THE COLORIMETRIC DETERMINATION OF 
HEM OGLOBIN.* 
By WALTER W. PALMER. 
(From the Hospital of The Rockefeller Institute for Medical Research.) 
(Received for publication, November 30, 1917.) 
_ The method to be described for the determination of hemoglobin 
depends upon the comparison, in a colorimeter, of carbon mon- 
oxide hemoglobin solutions, one of which has a known hemo- 
globin content. Hoppe-Seyler (1) was the first to describe car- 
bon monoxide hemoglobin and to make use of this stable com- 
bination for estimating the hemoglobin content of blood. He 
devised a “‘double pipette’ for comparing the unknown carbon 
monoxide hemoglobin solution with the standard carbon monoxide 
hemoglobin solution, prepared from hemoglobin crystals; but the 
method never came into general use, because of the many tech-. 
nical difficulties involved. Haldane (2) suggested a much simpler 
method for comparing carbon monoxide solutions, using the appa- 
ratus employed by Gowers (3) for comparing oxyhemoglobin 
solutions with a picro-carmine standard. This apparatus was 
later employed by Sahli (4) who prepared an acid hematin 
standard by adding dilute hydrochloric acid to blood. 
A critical discussion of the various methods in use for the 
estimating of hemoglobin is beyond the scope of this paper. As 
Haldane (2) has pointed out, artificially colored solutions and 
tinted glass present great difficulties in standardization with a 
definite strength of hemoglobin solution. With a certain strength 
of color solution or tinted glass, it is possible to imitate quite 
perfectly the tint of a given hemoglobin solution provided the 
quality of light remains the same. Any variation from these 
standard conditions, either in quality of light or strength of 
hemoglobin in solution leads to serious errors. Haldane at- 
* A brief report of this paper is given in Proc. Soc. Exp. Biol. and Med., 
1916-17, xiv, 175. 
119 
