summer solstice; assuming this to reflect a rule that was adhered, in Athens, in
practice, many scholars have inferred that the Athenian year was regularly
intercalated in such a way that Hekatombaion, thefirst month, always began
after the solstice. The same is implicit, indeed, in the works of slightly later
Athenian authors.^57 These sources, however, do not confirm how regularly or
rigorously this rule—if indeed it was a rule—was followed.^58
Attempts have been made by Merritt and other scholars, in particular
Müller (1991) and Osborne (2009), to infer from epigraphic evidence that
the intercalation of months at Athens was not only regular but even followed a
fixed 19-year cycle (the Metonic cycle, on which see in more detail below,}3).
Ordinary and intercalated years attested in Athenian inscriptions between the
late second centuryBCEand the late second centuryCEare tabulated by Müller
(reproduced in Table 1.1) and would seem to indicate, atfirst sight, compati-
bility with such a cycle. The table is laid out in rows of 19 years, starting from
126/5BCEas year 1 of thefirst 19-year period. A consistent pattern appears to
emerge, with intercalated years occurring in years 3, 6, 8, 11, 14, 17, and 19 of
the 19-year periods. The only deviations from this pattern are thefirst entry of
the table, 126/5BCE, which was intercalated (instead of ordinary), and the last
entry, 211/12CE, ordinary (instead of intercalated); Müller identifies these two
years as the outer limits of his period, which is to say that the 19-year cycle was
only instituted at Athens from after 126/5BCE, and abandoned by 211/12CE.
All the data in Müller’s table are drawn from Athenian epigraphical sources,
with the exception of the intercalated years 91/0, 89/8, and 75/4BCE, which are
from coins of the kingdom of Pontus. The latter are clearly irrelevant to the
calendar of Athens, and should therefore be ignored (I have bracketed these
years in the table; the calendar of Pontus will be discussed in Chapter 5). This
leaves us with two very distinct periods: one extending from 125/4 to 95/4BCE,
reasonably well documented and apparently attesting the 19-year cycle, and
the other in the second centuryCE.^59 Although the data for the later period are
sporadic (only eleven entries for 111/12—177/8CE), Müller argues that their
compatibility with the 19-year cycle proves that the cycle was also then in use,
since there would have been a less than 0.1% probability for the data to agree
with the cycle purely at random. Finally, Müller extrapolates from these two
periods and concludes that the cycle must have been continuously in use from
the late second centuryBCEand to the late second centuryCE.
(^57) Aristotle,Hist. Anim. 543 b12, Theophrastus,Hist. Plant.4. 11. 4–5 (cited in Jones 2007:
150). See Samuel (1972) 64 (citing earlier scholarship) and Hannah (2005) 72, but some
reservations in Bickerman (1968) 37.
(^58) Plato’s reference further inLaws(7. 809D) to the calendar conforming to the courses of the
heavenly bodies is certainly not a realistic account of the Athenian calendar.
(^59) For the earlier period, Müller draws mainly on the works of Pritchett; for the later, on Follet
(1976) 354–6, 363–6, with evidence from ephebic lists of gymnasiarchs and prytanies of the 2nd
to the early 3rd cc.CE.
Calendars of AncientGreece 39