of the Ionization Potential of Helium. 



567 



as the most probable value. All of these values are too low 

 to fit into Bohr's theory. Bohr has suggested that the effects 

 observed in the experiments might have been due to ioniza- 

 tion o£ impurities by radiation from the helium under 20*5- 

 volt bombardment, or to the liberation of electrons from the 

 metal parts of the apparatus from the same cause. The fact 

 that the characteristic frequency of helium as calculated by 

 Cuthbertson * from dispersion data (59 x 10 15 ) corresponds 

 very closely to 20 5- volt impacts, lends some support to this 

 explanation. Further, certain experiments of Bau's f in 

 which he finds 30 volts necessary to excite the ordinary 

 many-lined spectrum of helium and 80 volts necessary to 

 excite X4686 have been taken, although without sufficient 

 justification J, to imply that Bohr's calculated value is the 

 correct one. In view of the remarkable agreement with fact 

 of many of the deductions from this theory of Bohr's, it has 

 seemed worth while to re-determine the ionization potential 

 of helium, avoiding the suggested possibilities of error, and, 

 further, in view of the extraordinary susceptibility of helium 

 to disturbances due to minute traces of impurities, with an 

 arrangement such that all traces of occluded gases could 

 positively be got rid of in the first place and the gas sub- 

 sequent]} be maintained absolutely without contamination. 

 The apparatus used in this investigation has been designed 

 with these ends in view. 



Fig. 1. 



The experimental tube, which is constructed of transparent 

 quartz, is shown in fig. 1. The filament F, which is of 



* Proc. Eoy. Soc. 1010. 



f Sitz. der Phi/s. Med. Ges. zu JJlirzhio-r/, Feb. 1914. 



X Richardsou & Bazzoni, i Nature/ Sept. 7, 1015, 



in 



.! 



