7 
being applied as a correction to the other experiment, gave 
consistent results. 
If we write p in the form k p , the formula to be used may 
be written (lc p k^) l ~ (k p k q, )l'. This leads to the relation 
The standard solution contained 1*2 cub. c. of a strong 
carbon diffusion in 500 cub. c. of water, and the length of 
column was 21 ‘2. Comparing with this a solution contain- 
ing 9*6 cub. c. in 500 cub. c. of water, on one occasion I 
made 2*94 the equivalent column, on another occasion 2*87, 
and on a third occasion 2*8. The mean of these three 
results is 2 -87. Hence the corrected formula will be 
(Q + 0\L14)Z = (Q' + 0*114)Z' 
Q and Q' denoting the number of cubic centimetres of the 
strong diffusion added to each cylinder. In the following 
table I have recalculated the results given on page 197, 
Yol. XIX. of the Proceedings : 
A 
B 
C 
2-02 
1*92 
.... 177 
1*59 1-44 1-32 
1-22 115 1-06 
1-05 
0*96 
0-88 
089 ..... 
.... 0-82 
076 
078 
.... 0-72 
0*66 
A denotes length of column by experiment, B denotes 
theoretical length calculated by corrected formula, C the 
theoretical length deduced from uncorrected formula. It 
will be noticed that the discrepancies between A and B are 
less than those between A and C. 
“Note on the Colour Relations of Nickel, Cobalt, and 
Copper,” by James Bottomley, D.Sc., F.C.S. 
