molecules to prevail, resulting in the inception of crystallization. To
discover this requires the keenest observation. When beginning to form,
the crystals are too minute to show either form or size, even when
viewed through a strong magnifying glass. There is to be seen simply a
very delicate cloud. The inexperienced observer would entirely overlook
this cloud, his attention probably being directed to some curious
globular and annular objects, which I have nowhere seen explained. Very
soon after the sample from the pan is placed upon glass for observation,
the surface becomes cooled and somewhat hardened. As the cooling
proceeds below the surface, contraction ensues, and consequently a
wrinkling of the surface, causing a shimmer of the light in a very
attractive manner. This, too, is likely to attract more attention than
the delicate, thin cloud of crystals, and may be even confounded with
the reflection and refraction of light, by which alone the minute
crystals are determined. The practical operator learns to disregard all
other attractions, and to look for the cloud and its peculiarities. When
the contents of the pan have again reached the proper density, another
portion of sirup is added. The sugar which this contains is attracted to
the crystals already formed, and goes to enlarge these rather than to
form new crystals, provided the first are sufficiently numerous to
receive the sugar as rapidly as it can crystallize.

The contents of the pan are repeatedly brought to the proper density,
and fresh sirup added as above described until the desired size of grain
is obtained, or until the pan is full. Good management should bring
about these two conditions at the same time. If a sufficient number of
crystals has not been started at the beginning of the operation to
receive the sugar from the sirup added, a fresh crop of crystals will be
started at such time as the crystallization becomes too rapid to be
accommodated on the surfaces of the grain already formed. The older and