is no occasion for the
action of water in dissolving siliceous substances in one place, in
order to concrete and crystallise it in another.

In these cavities of the solid granite rock, where crystal is found
regularly shooting from a basis which is the internal surface of the
cavity, we find the other constituent substances of the granite also
crystallised. I have those small cavities, in this rock, from the island
of Arran, containing crystal, felt-spar, and mica, all crystallised in
the same cavity[34]. But this is nothing to the _druzen_ or crystalline
concretions, which are found in a similar manner among metallic and
mineral substances in the veins and mines; there, every species of
mineral and metallic substance, with every variety of mixture and
composition, are found both concreted and crystallised together in every
imaginable shape and situation.

[Note 34: The Chevalier Dolomieu makes the following observation.
Journal de Physique, Juillet 1791.

"J'ai ete etonne de trouver au centre d'un enorme massif de granit, que
l'on avoit ouvert avec la poudre pour pratiquer un chemin, des
morceaux, gros comme le poing et au dessous, de spath calcaire blanc,
tres-effervescent, en grandes ecailles, ou lames entrecroisees. Il
n'occupoit point des cavites particulieres, il n'y paroissoit le
produit d'une infiltration qui auroit rempli des cavites, mais il etoit
incorpore avec les feld-spath, le mica, et le quartz, faissoit masse
avec eux, et ne pouvoit se rompre sans les entrainer avec lui."

This great naturalist is convinced that the spar had not been here
introduced by infiltration, although that is the very method which he
employs to form concretions, not only of spar but of crystal, zeolite,
and pyrites, in the closest cavities of the most solid rocks of
basaltes. These four substances in this stone were so mixed together
that nothing but the fusion of the whole mass could explain the state in
which they appeared; but, thinking that such a supposition could not
be allowed, this naturalist, like a man of science when his data fail,
leaves the matter without any interpretation of his own. This however is
what he has not done in the case of basaltes, or that which he mistakes
for proper lavas, as I shall have occasion to show.]

Here is an infinite operation, but an operation which is easily
performed by the natural arrangement of substances acting freely in
a fluid state, and concreting together, each substance, w