Air in Closed Vessels and its Causes. obi 



the vertical. Opposite thi< plate, and occupying as nearly as 

 possible the same relative position as in the Wilson electroscope, 

 was suspended a gold-leaf. Through an ebonite plug iu the 

 bottom of the cylinder a vertical wire passed, to the end of 

 which a horizontal cross-wire was soldered, and this could 

 be turned round so as to make contact with the plate. This 

 charging rod was connected to one terminal of a battery of 

 -mall storage-cells, the other terminal of which was connected 

 to the ea>e ot' the instrument and to the gold-leaf. This 

 arrangement was the inverse of that of the AVilson electro- 

 scope. In this case the gold-leaf was kept at constant 

 potential (zero) while the potential of the plate was allowed 

 to vary. By altering the tilt of the instrument by means of 

 levelling-screws attached to the base, and by varying the 

 potential of the plate, the combination of tilt and potential 

 giving maximum sensitiveness was found. In this position, 

 a variation of one volt in the potential of the plate gave a 

 deflexion of the gold-leaf through about 18 divisions. The 

 potential of the plate which gave this sensitiveness was about 

 200 volts. It was charged up by means of the charging rod 

 and then insulated. As the potential of the plate fell the 

 gold-leaf was deflected from its initial position, and as this 

 position was in a region of uniform sensitiveness, the deflexion 

 was a measure of the change in potential of the plate, and 

 consequently of the ionization in the vessel. As the reliability 

 of the results depended entirely on the insulating power of 

 the quartz pillar, this was very carefully tested, and great care 

 was taken to have it as perfect as possible. It was obtained 

 fre-h from the blowpipe and put in position with forceps, 

 great care being taken not to finger it. A small capsule 

 containing calcium chloride was enclosed in the apparatus, 

 and it was found that with those precautions the quartz 

 insulated splendidly. It was tested as follows : The leak was 

 first of all determined in the ordinary way. Some fine wire 

 was now twisted several times round the quartz pillar, about 

 halfway down, and connected to the charging-rod. In this 

 case there could be no leak to earth by means of the quartz. 

 The leak was now redetermined, and the two values were prac- 

 tically identical. This new form of apparatus (fig. 2, p. 558) 

 Was not without new detect- of its own. In the first place, it 

 was extraordinarily sensitive to small changes of temperature. 

 The large volume of the vessel in which the gold-leaf was 

 now enclosed meant a great increase of convection-currents 

 whenever any non-uniform changes of temperature took 

 place. This de-feet was remedied by encasing the instrument 

 in asbestos, and surrounding it with a cardboard screen fitted 



