REGNAULT’S HYGROMETRICAL RESEARCHES. 637 
be heated by a spirit-lamp, when it is desired to keep the water 
surrounding the bell at a stationary temperature higher than 
that of the ambient air. 
I placed four hygrometers under the bell; two hygrometers 
mounted with hairs prepared with the solution of carbonate 
of soda, one hygrometer with a hair cleansed by zther, and 
lastly, a fourth hygrometer mounted with a filament of silk. 
The vacuum is made the first time by leaving the stopcock r 
open; then the stopcock 7 is closed, and a vacuum formed 
several times, by each time letting in dry air very slowly: the bell 
is thus dried perfectly. Finally, a vacuum is made for the last 
time as perfect as possible, and the air-pump is detached. The 
difference in the height of the two columns of the barometric 
manometer is measured by the cathetometer ; the elastic force of 
the dry air remaining in the apparatus is thus obtained. It is 
ascertained that the vacuum is maintained in an absolute manner. 
In order to introduce a small quantity of moisture into the 
bell, the stopcock 7 is opened for a few instants, and then closed 
again. The tension of the vapour introduced is measured by 
the increase of the difference of level of the two columns of 
mercury. The water of the bell is frequently stirred, and kept 
at a stationary temperature. The hygrometers assume their 
state of equilibrium very promptly. When it is well-established, 
their indications and the temperature of the thermometer are 
noted, and the elastic force of the vapour measured. 
To introduce a fresh quantity of vapour, the stopcock 7 is 
opened, and is then closed; the same determinations are recom- 
menced, and so on, until the space is brought to a state of satu- 
ration. The temperature not having perceptibly changed during 
the experiments, a table of degrees of the hygrometer is obtained 
for the different fractions of saturation corresponding to one and 
the same temperature. It is easy to make a second series of 
determinations at a higher temperature, drying the bell again, 
and operating in the manner just described. It may then be 
ascertained whether the two tables obtained are identical. 
When it is only required to make the table of one hygrometer, 
the bell need not be kept in water, and the experiments may 
be made at the temperature of the ambient air, provided that 
this temperature is somewhat high, for at too low temperatures 
the elastic force of the vapour presents very feeble absolute 
values, which cannot be measured with sufficient precision. 
