kilometres between the points of melting. This we can well understand when we know how irregularly the cone is constructed, and how buried coulées of lava may derange the direction of the fracture, such as we exaggerately see illustrated in some old denuded trap dikes, threading their way along planes of least resistance. There is another source of error—that is, that so little of the projecting edge of the dike is exposed to accurately take its strike, thus rendering us unable to determine by this means the locality of an old volcanic axis.
If we look at the figure, at the surface C' C'' of the subjacent rock, we observe it forms a wave-like line in section. It is again to Mr. Mallet[1] that credit is due for the explanation of this somewhat anomalous appearance. It is known that the ground under high towers and other heavy structures is gradually compressed by the immense superincumbent weight. At the same time a corresponding elevation takes place around the base of the structure. This is just what occurs in a volcanic mountain. The immense pressure of superposed material compresses, to a variable degree, the subjacent rock, according to its yielding power. This will be greatest where the column of materials is highest, that is to say, exactly under the crater edge as at C', in the diagram. This causes a corresponding rim-like elevation around the base, or at the toe of the cone as at C'', in the diagram.
The materials which go to form the cone are the subjects of our next consideration.
Taking as our standpoint the old but useful division of lavas into basaltic or basic, and trachytic or acidic, let us look at the characters presented by these two great classes of rocks. Basalt and its congeners are generally heavy, compact, dark-colored, more or less crystalline. Very rarely vitreous in structure, and only in small patches. Excessively fluid in the molten state, losing heat and fluidity slowly, and then passing rapidly from the liquid to the solid state, the liquid fragments of which, when ejected from the crater, generally fall still plastic, and, when cold, form an excessively ragged, hard, angular mass. The surface or scoria of the lava-stream also is hard, and not easily broken, the main mass itself being very apt to form the well-known columnar structure. On the other hand, the trachytic or acidic lavas, when molten, are very viscid, which condition increases rapidly as it loses its heat, so that it flows very short distances, often stopping midway down the steep side of the cone, as in the island of Vulcano, or forming a large, boss-shaped mass around the vent.[2] When cooled slowly it crystallizes, but it is much more liable to form a vitreous mass or obsidian than the basaltic rocks, resulting probably from its high percentage of silica. In fact, it behaves very much like glass or slag in its physical transformations. As on the surface of the glass pot is
