result of the pressure of light. The reader will recall the
applications of this theory to the Zodiacal Light and the Aurora. In
the form in which we now have to deal with it, the supposition is made
that as a comet approaches the sun eruptions of vapor, due to the
solar heat, occur in its nucleus. These are naturally most active on
the side which is directly exposed to the sun, whence the appearance
of the immense glowing envelopes that surround the nucleus on the
sunward side. Among the particles of hydro-carbon, and perhaps solid
carbon in the state of fine dust, which are thus set free there will
be many whose size is within the critical limit which enables the
light-waves from the sun to drive them away. Clouds of such particles,
then, will stream off behind the advancing comet, producing the
appearance of a tail. This accounts for the fact that the tails of
comets are always directed away from the sun, and it also explains the
varying forms of the tails and the extraordinary changes that they
undergo. The speed of the particles driven before the light-waves must
depend upon their size and weight, the lightest of a given size
traveling the most swiftly. By accretion certain particles might grow,
thus losing velocity and producing the appearance of bunches in the
tail, such as have been observed. The hypothesis also falls in with
the researches of Bredichin, who has divided the tails of comets into
three principal classes -- viz.: (1) Those which appear as long,
straight rays; (2) Those which have the form of curved plumes or
scimitars; (3) Those which are short, brushy, and curved sharply
backward along the comet's path. In the first type he calculates the
repulsive force at from twelve to fifteen times the force of gravity;
in the second at from two to four times; and in the third at about one
and a half times. The straight tails he ascribes to hydrogen because
the hydrogen atom is the lightest known; the sword-shaped tails to
hydro-carbons; and the stumpy tails to vaporized iron. It will be seen