IT is sometimes assumed that the oxide film formed on iron at room temperatures is uniform in thickness across a specimen, and that where several oxides exist together they are present as layers stacked neatly one above the other. In an electron microscope investigation of the oxidation of mild steel using oxide films stripped by the iodine/methanol technique1 the only oxide that consistently conforms with this idea is magnetite, Fe3O4. Fig. 1 shows magnetite with its diffraction pattern on a film stripped from mild steel. The metal specimen was electropolished in a 5 per cent perchloric acid/95 per cent acetic acid mixture, washed in methanol and given no additional air exposure before the surface layer was stripped. It is interesting to note that these stripped oxide films remain quite stable if kept in a desiccator for a year and still give the same diffraction results after this length of time. Magnetite typically appears as an almost featureless layer, though with experience its presence may be recognized without examination of the very characteristic orientated arced diffraction patterns it produces. The (001) projection, as in Fig. 1, occurs most frequently although other orientations have been observed on the surface layers produced on thin foil corrosion specimens and there is evidence that the orientation changes from grain to grain. The usual difficulty in distinguishing between Fe3O4 and γ Fe2O3 was experienced in this work2 since these oxides have the same crystal structure and lattice parameters. It is noticeable that the (200) reflexions are absent and that no mixed index reflexions appear–these facts point to Fe3O4. There are, however, eight (311)3 reflexions that would not be expected and these consistently appear with this type of pattern on foils as well as on stripped oxide films. They suggest that the film is of the order of one unit cell in thickness, approximately 10 A, and that the reflexions could arise by streaking4.