The cabochon form may be of three different kinds. In the first, the double cabochon (Fig. 40), both the upper and the under sides of the stones are curved. The curvature, however, need not be the same in each case; indeed, it is usually markedly different. Moonstones and starstones are generally cut very steep above and shallow underneath. Occasionally a ruby or a sapphire is, when cut in this way, set with the shallow side above, because the light that has penetrated into the stone from above is more wholly reflected from a steep surface with consequent increase in the glow of colour from the stone. Opals are always cut higher on the exposed side, but the slope of the surface varies considerably; they are generally cut steeply when required for mounting in rings. Chrysoberyl cat’s-eyes are invariably cut with curved bases in order to preserve the weight as great as possible. The double cabochon form with a shallow surface underneath merges into the second kind (Fig. 41) in which the under side is plane, the form commonly employed for quartz cat’s-eyes, and occasionally also for carbuncles. In this type the plane side is invariably mounted downwards. In the third form (Fig. 42) the curvature of the under surface is reversed, and the stone is hollowed out into a concave shape. This style is reserved for dark stones, such as carbuncles, which, if cut at all thick, would show very little colour. A piece of foil is often placed in the hollow in order to increase the reflection of light, and thus to heighten the colour effect. [Illustration: FIG. 42.—Double (Concavo-convex) Cabochon.] In early days it was supposed that the extreme hardness of diamond precluded the possibility of fashioning it, and up to the fifteenth century all that was done was to remove the gum-like skin which disfigured the Indian stones and to polish the natural facets. The first notable advance was made in 1475, when Louis de Berquem discovered, as it is said quite by accident, that two diamonds if rubbed together ground each other. With confident courage he essayed the new art upon three large stones entrusted to him by Charles the Bold, to the entire satisfaction of his patron. The use of wheels or discs charged with diamond dust soon followed, but at first the lapidaries evinced their victory over such stubborn material by grinding diamond into divers fantastic shapes, and failed to realize how much might be done to enhance the intrinsic beauty of the stones by the means now at their disposal. The Indian lapidaries arrived at the same discovery independently, and Tavernier found, when visiting the country in 1665, a large number of diamond cutters actively employed. If the stone were perfectly clear, they contented themselves with polishing the natural facets; but if it contained flaws or specks, they covered it with numerous small facets haphazardly placed. The stone was invariably left in almost its original shape, and no effort was made to improve the symmetry.