Scientific American 3, 17.1.1863
The London Mechanics' Magazine contains the following interesting remarks on this subject:-
From very ancient times it has been the practice to divide the ounce troy into twenty-four imaginary parts. An ounce of pure gold, therefore, and what is called twenty-four carat gold, are identical terms. Anything less that twenty-four carat gold indicates that in an ounce of that substance there are so many twenty-fourth parts of gold, and that the remaining portions of an ounce of it is made up of an alloy of some inferior metal, usually copper. Except for wedding rings, which are, or should be, made of standard gold, consisting of twenty-two parts or carats of pure gold, and two of copper alloy, jewelers seldom use gold of a higher rate of purity than eighteen-carat. How often they use it of a much lower degree of fineness they know best. There is no doubt that it is possible to give almost any color to gold, by the addition of particular alloys; and of late ingenuity has been at work to give the sixteen-carat gold the appearance of pure gold. This is done by the aid, partly, of what is techically known as the coloring pot, in which the metal is treated to an acid bath of a certain amount of strenght. By the judicious use of this contrivance, twenty-two carat gold may be made to resemble very closely native gold, as found in the shape of nuggets in California or Austr[i]alia. Nuggets are never, however, found to be pure gold - they consist for the most part of about twenty-three carat gold, the fraction being made up of an alloy of some inferior metal. In order to test the purity of gold the application of heat is, perhaps, one of the simplest means. Pure gold will not be in the least discolored by it, while twenty-three carat gold will take a slightly red tint. There is this disadvantage about the mode of testing suggested, it will certainly discolor very materially all gold of degrees of fineness inderior to that of twenty-three carat, and nothing but the acid bath will restore it to its original hue. In order to effect this latter operation on a small scale, nothing more is necessary than to obtain an earthen pipkin or gallipot, place the article in it, cover it with nitric acid, and hold it over a spirit or gas lamp, or even a candle, until the acid boils. The result will be that the metal will be restored to its original brightness.
As regards the testing of the genuineness of gold coin, there is nothing equal in simplicity or efficacy to weighing it. It is impossible that aby metal inferior to stand gold can be used in the manufacture of counterdelt sovereigns or half sovereigns which will give pieces, of the same size, of equal weight.
The testing of "jewelers' gold," when used in the manufacture of many kinds of trinkets, is a matter attended with considerable riska and difficulty; and perhaps the only means readily at the command of the purchaser, for insuring a proper relationship between wuality and price in such cases, is to ascertain the respectability of the seller.