The treatment of patients with iron compounds is of very old date, but its value is still much disputed. Bunge, for instance, was of the opinion that the effect of iron compounds on chlorosis was suggestive. Iron has at times been administered in inorganic compounds, at times in organic ones and at other times as reduced iron. Later years have brought several new points of view, based on clinical and experimental investigations. Ferrous compounds are more effective in chlorosis than ferric compounds. Heubner was of the opinion that patients with chlorosis are not able to convert the trivalent iron of the food to the divalent form of hemoglobin. Amatsu1proved that ferrous tartrate, administered intravenously, is far more poisonous than ferric tartrate, and also that it is far more effective in animals with anemia after hemorrhage. Later Starkenstein and his collaborators,2who assume that a substance is effective