Secondly: as to _weight_. This is the most accurate method, but the stone must be taken from its setting. The mode of taking the gravity has already been given (page 13), and the amount of each is stated in Chapter II. Garnet is the heaviest of gems; weighed in water it loses only one-fourth of its weight; _i. e._, if a red garnet be suspended by a fine thread from a delicate balance and immersed in a glass of water under it, one-quarter of its ordinary weight in air must be added to the pan from which it is suspended to restore the equilibrium. In like manner, ruby and sapphire lose a little more. The diamond and white topaz lose two-sevenths of their weight. Rock-crystal, amethyst, carnelian and agate lose five-thirteenths; and opal about one-half, being the lightest of gems. The emerald loses more than one-third. As “paste” _can_ be made so as to have the same specific gravity as the genuine article, this test alone can not be relied upon; but very few of the imitations are so carefully made. The test is very convenient in distinguishing gems of like color from each other, as oriental ruby, spinel ruby and red tourmaline, and green tourmaline and emerald. Thirdly: characteristics depending on _light_ and _electricity_. It is not easy to look through a diamond of the first water, while imitations readily permit objects to be seen through them. A very delicate and perfect test of a diamond, distinguishing it from all colorless gems, as white topaz, white sapphire and white zircon, but not from “pastes,” is to look through it at a pin-hole in a card. This requires some dexterity, and the gem should be fixed to a steady object by a bit of wax at a proper distance. A true diamond will show but one hole, all the others will show two. As white topaz, when large, is a magnificent stone, it is often palmed off for a diamond of great value; but this test is invariably certain. A true diamond retains its brilliancy under water. When a colored stone is placed in the path of the solar spectrum (the row of seven colors into which sunlight is separated by a prism), its color will vary with the portion of the spectrum which falls upon it; and two stones of the same color, but of a different nature, will exhibit different effects. Thus, a paste placed beside a fine colored gem, betrays its worthlessness. A simpler method of testing stones is to look at them through a bit of glass, colored red, yellow, blue or green. Every stone will exhibit, under this test, properties peculiar to itself, and by which its [Illustration: SEARCHING FOR DIAMONDS.] nature may be recognized. This is also a severe test for the purity of tint; for if pure and unmixed, the stone will appear completely black in every other light but its own color. Milky and turbid stones can not bear this test.