It may here be worth while to mention that it is not invariably true that a substance gets contracted, and the molecules more and more pressed together, as it assumes a solid form.  There is at least one exception.  If we take 1,700 pints of steam, the water, as I said, on becoming cool enough to lose the vapourous form, will shrink into a measure holding a single pint; if we cooled lower still, it will get smaller and smaller in bulk (though of course not at all at the same rate) till it arrives at a point when it is just going to freeze; then suddenly (7 degrees above the freezing point) it again begins to expand.  Ice occupies more space than cold water; its molecules get arranged in a particular manner by their crystallization.

On the admission of an intelligent Creator providing, by beneficent design, the laws of matter, it is easy to give a reason for this useful property.  It prevents the inhabitants of northern climates being deprived of a supply of water.  As it is, the solid water or ice expands, and, becoming lighter, forms at the top of the water, and the heavier warmer water remains below.  But if ice always got denser and sank, the warmer liquid would be perpetually displaced and so come up to the surface, where it would freeze and sink in its turn.  In a short time, then, all our water supplies would (whenever the temperature went down to freezing, which it constantly does in winter) be turned into solid ice.  This would be a source of the gravest inconvenience to the population of a cold climate.  If we deny a designing mind, the alternative is that this property of water is a mere chance.

But to return to molecules.  Molecules are endowed with an inherent faculty of motion; only under the conditions of what we call the solid, they are so compressed, that there is no room for any motion appreciable to the senses.  Even if the solid is converted into vapour, the molecules are still much restrained in their movements by the pressure of the air.  But of late years, great improvements (partly chemical, partly mechanical) have been made in producing perfect vacua; that is to say, in getting glass or other vessels to be so far empty of air, that the almost inconceivably small residue in the receptacle has no perceptible effect on the action of a small quantity of any substance already reduced to the form of gas or vapour introduced into it.  Dr. W. Crookes has made many beautiful experiments on the behaviour of the molecules of attenuated matter in vacua.  The small quantity of vapour introduced contains only a relatively small number of molecules, which thus freed from all sensible restraint within the limits of the glass vessel used, are free to move as they will; they are observed to rush about, to strike against the sides of the vessel, and under proper conditions to shine and become radiant, and to exhibit extraordinary phenomena when subjected to currents of electricity.  So peculiar is the molecular action thus set up, that scientific men have been tempted to speak of a fourth condition of matter (besides the three ordinary ones, solid, liquid, and gaseous), which they call the ultra-gaseous or radiant state of matter.