Abstract
Centrifugalization of blood at a relative centrifugal force of 1800 for 1 hr. (as usually performed for hematocrit detns.) did not separate cells and plasma completely, and considerable amts. of plasma were left in the interstices of the cell mass, thus giving a falsely high hematocrit value. By adding a known amt. of Evans Blue dye (T-1824) to a known vol. of blood and determining the degree of dilution of the dye by colorimetric methods, the exact vols. of plasma and cells were calculated since the dye was distributed through the plasma and did not enter the erythrocytes. The validity of this dye dilution method was verified by methods employing the dilution of plasma protein in a known vol. of whole blood when a known vol. of saline was added; and by determining the dilution of erythrocytes "tagged" with radioactive Fe when they were added to whole blood. In all instances the true % of cells was quite constantly 8-9% lower than the centrifuge hematocrit. A positive error of this degree in the centrifuge hematocrit results in an error of 15-20% in the total erythrocyte vol. as detd. in vivo by the Keith, Rowntree, Geraghty method, and leads to erroneous conclusions as to the actual size of erythrocytes, and as to the concs. of various substances in erythrocytes.