752 Prof. F. Horton and Miss A. C. Da vies 



on 



observations o£ Fricke and Lyman * on the extreme ultra- 

 violet spectrum of helium. 



The experiments to be described in this paper have yielded 

 results which are not in entire agreement with the con- 

 clusions of Franck and Knipping, and further reference to 

 their work will be made. It is sufficient to mention here 

 that there does not seem to the authors to be any justification 

 for the assumption that the bends in their curves at 20*4 volts 

 and 21*2 volts are due to the measurement of radiation 

 currents rather than ionization currents, and in view of the 

 results recorded in this paper it seems probable that the 

 detection of a bend at 20*4 volts in impure helium while no 

 marked bend is obtained in the pure gas until 21*2 volts is 

 reached, is due to the fact that in the former case the 

 20*4 volts radiation ionized the impurity, and led to the 

 detection of a large ionization current, while in the latter 

 case the detection of the radiation depended upon the photo- 

 electric effect it could produce on the collecting electrode, 

 which effect may have been too small to be easily measurable 

 with the galvanometer used. Reasons are given in the course 

 of the present paper for the view that in both cases the 

 bend at 21*2 volts denotes an increase of ionization. 



As in the experiments described in our earlier paper 

 we undoubtedly obtained radiation at 20'4 volts in helium, 

 which, from the precautions taken, must have been quite as 

 pure as that used by Franck and Knipping, we determined 

 to reinvestigate the phenomenon, looking carefully for a 

 second radiation-point near to the first one, and in addition 

 to investigate the magnitude of cumulative effects. In our 

 earlier research the experiments to test for ionization were 

 all performed with a view to detecting it as a direct result of 

 a single electron impact, and the production of ionization as 

 a result of cumulative effects was not investigated. The new 

 experiments have confirmed our original conclusion that 

 radiation is produced in pure helium by the impact of elec- 

 trons with 20'4 volts energy. Thev have also verified the 

 conclusion of Franck and Knipping that there is a second 

 radiation point at 21*2 volts, and have further shown that 

 both types of radiation can be absorbed by normal helium 

 atoms and re-emitted, so that when radiation is produced in 

 helium it is handed on trom atom to atom throughout a 

 volume of the gas. The possibility of the occurrence of 

 this effect was first drawn attention to by Richardson and 

 Bazzonif, but we believe that it is demonstrated here for 



* H. Fricke and T. Lyman, Phil. Mag. xli. p. 814 (1921). 



t 0. W. Richardson and C. B. Bazzoni, ' Nature,' xeviii. p. 5 (1916). 



