The explanation lies in this, that, though the rectangular mass of coal appears extremely uniform in its structure, it is truly a stratified mass; it is therefore affected, by the sudden approach of fire in a very different manner, according as the edge of the stratum, which is seen in four of the sides of this supposed cube, shall be applied to the fire, or the other two sides, which are in the line of the stratum, or parallel to the bed of coal.  The reason of this phenomenon now remains to be considered.

When the edge of the coal is exposed to the fire, the stratification of the coal is opened gradually by the heat and expanding vapours, as a piece of wood, of a similar shape, would be by means of wedges placed in the end way of the timber.  The coal then kindles quietly, and quickly flames, while the mass of this bituminous schistus is opening like the leaves of a book, and thus exhibits an appearance in burning extremely like wood.  But let the fire be applied to the middle of the bed, instead of the edge of the leaves, and we shall see a very different appearance; for here the expanded aqueous vapours, confined between the laminae, form explosions, in throwing off splinters from the kindling mass; and this mass of coal takes fire with much noise and disturbance.

The ashes of this coal may be determined as to quality, being in general a subtile white earth; but, as to quantity, the measure of that earth produces an indefinite variety in this species of coal; for, from the kennel or parrot coal, which is valuable for its burning with much flame, to that black schistus which our masons use in drawing upon stone, and which, though combustible in some degree, is not thought to be a coal, there is a perfect gradation, in which coal may be found with every proportion of this earthy alloy.

Among the lowest species of this combustible schistus are those argillaceous strata in Yorkshire from whence they procure alum in burning great heaps of this stone, which also contains sulphur, to impregnate the aluminous earth with its acid.  We have also, in this country, strata which differ from those aluminous schisti only in the nature of the earth, with which the bituminous sediment is mixed.  In the strata now considered, the earth, precipitated with the bituminous matter, being calcareous, has produced a limestone, which, after burning especially, is perfectly fissile.

Therefore, with regard to the composition of mineral coal, the theory is this.  That inflammable, vegetable, and animal substances, in a subtilised state, had subsided in the sea, being mixed more or less with argillaceous, calcareous, and other earthy substances in an impalpable state.  Now, the chemical analysis of fossil coal justifies that theory; for, in the distillation of the inflammable or oily coal, we procure volatile alkali, as might be naturally expected.

Thus we have considered fossil coal as various, both in its state and composition; we have described coal which is of the purest composition, as well as that which is most impure or earthy; and we have shown that there is a gradation, from the most bituminous state in which those strata had been formed in being deposited at the bottom of the sea, to the most perfect state of a chemical coal, to which they have been brought by the operation of subterranean fire or heat.