32 Larmor, Physical Aspect of tJie Atomic Theory. ' 



the presence of hydrogen free or combined as a potent 

 stimulus to ternary reaction.* 



lojiisation and Solution : Available Energy and 

 BertJielofs rule. 



At first sight it might appear that the principles of 

 statistical equilibrium in dilute systems would afford a 

 criterion as to the intimate process of dissociation, 

 whether into ions or molecules. Thus, to fix the ideas, 

 the two reactions, HC1 = H + C1 and 2HC1 = H2 + Cl 

 would come to different equilibria, of the types n = k .n^n^ 

 and ir = k'.n\n\. But in reality the subsequent aggre- 

 gation of H and H into H., itself involves an equilibrium 

 ni= J{Kn\), so that the discrimination is not possible on 

 these considerations alone. Nor is it ever possible in this 

 way on the thermodynamic theory, which can be seen 

 (c/. Appendix) to be consistent with separate independent 

 equilibria as regards every type of reaction that is formally 

 possible in the system. 



The process of ionisation in a liquid solvent is 

 obviously very different from ordinary gaseous dissociation. 

 The view that some such special type of dissociation is 

 required in order to form a coherent mental picture 

 of Faraday's electrolytic results, must really in strict logic 

 go back to Clausius' ideas of about fifty years ago. It is 

 true that he did not venture to suggest more than 

 extremely slight ionic dissociation. But once the mere 

 possibility is granted, there is no ultimate escape from 

 the permanent ionic separation of Arrhenius : for it is 

 only a question of making the solution more and more 

 dilute in order to diminish indefinitely the chance of 



* I have ventured to add an abstract discussion on the formal possibilities 

 that are open, which was drawn up about a year ago, as an Appendix. 



