The Calibration of Diffusion Membranes and the Calculation of Molecular Volumes from Diffusion Coefficients

In the membrane method for measuring diffusion coefficients (Northrop and Anson (1928-29)) diffusion takes place through a porous disc. Under these conditions convection currents which usually make diffusion measurements difficult are avoided, the time required for a measurement is greatly shortened, and it becomes possible to remove the diffused substance for estimation. Under suitable conditions the rate of diffusion of one substance is independent of other substances present. To calculate the diffusion coefficient it is necessary to know only the per cent, not the absolute amount, of the original material which passes through the disc in a given time. The method can be applied, therefore, even to impure biological substances which can be estimated only by activity measurements. Since the dimensions of the pores are unknown, the membrane is calibrated by measuring the rate of diffusion through it of a substance of known diffusion coefficient. The first part of this paper describes the calibration of diffusion membranes with sodium chloride, potassium chloride, and hydrochloric acid and summarizes the evidence that the membrane method yields correct diffusion coefficients. The second part of this paper states the assumptions made in calculating molecular volumes from diffusion coefficients by Einstein's law, states the consequences if these assumptions are not valid, and discusses the possible reasons why molecular volumes of various proteins calculated from their diffusion coefficients are higher than the molecular volumes calculated from osmotic pressures and sedimentation data. The third part