Calendars in Antiquity. Empires, States, and Societies

The other double dates, from the Seleucid period, are inferable from three
third-centuryBCEastronomical reports cited in Ptolemy’sAlmagest(mid-
second centuryCE, but probably using earlier, intermediary sources). The
actual reports are dated by a year number‘according to the Chaldaeans’(i.e.
the Seleucid era), followed by a Macedonian month-name and day number.
Ptolemy then provides the equivalent year number in the era of Nabonassar
(his standard chronological scheme) and an Egyptian date, which we can
easily convert into Julian calendar dates (see Table 5.1).^20
The Macedonian month-names (Apellaios, Dios, and Xandikos) can be
securely established as really referring to Babylonian months, for the following
reasons:

1. Ptolemy (or his source) calls these dates‘Chaldaean’(although this may
   arguably refer only to the Seleucid year, not to the months and days).


2. The distinctive language and form of the astronomical reports cited here
   by Ptolemy suggest that they are direct translations of Babylonian
   sources such as the Astronomical Diaries (A. Jones 2006: 257). In their
   original language, the dates of these reports would obviously have been
   recorded according to the local Babylonian calendar.


3. These dates imply months beginning atfirst visibility of the new moon,
   as in the Babylonian calendar.^21

Table 5.1.Double dates in Ptolemy’sAlmagest

Source in
Almagest

‘Chaldaean’year, and

Macedonian date

Nabonassar year, and

Egyptian date

Julian equivalent

(not in text)

9. 
   
   
   
   7. 10 67, Apellaios 5 504, Thoth 27a 245 BCENov. 18/19
   
   
   
   


10. 
    
    
    
    7. 9 75, Dios 14 512, Thoth 9 237 BCE, Oct. 29/30
    
    
    
    


11. 7 82, Xandikos 5 519, Tybi 14 229 BCE, Mar. 1/2

aSince the Egyptian day starts in the morning,‘ 27 ’means (in the context of astronomical observations) the
night running from the 27th in the evening through to the next morning, at the turn of the 28th (by contrast
the Julian day starts at midnight, and therefore the night period straddles two dates, here 18/19). In the
Babylonian calendar, the day starts in the evening. This report actually relates to an astronomical observation
in the morning, i.e. at the turn of the 28th (Jones 2006: 258, 264–5).

controversial and not very convincing. See now Bennett in http://www.tyndalehouse.com/
Egypt/ptolemies/ptolemy_i_fr.htm (accessed 30 June 2010), n. 6.

(^20) Ptolemy,Almagest9. 7. 10 (Toomer 1984: 452), 9. 7. 9 (ibid.), and 11. 7 (ibid. 541), cited in
Hannah (2005) 92. See Samuel (1962) 47–8, but most importantly A. Jones (2006).
(^21) So at least thefirst two cases: thus the second (of 237BCE) implies that Dios began on the
evening of 16 October, when the new moon was indeedfirst visible (paceParker and Dubberstein
1956, followed by A. Jones 2006: 264, who take 17 October asfirst evening of new moon
visibility; in fact the date of this report appears to be confirmed by Sachs and Hunger
(1988–2006) v. 105–6 no. 39, even if—as pointed out by A. Jones (2006: 264 n. 15)—we cannot
assume its textual reliability). The third case, of 229BCE, is more problematic as it implies a
month beginning one day later; see discussion below, n. 26.
Fragmentation: Babylonian and Julian Calendars 239