Radiation and Tonization of Helium, 769 



Direct test of the influence of impurity in producing radiation 

 from 20*4: volts electron collisions. 



Franck- and Knipping state that their experiments show 

 that in pure helium no bend occurs in the current-potential 

 difference curve until the electron energy reaches 21*2 volts, 

 while in impure helium a marked bend in the curve occurs 

 when the electron energy is only 20*4 volts. Our experi- 

 ments, though they do not confirm the conclusions of Franck 

 and Knipping, show that the bend which we obtain at 

 20'4 volts is much more strongly marked in the presence of 

 impurity than it is in pure helium, a fact which we explain 

 as being due to the ionization of the impurity by the 

 20°4 volts radiation. According to Franck and Reiche, and 

 to Franck and Knipping, the absence of a marked bend at 

 20'4 volts in their pure helium curves is due to the fact that 

 the outer electron of the helium atom, which, as a result of a 

 collision with an electron having 20*4 volts energy, is 

 removed to an orbit coplanar with that of the inner electron, 

 cannot revert directly to the normal condition and emit 

 radiation. Indirect methods by means of which the return 

 of the electron to the normal condition can be brought about 

 are the following : — 



1. By the ionization of the metastable atoms and sub- 



sequent recombination. 



2. By the metastable atoms forming short-lived chemical 



compounds, the formation of which generates electric 

 fields of sufficient magnitude to affect the limitations 

 of the selection principle. 



The second of these indirect methods is supposed to occur 

 to a considerable extent in the presence of impurity, but in 

 pure helium the only combination possible is the formation 

 of He 2 molecules, and, according to Franck and Knipping, 

 only weak electric fields would be generated by this combi- 

 nation, so that the reversion of coplanar atoms to the normal 

 condition would not occur. The experiments of these 

 writers, and those already described by us, have shown that 

 the first indirect method of reversion to the normal con- 

 dition, given above, does not occur to any extent as the 

 result of electron impacts on 204 volts abnormal atoms. 



In pure helium, bombarded by electrons with energies 

 intermediate to 20*4 volts and 21*2 volts, there must there- 

 fore be, on the view of the writers mentioned above, an 

 accumulation either of metastable atoms or of He 3 molecules. 

 The admission of a little impurity to an accumulation of 



