If radium were really able (as some people have wrongly inferred from the French experiments) to change the chemical nature of corundum and convert it into topaz and emerald, the case would be very different from that of merely changing the colour of the corundum. What is to-day called “topaz” is a sherry-yellow crystal consisting of silicate of alumina and of fluoride of alumina. It turns pink when heated, and is also known of a blue colour and colourless. The topaz of the ancients from the coasts of the Red Sea is of a different chemical nature, and is now called peridot. Yellow corundum is sometimes wrongly called Oriental topaz, and the yellow-brown quartz crystals properly known as cairngorms are sometimes wrongly called Scotch topaz. So that the word “topaz” is used loosely as well as strictly, and confusion results. Emerald is widely distinct from corundum, sapphire, and ruby. It is a silicate of alumina and beryllium, and in its coarse and pale-coloured variety is known as beryl.
From all this it appears that some names of precious stones indicate substances quite distinct from one another chemically, built of differing elements, and also _per contra_ that what is actually one and the same kind of precious stone in chemical composition and native crystalline form may present examples possessing various colours and degrees of transparency, each variety being called by a distinct name, and regarded popularly as a distinct kind of stone. Radium rays can convert colourless alumina or corundum into blue alumina (sapphire) or red alumina (ruby), but they cannot change alumina into beryllia (that is into emerald), nor into fluoride (that is into topaz).
One naturally asks, “To what is the colour of these precious stones due?” The answer is difficult, because very minute traces of chemical impurity, such as iron, cobalt, manganese, or chromium may suffice to tint an otherwise transparent, colourless crystal with the brightest red, yellow, blue, violet, or green. Moreover, it is certain from what we know of traces of metallic impurity in artificial glass that it may exist in such a state of chemical combination as to give no tint whatever to the glass, but after prolonged exposure to light or other agencies, the minute impurity may combine chemically with oxygen present in the glass and develop colour. Thus, for instance, old window-glass often assumes a violet or amethystine tint after long exposure. This varying colour of the combinations of metals according to whether they are oxidised or not, and the degree of oxidation, or the special salt which they may form, is in itself an unexpected thing to those who are not chemists. The metal chromium, for instance, gives rise to colourless, to yellow, red, green, and blue combinations. Manganese, a metal commonly associated with iron, gives rise to brilliant green, to violet, and to wine-red combinations, and if scattered as microscopic particles of black oxide in glass would produce no colour effect at all. From what we know of glass and the ease with which it is coloured to every shade of the rainbow by the admixture of traces of metallic impurities--so that “paste” or glass gems of all colours can be manufactured--it is not surprising to find that natural crystals, transparent and often devoid of colour (such as corundum, diamond, quartz, and topaz), are yet also found more or less frequently coloured in various tints. Nevertheless, it is the fact that in very few cases have chemists been able to prove by analysis what precisely is the cause of the colour in any given crystal or precious stone, although they may strongly suspect this or that as the colour-giving impurity. The actual quantity of a metallic impurity sufficient to give a tint is so excessively minute that the chemist finds it impossible to determine what it is by examining one small precious stone. He has not a sufficient bulk of material to operate on.