Attempts to examine correlations between elements within the data through simple biplots and correlation analysis yielded very limited results. On the whole, in the combined data, there were very few strong positive or negative correlations. While there were stronger positive correlations in each of the artefact datasets, with several elemental associations perhaps indicating a corrosion 'package', ultimately there were too few data points taken to have any confidence in these associations. While more data points would have been useful in clarifying these connections, ultimately it is clear that both artefacts have significant surface enrichment or depletion of certain elements. The main goal of course is too provenance the two marbles, not explore how their surfaces have been altered, and so the limited number of points analysed for each were sufficient.