As a follow-up to previous gas analysis experiments in which pyrite was heated to 681 K in an anoxic (oxygen starved) atmosphere, the first oxidation product, FeSO4, was studied as a bulk material. No decomposition of FeSO4 to Fe3O4 was observed in the temperature range studied. The lack of decomposition of bulk FeSO4 to Fe3O4 suggests that FeS2 oxidizes directly to Fe3O4, or that FeSO4, FeS2 and O2 react together to form Fe3O4. Magnetic susceptibility and magnetization measurements, along with magnetic hysteresis curves, show that small particles of Fe3O4 form on the pyrite surface, rather than a continuous layer of bulk Fe3O4. A working model describing the oxidation steps is presented. ?? 1990.