coincidence of Sunday and the beginning of a lunar month.^67 Indeed, it may
have been in anticipation of this triple coincidence that this calendar was
especially prepared.^68
If the seven-day column in this calendar began, therefore, on Sunday, it
need not have represented anything but the Judaeo-Christian week; there is no
reason to read anything astrological into it. The same applies to the lunar
calendar: if anything, analogy with the seven-day column would suggest that
the lunar column had some Christian purpose, e.g. to facilitate the calculation
of the dates of Easter (which are themselves listed in section IX of the same
codex),^69 rather than an astrological meaning of which there is no hint in this
section of the codex.^70
The astrological interpretation of thelunadates in funerary and other
inscriptions is also based on only tenuous evidence. One of the arguments is
that in at least two of thelunainscriptions, the date includes not only the lunar
(^67) Hannah (2005) 140–1 (but ignoring the coincidence with Sunday). The lunar conjunction
was actually on 31 December 354CE, but this does not mean that 1 January 355 could not have
been taken as the beginning of the lunar month. According to thesupputatioRomana(Christian
Easter cycle, on which see below), the epact (lunar date on 1 January) of 355 would have been
2 (the lunar month thus beginning on 31 December 354: Mosshammer 2008: 210); however, we
should not assume that the author of the Codex necessarily followed it. Epacts are listed in the
consular table in section VIII of the Codex, but the list unfortunately ends in 354.We can
establish, however, that this list does not consistently follow thesupputatio: e.g. the epact for 354
is given as 22, whereas according to thesupputatioit should have been 21; on the other hand, an
epact of 22 in 354 seems to preclude an epact of 1 in 355, so that section VIII of the Codex does
not clearly support my interpretation of the lunar column of section VI. On balance, however, it
seems to me that the triple coincidence (if only approximate) cannot be ignored.
(^68) Lehoux (2007) 192–4 argues that the lunar column was not tied to any particular year, with
A on 1 January representing the beginning of a lunar month; the lunar letters could be used in
fact in any year, as long as it was known that A represented each year a different day of the lunar
month, and as long as this day (the‘epact’) was known—just as the nundinal letters (from A to
H), in all Romanfasti, were used.Whilst Lehoux is right, in general, about the function of
nundinal letters in Romanfasti, and although the lunar letters A–K in this calendar are
reminiscent of nundinal letters, aflexible use of the lunar letters in this manner would have
disrupted, in some years, the correct distribution of 29- and 30-day months within the calendar
(in the calendar of Philocalus, as in the Roman Easter cycles, thefirst lunar month ending in
January must have 30 days, followed by 29- and 30-day months in alternation). It seems
preferable to argue that although the calendar could have been reused in other years (and may
even have been intended for this purpose), it was primarily designed for 355CE, when 1 January
coincided with the beginning of a lunar month appropriately represented with the lunar letter A.
(^69) Mommsen (1892) 62–4, on which see Salzman (1990) 39–41. Note that the weekdays (in
the calendar, section VI) would also have been essential for the calculation of the date of Easter,
which must always be a Sunday: H. Stern (1953) 57. Below I shall argue, however, that the main
purpose of the lunar column is more likely to have been calendrical (i.e. to be used as a dating
method).
(^70) PaceHannah (2005) 141. Although astrological interests are evident in some other sections
of the codex (section IV on the planets, and section V on the position of the moon in the zodiac;
Mommsen 1892: 42–7), this should not determine our interpretation of the calendar in section
VI.
320 Calendars in Antiquity