have been made on the provisional assumption that the month would be full
(30 days). If the month turned out to be hollow, the 29th day became the last
day of the month and was re-named‘old and new’(instead of‘2nd’, as it would
have been named in a 30-day month), and‘2nd’was purely and simply
suppressed from the sequence.^68 Strepsiades’countdown from 5 to‘old and
new’without suppression of the 2nd does not necessarily imply, therefore, that
he knew in advance that this month would be full. His apparent foreknow-
ledge is only the reflection of a conventional counting system which was built
on the assumption—later to be corrected if necessary—that the month was
always full.^69
It has been argued, however, that if Strepsiades was dreading the coming of
‘old and new’, he should have feared, in his countdown, the possibility of a
hollow month and hence of‘old and new’coming one day earlier. This would
support the view that length of the month—in this case, full—had been set in
(^68) Pritchett and Neugebauer (1947) 3–4, 24–5, Samuel (1972) 59–61, Pritchett (1982), (2001)
41 – 87, 117–25, Hannah (2005) 43. This is against the theory of Meritt (1961), following earlier
studies (for other references see Bickerman 1968: 100 andWalsh 1981, followed byWoodhead
1992: 120–1; 1997: 113) that the length of the Athenian month was determined in advance by
observation of the gibbous moon around the 20th of the month, and not (as one generally
expects in the context of the ancient world) on the basis of observation of the new moon (or
alternatively of the disappearance of the old moon). According to this theory, the 21st day of the
month (i.e. 10th from the end) was the day omitted if it was decided that the month should be
hollow. However, this bizarre theory is convincingly refuted by Pritchett and Neugebauer (1947)
23 – 31, Pritchett (1982), (2001) 41–87, 117–25, for a number of reasons. Firstly, regulation of the
moon according to the gibbous moon would have been highly impractical, because without
precise instruments or an accurate lunar theory, the shape and size of the gibbous moon would
have been very difficult to measure empirically or to calculate. It seems highly unlikely, therefore,
that such a method was ever thought of or used. Secondly, the theory of the gibbous moon
calendar has no positive evidence to support it, whereas Pritchett can cite as evidence at least a
fragment attributed to Proclus in the scholion to Hesiod,Op. 765 – 8, which states that it is the
penultimate day that is omitted in a hollow month. In some places outside Athens, furthermore,
the evidence confirms that it was the penultimate day that was normally omitted: the last day of
the month was calledtriakas(literally‘30th’) regardless of whether the month was hollow or full,
whereas the penultimate day (‘2nd’in the Athenian calendar) was calledprotriakas(‘before the
30th’) and omitted in hollow months: Samuel (1972) 69 (Boeotia), and 110 (Rhodes).Walsh
(1981) 110, 112 n. 9 predictably argues that Athens was different, basing himself on the
assumption that Solon would have attempted to make the calendar more predictable, by
instituting a method to determine month-lengths in advance; but such a concern for predict-
ability is modern, and not necessarily relevant to ancient Athens. Athenian society was surely
adapted to an unpredictable calendar, especially as the calendar was subject to archontic
tampering (as I shall argue further below); there is no evidence that calendrical unpredictability
was ever considered at Athens, in itself, a problem.
(^69) Note also the common assumption of 30-day months for the calculation of lengthy periods,
particularly for accounting purposes, on which see Hannah (2005) 34–5, referring to an
Athenian record of wages (CIAii. 2, 834c, l. 60) and passages from Hippocrates (On Flesh19.
27 – 8) and Ps.-Aristotle (Economics 1351 b 11 – 15); also Pritchett (2001) 113, 151 n. 22 and
Bickerman (1968) 37. The use of 30-day months for accounting purposes is also attested in
Babylonia, as we shall see in Ch. 2.
Calendars of AncientGreece 45