be only a
comparatively faint condensation, and may or may not develop into a
point of light as the comet approaches the sun. A tail also is generally
not seen at great distances from the sun, but gradually develops as the
comet approaches perihelion, to fade away again as the comet recedes
from the sun.

A few comets are known to revolve in orbits with a regular period,
while, in the case of others, no evidence is afforded by observation
that the orbit deviates from a parabola. Were the orbit a parabola or
hyperbola the comet would never return (see ORBIT). Periodicity may be
recognized in two ways: observations during the apparition may show that
the motion is in an elliptic and not in a parabolic orbit; or a comet
may have been observed at more than one return. In the latter case the
comet is recognized as distinctly periodic, and therefore a member of
the solar system. The shortest periods range between 3 and 10 years. The
majority of comets which have been observed are shown by observation to
be periodic; the period is usually very long, being sometimes measured
by centuries, but generally by thousands of years. It is conceivable
that a comet might revolve in a hyperbolic orbit. Although there are
several of these bodies observations on which indicate such an orbit,
the deviation from the parabolic form has not in any case been so well
marked as to be fully established. Circumstances lead to the
classification of newly appearing comets as _expected_ and _unexpected_.
An expected comet is a periodic one of which the return is looked for at
a determinate time and in a certain region of the heavens. When this is
not the case the comet is an unexpected one.

_Physical Constitution of Comets._--The subject of the physical
constitution of these bodies is one as to the details of which much
uncertainty still exists. The considerations on which conclusions in
this field rest are very various, and can best be set forth by beginning
with what we may consider to be the best established facts.

We must regard it as well established that comets are not, like planets
and satellites, permanent in mass, but are continuously losing minute
portions of the matter which belongs to them, through a progressive
dissipation--at least when they are in the neighbourhood of the sun.
When near perihelion the matter of a comet is seen to be undergoing a
process in the nature of evaporation, successive envelopes of vapour
rising from the nu