566 REGNAULT ON THE 
tween two observations, or it is allowed to get lower, and then 
again made stationary by regulating the lamp. It was thus easy 
to observe that the movements of the mercurial column followed 
the slightest changes in the thermometer, and precisely the same 
tensions were observed when the thermometer indicated the 
same temperature. 
By this mode of experimenting, the columns of mercury in 
the two barometers are not heated throughout their whole height, 
but those portions which rise above the vessel are subject to 
exactly the same conditions. The difference in the height of the 
two columns is taken at the temperature of the bath; this dif- 
ference, brought down to zero, gives the tension of the aqueous 
vapour. 
The question might still be raised, whether the whole of the 
chamber filled by the vapour has exactly the same temperature 
as the bath; whether the surface of the mercury, for instance, 
has not a lower temperature, in consequence of its connexion 
with the mercury below, which is colder. This last circum- 
stance certainly takes place, if the level of the mercury is allowed 
to sink nearly to the bottom of the vessel; but in my experi- 
ments the level always remained several decimetres above the 
bottom. It appeared however desirable to assure myself of this 
circumstance by direct experiment. I fixed in the circular holes 
of my sheet-iron vessel two tubes, 14 millimetres in internal dia- 
meter, and closed at bottom; the open extremity of these tubes 
extended above the level of the water in the vessel, and their 
closed extremities were fixed at the height of the level of the 
mercury in the cup U. One of these tubes contained mercury 
which ascended to the same height as the level of the mercury 
in the real barometer, the place of which it occupied. The mer- 
cury in the second tube only rose to the lowest point that had 
been observed in the wet mercurial barometer during the ex- 
periments on the elastic force of aqueous vapour. The vessel 
V V’ having been filled with water at the surrounding tempera- 
ture, after the expiration of some hours the difference of level 
was observed by the cathetometer; then the temperature was 
raised successively up to 50°, and from time to time the differ- 
ence of level in the columns, after having rendered the tempera- 
ture stationary, was read off, as in the actual experiments on the 
tension of vapour. 
The differences observed in the level under these different 
conditions, reduced by calculation to 0°, ought to remain con- 
