On Electrolytic Convection in G ex s si er- tubes. 307 



electric field, the experiments show that the ionization of 

 argon and helium is somewhat more easily effected than that 

 of other gases. 



There are other reasons for thinking that the atomicity of 

 an elementary gas is not an important factor in its facility 

 of ionization. 



Thus mercury vapour was found by Thomson and Ruther- 

 ford to conduct much more freely than air under the influence 

 of the rays. Helium, on the other hand, conducts only about 

 half as well as air. 



There is one peculiarity which seems to belong to mona- 

 tomic gases only. 



It is only in such gases that aluminium cathodes " splutter/'' 

 This effect is observed in mercury, argon, and helium. 



This, however, is a subsidiary point. There is no means 

 of interpreting it in the present state of our knowledge of the 

 subject. 



XXY. On an Experiment bearing on the Hypothesis of Electro- 

 lytic Convection in Geissler-tubes. By H. Morris-Airey, 

 B.Sc* 



PROF. J. J. THOMSON, in his book 'Discharge of 

 Electricity through Gases,' describes an experiment to 

 show the convection of chlorine through hydrogen under the 

 influence- of an electric discharge. 



A small quantity of chlorine is introduced into a capillary 

 tube filled with hydrogen at a low pressure, and the discharge 

 from an induction-coil allowed to take place through the tube. 

 On examining the tube with a spectroscope the chlorine is 

 seen only at the anode, and on reversing the current the 

 chlorine appears at the new anode, and so on as often as the 

 current is reversed. 



The conclusion Prof. Thomson draws from this is, that the 

 chlorine is actually transported through the hydrogen from 

 one end of the tube to the other after each reversal of the 

 current. 



This conclusion, however, does not seem to be justified, 

 since the difference of temperature known to exist at the two 

 electrodes may be sufficient to account for the effect, the 

 chlorine being seen only at the hottest part of the tube. It 

 seemed interesting therefore to see if the motion of the 

 chlorine could be detected in some other way. 



An experiment was made with a tube of the shape shown 

 * Communicated by Prof. Schuster, F.R.S, 



