PHYSICAL CHEMISTRY 369 



are completely ionised at all concentrations. Maclnnes com- 

 putes the changes in the mobility of the ions which would be 

 required to explain the variations in conductivity on this 

 assumption. He finds that between o-oiN and i-oN, the 

 change in conductivity of the chlorine ion, Tq^ is from 61 -6 

 to 47, and that the value of this product is independent of the 

 nature of the univalent cation present. The agreement is 

 especially good if the viscosity of the solution be taken into 

 account. The conductivity of an aqueous solution is thus 

 given by the sum of two conductivities, one due to the anion, 

 and one due to the cation, i.e. the Kohlrausch law is extended 

 to solutions of electrolytes up to i-oN. 



Much more attention is now being paid to the alternative 

 view that solutions of electrolytes are completely ionised. The 

 theoretical work of Sutherland, Bjerrum, Milner, and Ghosh 

 in this field has shown that conductivity can be adequately 

 explained on this assumption. A paper published by Ghosh 

 (Zeit. Phys. Chem., 1921, 98, 211) gives a summary of his 

 views on ionisation. In this theory the undissociated molecules 

 of the Arrhenius theory are replaced by a group of ions held 

 together by electrostatic forces, those ions becoming active 

 which possess a sufficiently high kinetic energy to overcome the 

 electric field. The ratio of the active to the inactive ions is 



A 



given by the relation a = e "^^ where A the work required to 



NE* 

 separate the electrical group is equal to ^r^ where r is the 



distance between the ions, when these are arranged in a cubic 

 lattice, D the dielectric constant, and E the electrical charge. 

 The degree of ionisation a calculated from this formula is 

 accurate between coi and 0-0002 N. The author does not 

 apply his equation to concentrated solutions. 



The thermodynamic treatment of dilute solutions possesses 

 a much wider range of application than any theory based on 

 kinetic grounds. The thermodynamic degree of dissociation 

 has the same value, from whatever physical property it is 

 derived. In this respect it possesses a considerable advantage 

 over the degree of dissociation calculated from conductivity 

 data. G. N. Lewis and Merle Randall {J.A.C.S., 1921, 43, 

 1 1 50) have published a summary of several chapters on this 

 question in their new book on Chemical Thermodynamics. 

 They point out that the " true " degree of dissociation cannot 

 be logically defined, since the fixing of the limiting distance 

 within which two ions are associated and beyond which they 

 are dissociated — is quite arbitrary. On this account it would be 

 expected that the degree of dissociation would be dependent 

 on the experimental methods which were employed in its 



