Soil forensics as a tool to test reported artefact find sites

The reported find sites for archaeological artefacts such as coin hoards, can in some cases be either accidently mistaken or potentially deliberately fabricated. However, testing the veracity of such reported find sites can be difficult. Advances in the analysis of soil samples for both criminal and environmental forensic investigations, is allowing the characterisation of very small soil samples to be achieved. In this study forty three soil samples were analysed from six groups of coins, each of which had been reported as an individual coin hoard collected at different locations in Devon and Somerset, UK. In-situ soils were removed from the surface of the coins and mineralogically analysed using automated scanning electron microscopy and energy dispersive analysis. The mineralogical data show that five of these six coin groups could not have been derived from individual find sites. The mineralogical data for one of the groups was indicative that the coins making up that group could potentially have been derived from a single location. Subsequent and independent to the mineralogical assessment of the coins, a numismatic inspection of the coins led to the same conclusions. Automated mineral analysis, which can be carried out on very small soil samples, may prove to be a useful technique for the assessment of the reported provenance of archaeological artefacts.