New field and geochemical evidence from vitrified forts in South Morar and Moidart,NW Scotland: further insight into melting and the process of vitrification

New field and laboratory work has been carried out on two vitrified forts, Rhubh Aird Ghamhsgail near Arisaig and The Torr near Acharacle. Fieldwork confirms that both are constructed from rubble largely comprising local psammitic Moine Supergroup rocks and that their walls are relatively well preserved with large portions of vitrified material remain, particularly at The Torr, where remnants of up to four courses of stone facing occur. The preserved wall fills comprise mostly psammitic rubble set in a dark brown, vesicular, glassy to aphanitic matrix that in many places preserves evidence of downward flow. Conventionally, vitrification is explained through combustion of timber interlacing within the walls leading to melting that has been modelled in the granite system. However, considering Iron Age smelting technology, this leads to problems in obtaining high temperatures (≥1000 °C) over tens of metres of wall length. Geochemical analysis of the vitrification indicates that modelling the melting within the granite system is incapable of providing a partial melt matching the composition of vitrification. At The Torr, pelite fragments have been found near the base of the wall where partial melt occurs as black bubbles. In these fragments it appears that micas (largely biotite) have formed the melt via the reaction biotite + quartz = sanidine + orthopyroxene + liquid which takes place at temperatures c. 850 °C. Quench crystals in the melt include orthopyroxene, plagioclase, spinels, ilmenite, magnetite, anatase and apatite confirming that it was not a granite melt. These melted fragments are found next to material with biotite that, superficially, appears unmelted, suggesting that some of the heating took place perhaps away from the edifice and unmelted residue was utilised as part of the rubble fill, suggesting that in situ partial melting of the rubble is unlikely. However, in other parts of the edifice heat affected, yet unmelted, fragments are dominant. The new evidence indicates that vitrification occurred at lower temperatures than previously modelled and thus the melts could have been achieved more easily than previously thought.