Supplementary Material to the paper “Fe-Mg partitioning between olivine and high-magnesian melts and the nature of Hawaiian par

In this Supplement, we provide four sections intended to amplify and/or expand on specific topics raised in the main text. The first section is an assessment of the data underlying Roeder and Emslie (1970) as given in an unpublished table made available by the senior author. We assess general quality and consider possible sources for the discrepancy in values between the canonical 0.30 obtained by Roeder and Emslie (1970) and the ~0.34 preferred in this study. Section 2 provides a detailed description of the projection technique used in the main text to reduce compositional dependences of on melt and olivine composition for literature data. In Section 3, we evaluate the sources of discrepancies between predictions of the Ford et al. (1983) and Toplis (2005) models for and our experimental data. Finally, Section 4 is a listing of all of the references for data incorporated into our database of olivine-liquid pairs from the literature. D,Fe+2 -Mg K