Cady — On the Energy of the Cathode Hays. 13 



direct ; moreover, since an induction coil was used as source of 

 discharge, one would expect the formation of a " magnetic 

 spectrum,"* which would greatly complicate the appearance of 

 both spots. Thus it is easy to see that a systematic error may 

 have crept into the results. 



Whether such an error can have been great enough to give 

 the value r' = 1 instead of r' = 0*7 is, of course, not certain ; 

 but it may be added that the energy of the cathode rays is pro- 

 portional to the square of their deflectibility, hence an error of 

 n per cent in the measurement of the deflection would corre- 

 spond to an error of Sn per cent in the energy. In any case 

 it is desirable to determine both the deflectibility and the 

 energy of reflected cathode rays by as accurate a method as 

 possible. 



My thanks are due to Prof. Warburg for his kind advice and 

 assistance during this investigation. 



* K. Birkeland, Comptes Rendus, cxxiii, p. 492, 1896. 

 Providence, E. I., May 5, 1900. 



