Resonance Potentials for Electrons in Metallic Vapours. 45 



the ionizing potential of a metallic vapour and the limiting 

 wave-length for photoelectric sensibility of the solid. The 

 data available at the time his paper was written were " too 

 meagre to warrant any importance being attached to the fact 

 that the difference between V , the voltage corresponding to 

 the limiting line for photoelectric sensibility of the solid 

 metal and V c the ionization potential for the vapour, is nearly 

 constant, and equal to 6 volts/' If such a relation did exist 

 as Van der Bijl points out, it would mean that a material 

 having an ionization potential of less than 6 volts would be 

 photoelectrically active in the dark. Both sodium and potas- 

 sium have ionization potentials less than 6 volts. In the 

 case of lithium, rubidium, and caesium we should expect the 

 ionization potentials to be 5*37, 4*16, and 3*88 volts re- 

 spectively, although these values have not been determined 

 experimentally*. All of these materials accordingly would be 

 photoelectrically active in the dark. That the above empirical 

 relation is not justified is seen from the following table of 

 experimental data : — 



Table V. 



Element. Photoelectric Ionization 



V . Ye. Ya-V . 



Mg 3-08 7-75 4-67 



Cd 3-49 8-92 5-43 



Zn 3-77 95 5-73 



Na 1-81 513 3-32 



Hg 4-44 1035 591 



Thus, V c — Y is definitely not equal to 6 volts and no 

 simple accurate relation is apparent. 



Hebb f has shown that ionization of mercury vapour is 

 not produced by the radiation A = 2537 acting photo- 

 electrically. This fact is in accord with the above discussion. 

 "We should expect the long wave-length limit to be X = 1188 A. 

 That radiation of wave-length \ = 2537 cannot ionize mercury 

 vapour is evidenced by the failure of Tate t, Davis and 

 Goucher §, and several other investigators to detect any sign 

 of ionization of a normal mercury atom suffering resonance 

 collision, as during such collisions the line \ = 2537 is 

 emitted. 



Giibreath || has concluded that potassium vapour is acted 



* Note added Nov. 11. — We have since experimentally determined and 

 confirmed these values for Kb and Cs. This paper including data for As 

 will appear in Phys. R. 



t Phys. R. xi. p. 179 (1918). 



% Tate, Phvs. R. vii. p. 686 (1916) ; idem. x. p. 81 (1917). 



§ Davis and Goucher, x. p. 101 (1917). 



|| Phys. R. x. p. 166 (1917). 



