RICHARDS. 



CALOMEL ELECTRODE. 



19 



It is evident that there is a general tendency towards the fulfilment of 

 the conditions required by the formula. The agreement in the case of 

 cadmium, where the effect of irregular dissociation is most apparent, is 

 quite remarkable. The agreement in the cases of sodium and zinc is not 

 bad ; but the deviations in the case of barium on the one hand and that of 

 potassium and lithium on the other are quite too large to be accounted for 

 upon the basis of experimental error. It is obvious that in these cases we 



have some modifying circumstances ; our assumption that = -^ 



P P 



is evidently not justified. The attempt to penetrate further into the 



