'22S Research Staff of the G. E. C., London, on the 



electrons may be great compared with the length of the 

 .greatest straight line that can be placed in the vessel. 



6-14. It is shown that the conversion of CO to C0 2 , 

 which plays a predominant part in the disappearance of CO, 

 is simply related to ionization. The reaction 2CO = C0 2 + C 

 occurs whenever a positive CO ion collides with a neutral 

 CO molecule. Accordingly, in ideal conditions, the absorp- 

 tion of CO and the current flowing to an electrode which 

 receives all the positive ions and no other charged particles 

 would obey Faraday's law for a " semivalent " ion. But it 

 is only in carefully chosen conditions (par. 12) that any 

 approach to this law can be exhibited experimentally, and 

 the perfect proportionality seems impossible to obtain. The 

 difficulties arise from (1) the reversibility of the reaction ; 

 (2) the recombination of the ions in the gas ; (3) the arrival 

 of positive ions at the cathode, where the current they carry 

 is indistinguishable from that carried by the emergent 

 electrons. The first difficulty may be peculiar to the 

 reaction studied ; the second and third arise in all cases, and 

 w r ould seem to make illusory most of the cases of Faraday's 

 law in gases that have been recorded hitherto. 



15. Similar preliminary experiments on nitrogen have 

 been made. Here, again, absorption is proportional to ioniza- 

 tion, but the reactions involved are much more complex. 

 The " clean-up " of hydrogen, described by Langmuir, seems 

 entirely independent of ionization. 



16. The experiments seem to show that a luminous 

 discharge can be obtained, showing the "manv-lined" 

 spectrum of the gas, without any appreciable recombination 

 taking place. This fact may be important for the theory of 

 spectra. 



Ionization and Absorption. 



1. In the first paper of this series (Phil. Mag. xl. p. 585), 

 it was established that most gases do not disappear in the 

 electric discharge with thermionic emission from the cathode, 

 unless the potential between the electrodes exceeds the 

 " glow potential. - " The glow potential is reached when 

 the space charge round the cathode is neutralized by posi- 

 tive ions, and the current through the gas becomes 

 proportional to the saturated thermionic emission. In the 

 vessels used, which had both electrodes of thin wire 

 separated by distances large compared with their diameters, 

 the attainment of the glow potential was accompanied by 

 a very great, and partially irreversible, increase of current. 



It appeared natural to attribute the beginning of absorption 

 of the gas, as well as the appearance of luminosity, to a very 



