DR, ANDREWS ON THE GASEOUS STATE OE MATTER. 
425 
Although this latter result can only be regarded as an approximation to the truth, 
from the limited range of pressure employed, as well as from the uncertainty of some 
of the principles upon which the mathematical expressions depend, it may be taken to 
be sufficiently accurate for our present purpose. Combining it with the mean of the 
two foregoing numbers, we obtain for the change of internal capacity of the glass tube 
for one atmosphere of pressure 
K=0-0000036. 
From other experiments it would appear that the value of K diminishes as the 
pressure increases, a result which might be expected from the change of volume being 
chiefly due to the yielding of the inner layers of the glass of which the tube was 
composed. For 100 atmospheres the entire correction would, according to these experi- 
ments, amount only to -g^oo part. I have not thought it necessary to apply this correc- 
tion, as it falls in most cases within the limits of the errors of observation ; but the 
data I have given are sufficient to allow this omission to be readily supplied. 
I have deemed it important to attempt the solution of a problem of the highest 
interest in itself and of the greatest importance to the accuracy of many fundamental 
determinations in physics and chemistry, but which, so far as I know, has never been 
made the subject of direct investigation. It is to ascertain whether mercury has to any 
extent the property of absorbing gases in the same manner as water and alcohol. The 
fact that gases like ammonia, which are largely absorbed by water, may be readily col- 
lected and preserved over mercury, and the absence of any diminution of volume in 
gaseous mixtures standing for long periods of time over mercury, rendered it highly 
improbable that any such absorption w T ould take place. It was, however, important to 
ascertain whether under great pressures any indications of absorption could be found. 
The following is the way in which I have attempted to resolve this question. 
Having reduced the pressure to 10 atmospheres in the apparatus, mounted with an 
air-manometer and carbonic acid tube, I allowed it to remain undisturbed at this low 
pressure for ten days. At the end of this time, on a sunless day, which favoured accu- 
rate observations, water was made to circulate from a large cistern, at a steady tempera- 
ture of 8°*39, through the rectangular vessels which enclosed the tubes ; and the pressure 
was then quickly raised from 10 atmospheres till half a millimetre of liquid was formed 
in the carbonic acid tube. The air-manometer was then read, and the apparatus left 
at the new pressure for five days. It was again adjusted carefully, so that half a milli- 
metre of liquid carbonic acid was formed as before. The temperature at this reading 
was 8 0, 45. The calculated pressures, from the indications of the air-gauge, were on the 
first occasion 43*94 and on the last 44*00 atmospheres. A difference of pressure of 0*06 
atmosphere corresponded therefore to a difference of temperature of 0°*06. From 
experiments to be described in a future communication, it will appear that at this pressure 
a change of 0*06 atmosphere corresponds to 0°*055. The agreement is complete, and 
the air in the manometer had therefore undergone no diminution of volume or absorp- 
tion, after an interval of five days, from a change of pressure from 10 to 44 atmospheres. 
