which was probably conceived in the early fourth century, as its 84-year cycle
started in 298CE; it was probably in use in Rome by the 340s, and became
standard among Roman Christians until the mid-fifth century.^94
All these cycles were inventive and innovative, but in some respects,
continuous with pre-Christian traditions of lunar calendar reckoning.^95 In-
deed, they went well beyond the strict purpose of calculating the date of Easter,
for which all that would have been needed were the dates ofluna XIVand of
the following Sunday in every year of the cycle: the Easter cycles provided, in
fact, information required for the construction of full lunar calendars. The
table of Hippolytus, to begin with, marks with the Greek letter ̄the years in
the cycle that are embolismic, i.e. comprising thirteen lunar months; this
implies‘a properly theorized calendar rather than a mere sequence of dates’
(Holford-Strevens 2008: 170). ThesupputatioRomanaand all subsequent
Roman Easter cycles provide, in addition to the dates ofluna XIVand Easter,
the lunar dates of every 1 January, which are known as‘epacts’.^96 The epacts
are not essential for the calculation of the dates ofluna XIVand Easter Sunday.
Their inclusion in tables of thesupputatioRomanaimplies a scheme for
calculating the lunar dates of the whole year, as becomes explicit in late-
antique accounts of thesupputatio(as of later Roman Easter cycles), which
lay out a continuous lunar calendar with pre-determined,fixed-length months
or perhaps 297 by analogy with thesupputatioRomana); this early date might support a 3rd-c.
dating (as assumed by Krusch 1880: 4–23, Jones 1943: 15–16, Richard 1974: 316, Lejbowicz 2006:
21, and others: seeWarntjes 2007: 70 n. 113). The reconstruction of Augustalis’Laterculusis also
contentious: see Krusch, E. Schwartz, and (favouring Schwartz)Warntjes loc. cit.
(^94) ThesupputatioRomanais mentioned by name in a letter of Paschasinus to Leo in 444CE
(Krusch 1880: 248, E. Schwartz 1905: 41, Mosshammer 2008: 204), and described in detail in
early medieval manuscripts published by Krusch (1880: 227–44, with discussion on 31–115; see
E. Schwartz 1905: 40–58, Mosshammer 2008: 206–13, Holford-Strevens 2008: 173–8). However,
its earliest attested use is in the Codex of Philocalus of 354CE, where section VIII includes a list of
epacts running from the foundation of Rome until 354CEthat broadly conform to the epacts of
thesupputatio(with some exceptions: see Mosshammer 2008: 213–16 and above, n. 67) and
cannot be reconciled with any cycle other than that of thesupputatio. This does not prove, but
suggests at least, that thesupputatiowas in use in Rome by the 340s. The epoch of 298CEsuggests
that it was designed no later than the early 4th c.: Mosshammer (2008) 238.
(^95) Holford-Strevens (2008) 175. It has been argued, for example, that the lunar cycle of the
supputatioRomanawas originally designed not by Christians but by astrologers (H. Stern 1953:
55 – 7, followed by Salzman 1990: 36), on the grounds that thesupputatioRomanabegins on a
Saturday (1 Jan. 298,luna I), which is thefirst day of the planetary week, and not on a Sunday,
thefirst day of the Judaeo-Christian week (see further above, n. 65).
(^96) On the term‘epact’see Lejbowicz (2006) 27 n. 69. It isfirst attested in Ps-Cyprian,De
Pascha Computus,7(PL4. 1031B,epactae lunares); however, the word is missing in one of the
two manuscripts (Ogg 1955: 28 n. 1, Strobel 1984: 47), and its meaning in this passage is unclear
(see interpretations in E. Schwartz 1905: 40, Holford-Strevens 2008: 169 n. 13). The lunar dates
of 1 January are listed in section VIII of the Codex of Philocalus of 354CE(above, nn. 67, 94),
suggesting that by then January epacts were incorporated into the Roman Easter computation.
328 Calendars in Antiquity