220 
be a function of the quantity of colouring matter only. 
Occasionally in chemistry we have to deal with quantities 
so minute that we can no longer use the balance for their 
determination, and yet those small quantities when they 
yield coloured solutions are evidently comparable among 
themselves. In a short communication which I made to 
this Society (vol. xv., 63), I suggested that such a process as 
above indicated might be applied for the measuring of such 
traces, assuming that the depth of the disc would be 
inversely as the quantity of colouring matter present. Some 
experiments were made on colouring matter in solution. 
Glass cylinders were used about lOin. high and 2in. dia- 
meter ; to one side of each cylinder was pasted a strip of 
paper marked with an arbitrary scale (each degree of the 
scale I afterwards found to be 27mm.), the white surfaces 
used were crucible lids ; a piece of glass rod resting on one 
of the cylinders, having the string of the moveable disc 
attached to it, served as a windlass for raising or depressing 
it. To avoid the use of many ciphers I have taken 
•0001 grm. as the unit of measurement. In the following 
experiments column A denotes the amount of the colouring 
salt present, B the depth of the disc, and C the amount of 
the colouring salt thence derived by calculation. Standard 
solution 35-2 of permanganate of potash in 500 cc. of water 
depth of disc, 8 -3. 
ABC 
53 6-1 48 
70 4-1 71 
Experiments were made with still stronger solutions but 
the differences were considerable. 
Standard solution 24 of permanganate of potash in 500 cc. 
of water, depth of disc, 8’3. 
A 
B 
C 
36 
5-4 
37 
48 
3-6 
55 
60 
2-8 
71 
