221 
Still stronger solutions were used with this standard, but the 
discrepancies between the true and the calculated quantities 
were considerable. Standard solution 5-8 in 500 cc. of 
water, length of column, 8 '3. 
ABC 
7-2 5-7 8-4 
Some experiments were made with solutions of perman- 
ganate of potash, using a shorter column for the standard 
solution. Standard solution 2d in 2 1 4*6 cc., length of 
column 3*8. 
ABC 
3-6 2-2 4-1 
An experiment was also tried with a still weaker solution. 
Standard solution 1*4 in 214 , 6 cc. of water, length of 
column 3 ’8. 
ABC 
2-4 2-6 2-1 
Some experiments were also made by looking through 
the column of fluid at a white surface external to the 
cylinder. Standard solution 12 in 500 cc., length of 
column 8 ’4. 
ABC 
14-4 6-5 15*5 
28-8 3d 32-5 
16-8 5-3 19 
I prepared a fresh standard solution of the permanganate 
and compared the last solution with it. The new standard 
was similar to the last in composition. 
ABC 
16-8 5*7 17-6 
Thus the result was somewhat more favourable with a 
newly-prepared solution. I used the permanganate on ac- 
count of its great tinctorial power, but it is a salt of consider- 
able instability, especially in contact with organic matter. 
The water used in the experiments was obtained by dis- 
tilling Manchester town’s water from a copper still lined 
with tin, and condensing in a tin condenser. Now such 
