ON EVAPORATION AND DISSOCIATION. 
91 
The error in pressure is thus seen to be less than 0’5 per cent., and corresponds to an 
extremely minute difference in temperature ; in the last instance it is barely 0 - 20°. 
§ 44. In experiments on the vapour-pressures of ammonium chloride it was unneces¬ 
sary to boil out the tube A, for it was possible to ensure absence of air by raising the 
pressure, heating the substance, and then reducing pressure so as to allow gas to 
escape. The mass of solid ammonium chloride could not pass the constriction. At a 
temperature of 320° it was considered certain that all air had been removed when this 
operation had been repeated six times, a considerable interval of time elapsing between 
each escape. A large number of preliminary experiments show 7 ed that the pressure 
rose steadily, and as this was considered at first to be due to dissociation, time-curves 
were constructed. In order, as we imagined, to shorten the time required, the pres¬ 
sure in the differential gauge was raised before each reading, so as to keep the volume 
of gas constant. 
§ 45. For the sake of clearness the experiments are arranged in the order of 
temperature. At 290° vapour was allowed to escape, and a reading was immediately 
taken, with the following result:— 
Temperature. 
Pressure. 
O 
millims. 
290 
188-5 
300 
250-6 
252-5 
310 
342-4 
320 
342 1 
445-9 
443-7 
330 
597-6 
596-0 
§ 46. A considerable number of experiments were made at 320° and 340°, pressure 
being allowed to rise. 
At 320°, the following initial readings were obtained :—4497 millims. and 4477 
millims. Having noticed that pressure steadily rose, after these readings had been 
taken, it was resolved to construct time-curves, so as to ascertain the rate of increase, 
and the final value. But it was found that this rate was sensibly constant during any 
one experiment, although it diftered in each experiment, and was always slower after 
each escape of vapour. Since the explanation of the results which we propose to give 
was only arrived at after careful examination of these time-curves, it is necessary to 
give the numbers in detail. The third column will be explained later. 
