354 



MR. W. C. D. WHETIIAM ON THE IONIZATION OF 



solvent, and the curve shows the same drop at extreme dilution as do the curves 

 for acids, though in a less marked manner. But, in this case, a correction made on 

 the lines indicated above gives the results : 



I'OOO 

 1-002 

 0-996, &c. 



a 2 = 1-002 

 a s = 1-001 



a 3 = 1-000 

 a 8 = 0-999 



In this series the ionization of the first six solutions is constant, when corrected, to 

 an accuracy of two parts in a thousand, which is within the limits of experimental 

 error. 



It is probable, therefore, that in the case of permanganate, the drop in the curve is 

 caused by interaction between the salt and the residual impurities of the solvent, and 

 that this action is completed by the first quantity of salt added. In the case of acids, 

 however, some other explanation is needed. 



This result is confirmed by, another phenomenon which would not have been dis- 

 covered if the usual method of beginning with a strong solution and diluting it had 

 been adhered to. 



When the first quantity of stock solution is added to the solvent and mixed with 

 it, in the case of solutions having normal curves the resistance settles at once to its 

 final value, but when the drop occurs in the ionization curve, the resistance is found 

 to rise for some considerable time after mixing is complete, as though the action 

 which caused the diminution in conductivity required time for its completion. This is 

 illustrated by the following numbers : 



Barium Chloride. 



Copper Sulphate. 



