experiment has shown us up to the present that there are really, in this earth and throughout the mighty universe, a number of different kinds of matter. To be quite simple in our conceptions, we might wish to say that all elements are made up of different arrangements of particles of one firstformed matter; but that would be merely guess-work, and at present we cannot get beyond our seventy elements.

If we conceive the smallest particle of an element that can independently exist, we may call such a particle an atom. It will have a certain weight, and the atoms of different elements differ in their weights. But any quantity of matter that we can see, even with a microscope, must consist of enormous numbers of atoms.

The atoms are said to be grouped together in molecules, or "little masses. Chemists regard the molecule of certain elements as consisting of only one atom; in others, two or even six atoms may form the molecule. A molecule has been defined by Professor Tilden as "the smallest quantity which is able to take part in or result from a chemical change."1

A chemical compound is built up of molecules, each of which consists of atoms of more than one kind. In the simplest compounds the molecule consists of only two atoms, one atom of one element being combined with one of another. The elements have been shown to combine together in fixed proportions, forming compounds with distinctive properties. We may hence say that the molecule of a compound is the smallest particle into which the substance could be divided while yet retaining its peculiar chemical constitution. Further splitting might give us certain simpler compounds; but ultimately the actual atoms, or groups of similar atoms, of each molecule would be set free; in other words, the compound would separate into its elements.

All this may seem beyond human observation; but large quantities of molecules of a compound can be dealt with in our experiments, and large quantities of elementary molecules can be extracted by chemical processes from them. Or, on the other hand, we can cause large quantities of elementary molecules of different kinds to combine together to form large quantities of molecules of a chemical compound; and 1 "Introduction to the Study of Chemical Philosophy," 1876, p. 4.

in either case, by weighing the quantity of each substance dealt with, we can find out in what proportions the elements are combined in any particular compound.

We may now glance backwards, and define an atom as the smallest quantity of an element that can be driven out of or caused to enter into a molecule of a compound in the subtle process of destroying or producing any of the known chemical combinations of that element. Here we have got down as far as we can go in the chemical constitution of a mineral. Minerals are made of chemical substances, which are made of molecules; these molecules consist of atoms either (a) of different kinds, in which case we are dealing with a compound, or (b) of only one kind, in which case our substance is an element.

Now we may try to arrive at a definition of a mineral. Let us look at those which we have separated from the granite-three materials evidently requiring distinct names, evidently differing markedly in their physical characters.

The clear little lumps are called Quartz, and do not tell us very much by their external appearance. They break across with irregular curving fractures, much as glass does; and they have in this case no regular shape. But they are transparent, and also very hard; for we cannot scratch them with a knife, and they, on the other hand, will scratch glass. Moreover, they will not soften and melt, as glass does, when held with a pair of forceps in the flame of a Bunsen burner -the gas-burner used in chemical laboratories or in ordinary gas-stoves. Clearly, they are not grains of glass.

If we give them to a chemist, he will find in them the two elements, Silicon and Oxygen, always in the proportion of one atom of silicon to two atoms of oxygen. Quartz is made, in fact, of the oxide of silicon, commonly called Silica. We may take any of these grains from granite rocks found anywhere in the world, and yet their chemical composition will be the

same.

The dull white or pink bodies are called Orthoclase, or Orthoclase Felspar; these, as we have already noticed, have the property of breaking across regularly in certain directions-in two directions, at any rate-when struck with a hammer. This shows that their internal structure, the way in which the molecules are grouped together, differs from that of quartz. Moreover, we can see something of their

external shape; it is often regular, and, even on roughly broken surfaces of the rock, the larger orthoclases are seen to approach the form of small flat bricks. Then, in turning about the rock-specimen in the hand, some of these orthoclases seem to be built up of two parts; that is, one half of the orthoclase catches the light and gives a bright reflection, while the other half looks dull; clearly the planes of fracture are differently sloped in the two parts of the mineral, and it has not a perfectly simple internal structure.

A good knife will just scratch the orthoclase, if drawn firmly across it; but orthoclase will scratch ordinary glass; we thus learn that it is harder than glass, but not so hard as quartz. A little splinter of orthoclase can just be melted after holding it for some time in the flame of a Bunsen gas-burner.

Our chemist tells us that this mineral is more complex than quartz. He finds in it Silicon, Oxygen, Aluminium, Potassium, and generally Sodium. The proportion of these elements to one another is always the same, except in the case of the two last named. Sometimes there is only potassium present, and no sodium; sometimes sodium occurs, and then there is less potassium. We say that the potassium may be replaced in part by sodium without the principal characteristics of the mineral being altered. In such cases it must be a matter of general consent and judgment as to whether we are to call the mineral by a new name when such a replacement occurs. The atoms of sodium are lighter than those of potassium; but any difference in the weight of a given bulk of orthoclase, according as it contains potassium, or both potassium and sodium, would be very trifling. Cases often occur, moreover, where a lighter element replaces a heavier one, and yet the resulting variety of the mineral is actually heavier than the ordinary form. The molecules containing the lighter element must in such cases lie closer together, i..., there must be more of them in a given space, than occurs when only molecules containing the heavier element are present.

So we may note that the chemical composition of some minerals may vary slightly, varieties being set up which are grouped together under the common mineral name. A molecule of our typical orthoclase contains 2 atoms of potassium, 2 of aluminium, 6 of silicon, and 16 of oxygen. But in most

specimens of orthoclase there are some molecules in which two atoms of sodium take the place of the two atoms of potassium. When such molecules are shown by a chemical analysis to be numerous, we call the mineral variety Soda-Orthoclase.

Our third mineral in the granite is called Mica-the platelike substance which is easily dug out with the knife, and which gives such a shining appearance to many specimens of the rock. Some micas are pale and silvery; others are quite dark, being bronze-coloured or black; but they have many characters in common. Again and again their proper form can be seen; their molecules are clearly capable of building up six-sided plates. This structure is so common that it cannot be accidental; and careful measures show that the angles of the hexagonal figure are kept the same in the same chemical variety. Such a structure is called a crystal; and a little observation will show us that the orthoclase also is in the form of crystals, though they have probably not been able to grow to such perfection. The mica splits very easily in one direction, so that we can flake off thin flexible plates; and even the blacker micas are no longer opaque when one of their thin plates is held up to the light. The mineral is quite soft, and can often be scratched by the thumb-nail.

The chemical composition of the micas is still more complex than that of the orthoclase felspar, but all common varieties contain Silicon, Oxygen, Aluminium, Potassium, Magnesium, Iron, and Hydrogen. The pale varieties contain little iron and usually little magnesium; the dark varieties are richer in these elements and poorer in silicon and aluminium. There may be a good deal of replacement of one element by another in the mica series, the general external characters being retained; and several different kinds or species of mica are recognised, and are known by distinct mineral names.

If we are so fortunate as to find a granite with a number of small hollows in it, perhaps half-an-inch or an inch across, such as occur in the famous rocks of the northern part of the Mourne Mountains, then we may see that all the minerals which we have discussed are capable of forming crystals, wherever freedom is given to them and when they are not too much squeezed together. In such hollows we find the mica, and the orthoclase, and the quartz itself, delicately crystallised, beautifully and regularly shaped, so

that persons looking at the handsome crystals in public collections often think that they have been artificially cut and polished. But you have only to go into the hills, and search in cracks and hollows for yourselves, to find that crystals occur naturally, and that they are, indeed, the form frequently adopted by natural chemical substances. And, further, the crystals of one compound usually differ from those of another; the sides make different angles one with another, even if the forms are very much alike; so that, by careful observation and measurement of the angles in various crystals, we can use the outer shape to help us in determining the nature of the mineral.

And now we can at last arrive at our definition. A mineral is a natural substance, formed without the action of plants or animals. Its chemical composition is constant, or varies only by a well-defined series of chemical replacements. Under favourable conditions, it assumes a crystalline form.

One mineral is known from another by a number of characters which must be considered all together; and these will be found stated in any text-book of mineralogy.1 Let us run over the most important of these characters

here.

1. Colour.-When there are several minerals in a rock, this character often clearly marks out one from another. But it is of far less importance than might be supposed, since many common minerals are coloured by some trifling impurity; a sort of stain runs through them, as it were, and in their pure condition they are colourless. Metallic ores, however, usually have characteristic colours; thus Iron Pyrites, the common iron sulphide, is brass-yellow, and Native Copper, the natural element, is copper-red. But, on the other hand, the red gem Ruby and the blue Sapphire are mere varieties of the same mineral species, Corundum, the composition of both being aluminium oxide coloured by a substance very insignificant in amount. Hence we must

1 Such as Hatch, "Mineralogy" (Whittaker & Co.); Rutley, "Mineralogy" (T. Murby).

2 Artificial corundum has been coloured red (ruby) by adding chromium fluoride to the materials employed; and, curiously enough, the addition of this same substance in varying proportions has given rise to sapphires and to a green variety. Rubies have also been made by adding potassium bichromate, and sapphires by adding cobalt oxide (Fouqué and Lévy, Synthèse des Minéraux et des Roches, pp. 220 and 222).