202 GENERAL DISCUSSION OF RESULTS. 



from 12 to 25, we should expect to find the difference in question greater from 

 to 12 than from 12 to 25. Here, again, exactly the opposite is true. 



The difference between the conductivity of the mixed salts and the sum of the 

 conductivities of the constituents can not, then, be accounted for as due solely to a 

 change in dissociation caused by each salt driving back the dissociation of the other 

 with a common ion. Stine pointed out that there are three other factors which 

 must be taken into account: (1) change in hydration, which changes the size and 

 mass of the ion; (2) change in the viscosity of the solution with change in tempera- 

 ture, which changes the friction of the ions in moving through the solution; (3) change 

 in the number of dissolved particles ions and molecules. 



In the cases of potassium and ammonium chlorides, which are very little hydrated, 

 the first factor (change in hydration) plays a very minor role; the third undoubtedly 

 plays some part, since a becomes smaller the higher the temperature; the second, 

 or change in viscosity with rise in temperature, is undoubtedly the most important. 



The effect of the viscosity of the solvent on the conductivity of electrolytes dis- 

 solved in that solvent was thus clearly recognized by Stine, and pointed out by him 

 a half-dozen years ago. 



This work was strong evidence for the solvate theory of solution. It showed that 

 the effect of one salt on the hydration of another salt with which it was mixed was 

 what would be expected from the law of mass action. This was worked out for a 

 sufficient number of typical substances, with very different hydrating powers, to 

 enable us to draw a general conclusion as to the correctness of the solvate theory 

 of solution. In some cases there are apparent discrepancies between the results 

 obtained by Stine and those found by subsequent workers with the same substances. 

 It must be remembered that the subsequent work with both the freezing-point and 

 the conductivity methods was done after these methods were both greatly improved. 



Further, in the work of Stine the aim was to obtain comparative values for the differ- 

 ent substances rather than the highest degree of accuracy for the individual compounds. 



RESULTS OBTAINED BY PEARCE. 



The work which had already been done 1 in this laboratory on the freezing-point 

 lowerings of water produced by electrolytes in general, had shown that they prac- 

 tically all give so-called "abnormal lowerings" that is, lowerings much greater 

 than can be accounted for from their dissociation. These freezing-point lowerings 

 become more and more abnormal the more concentrated the solution. 



On the other hand, Jones, 2 when working in Ostwald's laboratory, had used the 

 freezing-point method to measure the dissociation of dilute solutions of electrolytes. 

 This was done for the purpose of seeing whether the freezing-point and conductivity 

 methods gave the same or different values for the dissociations of electrolytes in 

 dilute solutions, or whether the so-called solubility method of measuring dissociation 

 gave the true values. It was subsequently shown that the last-named method gave 

 incorrect results. 



The freezing-point method, as used by Jones in 1892, gave values for dissociation 

 that agreed fairty well with those calculated from the conductivity measurements 

 of Kohlrausch. It was not known at that time that concentrated solutions of these 



'Carnegie Institution of Washington Publication No. 60. =Zeit. phys. Chem., II, 110, 529; 12, 622 (1893). 



