spread of the Julian calendar was almost conterminous, indeed, with that of
the Roman Empire, and thus may be seen as a clear marker of‘Romanization’
in the Near East.
In the case of the province of Arabia, the transformation of the calendar was
particularly rapid. Before the creation of Roman province of Arabia in 106CE,
Nabataean documents and inscriptions were dated with Babylonian month-
names,^157 suggesting that the Nabataean kingdom had retained the lunar
Babylonian calendar of the Seleucid period. But Judaean Desert documents
from soon after 106CEand dated‘according to the province of Arabia’assume
a completely new system: Julian and Macedonian month-names (in the Greek
texts) and Babylonian month-names (still used in the Nabataean and Aramaic
texts) are consistently synchronized with each other in such a way as to
confirm, beyond any doubt, that the calendar of Arabia (with either Macedo-
nian or Babylonian month-names) had been Julianized and was identical,
already then, with what is later attested in epigraphic sources and the
hemerologia.^158
This new calendar was structurally the same as the Alexandrian calendar,
with twelve 30-day months andfive epagomenal days, and a sixth epagomenal
day in Julian leap years;^159 but unlike the calendars of Ascalon and Gaza,
its months were not conterminous with those of the Alexandrian calendar.
In Arabia thefirst month, Xandikos, began on 22 March, whereas the
Alexandrian month nearest to this date, Pharmouthi, began on 27 March.^160
We must reject again the argument that since this calendar did not con-
form, in structure, to the Julian model, its institution would have preceded the
annexation of Arabia into the Roman Empire. There is no reason why the
Nabataean kingdom should have forsaken, at any stage, the lunar Babylonian-
type calendar it had inherited from the Seleucid period. Considering the
radical changes that occurred to the province of Arabia soon after 106CEin
terms not only of official language but also of political and legal structures (see
(^157) See the Nabataean inscriptions published by Yadin (1962) 239 =P.Yad. 1 – 3 in Yadin
(1989), Healey (1993) and (2009), all from the 1st c.CE(except Healey 2009: 113–15 no. 14, dated
34 BCE). Attested month names include Nisan, Ab, Elul, Kislew, Tebet. Further references in Stern
(2001) 38 n. 160.
(^158) See Stern (2001) 39–42. Later epigraphic sources include, in major part, the Greek
tombstone inscriptions of Zoar (4th–6th cc.), in Meimaris and Kritikakou-Nikolaropoulou
(2005), (2008). 159
Evidence that the intercalary day in leap years was counted as an extra epagomenal comes
from an inscription from Zoar (eid. 2005: 243–5 no. 148) dated Tuesday, 2nd epagomenal in the
year corresponding to 436CE(a leap year); the weekday indicates that this was 17 March (and not
18 March as in non-leap years).
(^160) Samuel (1972) 177; a discrepancy of 5–10 days between the two calendars maintains itself
throughout the year. On Xandikos asfirst month of the calendar of Arabia, see above, n. 8.
Fragmentation: Babylonian and Julian Calendars 291