208 GENERAL DISCUSSION OF RESULTS. 



temperatures, and the temperature coefficients calculated. The curves for conduc- 

 tivity and fluidity could then be compared, and we could see whether the minima 

 in the one corresponded to the minima in the other. 



The questions raised were: Will those salts with small molecular conductivities 

 when completely dissociated gives value of n m which are inversely proportional 

 to viscosity; whether the product of viscosity is a constant for mixed solvents and 

 at different temperatures ? Is this value the same for different electrolytes ? 



The salts used were lithium nitrate and cadmium iodide. It is well known that 

 acetone has very considerable dissociating power. It was found that lithium nitrate 

 was practically completely dissociated in acetone at a volume of 100,000 liters. 

 The Ostwald dilution law was tested as to its applicability to the results in acetone, 

 and found, for the more dilute solutions, to hold approximately. The molecular 

 conductivity was multiplied by the viscosity to see whether the product is, or is 

 not, a constant. It was found that the product is practically constant and has the 

 value of 0.70. This agrees with Walden's value in many organic solvents. In 

 mixtures of water with acetone this value ranges from 0.60 to unity. The deviation 

 of the value from 0.70 is probably due to the fact that complete dissociation was not 

 always reached in acetone. 



The molecular weight of lithium acetate in acetone was then determined by the 

 boiling-point method, using the apparatus devised by Jones. 1 It was found for the 

 most dilute solution that could be studied by this method that the molecular weight 

 was greater than the theoretical for the simplest chemical formula. This shows that 

 even in such solutions there is considerable polymerization of the molecules. This 

 is in keeping with the earlier work of Jones, in which he determined the molecular 

 weights of certain salts dissolved in acetone, and found that they were always 

 polymerized by this solvent. 



Cadmium iodide, which had also been found by Jones to be polymerized in ace- 

 tone, was brought within the scope of this work. Although cadmium iodide is so 

 little dissociated in water, we seem to have reached very nearly the valye of fx rj0 in 

 acetone. The Ostwald dilution law holds roughly for dilute solutions of cadmium 

 iodide in acetone, showing that fjL x was nearly attained. This, however, required a 

 dilution of several hundred thousand liters. 



The dissociation of cadmium iodide in acetone is much greater at zero than at 

 25. From our results it seems that the dissociation would probably be com- 

 plete at about 600,000 to 700,000 liters. At this dilution the curves of molecular 

 conductivity would become similar to those of fluidity, molecular conductivity 

 and viscosity being inversely proportional to one another. The product of the two 

 would then be a constant. 



RESULTS OBTAINED BY SCHMIDT. 



The work up to this point had had to do with solvents like the alcohols and ace- 

 tone, which have viscosities of the same order of magnitude as water. The relations 

 already discussed in this and in the preceding monograph 2 dealing with this same 

 general subject had then been worked out for solvents having comparable viscosities. 



J Amer. Chem. Journ., 19, 581 (1897). 5 Carnegie Institution of Washington Publication No. 80. 



