IN THE PRESENCE OF DUST-FREE AIR AND OTHER GASES. 
289 
law is sufficiently nearly obeyed to justify us in saying that condensation begins 
when lies between 1’35 and 1’37. Eegnault founcP that if the pressure of 
the CO 3 was changed from 1 atmosphere to 2 , the ratio of the initial to the final 
volume was 2 X 1’0076. The difference between P]/P 3 and is therefore well 
within 1 per cent. 
Cloud-like Condensation in COg. 
Date. B. 
f C. 
7r. 
Pi- 
Pi- 
P-2- 
P 2 - 
P 1 /P 2 - 
Result. 
]\Iay 25. 775 
17-5 
15 
7.38 
1487 
223 
972 
1-530 
Rain 
„ . 775 
17-5 
15 
734 
1503 
230 
979 
1-535 
Fog- 
All the above results were obtained with the same sample of CO 3 . On absorbing 
the gas by KOH a bubble amounting to less than 1 part in 1000 of the whole 
remained. The gas had been in the expansion apparatus for three days and 
50 expansions had been made. The contamination which takes place by air diffusing 
through the lubricating water round the piston is therefore certainly very slight. 
In the experiments with CO^ the lubricating water was less frequently renewed, 
and a smaller quantity run in at a time than with the less soluble gases. There 
was, therefore, even less chance of contamination of the latter than of the CO 3 . 
Hydrogen. 
This was prepared by passing steam over sodium. This method was used by Scott 
in his experiments on the composition of water.t 
The apparatus for the preparation of the gas is shown in fig. 3. 
The water was contained in the bulb A and the sodium in B, which was prolonged 
into a narrower tube C, fused directly to H in fig. 2 . The vertical tube D served for 
the introduction of the water. 
The sodium was previously heated in a tube, kept exhausted by the water pump. 
The tube was held for several minutes in a Bunsen flame*, and the sodium then 
poured off into the clean part of the tube. 
The water to be introduced into A was obtained free from dissolved gases by 
boiling rapidly down in a flask to about one-sixth of the original volume. This was 
drawn up through D without allowing it to cool, the end of D dipped under mercury 
and the apparatus immediately pumped out. The mercury rose in D, which now 
served as pressure-gauge and safety-tube. 
The bulb A was now warmed till the pressure exceeded that of the atmosphere, 
and bubbles began to escape through the mercury in E. 
The fame was then removed, and the apparatus again pumped out, to as low a 
* ‘ Comptcs Renclus,’ vol. 23, p. 794 (1846). 
t ‘ Phil. Trans.,’ 1893, p. 543. 
2 P 
MDCCCXCVII.—A. 
