NUK - logo
E-resources
Full text
Peer reviewed Open access
  • Portable XRF investigation ...
    Pearce, Nick J.G.; Bevins, Richard E.; Ixer, Rob A.

    Journal of archaeological science, reports, August 2022, 2022-08-00, Volume: 44
    Journal Article

    •Portable XRF analysis was applied to provenancing the Stonehenge bluestones.•In situ analysis was undertaken on Stone 62 at Stonehenge and on key Preseli outcrops.•The source of Stone 62 is attributed to the Garn Ddu Fach Preseli outcrop.•A robust analytical strategy was developed and implemented to take into account a range of factors possibly expected to influence the analyses. The doleritic bluestone monoliths at Stonehenge have long been known to have been sourced from the Mynydd Preseli area in west Wales, some 225 km away. On geochemical grounds, based on a range of major and trace elements determined by laboratory-based X-ray fluorescence spectrometry, they have been divided into three groups (Groups 1–3). Subsequently, rare earth element data obtained by solution nebulization ICP-MS showed Group 2 Stone 45 to have been sourced from Cerrigmarchogion with Group 2 Stone 62 showing similarities to the outcrops of Carn Ddafad-las and Garn Ddu Fach. In order to test this possible link, portable XRF (pXRF) analyses were obtained in situ from Stone 62 at Stonehenge and from key outcrops in the Preseli. To obtain reliable results from heterogeneous coarse-grained igneous rocks using pXRF to enable comparisons between orthostats and possible sources, a robust analytical strategy was developed. For outcrops this involved a series of horizontal traverses through individual outcrops comprising 10 to 15 analyses a few centimetres apart, giving 60–175 independent analyses per outcrop. This gives a large, analysed surface area per outcrop providing representative data and also has potential to show any vertical variation in source outcrops. Analyses were taken from similarly weathered surfaces on orthostats and outcrops to minimise compositional changes from surface weathering, and a minimum of 20 analyses were taken from orthostats at various locations across the stone. This approach provides a set of well-determined elements which can reliably be used for provenance studies (including K, Fe, Mn, Zn, Rb, Sr, Zr, Nb, Ba). Nickel, Ti and V are affected slightly more by weathering but still prove useful in comparisons, but Cr, used in earlier studies as a compositional discriminant with Ni, here has poor accuracy by pXRF and is not used. The pXRF analyses show that Stone 62 sits within the same ‘compositional space’ (for multiple element concentrations and ranges) as analyses from Carn Ddafad-las and Garn Ddu Fach, two outcrops of the same intrusion which differ from all other analysed outcrops. More specifically, analyses from Stone 62 and Garn Ddu Fach overlap, these forming a subset of the analyses from Carn Ddafad-las, which shows a more extensive range of those same elements. On this basis it is suggested that Garn Ddu Fach is the likely source of Stonehenge non-spotted dolerite Stone 62, a suggestion supported by indistinguishable petrographic characteristics between Stone 62 and dolerite sample PGDF24 from Garn Ddu Fach.