originals are kept in the strong room of a bank or safe-deposit. In all cases, however, a hard file will abrade the surface of the false stone. In chapter VII. we found that quartz is in the seventh degree of hardness, and an ordinary file is but a shade harder than this, so that almost all stones higher than No. 7 are unaffected by a file unless it is used roughly, so as to break a sharp edge. In order to prepare artificial diamonds and other stones for the file and various tests, they are often what is called "converted" into "doublets" or "triplets." These are made as follows: the body of the glass is of paste, and on the "table" (see last chapter), and perhaps on the broader facets, there will be placed a very thin slab of the real stone, attached by cement. In the case of the diamond, the body is clear, but in the coloured imitations the paste portion is made somewhat lighter in shade than the real stone would be, the portion below the girdle being coloured chemically, or mounted in a coloured backing. Such a stone will, of course, stand most tests, for the parts usually tested are genuine. A stone of this nature is called a "doublet," and it is evident that when it is tested on the underside, it will prove too soft, therefore the "triplet" has been introduced. This is exactly on the lines of the doublet, except that the collet and perhaps the pavilions are covered also, so that the girdle, which is generally encased by the mounting, is the only surface-portion of paste. In other cases the whole of the crown is genuine, whilst often both the upper and lower portions are solid and genuine, the saving being effected by using a paste centre at the girdle, covered by the mounting. Such a stone as this last mentioned is often difficult to detect without using severe tests and desperate means, e.g.:--(a) by its crystalline structure (see Chapter III.); (b) by the cleavage planes (see Chapter IV.); (c) by the polariscope (see Chapter V.); (d) by the dichroscope (see Chapter VI.); (e) by specific gravity (see Chapter VIII.); (f) cutting off the mounting, and examining the girdle; (g) soaking the stone for a minute or so in a mixture said to have been originally discovered by M. D. Rothschild, and composed of hydrofluoric acid and ammonia; this will not answer for all stones, but is safe to use for the diamond and a few others. Should the jewel be glass, it will be etched, if not completely destroyed, but if genuine, no change will be apparent; (h) soaking the diamond for a few minutes in warm or cold water, in alcohol, in chloroform, or in all these in turn, when, if a doublet, or triplet, it will tumble to pieces where joined together by the cement, which will have been dissolved. It is, however, seldom necessary to test so far, for an examination under the microscope, even with low power, is usually sufficient to detect in the glass the air-bubbles which are almost inseparable from