which is too early for 1010 and unrelated to it. Every prop in the Battle of Barry story turns out, on inspection, to belong to a different play.

And yet it persists. The site of the supposed battle still appears on early Ordnance Survey maps. Local tradition tells it. Books retell it. The Keith family still bears the red and gold stripes Malcolm allegedly painted with his fingers. There is a real history beneath the legend - the area was Pictish, then Scottish, and did face occasional Viking raids - and a fictional history layered over it that has become its own kind of fact. Standing on the dunes at Carnoustie now, you can hear the wind in the marram grass and the distant pulse of golfers on the Carnoustie Championship Course nearby. The bones beneath the sand belonged to people who lived and died here, going about ordinary lives, long before anyone needed to invent a battle to explain them.