time of December.^49 This contrasts with the Seleucid calendar, in which the
Macedonian month of Dios (equivalent to the Babylonian Tishrei, month VII)
occurred around the time of October, thus two months earlier.^50 The only
explanation is that Ascalon had made, at some stage beforehand, two excessive
intercalations.^51 But in the calendar of Gaza, its southern neighbour, Dios
occurred around the time of November,^52 and was thus discrepant from the
Seleucid calendar by only one month; which suggests that only one excessive
intercalation had previously been made. Further north along the coast, the
calendar of Caesarea seems to have had Dios in November,^53 thus again
implying one excessive intercalation. Continuing northwards, the calendar
of Tyre had Dios in November or December, thus one or two excessive
intercalations.^54 Still further north, the calendar of Antioch had Dios in
November, thus one excessive intercalation.^55 The greatest discrepancies
appear to have been in the calendars of Heliopolis and Sidon. The calendar
of Heliopolis (Baalbek) placed Thisirin(Tishrei) at the time of January,
considerably later than the Babylonian Tishrei and suggesting that three
excessive intercalations had previously been made.^56 In Sidon, the month of
Dios occurred in January, suggesting again three excessive intercalations.^57
(^49) This can be inferred from the fact that when the calendar of Ascalon was adapted to the
Julian calendar, at about the end of the 1st c.BCE, 1 Dios wasfixed at 27 November (Samuel 1972:
177). This is presumably a reflection of the time of the year when the lunar month of Dios
generally occurred. 50
See Table 5.2. As has been shown in Ch. 2, even if it is not proved that the Babylonian
calendar still followed in this period the Saros Canon 19-year cycle, it is fairly certain that
intercalations were sufficiently regular (i.e. with an even distribution of seven intercalations over
any period of 19 years) for the relation between the Babylonian calendar and the seasons (or the
solar year) to have remained stable. Indeed, astronomical diaries stretching until the mid-1st c.
BCEconfirm that throughout the Achaemenid, Seleucid, and early Parthian periods, month I of
the Babylonian calendar consistently remained thefirst lunar month after the vernal equinox
(i.e. mostly in April). This means that month VII would have corresponded roughly to October. 51
Ascalon became independent from the Seleucids in 104BCE(Schürer 1973–87: ii. 90–1,
106 – 7); the excessive intercalations would therefore have been made at some point during the 1st
c.BCE.
(^52) The inference is the same as for Ascalon (see above, n. 49), except that in Gaza, 1 Dios was
fixed at 28 October when it was Julianized. 53
Since in Caesarea, 1 Dios wasfixed at 3 November (so according to the reconstruction
of Rey-Coquais 1978, which, however, remains uncertain; see below, n. 142). 54
In Tyre, 1 Dios wasfixed at 18 November (Samuel 1972: 176). See further below, after
n. 141. 55
In Antioch, Dios of the Julianized calendar was assimilated to the month of November (this
calendar eventually became dominant in late Roman Syria; ibid. 174). The excessive intercalation
which this implies would have been made after 47BCE, when the calendar of Antioch appears to
have still conformed to the Seleucid equation of Macedonian and Babylonian calendars, with
Dios occurring around October (see above, n. 27).
(^56) In Heliopolis, where Babylonian month-names were exceptionally used, 1 Tishrei wasfixed
at 23 December (ibid. 176); see below and Table 5.7.
(^57) In Sidon, Dios of the Julianized calendar was assimilated to the month of January: ibid. 174.
Sidon was independent from the Seleucid Empire since 111BCE(Schürer 1973–87: ii. 90–1), so
that the three excessive intercalations would have been made during the following one hundred
Fragmentation: Babylonian and Julian Calendars 251