ELASTIC FORCES OF AQUEOUS VAPOUR. 567 
stant, if the two columns are equally influenced by the corre- 
sponding parts. This, in fact, was found to be the case, no dif- 
ference having been observed which exceeded 0:07 millimetre. 
The water in the vessel being kept stationary at a tempera- 
ture of 45° by means of the lamp, the reservoir of a very deli- 
cate thermometer was immersed in the mercury of the tube 
which contained this liquid at the lowest level. After the lapse 
of two or three minutes, it indicated an absolutely stationary 
temperature, it was then quickly taken from the tube and plunged 
into the water of the vessel; not the least elevation of tempera- 
ture was perceptible, though the thermometer was capable of 
showing with accuracy about the th of a degree. A second 
thermometer immersed constantly in the water of the vessel 
proved, besides, that the temperature of the water had not 
changed during the interval employed in reading the indications 
of the first thermometer. 
The surface of the mercury in the wet barometer was covered 
by a layer of liquid 3 or 4 millimetres in height; this layer, by 
its weight, would depress the level of the mercury, but, on the 
other hand, it would tend to heighten it by its capillary action. 
By measuring with the cathetometer the distance between the 
lowest point of the concave meniscus formed by the water and 
the summit of the convex meniscus of the mercury, and dividing 
this distance by 13°5,the density of mercury compared with water, 
we obtain the little height of mercury equivalent to the column 
of water thus estimated. The capillary action of the water was 
determined by direct experiment. To the upper extremities of 
the two barometric tubes, used in the preceding experiments, 
two tubes of small calibre, previously cemented into a copper 
tube having three arms, were joined. The third arm was fur- 
nished with a glass tube communicating with the air-pump. 
Between one of the barometric tubes and the copper tube was 
placed a U-shaped tube containing pumice-stone moistened with 
sulphuric acid. The two barometers dipped into the same cup 
of mercury. A vacuum was produced several times with the 
air-pump, and the air admitted to dry the sides of the tube in 
connexion with the sulphuric acid. Lastly, a vacuum was again 
made, and the tube communicating with the air-pump sealed by 
fusion; the level of the two columns being then the same, a 
small quantity of water, about as much as was used in the ex- 
periments upon vapour, was passed up into one of the tubes. 
There were now two imperfect barometers communicating with 
