l manifestly erroneous, for
there never could have been any real fixed or steady correlation between
the constellations and the months. Similarly, the theories which claim
that the ancient names for the months were derived from the
constellations are equally untenable. Some writers have even held both
classes of theory, overlooking the fact that they mutually contradict
each other.

But there came a time when the inconvenience of the unequal division of
the zodiac by the constellations was felt to be an evil, and it was
remedied by dividing the ecliptic into twelve equal parts, each part
being called after the constellation with which it corresponded most
nearly at the time such division was made. These equal divisions have
been called the _Signs_ of the zodiac. It must be clearly understood
that they have always and at all times been imaginary divisions of the
heavens, that they were never associated with real stars. They were
simply a picturesque mode of expressing celestial longitude; the
distance of a star from the place of the sun at the spring equinox, as
measured along the ecliptic,--the sun's apparent path during the year.

The Signs once arranged, the next step was an easy one. Each sign was
equivalent to 30 degrees of longitude. A third of a sign, a "decan," was
10 degrees of longitude, corresponding to the "week" of ten days used in
Egypt and in Greece.

This change from the constellations to the Signs cannot have taken place
very early. The place of the spring equinox travels backwards amongst
the stars at the rate of very little more than a degree in 72 years.
When the change was made the spring equinox was somewhere in the
constellation _Aries_, the Ram, and therefore Aries was then adopted as
the first Sign, and must always remain such, since the Signs move
amongst the stars with the equinox.

[Illustration: POSITION OF SPRING EQUINOX, B.C. 2700.]

We cannot fix when this change was made within a few years, but it
cannot have been _before_ the time when the sun at the spring equinox
was situated just below _Hamal_, the brightest star of the Ram. This was
about 700 B.C. The equal division of the zodiac must have taken place
not earlier than this, and with it, the Bull must have been deposed from
the position it had always held up to that time, of leader of the
zodiac. It is probable that some direct method of determining the
equinox itself was introduced much about the same time. This new system
invo