930 Prof. Horton and Miss Dalies on Electron Velocities 



a sudden large drop in the magnitude of the current occurred. 

 The brighter principal series lines were however still visible, 

 and these did not all disappear until the electron velocity was- 

 20*3 volts, when the yellow line X 5852 vanished. With an 

 electron velocity of 24 volts the green lines X 5331 and 

 X 5341 which are two o£ the brightest of the subordinate 

 series lines, were as bright as the principal series green line 

 X 5401. From curve A it may be seen that X 5401 was 

 visible when the current measured in arbitrary units was Q'^S, 

 whereas X 5331 and X 5341 were still not visible when the 

 current measured in the same units was 600. The fact that 

 X 5331 and X 5341 and the other subordinate series line& 

 require a higher electron velocity for their production than 

 do X 5401 and the other principal series lines, is thus clearly 

 not due to a mere difference in the relative intensities of the 

 two sets of lines. The results of many series of observations- 

 of the changes in the spectrum with increase of the electron 

 velocity seemed to indicate that a minimum velocity of about 

 23 volts was required to produce the subordinate series lines. 



The set of lines which appeared at the higher electron 

 velocity contained, in addition to the brighter lines (given in 

 Table IIL), all the other subordinate series lines in Table I.. 

 These made their appearance a small fraction of a volt after 

 the brighter lines were visible, when the general brightness 

 of the luminosity was high. Thus, although the principal 

 series lines are produced at 20 volts and the subordinate series 

 lines at 23 volts, the fainter members of these series in each 

 case require for visibility a greater concentration of radiant 

 atoms than exists when only the swiftest members of the 

 electron stream have the necessary minimum velocity. 



From the results of these experiments, and from our 

 previously determined values of the ionization velocities for 

 electrons in neon, it seemed probable that the appearance 

 of luminosity in the gas was connected with the 200 volt 

 ionization, but that the production of the complete neon 

 spectrum required the third type of ionization, found in our 

 earlier work to begin at 22'8 volts. Some confirmation of 

 this view is provided in the curves already given in fig. 2. 

 In both of these curves the limiting electron velocity for the 

 visibility of any. subordinate series lines is practically coin- 

 cident with the beginning of an alteration in the rate of 

 change of the ionization current with change of electron 

 velocity, such as is usually taken to indicate a change in the 

 type of ionization occurring. Moreover, the bend in the 

 curve occurs at about 23 volts, the velocity at which the third 

 type of ionization begins. In order to test further our view 

 regarding the connexion between the ionization velocities 



