NOYES AND COOLIDGI']. ELECTRICAL CONDUCTIVITY. 197 



spends to the observed cell-constant ratio. This volume was obtained 

 by interpolation from a plot made as described in Section VI. The 

 actual volume occupied by the solution at the higher temperature is 

 greater than this by an amount equal to the expansion of the bomb upon 

 heating from 135° to that temperature. As there stated, the temper- 

 ature-coefficient of volume expansion of the steel shell of the bomb is 

 assumed to be 0.000045 per degree. 



The seventh column gives the specific volume of the solution at the 

 temperature of observation. It is obtained by dividing the values of the 

 preceding column, after correcting them for the expansion of the bomb 

 as just described, by the weight of solution given in the fifth column. 



The last column gives the ratio of the specific volume at the round 

 temperatures 218°, 281°, and 30G°, to that of the same solution at 4°. 

 Thus, this ratio shows the volume occupied by that quantity of solution 

 which at 4° has a volume of 1 ccm. The values are obtained from those 

 of the preceding column by reducing them to these temperatures by 

 means of the temperature-coefficient obtained from our specific-volume 

 values, and then dividing the results by the specific volumes of the 

 solutions at 4°. These specific volumes are as follows: 0.9958 for 0.1 

 normal, and 0.999G for 0.01 normal sodium chloride; and 0.9954 for 0.1 

 normal, and 0.9995 for 0.01 normal potassium chloride.* 



X. The Results of the Specific-Volume Determinations. 



The final results are brought together in the following table. The 

 value at 140° is that found by llirn f for pure water reduced from the 

 higher pressure which he employed to the vapor-pressure. 



The re:^ults with the 0.002 normal solution may be regarded as com- 

 pletely identical with those that would be obtained with pure water ; for 

 this solution contains only about 0.01 per cent of salt; and, moreover, 

 the experiments themselves show that there is no difference between the 

 specific-volume ratio of the 0.002 and 0.01 normal solutions, and that the 

 difference between the latter and that of the 0.1 normal solution is some- 

 what less than one per cent, which indicates that the order of magnitude 

 of the difference between pure water and the 0.002 normal solution is 



* Tliese values were computerl from the densities given by Kohlrausch and Hall- 

 waclis (AVied. Ann., 50, 122, 1893) for NaCl at 18°, and from that given by Kolil- 

 rauscli (Leitvermogen der Elektrolyte, 76) for a normal KCl solution at 18°, under 

 tlie assumptions that the change in density is proportional to tlie concentration and 

 that the expansion is the same between 4° and 18° for these solutions as for water. 



t Hirn, Ann. chim. phys. (4) 10, 32 (1867). 



