It is hardly necessary to insist that as in the case of the rocks composed of single minerals, already referred to above, the predominant mineral or minerals of compound rocks determine the facility of weathering, as well as the quality of the soil resulting therefrom. Since rocks are named essentially in accordance with the _kinds_ of minerals that constitute their regular mass, the _proportion_ in which the several constituents stand to each other may vary greatly. Thus a _granite_ may consist, over considerable areas, mainly of a mixture of potash feldspar and quartz; in others, mainly of quartz and mica with little feldspar. Very frequently, hornblende replaces mica partially or wholly. The latter will weather much more slowly than feldspar or hornblende, and will produce an inferior soil when decomposed. Allowing for such variations, a fairly approximate general estimate of the quality and peculiarities of soils from crystalline rocks may nevertheless be made. To some extent such estimates must make allowance not only for the chief ingredients, but also for those which are called “accessory” or characteristic, and which while not present in large amount, may nevertheless exert a considerable influence upon the quality of the soil. _Soils from granitic and crystalline rocks._—In the case of the (potash-feldspar) granite soils it is generally admissible to expect that they will be fairly supplied with phosphoric acid, because in the great majority of cases, minute crystals of apatite (phosphate of lime) are more or less abundantly scattered through it. From the potash feldspar present, granite soils may always be relied on for a good supply of potash for plant use; on the other hand, unless hornblende be present, they are pretty certain to be deficient in lime, since neither lime, feldspar nor calcite are probable accessory ingredients of this rock. Granite is exceedingly apt to weather by mechanical disintegration far in advance of its chemical decomposition. It is therefore common to find in sedentary soils overlying granite, a gradual increase of grains of its component crystalline minerals as we descend in the subsoil; until finally the latter grades off into rock almost unchanged save in lacking coherence. This is seen strikingly in the southern Appalachians, as well as in the Sierra Nevada and Sierra Madre of California; at Cintra in Portugal, at Heidelberg in Germany, and elsewhere. But of the rocks that resemble granite and are popularly so called, a good many are not “true to name” and therefore form soils differing materially from the type just mentioned.