th the confusion
of an unorganised mob, and not with the steady accuracy of a
disciplined host. In this mass of bismuth also we have an example of
confused crystallisation; but in the crucible behind me a slower
process is going on: here there is an architect at work 'who makes no
chips, no din,' and who is now building the particles into crystals,
similar in shape and structure to those beautiful masses which we see
upon the table. By permitting alum to crystallise in this slow way,
we obtain these perfect octahedrons; by allowing carbonate of lime to
crystallise, nature produces these beautiful rhomboids; when silica
crystallises, we have formed these hexagonal prisms capped at the ends
by pyramids; by allowing saltpetre to crystallise we have these
prismatic masses, and when carbon crystallises, we have the diamond.
If we wish to obtain a perfect crystal we must allow the molecular
forces free play; if the crystallising mass be permitted to rest upon
a surface it will be flattened, and to prevent this a small crystal
must be so suspended as to be surrounded on all sides by the liquid,
or, if it rest upon the surface, it must be turned daily so as to
present all its faces in succession to the working builder.

In building up crystals these little atomic bricks often arrange
themselves into layers which are perfectly parallel to each other, and
which can be separated by mechanical means; this is called the
cleavage of the crystal. The crystal of sugar I hold in my hand has
thus far escaped the solvent and abrading forces which sooner or later
determine the fate of sugar-candy. I readily discover that it cleaves
with peculiar facility in one direction. Again I lay my knife upon
this piece of rocksalt, and with a blow cleave it in one direction.
Laying the knife at right angles to its former position, the crystal
cleaves again; and finally placing the knife at right angles to the
two former positions, we find a third cleavage. Rocksalt cleaves in
three directions, and the resulting solid is this perfect cube, which
may be broken up into any number of smaller cubes. Iceland spar also
cleaves in three directions, not at right angles, but oblique to each
other, the resulting solid being a rhomboid. In each of these cases
the mass cleaves with equal facility in all three directions. For the
sake of completeness I may say that many crystals cleave with unequal
facility in different directions: heavy spar presents an e