[DAWES] SPECIFIC HEAT OF A GAS AT CONSTANT PRESSURE 191 



degree in the thermometer reading. Besides its greater sensi- 

 tiveness the Bunsen calorimeter possesses the advantage of not requiring 

 any correction for radiation errors since no heat is communicated by this 

 means on account of the surrounding ice jacket. 



III. — New Method. 

 A. — Apparatus. 



In the experiment described below a method was devised by which 

 this calorimeter could be used with special advantage in determining the 

 specific heat of a gas, and at the same time some of the defects of the 

 Eegnault an-angement avoided. Fig. 2 is a diagram of the arrangement 

 of the apparatus as finally adopted after considerable development. The 

 gas was stored under pressure in a reservoir A and kept at zero tempera- 

 ture by means of ice in the vessel surrounding it. The flow of gas was 

 regulated by a valve B, and its pressure was indicated by a water mano- 

 meter C. A phosphoric pentoxide drying tube D was inserted in its 

 path to absorb any moisture coming from the manometer. The gas was 

 heated as it passed through a tube in a water bath E and was kept at a 

 temperature of 100" as far as the mouth of the calorimeter by means 

 of a steam jacket. It passed through the test tube of the calorimeter F 

 in a copper tube of special construction shewn on a larger scale in Fig. 3 

 The gas entered this tube through the inlet d and issued from it by the 

 outlets a and h, each of which could be closed by a valve. The lower 

 part of the tubing was coiled, as shewn in the figure, and immersed in 

 water to the height e. With the valve h open and a closed the gas passed 

 directly out without going through the bent portion of the tube. The 

 différence of temperature between the points c, d, was measured by 

 means of a copper-iron thermocouple, the wires of which passed out 

 through air-tight caps at a and h, and thence to a galvanometer. 



B. — Calibration. 



(1) The TTtermocoupIe. 



The thermocouple was calibrated before the wires were sealed into 

 the tubes, one junction being kept in melting ice and the other placed in 

 a water bath along with a standard thermometer. The deflections of the 

 galvanometer were observed for a series of dift'erent temperatures, and 

 the results are shown in Fig. 4. 



(2) The Gauge. 



The relation between the quantity of gas which passed out of the 

 reservoir and the corresponding fall of pressure was found in the follow- 



