gh the fissures into
which they find their way are seldom riven up to the surface. In the
same way beneath the ground in non-volcanic countries we may discover
at a great depth in the older, much-changed rock a vast number of
these crevices, varying from a few inches to a hundred feet or more in
width, which have been filled with lavas, the rock once molten having
afterward cooled. In most cases these dikes are disclosed to us
through the down-wearing of the earth that has removed the beds into
which the dikes did not penetrate, thus disclosing the realm in which
the disturbances took place.

Where, as is occasionally the case in deep mines, or on some bare
rocky cliff of great height, we can trace a dike in its upward course
through a long distance, we find that we can never distinctly discover
the lower point of its extension. No one has ever seen in a clear way
the point of origin of such an injection. We can, however, often
follow it upward to the place where there was no longer a rift into
which it could enter. In its upward path the molten matter appears
generally to have followed some previously existing fracture, a joint
plane or a fault, which generally runs through the rocks on those
planes. We can observe evidence that the material was in the state of
igneous fluidity by the fact that it has baked the country rocks on
either side of the fissure, the amount of baking being in proportion
to the width of the dike, and thus to the amount of heat which it
could give forth. A dike six inches in diameter will sometimes barely
sear its walls, while one a hundred feet in width will often alter the
strata for a great distance on either side. In some instances, as in
the coal beds near Richmond, Va., dikes occasionally cut through beds
of bituminous coal. In these cases we find that the coal has been
converted into coke for many feet either side of a considerable
injection. The fact that the dike material was molten is still further
shown by the occurrence in it of fragments which it has taken up from
the walls, and which may have been partly melted, and in most cases
have clearly been much heated.

Where dikes extend up through stratified beds which are separated from
each other by distinct layers, along which the rock is not firmly
bound together, it now and then happens, as noted by Mr. G.K. Gilbert,
of the United States Geological Survey, that the lava has forced its
way horizontally between these layers, graduall