THEEMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES. 357 



The muff was then 



It will be seen that the temperature of the entering gas is measured just before it 

 enters the body of the calorimeter, and the temperature of the issuing gas just after 

 it leaves the body of the calorimeter ; in neither case can the temperature of the gas 

 in the calorimeter be affected by conduction along the pipes. Most of the experiments 

 were made with the issuing gas at atmospheric temperature, so that the outer steel 

 shell was at atmospheric temperature and could neither give out nor receive heat. To 

 make this doubly sure the calorimeter was wrapped in cotton wool and placed inside 

 a muff filled with water kept at approximately the same temperature as the shell of 

 the calorimeter. The only remaining path by which heat can enter the calorimeter 

 is through the left-hand vulcanite end ;-the amount flowing in by this path must be 

 extremely small. There is a Hoskins alloy thermocouple clamped to the outside of 

 the steel t\ibe and connected to a very sensitive galvanometer which gives a deflection 

 of 20 mm. per degree change of temperature of the junction. The electric power was 

 regulated during an experiment so as to keep the temperature indicated by the 

 galvanometer as steady as possible. 



A few experiments were made with the issuing gas at +30 0. 

 heated to about the same temperature, but it is 

 possible that a little heat leaks out at the ends 

 of the muff, so that these experiments may be 

 slightly wrong. 



The passages through this calorimeter arc 

 rather small, so that there was an appreciable 

 drop of pressure in the gas as it passed through 

 it. The difference of pressure between the inlet 

 and outlet was measured by means of a differen- 

 tial pressure gauge, the construction of which 

 is shown in fig. 2. An inner glass tube dips 

 into mercury contained in the outer steel tube. 

 The difference of level of the mercury inside and 

 outside the glass tube, which is proportional to 

 the difference of pressure is measured by the 

 change of resistance of a wire stretched down 

 the centre of the glass tube. Several sorts of 

 wire were tried ; some made uncertain contact 

 with the mercury, but a bright hard steel banjo 

 wire answered very well. The resistance of the 

 wire was measured with a simple Callendar and 

 Griffith bridge. The gauge was calibrated by 

 direct comparison with a glass U-gauge containing mercury. The arrangement of 

 the valves enabled the pressure to be applied and released quickly, so that the zero 

 and pressure readings could be repeated several times as a check on their accuracy. 



HIGHER 



"PRESSURE 



LOWER 

 PRESSURE 



SCALE 



1 INCHES 



Fig. 2. 



