result of setting the beginning of the month of Ab on Augustus’birthday on
23 September. In fact, the 31st days do not correspond exactly to equinoxes
and solstices: there is no evidence that 22 September, for example, would have
been regarded as the exact day of the autumn equinox.^146 The parallelism, on
balance, is thus very weak.
Tubach and others, moreover, have surprisingly overlooked the variant
reading of thehemerologia. The Vatican MS, indeed, has a slightly different
sequence with the month of Eiar beginning on 23 June instead of the 24th
(see Table 5.7).^147 This small difference is significant, because it completely
alters the structure of the calendar of Heliopolis: according to the Vatican
MS, the year consists simply of an alternation of 30- and 31- day months,
with the only exception, in October and again in January, of two consecutive
30-day months (unavoidable in a 365-day year). This calendar structure,
which differs completely from the calendar of Enoch, is equally as rational
and as historically plausible as that of the Florence manuscript.^148 This
significant divergence between the manuscripts casts further doubt on
Tubach’s theory.
The early Aramaic of the month-names—if indeed it is early Aramaic—
does not prove that the calendar of Heliopolis, as represented in thehemer-
ologia, predates the Roman period: for as we have seen, most calendars of the
Roman Near East preserved their old month-names even after converting to a
Julian-type calendar. Given the broader context of ancient Near Eastern
calendars, it seems most probable that the calendar of Heliopolis was originally
lunar, and that during the Achaemenid and Seleucid periods it corresponded,
as its month-names clearly indicate, to the standard Babylonian calendar.^149
Conversion to a 365-day solar year is most likely to have occurred much later,
in the early Roman period, at the time when the cities of the Roman Near East
were switching to a variety of Julian-type schemes.
(^146) Note also that the commonly held assumption that the four 31st days in the calendar of
Enoch, or their equivalent in the calendars of Jubilees and Qumran, correspond to equinoxes and
solstices is not explicitly stated in Enoch, Jubilees, or Qumran sources: see Ch. 4 n. 96.
(^147) The Leiden MS lacks the month of June; it is therefore impossible to use it as evidence in
this context. In the Vatican MS day 8 of Eiar is duplicated on 30 June and 1 July. This is probably
the result of dittography (as elsewhere in the manuscript: see Table 5.6 n.f), which should not
affect the dates of the beginning of the months.
(^148) The argument that scribal errors are more common in the Vatican than in the Florence
MS (see above) is not sufficient, in my view, to favour the latter, because there is nothing in the
text of the Vatican MS to indicate, in this case, any scribal error.
(^149) The city’s Greek name, which means‘sun-city’, and the apparent centrality of the sun-god
in the city’s public cults during the Roman period (on which see Millar 1993: 285, Ball 2000:
39 – 47), need not have any bearing on the nature of the calendar that was used in the Hellenistic
period (paceTubach 1994: 188–9).
288 Calendars in Antiquity