In the course of the Anabasis Arrian gives the Attic month for six events
1. II 11 x, Battle of Issos, Maimakterion
2. II 24 v, Fall of Tyre, Hekatombaion
3. III 7 i, Reaches Thapsacus, Hekatombaion
4. III 15 vii, Battle of Gaugamela, Pyanepsion
5. III 22 ii, Death of Dareios, Hekatombaion
6. V 19 iii, Battle of the Hydaspes, Munychion
This would require no further comment were it not for the fact that Plutarch twice dates the Battle of Gaugamela to Boedromion (Life of Alexander 31, vi and Life of Camillus 19, v) not only that two fragments of a Babylonian Astronomical Diary date the battle to 24 Ululu and since they also record the lunar eclipse that preceded the battle modern astronomers can work out that the battle occurred on the 1st October in the Julian calendar.
Since the Babylonian calendar was at this time, regulated by lunar observation and working on a strict Metonic cycle of seven intercalary months in each nineteen year period (intercalations in years 1, 3, 6, 9, 11, 14 and 17; the first an additional month VI Ululu the others additional month XIIs Addaru), we can retrocalculate that in 331BC Ululu ran 8 September to 8 October (see Parker-Dubberstein 1956 and Chris Bennet 2003). More specifically, from the evening to the evening of these dates; the Babylonians reckoning a day from sundown to sundown. This also explains why Plutarch places the battle on the 26th and the tablets on 24th
ULULU
E 13 14 15 16 17 18 19 20 21 22 23 24 B
N D N D N D N D N D N D N D N D N D N D N D N D
14 15 16 17 18 19 20 21 22 23 24 25 26
BOEDROMION
In 323 we know that the Macedonian calendar synchronised with the Babylonian and that Daisios was the equivalent of Aiiaru, an equivalence which makes Ululu equal Hyperberetaios. The Attic equivalent for Hyperberetaios is Boedromion. Pyanepsion is the next month.
We can be certain that the correct month is Boedromion and that no tampering with the calendar would have dragged the battle into Pyanepsion if the Eclipse occurred, as Plutarch says ‘about the time of the beginning of the Mysteries’. The Mysteries always began on the fourteenth day of Boedromion and ended ten days later on the twenty-fourth or ‘the seventh waning’. These dates were immutable. If the archon needed to postpone a Festival the practice was to repeat days not to move the Festival ie. They would have 24 Dec again, 24 Dec for the third time, etc. rather than declare that Christmas would be on Dec 27, say. These additional days would then be lost usually towards the end of the year when the Festivals were less important.
The effect of these embolismic days was to retard the month against the lunar norm. None of the inscriptions which preserve embolismic dates demonstrate a single case where the Archon/Festival/Sacred Calendar is in advance of the Civic/Pryttany or the later ‘kata Theon’ date. Nor, given that in a lunar month eclipses may only occur mid-month, when the moon is full, is there any case for thinking that any great discrepancy had arisen so early in this year (Boedromion was the third month of the Attic calendar). There is the remotest of possibilities that the Athenians chose to drop the last five days of the month but there can have been no need to correct for embolismic days (14 Boedromion accords well with the mid-month eclipse) and hence the move would be unique.
So why does Arrian place both the eclipse and the battle in the month of Pyanepsion? And he is quite explicit on the matter, II 15 vii ‘Aristander’s prophecy proved true, that Alexander’s battle and victory would occur in the same month in which the moon was partially eclipsed.’ And when an ancient said in the same month they meant in the same named month not some notional synodic/lunar month. We can demonstrate this from Plutarch’s description of Alexander declaring a ‘hollow’ month ‘full’ in order that Aristander’s prophecy that Tyre would fall in that month would not fail, Alexander 25.
Arrian states his two major sources as the Macedonian Ptolemy and the Macedonian, even if ‘naturalised’, Aristoboulos. It would be odd if they did not give Macedonian months, indeed the account of Plutarch would seem to indicate that this was certainly true of Aristoboulos (Granikos in Artemesia 16 ii, death in Daisios 75 vi). We must assume that Arrian believed Hyperberetaios to be the equivalent not of Boedromion but of Pyanepsion. There is evidence for this calendar shift from literary, epigraphic and numismatic sources.
Evidence for the Calendar Shift
The numismatic evidence has been most conveniently collected and presented by G R F Assar in his article ‘Parthian Calendars at Babylon and Seleucia on the Tigris’ Iran Vol 41 (2003) pp171-191.
A series of dated tetradrachms from Seleucia include the abbreviated months Gorpiaios embolimos and Dystros embolimos, where we would expect to find Hyperberetaios and Xandikos. Assar demonstrates that this retardation dates to 28/9 Sept 48 BC, which is almost synchronous with the new era at Antioch which was based from Caesar’s victory at Pharsalos : 1 Oct = 1 Hyperberetaios as the first month of the year rather than Dios. He goes on to show that a series of officina marks preserved on the coinage of Vologases II and Pacorus II can only be made to make sense if Dios had again become the first month of the year, certainly by 77/9 AD, the date of these coins, but more probably in 67AD, which would have been the start of a new nineteen year cycle.
Other evidence shows that, whilst reversed in the Parthian kingdom, the retardation persisted in the Roman East. Josephus, Antiquities of the Jews I 3 iii, equates the second Jewish month with Dios, formerly the first month of the Macedonian calendar, and a graffitto horoscope found in Dura Europa dating to 9 Panemos 176 AD which works out to July 3 when the position of the zodiac were worked out. Panemos should correspond to June. Also two inscriptions from Dura , SEG II 770 and 771, record the month of Dystros embolimos, rather than Xanthikos.
Against this shift being in use in Arrian’s day Zenobios, Centuria VI 30 might be adduced. He is a direct contemporary and clearly states that ‘Hyperberetaios is the last month of the Macedonians.’ But aside from the evidence of Dura from 176 AD we must consider the nature of Zenobios’ work. It is not a History, it is a list of proverbs and turns of phrase with explanations of their origins. It is more than likely that it represents a compilation of previous compendia in which case the explanations came with the phrases and only reflect the state of affairs when they were originally written not that when Zenobios collected them.
The Problem of Hekatombaion and the Archon Year
Hekatombaion was the first month of the Athenian Calendar and the one in which the Eponymous Archon changed; if Arrian is getting his equivalents wrong then his Hekatombaion should really represent the Skirophorion of the previous Archon year, yet his Archon years are correct.
This could be a telling point, but it only really follows if Arrian obtained his months and Archon years from the same source and this does not seem likely, after all why would his Macedonian sources date according to Athenian magistrates and Macedonian months? It would seem more likely that he drew his Archon years from a chronographic source such as the Marmor Parium. This sort of source records the events of a given Archon year with no reference to the months or even seasons.
EVENT ARRIAN DIODOROS MAR.PAR POXY_12 DINSMOOR
ISSOS NIKOKRATES NIKOKRATES KTESIKLES KTESIKLES NIKOKRATES
TYRE * NIKETES NIKERATOS NIKOKRATES NIKETES NIKETES
GAUGAM ARISTOPHANES ARISTOPHANES NIKETES ARISTOPHON ARISTOPHANES
D.DARIUS* ARISTOPHON ARISTOPHON ARISTOPHON KEPHISOPHON* ARISTOPHON
HYDASP HEGEMON CHREMES - - HEGEMON
As can be seen there was confusion as to in which archonship events occurred, however, the fault almost certainly lies with Diodoros and the chronographers. For instance the Marmor Parium says
This is clearly nonsense as more than a year separates Granikos from Issos. Diodoros is notorious for muddling his archons and consuls.3) Apo tes Alexandrou diabaseos eis ten Asian kai maches peri ton Granikon, kai apo tes en Issoi maches Alexandroi pros Dareion, ete LXX, archontos Athenesi Ktesikleious
3) From Alexander’s crossing into Asia and the battle near the Granikos, and from the battle against Dareios at Issos, 70 years, when Ktesikles was archon at Athens.
The case of the fall of Tyre shows the Marmor Parium predating events due to its losing the year of Issos, and not specifically dating the fall of Tyre to this year, although it must be the fall of Tyre that marks the conquest of Phoenicia
The Oxyrhynchus Chronicle is more specific4) From when Alexander seized Phoenicia and Cyprus and Egypt, 69 years, when Nicocrates was archon at Athens
We have one story of the Fall of Tyre in Plutarch that the month in which it fell was ‘hollow’, Alex 25. If the Macedonian calendar regularly alternated ‘full’ and ‘hollow’ months then we can count back from a known ‘hollow’ month, Daisios 323, if we do then we find that Skirophorion was ‘hollow’ and not Hekatombaion. Unfortunately, such an assumption is unwarranted as we cannot be certain how the Macedonian calendar was regulated.[7] In the 112th Olympiad [332 B.C.] Gryllus of Chalcis won the stadion race, and the archons at Athens were Nicetes, Aristophanes, Aristophon and Cephisophon. In the first year Alexander the son of Philippus captured Tyre and took possession of Egypt, where the natives willingly received him because of their hatred of the Persians. Then he ordered [? the foundation of the city of Alexandria] ... He made an expedition to the temple of Ammon, and on his way he founded the city of Paraetonium. In the third year Alexander won another victory over Dareius, in a battle at Arbela. After that Dareius was treacherously killed by his own friends, and the empire of the Persians came to an end; it had lasted 233 years from Cyrus, who established it.
It is possible that there was a record in Athens of events in the inaugural month of the Eponymous Archon and Arrian obtained his Hekatombaia from there. The only exception being the notice of Alexander reaching Thapsakos, this surely comes from the Macedonian sources and is a translated month and therefore Skirophorion Arrian has has simply placed it in the same archon year as the battle of Gaugamela because he knows Hekatombaion is the start of the year.
Issos; Maimakterion (Nov/Dec) or Pyanepsion (Oct/Nov)?
There is one factor in the narrative of the aftermath of the battle which makes it almost certain that the correct month should be Pyanepsion. This is the adventures of Amyntas and his comrades. Arrian II 13 ii ff.
Maimakterion marked the cessation of sea travel due to stormy weather, now given the extreme circumstances maybe one would risk the two day run to Cyprus from Tripolis at the beginning of the close season, especially with ad- hoc crews and heavily laden ships. However, if the battle took place in Maimakterion these fugitives still have to get to Tripolis which is at least 180 miles away.
These troops were infantry and would have had to flee North adding another five days to their journey; 180 miles @15m/day would take twelve days or fifteen with rests so at least two weeks into the no sailing season, probably nearer to three. This becomes even less likely if we consider the crossing to Egypt took place in 333, which would take another three days.
It is surely much more likely that an infantry force would risk the seas just before the stormy season rather than almost a month into it especially when one considers their route is across the open sea not city hopping down the Phoenician coast; also an option for an overland escape.
Conclusion
It would seem that there are good reasons for doubting Arrian’s monthly datings; a reasonable explanation being the shift in the Syrio-Macedonian Calendar in the First century. There still remains the Hekatombaion problem but given the definite mis-dating of Gaugamela and the highly probable one of Issos it seems pertinent to treat his monthly notices with caution.