Lerrs anp Bhake—On Atmospheric Carbonic Anhydride. 441 
by raising or lowering the mercury reservoir to the proper level. 
When detaching the stock bottle of potash solution the rubber 
juncture of C is pinched between the finger and thumb, and then 
plugged with a piece of glass rod so that C remains full of liquid. 
Collecting and Absorbing Vessels.—The vessels we employed for 
this purpose in our test experiments were ordinary 6-litre flasks, 
fitted either with ordinary corks, well soaked in hot paraffin wax, or 
india-rubber corks; but in either case a layer of paraffin wax, 
about {th inch in thickness, was melted on to their lower surface 
Two tubes passed through the cork attached to each receiver, a 
long one reaching about three-quarters of the way down the vessel, 
and a short one which terminated flush with the lower surface 
of the cork. The upper ends of these two tubes were plugged, 
when necessary, with short pieces of india-rubber tube attached to 
glass rods. But for use by a scientific expedition, we have had 
a different vessel constructed which we shall describe presently. 
Apparatus for Liberating and Measuring the Absorbed Carbonic 
Anhydride.—For this purpose we employed the excellent and reli- 
able apparatus described by Dr. W. E. Adeney in the Transactions 
of the Royal Dublin Society, Vol. 5 (series 2), Part 11 [1895], 
p. 548, and used by him in his important research on the dissolved 
gases in natural and polluted waters—slightly modified by our- 
selves. Constant experience of this apparatus extending over some 
years has proved to us that it is most serviceable for boiling out 
the dissolved gases from a liquid im vacuo, and yields very accurate 
results in their subsequent analysis. ‘The error in the latter is 
probably less than 0:01 c.c. 
The Process.—The absorbing vessel is first filled with the air- 
sample to be examined; and in order to do this, the method of 
exhaustion is preferable, the operator being at a sufficient distance 
from the collecting area to avoid the possibility of breath containi- 
nation. An exhausting syringe is attached to one end of an 
india-rubber tube about six feet long, while the other end of 
this tube is connected with either of the glass tubes passing 
through the cork of the receiver. The syringe is set in motion, 
and pumping continued until 20-30 litres of air have passed 
through the receiver. The india-rubber tube is then detached, 
and the receiver plugged, a note being taken of the readings of 
the barometer and of the wet and dry bulb thermometers. 
