Dio Cassius 51.19.6 states that the Senate had decreed that the date of the fall of Alexandria should be regarded as the start of the reckoning of time, i.e. of Augustus' regnal year in Egypt. From the Fasti Antiates, we know that Alexandria fell on Kal. Sex. A.U.C. 724 = 30. Thus, if Dio is correct, Augustus' regnal years in Egypt were initially reckoned to begin on Kal. Sex.
The earliest document that shows a regular Egyptian regnal year for Augustus is SB 16.12469, the lease of a red cow in year 5 = 26/5, which runs from Hathyr to 30 Mesore without a change in year number. From this it is clear that any attempt to implement the Senate decree failed no later than year 5.
In several studies, T. C. Skeat cited contemporary evidence in support of this thesis that Augustus' year began on the anniversary of the fall of Alexandria before that year:
T. C. Skeat, JRS 43 (1953) 98 noted that Canon of Ptolemy gives 22 full years to Cleopatra VII, hence by Ptolemy's time her year 22 = 31/0 was formally accounted as part of her reign and not part of Augustus'. Thus Augustus certainly did not follow Egyptian custom by annexing her last year as his first.
Positive support for Dio's statement comes from pOxy 12.1453, a contract by four temple lamplighters to provide oil, the term of which was read by Grenfell and Hunt as "from 1 Thoth to Mesore [Epagomone] 5(?) of the present first year of Caesar" as they had in the preceding year 22 = 7 (i.e. of Cleopatra). In T. C. Skeat, JRS 43 (1953) 98, he proposed that this confirmed that Augustus began his reign in Egypt on 1 Thoth of the year following Cleopatra VII. However, in T. C. Skeat, ZPE 53 (1983) 241, he decided that the end date of the contract should actually read Mesore 7(?), and that the Epagomene interpolated by Grenfell and Hunt was incorrect. 6(?) Mesore was also noted as a possible reading of the end date by Grenfell and Hunt, and was regarded as paleographically equiprobable to 5(?) Mesore by G. Geraci, Genesi della provincia romana d'Egitto 160. Skeat argued that the contract runs from 1 Thoth to the end of year 1 of Augustus, which implies the equation prid. Kal. Sex. A.U.C. 725 = [5 or 6 or 7] Mesore year 1 of Augustus.
In T. C. Skeat, CdE 69 (1994) 308, he cited two other items. First, the dedicatory inscription at the temple of Dendera states that the goddess "took possession" on "the 19th day of Caesar's reign". This highly unusual form of dating suggests that the normal form of dating would be ambiguous. On Skeat's theory, the 19th day of his reign was 26 Mesore = 21 August 30. He argued that to give this as 26 Mesore year 1 made it ambiguous -- especially after year 5 = 26/5 -- as to whether the dates was in 30 or 29.
In the same article, and of most relevance to Roman chronology, Skeat noted pRylands 4.601, an annual lease ending on 7 Mesore year 4 = 1 August 26. He noted that the day number appears to have been a later insertion. He argues that this indicates that there was doubt as to when the year was actually due to end.
Skeat argued that this data should be interpreted to mean that the Augustan regnal year ran from 8 Mesore to 7 Mesore in his years 1 to 4. This is certainly possible. However, he did not consider the relationship of such a regnal year to the Roman calendar. The original senatusconsultum was surely dated according the Roman calendar, not the Egyptian one. Thus, Roman leap days will have caused uncertainty about when the regnal year ended on the Egyptian calendar.
If Skeat's analyses are correct, we may therefore infer that:
pOxy 12.1453 implies the equation prid. Kal. Sex. A.U.C. 725 = 5/6/7 Mesore year 1 = 30 or 31 July or 1 August 29. On the Scaligerian model, prid. Kal. Sex. A.U.C. 725 = 1 August 29 = 7 Mesore year 1. On the model proposed here, prid. Kal. Sex. A.U.C. 725 = 31 July 29 = 6 Mesore year 1. Thus this papyrus is actually consistent with either scheme.
pRylands 4.601 implies the equation prid. Kal. Sex. A.U.C. 728 = 7 Mesore year 4 = 1 August 26. On the standard model, prid. Kal. Sex. A.U.C. 728 = 2 August 26.
On the standard model of the triennial cycle, a leap year occurred in A.U.C. 727 = 27 = year 3; on the system proposed here, Roman leap years occurred in A.U.C. 725 = 29 = year 1 and A.U.C. 728 = 26 = year 4. The first occurred too early in the Roman regime to have caused calendrical difficulties in Egypt, while the second appears to be reflected in the "uncertainty" that Skeat detected on the date in pRylands 4.601.
In a recent article, E. Grzybek, in Y. Perrin (ed.), Neronia VII, 145 at 148-150 has challenged Skeat's arguments on pOxy 12.1453 and pRylands 4.601:
On pOxy 12.1453, Grzybek accepts that the syntax of the contract shows that it was negotiated before it began, on 1 Thoth, and hence that the reference to "the current first year" of Caesar implies that year 1 began before 1 Thoth. However, he believes that that year ended on 30 Mesore, i.e. that the long year envisaged by Skeat was year 1, not year 5. He says "[p]ar la suite, Octavien doit avoir compté ses années commes les Lagides ... du 1er Thoth au denier jour épagomène. En tout cas, les sources de son époque n'autorisent pas à admettre d'autres dispositions ou modifications que celles qui concernaient sa première année."
I find this argument circular and ex cathedra. I think it also conflicts with the evidence. Since the contract only began on 1 Thoth year 1, the preexisting contract for year 22=7 must have run to term on 5 Epagomene, even though, on Grzybek's own recognizance, that year had finished early, in early Mesore. Hence the service contract was clearly an annual contract, not an anniversarial one. The fact that a date in Mesore was chosen to end it in year 1 surely indicates, as Skeat argued, that that date was understood, at the time the contract was agreed, to mark the end of the first year of Caesar.
It is not impossible that, after this contract was agreed, year 1 was realigned to end on 5 Epagomene. But Grzybek adduces no evidence to show that it did, he simply says that it must have done. On the contrary, by far the simpler inference is Skeat's: that year 1 ended on [5-7] Mesore and that Augustus' years were realigned in year 2, 3, 4 or 5.
On pRylands 4.601, he argues, in effect, that Skeat is overreading the evidence, and that the delay in inserting the date merely means that the documents were prepared a little in advance of an uncertain date of registration, as in many modern contracts.
The difficulty here is that the least itself is dated in Mesore, with no indication that it does not take immediate effect, and the day number is omitted in both the lease and the grapheion copy. While in isolation Grzybek's objection is not without merit -- there could well be another, local, explanation of this -- the facts that the realignment almost certainly took place after year 1 and that the Alexandrian reform was almost certainly introduced in year 5, certainly makes Skeat's explanation viable, if not ironclad.
In short, I think Grzybek's argument on pOxy 12.1453 fails and that Skeat has the more likely case on pRylands 4.601. As far as Roman calendrics is concerned, the key point is that this data is not central to the case for a different triennial cycle in the early Julian calendar than has generally been assumed. Rather, Skeat's theory is consistent with the revised cycle, and the revised cycle fits in better with Skeat's analysis of the early Augustan calendar in Egypt than the traditional one does.
Website © Chris Bennett, 2001-2007 -- All rights reserved