Thermodynamic Properties of Air. 7 



represented in fig. 1, provided with screw stopcocks. These 

 were weighed on a large chemical ba- Fig. 1. 



lance carrying some 1500 grammes. 

 By gradually opening the stopcocks the 

 flasks yielded during the calorimetric 

 experiment a perfectly uniform current 

 of air. They were then again weighed, 

 whence, by applying the usual correc- 

 tions, the mass of gas passed through the 

 calorimeter could be accurately calcu- 

 lated. In the three flasks used it was easy 

 to store up some 50 or 60 litres of gas. 

 The tightness of the flasks and stop- 

 cocks was controlled by the weigh- 

 ings themselves ; even a small leakage 

 renders an accurate weighing im- 

 possible *. 



The mode of preparing and compres- 

 sing the air was exactly the same as 

 described in Part I., § 5. 



§ 7. I shall now describe the calori- 

 metric apparatus, namely that form of it 

 which I usedin the low-temperature work. 

 In my first experiments I constructed 

 a calorimeter on the lines of that of 

 Reguault and Wiedemann ; the cooler, 



i. e. the system of metallic tubes destined to transmit heat 

 from the gas to the water, was attached to the calorimetric 

 vessel by solder, while the exit-tube of the heater was intro- 

 duced near the bottom of the vessel and fixed there by means 

 of a cork. This arrangement proved inconvenient, and more- 

 over it rendered difficult the intended measurement of the 

 initial temperature of the gas. Accordingly I modified the 

 apparatus in such a manner as to make the heater entirely 

 independent of the calorimeter proper. In fig. 2 (reduced 

 in the ratio 2-J : 1) DP R/R represents the heater (I shall 

 continue to call it so, although, when used at low tempera- 

 tures, it acts really as a cooler) ; S the cooler ; both are fixed 

 to a separate support E, which stands on a heavy base E'. 

 This part of the apparatus may thus be used in conjunction 

 with different calorimetric vessels. 



* I think that the method described here, of weighing gases in a com- 

 pressed condition, could be advantageously employed in accurate deter- 

 minations of the density of some gases under atmospheric pressure. For 

 that purpose the weighed flasks ought to be dischnrged into an empty 

 receiver of known capacity and temperature, provided with a mercury 

 manometer. 



