However, Julius Africanus probably only means that on average, over lengthy
historical periods, intercalations in the Greek (and Hebrew) calendars can be
assumed to occur at the rate of three in eight years (Stern 2001: 23). The theory
that the dates of the Olympic and Pythian games, and hence of the Elean and
Delphic calendars respectively, were regulated by the octaeteris is only a
tenuous inference from late sources (mainly Censorinus 18. 6) which, if
correct, reflects only the concerns of their astronomically minded late-antique
authors.^95 Epigraphic sources prove that at Athens, the octaeteris was not used
from the Classical period until at least as late as the second centuryCE(Follet
1976: 353–5, 359, 365–6).
The use of the 19-year cycle is sometimes inferred from a passage of the
historian Diodorus about Meton, where he comments that still in his day (i.e.
mid–latefirst cent.BCE)‘most Greeks’were using Meton’s 19-year cycle.^96
Diodorus cannot be referring to afixed, fully constituted astronomical calen-
dar with pre-determined month-lengths, because this would contradict his
own comment elsewhere (1. 50. 2) that‘most Greeks’again, seemingly also in
his day, still practised the intercalation and suppression of days. He probably
only means a sequence of intercalated years (i.e. seven intercalations in a cycle
of 19 years), which anyway is possibly all that Meton instituted (see above).
Although such a cycle was not regularly followed at Athens, the Greeks of Asia
Minor and the East in thefirst centuryBCEdid quite possibly follow it, and in
his period, Diodorus may have had good reason to refer to them as the
‘majority of Greeks’. However, this was not because they had adopted the
astronomical calendar of Meton (as Diodorus misleadingly suggests). It was
rather because as former subjects of the Seleucid Empire, they used the
Seleucid calendar (or derivative versions of it) which itself was assimilated to
the Babylonian calendar and based on a 19-year cycle.^97 As has been noted
above (}1), Greek calendars appear generally to have increased in regularity
during the Hellenistic period, most likely under Seleucid influence. This was a
political process, not the result of exposure to astronomical schemes.
Although astronomical cycles and schemes were not adopted whole piece as
civil calendars, they may still have been consulted for the purpose of regulating
the civil calendars and maintaining their conformity to the moon’s cycle and
the annual seasons. The city magistrates’ad hocdecisions to make months full
or hollow, to intercalate or suppress days, or to intercalate extra months in the
year, could thus have been based at least in part on the stable scheme that
(^95) So rightly Hannah (2005) 35–41, who, however, gives this theory far more attention than it
deserves.
(^96) Diodorus 12. 3, in the context of his account of Meton’s cycle:ìÝåæØ ôHí ŒÆŁ’ìAò åæüíøí
ïƒ ðºåEóôïØ ôHíEººÞíøí åæþìåíïØ ôfiB KííåÆŒÆØäåŒÆåôÅæßäØ ïP äØÆłåýäïíôÆØ ôBò IºÅŁåßÆò.
(^97) See Ch. 2; it is at least a reasonable assumption that the Babylonian calendar followed a 19-
year cycle of intercalations in this period, as it had done in earlier centuries. On the Seleucid and
post-Seleucid calendars, see Ch. 5.
52 Calendars in Antiquity