290 INFRA-RED EMISSION SPECTRA. 



unknown. Since the effect upon the phosphorescent plate is cumulative, 

 as is true of the photographic plate, this method is highly desirable for 

 detecting weak lines, which could not be detected with a bolometer or 

 thermopile. For measuring the energy of the different lines, the bolo- 

 meter, radiometer, or the thermopile is the most useful. Julius,' using 

 a bolometer, applied this method to explore the emission spectrum of a 

 Bunsen flame, when different vapors, e. g., CS2 and Br, were burned in 

 it. His results are of considerable interest, especially for CS2, for 

 which he found emission bands whose maxima coincide with those of 

 the absorption bands, and will be noticed in the present work on the 

 radiation from gases in the vacuum-tubes. 



The infra-red spectra of the alkali metals were investigated by 

 Snow," using a glass prism and a bolometer. He used the chlorides 

 of the metals in an arc between hollow carbon electrodes. His observa- 

 tions extend to about 2 /x, and show no strong lines beyond 1.4 ju,. The 

 metals Rb and Cs show small lines up to i.y /x where Lehmann (loc.cit.) 

 found small lines for these same metals. Whether this absence of lines 

 beyond 1.5 /a was due to the opacity of the prism was not determined by 

 Snow. Lewis,' using a radiomicrometer and a large grating, examined 

 several alkali metals in the carbon arc. For Xa he found a doublet at 

 A = 0.81837 and A = 0.8194 /x. 



The emission bands of CO, and H^O (vapor) have been investigated 

 by Paschen.^ He found a strong emission band for COo which shifts 

 toward the long wave-lengths with rise in temperature, the maximum 

 being at 4.27/1, at 17° (absorption), at 4.3/^ for a temperature of 600°, at 

 4.388 fji for 1000° and 4.40 /* for the Bunsen flame, which has a tempera- 

 ture of at least 1800°. His method of observation consisted in heating 

 the CO, by passing it through an electrically heated platinum spiral, 

 and finding its emission just as it leaves the spiral. 



Rubens and Aschkinass' have also investigated CO2 and H^O (vapor) 

 to 20 /x, using a sylvite prism and a linear thermopile. They found a 

 small emission band of CO2 at 14. i /x. The gas was heated by passing 

 it through a metal cylinder, open at the ends and heated by means of 

 Bunsen flames. 



Other emission bands of HoO and COo will be referred to in the text, 

 and it will be sufficient to add that where one shifts toward the long 

 wave-lengths another shifts toward the short wave-lengths. 



^ Julius : Licht. u. Warmestrahluag verbannter Gase. Berlin, 1890. 



^Snow: Phys. Rev., i, p. 35, 1893. 



^Lewis : Astrophys. Jour., 2, p. i, 1895. 



*Pashen : Ann. der Physik (3), 53, p. 324, 1894. 



^Rubens & Aschkinass: Ann. der Physik (3), 64, p. 584, 1898. 



