154 



Mr. C. T. E. Wilson. 



vessel containing only 163 c.c. This made it niuch more easy to 

 ensure the freedom of the air from dust particles. Geitel worked with 

 volumes amounting to about 30 litres ; his observations show the 

 interesting phenomenon of a gradual increase of the conductivity of the 

 air in the vessel towards a limiting value, which was only attained 

 when the air had been standing in the vessel for several days. This, 

 as Geitel points out, is to be explained by the gradual settling of the 

 dust particles, the conductivity of the air being greatest when there 

 are no dust particles present to entangle the ions. 



The principal difficulty in the way of obtaining a decisive answer to 

 the question whether any leakage of electricity takes place through 

 dust-free air is the fact that one is so liable to be misled by the leak- 

 age due to the insulating support. As will be seen from the descrip- 

 tion which follows, this source of uncertainty was entirely eliminated 

 in the method which I adopted. It had, moreover, the advantage of 

 reducing to the smallest possible value the capacity of the conducting 

 system in which any loss of charge is measured by the fall of 

 potential. 



The conducting system, from which any leakage is to be detected 

 and measured, consists solely of a narrow metal strip (with a narrow 

 gold leaf attached to indicate the potential), fixed by means of a small 

 bead of sulphur to a conducting rod which is maintained at a constant 

 potential, equal to the initial potential of the gold leaf and strip. 

 With this arrangement, if any continuous fall of potential is indicated 

 by the gold leaf, it can only be due to leakage through the air ; any 

 conduction by way of the sulphur bead can only be in such a direction 

 as to cause the leakage through the air to be under-estimated. 



The form of apparatus used in all the later experiments is indicated 

 in fig. 1. The gold leaf and thin brass strip to which it was attached 

 were placed within a thin glass bulb of 163 c.c. capacity; the inner 

 surface of the bulb being coated with a layer of silver so thin that the 

 gold leaf could readily be seen through the silvered glass. The upper 

 end of the strip had a narrow prolongation, by means of which it was 

 attached by a sulphur bead of about 2 mm. in diameter to the lower 

 end of the brass supporting rod. The latter passed axially through 

 the neck of the bulb, its lower end just reaching to the point where 

 the neck joined the bulb. The interior of the neck of the bulb was 

 thickly silvered to secure efficient electrical connection between the 

 thin silver coating of the inside of the bulb and a platinum wire sealed 

 through the side of the tube. The platinum wire was connected to 

 the earthed terminal of a condenser consisting of zinc plates embedded 

 in sulphur, the other terminal of the condenser being connected to the 

 brass supporting rod and maintaining it at a nearly constant potential. 

 An Exner electroscope connected to the same terminal of the con- 

 denser was used to test the constancy of the potential; and any loss 



