174 WORK OF J. SAM GUY. 



in temperature; hence, its power to dissociate an electrolyte into its ions has been 

 diminished. It is, however, true that the theory of Dutoit and Aston is only an 

 approximation. 



The decrease in velocity of the ions with rise in temperature must then be the one 

 conditioning cause of increase in conductivity. This change in velocity of the ions 

 may be due to either or to both of the following causes: First, change in the viscosity 

 of the medium through which the ions move; second, as Jones 1 and his coworkers 

 have shown, to the change in complexity of the solvates which surround the ions. 



In no other solvent is the change in conductivity with change in temperature so 

 pronounced as in the one which chiefly concerns this investigation, viz, glycerol. 

 The chief cause of this change is largely the change in the viscosity of the solution, 

 while we believe that there is some evidence brought out in this investigation that 

 indicates the presence of glycerolates. 



Table 115 gives the molecular conductivities at 25, 35, and 45 of all the elec- 

 trolytes which we have studied in pure glycerol as a solvent. It is seen that in all 

 cases the values for ju are extremely small, but show, in general, a regular increase, 

 both with increased dilution and with rise in temperature. 



Associated with the conductivity of every substance are the temperature coeffi- 

 cients of conductivity, both in per cent and in conductivity units. Since the latter 

 show the actual increase in conductivity per degree rise in temperature, a discussion 

 of these data will bring out the most interesting points of this part of the work. 



Although the temperature coefficients of conductivity, when expressed in con- 

 ductivity units, show, in general, a regular increase with increased dilution, this is 

 much more marked with ternary than Math binary electrolytes. This fact has been 

 observed by Jones for aqueous solutions in a discussion of the Avork of West. 2 



Results of the present investigation show that in gtycerol the temperature coeffi- 

 cients of conductivity of any given substance, at high dilution, are larger than at 

 lower dilution, and that the relative increase is greater with salts of barium, stron- 

 tium, calcium, and cobalt than with salts of sodium, potassium, and ammonium. 

 These facts may be explained in terms of the theory of solvation. That solvation 

 takes place in aqueous solution has been shown beyond reasonable doubt by Jones 

 and his coworkers; and, indeed, Jones and Strong have obtained abundant spectro- 

 scopic evidence for solvates in glycerol as a solvent. 



If there is solvation, then, according to the mass law, in the more dilute solutions, 

 where the amount of solvate per ion is greatest, we should expect to find the most 

 complex solvates. Any change in temperature would produce the greatest effect 

 where the solvation was greatest, that is, in the most dilute solutions. Again, this 

 change in solvation should be more apparent in those salts which have the greater 

 power of combining with the solvent, or, in the case of water, with those salts that 

 have the largest number of molecules of water of crystallization. 



We can not, of course, say that salts of barium, strontium, calcium, and cobalt 

 possess a power of combining with glycerol similar to that which they manifest 

 towards water, but it is not surprising to find solvation more marked with these 

 salts than with salts that have very slight hydrating power, such as the salts of 

 sodium, potassium, and ammonium. 



lAmer. Chem. Journ., 35, 445 (1906). mid., 34, 357 (1905). 



