Within the last few years, a number of workers in the field of medicine have expressed dissatisfaction with the methods commonly used in determining the amount of hemoglobin in the blood. They have pointed out not only that the different types of hemoglobinometer use different values of concentration of hemoglobin for 100 per cent on the scale, but also that the methods themselves are for the most part not sufficiently accurate for use in diagnosis. Of these two objectionable features, the latter is, of course, by far the more serious. The confusion and misunderstanding caused by dissimilar scale values can be avoided simply by the adoption of a uniform method of reporting hemoglobin, as recommended recently by Lindsay, Rice and Selinger;1 but if the method itself is not sufficiently accurate, errors in diagnosis are unavoidable.
ACCURACY OF METHODS IN COMMON USE
In order to obtain some idea concerning the accuracy