[hughes] 



RADIATION FROM BORON 



The important point, however, is that no matter whether the dis- 

 continuity was pronounced or not, it always appeared at the same 

 place. Fig. 2 gives the results obtained on one occasion. It will 

 be seen that on this day a progressive increase in the effect took place. 

 Whether this was due to an increase in the amount of radiation 

 emitted per unit electron current, or w^hether the nickel plate increased 

 in sensitivity, is not known. However, the discontinuity in the curve 

 always occurred at about 150 volts. The upper curve is the mean of 

 the four runs shown and shows a well-marked discontinuity at 150 

 volts, which corresponds to a wave length X82.3. 



I 20 - 



mo - 



150 aoo 



F/G. 2 



Z50 



yOLTS 



Discussion. — In Fig. 3 the square roots of the voltages corre- 

 sponding to the K-critical absorption wave lengths of Mg, Al, P, and 

 S, as determined by the usual crystal grating methods, are plotted 

 against the atomic numbers, giving almost exactly a straight line. 

 This line is produced (assuming the same very slight departure from 

 the linear relation to hold for elements lighter than magnesium as 

 for those heavier) so as to indicate the values to be expected for the 

 lighter elements. The value to be expected for boron is 149 volts 

 which is in good agreement with that obtained in this investigation. 



—23 



