We have seen how remarkably the properties of iron are affected by minute additions of carbon which may be assumed to enter into chemical union with the metal. The properties of other metals may be influenced by minute quantities of added elements, although in quantities so small as to preclude the possibility of their forming ordinary chemical compounds. It by no means follows, however, that the atom of an added element does not exert a direct influence. In Professor Roberts-Austen’s laboratory, in London, two ladles were filled with exceptionally pure bismuth; into one ladle a tiny fragment of tellurium was placed. The ladles were poured each into a separate mold, and when the metal became cold it was fractured by a hammer. The bismuth to which the tellurium was added had become minutely crystalline; while that which remained pure had crystallized in broad mirror-like planes. One reflected light as a mirror; the other, containing the tellurium, scattered the light it received. With no guidance but that of mere inspection, one would have said that the two substances were distinct elements, and yet the only difference was that one contained 1/2000 part of tellurium and the other no tellurium at all. Submarine telegraphy presents us with a case as striking: were its copper wire to contain but one-thousandth part of bismuth, the line would be so much reduced in conductivity as to be commercially worthless: quite as harmful are mixtures of antimony. In coining, the addition to gold of one five-hundredth part by weight of bismuth produces an alloy which crumbles under the die and refuses to take an impression. In the manufacture of such dies it is necessary to employ a steel containing 0.8 to 1 per cent. of carbon and no manganese. It is usual, says Professor Roberts-Austen, to water-harden and temper it to a straw color, and a really good die will strike 40,000 coins without being fractured or deformed, but if the steel contains 0.1 per cent. too much carbon, it would not strike 100 pieces without cracking, and if it contained 0.2 per cent. too little carbon, it would probably be hopelessly distorted and its engraved surface destroyed in the attempt to strike a single coin. As in coining so in steam-engineering. A little arsenic added to copper improves it for the fire-boxes of locomotives. Boilers of old, formed of copper slightly admixed with sulphur, lasted longer than modern boilers built of copper free from sulphur. Antimony behaves like arsenic, and in due proportion strengthens copper; bismuth, on the contrary, weakens copper, and a perceptible effect is wrought by a mere trace. Nickel is made malleable by adding extremely small quantities of phosphorus, magnesium, or zinc. BOOKS ON IRON AND STEEL Chosen and annotated by Professor Bradley Stoughton, School of Mines, Columbia University, New York.