In differentiating types of anemia, clinicians are constantly making use of the relationship between the number of erythrocytes and the hemoglobin. This relationship is usually expressed as the color index. In view of its wide use, it is surprising how vague are the ideas of the factors on which the color index is fundamentally based. This is due, in some measure, to the common use of only relative methods for the determination of hemoglobin. To arrive at precise information, it is necessary to have the hemoglobin content expressed in exact terms of grams per hundred cubic centimeters. Suggestions for a normal hemoglobin standard on this basis have been pointed out elsewhere.1 Very few careful studies of color index have been made. It is not even universally agreed that the color index of normal blood is always 1.00. Meyer and Butterfield,2on the basis of a very few observations with hemoglobin