196 
neglected to do so. The tints so obtained were in every 
case satisfactory, until I reached Experiment XIV. In this 
and in the succeeding experiments I thought that columns 
of the theoretical lengths gave tints slightly lighter than 
the standard diffusion. On some occasions I have wavered 
in my opinion that the theoretical column gave too light a 
tint in the fourteenth experiment. At another time I got 
the following result : — 
Expt. XVII. — Comparison diffusion contains 9 '6 cub. c. 
in 500 cub. c. 
A B C D 
2-8 2-95 2-87 2*65 
Here again the number under C is a little greater than that 
under D. I afterwards repeated the experiment with this 
variation — that the standard diffusion was contained in the 
comparison cylinder, and the comparison diffusion in the 
standard cylinder. 
Expt. XYIII. — Standard diffusion contains 1*2 cub. c. in 
500 cub. c. ; length of column, 22'6. 
Comparison diffusion contains 9 *6 cub, c. in 500 cub. c. 
A B c D 
2-85 
3-07 
2-96 
2-82 
On another occasion I got — 
A 
B 
c 
D 
2'85 
815 
8-0 
2-82 
Here again there still remains a tendency to make the 
column a little longer than the theoretical. 
I repeated the experiments in which the strength of the 
comparison diffusion is several times a multiple of the 
strength of the standard diffusion. The strength of the 
standard diffusion was the same as in the previous experi- 
ments, and the length of the column was 21*2. 
Expt. XIX. — Comparison diffusion contains 9 '6 cub. c. in 
500 cub. c. 
A B C D 
2-63 2-98 2-8 2-65 
When I actually tried the theoretical column I thought 
that tlie tint was perhaps slightly lighter. 
