CHEMISTRY: D. A. M'INNES 
529 
This result indicates, at any rate in the case of solutions of concentra- 
tions up to 0.03 normal, the correctness of the two assumptions in- 
volved in the calculations, namely, that the osmotic work of transferring 
equivalent quantities of the ions is nearly the same for positive ion and 
negative ions, and that the osmotic work of transferring chloride-ion 
from one salt concentration to another is nearly the same whether the 
cation is potassium, sodium, or hydrogen. It is probable that the 
first of these assumptions is substantially correct up to much higher 
concentrations than it can be tested by the method here employed, since 
this method involves also the second assumption just stated, which 
doubtless becomes inexact at higher concentrations. Equation (1) 
here considered must in any case give a much more exact value of the 
liquid potential than does the Nemst equation 
RT(l-2n.) 
F C2 
How much the values calculated by the latter differ from those cal- 
culated by the former equation in the case of the more concentrated of 
these dilute solutions is shown in the following table: 
LIQUID POTENTIAL IN MILLIVOLTS 
SUBSTANCE 
MOLS PER LITER 
DIFFERENCE 
By equation (1) 
By Nernst equation 
HCl 
0.01665 
0.001665 
-36.94 
-38.03 
1.09 
NaCl 
0.01673 
0.001673 
11.46 
11.72 
0.26 
KCl 
0.01670 
0.001674 
00.44 
00.45 
0.01 
HCl 
0.03330 
0.003329 
-36.64 
-38.23 
1.59 
KCl 
0.03347 
0.003347 
0.43 
0.45 
0.02 
The difference between the values for the liquid junction potential as 
calculated by the two methods is small for salts having ions with nearly 
the same mobility, but the difference increases rapidly with increased 
concentrations of the solutions, since the Nernst equation contains the 
assumption that the ions are 'perfect' solutes, an assumption which is 
not even approximately true at higher concentrations if the ion con- 
