at the Electrodes in a Solution. 



55 



symmetry) constantly c ; the difference between real and 

 calculated values therefore is 



</ 



l_ f ' dt 



J? 



16Kt' 



It is easy to see that as 

 function is smaller than 



lon°r as t< 



8K 



the value of this 



Vi 





and from statement No. 2 we can conclude that the expres- 

 sion just given is also an upper limit for the differeuce 

 between the real and calculated values at the cathode. In 

 all the experiments carried out, this quantity is absolutely 

 negligible. 



We now come to the reconsideration of our assumption 

 that diffusion in a solution is unaffected by the passage of a 

 current through it. This question was first tested experi- 

 mentally by H. F. Weber in the case of a zinc sulphate 

 solution, and answered by him in the affirmative. It has 

 recently been very thoroughly examined theoretically by 

 Kohlrausch *. As shown by him, changes of concentration 

 are in general brought about by the passage of a current 

 through a solution of non- uniform concentration. These 

 changes are in the case of a cylindrical vessel given for each 

 ion in our notation by the formula 



3<? 



a*" 



= + 





(9) 



96540 ~di 



in which c is the concentration of the ion under considera- 

 tion, a is its mobility, and k the conductivity of the solution. 

 The positive or negative sign must be taken according as the 

 ion is positive or negative (in contrast to Kohlrausch's nota- 

 tion, where the positive ions move from smaller values of 



CLC 



x to greater). As is known, — = n is Hittorfs transport 



value of the ion under consideration, and we can Write the 

 above equation in the form 



Be _ i 'dn'dc 



~dt ~ ± 96540 Tc 5? 



(10) 



from which it can be concluded that in the interior of a 

 solution of a single salt in which the transport values are 



Wied. Ann. lxii. p. 209 (1897). 



