3-0 



2-0 



1-0 



Disappearance of Gas in the Electric Discharge. 233 



a large number of ions may be produced in these circum- 

 stances from a small amount of gas. 



4. The importance of these facts for our immediate 

 purpose is that they remove any objection to supposing that 

 the absorption of gas which is characteristic of the discharge 

 is associated with ionization of the absorbed atoms. But 

 there are other features of the discharge in vessels of this 

 type which are interesting. Thus we may study the 

 variation with the pressure and with the potential of the 

 ratio iff/ie, when the glow is developed. (The ratio is 

 independent of V within fairly wide limits ; in what follows 

 it is always to be assumed that V lay within these limits.) 

 Fig. 4 shows the results obtained in the vessel of fig. 3. 



Fio-. 4. 











IS 



: th p for 







v = a 



JO volts 













Variatior 



vessel I 



n CO 





















































V=IOO<. 



oits 















•Gl 



02 



■03 



■04 



■05 



•08 -07 



Pressure (mm) 



■08 



■09 



It is remarkable that the ratio hardly varies with the 

 pressure until it falls so low that the applied potential is 

 below V ff ; then, as fig. 3 shows, it drops suddenly. More- 

 over, at the lowest pressures represented, i g ji a is still greater 

 than 2 for V = 250, while the m.f.p. of a molecule in the gas 

 is more than 3 cm., and presumably the m.f.p. of an electron 

 still larger. Now the greatest distance of the cathode from 

 the walls is 4 cm., and an electron would not usually hit two 

 molecules if it went straight to the walls. Accordingly, i£ 



Phil. Mag. S. 6. Vol. 42. No. 248. Aug. 1921. B 



