Equilibrium in Salt- Solutions. 153 



formed to the partial or total exclusion of other salts contain- 

 ing either of its constituents. The quantities, therefore, with 

 which we have to do are comparatively small, and depend on 

 the different solubility of the various salts in the way I have 

 described in the paper cited above. 



My method of experiment is as follows: — 



Having prepared solutions of all four salts of the same mole- 

 cular composition, I determine their specific gravities, and 

 from these calculate the molecular volumes. Representing 

 these volumes by a, b, c, cl, we have 



(1) MR+M / R / =a + 5=A. 



(2) MR' + M'B=c + d:=B. 



A and B are the molecular volumes of the first and second 

 couples of salts respectively if no change of volume occur on 

 mixing the two solutions. Next, the molecular volumes of 

 (1) and (2) after mixture are determined, and found to be V, 

 the same in each case. In all my experiments I have met with 

 no case in which A=B; nor, if the solutions be not too 

 concentrated, any in which V is greater than A or B. If 

 now V < A and A < B, then the/e can be none of B, the couple 

 with the larger molecular volume, present in the solution, and 

 of course, conversely, V < B and B < A, then no A is present. 

 The fact that the molecular volume V is the same in each case, 

 shows that rearrangement has taken place in one case or the 

 other, while V, being less than the lesser of the two theoretical 

 volumes, appears to me to prov^e the entire absence of the 

 couple which has the larger molecular volume. 



Table I. contains the results I have obtained with various 

 salts in solutions of different strengths. These solutions con- 

 tained almost exactly the correct quantities of the salts; and 

 that this was the case may be seen by the very close agree- 

 ment of the molecular volumes of the mixtures, although the 

 number of weighmgs for each set of four salts amounted to at 

 least thirty. In Table II., on the other hand, are given the 

 results of similar experiments with solutions not so accurately 

 prepared: as a necessary consequence the molecular volumes of 

 the two mixtures do not agree so closely as those in Table I.; 

 I but in the cases given such disagreement is of no importance, 

 1 as the differences between A and V, or B and V, are too great 

 to be appreciably affected by the differences in concentra- 

 1 tion of the solutions employed; it is only in cases such as the 

 : more dilute solutions in Table I. that this can become a serious 

 source of error. 



