The solubilities of Platinum (Pt) and Rhodium (Rh) in a haplobasaltic melt (anorthite-diopside eutectic composition) have been determined experimentally by using the mechanically assisted equilibration technique at 1300 degrees C, as a function of oxygen fugacity (10(-12) < fO(2), less than or equal to 1 bar), imposed by CO-CO2, N-2-H-2-H2O, Ar-O-2, and air gas mixtures. Samples were analyzed by sample nebulization (SN) inductively coupled plasma-mass spectrometry and, using some of these samples as standards, also by laser ablation (LA) inductively coupled plasma-mass spectrometry. The latter is a true microanalytical technique that allows small-scale sample heterogeneity to be detected. At each oxygen fugacity step, a time-series of samples was taken, to demonstrate that the solubilities converge on a constant value. In addition, solubilities were measured after both increasing and decreasing the imposed fO(2). The results fall into three groups, according to oxygen fugacity. At high fO(2)s (fO(2) greater than or equal to 10(-2) bars), samples are homogenous at all sampling scales. Both Pt and Ph predominantly dissolve in the silicate melt as 2+ species, with some evidence for Pt4+ and Rh3+ at the highest fO(2)s studied (air and pure O-2). From these data, we obtained the following expressions for the solubilities of Pt and Ph: Pt/ppb = 2100(fO(2)) + 10980(fO(2))(1/2) Rh/ppb = 68630(fO(2))(3/4) + 31460(fO(2))(1/2) At fO(2) < 10(-5) bars, the true solubilities of Pt and Rh appear to be obscured by Pt-Ph micronuggets, which remain suspended in the melt despite stirring on time scales of 10(3) h, resulting in samples that are heterogenous on the laser sampling scale. Samples at intermediate fO(2) (10(-2) to 10(-5) bars) are affected by the micronugget problem on the sampling scale of the conventional SN-inductively coupled plasma mass spectrometry, but these can be filtered out by analyzing on the laser sampling scale. Copyright (C) 1999 Elsevier Science Ltd.