On the Spontaneous lonisation of Gases. 277 



' On the Spontaneous lonisation of Gases." By C. T. R WILSON, 

 M.A., F.R.S., Fellow of Sidney Sussex College, Cambridge. 

 Received October 22, Read December 5.. 1901. 



It was shown by Geitel,* and independently by the writer of this 

 paper,! that there is a continuous production of ions in air contained 

 in a closed vessel, even when it is not exposed to any known ionising 

 agent. The present communication contains the results of measure- 

 ments of the relative rates of production of ions under such conditions 

 in air and certain other gases. 



The relative ionisations (the term " ionisation " being used in the 

 sense of " rate of production of ions ") were obtained by a comparison 

 of the " saturation " currents in the various gases. The apparatus used 

 for this purpose differed only in minor details from that described and 

 figured in a previous paper 4 



The vessel in which the ionisation was measured was a glass bulb 

 5 '5 cm. in diameter, thinly silvered internally. The conducting system, 

 from which took place the loss of charge to be measured, consisted of a 

 brass wire with a narrow strip of gold leaf attached, the wire being 

 fixed by means of a small bead of sulphur to a brass rod passing 

 through a plug of sealing-wax in the neck of the bulb. This supporting 

 rod was kept at a constant potential by means of a condenser, con- 

 sisting of zinc plates embedded in sulphur. Electrical connection 

 could be made when required between the leaking system and the 

 supporting rod by means of a strip of watch balance-spring, soldered 

 at its upper end to the supporting rod, and with its lower end bent so 

 as to form a bridge over the sulphur bead. By approaching a magnet 

 to the neck of the bulb the free end of the spring could be brought into 

 contact with the upper end of the brass wire carrying the gold leaf. 



The position of the gold leaf was read by means of a microscope 

 provided with an eye-piece micrometer-scale. On the portion of the 

 scale used, one scale division corresponded to 0*31 volt, as tested by 

 means of a Clark cell. An initial potential of about 90 volts was used 

 in the experiments. 



To make a determination of the rate of leak, the brass wire and gold 

 leaf were brought to the potential of the supporting rod by making 

 momentary electrical connection by means of the magnetic 'contact 

 maker. The time indicated by a watch at the moment of transit of 

 the gold leaf across the next scale division was then observed ; the 

 moment of transit across a lower scale division about one hour later 

 was also observed. Contact was then made a second time and the 

 position of the gold leaf again read. The time taken by the gold leaf to 

 traverse a certain whole number of scale divisions was thus deter- 



* ' Physik. Zeitschrift,' 2 Jahrgang, No. 8, pp. 116-119. 



t ' Camb. Phil. Soc. Proc.,' vol. 11, p. 32 ; ' Roy. Soc. Proc.,' vol. 68, pp. 151-161. 



t ' Koy. Soc. Proc.,' vol. 68, p. 155. 



