An absolute astronomical synchronism is given by Livy 44.37.8 which describes a lunar eclipse on a.d. III Non. Sept. A.U.C. 586 = 21 June 168, on the eve of the battle of Pydna. (The fourth century historian Eutropius 4.7.1 gives a.d. III Non. Sept. as the date of the battle itself, but Livy's account is clearly to be preferred. Livy actually dates the event to "the night before the day of prid. Non. Sept.", and relates the battle as happening on the day after the eclipse.)
However, the synchronism has been doubted. One problem is that Livy 44.37.8 and Plutarch, Aemilius Paullus 17.7, the other account of the eclipse, describe the eclipse as occurring well into the night, when modern calculations show it to start occurring well before sunset, with the moon rising at Pydna already eclipsed. This suggests that the battle may be misdated, e.g. that it should be identified with the eclipse of 11/12 June 167 or that of 2 September 172. These alternatives may both be dismissed. The year is certain because it is dated by the consular fasti. The date is bounded by an Athenian inscription (ISE 35; B. D. Merritt, Hesp. 3 (1934) 18 no 18, Hesp. 5 (1936) 429), which records that a certain Kalliphanes reported the Roman victory over the Macedonians to an extraordinary session of the Athenian assembly convened by the Athenian generals in Piraeus on the last day of Skirophorion in the archonship of Eunikos. That the battle is Pydna is certain because the inscription also mentions Attalus, brother of king Eumenes II of Asia, who according to Livy was present at the battle as a legate to the Romans. The assembly was therefore convened on the last day of the Athenian year ending in A.U.C. 586 = 169/8. Since 30 Skirophorion is the last day of the Athenian year, it can be fixed reasonably closely astronomically by the date of the first new moon after the summer solstice, and corresponds to c. 7 July 168. The battle must therefore have been fought shortly before this date.
From the synchronism known for A.U.C. 564 = 190, there were 12 intercalations between 190 and 168, 3 of 22 days and 9 of 23, if the Roman civil date of Pydna is correct. Fewer intercalations would place the battle well before the eclipse of 21 June. More intercalations would place the battle in or after 14 July, or 6 July at the very earliest, if all intercalations were supposed to be 22 days long except the one known to be 23 days long. This leaves insufficient time for the news to reach Athens, let alone for an assembly to be convened. Hence there must have been 12 intercalations between 190 and 168 in order for the Roman civil date to be consistent with the Athenian data.
There is one other twist to this. Livy 44.36.1 indicates that the battle was fought after the summer solstice, which at this time was 26 June (Julian). On this basis, V. M. Warrior, AJAH 6 (1981) 1 at 39, judging this datum to have come from Polybius and therefore more plausible than the eclipse datum, which must derive from an annalistic source, supposes that the battle was actually fought several days after the eclipse, even though she accepts the eclipse synchronism. That is, she requires that Livy's date is for the eclipse, not the battle. This is not implausible. Livy's account of the eclipse is clearly tendentious: it is inaccurate in astronomical detail, and it emphasises Roman ability to predict the eclipse and the inability of the Macedonians to do so (the reverse is more likely to have been the case). It is perfectly possible that the source of this account eliminated the interval between the battle and the eclipse to further improve its dramatic effect. But Livy's date for the battle, prid. Non Sex. A.U.C. 586, cannot be equated to any Julian date between 26 June and 6 July 168, given the eclipse synchronism for A.U.C. 564 = 190. Hence the ultimate source for it must have tied it to the eclipse, regardless of whether the battle was actually fought on the following day.
The other argument against the synchronism is that of P. Marchetti, BCH 100 (1976) 401. Marchetti accepts the Julian date of the battle (i.e. that it was on the day following the lunar eclipse) but rejects the synchronism with the Roman civil date a.d. III Non. Sept. He argues that Polybius gives indications that Id. Mart. in A.U.C. 576 = 178, A.U.C. 595 = 159 and A.U.C. 597 = 157 and was at the end of the Polybian summer, i.e. around late October, therefore it is not possible that Id. Mart. A.U.C. 586 = 168 B.C. was at the start of January 168, which the synchronism of Livy 44.37.8 would require. He notes that Livy 45.1 describes the news as reaching Rome on a.d. XIV Kal. Oct., on the second day of the Ludi Romani, held annually in honour of Jupiter Optimus Maximus. He argues that Livy frequently describes the games shortly after giving election results, and therefore that at this time they were actually held towards the end of the year. The dates in September are said to be a projection of the custom prevailing in Livy's own time. Hence he supposes that Livy's source only noted that the news of Pydna arrived in Rome 13 days after the victory, on the second day of the games, and that it was Livy himself who pinned these events to September.
This analysis is open to several objections:
All the alleged seasonal synchronisms in Polybius are vague and questionable. Indeed, by the same weak standard of argument, Walbank supposes that Polybius shows that A.U.C. 599 = 155 began at the end of the Polybian winter.
- There is no reason to believe that the Ludi Romani were moved from the end of the year to September. P. Brind'Amour, Le calendrier romain 352 justly notes: «C'est demander à Jupiter de changer son adresse, pour conserver un renseignement de Tite-Live dont on s'efforce justement de montrer qu'il est erroné en ce qui a trait à la date du 3 septembre!»
- If Livy was guilty of projecting the imperial calendar onto republican events, there ought to be other traceable examples.
All testable indications are, therefore, that the synchronism is valid. We must suppose that the mismatch between the description of the eclipse and modern calculation is due to embroidering for literary effect. Other signs of this are clear, such as Plutarch's story (Aemilius Paullus 17.9) that Aemilius Paullus know the eclipse in advance, and his comment (Aemilius Paullus 16.9) that it was the end of summer -- appropriate if a.d. III Non. Sept. is interpreted as a Julian date, but not for a date in June.
Given that the synchronism is valid, this allows us to fix the Julian dates of this year with certainty. In particular, Kal. Mart. A.U.C. 586 = 21 December 169. We have no way to determine directly whether the year was intercalary. However, on the reconstructed Lex Acilia proposed here, since the following year, A.U.C. 587 = 167, was 377 days long, this year must have been a regular year.
Website © Chris Bennett, 2001-2007 -- All rights reserved