84 Dr. Norman Campbell on 



If the pressure is too great the electrons ejected in ionization 

 lose their initial velocities by collision with the molecules 

 before they can make their presence felt at the electrodes ; 

 if the electric field is too great their initial velocities are 

 inappreciable compared with those that they acquire in- 

 moving under the held. Evidence for the existence of 8 rays 

 in gases would be best obtained by studying the ionization 

 of a gas at a very low pressure in a field such that the 

 greatest difference of potential is not great compared with 

 that of the initial velocity of the 8 rays. 



2. Let us consider what should be the relation between i, 

 the current through a gas contained in a parallel plate con- 

 denser, and V , the potential difference between the plates, 

 when the pressure of the gas is so low that collisions between 

 the electrons liberated and the molecules of the gas are 

 negligibly rare. The current will consist of three parts. 

 (1) There will be the current carried by the 8 rays from the 

 electrodes. If the same number, N, of 8 rays are liberated 

 from each electrode, and if / (V) is the fraction of these rays 

 emitted in such a direction and with such a velocity that 

 they cannot reach the opposite electrode against an opposing 

 potential V, then this part of the current is N^./(V ). IS" 

 and/(V ) can be determined by measurements when the 

 pressure of the gas is zero *. (2) There will be the current 

 carried by the positive ions which are left when the molecules 

 are ejected. The "initial velocity" of these positive ions, 

 due to the reaction from the electrons ejected, will, if: the 

 velocity of those electrons is nearly the same as that of the 

 8 rays from metals, be less than 0*001 volt and small com- 

 pared to their velocity of thermal agitation at room tempera- 

 tures (0 033 volt). Since in the circumstances considered 

 recombination will not occur, the number of these ions 

 arriving at the oppositely charged electrode will be determined 

 by their diffusion ; this part of the current will be similar to 

 that studied in a recent paper t- If n is the number of such 

 ions formed per second this part of the current will be 



nefl-S/i/Yt), 



where jx is about 0*025 volt at room temperatures. (3) There 

 will be the current conveyed by the 8 rays from the gas. 



* Some of the 8 rays arriving- at the electrodes are reflected there (cf. 

 Phil. Mag. Aug. 1911, p. 276 and Jan. 1912, p. 46), but, if we may assume 

 that reflexion affects the 8 rays from gases in the same way as those 

 from metals, the following considerations are independent of the effects 

 of reflexion. 



f Phil. Mag. Nov. 1913, p. 912. 



