448 Scientific Proceedings, Royal Dublin Society. 
They prove that the process sufficiently fulfils the conditions 
of accuracy, which we consider to be necessary for tracing the 
fluctuations in the amount of atmospheric carbonic anhydride. 
It is also very probable that part at least of the errors arose, 
not from the process of determination, but in the preparation of 
the artificial mixtures of purified air and carbonic anhydride, the 
accurate measurement of the small volumes of the latter employed 
for the purpose being somewhat difficult. 
As a final test of the new method we made duplicate determina- 
tions of carbonic anhydride in ordinary fresh air collected in two 
different receivers. 
Determination of Carbonic Anhydride in Air taken from the 
Grounds of the Queen’s College, Belfast. 
Temperature, 16-94° C. (dry bulb), 13°61° C. (wet bulb). 
Barometer, 757:5. Weather fine, with ight S. W. breeze. 
(1) Volume of air taken, 6584 cc. = 61789 cc. at N.T.P. 
without correction for moisture, or 6098°2 with correction for 
moisture. 
CO, found at N.T.P. = 1822 ce. 
(2) Volume of air taken, 6672 c.c. = 62615 cc. at N.T.P. with- 
out correction for moisture, or 6179-7 with correction for moisture _ 
CO, found at N.T.P. = 1°872 e.c. 
COz2 found in 10,000 CO2 found in 10,000 
No. of air without cor- of air with cor- 
rection for moisture. rection for moisture. 
GIA elated ke 20018) Re ne alien cee neat C9 Si 
(2.) 5 é ‘ 2-989 S e 5 5 3°028 
Difference, : Se On04 ; : : . .0:041 
Modification of the process proposed for Field Work or for employment 
by a Scientific Expedition. 
For these purposes we propose that a series of sealed tubes 
should be prepared in the laboratory, each tube containing a 
charge of exactly 50 c.c. of the weak potash solution. 
The only operations to be performed at the place of observation 
will be the collection of the air sample in a suitable receiver, the 
transfer of the contents of one of the sealed tubes to the latter, 
