of Ediiiburgh, Session 1882-83. 
27 
air initially filling A had been displaced. Meantime water had been 
kept running through both compartments of the jacket. The water 
from the laboratory reservoir was found to be practically invariable 
in temperature during the course of any one experiment. When 
the absorber had been filled, and the water had been running 
through the jacket for some time, the stopcocks B and C were 
closed and D opened. The water was then kept running through 
the jacket until the gas in the absorber had attained a practically 
constant temperature, until the liquid in the manometer therefore 
indicated a practically constant pressure, The barometer and 
manometer were then read, and the temperature of the water jacket 
noted. Then L was put in communication with the boiler, and 
the water in the radiator was displaced by steam, after which the 
variation of the manometer with time was noted until it had again 
reached a constant state, when the barometer was again read. If 
the readings of the barometer before and after any set of observa- 
tions were not the same, it was supposed to have varied uniformly 
with time in calculating pressures. In all cases the difference was 
very slight, the sets of observations lasting only from 20 to 30 
minutes. From the readings of the barometer and manometer the 
j^ressures were easily calculated. 
No precautions were taken specially to exclude dust, so that, when 
water was present in the gas operated on, it was not necessarily all 
in the state of vapour. 
The following are details of some of the experiments made. The 
unit employed in the measurement of pressure is that due to a 
millimetre of the dilute acid solution used in the manometer. Its 
density was 1T038. Zero of time is the moment at which com- 
munication was broken between the radiator K and the water 
supply, and established between the radiator and the boiler. 
Shortly after that the upper end of the absorber begins to rise in 
temperature, and the pressure of the gas contained in it to increase. 
When the steam issues freely from the radiator, it may be supposed 
to have very nearly reached a constant temperature. 
I. Dry Air . — Air had been drawn from the drying tubes through 
the absorber for 2 hours. The following table shows the pressures 
after given intervals of time : — 
