for Production of Luminosity in Atmospheric Neon. 933 



the luminosity when the magnetic field was not on, extended 

 throughout the whole of the space between the grid and 

 anode instead of being concentrated into a bright central 

 column. r l he lines seen through the spectroscope were 

 very much less intense in the absence of the magnetic 

 field, and in some cases, even though the luminosity could 

 be seen faintly at electron velocities between 20 volts and 

 21 volts on observing the apparatus directly, no line but 

 \ 5852 could be seen through the spectroscope until the 

 ionization velocity 22"8 volts was exceeded and the second 

 group of lines appeared ; the remaining lines of Table II. 

 then became visible also. This was the case in the series 

 of observations represented in fig. 4. The curve shows 

 ionization beginning at about 17 volts, an increased rate 

 of production of ionization beginning at 20*2 volts and 

 -another such increase at 23T volts. At 20'8 volts a faint 

 luminosity was observed in the tube, and the yellow line 

 X 5852 was seen through the spectroscope at 2T9 volts. No 

 other lines appeared until 23*4 volts, at which velocity all the 

 brighter green and red lines of Tables II. and III. could be 

 seen faintly. After the completion of these observations, 

 the spectrum was observed with the magnetic field on and 

 with an electron velocity of 22*5 volts, i. e. with the velocity 

 just below that at which the third type of ionization begins. 

 All the brighter principal series lines, including the green 

 line X 5401, were distinctly visible and were much brighter 

 than they had previously been at 24 volts ; but none of the 

 subordinate series lines were present. From these and 

 other similar curves it was concluded that a minimum 

 electron velocity of 22'S volts is required for the production 

 of the subordinate series lines, but that the principal series 

 lines can be stimulated by electrons having a smaller velocity 

 than this. 



The fact that the lines of Table II. are generally first 

 observed at some velocity after the 20'0 volts ionization 

 has been detected, but before the 22*8 volts ionization 

 occurs, combined with the fact that no luminosity was ever 

 observed below 20*0 volts, constitutes fairly conclusive 

 •evidence that the stimulation of these principal series lines 

 results from the occurrence of the 20*0 volts type of 

 ionization and not from the 16*7 volts type of ionization. 

 Further evidence on this point was obtained from several 

 series of simultaneous observations of the ionization current 

 and the spectrum of the luminosity taken with different 

 intensities of electron stream and gas pressure. In one of 

 these series the value of the ionization current when there 



