Egyptians’.^152 Also significant may be the likelihood (and indeed, common
assumption) that the astronomer Sosigenes, whom Julius Caesar called upon
to design his calendar,^153 came from Alexandria.
The structural differences between the Julian and Egyptian calendars reflect
an attempt, on the one hand, to accommodate traditional elements from the
Roman Republican calendar, and on the other hand, to improve the existing
structure of the Egyptian calendar. Thus the Julian calendar retained from the
Republican calendar its month-names^154 as well as an idiosyncratic (albeit
different) sequence of month-lengths, that has survived in our calendar until
this very day. It may seem strange that Julius Caesar did not take the
opportunity to rationalize the calendar and institute, for example, a regular
alternation of 30- and 31-day months.^155 He may have been constrained by
religious considerations, as certain changes in the month-lengths may have
affected the dates of festivals.^156 Still, the changes that were made to Republi-
can month-lengths so as to produce the Julian months were not without
(^152) Appian,CivilWar2. 154. Appian may be slightly overstating the case, out of native
Egyptian patriotism. See also Lucan,CivilWars, 10. 172–218, where Caesar while in Egypt in 48
BCEis depicted as conversing with an Egyptian Hellenistic scholar about his knowledge of
astronomy.
(^153) So at least according to Pliny,NaturalHistory18. 211–12, who adds that Sosigenes
composed three astronomical treatises on this subject, but does not indicate Sosigenes’place of
origin (see also ibid. 2. 39). However, thefigure of Sosigenes is not attested outside Pliny.
According to Macrobius,Saturnalia1. 14. 2, the Julian calendar was designed with the help of
a local individual, the scribe M. Flavius. On the possibility that thisscribawas a minor pontiff
(see above, n. 137), see Rüpke (1995) 372–3. It seems to me, however, that the similarity of this
name with that of the scribe Cn. Flavius (mentioned ibid. 1. 15. 9; see above, near n. 147)—albeit
not an uncommon name—should be regarded as suspicious.
(^154) After Caesar’s murder in 44BCE, the month Quintilis was renamed Julius, as described in
Dio 44. 5. 2. SeeWeinstock (1971) 152–8, Rüpke (1995) 394–5, 405–6, the latter with discussion
also of the later change of Sextilis to Augustus, on which see references below, n. 161. 155
Blackburn and Holford-Strevens (1999) 98 and Bennett (2003) 225 and n. 33 rightly
denounce the commonly accepted, but completely groundless theory, that the Julian calendar
was originally instituted as a regular alternation of 30- and 31-day months (except for Februarius
in a non-leap year). This theory was originally put forward by Sacrobosco (13th c.), who gave the
original sequence of months as follows: Januarius (31), Februarius (29/30), Martius (31), Aprilis
(30), Maius (31), Junius (30), Julius (31), Sextilis (30), September (31), October (30), November
(31), December (30). According to Sacrobosco, the length of Sextilis was increased to 31 days
when it was renamed Augustusc.8BCE. As a result, Februarius had to be reduced by one day; and
then, to eliminate a sequence of three 31-day months from Julius to September, the order of 30-
and 31-days months from September to December was reversed. This theoryfinds no support in
any of the sources: quite on the contrary, the accounts of Julius Caesar’s calendar reform in
Censorinus (20. 9) and Macrobius (Saturnalia1. 14. 7) imply the same month-lengths as were
later in use in the Julian calendar. Furthermore, as Bennett points out, sources contemporary or
close to the institution of the Julian calendar provide clear evidence against this theory: P.Oxy.
LXI 4175, a papyrus dating from 24BCE, assumes already 31 days for the month of Sextilis (A.
Jones 2000b: 165); whilst Varro (DeReRustica1. 28), writing no later than 37BCEand referring
to the newly instituted calendar (dies civiles nostrosquinunc sunt) provides dates and lengths of
the seasons that are only compatible with the Julian calendar as later known. 156
See Macrobius,Saturnalia1. 14, Rüpke (1995) 376–8), Feeney (2007) 152–6.
212 Calendars in Antiquity