Krogh's (1919a) measurements of the rate of diffusion of 02, N2, CO and CO02 through animal tissues have formed the basis of many calculations of the role of diffusion in physiological processes (see, for example, Krogh, 1919b;Hill, 1928;Roughton, 1932;Jacobs, 1935). No information, however, appears to exist as to the rate of diffusion of these dissolved gases through homogeneous solutions of proteins: the value of the diffusion coefficients of 02 and CO in concentrated haemoglobin solutions has, in particular, been a desideratum in the theoretical investigations of Roughton (1932) and Nicolson & Roughton (1951) on the rate of exchange of 02 and CO between the interior of the red blood cells and the surrounding fluid. At the concentrations found in the red cells, the haemoglobin molecules must, according to Perutz (1948), be practically touching and there should accordingly be both marked geometrical obstruction and viscous resistance to the passage of small molecules of dissolved gases between the large protein molecules. The relation between the value of the diffusion coefficients and the protein concentration is therefore also of some theoretical interest.In the present paper we have determined the diffusion coefficients of N2 and CO in haemoglobin solutions ranging from 0 to 45 g haemoglobin/100 ml., the latter concentration being appreciably higher than that usually found in the red cells. In the case of N2 there has been no complication from any simultaneous chemical reactions of the penetrating substance: with CO, on the other hand, there has, from time to time, been trouble from side reactions even when the concentrations of CO have been such as to ensure that the haemoglobin is at least 99.99 % saturated with CO. Fortunately for our present purpose, N2 and CO have the same molecular weight and are very similar in many of their physical properties, including solubility. A priori, their diffusion coefficients in haemoglobin solutions should be about the same: molecules has also to be taken into account. Perutz's work would appear to make the need for such an allowance very unlikely, but fortunately we have found it possible, by means of an alternative method to our main one of measuring the diffusion coefficients of CO in haemoglobin solutions, to obtain direct experimental evidence on this point.
METHODS I. Diffusion apparatusIn most of the experiments diffusion was measured between a gas phase containing CO or N2 at 1 atm pressure, and haemoglobin solution containing these gases in solution, but at different partial pressures from those in the gas phase. The apparatus, which was specially designed for use in conjunction with the Van Slyke gasometric methods, is shown in Fig. 1. It consists essentially of a large glass tap, T (bore about 1 cm), one lead of which is closed at the end whilst the other is fused to a straight capillary bore tap, U, and also to a 3-way capillary bore tap, V, the 265 266 I. S. LONGMUIR AND F. J. W. ROUGHTON leads from which are inclined at angles of 1200 to each other. On the...