Ionization of Platinum by Cathode Rays. 287 



speed, falls to a minimum at 10 or 11 volts, increases again 

 rapidly to a maximum at about 200 volts, and then falls 

 gradually and continuously. When the speed of the incident 

 rays is between and 11 volts, the speed of the electrons 

 leaving the plate depends wholly on that of those rays; the 

 number of electrons with any speed between and that of 

 the incident rays is approximately the same, but increases 

 slightly as the speed approaches that of the incident rays. 

 On the other hand, when that speed is greater than 40 volts, 

 the speed of the electrons leaving the plate is independent of 

 that speed ; almost all the electrons have speeds less than 

 10 volts, and the number of these electrons with any speed 

 increases as the speed decreases. There is a sudden change 

 in the " speed curve " o£ the electrons when the speed of the 

 incident rays passes through 11 volts. 



These observations are explained by the supposition that 

 there are two distinct processes to be taken into account, 

 a reflexion of the incident rays and an excitation by them 

 of secondary rays. The first process alone is operative when 

 the speed of the incident rays is less than 11 volts, the second 

 comes into operation at that speed, and becomes of increasing 

 importance relatively to the first as the speed increases, until 

 the maximum at 200 volts is passed. It has been shown that 

 the distribution of the speed of the electrons leaving the plate 

 under the action of incident rays having a speed of several 

 thousand volts, is precisely the same as that of the "delta 

 rays " which leave the plate when alpha rays fall on it. This 

 observation and evidence from other sources has led to the 

 belief that the emission of the secondary rays by fast 

 primary rays is a process altogether analogous to the ioni- 

 zation of a gas by such rays; the critical speed 11 volts 

 represents the "ionization potential" of the metal, a quantity 

 precisely analogous to that which Franck and Hertz * have 

 recently determined with great care for several gases; the 

 maximum at 200 volts represents the speed of the rays which 

 have maximum ionizing power. 



2. Since the ionization potential varies considerably (from 

 20*5 to 7*5 volts) with the nature of the gas, while the 

 ionization potential of metals appears to be independent of 

 the nature of the metal, the suggestion naturally occurs that 

 the ionization which occurs at the surface of the metal takes 

 place, not in the atoms of the metal itself, but in a layer of 

 gas, the same for all metals, condensed on the surface; and 

 this suggestion seems to be confirmed by the observation 



* Ber. DeutscJt. Phys. Gesell. xv. 2, pp. 34- 44 (1913). 



