In quite a different class from the imitation gems made of cut flint-glass are the artificial gems, which in nature and composition are exact reproductions of natural gems, but which have been produced by artificial processes. As far as the writer is aware these are only found in any large numbers in the case of the ruby, but in that case, at all events, it is said that the production of the artificial crystals is at least as costly as the purchase of the natural stones. There can, however, be very little doubt that as the processes of fusion and crystallisation become better known and understood, and the chemistry of silicate minerals is developed, the artificial production of mineral crystals in, at all events, moderate sizes will become increasingly possible; it is even to be hoped that their production will be so far perfected as to place their really valuable properties at the service of man. _Chilled Glass._--In all the processes of glass manufacture described in the present book, annealing has always played an important part. The glass, after it has undergone its last treatment under the influence of heat, is subjected to a gradual cooling process with the object of freeing it from the internal strains which it would otherwise retain, and which would, ordinarily, endanger its existence and interfere with its use. It is, however, well known that surfaces of glass subjected to such internal strains as result in a compressive stress on the glass near the surface, are less liable to injury, and are apparently stronger than when the glass is annealed and the stresses are removed. On the other hand, glass surfaces under tension are extremely delicate and fragile. In some respects, therefore, glass which has not been annealed may appear to be stronger than the annealed product. The well-known case of the Rupert’s drop is an example of this kind. Rupert’s drops are produced by dropping molten glass into water; they generally take the form of a more or less spherical body having a long tail, tapering off into a thread, attached to it. Such a Rupert’s drop may be struck with a heavy hammer, and will safely resist a blow that would splinter a similar body made of annealed glass. If, however, the surface be scratched, or the tip of the tail be broken off, the entire “drop” breaks up, sometimes with a violent explosion, into minute fragments. Numerous inventors, among whom De la Bastie and Siemens figure most conspicuously, have endeavoured to utilise these properties of chilled glass, not exactly by endeavouring to produce that extreme degree of internal strain which is characteristic of the Rupert’s drop, but by producing what they describe as “tempered” glass, in which the internal strains have been reduced by less violent cooling to such an extent as to retain some of the advantages of the hardened, internally strained condition while approximating more or less to the safer state of annealed glass.