506 



UNIVEESITY OF VIRGINIA PUBLICATIONS 



sociated molecule must be larger, ^(nless the ion is hj^drated to a greater 

 extent than the molecule. Ohohn has made a special studj^ of the rate 

 of diffusion of flectroljd^es, comparing the diffusion constant calculated 

 from Nernst's formula with experimental values. In most cases the 

 observed rate of diffusion was less than the calculated, and the discrepancy 

 was far greater for weak than for strong electrolji;es. This is easiest 

 explained by assuming that the undissociated molecule migrates more 

 slowlj^ than the ions, hence, the diffusion constant calculated for complete 

 dissociation wall alwaj^s be too great unless the dissociation be complete. 



These observations are not in accord with the experimental facts for our 

 case, and they apparently contradict the possible explanation suggested some 

 pages back; however, the theory of Nernst and the experiments of Oholm 

 have value for water solutions only and need not hold for KI solutions. 



Influence of an added electrolyte having a common ion. It has been 

 shown by Abegg and Bose (38) that the effect of an added electrolyte 

 having an ion common -n-ith the diffusing electrolj-te is to make the rate of 

 diffusion approach that of the ion that is not common. The following 

 scheme, taken from the article bj^ Abegg and Bose, will make clear why 

 this should be so, and as a matter of experimental fact it is so. 



Let us consider two diffusion sj^stems. In the first let there be two 

 layers of the same electrolj^te. A.B. having in one layer the concentration 

 c and in the other laj'er the concentration c + dc. In the second system 

 let there be the same electrolj-te having the same concentration, and also 

 the added electrol^'te A.X ha-ving the same concentration C in both layers: 

 C being very great in comparison vAth. c. Then vnll the scheme given 

 below represent the concentration of all kinds, supposing the dissociation 

 to be complete. We see that in the second sj'stem, the common ion A 

 has practically the same concentration in both layers, and hence it diffuses 

 very slowly. This leaves the uncommon ion as the onlj^ factor of impor- 

 tance in the rate of diffusion. Of course in practice these conditions are 

 not realized, but the rate of diffusion approaches that of the uncommon 

 ion as the concentration of the added electrohd:e increases in comparison 

 with that of the original diffusing electrolji;e. 



Scheme copied from Abegg & Bose (38). 



