ATMOSPHERIC PRESSURE BETWEEN 104 C. AND 115 C. 413 



Determination of (he Weight of Condensed Steam. 



Prof. CALLENDAR has shown the importance of obtaining steam free from impurities, 

 hence only distilled water was used in the boiler. The steam passed to the calori- 

 metric apparatus through block tin tubes, and the various joints were made in such 

 a way that the steam came in contact with as little rubber as possible. After its 

 passage through the calorimeter, the steam was led into a brass condenser more than 

 1 metre long. The water was collected in half-litre flasks, placed in such a manner 

 that the end of the condenser was always 2 or 3 inches inside the flask. In all 

 experiments the flasks were closed immediately after their removal from the condenser 

 and then weighed. The question of the evaporation of the water during its conden- 

 sation and collection was carefully considered. The air in the flasks was always fully 

 saturated, since a certain amount of water remained in them from previous experi- 

 ments, so that one would not expect any appreciable amount of evaporation during 

 the actual collection of the water. Experiments were made in which, immediately 

 after weighing a flask, it was uncorked, placed near the end of the condenser for a 

 known time, and then weighed again. In no case \vas there a change in weight 

 amounting to 0'03 gr. in 15 minutes. In some experiments the collecting flasks were 

 fitted with corks, through which passed (a) a drying tube filled with calcium chloride ; 

 and (6) a tube which was only just wider than the nozzle of the condenser. Again 

 no systematic differences were observed and, since the rapid adjustment of a flask 

 under the end of the condenser was made less facile, in most of the experiments the 

 flasks have been used without drying tubes. The weights were obtained by means 

 of a large Oertling balance, sensitive to O'Ol gr. with 500 gr. on the scale-pan. It 

 was sufficiently accurate to note the weights to 0'05 gr., care being taken that the 

 total error on the difference between the weights (i.e., flask alone and flask + water) 

 was not cumulative. The maximum error on the weight determinations is not likely 

 to be greater than 0'05 gr., or 1 part in 3000 on the minimum flow. The weights of 

 water have been reduced to weights in vacuum ; the air displaced being always fully 

 saturated, its density under average conditions, viz., 760 mm., and at 15 C., has 

 been taken as 0'00121 gr. per cubic centimetre. 



Adjustment of the Flow. 



The amounts of steam flowing through the calorimeter were adjusted by means of 

 throttles, fig. 7, T. Each throttle consisted of a short length of glass tube which 

 was tightly fixed, by means of a rubber tube and wire, into the outflow tube 

 connecting the steam separator with the side-heating tube. The diameters of the 

 throttles were such that the flows used were of the order of 0'8, 0'4, and 0'2 gr. per 



second. 



Method of Making an Experiment. 



The tap, H, fig. 4, between the two main limbs of the regulator was opened slightly, 

 so that the level of the mercury in the tube in which the gas current is adjusted was 



