LE ROY CG. COOLEY. 85 
wat 0.0000018198 (3) 
in which @ is the value sought for, viz.: the actual weight 
of one of the particles into whichasolid coloring substance 
may be divided when the eye is able to detect the faint- 
est tint of its color in solution. The values of 2, 2 and 
v are to be determined by experiment for each different 
substance: the ‘‘limit of visibility for mass’’ of these 
can then be calculated. 
The experimental work has been carried out for several 
substances by means of an observation tank, rectangular 
in form, measuring about 20 cm long, about 15cm high, 
and exactly 4 cm inside between its walls, which consist 
of colorless plate glass. A vertical partition between 
these walls divides the tank into two exactly similar com- 
partments. <A sheet of ground glass is fixed against the 
outside of one wall. A carefully measured volume of 
pure water is placed in one compartment and about the 
same quantity in the other. The tank is placed in 
front of a window looking northward, and a standard so- 
lution of the coloring matter is then slowly run from a 
burette into the compartment containing the known vol- 
ume of water. The light from a clear northern sky, soft- 
ened and diffused by the ground glass, traverses the 
liquid and enters the eye. With the pure water in the 
other compartment always in view for comparison, the 
observer is able to judge accurately when the water be- 
gins to show tint by the addition of the colored solution. 
To illustrate the process and the calculations, the fol- 
lowing description of the experiment with potassium 
chromate will suffice. 
Thirty milligrams of the salt were carefully weighed, 
dissolved in pure water, and the solution made up to ex- 
actly 300 cc. Two hundred cubic centimeters of pure 
water were measured into one compartment of the obser- 
vation tank, and the other compartment was also filled 
35 
