Other evidence may be sought for in the spectrum of argon, which was carefully examined by Mr. Crookes. It consists of a great number of lines, extending all through the spectrum, from far down in the red to far beyond the visible violet; the invisible lines were examined by the aid of photography, for ultraviolet light, although invisible to the eye, impresses a photographic plate. The most striking feature of this spectrum is the change which can be produced in it by altering the intensity of the electric discharge which is passed through the tube containing argon at a low pressure. By interposing a Leyden jar between the secondary terminals of the induction-coil from which sparks are taken through the gas, the colour of the light in the tube changes from a brilliant red to an equally brilliant blue. A large number of lines in the red spectrum disappear, on interposing the jar, while many lines in the blue-green, blue, and violet part of the spectrum, invisible before, shine out with great brilliancy. There is no other gas in which a similar alteration of intensity of discharge produces such a marked difference, although in many gases, supposed to be simple substances, similar changes may be produced. So far as we know at present, however, such a change cannot be definitely ascribed to the presence of a mixture of two elements, although it is in itself a very remarkable phenomenon. On the other hand, Professors Runge and Paschen, in a paper communicated to the Royal Academy of Science of Berlin in July 1895, have adduced reasons for concluding that helium, the gas from clèveite, is a mixture; it appears to show lines belonging to two spectra, each series of lines exhibiting certain regularities. But this, although an important conclusion in itself, has no direct bearing on the question of the simplicity of argon. One method of separating the constituents of a mixture is by taking advantage of their different solubilities in water, or in some other appropriate solvent. And as argon was found to have the solubility of 4 volumes in 100 of water, while helium is very sparingly soluble, only 0·7 volume per 100, it is not unreasonable to suppose that, if argon consisted of a mixture of elements in argon, one should be more soluble than another. Exhaustive experiments in this direction have still to be carried out; but Lord Rayleigh has made experiments which render it very improbable that any separation into its constituents, if it be a mixture, can be thus effected. Wishing to ascertain if there were any helium in the air, he shook up atmospheric argon with water, until a very small fraction remained undissolved. The spectrum of this small residue was identical with that of the original argon, from which it would appear that this method, at least, is incapable of effecting any separation.