long-term time-reckoning frame; they are well attested in astronomical
sources from the early third centuryBCEto the mid-first centuryCE.^90
Astronomical and civil calendarsIt has been argued by some scholars that astronomical calendars, particularly
the Metonic and Callippic cycles, were adopted at Athens to regulate, and
possibly even determine, the festival (archontic) calendar.^91 The evidence,
however, does not support this theory. Deviations from the Metonic and
Callippic cycles are sufficiently attested in the epigraphic record at Athens to
conclude that in practice, a 19-year cycle was never consistently used.^92 Strict
adherence to the Callippic cycle would have been impossible without abolish-
ing the practice of intercalation and suppression of days, which is attested at
Athens long after the period of Callippus (see above,}1). Outside Athens,
likewise, there is no evidence that the Metonic or Callippic cycle was ever
adopted by civil calendars.^93
The use of the eight-year cycle (octaeteris) in Greek cities is often inferred
from the statement of the chronographer Julius Africanus (early third century
CE) that Greeks (and Jews) make three month-intercalations in eight years.^94
(^90) A. Jones 2000aand, for a summary, (2007) 149. From the 1st c.CE, the Egyptian calendar
came to be favoured instead.
(^91) Especially Meritt (1964); also Müller (1991), cited favourably by Jones (2000a: 156 n. 46),
but see my critical comments above (section}1);Woodhead (1997), inasmuch as he assumesa
priorithe Metonic cycle for the interpretation of all Athenian epigraphic datings (see above,
n. 44); and Osborne (2008), (2009), see above (}1).
(^92) This has been widely acknowledged by other scholars, including Samuel (1972) 55–61,
Woodhead (1992) 118–19, Bowen and Goldstein (1994), Pritchett (2001).
(^93) The use of local civil calendar month-names for the Metonic cycle in the Antikythera
mechanism (see above, n. 86) is taken by Freethet al. (2008) as evidence that‘it may have been
common for Greek civil calendars to follow the Metonic cycle by about 100BC’. ibid. 614); and
further,‘it is difficult to avoid the conclusion that the way that the calendar is structured on the
Mechanism...reflects the local practices of the time, since otherwise the information displayed
on the Mechanism would have frequently been in conflict with reality’(ibid. 17;‘local practices’
and‘reality’presumably mean here the local civil calendar). This inference is excessive and
unwarranted. The use of local month-names, rather than the normal Athenian names, in the
astronomical calendar of the Antikythera mechanism is most likely only an expression of local
patriotism, but without any bearing on how the local civil calendar was actually reckoned.
Perhaps it shows that the Mechanism’s authors believed, or rather liked to think, that their
local civil calendar could be rationalized as a Metonic cycle; but we need not assume an
expectation, on their part, that the Mechanism would consistently display actual civil calendar
dates—indeed, if their civil calendar was irregular and unpredictable, they must have realized
that no mechanism could possibly conform to their local calendar, and would therefore not have
regarded a discrepancy between the Mechanism and their local calendar as unexpected or
problematic.
(^94) e.g. Bickerman (1968) 30. The passage from Julius Africanus’Chronography,book 5,
appears in Eusebius,Proof of theGospel,8. 2 (390D.PG22. 609D–612 A; translation in
Mosshammer 2008: 390).
Calendars of AncientGreece 51