of Gases e.ecited by Ultra-Schumann Waves. 905 



Fig. 11 (PL XI V.) shows the water-band streamer obtained 

 with a five minutes exposure when the jet-chamber was filled 

 with nitrogen. With the chamber filled with air and an 

 exposure of twenty-five minutes (fig. 12), we find that the 

 streamer is less intense. The presence of oxygen in the 

 nitrogen may be the cause o£ the water-band, but if much 

 oxygen is present (as in air) the band is enormously weakened. 

 Strutt has found that the "after-glow" of nitrogen disappears 

 if ox)rgen is present, and the fluorescence of iodine excited by 

 ultra-violet light is destroyed also by oxygen. 



Fluorescence of Moving and Stationary Gases. 



In our experiments of three years ago an attempt was 

 made to blow the jet to one side by a blast of air, for we 

 were of the opinion at that time that the luminosity might 

 be due to luminous corpuscles projected from the spark, or 

 carried up in the gas blasts projected through the slit by 

 the explosive discharges. 



It was found, however, that the air current merely inter- 

 rupted the jet, the luminosity remaining visible above the 

 moving stream of air. This effect is shown by the photo- 

 graph reproduced on PI. XV. fig. 13, made with higher 

 dispersion than the figures previously mentioned. The 

 streamer, which is seen distinctly cut in two by the air- 

 current, is that of the so-called fcC water-band," the fainter 

 nitrogen bands not showing. The position of the tube 

 delivering the air current is indicated. As we are now quite 

 sure that we are dealing with fluorescence produced by ultra- 

 Schumann waves, it will be unnecessary to speculate about 

 this phenomena from a corpuscular view-point. The air in 

 the moving current was the same as the air in the jet- 

 chamber, and the experiment establishes the fact that air is 

 fluorescent only when it is stagnant. In other words, it 

 seems as though a given mass of air must be acted upon by 

 the radiations from a number of successive sparks to attain 

 its full luminosity. This appears to be true, however, for 

 the water-bands only. For the nitrogen bands the reverse 

 is true : they are brighter if the gas is in motion. 



If a current of nitrogen is directed across the spark jet, 

 the water-band streamer is interrupted and a strong patch 

 of luminosity appears displaced towards one side (fig. 5, 

 PL XIV.). The direction of the displacement is inde- 

 pendent of the direction of the moving current of gas, i. e. 

 it is produced by the prism, as is shown by a comparison of 

 figs. 5 and 6. The jet-chamber was filled with the same 

 nitrogen as that which was flowing in a stream across the 



