of Bright Lines in the Solar Spectrum. 59 



bable that the vast thickness of this gas which must be tra- 

 versed by a ray of light emitted by the photosphere should be 

 barely sufficient to reverse its spectrum. If the existence of 

 a zone of combustion be granted, however, this region becomes 

 the source of radiation of all gases which extend so far] Thus 

 in the case of the D 3 element, which reaches nearly the same 

 level in the sun's atmosphere as hydrogen, the stratum of gas 

 exterior to the zone of combustion is, on the present view, 

 alone concerned in reversing the line under consideration ; and 

 this stratum may be of insufficient thickness to produce any 

 marked absorption. The " 1474" substance, however, which 

 rises far above hydrogen, appears to exist in sufficient quan- 

 tity exterior to the supposed region of combustion (or its spe- 

 cific absorptive power is sufficiently great) to produce a 

 marked reversal in the solar spectrum*. 



The hypothesis advanced in the present paper does not ne- 

 cessarily imply (at least under existing solar conditions) the 

 production and accumulation of large quantities of compound 

 bodies in the higher regions of the sun's atmosphere. The 

 zone of combustion may be, so to speak, only a local pheno- 

 menon confined to a thin shell of the sun's outer envelopes ; 

 and compounds formed would be rapidly decomposed both by 

 dissociation and chemical reduction by being swept down into 

 the underlying hotter regions by the convection-currents which 

 take place on such an enormous scale in the sun's atmosphere. 

 The heat of the zone of combustion may also contribute to 

 the dissociation of compounds formed therein "(*. 



* Lockyer lias recently shown (Comjrt. Rend. Ixxxvi. 819 ; Proc. Roy, 

 Soc. xxvii. 282) that the blue line of lithium (w.-l. 4603) is represented 

 in the solar spectrum, while the red line (w.-l. 6705) has not hitherto 

 been detected. The question suggests itself whether the absence of this 

 last line may not also be connected with the existence of a region of com- 

 bustion. The low atomic weight of lithium would lead to the belief that 

 this element extends to a great height in the solar atmosphere. Thus the 

 zone of combustion might be the source of lithium-radiation, and at the 

 temperature of the sun the blue line may be the longest (as appears pro- 

 bable from the fact that this line requires a high temperature for its de- 

 velopment) ; so that the vapour above the region of combustion may be 

 sufficient to reverse the blue, but insufficient to reverse the shorter red 

 line. I would here ask whether the bright red line so frequently seen in 

 the spectrum of the chromosphere by Lockyer (Phil. Trans. 1869, pp. 

 428 and 129), and described as being less refrangible than C, may not be 

 the missing lithium-line ? I may add that a line less refrangible than C 

 has also been frequently seen by Respighi at the base of prominences. It 

 is highly significant that during the eclipse of 1808 a blue line between F 

 and G was seen by Rayet in the spectrum of a prominence. This is the 

 position that would be occupied by the lithium-line w.-l. 4603. 



t See Bnnsen's experiments on the combustion of different mixtures of 

 CO and II with O (Pogg. Ann. exxxi. 161); also Berthelot "On the 

 Chemical Equilibrium of C, II, and O " {Bull, Soc, Chim. [2] xiii. 99). 



