"The nearest approach that has been made, of an exact and scientific kind, toward determining the constitution of the sun's surface has resulted from an examination of the _solar spectrum_. A ray of light, by passing through a triangular prism of glass, is, as you know, divided into its elements, or constituent colors. The ray of light is spread out like a half-open fan. This divided and expanded ray, thrown upon a screen, is called the spectrum. An examination of the solar spectrum by a microscope shows certain fine dark lines across it. The lines are invariably the same in their position and grouping. The spectrum of the stellar light is found to differ from that of the solar light, and the light of one star differs from that of another star. Light from incandescent metallic vapors gives bright lines across the spectrum. Each metal has its own number, position, grouping, and color of these spectral lines. By comparing the solar spectrum with the spectra of the various metals--the processes are curious and the explanation difficult to be understood--corresponding lines are discovered, and the conclusion is reached that the sun's atmosphere contains the vapors of several of our well-known metals, as iron, nickel, sodium, potassium, and others. This is a most curious and marvelous scientific feat, to make an approximate chemical analysis of the sun and stars by means of their light. The conclusions, however, seem trustworthy. "Can you tell us, Ansel, whether the earth receives heat from the moon and stars?" "I cannot, sir." "I should be glad, Mr. Hume, to have you instruct us upon this point."