One finds analogous facts in the other systems of crystals.
Whence come these apparent exceptions to that regular-
ity which, really without knowing just why, we attribute
to the laws of nature?

Haüy was very familiar with these phenomena of the
hemihedron and if he did not attribute to them any
great importance it is because his theory led him to
a somewhat distorted view of them, as I have just said.
According to his conception the form of the integral
molecule was, first of all, that which cleavage, the natural
division of the crystal, gave to it. A cubical crystal of
marine salt produces cubes by cleavage; a rhombohedral
crystal of Iceland spar gives in the same way rhombo-
hedrons. The rhombohedron was, therefore, for Haüy
a primitive form. When we intersect the six lateral
angles by planes having the same angle of inclination
to the faces of the rhombohedron, we obtain by a
perfectly regular process of derivation, the hexagonal
prism of quartz. And so for the other cases. This
conception formed a logical and coherent whole, but left
Haüy indifferent to the questions of the hemihedron.

In order to understand the hemihedral character
of the rhombohedron, it is necessary to reverse the
order and take the hexagonal prism as the primitive
form. Then the rhombohedron can be derived from
it only by way of the hemihedron. The same is true
in the other systems. Weiss, the mineralogist, did
this and straightway the hemihedron appeared to
be a phenomenon more frequent than was supposed,
and there arose a problem requiring solution. Why
this deviation from the law of symmetry?

This is what Delafosse tried to explain in 1840,
by the aid of a deceptive hypothesis which to-day
seems very childish. "In the prismatic quartz," he
said, "the hemihedral constitution exists without being

-6-