558 Mr. A. Wood on Spontaneous Ionization of 



with mica windows to shield it from draughts and air- 

 currents. Then, again, the field inside the vessel was far 

 from being uniform, as the plate was not symmetrical with 

 respect to the cylinder. This, probably, accounted for a small 



Fiff. 2. 



B,7TT£/?y 



A. Gold-leaf. B. Brass Plate. C. Quartz Pillar. D. Charging Kod. 



E. Earth. 



oscillation of the leaf which was noticed. On the other 

 hand, the apparatus was complete in itself, very easy to sur- 

 round with screens of different materials, and sufficiently 

 sensitive to give a leak easily measurable in an hour. It was 

 found that the following method of experiment gave the most 

 satisfactory results. The plate was charged up, the charging- 

 rod left in contact with it, and after a short time the position 

 of the gold-leaf was read every half-minute for five minutes. 

 The mean of these readings gave the zero position of the 

 gold-leaf, and eliminated small errors due to its vibratory 

 motion. The plate was then insulated, and the position of 

 the gold-leaf again read in the same way. The mean of these 

 readings gave its initial position, and a similar set at the end 

 gave its final position, and so the deflexion for the period. 

 At the end of the experiment the zero was redetermined by 

 bringing the rod again into contact with the plate^ and any 

 small change was allowed for. With this arrangement the 

 experiment was repeated. Blocks of wood were piled round 

 the instrument to a depth of 29 cm., and a small but distinct 

 increase of the ionization was observed. 



