1898-99.] Sulphuric Acid and Sulphates in Solution. 
505 
employed, the greatest “ weight ” must be attached to the first 
experiment in this series, the smallest weight to the last, or in 
other words, a small experimental error is exaggerated in calcula- 
tion very much more in the last experiment than in the first. 
The mean value of m was 0*85. Now, in the second series, the 
variation in C 1 (l - a 4 ) is small, while C 2 (l - a 2 ) varies consider- 
ably from the first to the last experiment. Such conditions are 
most favourable for the calculation of a new value for the exponent 
n. When this was done in the same manner as above, m having 
the value 0‘85, the value obtained for n was 0*89. It appears, 
therefore, then m and n are nearly equal, and if we put 
m = n = 0'87 the results of the experiments will be expressed 
with considerable accuracy. 
But having decided that m = n, their value can be calculated 
from each series independently according to the simplified expression, 
for we have 
and 
and therefore 
CjxCa^C*; 
. Ci * c, 
' 3 - Kj 
I? _ Cj x C n 
'ni“ k; 
x = log. 
c lX c 2 
Cl X C„ 
, c 3 
'-'nr 
The mean value of x obtained from the first three experiments 
of the first series was 1*28 ; from the second series 1T1 ; and from 
the third TO 7. The mean of these is IT 5. We would expect 
this to be nearly equal to that already found by the first method 
of calculation, and as it happens, it is exactly the same, for 
£c = -i-= -^==1*15. The expression from the equilibrium may 
now be written 
C.H 2 S0 4 (1 -cq) X 0.K 2 S0 4 (1 -a 2 ) = K{C.KHS0 4 (l -ag)} 1 - 15 . 
By introducing into this formula the corresponding concentra- 
tions of the undissociated portions of free acid, neutral and acid 
sulphates given on p. 503, the value of K for each experiment was 
VOL. XXII. 21/4/99. 2 K 
