« A.U.C. 537 = 217 B.C. »
We have three synchronisms that enable us to bound the dates of this year:
Livy 22.1 reports that news of a number of portents were received in Rome about the time that the new consuls took up their office, Id. Mart. A.U.C. 537. One of these was of a solar eclipse, which must be the eclipse of 11 February 217, which reached a magnitude of about 60% in Sicily. Therefore, Id. Mart. A.U.C. 537 cannot have been before mid February 217.
Ovid, Fasti 6.766, gives the date of the battle of Lake Trasimene as either a.d. X or a.d. VIII Kal Quin.; for the purposes of determining the number of intercalations we can safely assume the former. Polybius 5.101.3 reports that the battle occurred while Philip V of Macedon was besieging Thebes.
It is possible to date this siege independently. Polybius 5.95.5 notes that Aetolian operations that year began shortly before the start of the harvest, i.e. in late May, and in Polybius 5.97.1 that Philip occupied Byalazora in Paeonia and turned his attention to Greece at about the same time; this synchronism is credible since Philip must be allowed some time for his Paeonian campaign. After assembling his levies he spent a week marching to Melitea, which he attempted to take by surprise but failed because his scaling-ladders were too short for the walls. He then regrouped and marched to Thebes, which be besieged for about two weeks. Thus, the siege of Thebes began about three weeks after the start of the harvest. This dates the battle of Lake Trasimene to some time in mid-late June 217.
Polybius 5.101.6 also gives a terminus ante quem for the battle of Lake Trasimene by reporting that Philip V received news of the battle just after arriving at Argos for the Nemean Games.
All we need to establish to use this synchronism is the Julian date for the Nemean Games. These were one of the four great pan-Hellenic games, and were normally held every other year, in what for us is the odd-numbered year. The primary datum on their date is a scholium on Pindar's Odes, schol. Pind. Hyp. Nem. d, e, which states in two different MS traditions that the games were held on 12 or 18 Panemos. The problem, then, is to determine how this date relates to the Julian calendar. This is, of course, a matter of considerable controversy.
P. Brind'Amour, Le calendrier romain 161 argued as follows: Inscriptional evidence quoted by A. E. Samuel, Greek and Roman Chronology 90, showed that Argive Panemos was the second month before Argive Karneios (the intervening month being Argyeos); Thucydides 5.54 that Argive Karneios corresponded to Spartan Karneios; and Plutarch, Nicias 28.1 that Spartan Karneios corresponded to Athenian Metageitnon. From this chain of correspondences, it follows that Argive Panemos corresponded with Athenian Skirophorion, the last month of the Athenian year, which ended with the first new moon following the summer solstice. For this year, Argive Panemos therefore began about 9 June, meaning that Philip received the news of the battle on about 20 or 26 June. Therefore, a.d. X Kal. Quin. A.U.C. 537 was some time shortly before 12 June 217 (allowing a couple of weeks for the news to reach Philip); by extension, Id. Mart. A.U.C. 537, 97 days earlier, was on or somewhat before 7 March 217.
There are three problems with this line of argument which are immediately obvious from reading the cited texts:
According to Polybius, the following events occurred between the battle of Lake Trasimene and the Nemean Games: Philip V completed the Theban siege; organised the sale of its inhabitants into slavery and founded a Macedonian colony there; dealt with a set of embassies; unsuccessfully pursued an Illyrian fleet; dragged part of his own fleet across the Isthmus of Corinth; and only then headed off to Argos. Thus we are certainly looking at an interval of several weeks between a.d. X Kal Quin. A.U.C. 537 and the Nemean Games. If the Nemean Games were actually held in Athenian Skirophorion, the battle ought to have been fought in mid-late May, but, given the harvest synchronism of Polybius 5.95.5, which Brind'Amour did not consider, it seems more likely that the Games were held some time in mid July -- Athenian Hecatombaion -- not the latter part of June -- and that the battle was fought in mid-late June.
It is clear that the first mention of Karneios in Thucydides 5.54 is using the Spartan calendar. Thucydides also states that the month of the festival of the Karneia was sacred to all Dorians, and Sparta and Argos were both Dorian states. This does suggest that the month concerned was at least roughly synchronised for all Dorians. But it is not at all clear that Thucydides understood that it also had the same name in both the Argive and Spartan calendars. If the names were in fact different, it would be confusing for the reader if Thucydides then switched from the Spartan name of the month to the Argive as he switched his frame of reference from the Spartans to the Argive. His account would be equally sensible if he were solely referring to the Spartan Karneios in both instances. That is, his reason for reusing the name has nothing to do with the Argive calendar, so the passage proves nothing about the Argive name for Spartan Karneios.
Plutarch, Nicias 28.1 equates Syracusan Karneios to Athenian Metageitnon, not Spartan Karneios. Since Syracuse was a Corinthian foundation (Strabo 6.2), and Corinth was an Achaian city, there is no obvious reason from this passage to equate the Syracusan calendar with the Spartan.
Brind'Amour's argument is a variant of that of A. Boethius, Der argivischer Kalendar 1-48, who first established that Argive Karneois was the second month following Argive Panemos, but simply equated Syracusan Karneios with Argive Karneios without giving any justification for it.
I have found three more recent studies on this question, which are the basis of the following discussion: P. Charneux, BCH 81 (1957) 181 at 197ff; P. Perlman, Athenaeum 67 (1989) 57; and S. D. Lambert, ZPE 139 (2002) 72. Perlman's article contains a detailed survey of the evidence and previous work published to 1989.
The first point to be established is the calendar to which the "Panemos" of schol. Pind. Hyp. Nem. d, e belongs, since this month name occurs in a large number of Greek calendars.
P. Perlman, Athenaeum 67 (1989) 57 at 62f. collects all the Argive decrees honouring foreign benefactors. Three of these (SEG XIII.240, SEG XVII.144 and SEG XIII.242) are dated 24 Panemos, one (SEG XXX.360) 30 Panemos; three (SEG XXXIII.279 (twice), SEG XXXIII.280) 29 Argyeos, adopting motions proposed in Panemos, and three more can be restored this way; one (SEG XIII.243) is dated 6 Agyeos, probably on a motion proposed in Panemos; one (SEG XXX.357) on 4 Karneios; one (SEG XVII.143) 29 Hermaios on a motion proposed in Karneios; one (SEG XI.1084) 9 Arneios held over from Teleos; one (SEG XXX.359) 26 Arneios on a motion proposed earlier that month; and four more whose months cannot be recovered. Thus well over half of these decrees were passed or moved in late Panemos. Since the Nemean Games were a major pan-Hellenic festival, this strongly suggests that most of the decrees were passed to honour benefactors who had come to Argos for the games. We may therefore reasonably conclude that the games were in fact held in Argive Panemos.
These decrees, and other Argive data, establish that the months Panemos, Agyeos, Karneios and Hermaios ran in sequence, and that Teleos was followed by Arneios. Additional Argive month names are known from other decrees: Gamos, Amyklaios and Apellaios. However, the complete set of Argive months, their sequence, and their alignment to months in better known calendars such as the Athenian all remain unknown.
The second point to be confirmed, to the extent possible, is the date in Panemos.
Schol. Pind. Ol. 7, 147c, states that the Rhodian Halieia was held on 24 Gorpiaios, six days from the Nemean Games. Since Greek months in many calendars were approximately lunar, and therefore at least roughly synchronised, this is usually held to support the date of 18 Panemos given in schol. Pind. Hyp. Nem. d, e, i.e. that the date of the Rhodian Halieia, 24 Gorpiaios, was six days after the Nemean Games.
P. Perlman, Athenaeum 67 (1989) 57 at 58f. argues that the actual date of the games in Panemos corresponded to 30 Gorpiaiosaios. Her arguments are as follows:
Greek months were not tightly synchronised between different calendars, see W. K. Pritchett, BCH 81 (1957) 269. For example, a truce between the Spartans and the Athenians is dated to 14 Elaphebolion (Athenian) in Thucydides 4.118.12 and to 12 Kerastios (Spartan) in Thucydides 4.119.1; and Plutarch, Aristides 19.7 equates 4 Boedromion (Athenian) to 27 Panemos (Boeotian). Perlman concludes that it is highly improbable that days in an Argive month would have the same number in the month of another calendar, and that the date of "18 Panemos" was actually an ancient supposition derived from Schol. Pind. Ol. 7, 147c.
This would be a very fair point if the scholiast were giving a specific synchronism, i.e. giving the relative dates for the Halieia and the Nemean Games in a specific year. Indeed the calendric situation is even worse than Perlman describes, since Pritchett showed in the same article that days could be repeated in festival months on a completely arbitrary basis as was convenient for local circumstances. As examples he cited an Athenian inscription dated "4th intercalation of 9 Elapethelion", and a Euboean decree which authorises each of the four Euboean cities to intercalate up to three days as needed locally to delay the celebration of the Dionysia in their city, to ensure that a troupe of travelling actors hired to perform for it would actually be able to appear. (Greek contractors must have loved this: "No way is the Parthenon not going to open on 21 Thargelion guv, we just have to repeat 20 Thargelion 15 times.")
But what the scholiast is giving is a general statement. Since it was well-known that in practice the celebration dates of local festivals could vary considerably, the statement can only make sense if it is understand as giving an ideal or nominal relationship. As Pritchett showed, the existence of such an ideal calendar is demonstrable in early 2nd century Athens, where dates could be given as [date] "of the archons" or [date] "of the god" -- i.e. according to the moon.
If one accepts that the two dates reflect an ideal relationship, then the relationship can only be considered coincidental if it can also be shown that at least one of the two calendars involved was not, at least in the ideal, a lunar calendar. But Perlman herself supposed that 24 Gorpiaios was a date in the Seleucid calendar, which was lunar, and as far as I can tell we have no reason to suppose that the Argive calendar was not notionally lunar. Indeed Perlman herself concludes that the games fell on 1 Panemos -- 6 days after 24 Gorpiaios -- i.e. that both Gorpiaios and Panemos were lunar months and hence aligned! Recorded Argive decrees were passed on 29 Agyeos, 29 Hermaios, and 30 Panemos; these make a lot of sense as being the last days of lunar months.
The term apecw used in schol. Pind. Ol. 7, 147c is rarely used to denote temporal distance, but in the two passages Perlman could find it specifically meant "before". Thus she argued that the scholium should be interpreted to mean that the date of the Rhodian Halieia, 24 Gorpiaios, was six days before the Nemean Games, which therefore occurred on 30 Gorpiaios/1 Hyperberetaios. The scholiast of schol. Pind. Hyp. Nem. d, e supposedly derived the date 18 Panemos by making the same "mistake" as modern scholars.
A grand total of two known examples does not make a strong case for general usage. Perlman's argument requires that the scholiast found ancient Greek as ambiguous as she does and was as ignorant of the Argive date of the games as we are. Since both assumptions are highly improbable, it seems to me far more likely that the dates given in the two scholia represent independent records, just as they appear to be. As she herself notes (P. Perlman, Athenaeum 67 (1989) 57 at 63) the distribution of honorary decrees, with a strong bias towards late Panemos, suggests that the Argive alaia was busy with the games in the earlier part of the month, just as one would expect if the date was 18 Panemos.
Pindar, Nemean Odes 4.35, suggests that the Nemean Games were held at the New Moon. If Gorpiaios was a month in a lunar calendar, 1 Hyperberetaios would be a new moon date. Perlman suggests that this new moon marks the start of the games, which were therefore held on 1 Panemos.
S. D. Lambert, ZPE 139 (2002) 72 at 72 n. 4 notes that Pindar, Nemean Odes 4.35 is not normally understood to refer to the Games but to the festival at which the ode itself was recited, and certainly such an interpretation must first be excluded before the verse can be interpreted as referring to the Games. Further P. Perlman, Athenaeum 67 (1989) 57 at 59 n. 8, herself notes that the interpretation of this passage is controversial, and may not even indicate that the occasion of recital was a new moon.
Perlman's suggestion that this new moon implies a start date of 1 Panemos for the games rather contradicts her earlier argument that the Greek months were not synchronised.
In short, I find Perlman's arguments on this point completely unconvincing. To the extent that the contemporary data is applicable, it tends to support a date of 18 Panemos, as being moderately close to the Argive decrees, and that the Games were not held on the new moon. Thus in this instance I am inclined to accept the idealised equation 18 Panemos (Argive) = 18 Gorpiaios, in whatever calendar Gorpiaios belongs to.
With this background, we can now examine the three analyses on the Julian date of the games.
P. Charneux, BCH 81 (1957) 181, at 198, noted that Plutarch, Moralia 245E, in giving a story reported by Socrates of Argos, equates Argive Hermaios with "the month now known as the fourth month". Since Charneux also established that Hermaios was the third month after Panemos, we can infer that Argive Panemos would have been "the first month". But, naturally, Plutarch did not bother to tell us what calendar he was counting in.
Charneux argued that the calendar in question was that of the Achaian League, which admitted Argos in 229. At that time, according to Polybius 5.1.1, the term of the Achaian strategos began at the rising of the Pleiades, i.e. c. 13 May. Interpreting this event as marking the start of the Achaian year, Argive Panemos would therefore correspond to Achaian month 1, which would normally = the lunar month for May/June, i.e. roughly equivalent to Athenian Skirophorion.
P. Perlman, Athenaeum 67 (1989) 57 at 69, noted that there was a second Achaian calendar, adopted later, which began the year with the autumnal equinox; this was in fact the standard basis for estimating the date of Plutarch's story before Charneux redated it. The first month on this calendar would equate Argive Panemos with about October/November. Moreover, she notes that Plutarch does not explicitly say that his equation comes from Socrates of Argos; it may well represent whatever calendar Plutarch himself used. Perlman suggests that he was here using the Phocian calendar, which was widely used at Delphi. Delphic inscriptions give equations between the Phocian, Delphic and Athenian calendars as follows:
SGDI 1715: Tritos (Phocis; i.e. the third Phocian month) = Poetropios (Delphi)
- SGDI 2089: Poetropios (Delphi) = Poseidion (Athens), the sixth Athenian month
Therefore, on the Phocian calendar, the fourth month corresponded to Athenian Gamelion and the first to Athenian Pyanepsion. Since ex hypothesi the first Phocian month corresponded to Argive Panemos, we can equate Argive Panemos to Athenian Pyanepsion, which allows us to place it at roughly September/October in the Julian calendar.
Perlman's conjecture that Plutarch was probably using his own normal calendar is a reasonable one, and the fact that he says that Hermaios is "now" (i.e. second century A.D.) known as the fourth month is an argument against assuming the Achaian calendar of the third century B.C. But it is far from obvious that Plutarch's own calendar was the Phocian one, and Perlman gives absolutely no justification for making this choice. Plutarch was a native Boetian, educated at Athens, resident in Rome for many years, and retired as a priest to Delphi when he wrote most of his works. I can't claim to have studied Plutarch's calendrics in any detail, but where he gives synchronisms that I have noticed, as in Plutarch, Nicias 28.1, they are to the Athenian calendar, which was surely the Greek calendar that was best-known to the intelligentsia in Plutarch's time. On this calendar, Argive Panemos should correspond to Athenian Hecatombaion -- the first month of the Athenian year, starting with the first new moon after the summer solstice.
That being said, it seems to me that the most natural reading of the text is that Plutarch is only saying that the fourth Argive month was no longer called Hermaios in his time. If so, then this statement establishes that Panemos was the first month of the Argive year but there is no calendrical synchronism that can be used to establish when that was.
P. Perlman, Athenaeum 67 (1989) 57 at 69 ff. proposed to identify the date of Argive Panemos through the synchronism with the Rhodian Halieia given by schol. Pind. Ol. 7, 147c, assuming that this festival occurred six days before the Nemean Games.
Since the Rhodian calendar does not have a month Gorpiaios, this synchronism takes a little work to resolve.
It should come as no surprise to any reader who has managed to follow the thread this far that there is no ancient source which gives the date of the Halieia in the Rhodian calendar. The major Rhodian festival in honour of Helios was the Dipanamia, celebrated in Panemos II, which was intercalated between Pedageitnos and Diosthios, the sixth and seventh Rhodian months, down to the second century, and after Thesmophorios, the last month, thereafter. But, studies of the Rhodian priest-lists have shown that the Dipanamia was not the Halieia. However, there are several dates recorded for other sacrifices to Helios. Perlman cites Rhodian inscriptions which reveal that Helios received sacrifices on different dates in different Rhodian cities: 1 and 20 Dalios and before 20 Panemos at Kamiros and on 14 Hyacinthios at Lindos. But these appear to be local sacrifices, and none of them are identified as the Halieia. So, until someone publishes a Rhodian inscription that actually dates the Halieia, this line of inquiry is at a dead end.
We therefore have to fall back on identifying the date through the synchronism of the Halieia to 24 Gorpiaios. The trouble is that, while there is a month of this name in at least 20 known Greek calendars, these do not include the Rhodian one. Perlman proposes to narrow the choice by trying to identify the scholiast, and by supposing that he intended the scholium to be understood by the widest possible audience.
We have two recensions of the Pindaric scholia: the Vatican and the Ambrosian. Schol. Pind. Ol. 7, 147c is only given in the Vatican recension, which does not identify the author for this or any other scholium. But the scholium in the Ambrosian recension for Pind. Ol. 7, 147b cites Istros as the authority for the Halieia. We may conclude that Istros was also the authority for Schol. Pind. Ol. 7, 147c.
Istros was a pupil of the famous Alexandrian poet Callimachus who lived in the latter part of the third century. Perlman therefore argues that the choice is between the Seleucid Macedonian calendar and the Ptolemaic Macedonian calendar. She argues against the Ptolemaic calendar on the grounds that it was not well regulated against the seasonal year and therefore would not have given the reader a good sense of when the Halieia occurred.
Supposing that Gorpiaios was indeed a Seleucid month, Perlman notes that at this time Seleucid Gorpiaios corresponded to Babylonian Abu and Athenian Hekatombaion (recte Metageitnion), i.e. the second month following the summer solstice, roughly August/September. In fact we could date the nominal date of the Rhodian Halieia in A.U.C. 537 = 217 to within a couple of days: 24 Gorpiaios S.E. 95 = 1 September 217. Supposing that the Games were held on 1 Panemos, and that Panemos was equated to Metageitnion (recte Boedromion), the nominal date should be 1 Panemos (Argive) = 7 September 217.
I think Perlman's conjectural identification of Istros as the author of the scholium is reasonable, but for from certain, and her inference that he would have used the Seleucid calendar seems very weak. Its true that we don't understand very well how the Ptolemaic Macedonian calendar worked after Ptolemy II. But that doesn't mean that Istros would not have used it -- nor that his contemporaries did not understand how it was regulated. Moreover Istros was not a native Alexandrian. According to the Suda, IstroV, he was from Cyrene or Paphos; the latter is considered more likely by modern scholars.
A stronger objection to identifying Gorpiaios as the Alexandrian month is that the available data for the reign of Ptolemy III synchronises Gorpiaios to the period November-February. It does seem a reasonably safe bet that the Nemean Games did not occur in mid-winter.
While the argued synchronism with Metageitnion (recte Boedromion) is certainly possible, it remains inconclusive.
S. D. Lambert, ZPE 139 (2002) 72 published a re-examination of fragment F2 of the sacrificial calendar of Nikomachos from Athens. This included in lines 10-11 an offering of a ram to a deity which Lambert reconstructs as being [D]ii Ne[mewi], i.e. the Nemean Zeus. This was previously reconstructed as [D]ii Ne[aniai], which can be understood as either "Zeus as a young man" or "Zeus as the hero Neanias". Lambert argues against this restoration on the grounds (a) that Zeus is not otherwise connected with adolescence rituals and (b) while there is a well-known Attic hero Neanias, who is the object of a sacrifice in Pyanopsion, there was no obvious reason to associate him with Zeus. He cites Demosthenes 21.115 as evidence of Athenian participation in the cult of Nemean Zeus, and notes that it was "normal for member states of various amphictyonies ... to have filials of the amphictyonic cult at home."
Line 2 can be reconstructed as the 11th or 12th of a month; with less certainty line 9 can be reconstructed as the 12th. Lambert argues that the column gives the sacrifices of an odd-numbered year, and that in order to maintain a uniform density of rituals down the column the month must be assumed to be the second month, which would make it Metageitnion.
Assuming that the Athenian sacrifices to the Nemean Zeus occurred at the same time as the games, this gives the Games an Athenian date for the games of 1[2? Metageitnion], i.e. the second month following the summer solstice, roughly August/September. This would nominally date the Nemean games in A.U.C. 537 = 217 to around 18 August 217.
I hardly have the expertise to evaluate Lambert's reconstruction of the overall structure of the sacrificial calendar, but the restoration of 1[2? Metageitnion] sounds reasonable, and the argument for associating the sacrifice to Nemean Zeus seems plausible. However, it is not so obvious to me that this necessarily implies that the Nemean Games took place at the same time. It could well be, for example, that the sacrifice was performed after the return of the Athenian delegation to the Games, which would imply that they were held in Hekatombaion, not Metageitnion.
Perlman finally provides a very useful survey of the literary references to the Games, the aim of which is to show that these are not inconsistent with her proposed date of c. September. Most of these are consistent with several possible dates. The following seem to me to have probative value:
Demosthenes 21.115 notes that Meidias, who has accused Demosthenes of murder, had allowed him to conduct initiatory rites for the Athenian council and to lead the Sacred Embassy to the Nemean Games. These are presumably in chronological order, and the first event occurred at the beginning of the Athenian year, i.e. in Hecatombaion. Therefore the Nemean Games were probably after the beginning of Hecatombaion, not in Skirophorion.
Polybius 2.65 recounts that Antigonus Doson, king of Macedon, and his Achaian allies advanced into Laconia early in summer, where, after a few days, he defeated the Spartan king Cleomenes at the battle of Selassia, and immediately occupied Sparta. According to Polybius 2.70.1, he left Sparta after a few days and went back to Macedonia via Argos, where according to Polybius 2.70.4 he celebrated the Nemean Games. This entire sequence of events can't have taken much longer than 2-3 weeks, hence the Games were celebrated early-mid summer, not late in the summer as Perlman requires.
Of course, there is a catch. Cleomenes fled to Egypt where according to Plutarch, Cleomenes 32, he was received by Ptolemy III shortly before his death. It is now generally agreed that the Egyptian data shows that Ptolemy III died at the end of 222. Hence Selassia was fought in the summer of that year. But the Nemean Games were normally held only in alternate years -- the odd years by our reckoning -- so there should not have been a Nemean Games in 222; largely for this reason it was long argued that Selassia must have been fought in 221. Perlman argues that these Games were therefore abnormal, and their date therefore cannot be used to indicate the normal date of the Games.
Polybius 5.101 -- the very passage under discussion here -- suggests, as noted above, that the Nemean Games were in July, by reason of the Argolid harvest. Perlman acknowledges this as the most difficult citation to reconcile with her dating. However, Polybius' account of subsequent events in Greece -- continued pressure on the Aetolians and the conclusion of the Peace of Naupactos -- is followed by statements (Polybius 5.106 and Polybius 5.107.5) that the Achaeans and Aetolians elected the strategoi for the following year, events that normally took place sometime between September and November. She argues that a c. July date for the Games leaves too much time for peace to be concluded.
However, it seems to me that everything fits perfectly if the Peace of Naupaktos was concluded in mid August, especially if the Achaian elections were held closer to September than November. Moreover, Perlman overlooks that Philip V himself still had to return to Macedon and fight a campaign against the Illyrian king Scerdilaisas, as recounted in Polybius 5.108. A date in October for the Peace of Naupaktos gives him at best an absolutely minimal amount of time to fight this campaign before the onset of winter.
All things considered, it seems to me that there is a much better case for equating Argive Panemos with Athenian Hekatombaion than with Skirophorion, Metageitnion, Bodroemion or Pyanepsion, although Metageitnion is not excluded, and indeed may occasionally have corresponded due to phase differences between the Argive and Athenian calendars in the insertion of embolimos months. Assuming Panemos (Arg.) = Hekatombaion (Ath.), we can date the Nemean Games in this particular year, 18 Panemos (Argive), approximately to 25 July 217.
We are finally in a position to determine the number of intercalations between A.U.C. 537 = 217 and A.U.C. 564 = 190, the most recent year whose dates are certain. The estimated dates for Lake Trasimene for different intercalation values are given in the following table:
Number of Intercalations Number of intercalated days Julian date of a.d. X Kal Quin. A.U.C. 537
5 110-115 2-7 August 217
6 132-138 10-16 July 217
7 154-161 17-24 June 217
8 176-184 25 May - 2 June 217
9 198-207 2-11 May 217
The value that places the battle of Lake Trasimene in mid-June, as required by the Polybian evidence, is seven intercalations. This is the same conclusion reached by P. Brind'Amour, Le calendrier romain 161, but on a fallacious dating of the Nemean Games, which gave far too little time to allow the news to reach Argos by 26 June, let alone for Philip V to complete the series of actions ascribed to him by Polybius. Six or less can be ruled out for the same reason on the dating determined here for the Nemean Games. Eight could be possible, but if the Polybian synchronism with the siege of Thebes is correct it would place the siege too close to the Argolid harvest. Nine intercalations (per P. S. Derow, Historia 30 (1976) 265 at 275f.) places the battle well before that harvest. It also places Id. Mart. A.U.C. 537 in the range 16-25 January 217 which is too far before the Sicilian eclipse.
Id. Mart. A.U.C. 537 therefore lies in the range 3-10 March 217 and the battle of Lake Trasimene was fought between 17 and 24 June 217. If the conjecture offered here for the Eclipse of Ennius is correct, the actual dates were 3 March and 17 June 217 are assumed here.
Derow's reconstruction was based on the a priori assumption that intercalation in this period exactly followed the principle of intercalations of alternating length in alternating years described in Macrobius, Saturnalia 1.13.11-14. He then argued that the accounts of Hannibal's campaigns given by Livy and Polybius supported this scheme. However, the indications they give for the time taken for key steps in the sequence of events are too imprecise to accept this analysis as definitive. Indeed, since Derow's scheme has Hannibal starting to prepare his winter quarters in July/August -- high summer -- one could reasonably argue that in fact the same data is sufficient to show that Derow's scheme is too early.
Nevertheless, there is one contraindication that, taken in isolation, does appear to favour Derow's scheme. According to Livy, Q. Fabius Maximus was granted dictatorial powers shortly after Lake Trasimene (Livy 22.8), and gave up command to the consuls of that year "while there was still some time left in autumn" (Livy 22.32). Since the term of a Republican dictatorship was fixed at 6 Roman months, this corresponds, at the very earliest, to a date in late December 217 on the scheme proposed here. By no stretch of the imagination can such a date be regarded as a date in autumn. However, Polybius 3.106.1 says that Fabius gave up his command after the consuls of the next year had been elected, and that the incoming consuls appointed the consuls of the previous year, as proconsuls, to command the armies against Hannibal. The simplest way to explain this apparent contradiction is to suppose that Livy, who used Polybius as a primary source, overlooked this nuance when composing his account, and interpreted the fact that the proconsuls took up active command immediately to mean that the campaigning season of the previous year was not yet over.
The same datum suggests that this year was not intercalary. The term of Fabius' dictatorship must have expired in Ianuarius or Februarius. The fact that the new consuls had already been elected and were ready to take up office suggests that the interval between the end of his dictatorship and the start of the next consular term was short, which would not be the case if there was an intercalary month at this point.
Website © Chris Bennett, 2001-2011 -- All rights reserved