142 
than this experiment had yielded might be obtained. The 
excesses of ammonia used in the experiments were nearly 
proportional to the quantities of sulphate of copper in solu- 
tion ; but if we regard water as an agent whose tendency is 
to diminish the intensity of the colour, and ammonia as an 
agent whose tendency is to restore the colour, it would 
seem reasonable that the ammonia should be proportional 
to the water. The difference of the excesses of ammonia 
in the last two solutions was large, being 22 cub. c. I pre- 
pared fresh solutions, one containing 4 cubic c. of the copper 
solution with 30 cub. c. of additional ammonia and suffi- 
cient water to make 500 cub. c. The other solution con- 
tained 15 cub. c. of the copper solution with 30 cub. c. of 
additional ammonia, and sufficient water to make 500 cub. 
c. The quantities of the copper solution taken should cor- 
respond to 1600 and 6000 of copper sulphate ; to guard 
against imperfect measurements from the burette; I also 
weighed the solutions. The 4 cubic centimetres weighed 
3‘9854 grams, and the 15 cubic centimetres weighed 14-99 
grams. The ratio of the volumes is 375, and the ratio of 
the weights is 3-761, so that the error of measurement 
would be but small. 
With discs outside the results of experiments were as 
follows. Standard solution 1600 in 500 cub. c. of water, 
length of column 21*2. 
A B C 
6000 6 5653 
The number under B was the result of eight trials ; also the 
standard solution was on the left hand. With the standard 
solution on the right the results were 
ABC 
6000 5-4 6283 
The number under B was the mean result of eight trials. 
In one case the value got by experiment is too high, and 
in the other too low. The theoretical length is 5-65, and 
