to have begun the year in the autumn. Thus a document from Dura-Europos
dated 87CEin the month of Panemos refers to a transaction that had been
made the same year in the month of Dios; this is only compatible with an
autumn New Year.^9 An autumn New Year also explains the inconsistent
correlation between Seleucid era years and Arsacid era years in Parthian
documents in Greek from Babylonia (111/10BCE), Susa (21CE), and Dura-
Europos (87CE, 121 CE): consistency can be achieved if one assumes that the
Arsacid era began in the spring (following Babylonian usage), and the Seleucid
era in the autumn (Assar 2003: 177). A Macedonian autumn New Year is
further confirmed by Parthian numismatic evidence. Parthian coins from the
latefirst centuryBCE–earlyfirst centuryCE, which are dated according to the
Seleucid calendar and refer occasionally to intercalary months, appear to
assume a year beginning in the autumn;^10 officina letters (mintmarks) on a
series of Parthian coins dating from 77/8–78/9CEsuggest a year beginning in
the month of Dios.^11 By late Antiquity, the autumn New Year and Seleucid era
from 312BCEwere standard in the Syrian and Jewish calendars.^12
But although the preponderance of an autumn New Year in post-Seleucid
Asia Minor and the Near East suggests, as I am arguing, a direct survival from
the Seleucid period, it should be noted that autumn New Years existed in the
ancient Near East long before the arrival of the Seleucids. Even in Babylonian
hemerologies from the second millenniumBCE, it appears that a‘second New
Year’was celebrated at the beginning of month VII.^13 A neo-Assyrian omen
text states, albeit obscurely, that‘months XII and VI are the beginning of the
year, because months I and VII are at the beginning of the year’.^14 The main
New Year celebrations at Ugarit appear to have been in the autumn, although
there are indications also of a secondary New Year in the spring (de Moor
1971: 56–62, 77–80). A similar multiplicity of New Years is attested much later
term“Arab”’), but this explanation is speculative. The New Year at Miletus, from the 3rd c.BCE
onwards, is thought to have been in the spring; but inasmuch as non-Macedonian month-names
were used in this city, the‘Seleucid’nature of its calendar is generally questionable (see Samuel
1972: 114–18, and below, n. 35).
(^9) Hannah (2005) 91; the reference should beWelles, Fink, and Gilliam (1959) no. 18, ll. 13,
- By this period, the New Year may have already shifted from Dios to Hyperberetaios; see
discussion below.
(^10) However, this is only true if one assumes that the intercalations were based on thefixed 19-year
cycle of the earlier Babylonian calendar, which is far from certain: see discussion in Ch. 2.
(^11) Sellwood (1983) 282, 292; Assar (2003) 183. However, the geographical and chronologi-
cal disparity between these various sources allows for other possible explanations.
(^12) Even in Babylonia, the Jews assumed that the Seleucid era began in Tishrei (i.e. the
autumn): Babylonian Talmud,Avodah Zarah10a.
(^13) Labat (1939) 27; Cohen (1993) 6–7;Wagenaar (2005) 119.
(^14) Hunger (1992) no. 165. Thefirst clause may be a citation from an older Mesopotamian
source, interpreted in the second clause. For an alternative interpretation, see Parpola (1970–83)
ii. 186–7 (with further evidence of a New Year in month VII, Tashritu) and Cohen (1993) 400. It
is also possible that‘beginning’in thefirst clause is a scribal error.
Fragmentation: Babylonian and Julian Calendars 237