726 Elecbrodeless Ring Discharge in certain Gases. 



the highest electric intensities used, the band spectrum at 

 intermediate intensities, and the compound line spectrum at 

 the lowest intensities. The only possible difference between 

 the results obtained by the two methods is connected with 

 the continuous spectrum, which in the discharge tube appears 

 at the lowest intensities and in the ring discharge at an 

 intensity between those for the compound line and the band 

 spectra. As considerable doubt has been cast upon the 

 existence o£ a continuous spectrum in pure oxygen in a 

 discharge tube *, it is probable that the one observed in the 

 ring discharge is independent of that in the tube discharge. 

 The existence of a band spectrum occupying a position 

 intermediate between two line spectra o£ the same element 

 is interesting in view of suggestions that have been made to 

 ascribe line spectra to atomic, and band spectra to molecular, 

 systems. 



Another case which is interesting from the point of view 

 of the ring discharge is that of the Swan spectrum and the 

 carbon band spectrum. The Swan spectrum only appears in 

 carbon monoxide when it is very nearly pure, and then at a 

 high intensity, and we may have the same bulb showing the 

 carbon band spectrum at a low intensity and the same, mixed 

 with the Swan spectrum, at a higher intensity. As, under 

 the influence of the discharge, carbon dioxide would dissociate 

 into carbon monoxide and- oxygen, it would obviously be very 

 difficult in this case to get a sufficient electric intensity in the 

 carbon monoxide itself to cause the appearance of the Swan 

 spectrum. This would also be the case with any impure 

 carbon monoxide, and therefore it would seem possible that 

 the carbon band spectrum and the Swan spectrum may be the 

 low and high intensity spectra of the same substance, though 

 these experiments afford no clue as to whether that substance 

 is carbon monoxide, or carbon in some other form. 



In the cases of nitrogen and residual air the ring discharge 

 is peculiar in that it has never shown any clear evidence of 

 the presence of the negative band spectrum, the variation 

 from the positive band spectrum to the line spectrum taking- 

 place directly. A bulb containing cyanogen, which was not 

 sufficiently pure to give its own band spectrum, showed a 

 change from the positive to the negative nitrogen band 

 spectrum with increasing electric intensity, but this was 

 never observed in nitrogen itself. The general results agree 

 with those given by discharge tubes in assigning the positive 

 band spectrum in nitrogen to the weakest ionizing fields, and 



* Baly, ' Spectroscopy,' p. 443. 



