On the distribution of red cell volumes.

A large number of red cell volume distribution curves can be generated on the same sample of blood by manipulating the length of the Coulter Counter aperture and the shape of the red cell. Curves derived from sphered red cells flowing through an aperture 70 x 98µ long, or longer appear to be relatively free from artefacts and therefore the most likely to be "real." The use of long apertures to size sphered red cells will give a volume distribution curve that is approximately Gaussian and with a mean that is large relative to the variance. Whether red cells are Gaussian or log normal in their volume distribution is very difficult to determine. On a particle population where volumes are controlled as closely as on red cells (i.e. with such a low variance), the normal and the log normal distributions become practically indistinguishable. If in the future this decision can be reached by some independent means, then the volume distributions using the Coulter volume transducer can be easily and simply fitted to the shape desired by appropriate manipulation of aperture length. These ideal volume distribution curves are not likely to differ markedly from those obtained under the conditions suggested in this paper. For the present, use of a 70 x 98µ aperture to size red cells sphered by suspension in isotonic Sodium Potassium Tartrate will contribute to considerably greater accuracy in identification of red cell subpopulations in abnormal blood.