of Bright Lines in the Solar Spectrum. 57 



theory of gases (and making cine allowance for the fact that 

 the characteristic lines of the spectra being compared may 

 occur in parts of the spectrum not visually comparable so 

 far as regards intensity), we should say that some kinds of 

 molecules can have certain internal vibrations more readily 

 excited than is the case with other kinds. From the relation- 

 ship which exists between radiation and absorption, it follows 

 that molecules which have the most sensitive radiative orga- 

 nization have likewise the most sensitive absorptive organi- 

 zation. 



11. The non-metals are distinguished, as a group, from the 

 metals by the greater complexity of their spectra (which more 

 resemble the band spectra of compound bodies), and also by 

 their comparative insensitiveness to the spectroscope. Many 

 of the metals are known to give band spectra at low tem- 

 peratures ; but these break up into line spectra at high tem- 

 peratures. On the other hand, the band spectra of many 

 non-metals bear temperatures high enough to break up the 

 band spectra of metals without being resolved into line spectra. 

 We might thus have a mixture of two vapours, one metallic and 

 the other non-metallic, at the temperature of incandescence, 

 the former giving a line spectrum and the latter a band spec- 

 trum. If we imagine the temperature of such a mixture of 

 vapours to be raised to the point at which the band spectrum 

 of the non-metal breaks up, we should get a line spectrum 

 from both elements ; but the metallic lines would be more 

 intense* than those of the non-metal, owing to the greater 

 sensitiveness of the metallic molecule. We should thus have 

 realized the conditions laid down in a former paragraph (7), 

 where A would then represent the metallic, and B the non- 

 metallic vapour. 



12. It now remains to show the applicability of the fore- 

 going principles to the case under consideration. 



The oxygen and hydrogen of the sun's atmosphere will, for 

 the sake of simplicity, be exclusively considered. These gases 

 represent the metallic and non-metallic vapours of the last 

 paragraph. The photosphere, reversing layer, and chromo- 

 sphere represent the " first system " of paragraph 7 — i. e. the 



* " In a tube containing both nitrogen and aqueous vapour, the lines of 

 hydrogen (spectrum II order) made their appearance at the same time as 

 the spectrum of hands (I order) of nitrogen, whence it follows that the 

 lines of hydrogen are visible in a temperature in which the lines of ni- 

 trogen do not appear" (Schellen's ' Spectrum- Analysis/ p. 171). So also 

 Frankland and Lockyer found that in a tube containing hydrogen and 

 nitrogen, the lines of the latter gas under certain conditions of pressure 

 could be made to disappear entirely, while the hydrogen-lines under all 

 conditions remained visible (Troc. Roy. Soc. xvii. p. 4o4). 



