The interpretation of low-resistivity anomalies in the lithospheric mantle of several cratonic regions has invoked hydrogen, or connected networks of graphite with iron-rich silicates, and/or metal sulfides. Electrical laboratory measurements are a powerful approach for exploring these alternatives. We report electrical measurements of two xenoliths (pyroxenite and dunite) from Tanzania; two metal sulfides (FeS and Fe-S-Ni); and several mixtures of metal sulfides (3.4-18.2 vol.%) with xenolith. A multi-anvil press was employed to maintain a 2 GPa pressure and temperatures up to 1,627 K. The addition of 3.4 vol.% FeS to the pyroxenite or dunite matrix has little effect on bulk resistivity, particularly for T > 800 K. However, the resistivity drops dramatically-by factors of up to 1,000, depending on temperature-upon addition of 6.5 vol.% FeS in the dunite. Addition of 18.2% FeS causes a further decrease of the same magnitude relative to the 6.5% sample. Scanning electron microscope images do not reveal the formation of a connected FeS network as part of the decreased resistivity. Possible explanations for the apparently conflicting results include connections of the sulfide that are not imaged in the back-scattered images, either because of limited resolution, or perhaps the inherent limitations of the 2-D perspective. The complete data set (xenoliths, metal sulfides, and mixtures) was modeled with a modified version of Archie's law, and we find satisfactory agreement over a truncated temperature range. We conclude that the low-resistivity anomalies in Tanzania, Kaapvaal, and Gawler cratons can be explained by the presence of a few vol.% of solid sulfide.

The apparent end of the internally generated Martian magnetic field at 3.6-4.1 Ga is a key event in Martian history and has been linked to insufficient core cooling. We investigate the thermal and magnetic evolution of the Martian core and mantle using parameterized models and considered three improvements on previous studies. First, our models account for thermal stratification in the core. Second, the models are constrained by estimates for the present-day areotherm. Third, we consider core thermal conductivity, kc, values in the range 5-40 W m(-1) K-1 as suggested by recent experiments on iron alloys at Mars core conditions. The majority of our models indicate that the core of Mars is fully conductive at present with core temperatures greater than 1940 K. All of our models are consistent with the range of k(c)=16-35 W m(-1) K-1. Models with an activation volume of 6 (0) cm(3) mol(-1) require a mantle reference viscosity of 10(19)-10(20)(10(20)-10(21)) Pa s.