218 PROCEEDINGS OF THE AMERICAN ACADEMY. 



XVI. Summary. 



In this article has been described the construction of a platinum-lined 

 bomb of 124 ccm. capacity with electrodes insulated by quartz-crystal 

 cylinders, by means of which the conductivity and specific volume of 

 aqueous salt solutions can be determined with an accuracy of 0.2 or 

 0.3 per cent, at least up to a temperature of 306°. The methods 

 employed for overcoming the serious mechanical difficulties may be of 

 assistance in other investigations requiring similar apparatus; for 

 instance, in calorimetric bomb work. 



Measurements of the conductivity and specific volume have been made 

 upon solutions of sodium and potassium chlorides at concentrations 

 varying from 0.0005 to 0.1 normal, at the temperatures 140, 218, 281, 

 and 306°. 



The results obtained with these salts show that even at the highest 

 temperature the dissociation calculated from the conductivity ratio does 

 not change with the concentration in much closer accord with the re- 

 quirements of the Mass- Action Law than at the ordinary temperature. 

 The various empirical functions which have been proposed for the ex- 

 pression of the change of conductivity with the concentration were tested 

 as to their applicability to the results at these widely different tempera- 

 tures by a graphical method by which the effect of the uncertainty in 

 the conductivity values (A ) for zero concentration was eliminated. It 

 was found that that given by Kohlrausch, A — A = KC% was most 

 satisfactory, the deviations being scarcely greater than the possible ex- 

 perimental error. Except at the highest temperature, the function of 

 Barmwater, A — A = K A i C*, also gave good results. To make 

 applicable the general function A — A = K A n C" _1 (of which the 

 van't Hoff function A — A = A'A* C- is a special case), it is necessary 

 to vary the exponent n not only with the nature of the salt, but also 

 with the temperature. 



The conductivity values extrapolated for infinite dilution, and there- 

 fore the migration-velocities of the ions, were found to be an approxi- 

 mately linear function of the temperature, throughout the whole range 

 of temperature in the case of sodium chloride, and up to 218° in the 

 case of potassium chloride, the deviations being moreover not very large 

 (3.5 per cent) even at the highest temperatures with the latter salt. 

 The temperature-coefficients referred to the values at 18° are 3.05 per 

 cent for sodium chloride between 18 and 306°, and 2.74 per cent for 

 potassium chloride between 18 and 218°. 



