414 Seientifie Proceedings, Royal Dublin Society. 
aud was found to work admirably. By means of a motor G con- 
nected by a strap to the cranked wheel F’, the hinged boards DD 
(provided with an india-rubber band E to act asa spring) were 
made to move continuously like an ordinary bellows, the move- 
ment actuating an india-rubber “spray” apparatus C placed 
between the boards, thus causing the air in the receiver A to 
circulate through the two tubes B and B’ containing weak baryta 
solution coloured with phenol-phthalein. The upper of these 
tubes (through which the air passed out of the receiver) served also 
to indicate when all the carbonic anhydride had been removed 
from the air—weak baryta solution (1 ¢.c. = 0:1 ¢.c. CO, at N.T.P.) 
being added from the burette H from time to time until the solu- 
tion remained of a faint pink colour for an hour or so. It was 
found that every trace of carbonic anhydride was removed from 
the air when the apparatus had been working for about six hours. 
In order to obtain an artificial mixture of the air thus purified 
with a known volume of carbonic anhydride, the receiver was 
detached from the two absorbing tubes, plugged (with the pre- 
caution already mentioned for preventing access or egress of air) 
and transferred to an apparatus for introducing a measured volume 
of carbonic anhydride (fig. 3). 
The tube A standing over mercury contained a supply of pure 
carbonic anhydride—obtained by heating sodium bicarbonate 
until a portion of the escaping gas was entirely absorbed by caustic 
potash solution—and dried by a layer of sulphuric acid in the 
tube itself. This tube was attached by an india-rubber junction 
at G, completely immersed in a mercury trap, to the measuring 
pipette B, accurately marked at its lower (capillary) part to contain 
a definite volume of gas. 
Previous to attaching the receiver of purified air, this measuring 
pipette was filled with mercury by raising the mercury reservoir 
D, closing the stop-cock of the manometer tube OC, and opening the 
three-way stop-cock H, so that F also was filled with mercury 
when the stop-cock was closed. The receiver of purified air 
clamped in an inverted position to a tall retort-stand (not shown in 
the figure) was next brought over F, and the latter attached to it - 
by the rubber junction iorming part of the plug of the short tube of 
the receiver. ‘The stop-cock of the manometer tube was now opened, 
the mercury reservoir lowered, and the stop-cock of the measuring 
