The vitrified rubble upper core of the rampart is visible in this trench, which is 2m wide. Three voids can be seen, spaced roughly 0.5m apart. These represent the locations of horizontal timber beams that once lay within the rampart wall. (IMAGE: FCS)

It remains one of the biggest archaeological mysteries: why do so many hillforts, particularly across Scotland, appear to have undergone a significant burning event that caused their stone walls to melt and ultimately fuse (see CA 133)? Was it done deliberately, either during an attack or as a ceremonial act, or was it accidental? One of these ‘vitrified’ forts, Dun Deardail in Glen Nevis, was recently excavated over the course of a three-year project funded by Forestry Commission Scotland and the Heritage Lottery Fund as part of the Nevis Landscape Partnership. This Iron Age hillfort was built in the middle of the 1st millennium BC, around 2,500 years ago, and was eventually destroyed in a catastrophic fire. Now excavations have shed light on its construction, occupation, and destruction.

Macroscopic, microscopic, and geochemical analyses of the site’s vitrified rocks were particularly enlightening. The results revealed that the timber-framed rampart was constructed of local stone, that the core of the fire probably reached temperatures between 850˚C and 1,100˚C, and that the upper parts of the walls fused in oven-like conditions, with little exposure to oxygen. As the vitrified core is only found in the upper section of the rampart, it may have been caused by the collapse and burning of a substantial timber-and- thatch superstructure built on top of the wall.

A reconstruction drawing of the night of the burning at Dun Deardail by Chris Mitchell. (IMAGE: FCS by Chris Mitchell 2016)

A strategic and comprehensive dating plan also helped pin down the timeline of the site. Radiocarbon analysis determined that Dun Deardail was constructed in the 5th century BC and that the fire probably occurred between 335 and 120 BC – this date was further refined through environmental evidence. A sediment core from a nearby peat bog showed that large amounts of charcoal were first present around 400 BC – which the team believes represents the clearing of the site. There is then a slight decrease in concentration but with enough charcoal still present that it may reflect daily life at the fort. Then, between 347 and 284 BC, a massive charcoal layer is deposited and, using Bayesian modelling, the team determined that the event occurred around 310 BC. After this point, the presence of charcoal sharply decreases, indicating that the site had been abandoned.

A popular account of the excavations, The Archaeology of Dun Deardail, sets the site in the wider historical context of hillfort investigations in the Highlands and explores both the science behind the investigation of vitrification and a range of creative responses to the archaeology. The booklet is available to download for free from https://scotland.forestry.gov.uk/images/corporate/pdf/the-archaeology-of-dun-deardail.pdf.

This article appeared in CA 340.

Leave a Reply