422 
DR. ANDREWS ON THE GASEOUS STATE OE MATTER. 
would be easy with very fine glass tubes to make accurate observations even at much 
higher pressures. 
As the metallic tubes, whether made of cold-drawn copper or of forged iron, which 
form the body of the apparatus are, as at present constructed, f- of an inch in internal 
diameter, I have been able to make an important improvement in the arrangements. 
The glass tubes containing the gases now dip into small mercurial reservoirs formed of 
thin test-tubes, which rest on ledges within the metal tubes. This arrangement has 
prevented many failures in screwing up the apparatus, and has given greater precision 
to the measurements. 
In the following experiments the glass tube was filled with the gas in a pure and dry 
state by passing a stream for a long time through it while in an upright position ; and 
when the air was entirely expelled, the upper end was hermetically sealed. The stream 
of gas being still maintained across the lower end of the tube, which was enclosed in a 
test-tube partly filled with mercury, the whole apparatus was left for half an hour in 
an apartment at a steady temperature, after which the gas was imprisoned at a known 
temperature and pressure, by bringing the lower or open end of the tube into contact 
with the surface of the mercury in the test-tube. By this process the original volumes 
of air in the manometer, and of carbonic acid in the carbonic acid tube, could be fixed 
with great accuracy. 
The capacities of the entire tubes and of their capillary parts were ascertained by a 
set of careful determinations of the weight of mercury which filled them at a known 
temperature. Before the introduction of the mercury, the interior of each tube was 
carefully cleansed by boiling nitric acid in it, and afterwards washing it with distilled 
water and absolute alcohol. No attempt was made to remove by boiling the thin film 
of air which, even in the most carefully, cleansed tube, is always interposed between the 
surface of the glass and the mercury drawn into the tube. Under the conditions of 
these experiments this correction must be very small ; and its estimation would be a 
matter of extreme difficulty, as in screwing up the apparatus air of different densities 
would be imprisoned between the glass and mercury. On a future occasion I hope to 
lay before the Society the results of a special investigation of this subject. The average 
capacity of the capillary part of the tube of the air-manometer used in the greater 
number of the following experiments was for 1 millimetre 0 - 00018121 cubic centimetre, 
and this tube bore a pressure of upwards of 200 atmospheres without bursting. I have 
completed a series of experiments at higher pressures, which I hope soon to commu- 
nicate to the Society, with a hydrogen manometer, whose capacity for each millimetre 
was only (M)00016861 cubic centimetre, or yj of the preceding. Such a tube would bear 
a pressure many times greater than the former, and no serious difficulty would arise in 
operating with even finer tubes. There is therefore scarcely any limit to the pressures 
which may be measured in glass tubes. The glass of which these tubes were made was 
of excellent quality, and was specially prepared for me by J. Powell and Sons. 
No pains were spared in calibrating the capillary portions of the tubes. For this 
