index, is raised to the head of the dragon;
conversely, when the little disk of the moon inclines to the tail, the
eclipse is taking place in the tail of the dragon.

And, accordingly, when you observe the little moon of the index inclined
to one or other section of the circle, so also in the heavens, the
eclipse of the moon is only partial and the northern or the southern
part of the moon is illuminated.

The current time will indicate whether the lunar eclipse is visible or
not. As the new moon ecliptic falls during the day, the eclipse will not
be visible, since the earth denies a sight of the moon which is below
the horizon. But, conversely, if there are no clouds, the eclipse will
be visible anywhere, if the luminous bodies are ecliptically in
opposition at night.

Since lunar eclipses appear to all people as being of the same magnitude
and duration, and begin and dissipate at the same absolute moment of
time, I decided to reveal another facet of this spectacle on the right
side of the center of the clock (see chapter III above). There, at the
time of the true ecliptic full moon, as the pointer of the third little
index shows, you can ascertain the hours, etc., of duration, and the
inches of greatest obscuration. The little moon attached to the index is
a model of the actual eclipsed moon.

IX

Thus, with the aid of this machine, solar and lunar eclipses of the past
can be recalled and future ones can be foreseen. Indeed, if the index of
prime minutes is speeded up by hand, whose wheel imparts motion to the
other indices and shields, then, the dials and openings will foretell
the year, month, day, hour, etc., of any future eclipse. I foresaw that
the times would thus be evolved too slowly, and that the clock wheels
would be considerably worn by repeated experiments (if, for instance, by
the rotation of the index of prime minutes, to whose period only a
single hour corresponds, the future new and full moon ecliptics were
being investigated). Therefore, I took care that the wheel which
immediately communicates motion to the index of the synodic moon should
be so fitted internally to the mechanism that by the reversal of any
external index, the wheel would be removed from its proper position;
whenever desired, it could be quickly and most accurately restored to
its proper place.

In this way, since the close meshing of the wheels is released, you can
extend the experiment for many years, even for many centuri