280 
MR. C. T. R. WILSON ON CONDENSATION OF WATER VAPOUR 
such a position tliat S3 could be worked by the observer while looking through the 
telescope. 
Sg was then opened very slightly so that the mercury flowed very slowly through 
it into thus lowering the pressure in C. Tlie slow descent of the piston was 
watched and S3 closed just as the piston reached its zero position. The pressure in 
C as indicated by the open mercury pressure gauge was then read to the nearest 
millimetre. 
To obtain the actual pressure iu A we have to add to the pressure indicated by the 
gauge the barometer reading and to subtract the pressure required to support the 
Aveight of the piston. This last term enters as a constant correction which was 
determined combined with any constant error of the manometer, by noticing the 
pressure required to support the piston, when there was free communication between 
the inside of A and the atmosphere. At the same time the freedom of the apparatus 
from errors in the pressure-readings due to friction between the piston and the walls 
of the tube Avas tested by taking readings first Avhile the piston Avas being raised and 
then Avhile it Avas alloAved to sink sloAvly down. No diflerence was detected. 
To obtain any desired initial pressure in A, S3 is opened, Avhile Eg is in its lower 
position in order to clraAV in air through Sg which is then closed and Sg opened. Eg is 
then hung on a support fixed as high as can be conveniently reached, and S3 is closed 
just as the pressure, as indicated by the manometer, reaches the desired amount. To 
make the expansion Sg is closed and the plug G suddenly pressed up. The mano¬ 
meter is again read after expansion. 
In the case of the more insoluble gases, Avheu the temperature AA’as steady, the 
measurement of the final pressure did not require to be determined after each expan¬ 
sion. Several expansions could under such conditions be made Avithout any sensible 
change in the pressure measured when the piston was at the bottom of the tube. 
In order that the rain-like condensation might be visible, the experiments Avere 
done in a dark room, and the same method of illumination AA^as employed as in the 
case of the larger apparatus. 
The glass, too, Avas kept clear inside by periodically removing the expansion appa- 
ra,tus and washing the inside with caustic potash and nitric acid and rinsing Avell Avith 
distilled water. 
Temperature Avas measured by a mercury thermometer hung beside the expansion 
apparatus. This method is sufficiently accurate for the purpose, as the result of an 
expansion of a given amount Avas sensibly constant throughout the ordinary range of 
room temperature. It is necessary, hoAvever, that the temperature should be knoAA'u 
sufficiently accurately to enable the A^apour pressure to be found Avithin the nearest 
millimetre of mercury. An error of half a degree in the temperature reading does 
not make a diflerence of much more than half a millimetre in the vapour pressure. 
