Forces on the Osmotic Pressure of Electrolytes, 751 



should expect B to be modified i£ the dissociation were not 

 complete. Evidently B has the better agreement. 



The upper curves in fig. 3 represent the results for the 



Fig. 3. 



+ NaNO, <• 



09 10 [% /+ 



Concentration (qm-nuA./Utre). 



IT 



nitrates of* NH 4 , Na, K, Rb, and Cs from Loomis 5 and 

 Biltz's * measurements. There is again a fair agreement 

 with the theoretical curve, although the possibility of a 

 small amount of association in the last three salts is not 

 absolutely excluded. The freezing-points of other mono- 

 valent salts have not been measured with sufficient accuracy 

 at dilute concentrations to serve as a strict test, but as far 

 as they go they support the theory. In the bottom curve 

 of fig. 3, all the lowerings of LiCl, RbCl, CsCl, KBr, 

 KI, KF, KCN, and LiN0 3 , given in Landolt and B6rnstein'*s 

 Tables, are plotted along with the theoretical curve. It is 

 striking that in all these salts the lowerings are of the 

 same order of magnitude, and this the same as that 

 theoretical curve for complete dissociation. Until more 

 measurements have been made it is difficult to say whether 

 the differences are due to experimental error or to the action 

 of subsidiary causes other than partial association. From 

 the point of view of the theory here considered, partial 

 association as an explanation is excluded by the fact that 

 most of the experimental points lie above instead of below 

 the theoretical curve. 



* Biltz, ZeiU Phys. Chem. xl. p. 185 (1902). 



