acted upon, while the gravitational attraction is proportional to the
volume, or the total amount of matter in the body. But the surface of
any body depends upon the square of its diameter, while the volume
depends upon the cube of the diameter. If, for instance, the diameter
is represented by 4, the surface will be proportional to 4 × 4, or 16,
and the volume to 4 × 4 × 4, or 64; but if the diameter is taken as 2,
the surface will be 2 × 2, or 4, and the volume 2 × 2 × 2, or 8. Now,
the ratio of 4 to 8 is twice as great as that of 16 to 64. If the
diameter is still further decreased, the ratio of the surface to the
volume will proportionally grow larger; in other words, the pressure
will gain upon the attraction, and whatever their original ratio may
have been, a time will come, if the diminution of size continues, when
the pressure will become more effective than the attraction, and the
body will be driven away. Supposing the particles of the corona to be
below the critical size for the attraction of a mass like that of the
sun to control them, they would be driven off into the surrounding
space and appear around the sun like the clouds of dust around a mill.
We shall return to this subject in connection with the Zodiacal Light,
the Aurora, and Comets.

On the other hand, there are parts of the corona which suggest by
their forms the play of electric or magnetic forces. This is
beautifully shown in some of the photographs that have been made of
the corona during recent eclipses. Take, for instance, that of the
eclipse of 1900. The sheaves of light emanating from the poles look
precisely like the ``lines of force'' surrounding the poles of a
magnet. It will be noticed in this photograph that the corona appears
to consist of two portions: one comprising the polar rays just spoken
of, and the other consisting of the broader, longer, and less-defined
masses of light extending out from the equatorial and middle-latitude