574 REGNAULT ON THE 
of water, but I made use of it frequently to determine the elastic 
forces of very volatile liquids, as ether, sulphuret of carbon, &c., 
which will be published in a future memoir. 
It is composed of a bent tube adc, 15 millimetres internal 
diameter, and terminated by a narrower bent tube ce. The 
closed arm a 6 is filled with mercury, which is carefully made 
to boil, to free it completely from air and moisture. When the 
mercury has cooled, a small quantity of the volatile liquid is in- 
troduced into the branch dc, and made to boil for a few moments 
to expel any air which it might hold in solution; the tube is 
then inclined in such a manner as to allow a small portion of 
the hot liquid to pass into the closed arm ad. The portion of 
liquid remaining in dc is then expelled by boiling; to effect 
this more completely, an imperfect vacuum is made by a small 
pump, if it can be easily managed. The arm dc thus remains 
filled with dry air. 
The tube adc is then fixed into the sheet-iron vessel V V! in 
a perfectly vertical position opposite the glass. The tube ce is 
cemented into a connecting piece of copper with three arms e df, 
the arm d of which communicates with the manometrical appara- 
tus hilk, furnished with a stop-cock 7. To the arm fis cemented 
a narrow tube fg, which in case of need can be brought into 
connection with a small air-pump. The two connected tubes 
ih, kl are completely filled with mercury ; the air contained in 
the apparatus escapes by the open tube fg: this tube is then 
closed by the lamp. In order to diminish the pressure, mer- 
cury is allowed to flow out of the manometer by opening the 
stop-cock 7; the air enclosed in the arm bc expands into a 
larger space, and its elastic force diminishes. The mercury is 
allowed to escape until the level in the closed arm ad descends 
to m. The tension of the vapour is then measured by the pres- 
sure of the atmosphere, less the column of mercury #f in the 
manometer and the column mz in the bent tube abc. The 
capillary influence of the small layer of liquid above the me- 
niscus in the tube a 4, is determined, at the close of the experi- 
ments, by direct measurement, after equality of pressure has been 
established in the arms a and dc by opening the tube at a. 
The vessel V V’ is filled with water, maintained at a constant 
temperature in the manner described at page 565. 
It is easily ascertained, by the following method, whether the 
apparatus is properly arranged, and whether the arm a 6 is com- 
