Experimental data on the stability of hercynite yield very large discrepancies in the derived values of its free energy. Attempts were made to resolve these discrepancies through an electrochemical study of the equilibrium 2 FeAl2O4 = 2 Fe + 2 Al2O3 + O2. These measurements proved, however, also to give a wide range of apparent free energies, so a detailed phase equilibrium study was made of the Fe-Fe2O3-Al2O3 system at 1 atm pressure. Our results demonstrate that difficulties in measuring the position of Equilibrium 1 in T-fO2 space are due to the appearance of a new phase, a defect spinel of approximate composition Fe0.5Al2.33O4 and a cell edge a0 = 8.00 Å. The data show that previous experiments at 1 atm measured either the hercynite-corundum-Fe equilibrium or various equilibria involving the defect spinel phase, and hence the apparent discrepancies in the hercynite free energy. The results also indicate that hercynite is more stable at 1 atm than would be predicted from high pressure phase-equilibrium experiments. This appears to be due to the solution of excess Al2O3 in hercynite at low pressure. -from Authors