Heritage Cases

Like most rock art, little is known about the age of the art on the Linton and Zamenkomst panels or the composition of the paints. Both of these points are of considerable interest to visitors to the museum (Black pers. comm.). The Museum therefore approached us to enquire whether the dating techniques we have developed (Bonneau et al. 2011, 2017a, 2017b) could be used to directly date the art on these panels. Our protocols involve detailed characterisation of paint composition ahead of AMS radiocarbon dating. The panels have black paint on them, and are thus potentially datable by these means. Whilst AMS radiocarbon requires very small samples of carbon, this nevertheless translates to a paint sample of approximately 1–2 cm2 extent, depending on how thick the paint is. Whilst small, we believe that this is an unacceptably large amount of damage to do to important and publically displayed pieces of rock art. We therefore propose an alternative. We propose to collect small (about 0.5 mm2) samples of paint from images on the panels for inorganic characterisation. We will also collect similar samples as well as larger samples for radiocarbon analysis from paintings remaining in the two sites from which the panels come. These materials can then be correlated and direct dates from the sites be used to infer dates for the removed panels. Sample collection Samples will be collected following protocols we have developed in the course of previous research (Bonneau et al. 2017a:SOM). An overriding constraint in collecting samples is the need to minimize damage to paintings. Wherever possible, samples will be taken adjacent to already damaged parts of paintings to minimize the visual impact. Samples will be taken so as to maintain the visual integrity of each image by not cutting its outline. Decisions about which images to sample and where to take the samples will have to be made on-site and in discussion with the Curator. Only inorganic characterization of paints is planned. Samples for these analyses are small, approximately 0.5 mm2 in area. Samples of this size make little visual impact on the paintings. These samples will be collected wearing nitrile gloves and using a sterile scalpel blade, and will be subsequently stored in gelatine capsules. Photographs of sampled paintings will be taken before and after sampling. We expect that sampling will take two days.   Characterisation analysis Paint characterisation analysis will be undertaken using protocols we have previously developed (Bonneau et al. 2012, 2017a, 2017b). Analysis will be undertaken by Bonneau at the Université de Sherbrooke, Canada. A multi-instrumentation protocol will be used to acquire all the information needed to characterise the pigment and its alterations through time. Samples will be analysed unprepared and in cross-section. They will be first observed with a microscope and then under a Scanning Electron Microscope (SEM). Analyses will be carried out with SEM- X-rays Energy Dispersive Spectroscopy (SEM-EDS), Raman spectrometry, and Fourier Transform Infrared (FTIR) spectrometry. All these steps will be undertaken on both unprepared samples and samples in cross-section, apart from the FTIR spectroscopy analyses, which will only be carried out on unprepared samples. Microscopic observation is used to assess the homogeneity of the sample: mixture, one material, weathering layer on top of the paint layer, alteration/weathering/paint layer under the paint layer, etc. The use of cross-sections make it possible to see the ‘stratigraphy’ of the sample (i.e., the different layers composing the paint) and thus to evaluate possible repainting or superpositioning of pigments. The SEM-EDS analyses have two objectives: to observe the shape of the particles and to analyse the chemical elements making up the sample. This is a first step in the identification of the colouring materials. Moreover, the shape of the particles is a determinant observation to distinguish carbon-based materials (long pieces with holes for charcoal, small balls for soot, flakes between 0.5 to 10 μm in diameter, for carbon blacks) and to understand the geological history of iron-bearing materials such as ochres. Following elemental analysis, molecular analysis will be carried out to characterize alterations of the pigment and weathering products. Raman spectroscopy will be employed for the molecular characterization of both pigments and weathering products. In the case of carbon-based pigments, it enables us to differentiate between carbon-based particles made of graphite (which have a Raman spectrum with only one peak centred about 1590 cm-1) from those consisting of carbon black, soot or charcoal (which exhibit broad peaks centered at 1350 and 1590 cm-1). Regarding other colouring materials, it identifies the exact molecular shape to correlate the paint to actual minerals and to some extent, geological provenance areas. When using Raman spectroscopy, molecules of some colouring materials and weathering products, such as clays and calcium oxalates, are difficult to excite with visible wavelengths. To overcome this problem and in order to complement the Raman analysis, we will undertake FTIR spectroscopy analysis. Results obtained from this multi-instrumentation protocol provide information on the nature of the pigment present in the samples analysed, its alterations (humic acids for example), and the nature of associated weathering products.