duplicated, scribal error cannot be invoked so easily.^129 The New Year of 24
September in the Asia column of the Florence MS shows no signs of error, as
the previous day (23 September) is consistently counted here as day 30 of
the previous month (Laodikon).^130 If this New Year date is not a scribal error,
the same just as well can be assumed for date of 22 September (in Cyprus,
Vatican MS).
The possibility of error may be apparent where calendars of the same name
(Asia, Bithynia, Cyprus, Crete) differ from onehemerologionto the next: for
example, in the Florence MS the calendar of Asia has the incongruous
New Year of 24 September, but in the other manuscripts, the New Year of
Asia is 23 September; similarly, in the Vatican MS the Cypriot New Year is 22
September, but in the Florence MS it is 23. Inconsistencies such as these arise,
no doubt, from thehemerologiabeing distinct literary compositions, laid out
in different ways and with calendars of not exactly the same cities and
provinces; thehemerologiamay have been redacted in different periods and
places, by authors of variable competence and knowledge, and on the basis of
variously reliable sources. But this does not necessarily mean that inconsis-
tencies between calendars of the same name are the result of error. It is quite
possible that thehemerologiaare referring, under the same name, to slightly
different calendars that were used in different parts of the same province or in
different periods. Indeed, the Florence MS tells us explicitly that the calendar
of Asia was not the same throughout the province, as it distinguishes between
the calendars of Ephesus and‘Asia’, the former identical with a calendar called
‘of Asia’in the Leiden MS, and the latter with a New Year on 24 September
and identifiable, on the basis of its non-Macedonian month-names, as the
calendar of the city of Smyrna (Samuel 1972: 175 n. 1). The inconsistent New
Year dates of the calendar of‘Cyprus’in the Vatican and Florence MSS could
also reflect different practices in either different cities of the island or in
different historical periods.^131 Indeed, some of the month-names of the
(^129) This applies in particular to the following variations (in bold in Table 5.6), where day 1 is
not duplicated: the Vatican MS Asia 23/7, Cyprus 23/7 and 23/8, Bithynia and Crete 23/3; The
Leiden MS Bithynia 23/3 and Cyprus 23/8. 130
In the present text, Laodikon is a 31-day month that begins with a double day 1 and
terminates on day 30. If we consider day 30 on 23/9 as a scribal error, and substitute the New
Year (1 Kaisarion) on this date, then Laodikon becomes a 30-day month beginning with a double
day 1 and ending on day 29 (on 22/9), which is unparalleled and makes little sense.We would
have to assume that the whole month of Laodikon is erroneously numbered, and amend it in
such a way that it has only one day 1 and that it ends on day 30; it is unlikely, however, that the
entire day-count of Laodikon is erroneous.
(^131) The dates of the calendar of Paphos (western Cyprus) of 16 Apogonikos = 8 November
and 14 Ioulios = 6 January, in Epiphanius (Panarion51. 24,Williams 1987–94: ii. 55), imply a
New Year on 23 September and a month beginning on 24 December, as in the Florence MS; but
this does not necessarily mean that the‘Cypriot’dates of 22 September and 23 December in the
Vatican MS are errors.
282 Calendars in Antiquity