﻿Emission of Electrons by X-Hays. 797 



In these experiments the surface of the cube was perpen- 

 dicular to the direction of propagation of the X-rays, and 

 therefore parallel to the direction of the electric vector in 

 the X-ray beam. It was possible, therefore, that there was 

 a large number of electrons whose initial direction was 

 nearly parallel to the surface of the cube. Such electrons, 

 even if their velocities were considerable, might be bent 

 back into the cube by electric or magnetic fields of relatively 

 sniill strength. This hypothesis was tested by comparing 

 the reducing effect of electric and magnetic fields-when the 

 anole of incidence was 90° and 45°. It was found that the 

 percentage reduction was the same in the two cases. Had 

 the effect been due to the bending back of electrons emitted 

 in directions approximately parallel to the face of the cube, 

 the reduction of the emission would have been larger in the 

 first case than in the second. 



The conclusion reached was that, in addition to the high- 

 speed electron emission, there exists also an emission of 

 electrons of low speed. After considerations such as these 

 had led the writer to this conclusion, it was found that 

 similar effects had previously been observed by Campbell * 

 in his work on delta rays. 



In what follows, first the properties of the high-speed 

 electrons and then those of the slow electrons will be 

 discussed. 



A. High-Speed Emission. 



(4) Relative Electron Emission from Metals. — 'The effect 

 of the low-speed electrons was eliminated by charging the 

 inner wire cylinder to a voltage sufficient to prevent their 

 escape from the surface of the cube. The residual effect was 

 then that due to the high-speed electrons. 



The metals investigated were Aluminium, Iron, Nickel, 

 Copper, Silver, Tin, Grold, Lead, and Bismuth. The X-ray 

 tube was operated under varying conditions, and it was found 

 that the relative values for the various metals depended very 

 little on the conditions of the tube. This point will be 

 discu-sed more fully later. 



Table I. shows the results obtained, the value for the tin 

 being taken as 100. 



Table I. 



Radiator Al. Fe. Ni. Cu. Ag\ Sn. Au. Pb. Bi. 



Electron Emission 11 47 53 56 94 100 184 189 194 



Atomic Number 18 26 28 29 47 50 79 82 83 



Atomic Weight 27 56 59 67 108 119 197 207 208 



* N. R. Campbell, Phil. Mag. xxiv. p. 783 (1912). 



