Mr Richardson, Note on the Diffusion of Hydrogen, etc. 27 



Note on the Diffusion of Hydrogen through Palladium. By 

 0. W. Richardson, M.A., Trinity College. 



[Received 15 November 1904.] 



The rate of diffusion of hydrogen through a palladium tube 

 has been examined quantitatively by Winkelmann*, who found, 

 contrary to what one might expect d priori, that the mass of gas 

 diffusing in unit time at constant temperature was proportional to 

 the gradient of a power of the pressure less than unity. Winkel- 

 mann explained this result by supposing that the hydrogen 

 dissociated and that only the atoms had the power of diffusing 

 through the metal. More recently an interesting research on the 

 same subject has been published by G-. N. St Schmidt -j-, whose 

 conclusions might tend to throw doubt on the advisability of the 

 assumption of dissociation. 



The diffusion of hydrogen through platinum was also ex- 

 amined by Winkelmann |, who again found that the pressure 

 variation could be explained on the hypothesis that dissociation 

 occurred. A more detailed examination of this subject has been 

 made by the author conjointly with J. Nicol and T. Parnell§. 

 The principal results of this investigation were that the rate of 

 diffusion varied, (1) as the gradient of the square root of the 

 pressure, down to pressure differences of 3 mm., and (2) as an 

 exponential function of the temperature of the tube. All the 

 results were found to be accurately expressed by a formula de- 

 duced theoretically on the hypothesis that the hydrogen inside the 

 metal was in a state of dissociative equilibrium and that only the 

 dissociated atoms diffused freely. The experiments yielded no 

 evidence of an appreciable amount of dissociated hydrogen outside 

 the metal. 



In view of the consistent explanation of the diffusion of 

 hydrogen through platinum furnished by the dissociation theory 

 and the general similarity in the phenomena as regards both 

 metals, it does not seem too much to suggest that a hypothesis 

 which appears to furnish a complete explanation of the one case 

 should not be lightly rejected as regards the other. The object of 

 the present note is to point out that some of the objections to the 



* Drude's Ann. Vol. vi. p. 104. t lb. Vol. xin. p. 747. 



X lb. Vol. viii. p. 388. § Phil. Mag. S. 6, Vol. vm. p. 1. 



