As the author has previously shown, a definite small part of blood Fe is demonstrable by treating blood with dilute acid ("easily separable" blood-Fe). From continued researches of the author, partially in agreement with the data of Lintzels, there results the following 1. The amount of easily separable Fe is independent of the hemoglobin content. 2. The Fe separation proceeds as a "complete" reaction. 3. Thf expected formation of hematoporphyrin by separation of iron from blood pigment does not occur. 4. There is a consistently fleeting course of separation for acid (Barkan, Lintzel) and for weakly alkaline (Lintzel) reactions. While Fe may be separated from fresh hemoglobin crystals even after repeated recrys-tallization, under the same conditions any separable Fe will not be obtained from old crystalline preparations. The interpretation of the hitherto unexplained reactions as the separation out of the blood pigment molecule, because of the above experimental findings 1 and 2 might be tenable only in the following ways: either a part of the Fe in the hemoglobin molecule must be different from and accordingly more lightly bound than the greater remainder; or, on the contrary, the greater part of the ordinary hemoglobin must be mixed (itself with strongly bound Fe) with a smaller and again variable amount of the same sort of lightly bound Fe compound. These hypotheses make difficulties which are heightened by the above further findings 3 and 4. The author holds his original theory as a working hypothesis. He assumes that the hemoglobin found in blood corpuscles contains a small amount of another Fe-containing compound, adsorbed; from which, by usual methods of handling, Fe may be separated. The artificial liberation from fresh hemoglobin crystals of easily separable Fe is not yet successful. Concerning the chemical nature of the accompaniments of blood pigment (as postulated by the author), characterized by their easily separable Fe, there is as yet nothing to say. There are hints in the old literature of the presence of Fe, not belonging to blood pigment, in the blood of mammals including man. In cattle was found an average of 17 mgm. per liter of readily separable Fe. About the same amount was found by the author in geese; for horse- and dog-blood the average was higher; for men and for rabbits, possibly somewhat lower. Values under about 8 mgm. per liter have not been observed. Of the easily separable Fe, by far the greatest amount is from the corpuscles. That amount, computed from the dry substance, corresponds in large degree to the non-hemoglobin Fe of other animal cells. A concentration of Fe in the corpuscles does not proceed in vitro. The washed, hemoglobin-free, erythro-cyte-stroma (cell fragments of Ellinger) contains its entire Fe in easily separable form; it amounts to only a fraction of that present in the uninjured corpuscles, but is enough to admit of analysis after combustion of the organic matter. Plasma and serum regularly contain a small amount of acid-soluble Fe, about 1/250 of the hemoglobin Fe in blood. On standing, coagulated blood shows a gradual increase of plasma Fe, limited by the transport out of the blood corpuscles. In conditions of anemia and blood regeneration there appears the increase of blood corpuscle respiration quite independent of the easily separable Fe and the hemoglobin content. Feeding higher Fe doses has an influence only when the Fe value is low, in which case it raises it to normal. The total easily separable Fe is regarded as belonging to the complex of "transport iron." Concerning the mechanism of Fe transport, an hypothesis is developed.