476 



Dr. Harry C. Jones on the 



Na*CO,= 106-12. 



1. 



2. 



3. 



4. 



5. 



Normal. 



Lowering. 



Gram-molec. 

 Lowering. 



i. 



Dissociation. 





o 







per cent. 



0002 



00111 



5-5500 



2-937 



96-8 



0-004 



0-0222 



5-5500 



2-937 



96-8 



0-006 



0-0331 



5-5166 



2-919 



95-9 



0-008 



0-0440 



5-5000 



2-910 



95-5 



001 



0-0546 



5-4600 



2-889 



94-4 



02 



0-1069 



5-3450 



2-828 



91-4 



0-04 



0-1968 



4-9200 



2-603 



80-1 



0-06 



0-2769 



4-6150 



2-442 



72-1 



0-08 



0-3625 



4-5312 



2-397 



69-8 



o-i 



0-4464 



4-4640 



2-362 



68-1 



Discussion of the Eesults. 



The chlorides of potassium, sodium, and ammonium have 

 been found to give the same dissociation within the limits of 

 experimental errors. The same fact was found by Noyes* 

 from his " Solubility " investigations. Two complete series 

 of measurements have been made with solutions of sodium 

 chloride, and the mean results are given as the most exact for 

 the three chlorides. A number of solutions of potassium and 

 ammonium chlorides have, however, been measured f. The 

 curve from the results with sodium chloride (PL IV. curves 1, 2) 

 can also be regarded as expressing the results in all three 

 cases. The dissociation of the chlorides and other salts de- 

 creases more or less regularly with the increase in concen- 

 tration. Certain apparent exceptions will be noticed here 

 and there, especially in the dilute solutions. Thus the dis- 

 sociation of 0*004 n sodium chloride is 93*1 per cent., while 

 the dissociation of 0'006 n is 93*7 per cent. This is caused 

 by experimental errors, which are much greater for the dilute 

 solutions, because the lowering of the freezing-point here is 

 the least. An error of 0*0001° at 0*001 n solution will 

 affect the dissociation some per cent. This error, however, 

 diminishes very rapidly with increase in concentration, and 

 could be entirely disregarded at 0*1 n concentration. 



As evidence that these small deviations are caused by ex- 

 perimental errors, when two series of measurements are made 

 with solutions of the same salt, they do not appear constant 

 to the two series. 



These three alkali chlorides dissociate into two ions each, 





* Noyes, Zeits. pliys. Chem. ix. p. 



f Jones, ibid. xi. p. 114. 



