of Bright Lines in the Solar Spectrum. 55 



compounds. If therefore the temperature of the solar atmo- 

 sphere above the spot layer is low enough to permit of che- 

 mical combination taking place*, even when the portions of 

 the atmosphere concerned are swept down into the subjacent 

 spot-cavity, it follows that the layer into which oxygen ex- 

 tends (which, as we have seen, must be far above the spot 

 layer) would likewise be cool enough to allow of the formation 

 of compounds. 



7. It will help to give greater precision to the hypothesis 

 of a zone of combustion, if we follow the course of a ray of 

 light supposed to be emitted by the photosphere and received 

 in the spectroscope of a terrestrial observer. Passing through 

 the reversing layer, the ray undergoes that selective absorption 

 which gives rise to the Fraunhofer lines ; and if its spectrum 

 could be examined immediately after its emergence from this 

 layer, the oxygen- (and nitrogen-) lines would appear dark, 

 but less conspicuous than the metallic lines, for reasons which 

 will be entered into later on in this paperf. After traversing 

 the chromosphere the ray reaches the zone of combustion, in 

 which region, owing to the increased temperature, the lines of 

 all elements which extend so far would tend to be reversed 

 into bright lines of radiation J. 



[I say "tend to be reversed," because whether they would 

 actually become so depends upon the specific absorptive power 

 of the elements concerned for the rays in question. Thus, let 

 there be two gases, A and B, of which the spectral lines are 

 A a , Ajg, A y and B a , B^, B y respectively ; and let the specific 

 absorptive power of A be greater at a given temperature than 



* If these band spectra are regarded as the spectra of elements in the 

 stage of molecular complexity corresponding to the molecule giving the 

 band spectrum of iodine, or Roscoe and Schuster's new spectra of Na and 

 K (Proc. Roy. Soc. xxii. p. 3G2), the argument remains unaffected, since 

 these band-spectrum-giving molecules are spectroscopically equivalent to 

 the molecules of compound bodies. 



t The question here arises as to what order of oxygen-spectrum we 

 should expect to find at the temperature of the reversing layer. Dr. 

 Schuster seems inclined to believe that the temperature may be such as 

 to give the "compound 7 ' spectrum of this gas ('Nature/ vol. xvii. p. 148). 

 The recent observations of Lockyer upon the calcium-spectrum (Proc. 

 "Roy. Soc. xxiv. p. 352) tend to show that the temperature of this layer 

 is intermediate in dissociation-power between that produced by a small 

 coil with jar and a large coil with jar, a temperature which I am disposed 

 to believe would produce a state of molecular dissociation corresponding 

 to the line-spectrum of oxygen. 



X It is possible that the temperature of the chromosphere may fall off 

 at some particular level, so as to give above such boundary the " com- 

 pound" oxygen-spectrum. Should this be the case, the higher portion of 

 the chromosphere may obviously be left out of consideration, so far as re- 

 lates to its absorbing action on the line-spectrum of oxygen, 



