WORK OF C. M. STINE. 13 



mass action, 707 chlorine ions should be driven back to re-form calcium chloride to 

 958 for potassium chloride. But since it requires two ions of chlorine to re-form each 

 molecule of calcium chloride, only 354 molecules of calcium chloride will be re- 

 formed to every 958 molecules of potassium chloride. Multiplying these values by 

 fj,oo for calcium chloride and potassium chloride, respectively, we get 5,222, approx- 

 imately, for calcium chloride and 7,695 for potassium chloride. These values repre- 

 sent the relative losses in conductivity upon mixing. From this the decrease in 

 conductivity can be apportioned between the two gaits. 



All conductivity measurements are given as specific conductivities in the mixtures. 

 The units employed are reciprocal centimeter-ohms. For purposes of comparison 

 with the previous work upon the "hydrate theory" our units may be transformed 

 by dividing /z by 1.069. 



The approximate composition of the hydrates in the single solutions was calcu- 

 lated by the method previously employed by Jones and his coworkers. 



EXPERIMENTAL WORK. 



Freezing-point. The more concentrated solutions were frozen by means of a 

 mixture of solid carbon dioxide and alcohol, the freezing temperature being deter- 

 mined by means of alcohol thermometers. These temperatures could not be meas- 

 ured more closely than 0.5. For the more dilute solutions freezing-mixtures of ice 

 and sodium chloride, or ice and crystallized calcium chloride, were employed. The 

 temperature was read by means of Beckmann thermometers, covering ranges of 

 6, 12, and 25, respectively. When the more concentrated solutions were frozen 

 they were removed from the bath of solid carbon dioxide and alcohol as soon as ice 

 began to separate, in order that they might not be surrounded by the very cold 

 freezing-mixture while temperature equilibrium was being established. The neces- 

 sary correction for the separation of ice and consequent increase in concentration was 

 introduced in every instance. The separation of some of the dissolved substance 

 was readily ascertained by the grit or sand-like character of the solid that separated, 

 which was easily detected. Since it was found that, for the volume of liquid em- 

 ployed in the freezing-point tube (about 25 c.c), a freezing-point determination 

 made with a large undercooling and, consequently, a copious separation of ice, gave 

 a slightly different result from one where the undercooling was less; care being taken, 

 in the freezing-point determinations of both the pure solvent and the solutions, to 

 keep the undercooling nearly the same throughout about a degree. This necessi- 

 tated inoculating the solution, in nearly every case, with an exceedingly minute 

 particle of the solid phase of the solvent. 



CONDUCTIVITY. 



Conductivities were measured by the method of Kohlrausch. 



The resistance coils, manufactured by Leeds & Co., were tested and found in no 

 case to involve a greater error than 0.04 per cent. 



The dilute solutions were measured in conductivity cells of the form employed 

 by Jones and Bingham, 1 the tubes carrying the platinum electrodes being sealed 



lAmer. Chem. Journ., 34, 481 (1905). 



