The colour is the least reliable character that may be employed for the identification of a stone, since it varies considerably in the same species, and often results from the admixture of some metallic oxide, which has no essential part in the chemical composition and is present in such minute quantities as to be almost imperceptible by analysis. Who would, for instance, imagine from their appearance that stones so markedly diverse in hue as ruby and sapphire were really varieties of the same species, corundum? Again, quartz, in spite of the simplicity of its composition, displays extreme differences of tint. Nevertheless, certain varieties do possess a distinctive colour, emerald being the most striking example, and in other cases the trained eye can appreciate certain characteristic subtleties of shade. At any rate, the colour is the most obvious of the physical characters, and serves to provide a rough division of the species, and accordingly in Table II at the end of the book the gem-stones are arranged by their usual tints. DICHROISM The two rays into which a doubly refractive stone splits up a ray of light are often differently absorbed by it, and in consequence appear on emergence differently coloured; such stones are said to be dichroic. The most striking instance is a deep-brown tourmaline, which, except in very thin sections, is quite opaque to the ordinary ray. The light transmitted by a plate cut parallel to the crystallographic axis is therefore plane-polarized; before the invention by Nicol of the prism of Iceland-spar known by his name this was the ordinary method of obtaining light of this character (cf. p. 43). Again, in the case of kunzite and cordierite the difference in colour is so marked as to be obvious to the unaided eye; but where the contrast is less pronounced we require the use of an instrument called a dichroscope, which enables the twin colours to be seen side by side. [Illustration: FIG. 29.—Dichroscope (actual size).] [Illustration: FIG. 30.—Field of the Dichroscope.]