212 SCIENCE PROGRESS 



would proceed equally rapidly in both media ; the decay of 

 ionisation, on the other hand, might be assumed to depend on the 

 electrical forces between the ions, and therefore on the inductive 

 capacity of the medium. The equation 



HCl^H + C1 

 would therefore apply equally well in both cases, the only 

 essential difference being that the reverse action would be, say, 

 forty times as rapid in benzene as in water. This would lead to 

 a reduction of the coefficient of ionisation from 80 per cent, in 

 water to, say, 2 per cent, in benzene. As, however, the solutions 

 in benzene are excellent insulators, whilst the aqueous solutions 

 are excellent conductors, it is evident that the difference in 

 ionising properties is out of all proportion to the difference in in- 

 ductive capacities, and that the solvent must not merely " permit " 

 ionisation but exert a positive action in promoting it. 



In view of the difficulties attaching to the theory that the 

 electrification of the atoms may be regarded as a source of 

 energy, it is of interest to follow out the alternative supposition 

 that the ions may be treated as conductors which require an 

 expenditure of energy for charging. This leads to the expression 



E = 2jn x io 9 x - where / = mobility, e = s.i.c. 

 e 



for the energy of electrification of one gram-equivalent of ions 

 when treated as spherical condensers, the radius of which may 

 be determined from their mobility by the application of Stokes 

 Law. 1 



This result, which is independent of any assumption as to the 

 dimensions of the atom or the absolute magnitude of the ionic 

 charge, may also be expressed in heat units by making use of 

 Joule's equivalent, when the above expression becomes 



■n- 277T X IO 9 / „ 1 , 



h. = — - = 2000 - calories. 



4 - 2 x io 7 e t 



In water, taking e = 80, typical values for the energy absorbed 



; n electrification are 



+ + 



H 7950 cal. Na 11 10 cal. 



+ 

 CI 1650 cal. K 1640 cal. 



1 For the details of the calculation and a fuller discussion reference may be 

 made to the original paper (Bousfield and Lowry, Trans. Faraday Soc. 1907, 3. 

 126-128). 



