50 Mr. lv. Meldola on a Cause for (he Appearance 



that of 13. Imagine A and B to be raised to incandescence, 

 and placed in front of a source of white light at a higher tem- 

 perature, and let this combination be called the "first system." 

 On examination we should see the continuous spectrum crossed 

 by dark lines, A a , A^, A y , B a , B^, and B v , of which the first 

 series would be darker than the second. Now conceive the ra- 

 diation of the whole system to be weakened by general absorp- 

 tion or by removal to a distance. The lines of B would first 

 disappear ; so that if we imagine a mixture of A and B ("second 

 system ") to be heated to incandescence and placed between 

 the first system and the observer, the B lines might appear 

 bright on a background of continuous spectrum, while the A 

 lines remained dark, although weakened by the radiation of 

 the second layer of mixed gases.] 



Thus, if the sun's envelopes exterior to the zone of combus- 

 tion could be stripped off, we should see the solar spectrum 

 with the lines of oxygen (and nitrogen) bright, and the hy- 

 drogen-lines probably dark but much fainter than now seen. 



8. The reversal of the oxygen- (and nitrogen-) lines into 

 bright lines by the increased temperature of the region of 

 combustion is rendered possible, even with the intense light of 

 the photosphere as a background [and if, as most probably 

 would be the case, the temperature of the said region of com- 

 bustion is lower than that of, the photosphere], because the 

 light radiated by the latter has undergone almost its maximum 

 amount of weakening before reaching the zone of combustion, 

 not only on account of the distance of this last region from the 

 photosphere, but also because of the absorption, both selective 

 and general, which the light has undergone in passing through 

 the intervening reversing layer and chromosphere. 



9. We have next to turn our attention to that part of the 

 sun's atmosphere exterior to the zone of combustion, in order 

 to account for the fact that the hydrogen-lines appear so in- 

 tensely dark while the oxygen-lines are bright. The explana- 

 tion which I venture to suggest is based upon a wide survey 

 of the general spectroscopic characters of the elements. 



10. At the temperature of incandescence, the characteristic 

 lines in the spectra of any elements which are compared may 

 be of very different intensities. Thus Cappel has shown*, by 

 a series of quantitative determinations made at the temperature 

 of a Bunsen burner and of an induction-spark, that very dif- 

 ferent amounts of the metals experimented upon can be de- 

 tected by means of the spectroscope. The characteristic lines 

 of an element are those which Lockyer has shown to be the 

 longest. Interpreting such facts by the aid of the molecular 



* Pog-g-. Ann. exxxix. p. G28 (1870). 



