68. Let us now leave the forces which animate large masses of matter, and proceed to discuss those which subsist between the smaller particles of which these large masses are composed. And here we must say one word more about molecules and atoms, and the distinction we feel ourselves entitled to draw between these very small bodies, even although we shall never be able to see either the one or the other. In our first chapter (Art. 7) we supposed the continual sub-division of a grain of sand until we had arrived at the smallest entity retaining all the properties of sand--this we called a _molecule_, and nothing smaller than this is entitled to be called sand. If we continue this sub-division further, the molecule of sand separates itself into its chemical constituents, consisting of silicon on the one side, and oxygen on the other. Thus we arrive at last at the smallest body which can call itself silicon, and the smallest which can call itself oxygen, and we have no reason to suppose that either of these is capable of sub-division into something else, since we regard oxygen and silicon as elementary or simple bodies. Now, these constituents of the silicon molecule are called _atoms_, so that we say the sand molecule is divisible into atoms of silicon and of oxygen. Furthermore, we have strong reason for supposing that such molecules and atoms really exist, but into the arguments for their existence we cannot now enter--it is one of those things that we must ask our readers to take for granted. 69. Let us now take two molecules of sand. These, when near together, have a very strong attraction for each other. It is, in truth, this attraction which renders it difficult to break up a crystalline particle of sand or rock crystal. But it is only exerted when the molecules are near enough together to form a homogeneous crystalline structure, for let the distance between them be somewhat increased, and we find that all attraction entirely vanishes. Thus there is little or no attraction between different particles of sand, even although they are very closely packed together. In like manner, the integrity of a piece of glass is due to the attraction between its molecules; but let these be separated by a flaw, and it will soon be found that this very small increase of distance greatly diminishes the attraction between the particles, and that the structure will now fall to pieces from the slightest cause. Now, these examples are sufficient to show that molecular attraction or _cohesion_, as this is called, is a force which acts very powerfully through a certain small distance, but which vanishes altogether when this distance becomes perceptible. Cohesion is strongest in solids, while in liquids it is much diminished, and in gases it may be said to vanish altogether. The molecules of gases are, in truth, so far away from one another, as to have little or no mutual attraction, a fact proved by Dr. Joule, whose name was mentioned in the last chapter.