calendar of‘Cyprus’differ in the Vatican and Florence MSS, suggesting that
these in fact might be different calendars.^132
Calendar diversity within the province of Asia in the earlyfirst centuryCEis
supported, furthermore, by epigraphic evidence. An inscription from the city
of Philadelphia, dating from 40CE, implies a New Year on 23 September,
conforming to the provincial decree of 8BCE; whereas an inscription from
Metropolis, dated palaeographically to the reign of Tiberius (14– 37 CE),
assumes a New Year on 24 September, as in the calendar of‘Asia’of the
Florence MS.^133 Although this New Year date of 24 September could go back
to the original institution of the calendar in Metropolis in 8BCE,^134 it seems
more likely that deviations from the standard New Year date of 23 September,
Augustus’birthday, developed after Augustus’death in 14CE, when celebra-
tions of his birthday began to lose importance and to wane.^135 Shifts of the
New Year to 24 September in Metropolis (and possibly other cities of Asia),
and perhaps also to 22 September in some parts of Cyprus, are thus quite likely
to have begun in Tiberius’reign.^136
The epigraphic evidence fromfirst-centuryCEAsia, to which may be added
further epigraphic deviations from the decree of Priene and thehemerologiain
later centuries, together with the inconsistencies between thehemerologia
themselves that have been noted above, suggest that although the calendar
of the province of Asia, instituted in 8BCEand widely diffused in and around
(^132) The most distinct example is the eighth month, which is called ̃ÞìÆæåïòin the Florence
MS and 133 ̃ÅìÞôæØïòin the Vatican MS: see Kubitschek (1915) 106.
Philadelphia inscription: see above, n. 105. Metropolis inscription, referred to above as a
two-columnhemerologion: Engelmann (1999) 142–3. The latter equates the 14th (of the local
month) with the Nones of October, the 15th with the eighth before the Ides of October, etc.
(^134) Unfortunately, the Metropolis fragment of the decree of 8BCE(Dreyer and Engelmann
2006: 175–82) lacks the list of month-lengths; we cannot tell whether it prescribed the same
sequence of months as the Priene inscription, or whether it assumed a New Year on 24
September.
(^135) Price (1984) 61–2 notes the tendency for birthday cults to be silently superseded after the
ruler’s death, although in the case of Augustus, celebration of his birthday is still attested at
Pergamum in the Hadrianic period, early 2nd c.CE(IGRiv. 353b =I.Pergamum374B, cited in
Laf 136 fi1967: 76).
The practice of extending Augustus’birthday celebrations to a two-day period on 23 and
24 September, which is attested in the East (with twoSebastedays at the beginning of the month,
as in the Asia–Pamphylia calendar of the Vatican MS; see Table 5.6 n.c) as well as in northern
Italy (CILxi. 3303, from Forum Clodi, 18CE) and Gaul (CILxii. 4333, from Narbo, 11CE), has
been explained as arising from possible confusion or uncertainty regarding the‘correct’date of
Augustus’birthday according to the Julian calendar (since he was born before this calendar was
instituted); his original birthday date was known asa.d. VIIIkal. Oct., but confusion arose as to
whether this was an old calendar-date (i.e. 23 September), which in the new calendar becamea.d.
IXkal. Oct.(as September acquired an extra day), or a new-calendar date (i.e. 24 September).
This would explain why the Forum Clodi inscription dates the birthdayto a.d. VIIIkal. Oct.,
against all Romanfasti(as well as the decree of the calendar of Asia) that date itto a.d. IXkal.
Oct.: so Degrassi (1963) 513, Michels (1967) 180–1, Hannah (2005) 124–5, Feeney (2007 (154).
Although this explanation is largely conjectural, it raises the possibility that alternative New Year
dates, especially that of 24 September, could go back already to the reign of Augustus.
Fragmentation: Babylonian and Julian Calendars 283