All these details of the process of crystallisation are very evident indications of a determinate figure in the ultimate atoms of the substances which are crystallised. But besides the substances which are thus reduced by art to the form of crystals, there are larger classes which naturally exist in that state. There are certain planes, called _planes of cleavage_, in the directions of which natural crystals are easily divided. These planes, in substances of the same kind, always have the same relative position, but differ in different substances. The surfaces of the planes of cleavage are quite invisible before the crystal is divided; but when the parts are separated, these surfaces exhibit a most intense polish, which no effort of art can equal. We may conceive crystallised substances to be regular mechanical structures formed of atoms of a certain figure, on which the figure of the whole structure must depend. The planes of cleavage are parallel to the sides of the constituent atoms; and their directions, therefore, form so many conditions for the determination of their figure. The shape of the atoms being thus determined, it is not difficult to assign all the various ways in which they may be arranged, so as to produce figures which are accordingly found to correspond with the various forms of crystals of the same substance. (18.) When these phenomena are duly considered and compared, little doubt can remain that all substances susceptible of crystallisation, consist of atoms of determinate figure. This is the case with all solid bodies whatever, which have come under scientific observation, for they have been severally found in or reduced to a crystallised form. Liquids crystallise in freezing, and if aëriform fluids could by any means be reduced to the solid form, they would probably also manifest the same effect. Hence it appears reasonable to presume, that all bodies are composed of atoms; that the different qualities with which we find different substances endued, depend on the magnitude and figure of these atoms; that these atoms are indestructible and immutable by any natural process, for we find the qualities which depend on them unchangeably the same under all the influences to which they have been submitted since their creation; that these atoms are so minute in their magnitude, that they cannot be observed by any means which human art has yet contrived; but still that magnitudes can be assigned which they do not exceed.