conclusive. Athanasius was bishop of Alexandria at various times between 328
and 373CE; his Festal Letters were written and disseminated annually, and
included an announcement of the dates of the forthcoming Easter. The Letters
are only extant until 348CE, after which only fragments have survived (unfor-
tunately, without the Easter dates). His dates of Easter, in the earlier period,
are mostly compatible with the Alexandrian cycle; however, this alone cannot
prove that the Alexandrian cycle was in existence or in use, because the same
Easter dates could have been reached using some other method.^113 Besides the
Festal Letters themselves, we also have an index of the Letters, which covers
the entire period of Athanasius’career and contains considerably more calen-
drical data than just the date of Easter: the Index provides, in particular, the
lunar date of every Easter Sunday and the epact of the year.^114 These lunar
dates and epacts correspond to those of the Alexandrian cycle and prove, this
time beyond doubt, that the Alexandrian cycle was used.^115 However, the
Index is a later work that is likely to have retrojected the Alexandrian
computation onto Athanasius’Easter dates; it cannot be treated as an authen-
tic record of how Athanasius himself, or his Alexandrian colleagues, reckoned
the date of Easter in the mid-fourth century.^116
(^113) Athanasius’Easter (Sunday) dates agree with those of the Alexandrian calendar, but he
does not give their lunar dates or (alternatively) the dates ofluna XIV(as these lunar dates would
not have been important to his addressees), which makes it impossible to establish on what basis
his Easter dates were reached. I write that Athanasius’Easter dates are‘mostly’compatible with
the Alexandrian cycle, because there are some exceptions; however, these exceptions could be
explained as deliberate deviations from the cycle for the sake of compromising with Rome, as will
be explained below (and as the Index explicitly states for the year 349).
(^114) In the Index and the Alexandrian cycle, the epact is the age of the moon on the last day of
the Egyptian year, i.e. 29 (or 30) August, as argued by Mosshammer (2008) 76–80.
(^115) See further Mosshammer (2008) 165.
(^116) Athanasius’Festal Letters have only survived in a Syriac translation, published by Cureton
(1848), of which there is a Latin translation inPG26. 1351–1444. A modern edition is still
awaited, but see Camplani (1989), Barnes (1993) 183–91, and Gwynn (2007) 45–8, 79–80. For
the Index, also only in Syriac, see Martin and Albert (1985). The headings of the Festal Letters
(which are also in Martin and Albert’s edition) sometimes contain the lunar dates of Easter, but
they are clearly a later, editorial addition to the Festal Letters, probably contemporary to the
Index with which the headings generally agree. The Index and headings are tentatively dated by
Camplani (1989) 115–29, 190–3toc.400CE, but this is only a guess; Martin and Albert (1985)
221 more cautiously say just that the Index was originally composed in Greek, and probably in
Alexandria.Whenever the Index and headings were composed, the Alexandrian cycle was clearly
well established; but this says nothing of how Athanasius himself reckoned his Easter dates. The
Index is also likely to have used the Alexandrian cycle to calculate the Easter dates of the years in
which Athanasius did not write a Festal Letter (more on this below). Earlier scholars (e.g.
E. Schwartz 1905: 24–5, Neugebauer 1979: 98–101), therefore, placed excessive reliance on the
Index as evidence of the Alexandrian cycle in the mid-4th c. It should be noted, furthermore, that
the scheme implicit in the data of the Index differs slightly from the 19-year cycle that became
normative in the 5th c., as itssaltuslunaeappears to have occurred one year later, in 343 instead
of 342CE(and similarly 19 years later): for the Index consistently gives the epact of 29 for 342 and
361 CE, whereas according to the Alexandrian cycle as later known, it should have been 30
(because of thesaltus). E. Schwartz (1905) 24–5 emends it accordingly, in both years, to 30; his
emendation is justified by the lunar date of Easter for 361CE, which is given in the Index as
Sectarianism andHeresy 403