Another property by means of which one may distinguish the various gems from each other is _hardness_. By hardness is meant the ability to resist scratching. The term "hardness" should not be taken to include toughness, yet it is frequently so understood by the public. Most hard stones are more or less brittle and would shatter if struck a sharp blow. Other hard stones have a p.r.o.nounced _cleavage_ and split easily in certain directions. True hardness, then, implies merely the ability to resist abrasion (_i. e._, scratching).

Now, not only is hardness very necessary in a precious stone in order that it may _receive_ and _keep_ a fine polish, but the degree in which it possesses hardness as compared with other materials of known hardness may be made use of in identifying it.

No scale of _absolute_ hardness has ever come into general use, but the mineralogist Mohs many years ago proposed the following _relative_ scale, which has been used very largely:

MOHS'S SCALE OF HARDNESS. Diamond, the hardest of all gems, was rated as 10 by Mohs. This rating was purely arbitrary. Mohs might have called it 100 or 1 with equal reason. It was merely in order to represent the different degrees of hardness by numbers, that he picked out the number 10 to a.s.sign to diamonds. Sapphire (and ruby) Mohs called 9, as being next to diamond in hardness. True topaz (precious topaz) he called 8.

Quartz (amethyst and quartz "topaz") was given the number 7. Feldspar (moonstone) was rated 6, the mineral apat.i.te 5, fluorspar 4, calcite 3, gypsum 2, and talc 1.

It may be said here that any mineral in this series, that is of higher number than any other, will scratch the other. Thus diamond (10) will scratch all the others, sapphire (9) will scratch any but diamond, topaz (8) will scratch any but diamond and sapphire, and so on.

It must not be thought that there is any regularity in the degrees of hardness as expressed by these numbers. The intervals in hardness are by no means equal to the differences in number. Thus the interval between diamond and sapphire, although given but one number of difference, is probably greater than that between sapphire (9) and talc (1). The numbers thus merely give us an order of hardness. Many gem minerals are, of course, missing from this list, and most of the minerals from 5 down to 1 are not gem minerals at all. Few gem materials are of less hardness than 7, for any mineral less hard than quartz (7) will inevitably be worn and dulled in time by the ordinary road dust, which contains much powdered quartz.

In testing a gem for hardness the problem consists in finding out which of the above minerals is most nearly equal in hardness to the unknown stone. Any gem that was approximately equal in hardness to a true topaz (8) would also be said to be of hardness 8. Thus spinel is of about the same hardness as topaz and hence is usually rated as 8 in hardness.

Similarly opal, moonstone, and turquoise are of about the same hardness as feldspar and are all rated 6.

Frequently stones will be found that in hardness are between some two of Mohs's minerals. In that case we add one half to the number of the softer mineral; thus, peridot, benitoite, and jade (nephrite) are all softer than quartz (7) but harder than feldspar (6); hence we say they are 6-1/2 in hardness. Beryl (aquamarine and emerald), garnet (almandine), and zircon are rated 7-1/2 in hardness, being softer than true topaz but harder than quartz. A table of the hardness of most of the commonly known gem-stones follows this lesson.

Having now an idea of what hardness means and how it is expressed, we must next inquire how one may make use of it in identifying unknown gems.

HOW TO APPLY THE HARDNESS TEST. In the first place, it is necessary to caution the beginner against damaging a fine gem by attempting to test its hardness in any but the most careful manner. The time-honored file test is really a hardness test and serves nicely to distinguish genuine gems, of hardness 7 or above, from gla.s.s imitations. A well-hardened steel file is of not quite hardness 7, and gla.s.s of various types while varying somewhat averages between 5 and 6. Hence, gla.s.s imitations are easily attacked by a file. To make the file test use only a _very fine_ file and apply it with a light but firm pressure lengthwise along the girdle (edge) of the unset stone. If damage results it will then be almost unnoticeable. Learn to know the _feel_ of the file as it takes hold of a substance softer than itself. Also learn the _sound_. If applied to a hard stone a file will slip on it, as a skate slips on ice.

It will not take hold as upon a softer substance.

If the stone is set, press a sharp corner of a broken-ended file gently against a _back_ facet, preferably high up toward the girdle, where any damage will not be visible from the front, and move the file very slightly along the surface, noting by the _feel_ whether or not it takes hold and also looking with a lens to see if a scratch has been made. Do not mistake a line of steel, left on a slightly rough surface, for a true scratch. Frequently on an unpolished girdle of real gem material the file will leave a streak of steel. Similarly when using test minerals in accordance with what follows do not mistake a streak of powder from the yielding test material, for a true scratch in the material being tested. The safe way is to wipe the spot thus removing any powder. A true scratch will, of course, persist.

A doublet, being usually constructed of a garnet top and a gla.s.s back, may resist a file at the girdle if the garnet top covers the stone to the girdle, as is sometimes the case, especially in the smaller sizes.

In this case the back must be tested.

One should never pa.s.s a file rudely across the corners or edges of the facets on any stone that may be genuine, as such treatment really amounts to a series of light hammer blows, and the brittleness of most gem stones would cause them to yield, irrespective of their hardness. It should be remembered that some genuine stones are softer than a file, so that it will not do to reject as worthless any material that is attacked by a file. Lapis lazuli (5), sphene (5), opal (6), moonstone (6), amazonite (6), turquoise (6), peridot (6-1/2), demantoid garnet (6-1/2) (the "olivine" of the trade), and jade (nephrite) (6-1/2), are all more or less attacked by a file.

TABLE OF HARDNESS OF THE PRINc.i.p.aL GEM-STONES

10. Diamond.

9-1/2. (Carborundum.) 9. Sapphire and ruby (also all the color varieties of sapphire).

8-1/2. Chrysoberyl (alexandrite).

8. True topaz and spinel (rubicelle, balas ruby).

7-1/2. Emerald, aquamarine, beryl, Morganite, zircon (jacinth and true hyacinth and jargoon), almandine garnet.

7-1/4. Pyrope garnet (Arizona ruby, cape ruby), hessonite garnet (cinnamon stone), tourmaline (various colors vary from 7 to 7-1/2), kunzite (7+).

7. Amethyst, various quartz gems, quartz "topaz," jade (jadeite).

6-1/2. Peridot (chrysolite), demantoid garnet ("olivine"), jade (nephrite).

6. Opal, moonstone, turquoise.

5. Lapis lazuli.

LESSON IX

HARDNESS--_Continued_

MINERALS USED IN TESTING HARDNESS. For testing stones that are harder than a file the student should provide himself with the following set of materials:

1. A small crystal of carborundum. (Most hardware stores have specimen crystals as attractive advertis.e.m.e.nts of carborundum as an abrasive material, or the Carborundum Co., Niagara Falls, N. Y., will supply one.)

2. A small crystal of sapphire (not of gem quality, but it should be transparent and compact. A pale or colorless Montana sapphire can be had for a few cents of any mineral dealer).

3. A small _true topaz_ crystal. (The pure white topaz of Thomas Mountain, Utah, is excellent; or white topaz from Brazil or j.a.pan or Mexico or Colorado will do. Any mineral house can furnish small crystals for a few cents when not of specially fine crystallization.)

4. A small quartz crystal. (This may be either amethyst or quartz-topaz or the common colorless variety. The fine, sharp, colorless crystals from Herkimer County, N. Y., are excellent. These are very inexpensive.)

5. A fragment of a crystal of feldspar. (Common orthoclase feldspar, which is frequently of a brownish pink or flesh color, will do.)

These five test stones represent the following degrees of hardness:

1. Carborundum is harder than any gem material but diamond. It will scratch sapphire and ruby, which are rated 9 in hardness, hence we may call carborundum 9-1/2 if we wish. It is, however, very much softer than diamond, and the latter will scratch it upon the slightest pressure.

2. Sapphire, of hardness 9, scratching any gem material except diamond.

3. True topaz, of hardness 8. It is scratched by sapphire (and, of course, ruby), also by chrysoberyl (which is hence rated 8-1/2), but scratches most other stones. Spinel (which is also rated as 8 in hardness) is really a bit harder than topaz.

4. Quartz, of hardness 7, and scratched by all the previous stones but scratching those that were listed above as of less hardness than a file.

5. Feldspar, of hardness 6, hence slightly softer than a file and yielding to it, but scratching the stones likewise rated as 6 when applied forcibly to them. Also scratching stones rated as less than 6 on slight pressure.

We must next consider how these minerals may be safely used upon gem material. Obviously it would be far safer to use them upon rough gem material than upon cut stones. However, with care and some little skill, one may make hardness tests without particular danger to fine cut material.

The way to proceed is to apply the cut stone (preferably its girdle, or if that is so set as not to be available, a corner where several facets meet) gently to the flat surface of one of the softer test stones, drawing it lightly along the surface and noting the _feel_ and looking to see if a scratch results. If the test stone is scratched try the next harder test stone similarly. _Do not attempt to use the test stone upon any valuable cut stone._ Proceed as above until the gem meets a test stone that it does not attack. Its hardness is then probably equal to the latter and perhaps if pressed forcibly against it a slight scratch would result, but it is not advisable to resort to heavy pressure. A light touch should be cultivated in this work. Having now an indication as to the hardness of the unknown gem look up in the table of the previous lesson those gems of similar hardness and then by the use of some of the tests already given decide which of the stones of that degree of hardness you have. _Never rely upon a single test in identifying a gem._

For further study of hardness and its use in testing gems see _Gem-Stones_, G. F. Herbert-Smith, Chap. IX., pp. 78-81, and table on p.

305; or see _A Handbook of Precious Stones_, Rothschild, pp. 19, 20, 21.

LESSON X

DISPERSION

Another property which may be made use of in deciding the ident.i.ty of certain gems is that called _dispersion_. We have seen in Lesson II.

that light in entering a stone from the air changes its path (refraction), and in Lesson III. it was explained that many minerals cause light that enters them, to divide and proceed along two different paths (double refraction). Now it is further true that light of the various colors (red, orange, yellow, green, blue, and violet) is refracted variously--the violet being bent most sharply, the red least, and the other colors to intermediate degrees. The cut (Fig. 7) represents roughly and in an exaggerated manner the effect we are discussing.

[Ill.u.s.tration: FIG. 7.]

Now in a cut stone this separation of light of different colors, or dispersion of light, as it is called, results in the reflection of each of the colors separately from the steep sloping back facets of the stone. If almost any clear, colorless facetted stone is placed in the sunlight and a card held before it to receive the reflections, it will be seen that rainbow-like reflections appear on the card. These _spectra_, as they are called, are caused by the dispersion of light.

With a diamond the spectra will be very brilliant and of vivid coloring, and the red will be widely separated from the blue. With white sapphire or white topaz, or with rock crystal (quartz), the spectra will be less vivid--they will appear in pairs (due to the double refraction of these minerals), and the red and blue will be near together (_i. e._, the spectra will be short). This shortness in the latter cases is due to the small dispersive power of the three minerals mentioned. Paste (lead gla.s.s) gives fairly vivid spectra, and they are single like those from diamond, as gla.s.s is singly refracting. The dispersion of the heavy lead gla.s.s approaches that of diamond. The decolorized zircon (jargoon) has a dispersion well up toward that of diamond and gives fairly vivid spectra on a card, but they are double, as zircon is doubly refracting. Sphene (a gem rarely seen in the trade) and the demantoid garnet (a green gem often called "olivine" in the trade) both have very high dispersive power, exceeding the diamond in this respect. As they are both colored stones (sphene is usually yellowish, sometimes greenish or brown), the vividness of their color-play is much diminished by absorption of light within them. So also the color-play of a deeply colored fancy diamond is diminished by absorption.