159. Lockyer, in his astronomical researches, has recently started this question. He argues that in the sun and stars, and more especially in the whiter stars, there are temperatures very much higher than any which have been here produced. He assumes too that simplicity of constitution accompanies a simple spectrum, an hypothesis which is consistent with the fact that compounds as a rule give spectra much more complicated than those of simple substances. Now it is a curious circumstance that the atmospheres of some of the whiter stars, such as Sirius, do not appear to contain anything but hydrogen; at least we have no indication that they do; other stars, again, of less whiteness, in addition to hydrogen, have such substances as iron, sodium, etc., while yellow, orange, and blood-red stars and variable stars, appear to contain in their atmospheres substances which are compounds. If then it be true that as a rule the atmospheres of the whiter stars contain the fewer elements and those of smallest atomic weight, and that as stars diminish in whiteness their atmospheres rise in complexity of structure, in fine, if we have reason to associate together whiteness and simplicity, this undoubtedly tells in favour of the power of high-temperature heat to split up the so-called elements. We conclude the whiter stars to be the hotter stars, from the fact that their spectra contain a greater proportion of the more refrangible rays than do those of yellow or red stars. In fine, a speculation of this nature is not to be summarily dismissed, but ought to be retained as a working hypothesis which may in time throw great light on the ultimate constitution of the chemical elements. Is it fanciful to suppose that the passage prefixed to Chapter III. may refer to this, since (literally translated) it stands—‘... the elements, intensely heated, shall be broken up....’?