614 REGNAULT’S HYGROMETRICAL RESEARCHES. 
A small quantity of water is introduced into the balloon, which 
is then put in communication with an air-pump; a tube con- 
taining pumice-stone moistened with sulphuric acid is inter- 
posed to prevent the aqueous vapour from penetrating into the 
parts of the air-pump. A vacuum is formed for a considerable 
length of time. The vapour which is incessantly produced in 
the balloon finally expels the whole of the air, upon which the 
stopcock is closed. The balloon is placed in a large vessel of 
galvanized sheet-iron B C D E, so that the stopcock 7 is in front 
of a tubulature T fitted to the vessel. This vessel, which con- 
tains a layer of water two decimetres thick, is heated over a fur- 
nace. The extremity of the tube dc is inserted in a brass tube 
cd which is itself soldered to a flexible leaden tube de forming 
part of the mounting N of a large flask F. The flask F is kept 
in the water at the ambient temperature. A small leaden tube 
¢ forms a communication between the flask F and a barometrical 
manometer. 
A partial vacuum is formed in the flask F, and when the water 
is in full ebullition in the vessel BC DE, the stopcock r is 
opened. The water of the balloon then distils over, and con- 
denses in the flask #. At the end of about an hour the elastic 
force of the vapour is measured on the barometric manometer, 
and the stopcock r is closed. The communicating tube dc is 
completely dried, and the balloon is weighed after having been 
left till the following day on the hook of the scales. In this 
weighing, likewise, the method is followed which I have described 
in my memoir on the determination of the density of gases *. 
The balloon is arranged anew in the vessel BC D &, placing 
it in communication with the flask F. The water is raised to the 
boiling-point and an almost perfect vacuum is formed. On open- 
ing the stopcock 7, the greatest part of the vapour contained in 
the balloon condenses in the flask F, and there only remains a 
very small quantity, which holds in equilibrium the elastic force 
remaining in the flask F. This is carefully measured when the 
equilibrium is established, and the stopcock r is closed. 
The balloon is weighed. The difference p of the two weights 
is the weight of the vapour which fills the balloon at the tempe- 
rature T of boiling water under a pressure equal to the differ- 
ence h of the elastic forces observed in the two experiments. 
The weight z of the dry air which fills the balloon at 0°, under 
* Annales de Chimie et de Physique, 3° Série, vol. xiv. p. 211. 
