What makes the site genuinely odd is not its age but its ambiguity: no tools, no finished objects, no discarded metal came out of the ground. Only heat and residue remained to suggest what had once been done there.

The site, designated Rossan 4, came to light during pre-motorway archaeological testing carried out by Ian Russell in 2002. When a wider excavation followed, archaeologists uncovered two subrectangular hearths set roughly ten metres apart along an east-west strip about sixty-nine metres long and seven metres wide. The smaller of the two hearths, measuring around 1.4 by 1.12 metres, had slag associated with it, the glassy waste material left over from smelting or forging metal, and a post-hole nearby suggested some kind of simple structure once stood close to it. The larger hearth, at roughly 2.08 by 1.3 metres and somewhat deeper, lay to the west. Further west again, about twenty metres on, excavators found a charcoal-filled pit and, between the two hearths, a substantial area of oxidised, heat-reddened clay covering roughly eight by six metres. Radiocarbon dating of charcoal from the smaller hearth with its slag returned a date of Cal AD 1030 to 1280, placing the metalworking activity somewhere in the high medieval period. The charcoal pit told a more complicated story: its samples produced dates spanning Cal AD 1460 to 1680, Cal AD 1770 to 1800, and even Cal AD 1940 to 1950, a spread that suggests that particular feature is probably modern and unrelated to the medieval activity nearby. The work was reported by Murphy in 2003.