BY MIXTURES OF ELECTROLYTES.—BARNES. 1H 
It is difficult to estimate the limit of error of the above 
observations. ‘lhe observed values are in all cases means of at 
least three observations, which were found to differ from their 
mean values in different cases by very different amounts up to 
0.001 degree. There are also many sources of error in the 
calculations and they do not admit of exact valuation. As a 
rough estimate the limit of error due to both observation and 
calculation may probably be put at 0.0015 degree. 
If this estimate is approximately correct, the above table 
shows that the agreement between the observed and calculated 
values is very satisfactory for both mixtures of potassium 
chloride and sodium chloride, and of sodium chloride and hydro- 
chloric acid. 
Mixtures of Solutions of Three Electrolytes. 
In the case of mixtures of the three electrolytes used as 
equal volumes of the simple solutions were mixed, and there 
was no change of volume on mixing, and as each molecule of 
these electrolytes breaks down into two ions, expression (2) 
becomes 
A=}[M, n, (1+a@,) + M, 2, (1+a@,)+M, , (1+@,)] .... (5) 
where the n’s are the concentrations of the constituent solutions, 
Thus in any mixture the n’s are known, the M’s can be obtained 
as above, and the a’s can be determined by the method given 
below. 
Determination of the Ionization Coefficients in Mixtures of 
Three Electrolytes. 
1 Professor MacGregor has shown how to obtain equations 
sufficient for finding the ionization coefficients in a mixture of 
any number of electrolytes having a common ion, and how to 
solve them by a graphical procedure. As in the case of mixtures 
of two electrolytes,* I have, in the present case also, transformed 
1 Trans. Roy. Soc. Can. (2), 2, 69, 1896-97. 
2 Trans. N.S. Inst. Sci., 10, 124, 1899-1900, 
