214 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
after repeated trials with other arrangements, that it was necessary, in order to 
separate the air from the water, to drive the latter into steam and condense it again 
in an air-free medium, and at the same time draw off the air before it had a chance 
to dissolve again. The method adopted finally was to exhaust the air over the oil in 
the two bottles by connecting with a water-pump while the boiling water from the 
copper tank was running in. In this way, by introducing a slight constriction in the 
siphon tube, such as a connection made from a small-bore rubber tube, the water in 
the tube separated into steam just past the constriction, and was condensed in the 
cold water already in the bottles from a previous filling. The air was at the same 
time sucked up through the oil and carried off by the pump. It was a matter of 
considerable surprise to see the amount of air thus drawn off. 
After adopting this method of filling the bottles, there was no further trouble from 
air appearing inside the calorimeter, even as high as 90° C. When working at high 
temperatures, the distilled water, before it passed into the constant temperature 
tank, was passed through a spiral of tin tube in a steam-jacket. Instead of increasing 
the heat-loss of the circulating system by introducing cold water into the tank, a 
small quantity of heat was supplied by this means to the tank by the water flowing- 
in from the steam-jacket. Moreover, the steam-jacket was a check on the state of 
the water, air bubbles being generated if the water was not perfectly air-free. The 
water which was run through the calorimeter was never used a second time, although 
it might just as well have been. Fresh distilled water was so easily prepared by 
means of a small water still in the laboratory that it was deemed unnecessary to use 
it twice. 
A constant-level head was arranged near the apparatus to supply water for the 
cooler for the outflowing water; at the same time it also supplied the water circula¬ 
tion for the standardized resistance oil-bath, for the constant-level device of the tank, 
and for the condenser on the steam-jacket used at the higher temperatures. 
Method of Making an Experiment .—Obviously various expedients were necessary 
in order to have the jacket maintained at a constant temperature at any point on the 
scale between 0° and 100° C. For the experiments near 0° C. the regulator was 
removed entirely from the tank, which was then filled with cracked ice and water. 
A wire sieve was placed over the stirrer in the bottom of the tank, so as not to have 
its action interfered with by the ice, as well as to prevent any ice from being drawn 
into the circulating tubes. Wonderfully steady conditions were produced in this 
way, and maintained without the variation of a hundredth of a degree for over an 
hour at a time. Between the measurements with each flow of water used, which 
lasted about an hour, the tank was replenished with ice. Not a great deal of ice was 
required for this replenishing, since only about one-quarter of the tank-full of ice was 
melted in the hour. From 50 to GO lbs. of ice were melted during a complete 
experiment, although considerably more was used to cool the apparatus down to the 
ice-point before the experiment was started. From 100 to 150 lbs. of ice was 
