For a long time it has been known that iron occurs in the blood in some other combination than with hemoglobin, and recent studies1 have shown that this serum iron varies in quantity during iron starvation in experimental animals. In dogs rendered anemic by long-continued hemorrhage, with a limited intake of iron, the normal level of serum iron is reduced about 33 per cent, and when iron-containing foods or medicinal iron salts are given, the normal level returns, with consequent regeneration of the blood.
The object of this paper is to report clinical studies directed toward the application of these results to the problem of iron metabolism in pernicious anemia and secondary anemia in man.
The method for determination of iron devised by Elvehjem and Hart2 was adopted for this work, since its accuracy for small amounts of iron in biologic materials is unquestioned. One source of error in