Now and then, but rarely, the student of dikes finds one where the bordering walls, in place of having the clean-cut appearance which they usually exhibit, has its sides greatly worn away and much melted, as if by the long-continued passage of the igneous fluid through the crevice. Such dikes are usually very wide, and are probably the paths through which lavas found their way to the surface of the earth, pouring forth in a volcanic eruption. In some cases we can trace their relation to ancient volcanic cones which have worn down in all their part which were made up of incoherent materials, so that there remains only the central pipe, which has been preserved from decay by the coherent character of the lava which filled it. The hypothesis that dikes are driven upward into strata by the pressure of the beds which overlie materials hot and soft enough to be put in motion when a fissure enters them, and that their movement upward through the crevice is accounted for by this pressure, makes certain features of these intrusions comprehensible. Seeing that very long, slender dikes are found penetrating the rock, which could not have had a high temperature, it becomes difficult to understand how the lava could have maintained its fluidity; but on the supposition that it was impelled forward by a strong pressure, and that the energy thus transmitted through it was converted into heat, we discover a means whereby it could have been retained in the liquid condition, even when forced for long distances through very narrow channels. Moreover, this explanation accounts for the fact which has long remained unexplained that dikes, except those formed about volcanic craters, rarely, if ever, rise to the surface.