BETWEEN THE FREEZING AND BOILING-POINTS. 
239 
complete experiments. They extended over a period of just a year, and divide 
themselves naturally into 8 separate series.'* 
Series 1. Nos. I. to XV. 
This series includes experiments with Calorimeter C, between 4° and 35° C., and 
Calorimeter D, at 28°. Both calorimeters were fitted with a stranded platinum heating- 
wire with the silk-covered rubber cord wound round. The distilled water supplied to the 
calorimeter was boiled before running under the oil, in bottles forming the head, but 
no special care was taken to keep it hot while running in. A large quantity of the 
air was driven off 1 in the process of boiling, but subsequent results have shown that the 
water in the head must still have contained a considerable quantity of dissolved air. 
Several of the experiments include other flows, besides the flows used throughout the 
entire series of experiments. These have been already summarized in Section 5. 
The correction for Thermometer E is that given under Group I., Section 3, c. 
Series 2. Nos. XVI. to XXVII. 
Between Series 1 and 2 several alterations were made to the apparatus, one of the 
chief being the introduction of 40 feet of tin tubing into the constant temperature 
tank to replace a similar amount of copper tubing used previously. This was found 
necessary owing to the gradual formation of copper rust in the tube. This rust was 
carried into the small rubber tube conveying water from the tank to the calorimeter, 
and gradually reduced the flow. The experiments were extended from 22° up as far 
as 60° C., where they had to be discontinued in order to further refine the regulating- 
attachments. The calorimeter was fitted with a solid platinum heating-wire, with 
silk-covered rubber cord. The agreement of the results at the lower points between 
22° and 35° with those in Series 1 is very satisfactory. Above 45° the results are 
not so consistent, probably on account of the fluctuations in gas-pressure supplying 
the main heat to the circulating system. The experiment at 60° was taken, however, 
when no other gas was being used in the building, and the conditions were unusually 
steady. Rises of temperature of 11° and 5° were tried as a check on the measure¬ 
ments. The correction for Thermometer E is that obtained in the test in Group II., 
Section 3, c. 
Series 3. Nos. XXVIII. to XXXII. 
This series includes another attempt to obtain the high temperatures, but nearly 
all the experiments were spoiled by the liberation of air inside the calorimeter. The 
experiment at 67° is given as an illustration of the effect produced by the appearance 
of air. In this series the air bubbles were found in the calorimeter water after the 
experiment, in spite of the fact that the distilled water was kej3t continuously boiling 
as it was supplied to the head bottles, and was cooled from 100° C. only by the cold 
* In the following cl is used instead of the 8 of Sec. 2. 
