SUMMAKY. 169 



not form more complex aggregations when mixed than when unmixed. 

 Here, again, conductivity has been shown to be connected with fluidity. 



(6) Lithium bromide, however, gives a pronounced maximum in con- 

 ductivity, in mixtures of acetone with methyl or ethyl alcohol. The same 

 phenomenon was observed in the case of cobalt chloride in mixtures of ace- 

 tone with ethyl alcohol. We believe this maximum is due, primarily, to a 

 diminution in the dimensions of the atmosphere about the ions. 



(7) We think, also, that the changes in the size of the ionic spheres should 

 be considered as a factor in causing the conductivity minimum as well as the 

 maximum. 



(8) We have determined the temperature coefficients of conductivity and 

 fluidity, and found them to be of the same order of magnitude. Lithium 

 bromide, in the mixtures studied, showed, with rise in temperature, a large 

 increase in conductivity, due to increase in fluidity. The temperature co- 

 efficients of lithium bromide in the above mixtures are, therefore, all positive. 



(9) Cobalt chloride, however, in some of the acetone mixtures, at ordinary 

 temperatures, gave negative temperature coefficients. We think this is due not 

 only to the effect of the diminishing dissociation more than overcoming the 

 effect due to the increasing velocity of the ions, but also to the fact that the 

 solvates formed in these cases may be more stable at the higher temperature. 

 We should, therefore, expect the reaction which gave rise to the solvates to 

 be endothermic. 



(10) We have found for our substances, in a given mixture of solvents, a 

 dilution where the temperature coefficient of conductivity is practically zero. 

 This corresponds to the maximum in conductivity observed by other workers 

 at elevated temperatures. 



(11) We have shown that the temperature coefficients generally increase 

 with the dilution, not only in aqueous solutions, but also in the non-aqueous 

 solutions thus far studied. 



