From a photograph by Messrs. W. A. Mansell.)]

But it had one drawback, which the ancients could not have been expected
to foresee. The effect of "precession," alluded to in the chapter on
"The Origin of the Constellations," p. 158, would be to throw the
beginning of the year, as thus determined, gradually later and later in
the seasons,--roughly speaking, by a day in every seventy years,--and
the time came, no doubt, when it was noticed that the terrestrial
seasons no longer bore their traditional relation to the year. This
probably happened at some time in the seventh or eighth centuries before
our era, and was connected with the astronomical revolution that has
been alluded to before; when the ecliptic was divided into twelve equal
divisions, not associated with the actual stars, the Signs were
substituted for the Constellations of the Zodiac, and the Ram was taken
as the leader instead of the Bull. The equinox was then determined by
direct measurement of the length of the day and night; for a tablet of
about this period records--

"On the sixth day of the month Nisan the day and night were
equal. The day was six double-hours (_kasbu_), and the night
was six double-hours."

So long as Capella was used as the indicator star, so long the year must
have begun with the sun in Taurus, the Bull; but when the re-adjustment
was made, and the solar tropical year connected with the equinox was
substituted for the sidereal year connected with the return of the sun
to a particular star, it would be seen that the association of the
beginning of the year with the sun's presence in any given constellation
could no longer be kept up. The necessity for an artificial division of
the zodiac would be felt, and that artificial division clearly was not
made until the sun at the spring equinox was unmistakably in Aries, the
Ram; or about 700 B.C.

The eclipse of 1063 B.C. incidentally proves that the old method of
fixing Nisan by the conjunction of the moon and Capella was then still
in use; for the eclipse took place on July 31, which is called in the
record "the 26th of Sivan." Sivan being the third month, its 26th day
could not have fallen so late, if the year had begun with the equinox;
but it would have so fallen if the Capella method were still in vogue.

There is a set of symbols repeated over and over again on Babylonian
monuments, and always given a position of eminence;--it is the so-called
"Triad of Stars," a