184 



Mr. W. W. J. Mcol on the Molecular 



Now, treating these results as the former ones were treated, 

 we find that the replacement of potassium by sodium in its 

 salts produces a diminution in volume, as shown in Table III. 



Table III. 



No. in 

 Table II. 



Had. 



n. 



w(K Had.). 



^(NaBad.). 



Diff. 



A- Diff - 



— n 



Aq. 



1-15. 



CI 



1 



1827-7 



1817-5 



10-2 



10-2 



0-0 



2-16. 



J5T0 3 



- 2 



1879-6 



1858-6 



21-0 



10-5 



0-0 



3-17. 



OH 



2 



1815-5 



1795-4 



201 



101 



5-3| 



4-18. 



C10 3 



1 



1846-5 



1835-9 



106 



106 



0-2 



5-19. 



so 4 



2 



1 



1818-8 



1808-3 



105 



105 



0-10 



6-20. 



A 



5 



2062-1 



2011-5 



50-6 



101 



0-3 



7-21. 



„ 



2 



1902-4 



1881-6 



20-8 



10-4 





8-22. 



, ; 



1 



1850-8 



1840-1 



10-7 



10-7 





9-23. 



Br 



1 



1835-1 



1824-8 



10-3 



10-3 



0-2 



10-24. 





 2 



2 



7246-1 



7225-1 



210 



105 





11-25. 



co 3 



2 



2 



3618-4 



3597-6 



20-8 



104 



H-10 



12-26. 



T 



2 



2 



3680-5 



3659-9 



20-6 



10-3 





13-27. 



H 2 P0 4 



X 



5442-2 



5431-2 



110 



11-0 



0-4 



Mean . . . 



A = 



10-4 



In 13 cases, including 11 different pairs cf salts, the mean 

 value of (K— Na)Il is 10*4, the extreme variations being 

 — *3 and + *6, or, omitting H 2 P0 4 , as the sodium salt was 

 not quite pure, we have A = 10*4 + '3, a difference readily 

 produced by the different solubility of the various salts. For, 

 as has been shown*, this has a marked influence, owing to the 

 attraction of the salt-molecules for one another. In order to 

 avoid this source of error, it would be necessary to work with 

 much more dilute solutions, in which ease the experimental 

 errors would be increased ; as it is, I believe that I am fully 

 justified in neglecting this difference and in speaking of the 

 value of A in sufficiently dilute solutions as a constant : thus 

 my former results are fully borne out. 



A glance at the last column of Table III. will show that 



* Phil. Mag. August 1883. 



