1911.] 



The Properties of Colloidal Systems. 



245 



actual " vacuum " for electrolytic diffusion, since water itself is ionised and 

 always contains impurities. For this reason, measurements are not reliable 

 when Ci is less than m/1000. 



The side of the membrane in contact with the more dilute solution will be 

 charged positively to the inner side, owing to the movement of the diffusible 

 positive ion under osmotic forces in this direction until these forces are 

 counterbalanced by electrostatic attraction of the negative ion. 



In the experimental testing of the equation for the potential difference 

 between the two sides of the membrane I have met with some difficulty. 

 The first experiments were made with an osmometer of the pattern described 

 by Eoaf,* constructed of glass and ebonite only. A T-tube in the course of 

 that leading to the manometer was connected with a tube containing Congo- 

 red solution of the higher concentration, as contained in the osmometer, but 

 made stiff with gelatin in order to withstand the pressure. In later 

 experiments, it was found unnecessary to use the gelatin. The osmometer 

 was immersed in the more dilute solution, from which a syphon tube dipped 

 into Congo-red solution of the same concentration as that within the 

 osmometer. Into this solution the tube of an Ostwald calomel electrode was 

 immersed, while a second similar electrode was in connection with the 

 content of the osmometer, with or without the gelatin protection. The 

 system was, therefore, symmetrical, with the exception of the presence of the 

 parchment-paper membrane between the two solutions of the dye on the one 

 side. The electromotive force was measured by compensation with a' 

 potentiometer in the usual way, using a capillary electrometer as indicator. 



It was found that the sign of the potential difference was as deduced from 

 theoretical considerations, viz., the side of the membrane in contact with the 

 more dilute solution was positive to the other side, or, as the stronger 

 solution was always inside the osmometer, the outside was positive to the 

 inside. But values corresponding to those calculated from the equation 

 given above were obtained, as a rule, only immediately after the apparatus 

 was put together. In one such case, solutions of 30 and 80 litres dilution 

 gave an electromotive force of 0"025 volt against the calculated one of 

 0-022 volt. This value soon fell to 0-0136 volt, although the manometer had 

 not perceptibly fallen. On the next day, however, it had risen again to 

 0-0165 volt. 



In another experiment, in which the relative ionic concentration of the 

 two solutions was determined, by conductivity measurements, to be m/32 

 and ra/150 respectively, the E.M.F. found was 0*025 volt, instead of 

 - 022 volt calculated. With more dilute outer solution, the E.M.F. observed 

 * ' Quart. Journ. Exp. Physiol.,' 1910, vol. 3, p. 80. 



