250 Mr. J. J. Thomson on the 



or substituting for n and q their values from (8) we have 



>/;_ i 



= i^(M-rj(N-r) .... (9) 



Let us first suppose that equal quantities of the two gases 

 hydrogen and chlorine are present, so that M = N; in this case 

 equation (9) becomes 



- {■**£?} *<*-'> • • • • (10) 



Now if the decomposition is due to the action of some 

 external disturbance such as light, radiant heat, or electricity, 

 the fs and t's will not be functions of the pressure, and we 

 shall get an equation for r/N, which does not involve ihe 

 pressure; thus the proportion of the compound dissociated 

 is independent of the pressure. This case offers, in this 

 respect, a remarkable contrast to the preceding one. But 

 even if the compound were dissociated by heat, and we were 

 to suppose the dissociation produced by the collisions, though 

 we gave reasons before for supposing this view to be erro- 

 neous, still since only t 1} t 2 , t 3 are affected by the number 

 of collisions, and as they enter into equation (10) in the 

 form t^/titz, we see that unless the ratio of these quantities 

 is altered by the collisions, the dissociation will be the same 

 at all pressures. Thus we should expect that in this case 

 the dissociation will be very much less influenced by pres- 

 sure than in the preceding one. This has been confirmed 

 by the researches of M. Lemoine on the vapour of hydriodic 

 acid; he found that the dissociation of this compound was 

 only very slightly affected by the pressure. 



Equation (9) enables us to find the quantity of hydro- 

 chloric acid produced when hydrogen and chlorine are mixed 

 in any proportion, though no quantitive experiments seem 

 to have been made on hydrochloric acid, yet M. Lemoine 

 [Annates de Chimie et de Physique, 5 me serie, t. xii.) has made 

 a very elaborate series of experiments on the analogous case 

 of the combination of hydrogen and iodine, and we can com- 

 pare the results of our theory with his experiments. The 

 experimental results given below were made at a temperature 

 of 440° C. We determine the value of t 3 2 T 1 T 3 /t 1 t 2 T2 2 by 

 making the result given by equation (9), when N = M, agree 

 with the experimental results : if we do this we arrive at the 

 following results: — 



