black in the solar spectrum. Suppose that, when viewmg the solar spectrum, the light of the solid or liquid photosphere of the sun were suddenly extinguished, the gaseous atmosphere still glowing, the colours which we at present see would be at once extinguished, and all the dark lines would become bright bands with their appropriate colours. We would now be viewing a gaseous incandescent atmosphere, and we would have the spectra of all the bodies in this atmosphere represented by these bands. The next total eclipse will present an opportunity of applying this test. The perfect conversion of the positive into the negative spectrum is that which gives its validity to this new analysis. We can read these lines and interpret their meaning, as distinctly as we read the symbols and formulae of the chemist. We have only to ascertain what is the positive symbol of any substance by diffusing it in the flame of a Bunsen's lamp, and then look to the sun for a similar symbol; the only difference is, that, in the former case, the letters are written in coloured inks, whereas, in the latter, they are all black; but the form, grouping, and position are such that their meaning cannot be mistaken.
But how is it that the spectrum of a substance is coloured in terrestrial flame, and black in the gaseous envelope of the sun?—simply from the circumstance, that behind the incandescent atmosphere of the sun
