2 THE ROYAL SOCIETY OF CANADA 



speed of the electrons was increased by increasing the potential 

 accelerating the electrons between T and B. The boron emitted 

 radiation (which may be regarded as extremely soft X radiation, or 

 as ultra-violet light of extremely short wave length, no distinction 

 can be drawn), part of which fell upon the nickel plate N. The nickel 

 plate emitted photo-electrons under the influence of the radiation, 

 the photo-electric current being measured by an electrometer. As 

 will be seen from the diagram the radiation had to pass through two 

 gauzes. The function of these gauzes was to prevent any electrons 

 or positive ions (if any) from the filament or boron, getting to the 

 nickel plate. The gauze D being at a negative potential with respect 

 to the filament and boron prevented any electrons getting to N, 

 while the gauze A, being at a higher positive potential than either 

 the boron or filament, prevented any positive ions passing to N. 

 Care was taken to secure as high a vacuum as possible by means of a 

 diffusion pump and a liquid air trap. The pressure during the 

 experiments could not be measured on the McLeod gauge and was 

 certainly less than 10'^ mm. 



Method of Experiment. — The procedure was to measure the 

 photo-electric current from N as the accelerating potential was 

 increased step by step, keeping the electron current constant, and 

 to look for a discontinuity in the curve connecting these two quantities. 

 From the work of Beatty and others on ordinary X-rays, the dis- 

 continuity is to be associated with the excitation of the radiation 

 characteristic of the material of the target. Below the critical 

 accelerating potential nothing but "general" radiation is excited, 

 above the critical potential, "characteristic" radiation is emitted 

 in addition. If the accelerating potential necessary to call out the 

 characteristic radiation is V, then the frequency v, and the wave 

 length X, of the radiation are given by 



Ve = hj' = hc/\ 

 where e is the charge on the electron, h is Planck's constant, and c 

 the velocity of light. Expressed in volts and Angstrom units, this is 



V=12331/X. 



Experimental Results.- — It was found impossible to get reproduc- 

 ible curves on difïerent days, even though the heat treatment of the 

 apparatus at the beginning of each set of observations and the method 

 of exhausting were always carried out in the way. On some occasions 

 very marked discontinuities in the curves were obtained, on others 

 the discontinuities were definite but not so well marked, while some- 

 times the curves appear to be almost free from any discontinuity. 



