e, but was able to retaliate on his enemies by
landing on the coast of Africa at a point near the modern Cape Bon, and
devastating the Carthaginian territories. The voyage thither occupied
six days, and the eclipse occurred on the second day. Though we are not
informed of the route followed by Agathocles, that is to say whether he
passed round the North or the South side of the island of Sicily, yet it
has been made clear by astronomers that the southern side was that
taken.

Baily, who was the first modern astronomer to investigate the
circumstances of this eclipse, found that there was an irreconcilable
difference between the path of the shadow found by himself and the
historical statement, a gap of about 180 geographical miles seeming to
intervene between the most southerly position which could be assigned to
the fleet of Agathocles, and the most northerly possible limit of the
path of the eclipse shadow. This was the condition of the problem when
Sir G. B. Airy took it up in 1853.[55] He, however, was able to throw an
entirely new light upon the matter. The tables used by Baily were
distinctly inferior to those now in use, and Sir G. B. Airy thought
himself justified in saying that to obviate the discordance of 180 miles
just referred to "it is only necessary to suppose an error of 3' in the
computed distances of the Sun and Moon at conjunction, a very
inconsiderable correction for a date anterior to the epoch of the tables
by more than twenty-one centuries."

It deserves to be mentioned, though the point cannot here be dwelt upon
at much length, that these ancient eclipses all hang together in such a
way that it is not sufficient for the man of Astronomy and the man of
Chronology to agree on one eclipse, unless they can harmonise the facts
of several.

For instance, the eclipse of Thales, the date of which was long and much
disputed, has a material bearing on the eclipse of Agathocles, the date
of which admits of no dispute; and one of the problems which had to be
solved half a century ago was how best to use the eclipse of Agathocles
to determine the date of that of Thales. If 610 B.C. were accepted for
the Thales eclipse, so as to throw the zone of total darkness anywhere
over Asia Minor (where for the sake of history it was essential to put
it) the consequence would be that the shadow of the eclipse of 310 B.C.
would have been thrown so far on to land, in Africa, as to make it out
of the question for Agathocl