Lerrs anp Buaxe—On Atmospheric Carbonic Anhydride. 439 
by Mintz and Aubin during the French Transit of Venus Expe- 
pedition to Central and Southern America in 1882 was of this 
kind, but had certain disadvantages, notably the large volume of 
air which had to be operated on (200-800 litres), and the time and 
trouble which had to be expended in preparing the long absorption 
tubes containing pumice moistened with caustic potash solution, to 
say nothing of the complicated apparatus and operations sub- 
sequently necessary for liberating and measuring the absorbed 
carbonic anhydride. 
The process which we have devised is simple, and as we shall 
show is accurate also. On the one hand, it resembles Pettenkofer’s 
in that a relatively small volume of air is examined; while on the 
other, Mintz and Aubin’s principle is adopted of absorbing the 
carbonic anhydride by caustic potash solution, and afterwards 
liberating it by an acid and measuring its volume. 
Details of the New Process. 
Absorbing Solution.—The solution which we use for absorbing 
the carbonic anhydride from a given air sample consists of weak 
caustic potash, containing about 3 grams of ordinary stick potash 
dissolved in 2 litres of distilled water. Such a solution is roughly 
, and is neutralized by about } of its volume of carbonic 
N 
40 
anhydride. 
As it is essential that no carbonic anhydride should gain 
access to this solution when once the amount which it always 
contains (as carbonate) has been ascertained, we keep it stored in 
a bottle A shown in the diagram (fig. 1). 
The tube B contains soda lime, so that when drawing off a 
charge of the solution through the long tube C, the air entering the 
bottle to take its place is freed from carbonic anhydride. The ends 
of the tubes D and C are closed when this operation is not in pro- 
gress by plugs consisting of short glass rods attached to pieces of 
india-rubber tubing, and the pinch-cock E is also closed to pre- 
vent evaporation of the liquid into the soda lime. 
Measuring a Charge of Absorbent.—To measure off a charge of 
the absorbent, we employ an apparatus similar in some respects to 
2K 2 
