of 30 and 29 days in alternation.^97 Although epacts are listed only in the
supputatioRomanaand later Easter calendars, it has been convincingly argued
that the epact of 1 January is implicit already in the cycle of Hippolytus, which
was clearly constructed on the basis of an epact ofluna Ion 1 January at the
beginning of the cycle in 222CE.^98 From all this it is clear that the creators of
Roman Easter cycles, starting from Hippolytus, had more than just an interest
in the date of Easter: they sawluna XIVas part of a lunar calendar that needed
to be constructed in full.^99
This interest in lunar calendars was related, perhaps, to the popularity of
lunadates in Christian funerary inscriptions, and both in turn were continu-
ous with the Roman (or Latin, or Italian) tradition of lunar reckoning which
has been discussed above. It is true that the Christians of third-century Rome,
some of whom may have identified as Jews and many of whom were Greek-
speaking (as is manifest from the very fact that Hippolytus’cycle was inscribed
in Greek), would not have been readily associated with local Latin traditions.
Nevertheless, the subcultural status of Christians in Rome may have led them
to associate themselves with practices such as Latin lunar dating which had
become, in the Roman imperial period, subcultural. The combination in the
Easter tables of a Judaeo-Christian tradition (the date of Passover) with a Latin
tradition of lunar calendars can be interpreted as a statement of solidarity
between various subcultural traditions in third-century Rome, resulting in
‘horizontal’hybridity between them. The use oflunadates in funerary in-
scriptions and the creation and use of lunar, Easter calendars is thus likely to
have reinforced in a very subtle and implicit manner the subcultural, dissident
(^97) Krusch (1880) 51, with texts on 233–5, 242–3. Thefirst lunar month (which ends at some
point in the month of January) always counts 30 days; in Julian leap years, the second lunar
month (normally 29 days) counts an extra day; in some cases, the last lunar month (ending in
December) is shortened by one day (for thesaltuslunae). For a concise exposition see Blackburn
and Holford-Strevens (1999) 801–3. The lunar column of the calendar of Philocalus (section VI
of the Codex) follows the same scheme.
(^98) Mosshammer (2008) 122–4 and Holford-Strevens (2008) 168–72, who comments:‘That
suggests we have to deal, not with oneluna XIIIdetermined empirically and the rest by
subtraction and addition, but a calendar based on calculation by lunar epacts.’This, incidentally,
makes it even less likely that the cycle was of Alexandrian origin: as Holford-Strevens demon-
strates, it is clear at least that Hippolytus 99 ’cycle was a completely original design.
My discussion here is restricted to the RomanWest, and more particularly to the city of
Rome, but a similarly argument applies to the Easter cycles that emerged in the later 3rd c. in the
East (see further Ch. 7). Indeed, the 19-year cycles of Anatolius (Mc Carthy and Breen 2003,
Lejbowicz 2006: 24–9) and of Alexandria (as argued by Mosshammer 2008: 76–80, though on the
basis on later evidence) also include an epact earlier in the year (on 1 January and on the last
epagomenal day respectively), implying continuous lunar calendars. These cycles, furthermore,
are clearly derived from earlier, pre-Christian local traditions: the 19-year cycle is derived from
Metonic astronomical calendars, and the lunar calendar itself from the tradition of Greek and
Seleucid civil calendars (on which see Ch. 1)—just as I am arguing that in Rome, Easter cycles
were inspired by local traditions of lunar reckoning.
Dissidence and Subversion 329