sequence is repeated. Moreover, the 19-year cycle that seems to have been used
from the 470s may still have been perceived, at the time, as‘loose’: this would
explain why an intercalation of VI 2 could still be made in 465/5. It seems more
likely, therefore, that it was in the reign of Artaxerxes I (which started in 464/3,
after the intercalation of VI 2 in 465/5) that we should consider a rigid 19-year
cycle, with intercalations of only XII 2 , to have become formally established.
Indeed, in view of the connection between regnal periods and intercalation
practices (which we have noted so far, stretching back to the reign of Nabo-
polassar), it seems likely to attribute the innovation of a rigid 19-year cycle to
Artaxerxes.^115 The rigid 19-year cycle would have represented a further step in
the evolution of Babylonian intercalation.
433/2 XII 2 31111
430/29 XII 2 31414
427/6 XII 2 317
425/4 XII 2 21919
aSee Table 2.2 n.a.
bThis intercalation can be interpreted either in relation to preceding years as year 3 of a cycle, or in relation to
the following years as year 14 of a cycle.c
dFor this and the next entry, see Table 2.2 nn.b,c.
So according to a number of unpublished British Museum sources listed byWalker, proving wrong the
conjecture of Parker and Dubberstein (1956: 6, 35) that the intercalation was XII 2 in 493/2, six months earlier.
According to Sachs and Hunger (1988–2006: v. no. 54, BM36823), the intercalation was XII 2 in 493/2; this,
however, is only a textual reconstruction (presumably, following Parker and Dubberstein) that is anyway
erroneous. The intercalation of VI 2 in 492/1 is also attested in the Old Persian calendar (on which see Ch. 4):
G. G. Cameron (1965) no. 19.
eAn intercalation of VI 2 in 484/3 is attested in sources from Persepolis (G. G. Cameron 1948, nos. 10-11), but
only in the Old Persian calendar. It is generally assumed that in this period, the Old Persian calendar was fully
assimilated to the Babylonian calendar, and hence that Old Persian dates can serve as evidence for the
Babylonian calendar (e.g.Walker 1997: 23–4); but this assumption needs further discussion (see Ch. 4).
fEvidence for this and the next entry is debatable but generally accepted: see Parker and Dubberstein (1956) 8,
Aaboeet al.(1991) 5, Britton (1993) 67.
gi.e. 8 in the sequence of foregoing years, and 19 in the sequence of forthcoming years. If we include the
Old Persian evidence for 484/3 (above, n.e), the preferable sequence (for the Saros Canon cycle) should be
14 – 17 – 19.
hAn intercalation of XII 2 in 474/3 is attested in Cameron (1948) no. 27, which would be compatible with year
8 in the Saros Canon and the loose 19-year cycles; but again, this is only in the Old Persian calendar (see
above, n.e).
Entries in italics designate intercalations that are only attested, according toWalker’s list (with my supple-
ments), in astronomical sources that I have identified as theoretical and non-observational.
(^115) However, the absence of evidence for the years 457/6 and 438/7 (see Table 2. 4; all the
other years are attested) remains cause for some concern, because it means we have no proof of
XII 2 in year 6 of the cycle for the entire reign of Artaxerxes I. The use of a single intercalary
month may reflect an ancient Babylonian tradition. A passage in MUL.APIN (II ii 18–20) states
that the intercalary month I 2 (i.e. a second Nisannu, no longer used in thefirst millenniumBCE)
belongs to reign ofŠulgi, XII 2 belongs to the reign of the Amurru (Amorites), and VI 2 to the
reign of the Kassites, which suggests that different kings or dynasties favoured specific intercalary
The Babylonian Calendar 109