the moon. According to many ancient authors, the Roman calendar in archaic
times had been strictly lunar: thus, Romulus (the founder of Rome) was said to
have begun the months on the day when the new moon wasfirst sighted,^135
and Numa (its second king) to have instituted a lunar calendar with a 20-year
cycle of intercalations.^136 Macrobius (late fourth–early fifth centuriesCE)
describes how, in ancient times, the new moon would be sighted each
month by one of the minor pontiffs, who reported his sighting to another
member of the pontifical college, therex sacrorum(‘king of sacrifices’); they
both made a sacrifice to mark the beginning of the month, and announced
to the assembled people how many days were to elapse between the Kalends
and the Nones.^137 Whatever we make of these legends and late-antique
accounts, the survival of lunar features in the Republican Roman calendar
supports the tradition that originally the Roman calendar was lunar.^138
(^135) Macrobius,Saturnalia1. 15. 5. Elsewhere Macrobius refers to the archaic Romans as using
a lunar calendar‘like the Greeks’(ibid. 1. 13. 8, also 1. 15. 20). But according to Plutarch (Numa
18), Romulus’calendar was a 360-day year with erratic month-lengths.
(^136) Livy 1. 19. 6; this 20-year cycle is not explained, but presumably Livy means really a
19-year cycle, counting thefirst years of the cycles inclusively (see Rüpke 1995: 203 n. 51). See
also Solinus 1. 37–8. Cicero (De Legibus2. 29) and more explicitly Plutarch (Numa18) disagree
with Livy and attribute to Numa the Republican system of intercalation; similarly Censorinus
(20. 4) attributes to either Numa or Tarquinius (Priscus, Rome’sfifth king) the 355-day year,
which probably means the Republican calendar. Intercalation, which in context refers again to
the Republican calendar, is attributed by others to Romulus, Numa, or Servius Tullius (Macr.Sat.
- 20). Julian claims that Numa’s calendar was simply solar (Hymn to KingHelios41, 155
A–B). Obviously, these conflicting legendary accounts are in themselves of limited historical
value. 137
Macrobius,Saturnalia1. 15. 9–12, partly following Varro,De Lingua Latina6. 27–8, who
mentions only the announcement of the‘pontiffs’and on that basis deriveskalendaefrom an
archaic Latin verbkalare(to call, to announce; Hannah 2005: 100, and with parallel sources
Rüpke 1995: 211 n. 85). Somewhat inconsistently, Varro also derivesnonaefromnov(us)/(a),
referring to the new month or new moon; his other suggestion, thatnonaemeans the ninth day
before the Ides, is far more plausible (Feeney 2007: 152, Lehoux 2007: 49). On the etymology of
Kalends, Nones, and Ides, see further Michels (1967) 19–21, 120 n. 5, 123, Brind’Amour (1983:
225 – 7), and Rüpke (1995: 210–14). On therex sacrorum, see Rüpke (1995: 313 n. 77) and Beard,
North, and Price (1998: i. 24–5). According to Livy, the minor pontiffs were originally called
‘pontifical scribes’(22. 57. 3scriba pontificius), which says perhaps something about their
original function within the pontifical college, but also implies that the title of‘minor pontiff’
in Macrobius’account is an anachronism (Lehoux 2007: 49).
- 20). Julian claims that Numa’s calendar was simply solar (Hymn to KingHelios41, 155
(^138) For a suggested reconstruction of this lunar calendar, see Rüpke (1995) 209–30, and of its
intercalation system, ibid. 295–319. On the question of whether the Roman New Year was at one
time on 1 March, see Michels (1967) 97–9, Samuel (1972) 164–5, and Brind’Amour (1983)
130 – 2, 225–7. Rüpke (1995) 193–7 plausibly argues for a multiplicity of New Year days in the
Roman year (for a similar phenomenon in ancient Near Eastern calendars, see Ch. 5, near nn.
13 – 14). On the related theory, already debated by the ancients (according to Censorinus 20. 2;
also Servius onGeorgica1. 43), that the Roman calendar consisted originally of a ten-month year
from March to December (possibly followed by an unnamed interval equivalent to two months),
see Michels (19670 207–20), Bickerman (1968) 44–5, Samuel (1972) 167–70, Rüpke (1995)
192 – 202, and Hannah (2005) 99; the theory has been much discussed, but remains completely
speculative.
208 Calendars in Antiquity