day after the conjunction, before the new moon was sighted or even theoreti-
cally visible.
Further ambiguities arise from a passage of Aratus,Phaenomena 733 – 9
(third centuryBCE), who writes that‘when the moon with slender horns is
sighted in the west, she declares a waxing month’(the text continues that when
moonlight begins to project a shadow, it is the 4th day of the month, etc.).
Geminus cites this passage as evidence that the days of the calendar month are
related to, and possibly determined by, the phases of the moon.^11 However, the
first sighting of the new moon, which Aratus is evidently referring to, is
associated here with a month that is already‘waxing’, which suggests some
time after the month’s beginning. Some commentators suggest that Aratus is
referring to an astronomical month that begins earlier on at the conjunction,
as opposed to the calendar month which began at the new crescent.^12 Alter-
natively, Aratus’term for‘waxing’,IåîïìÝíïØï, may be interpreted as referring
to thefirst day of the month, particularly if we assume it is interchangeable
withƒóôÆìÝíïØï, which designates, in the Athenian and many other Greek
calendars, thefirst ten days or‘decad’of the calendar month (including the
first day of the month).^13 Nevertheless, this passage does open the possibility
that the Greek calendar month began before thefirst sighting of the new
moon; Aratus’ambiguity is frustrating but perhaps, in itself, significant.
Modern scholars have generally accepted Geminus’statements that the
Greek month began atfirst visibility of the new moon.^14 However, it is clear
on Geminus’own evidence, as well as on the evidence of the other passages
just mentioned, that this rule was not strictly followed, and indeed, quite
possibly, that it did not constitute a rule. Moreover, we do not know of any
procedure that may have been used, and indeed that would have been neces-
sary, for such a rule to be enforced. There is no evidence, for example, in the
whole of Graeco-Roman literature of anyone sighting the new moon.^15 The
beginning of the month may have corresponded approximately tofirst visibil-
ity of the new moon—and this may have been, as suggested by Geminus and
(^11) Elem. Astr.8. 13 (Aujac 1975: 49); and so Pritchett (2001) 91–2.
(^12) J. Martin (1998) ii. 457. See also Bowen and Goldstein (1994) 696 n. 4, Kidd (1997) 426.
(^13) Samuel (1972) 57; on decads, see ibid. 60, 70, 101, 110. The termƒóôÆìÝíïØïwould be more
usual for a waxing moon (cf. Homer,Odyssey14. 162, 19. 307): Kidd loc. cit.
(^14) Bickerman (1968) 18; Samuel (1972) 57, 141.
(^15) Apollonius mentions the difficulty of sighting the new moon at the beginning of the month
(Argonautica4. 1479–80, cited in Dunn 1998: 218); but this says nothing of calendrical practice.
TheíïıìÅíØÆóôÆß(in Lysias fr. 195. 2, Carey 2007: 423) should be interpreted as a society meeting
on thenoumenia(first day of the month), rather than as a body of professionals in charge of
observing the new moon and/or setting the calendar at Athens (Pritchett 2001: 36). Observation
of the new moon is not listed as one of the archon’s duties, e.g. in Aristotle,Ath. Pol.56. Pritchett
(2001) 94–7 speculates on how the old and/or new moon may have been observed at Athens for
calendrical purposes; he deplores the lack of evidence, but without reflecting that this lack of
evidence itself may be significant. On Macrobius’account of new moon sighting in ancient, (pre)-
Republican Rome for purposes of setting the calendar month, see Chapter 4 near n. 137.
28 Calendars in Antiquity