correct, at least temporarily, the effect of the drift of the 365-day calendar, and
to make the New Year celebrations that were held during the epagomenal days
agree with the vernal equinox, which traditionally was deemed the season of
the New Year. At the beginning of the sixth centuryCE, the vernal equinox had
shifted to the beginning of the ninth month (Āäar); it is in this period (late
Sasanian), therefore, that the epagomenal days would have been repositioned
at the end of the eighth month.^42 The Sogdian, Choresmian, and Armenian
calendars thus preserved the scheme that had originally been instituted in
Persia, whereas the Persians altered it slightly at some time in late Antiquity.^43
The most likely context for the dissemination of the Persian Zoroastrian
calendar to these distant regions would have been the Achaemenid Empire. It
is only under Achaemenid rule that Sogdiana, Choresmia, and Armenia were
part of a single geopolitical entity, in which they are likely to have shared the
same calendar with Persia; whereas they became independent, and thus
politically disconnected from each other, during thefirst century of Seleucid
rule (third centuryBCE).^44 At the beginning of the Seleucid period, moreover,
when Sogdiana was still part of the Seleucid Empire, it came under intensive
Hellenistic influence through colonization and imperial governmental con-
trol,^45 which makes it quite unlikely that the Persian Zoroastrian calendar was
adopted there in this period. In the unlikely event that it was after the
Achaemenid period that the Persian Zoroastrian calendar spread to any
these regions, this would not have happened through political means, but
only through the religious medium of Zoroastrianism. But this depends on
how widespread or significant Zoroastrianism really was in these regions in
this period;^46 it also depends on the extent to which the calendar that I call
(^42) De Blois (1996) 47, inferring from the Syriac Martyr Acts of Gregory that this change could
not have been made after 518CE(for possible evidence that it was made before 482/3, see Sims-
Williams and de Blois 2005: 191). He also explains on this basis the origin of further anomalies in
the Zoroastrian ritual calendar: in particular, the celebration of a‘Great New Year’on the 6th of
thefirst month, which will have originated from a refusal by some to omit thefive epagomenal
days before the New Year in the year when thesefive days were officially shifted to the end of the
following eighth month. For an alternative theory as to the origins of the‘Great New Year’, see
Boyce (2005) 8: it may have arisen as a result of popular confusion regarding thefive epagomenal
days, or deliberate opposition to insert them in the calendar year, at the time when the Persian
Zoroastrian calendar was originally instituted.
(^43) Further minor changes were made to the Zoroastrian calendar during the Middle Ages, but
this is beyond my scope; see de Blois (1996).
(^44) So argued by Boyce (1975–91) ii. 244, and esp. de Blois (1996).
(^45) Briant (1990) 48–52, Sherwin-White and Kuhrt (1993) 103–13; also (on Bactria, which
formed a single satrapy with Sogdiana in this period) Boyce (1975–91) ii. 24, iii. 51–68, 152–93.
(^46) Evidence of Zoroastrianism in Sogdiana, Choresmia, and Armenia, as well as in Bactria and
Cappadocia (to be discussed below) in the Achaemenid and Seleucid periods is at best sporadic
and mingled with elements from other cults or religions; although this is a reflection of the
eclectic nature of Zoroastrianism in this early period, as well as of the general scarcity of
historical evidence for these regions in Antiquity. For Sogdiana, see Sims-Williams (2000)
8 – 12. Boyce (1975–91) provides some evidence for Choresmia (iii. 192–3), but very little for
180 Calendars in Antiquity