of Gases excited by Ultra- Schumann Waves. 907 



many o£ the apparently inexplicable phenomena found in 

 this preliminary investigation. 



It appears to be pretty definitely established that oxygen 

 has a tendency to destroy the fluorescence of gases with 

 which it is mixed. This accounts for the extreme Eaintnessof 

 the water-band streamer in oxygen. The oxygen molecules 

 destroy their own fluorescence, so as to speak, the phenomenon 

 being analogous to the one observed by one of us in the case 

 of iodine *. 



A study of the fluorescence of iodine vapour under the 

 influence of the spark rays has been commenced by one of 

 the writers and will be reported in a subsequent paper. 



Delivered in a stream of warm nitrogen across the spark- 

 jet it fluoresced with a bluish green light, while if a 

 current of warm air was used there was no trace of any 

 fluorescence. If a thin plate of quartz was placed over the 

 slit, the fluorescence was visible for a few seconds and then 

 rapidly faded away. If, however, the quartz plate was 

 moved a little, the fluorescence appeared again. It was 

 found that the opacity of the quartz was produced by an 

 almost imperceptible film deposited upon it by the spark. 



We had made many experiments with thin plates of quartz 

 and fluorite, but had failed to find any trace of the water- 

 band streamer, though we obtained faint indications of the 

 coal-gas streamer and possibly those of G0 2 . 



The observations made with iodine threw a new light on 

 the subject and made it necessary to repeat the experiments 

 under conditions which would preclude the formation of the 

 opaque deposit. This was accomplished by moving the 

 quartz plate about during the exposure. Using this pre- 

 caution the nitrogen streamer was obtained by giving an 

 exposure of only five minutes. Its intensity was equal to 

 that obtained without the quartz plate with an exposure of 

 one minute. No trace of the water-band streamer was 

 found, however. 



A fluorite plate 1 mm. thick was next tried, and a faint but 

 unmistakable image of the water-band streamer was obtained 

 with a fifteen minutes exposure. Without the fluorite plate 

 a stronger image of the streamer appeared with a one minute 

 exposure. A rough estimate indicates that the fluorite plate 

 reduces the intensity of the radiations which are responsible 

 for the excitation ol the water-band to about 5 per cent, of 

 their original value. This circumstance makes it seem 

 extremely probable that we are dealing with ultra-violet waves 

 much shorter than the Schumann waves, which pass readily 



* Wood & Speas : " A Photometric Studv of the Fluorescence of 

 Iodine Vapour," Phil. Mag. Mar. 1914, p. 531. 



f 



