DISCUSSION OF RESULTS. 67 



the phenomenon does exhibit itself in the other mixtures, and this is because 

 of the high conductivities in methyl alcohol, these being high on account 

 of the small viscosity of methyl alcohol and its relatively great dissociating 

 power. 



Considering all of the salts in the various mixtures, it is seen that, in general, 

 the effect of increased viscosity on conductivity is greatest in that mixture 

 in which the minimum in conductivity occurs. In some cases the maximum 

 effect occurs elsewhere. For example, for lithium nitrate (table 40) at 25 

 the maximum effect is in the 75 per cent mixture, while the minimum in 

 conductivity occurs for the most part in the 50 per cent mixture. 



The explanation of this is found in the fact that, although in the one 

 mixture the effect is greater, it is offset by the effect of the smaller fluidity of 

 the other mixture that of the 75 per cent mixture being 81.8, and that of the 

 50 per cent mixture being only 68.9. In cases where the minimum shifts 

 with increase in dilution, the probable explanation is to be found in variation 

 in increase of dissociation accompanying further dilution. 



The result of variation in the temperature is also shown in the tables 

 particularly in the case of the lithium salt (table 40). At the maximum 

 effect is in the 50 per cent mixture; at 25, in the 75 per cent mixture. 



In table 43 is given a comparison of the temperature coefficients of fluidity 

 and of conductivity for various electrolytes in the various mixtures of methyl 

 alcohol and water, and also for ethyl alcohol mixtures, for which the data are 

 more meager. 



From table 43 it is evident that the temperature coefficients of conductivity 

 and fluidity do not differ markedly, particularly for some salts. For potas- 

 sium iodide they are most nearly equal in the 65 per cent mixture, differing 

 by only 7 per cent. For the other salts in the 50 per cent mixture the agree- 

 ment is closest, and the differences are in no cases greater than for potassium 

 iodide in the 65 per cent mixture. In some instances the agreement is seen 

 to be much closer (SrI 2 , CdI 2 ). 



For the ethyl alcohol mixture of 50 per cent, the temperature coefficients 

 differ to about the same degree as in the methyl alcohol mixtures. In other 



words, = constant. 

 <t> 



These facts are significant, as will appear from the latter part of this section. 



The conclusion which can be drawn is, then, that the decrease in conductivity 

 of electrolytes in binary mixtures of various alcohols and water, which is in some 

 cases accompanied by the minimum conductivity observed by Zelinsky and 

 Krapiwin, Jones and Lindsay, and ourselves, is caused primarily by a diminution 

 in the fluidity of the solvent, and a consequent decrease in ionic mobility, and 

 secondarily by the effect of one associated solvent on the association of another 

 such solvent. 



