76 Colours and their Relations. [January, 



thus altered were the three magnesium lines b I} b 2 , b A , the line 

 63 referred to iron and nickel, and the line 1474 of Kirchhoff's 

 map also due to iron. But the strong magnesium line 5527 

 of Angstrom's scale and the numerous other iron lines under- 

 went no similar change. It might be hence fairly inferred that, 

 in the photosphere, the atoms constituting the ultimates of 

 the various chemical elements, whose characteristic lines 

 have been detected in the solar spectrum, all exist in a 

 disunited state, forming an intimate mixture of .highly 

 attenuated gases ; but that some of those gases occasionally 

 pass the limits of the photosphere, and are projected a short 

 way into the chromosphere, where they glow under the 

 influence of electrical currents. For the bright lines, thus 

 forming reversals of some of the dark lines of the photo- 

 sphere, are always much shorter than the other bright lines 

 of the chromosphere — a fact indicating that the substances 

 which produce them ascend only a short way beyond the 

 usual limits of the photosphere, in which these same lines 

 are dark. 



Subtle as, according to this view of their constitution, 

 these gases must be, they must be excelled in their tenuity 

 by others, which, extending beyond the chromosphere, form 

 the corona seen in total eclipses of the sun. It is remark- 

 able that the spectrum of the corona, as observed by Pro- 

 fessor Young during the total eclipse of 1869, consists of 

 three bright lines, so nearly coincident with those observed 

 by Professor Winlock in the spectrum of the aurora borealis 

 as to leave little doubt of their identity — thus indicating that 

 the gases constituting the solar corona exist also in the 

 region at the outskirts of the earth's atmosphere, where the 

 auroral flashes play. These spectral lines of the solar 

 corona and the aurora borealis have not yet been identified 

 with any known spectra produced by artificial means. But 

 the extreme lightness of the gases producing them renders 

 it probable that, like the gases of the chromosphere, they 

 consist of separate atoms not united into the ultimate of any 

 chemical element. Should the individual lines be hereafter 

 identified with any of those embraced in the spectrum of 

 any known element or elements, this view of their constitu- 

 tion would be confirmed. (See Schellen's " Spectrum 

 Analysis," pp. 361, 399, 404, 414). 



The remarkable circumstance that, in the spectra of 

 several of the nebulae, there is seen only one of the bright 

 lines of hydrogen — that, namely, corresponding to the line 

 F — might in like manner be explained by supposing that, in 

 these nebulae, the atoms composing the ultimate of hydrogen 



