722 Mr. H. Donaldson on the Spectra of the 



repeated, and the effect of variation in the length of the 

 spark-gap and the frequency of sparking was studied. 



In all cases, at pressures considerably above the highest at 

 which the ring discharge took place, a feeble electrostatic 

 discharge was seen in the bulb, due to the windings of the 

 surrounding coil not being all in one plane, but this was 

 never used in observations of the spectra. 



Results. 



Residual Air : — The apparatus was washed out .several 

 times with air before exhaustion. The spectrum of the 

 discharge was chiefly the positive band spectrum of nitrogen 

 at the highest pressures at whicli the true ring discharge was 

 obtained, and the colour of the ring was red. At lower 

 pressures many of the nitrogen lines appeared and became 

 very bright as the lower limit of pressure was reached. The 

 negative band spectrum never came out distinctly, even if it 

 were present at all. 



Air in which phosphorus had been burned to remove 

 oxygen was then tried and gave the same spectra, a spiral 

 immersed in liquid air being inserted between the inlet 

 apparatus and the bulb to remove the phosphorus and 

 mercury vapours. 



Both with residual air and with this deoxygenized air there 

 was an orange after-glow, which persisted brightly for about 

 half a minute after the electric field was removed, and faintly 

 for a much longer time. This afterglow was observed down 

 a tube about a metre long in line with the collimator of the 

 spectroscope. While the discharge was still running the 

 same orange colour could be seen in the centre of the bulb, 

 and its spectrum was continuous throughout the yellow-blue 

 region, but had two places of maximum intensity. As soon 

 as the discharge was cut; off, the continuous part of the 

 spectrum died out, but the regions of maximum intensity 

 remained as the spectrum of the true after-glow, in the form 

 of two bands, diffuse on either side and having their centres 

 approximately at wave-lengths 5760 and 5378 tenth-metres, 

 and having a breadth of about 10 such units. 



As the first occasion on which this afterglow was observed 

 was one on which there was a possibility of the presence of 

 phosphorus vapour, an entirely new apparatus was set up, 

 but the spectrum remained unchanged. 



Hydrogen : — The gas was prepared by the action of pure 

 dilute sulphuric acid, to which a trace of copper sulphate 

 solution had been added, upon pure metallic zinc. In the 

 outer of the two rings which constituted the discharge, all 



