the Electrical Conduction produced by it. 133 



to the total ionization, the above experiments show that the 

 rate of production of the ions at any point is proportional to 

 the pressure. It has been previously shown that the absorp- 

 tion of the radiation is approximately proportional to the 

 pressure. 



Let q = rnte of production of the ions near the uranium 

 surface for unit pressure. 

 A = coefficient of absorption of the gas for unit pres- 

 sure. 



The total number of ions produced between the plates, distant 

 d apart, per unit area of the plate is therefore easily seen to 

 be equal to 



f 



pq \ e-P*** da?; 



or to 



U 1 - 



-p\ Q d \ 



P 



since we have shown that the ionization and absorption are 

 proportional to the pressure. If there is a saturating electro- 

 motive force acting on the gas, the ratio of the rate of leak at 

 the pressure p 1 to that at the pressure p 2 is equal to the ratio r 

 of the total number of ions produced at the pressure p l to the 

 total number at pressure p 2 and is given by 



1 — e~ p ^o d 



Now pJ^Q is the coefficient of absorption of the gas for the 

 pressure p lt If the absorption is small between the plates, 

 p{k d and p^d are both small and the value of r reduces to 



rts Pi 



P? 



or the rate of leak when the pressure is small is proportional 

 to the pressure. 



If the absorption is large between the plates at both the 

 pressures pi and p 2 , the value of r is nearly unity — i. <?. the 

 rate of leak is approximately independent of the pressure. 

 Experimental results on this point are shown graphically in 

 fig. 7 (p. 138). _ 



For intermediate values of the absorption, the value of r 

 changes more slowly than the pressure. 



With the same distance between the plates, the difference 



