554 Mr. A. Wood on Spontaneous Ionization of 



potentiometer-wire reinserted, and the zero reading repeated. 

 If there was little change in the zero during the experiment, 

 the change was allowed for, but if the change was at all 

 large the reading was rejected. The sensitiveness was now 

 redetermined, and the voltage to which the electrode had 

 been raised was found from the deflexion and the mean of 

 the values of the sensitiveness before and after the twenty- 

 minute period. The capacity of the electrode system remained 

 the same, and so this voltage was a measure of the ionization 

 of the air in the vessel. The extremely small capacity of the 

 electroscope made it invaluable for measuring' the small 

 charges obtained. A rough determination of the capacity of 

 the electrode, connecting wire, and gold-leaf showed it to be 

 about one electrostatic unit. It was found that wheu the 

 field was first applied, the leak in the vessel was considerably 

 above the steady value which it afterwards attained ; and to 

 avoid possible errors from this source the field was usually 

 left on all night, and the connexions so arranged that the 

 screens used to cut down the ionization could be put in 

 position and removed without taking off the field. The 

 probable cause of this phenomenon will be discussed later. 

 The lead screen used in the experiments was cylindrical in 

 form, and consisted of six layers of sheet-lead, making in all 

 a thickness of about 1 cm. The ionization in the vessel w r as 

 first measured without any screen — the mean of six twenty- 

 minute periods being taken as the correct value ; and the 

 vessel was then surrounded with the screen and the ionization 

 again measured. The two measurements were repeated 

 alternately several times, and from the values obtained the 

 mean ionization with and without the screen was found. The 

 results of five such sets of observations are given below. The 

 first set is given in full, and the initial falling off in the leak 

 will be noticed. In this particular case it lasted for nearly 

 two hours, but usually the time was much shorter. 





Ves. 



el Unscreened. 





Zero. 

 14-0 



Reading. 



Deflexion. 



Sensitiveness. 



Leak. j 



655 



31-0 



30-00 



1-930 i 



320 



53-0 



210 



28-25 



0-730 



35-0 



52-5 



17-5 



28-50 



0-610 



33-5 



51-0 



175 



29-75 



0-590 



32-5 



49-0 



16-5 



29-25 



0-564 



320 



46-0 



140 



28-37 



0-493 



28-0 



40o 



12-5 



27-37 



0-464 



28-5 



41-5 



130 



26-00 



0-490 



28-5 



405 



120 



25-50 



0-471 



29-0 



410 



120 2600 



0-461 



28-0 



41-0 



13 25-50 



1 



0-510 



