82 LIMIT OF VISIBILITY FOR MINUTE MASSES. 
a sheet of white paper, by which the light was reflected 
upward through the plate glass bottoms. 
1. Fifty cubic centimeters of pure water was placed in 
each vessel, and three drops of a solution of aniline 
green were added to the contents of one. The green 
tint imparted to the water in this vessel could be detected 
and itsintensity judged by comparison with the pure water 
in the other, as they stood side by side, by looking down 
through the fluid upon the white paper below. Fifty cubic 
centimeters of water were then added to each vessel, and 
the tints again compared. No diminution of intensity 
could be detected in the green solution. Again 50 cc of 
water were added to both, and this was repeated until 
5x50 cc had been added, each time with the same result. 
No variation in the intensity of color could be detected, 
although at last the same quantity of coloring matter was 
dissolved in 300 cc of water as was at first dissolved in 
only one-sixth of that volume. 
2. The experiment was also varied as follows: Fifty 
cubic centimeters of water having been placed in each 
vessel, three drops of the aniline green solution were 
added to each instead of to one only as before. Pure 
water was then added to one only, and the colors of the 
two were compared as before after each addition. No 
variation could be detected ; 250 cc in oneand 50cc in the 
other contained the same mass of aniline green and were 
colored with equal intensity. 
These experiments confirm the inference that a given 
weight of coloring matter will produce color of the same 
intensity in any quantity of water, small or large, pro- 
vided only that it is held in a vessel whose cross section 
has a uniform area. And therefore our expression, = 
represents the least number of particles into which the 
mass m must be divided in order to impart color—the 
faintest that can be detected—to the volume 2’. 
B82 
