Prof. Thomson, On the ionization of Gases, etc. 
417 
On the ionization of Gases bv Ultra- Violet Light and on the 
evidence as to the structure of light afforded by its Electrical 
Effects. By J. J. Thomson, F.R.S., Cavendish Professor of 
Experimental Physics. 
[Read 28 October 1907.] 
The ionization of gases by ultra-violet light has been observed 
by Lenard (Ann. der Phy. I. p. 486, ill. p. 298): in view of the 
importance of this effect in Meteorology however further experi- 
ments may not be superfluous. In working with tubes with a 
Wehnelt hot lime covered cathode I observed that the electric 
discharge when passing through the tube gave out large quantities 
of ultra-violet light, and this method of producing the light was 
adopted in the following experiments. The arrangement used is 
shown in Fig. 1. G is the Wehnelt cathode, A the anode; a quartz 
window P A mm. thick was fastened to the tube by sealing-wax. 
The cathode was freshly covered with lime and raised to a white- 
heat. A potential difference of 300 volts, producing a current of 
about T ampere, was generally used. The portion of the tube on 
which the window was fixed was inserted in a cardboard box B. 
The gas to be experimented on, after passing through drying- 
tubes and a long cylinder of cotton wool to remove the dust, 
passed in front of the window and then between two concentric 
tubes E and F insulated from each other, the distance between 
the tubes was about 3 mm., the inner tube was connected with a 
tilted Wilson electroscope giving a deflection of about 40 scale 
divisions for a volt and the outer cylinder was charged up to 
100 volts. The natural leak of the electroscope was determined 
with air in the ionization vessel and no flow of gas. The stream of 
gas was then passed without any discharge of electricity through 
the tube and a new determination of the leak made, so as to be 
sure that the gas was not ionized initially. No material increase of 
the leak was found, the discharge was turned on and several deter- 
minations made with the stream of gas on and off. The leak was 
now so marked that there was no difficulty in at once telling by the 
electroscope when the current of gas or the discharge was stopped 
or started. After the discharge and current were stopped a re- 
determination of the natural leak was made ; this is necessary, 
