0/ Determining the Elementary Electrical Charge. 213 



mercury to 55 cm. of mercury. Within a second after the 

 expansion communication between the cloud chamber and 

 the manometer was opened by means of the cock c, the height 

 of the mercury in the manometer having been first adjusted 

 by trial so that there was no motion upon opening the cock. 

 The computed equilibrium temperature* after the formation 

 of the cloud was in this case 14°"2 C, or approximately 12° 

 below the temperature of the room (26° C). Hence, as the 

 cloud chamber returned to the temperature of the room, 

 there should have been a fall in the mercury in the left 

 arm of the manometer of approximately 3 cm. Under no 

 circumstances was there ever an actual change of more than 

 half a millimetre, even though the communication between 

 the manometer and the chamber was made within a second 

 of the time of expansion. This indicated that the tem- 

 perature of the cloud chamber, just as soon as one could 

 possibly begin to observe the fall of a cloud, say two or three 

 seconds after expansion, was not appreciably different from 

 the temperature of the room. 



The above method of observation gives only the mean 

 temperature of the cloud chamber. To test directly the 

 temperature of the air exactly at the point of observation ; 

 that is, midway between the plates, I made a thermo-couple 

 out of 1 mil (*025 mm.) iron wire and 3 mil platinum wire, 

 and tested as follows its ability to register the instantaneous 

 temperature of the gas in which it was immersed. 1 placed 

 it in the centre of a carboy of 100 litres capacity and 

 pumped air into this carboy with a common hand bicycle 

 pump. Although the barrel of the pump had no more than 

 1/1000 the volume of the carboy, every stroke of the pump 

 could be easily seen by the movement of the galvanometer in 

 series with the couple, the scale reading changing from 1 to 

 2 mm. at every stroke, even though the strokes followed 

 one another at intervals of no more than a second. This 

 indicated that the couple had a sufficiently small thermal 

 capacity to respond with very little lag to changes in the 

 temperature of the surrounding gas. This conclusion was 

 further confirmed by comparing the observed and the com- 

 puted values of the instantaneous fall in temperature produced 

 by an adiabatic expansion of the air in the carboy. The 

 computed fall was 1°'1C C, the observed fall was 1°'12 C. 



The couple was next placed inside the fog chamber, midway 

 between the plates p and n (sec figure), and an expansion 

 of 20 cm. of mercury produced. The galvanometer showed 



* For the method of computation see C. T. R. Wilson, Phil. Trans. A, 

 p. 299 (1907) or J. J. Thomson, Phil. Mag. ser. 5, vol. xlvi. p. 5138 (1898). 



