74 Dr. W. W. J. Nicol on the 



calculated solubility is satisfactory ; I need hardly add that 



but a slight alteration in the value of y would largely affect 



8 



x- that is the calculated solubility, while it would have little 



influence on the value of 



M.V. = 1800 + na + n 2 ^-n d y. 



The theoretical bearing of the above is still more interesting. 

 Saturation ensues at the apex of the parabola when the value 

 of a + nfi — n 2 <y is practically the same for any two values of n 

 which lie closely together ; that is, at a point at which the 

 farther addition of a small quantity of salt would be unattended 

 by change in the mean molecular volume ; but as a positive 

 change is produced by cohesion, and a negative change by 

 adhesion, it is evident that at this point the cohesion and adhe- 

 sion are in equilibrium, as I pointed out three years ago. 



The point at which saturation would theoretically take place 

 can be calculated as follows : — 



On the supposition that the molecular volume with which 

 the last molecule enters into solution is the same as the mean 

 molecular volume of the molecules already present, then 



(na + n 2 j3-n 3 y)-[(n-l)a + (n-iyi3-{n-l) 3 v'] 



— a-{-n{3—n 2 y y 



when ., /S + 7 



n = l, or n=^~. 



The value n=l is of course beside the question, but the other 



solution shows that saturation is complete when n = ■ _ ' . 



B 

 Experiment gives n = ^- approximately. 



The last columns in Tables I., II., III., and IV. contain the 

 values for the first half of the above equation and will make 

 this clearer. 



Section II. 



Saturation in a Solution containing two Salts. 



In the former paper I described experiments on the mole- 

 cular volumes of saturated solutions of two salts containing 

 the same base or the same salt radical. I showed that in 

 three cases out of the four examined, the resulting solution 

 possessed a volume equal to that calculated on the supposition 



