The determination of ferrous iron in minerals, such as staurolite and kornerupine, which cannot be effectively dissolved by dilute sulphuric and hydrofluoric acids has been made possible on the macro-scale by the work of H. P. Rowledge; the mineral is fused with sodium fluoride and boric oxide at 900°C. in a sealed pyrex glass tube, the fusion cake dissolved in hot dilute sulphuric acid in a current of carbondioxide, and the solution titrated with potassium permanganate. We have met with no difficulty in applying the method on the macro-scale, but when only small amounts of a mineral are available several difficulties arise, and modifications become necessary.The air present in the sealed tube of a Rowledge fusion, using the quantities and dimensions he specifies, is already sufficient to oxidize about 3 to 5 mg. of ferrous oxide. But when the weight of the sample is reduced to 10 mg., it is not possible to reduce the dead-space in the tube in proportion ; it can hardly be reduced below about ¼c .c. which is sufficient to oxidize 0·3 mg. or 3% FeO and is likely to be more effective in this direction than on the larger scale, since the surface of the fusion is proportionally greater.