422 BOTANICAL GAZETTE [DECEMBER 
energy. After numerous preliminary tests this possibility was 
achieved. 
The porous cups are now furnished in a dark-colored claves 
deep, grayish brown, and these cups show a marked increase in 
light-absorbing power over the ordinary white ones. 
3. Another light-absorbing cup was obtained by coating the 
white form with a thin layer of washed lampblack. Common 
lampblack is boiled in distilled water, allowed to cool and settle, 
and the water decanted as well as possible. This process is repeated 
three to five times, and furnishes a clean, insoluble, and impalpable 
black powder, in the form of an aqueous paste. The latter may be 
diluted and applied to the cups with a small brush. This applica- 
tion should be made after the cup is filled and ready for operation, 
as the absorption of water by the surface when thus arranged is 
sufficient firmly and quickly to fix the carbon layer in place, and the 
latter is never allowed to become drier than it is destined to be in 
the actual operation of the instrument. The cup cannot be handled 
by the coated surface without injury, but it is a simple matter to 
renew the coating if such injury occurs. These coated cups operate 
in essentially the same manner as the permanently colored ones. 
used in these tests, the white, brown, and black cups were in- 
stalled on burettes, essentially as figured in Publ. No. 50 of the - 
Carnegie Institution. 
4. The black bulb thermometer im vacuo (the one used was 
obtained from the Kny-Scheerer Co., New York) is essentially 
an ordinary glass-mercury thermometer, the bulb of which is 
blackened and inclosed in a thin glass vacuum bulb. It is exposed 
to the light for a short time period and the rise of the temperature 
of the bulb noted as the reading. The instrument must be shaded 
and must come to air temperature between observations. It is 
seen that the light absorbed by this instrument is made to do the 
work of expanding the mercury, the amount of expansion occurring 
in a specified time being the measure of the energy absorbed. 
The devices for light estimation thus far mentioned all depend 
upon the heating effect produced by the absorbed light. Another 
group of instruments, all following the principle of the Bunsen- 
Roscoe “photometer,” depend upon the chemical effect produced 
