Chemical composition determines the fusibility of a rock at a given temperature. The least fusible rocks are, on the one hand, those which contain large quantities of silica, 60 to 75%, and, on the other, those which contain less than 40% of silica. The most fusible rocks are those with an intermediate percentage of silica, and among these the fusibility increases, as the percentage of silica diminishes, until the lower limit is reached. The effect of chemical composition upon texture is seen in the rapidity with which the less fusible rocks chill and stiffen, and therefore the greater frequency with which they form glasses".

Chemical composition is, however, important in this connection chiefly through its effect upon the rate of solidification. We have already learned (p. 9), that solidification very generally takes place by a process of crystallization, and this requires time. Hence, very rapid cooling results in a glass, but the microscope reveals the incipient stages of crystallization in many of even the glassy rocks. A somewhat slower rate of solidification produces a cryptocrystalline rock, and successively slower rates bring about the porphyritic, microcrystalline, and granitoid textures. Large crystals form slowly, and other things being equal, the larger the component crystals of a rock, the more slowly has it consolidated.

Pressure is of importance in preventing the rapid escape of the vapours and gases contained in the molten mass, and hence frothy, scoriaceous, and vesicular textures cannot be produced under high pressures. Pressure is also believed to be necessary for the formation of many phenocrysts in porphyritic rocks.

The mineralizers, such as steam, hydrochloric acid, and other vapours, determine the crystallization of many minerals, which refuse to crystallize in the absence of such vapours. Variations in the quantity of mineralizers present in different parts of the same mass occasion corresponding differences in the local textures. The well-known Obsidian Cliff, in the Yellowstone National Park, is. formed by a great lava-sheet, made up of alternating layers of glassy and microcrystalline rock, a difference which is referred to varying proportions of mineralizers present in different parts of the molten mass.

It must not be supposed that a molten magma consists merely of a number of fused minerals, mechanically mixed together and having no effect upon one another. If such were the case, the minerals in cooling should all crystallize in the order of their fusibility, the least fusible forming first, and the most fusible last. This is not what we find, and many facts which cannot be discussed here have led petrographers to the belief that a molten magma is a solution of certain compounds in others, and that crystallization occurs in the order of solubility, as the point of saturation for particular compounds is successively reached by the cooling mass.

Similar phenomena may be observed among the metals. If strips of copper be thrown into a vessel of melted tin, the latter will dissolve the copper at a temperature far below that at which the copper would melt alone.

In a rock magma the crystallization of the more and more soluble minerals will proceed regularly, provided the pressure and rate of cooling continue constant. As these conditions are, however, subject to variation, it frequently happens that the more soluble minerals begin to crystallize before the less soluble have all been formed, and thus the periods of formation of two or more kinds of minerals partly overlap.

Usually, the order of formation of the different kinds of minerals in a solidifying magma is as follows. First to form are apatite, the metallic oxides (magnetite, ilmenite), and sulphides (pyrite), zircon, and titanite. "Next come the ferro-magnesian silicates, olivine, biotite, the pyroxenes, and hornblende. Next follow the felspars and felspathoids, nepheline and leucite, but their period often laps well back into that of the ferro-magnesian group. Last of all, if excess of silica remains, it yields quartz. In the variations of pressure and temperature, it may and often does happen that crystals are again redissolved, or resorbed, as it is called, and it may also happen that after one series of minerals, usually of large size and intratelluric origin, have formed, the series is again repeated on a small scale, as far back as the ferro-magnesian silicates. Minerals of a so-called second generation thus result, but they are always much smaller than the phenocrysts and are characteristic of the ground mass.

"It results from what has been said that the residual magma is u increasingly siliceous up to the final consolidation, for the earliest crystallizations are largely pure oxides. It is also a striking fact that the least fusible minerals, the felspars and quartz, are the last to crystallize." (Kemp).

A very considerable number of minerals are found in the igneous rocks, but comparatively few in any large quantity. It thus becomes necessary to distinguish between the essential minerals of a rock and the accessory ones. The essential minerals are those which characterize a given kind of rock, while the accessory minerals are those which occur in small quantities and which may be present or absent, without materially affecting the nature of the rock. The distinction is necessary and useful, but is rather arbitrary.

Another necessary distinction is that between original and secondary minerals. Original minerals were formed with or before the rock of which they are constituents, and secondary minerals are produced by the alteration or reconstruction of the original ones.

With comparatively few exceptions, the igneous rocks are made up of some felspar or felspathoid, together with one or more of the pyroxenes, amphiboles, micas, olivine, or quartz. Magnetite is also very common.