Granite is found to penetrate the stratified rocks in the form of veins. The following section (Fig. 3) will show the relation of granite veins to the granitic mass below. The granite which is quarried for architectural purposes is often in comparatively small quantities, disappearing at the distance of a few hundred yards beneath the stratified rock; or else it exists in the form of isolated dome-shaped masses. It is probable that, if they could be followed sufficiently far, they would be found to be portions of dikes coming from the general mass of granite below. Even the granite nuclei of the great mountain ranges may be considered as injected dikes of enormous magnitude. [Illustration: Fig. 4.] Granite is itself intersected with granite veins more frequently, perhaps, than any other rocks; but the vein is a coarser granite than the rock which it divides. It is not uncommon to find one set of dikes intercepted and cut off by a second set, and the second by a third. The substance of the dikes was, of course, in a liquid state when it was injected, and the first must have become solid before the second was thrown in; hence the dikes are of different ages. The dikes _a b c_, represented in Fig. 4, must have been injected in the order in which they are lettered. It is probable that, by the process of cooling, the liquid mass from which these dikes have proceeded has been gradually solidifying from the surface downwards. If so, it would follow that the granite nearest the surface (1, Fig. 2) is the oldest, and the newest is that which is at the greatest distance below (4). It is possible that at great depths granite may be still forming, that is, taking the solid form, though of this there can be no direct proof. There is, however, proof that it has been liquid at periods of time very distant from each other; for the dikes sometimes reach to the top of the coal formation (for example), and then spread themselves out horizontally, as at _a_, showing that the rock above the coal had not then been deposited. Another dike will extend through the new red sandstone, as at _b_, and spread itself out horizontally as before. These horizontal layers of granite, by their position in strata whose ages are known, indicate the periods when granite has existed in a liquid state. Granite veins have been discovered in the Pyrenees as recent as the close of the cretaceous period, and in the Andes they have been found among the tertiary rocks. There are several other rocks, of minor importance, often found in connection with granite. Hypersthene rock, in a few cases, forms the principal part of mountain masses. Greenstone is more frequently associated with the trappean rocks, but it sometimes passes imperceptibly into syenite and common granite. Limestone is found in considerable abundance, and serpentine in small quantities, as primary rocks, and have evidently been formed like granite, by solidifying from a state of fusion.