« A.U.C. 668 = 86 B.C. »

CIL I2 713 records the cleaning of an aqueduct at Cales in Campania: "L. Cornelio / Cinna cos.iter(um) / purgatum / mense INTR" The reading is apparently not entirely clear: CIL regards the final character as a combination of N, T and R, but E. H. Warmington, Remains of Old Latin IV.178, probably following Mommsen, sees it just as a combination of N and R. The question is what month is represented. If the character is NR, then "Mense INR" could represents "Mensis IaNuaRius" or "Mensis INteRcalaris"; if it is NTR only "Mensis INTeRcalaris" is possible. Scholarship is divided: Mommsen regarded the date as IaNuaRius, as does Warmington and Degrassi, but the editors of CIL have opted for "Mensis INTeRcalaris".

The two options can be distinguished by noting that only one consul is named. Cinna's colleague, C. Marius, died on a.d. XIV Kal. Feb. A.U.C. 668 (Plutarch, Marius 46.5), on the 17th day of his 7th consulship. The tessera nummularia CIL I2 891, from Capua, is dated Non. Feb. in the consulate of L. Cornelius (Cinna) and L. Valerius (Flaccus). This shows that Marius' successor had been appointed by early Februarius, in which case he would certainly have been in office in any intercalary month. It is unlikely that his appointment was known in Capua on Non. Feb. but unknown in nearby Cales on Kal. Int. This clearly support a reading of IaNuaRius for CIL I2 713, since only then was there a vacancy in the office. Therefore, pending other evidence, CIL I2 713 may be disregarded as evidence for the length of this year.


A synchronism is also available for this year. Plutarch, Sulla 14.6, tells us that Sulla sacked Athens on Kal. Mart. of this year, and cites Sulla's memoirs for the date. Plutarch notes that this date was "very near" to 1 Anthesterion in the Athenian calendar. J. W. Müller, ZPE 103 (1994) 128, has argued that the Athenian lunar calendar at this time was regulated according to a Metonic cycle, and so infers that 1 Anthesterion (in the lunar calendar -- kata qeon) corresponds to c. 4/5 February 86. He then proceeds to explain away the fact that this is not close to 1 March 86 -- but completely neglects to note that the Roman date is not Julian.

This synchronism is not as exact as Müller makes it appear; Plutarch specifically disavows an exact match. Müller's data does not conclusively show a Metonic cycle, though it does show a lunisolar alignment of the Athenian calendar to the solstice, which suggests that we need only worry about misalignment in years that begin or end with a new moon close to the summer solstice. This is not such a year.

However, the historical accuracy of the mention of Anthesterion has been challenged. Plutarch introduces the synchronism in order to retail an omen -- a cloudburst from a clear sky at the moment Curio captured the tyrant Aristion -- that is lent force because Anthesterion was the month in which flood rituals were performed. In an interesting article, A. T. Grafton & N. M. Swerdlow, JWCI 51 (1988) 14, note that many of Plutarch's exact dates are or appear to be omen-based retrocalculations linking them to appropriate festivals; this pattern is also observable for many apparent exact dates in early Roman historiography. In JWCI 51 (1988) 14 at 18, Grafton and Swerdlow identify this particular synchronism as an example. By implication, Plutarch (or his source) was working backwards: he had the story of the capture of Aristion and inferred that it must have occurred on a day of suitable omen in Anthesterion.

Grafton and Swerdlow further argue that Plutarch's dating of the event to "very near" to 1 Anthesterion arises from his habit of treating Roman months as though they were lunar, so the synchronism comes from Sulla's date of Kal. Mart. For comparison, they cite Plutarch, Camillus 19, in which he states that the Battle of the Allia, dated to a.d. XV Kal. Sex. (= 18 July (Roman)) A.U.C. 364 = 390 BC (Livy 6.1.11) took place "near the full moon after the summer solstice", a statement that would be true if Quintilis was equated to a lunar Hekatombaion, and Plutarch, Publicola 14.3, in which he states that the Ides of September nearly coincides with the full moon of Metageitnion. Grafton and Swerdlow argue that these lunar synchronisms are Plutarch's own deductions. In support of this proposal, they note Plutarch, Roman Questions 24 (269B-D), which discusses the origins of the Kalends, the Nones and the Ides as marking lunar phases, and explicitly justifies using the synchronisms as an approximation even though he knows it is not a strictly accurate method of proceeding.

Since pre-Julian Roman months were not lunar, and since it is vastly improbable in any case that such data would have been recorded or transmitted to Plutarch from such an early epoch, nor that there would have been any means for him to have accurately correlated the Roman date to either the moon or the Athenian calendar, this argument is almost certainly correct for the Allia and for the events in the life of Publicola. It is much less obviously true in the present instance. Sulla's sack of Athens took place in an age much closer to Plutarch's own time, the Athenians were (unfortunately for them) at the heart of the action, and Athenian historians surely existed who gave dates for the events recorded. While the synchronism was certainly recorded for its ominous associations, I don't see any reason to doubt that Kal. Mart. actually fell in early Anthesterion.

Finally, accepting Plutarch's synchronism as accurate, there is also the question of whether the Athenian date was in Anthesterion according to the festival calendar (kat 'arconta) or the lunar calendar (kata qeon). In the case of a synchronism to Anthesterion Kat 'arconta , the Julian equivalent could well be a few days before or after c. 4/5 February 86.

While an exact synchronism is not forthcoming, the synchronism should be good enough for us to estimate the number intercalations between this date and Kal. Mart. A.U.C. 696 = 24 February 58, as follows:

Number of Intercalations   Number of intercalated days      Julian date of Kal. Mart. A.U.C. 668
     A.U.C. 668-696                                                           

              12                            264-276                           7-19 March 86
              13                            286-299                           12-25 February 86
              14                            308-322                           20 January - 3 February 86    

Thus this synchronism implies either 13 intercalations, most or all of which are 23 days long, or 14, most or all of which are 22 days long, between Kal. Mart. A.U.C. 668 and Kal. Mart. A.U.C. 696. If the Athenian date is actually kat 'arconta, this could move the Julian equivalent later in February, favouring 13 intercalations rather than 14. 12 intercalations only arises as a possibliity if the events described took place so late in an Anthesterion kat 'arconta that it was actually Elaphebolion kata qeon

Either way, the density of intercalations (13 or 14 in 28 years) gives an average frequency of about one intercalation every other year. The analysis of A.U.C. 687 = 67 further supports this, since it shows that there were only three intercalations between A.U.C. 687 = 67 and A.U.C. 697 = 57, giving 10 or 11 intercalations between A.U.C. 668 = 86 and A.U.C. 687 = 67.

This justifies the key assumption made in the analysis of A.U.C. 677 = 77, and allows us to accept the two possible equations Kal. Ian. A.U.C. 677 = 17 December 78 and Kal. Ian. A.U.C. 677 = 10 January 77.

If Kal. Ian. A.U.C. 677 = 17 December 78, the number of intercalations to A.U.C. 697 = 57 is reduced by nine, all of 23 days, giving possible ranges of dates as follows:

Number of Intercalations   Number of intercalated days      Julian date of Kal. Mart. A.U.C. 668
     A.U.C. 668-676                                                           

              4                            88-92                                12-16 February 86
              5                           110-115                              20-25 January 86

Given the clear predominance of 23 intercalations in this period, only 4 additional intercalations are possible, for a total of 13.

If Kal. Ian. A.U.C. 677 = 10 January 77, the number of intercalations to A.U.C. 697 = 57 is reduced by eight, seven of 23 days and one of 22, giving possible ranges of dates as follows:

Number of Intercalations   Number of intercalated days      Julian date of Kal. Mart. A.U.C. 668
     A.U.C. 668-676                                                           

              5                           110-115                              13-18 February 86
              6                           132-138                              21-27 January 86

Given the clear predominance of 23 intercalations in this period, only 5 additional intercalations are possible, for a total of 13.

Hence there were 13 intercalations between this date and Kal. Mart. A.U.C. 696 = 24 February 58. In either possible scenario, the best match would be obtained if all were of 23 days, but the synchronism is not tight enough to require this solution.

For an argument that the intercalations were all 23 days long under the Lex Acilia see discussion under A.U.C. 563 = 191. If accepted. this fixes the Julian dates of A.U.C. 668 = 86. The only way to pack the intercalations is to place them in alternate odd-numbered years, which is consistent with the facts that A.U.C. 671 = 83 and A.U.C. 687 = 67 were also intercalary.

This result fixes the Julian conversion of every Roman date from Kal. Mart. A.U.C. 667 = 22 February 87 to Kal. Mart. A.U.C. 696 = 24 February 58.

26 Dec 2005: Note that Muller's Athenian alignments are not perfectly Metonic (thanks to Sacha Stern)
4 Sep 2006: Review Grafton/Swerdlow discussion on the orgins of the Kal. Mart ~ 1 Anthesterion synchronism.

Website © Chris Bennett, 2001-2011 -- All rights reserved