616 REGNAULT’S HYGROMETRICAL RESEARCHES. 
ture very near that at which saturation would occur; but it 
may equally arise from the glass of the balloon retaining con- 
densed water on its surface by virtue of its hygroscopic property. 
When, on the contrary, this method is employed in the deter- 
mination of the density of aqueous vapour under very weak 
pressures, the least errors in the weighings cause much more con- 
siderable ones in the numerical value of the density. Moreover 
this method does not admit of obtaining the density of the va- 
pour at other temperatures than 100°; it is important to deter- 
mine this density at temperatures more and more approaching 
that which would produce saturation. 
The following process admits of these new determinations 
being made with greater precision. A large glass balloon is 
gauged exactly by weighing it full of water. Into this balloon 
a small glass bulb is introduced hermetically closed, and con- 
taining a quantity of water accurately weighed. The neck of 
this balloon is cemented into a tubulature which establishes the 
communication with a barometrical manometer. The balloon, 
as well as the manometer, is arranged in a large vessel filled 
with water. By means of a piece of plate glass the manometer 
may be measured by the cathetometer. All the precautions are 
moreover taken which have been described at length in my me- 
moir on the elastic force of aqueous vapour*. 
The balloon is dried, as complete a vacuum as possible is then 
formed, and the elastic force of the small quantity of air which 
remained in the balloon is then measured very exactly. By 
means of a piece of incandescent charcoal the glass bulbis broken, 
and the temperature of the water which surrounds the balloon 
is raised above that at which the vapour would be in a state of 
saturation. This temperature is rendered stationary, and the 
elastic force which exists in the balloon is measured. On sub- 
tracting from this elastic force that which belongs to the air, 
we obtain the elastic force of the vapour. 
Whilst the temperature is lower than that at which the space 
would be saturated by the weight p of water contained ori- 
ginally in the glass bulb, we shall find, for the tension of the 
vapour, the maximum tension which corresponds to that tem- 
perature; but beyond that temperature the vapour will behave 
like a gas, and designating its elastic force by f, we shall have 
for its density relative to that of the air taken under the same 
* See preceeding Memoir, p. 570. 
