The rapid replacement of the circulating hemoglobin in the injected animals occurs chiefly through an increased production of corpuscular substance having the same color index as that found in uninjected, anemic controls. By color index in this connection is meant the relation of hemoglobin to the volume of the massed corpuscles.

Late in the period of bleedings and hemoglobin injections the demand for stroma for the new-formed blood is far greater than in control animals that have been merely bled, yet the circulating corpuscles show no lessening in resistance to salt solution, such as might perhaps be expected were there a stroma lack. The hematopoietic tissue of the injected animals undergoes an extensive increase—a fact which speaks strongly for the view that the elements out of which stroma is formed are still abundant.

The factor which determines the spread of red marrow during anemia is shown by our experiments to be the presence in the body of hemoglobin, or perhaps of its precursors, in excess of the amount which can be utilized by the marrow already existing. Numerous illustrations in support of the point can be adduced from human pathology. Two will suffice. The widespread "currant jelly" marrow of pernicious anemia is found in an organism rendered anemic but supplied with hemoglobin in excess; while the pale, restricted marrow of cases suffering from chronic anemia due to repeated hemorrhages is associated with depletion of the constituents necessary for pigment production.