Electrodeless Ring Discharge in certain Gases, 723 



four hydrogen lines appeared at every electric intensity that 

 could be used, though at the highest pressures for the 

 discharge the secondary spectrum was also quite bright. In 

 the inner ring at the pressure for the best discharge, using a 

 short spark, only EL and H were present. On increasing 



the spark length H^ and H^ appeared and became brighter 

 as the spark was still farther increased This association 

 of H a with H^ and EL with H had been observed previously 



on several occasions with preliminary residual air-bulbs, but 

 in those cases it was always H a and H^ which appeared with 

 the shorter spark gap. 



Carbon dioxide : — The gas was prepared by the action 

 between a solution of pure sodium carbonate and pure dilute 

 sulphuric acid. As in almost every gas, the discharge at the 

 highest pressures was always reddish in colour, probably 

 owing to the presence of a trace of water vapour, but the 

 true colour of the carbon dioxide discharge was dark blue, 

 changing to whitish blue as the pressure decreased, or as the 

 spark-gap and frequency of sparking increased. 



The spectrum observed was, in every case, the carbon band 

 spectrum although all possible variations of electric intensity, 

 by varying spark-length, spark frequency, and size of leyden- 

 jars were tried. 



Carbon monoxide : — The gas was prepared by the action 

 between pure strong formic and sulphuric acids, and on its 

 way to the discharge bulb passed through a spiral tube 

 immersed in liquid air, to remove such impurities as mercury 

 vapour. 



Until the apparatus had been washed out some few times 

 with carbon monoxide, the spectrum observed was always 

 the carbon band spectrum. When the carbon monoxide was 

 more pure, the spectrum of the discharge, when a small spark- 

 gap and small leyden-jars were used, was still the carbon 

 band spectrum, but as the spark-length was increased to 

 about 4 mm. the Swan spectrum began to appear and grew 

 in brightness relative to the carbon band spectrum as the 

 spark was still further lengthened. The appearance or 

 disappearance of the Swan bands could be obtained by 

 altering the spark-gap about I mm. at the critical length. 



At lower pressures, the Swan bands were obtained without 

 so great a lengthening of the spark-gap. The spectrum was 

 the same when large jars were used, but in this case the 

 spark-length could not be made so great and the Swan bands 

 were never quite so bright as with the smaller jars. 



Oxygen. — The gas was prepared by heating pure potassium 



