GASES IN VACUUM-TUBES. 309 



The dispersed infra-red radiation from gases in a vacuum-tube is of 

 considerable interest in comparison with the radiation of gases in a 

 flame. The work of Julius and of Paschen (loc. cit.) shows that by 

 mere temperature elevation gases emit characteristic discontinuous 

 spectra. From a theoretical standpoint such an investigation is also of 

 importance, since all the infra-red lines predicted by our spectral series 

 end in the short wave-lengths just beyond the red. This presents the 

 interesting question whether emission lines are to be found beyond the 

 region of 2 fi. Any information on this subject will aid in placing the 

 numerous spectral series formulae upon a firmer, less empirical basis. 

 In this connection a recent theoretical paper by Garbasso* is to be 

 noticed, in which he maintains that the series of Kayser and Runge have 

 a real existence. 



In the preliminary part of this work the vacuum-tube used by Drew 

 (loc. cit.) was employed. In this tube the electrodes were at right 

 angles to and at a short distance from the main part, which formed the 





Fig. 141. 



positive column. Several rubber joints were in series with the mercury 

 pump, McLeod gauge, and the vacuum-tube. Several gases were exam- 

 ined, all of which, except hydrogen, showed an emission band in com- 

 mon at 4.75 [x. Hoping to gain intensity of the radiation, a tube was 

 made similar to one used by Runge and Paschen (loc. cit.). (See fig. 

 141.) The aluminum electrodes were 2.5 cm. long, and 2.8 cm. in diam- 

 eter, through which the radiation passed axially into the spectrometer slit 

 of 8 mm. length and i mm. width. The radiometer slit was 0.7 mm. 

 The internal diameter of the main portion of the vacuum-tube was i cm. 

 and its length was from 15 to 18 cm. The window was of rock salt, 

 secured on the outside by means of beeswax covered with shellac var- 

 nish. The window was also secured by means of shellac which had 

 been boiled until it thickened. To avoid rubber joints the vacuum-tube 

 was sealed to the pump and the gases introduced through a barometric 

 column of mercury. The tube was airtight but, at a low pressure, it 

 would become so hot that vapors would be given off and the cathode 

 luminescence would disappear. This luminescence would reappear 

 after the tube had cooled. According to Travers (Study of Gases) this 



^Garbasso : Nuovo Cimento, 9, p. 113, 1905. 



