sun keeps to it
accurately, but the planets wander somewhat above and below it (fig. 9),
and the moon wanders a good deal. It is manifest, however, in order that
there may be an eclipse of any kind, that a straight line must be able
to be drawn through earth and moon and sun (not necessarily through
their centres of course), and this is impossible unless some parts of
the three bodies are in one plane, viz. the ecliptic, or something very
near it. The ecliptic is a great circle of the sphere, and is usually
drawn on both celestial and terrestrial globes.

The earth's equator also produced into the sky, where it may still be
called the equator (sometimes it is awkwardly called "the equinoctial"),
gives another great circle inclined to the ecliptic and cutting it at
two opposite points, labelled respectively [Aries symbol] and [Libra
symbol], and together called "the equinoxes." The reason for the name is
that when the sun is in that part of the ecliptic it is temporarily also
on the equator, and hence is symmetrically situated with respect to the
earth's axis of rotation, and consequently day and night are equal all
over the earth.

Well, Hipparchus found, by plotting the position of the sun for a long
time,[2] that these points of intersection, or equinoxes, were not
stationary from century to century, but slowly moved among the stars,
moving as it were to meet the sun, so that he gets back to one of these
points again 20 minutes 23-1/4 seconds before it has really completed a
revolution, _i.e._ before the true year is fairly over. This slow
movement forward of the goal-post is called precession--the precession
of the equinoxes. (One result of it is to shorten our years by about 20
minutes each; for the shortened period has to be called a year, because
it is on the position of the sun with respect to the earth's axis that
our seasons depend.) Copernicus perceived that, assuming the motion of
the earth, a clearer account of this motion could be given. The ordinary
approximate statement concerning the earth's axis is that it remains
parallel to itself, _i.e._ has a fixed direction as the earth moves
round the sun. But if, instead of being thus fixed, it be supposed to
have a slow movement of revolution, so that it traces out a cone in the
course of about 26,000 years, then, since the equator of course goes
with it, the motion of its intersection with the fixed ecliptic is so
far accounted for. That is to say, the preces