426 
DE. ANDREWS ON THE GASEOUS STATE OF MATTER. 
Having thus shown that air is not absorbed by mercury, it remained to extend the 
inquiry to the case of carbonic acid gas. For this purpose the apparatus was again 
left at a pressure of 10 atmospheres for some days, and the pressure was then quickly 
raised at 4°-99 to 35-89 atmospheres, at which pressure the carbonic acid was still 
wholly in the gaseous state. After reading the volumes of air and carbonic acid gas 
in the two tubes, the apparatus was left at the new pressure for two days, when the 
observation was repeated at the same temperature, the pressure having been adjusted 
as at the first observation. The readings of the two tubes were precisely the same as 
before. Now as the former experiments had proved that the air in the manometer 
under increased pressure undergoes no absorption, the last experiment evidently extends 
this conclusion to carbonic acid gas. In confirmation of this result, I may mention that 
after allowing the apparatus to stand for many hours at a pressure of above 100 atmo- 
spheres, and then suddenly reducing the pressure to 10 atmospheres, not the slightest 
evidence of the escape of gas from the surface of the mercury could be detected even 
with the aid of a strong magnifier. In filling a barometer-tube the air which is expelled 
by ebullition has not been dissolved in the mercury, but comes from the thin shell of 
air interposed between the mercury and the surface of the glass. 
Two questions still remain to be considered. 1. Do the glass tubes undergo a 
permanent increase of capacity when exposed for a long time to these high pressures 1 
2. Is there any absorption of oxygen in the air-manometer from its slow combination 
with the mercury \ To both questions I am able to give a satisfactory answer. The 
apparatus was mounted on October 30, and three days after the pressure at which 
carbonic acid liquefies when the, temperature is 8°‘41 was found in two experiments 
to be 43-90 and 43-94 atmospheres, the mean being 43-92 atmospheres. During the 
following two months the apparatus was in daily use for a long course of experiments, 
in which the pressures varied from 12 to 120 atmospheres, and the latter pressure was 
often maintained for the space of 24 hours. On the 1st of January the pressure 
required for liquefaction was again determined at the same temperature of 8 0, 41. In 
three experiments it was found to be 43-96, 43-96, and 43'94 atmospheres — the mean, 
or 43-95 atmospheres, differing only by 0-03 atmosphere from the former result. This 
is quite within the limits of error of observation, as it corresponded to a difference of 
less than 0-1 millimetre in the actual readings. It follows therefore that there has 
been no appreciable enlargement of the internal capacity of the tube, or reduction of 
the volume of the air from chemical combination or absorption, during a period of two 
months of active work. That no oxidation of the mercury had occurred was further 
shown by the bright metallic surface of the fine mercurial column, and the absence of 
any tendency to drag when the mercury rose or fell from change of pressure*. 
It is with regret that I am still unable to give the true pressures which correspond 
* While writing the above I have carefully examined the apparatus, and find it to be now in as perfect order 
as when it was mounted five months ago, the mercury in the air-manometer moving through its extreme range 
as readily and with as much precision as in a thermometer of the best construction. — [March 22.] 
