Accordingly, these late authors commonly distinguish the day ofsynodos
(‘conjunction’), the last day of the month, from that ofnoumenia(‘new
month’), thefirst of the next month.^6 Conjunction is when the moon, along
its orbit, passes between the sun and the earth; it is then at minimal illumina-
tion, and completely invisible to the naked eye. The reason why the month
does not begin then, but only one day later (on thenoumenia), is presumably
that this is when the new moon becomesfirst visible.
However, the matter is not completely clear-cut. The same Geminus states
elsewhere that the new moon can sometimes be sighted on the 1st of the
month, but sometimes not until the 3rd.^7 This possibly means that although
the month begins when the new moonshouldfirst be visible, in practice it is
sometimes not sighted until the 3rd because of extraneous factors such as bad
weather. This would imply that the calendar did not depend on actual sight-
ings of the new moon, but rather on calculations or estimates of when the new
moon wouldfirst be visible. But although this interpretation is completely
plausible, it is equally possible that Geminus means that the Greek month did
not strictly followfirst visibility of the new moon, and could sometimes begin
a little earlier.^8
There are difficulties, furthermore, with the traditional ancient assumption
that the day of conjunction is the last day of the old month. The typical
interval between day of conjunction and day offirst visibility of the new moon
is actually two days; thus if the month began when the new moon wasfirst
sighted, conjunction should have been typically on thepenultimateday of the
previous month.^9 It may be argued that association of the conjunction with the
last day of the month was a literary convention, rather than a precise astro-
nomical truth.^10 Alternatively, it is possible that the month often began one
controversy, but sunset should be preferred (as at Censorinus,De die natali21. 6): see more
recently Bowen and Goldstein (1994) 695–6 and nn. 3–5, 713–14.
(^6) Geminus,Elem. Astr.8. 14 (Aujac 1975: 49); also scholia on Hesiod,Op. 765 – 8, 769
(comments attributed to Proclus). Plutarch (Solon25. 3) writes that Solon ordained at Athens
that the day of thesynodosshould be reckoned as‘old and new’(last day of the old month), and
the next day as 7 noumenia(first day of the new month).
Geminus,Elem. Astr.9. 14 (Aujac 1975: 61). The possibility of the new moon being sighted
on the 1st of the month seems to rule out an astronomical interpretation of this passage,
whereby the passage would be referring to a theoretical, astronomical lunar month running
from conjunction to conjunction. 8
See the remarks of A. Jones (2000a) 153 and n. 36.
(^9) See Stern (2001) 99–103.
(^10) Bowen and Goldstein (1994) 710–13 argue on the basis of P.Oxy. 53. 3710 (col. ii, ll.
33 – 43), Geminus,Elem. Astr.9. 6, 16, 10. 6 (Aujac 1975: 59–62), Plutarch (Romulus12, on the
foundation of Rome when a solar eclipse occurred on the last day of the month), and other
sources that the dating of solar eclipses—which necessarily occur at the time of conjunction—on
the last day of the month was only a later literary convention, which went back to an erroneous,
over-exacting reading of the text of Thucydides. On the apparently contradictory passages in
Thucydides (2. 28. 1, 4. 52. 1–2), where solar eclipses occur on thenoumenia, see Bowen and
Goldstein (1994) 702–7; it is clear that Thucydides does not mean a precise dating.
Calendars of AncientGreece 27