An accurate determination of the reign of Herod the Great reconciles the biblical nativity and crucifixion accounts with the Jewish historian Josephus' reports and the intermingled history of Jesus Christ with Caesars Augustus and Tiberius, various Roman Governors, Procurators and Prefects, and various Jewish High Priests.
The date of Herod's death is related to New Testament accounts of the birth of Jesus and the early sequence of Roman governors over Israel. Luke records the birth of Christ when Caesar Augustus decreed a census, the first census while Quirinius was "hegemoneúo of Syria" (Luk 2:1-3). The nativity accounts (Mat 2; Luk 1-2) relate that a miraculous "Star of Bethlehem" lead the Magi first to Herod in Jerusalem and then to the infant Jesus (in either Bethlehem or Nazareth). Following the birth of Jesus, Mary and Joseph fled to Egypt to escape Herod's massacre of the infants and returned after Herod's death. Caesar Augustus made Herod's sons Archelaus an ethnarch and Antipas and Philip tetrarchs over Israelite territories following Herod's death (Ant. 17.11.4). Ten years later Archelaus was banished and Rome instituted the offices of Procurator and Prefect over Israel (one of whom was Pontius Pilate), and Quirinius, then Roman governor of Syria, liquidated Archelaus' property (Ant. 17.13.2; 18.1.1-2; Wars 2.7.3; Dio Rome 55.24.9, 55.27.6).
Prior to Herod's death, Josephus reports that a lunar eclipse occurred the night before a Jewish fast (Ant. 17.6.4), and a couple weeks before the following Passover Herod the Great killed his son Antipater and then died of a long-festering illness. Josephus double-dates Herod's reign as "34 years since killing Antigonus and 37 years since being made king by the Romans" (Ant. 17.8.1; Wars 1.33.8). Josephus also synchronizes Herod's reign with Pompey's capture of Jerusalem in 63 B.C. and Augustus' victory at Actium on 31 B.C. Sep 2.
Many researchers and scholars have sought to find a natural phenomena that coincides with the Star of Bethlehem, and they have suggested astronomical phenomena observed in comets, conjunctions, supernovae, as well as bright stars. At present it would seem that all such natural phenomena have occurred in years other than 5 B.C., and thus all the effort has focused on other years to reconcile Josephus' reports of events in the life of Herod the Great with dates of these suggested astronomical phenomena. Presently, all such efforts to reconcile Herod's death and Christ's birth with some natural stellar phenomena have failed in some aspect; largely because of incorrect dates for the birth and crucifixion of Jesus and an apriori dismissal of the Star of Bethlehem as a miracle (for which there won't be any astronomical evidence). That does not mean the Bible is wrong about the Star of Bethlehem, rather it simply reaffirms the Star of Bethlehem as miraculous.
Efforts to conclusively identify which lunar eclipse had preceeded Herod's death (sometimes termed "Herod's Eclipse") have often lacked a correct interpretation of Josephus' grammar about that night, correct dates of Jewish fasts, and a correct chronology of Herod's life. A correct chronology of Herod's life further depends on correctly reckoning the regnal dates and years as reported by Josephus. As Josephus' historical sources were Roman, Greek, and Jewish, several reckoning systems are possible and seemingly intermingled. While Josephus is often faulted for ambiguity (if not outright error) in his histories, in retrosepct it would seem they are largely self-consistent and correct when properly reckoned.
This analysis largely resolves the second phenomenon, that a Herodian Chronology reckoned by the Julian calendar best reconciles that Herod's eclipse was definitely in 5 B.C. and most likely that of March 23 (rather than that of September 15), and Herod died just before the following Passover in 4 B.C., and those are the events to which Josephus alluded when he reported the events leading up to Herod's death.
Herodian chronologies synchronized by various reckoning systems
Four entire chronologies of the Herods are developed in yearly and monthly detail, using non-accession reckoning by Nisan and Tishri calendars and by the Julian calendar with variations for each total lunar eclipse of 5 B.C.
Comparing Josephus' reports to Julian, Nisan and Tishri non-accession reckoning
For each of the three reckoning systems, specifically dated events reported by Josephus are compared to determine which reckoning has the fewest self-inconsistencies and discrepancies with historically dated events.
Herod's appointment as king by Rome in 40 B.C. and capture of Jerusalem from Antigonus in 37 B.C.
Herod's death is double-dated from these two events, and they in turn are synchronized to Pompey's capture of Jerusalem in 63 B.C., Antiochus IV defeat by the Maccabees in 164 B.C., and Titus' destruction of the Temple in 70 A.D.
Synchronisms fix Herod's death to 4 B.C.
The fixed, synchonized interrelationships between many key historically dated events is demonstrated, which precludes adjusting the date of any one event without simultaneously invalidating all other dated events. Josephus' reports of events in the Herodian dynasty are a self-consistent, cohesive collection and discrete events cannot be re-dated separately, nor can Josephus' entire account be re-dated en mass as it is synchronized to several independently dated events.
Having determined that Herod died in 4 B.C., the eclipse preceeding his death must have occurred in 5 B.C. and the events mentioned in the context of that eclipse (a Jewish fast and burning the rebels alive) provide a context against which Josephus' phrasing and grammar of Antiquities 17.6.4 can be analyzed and understood. Five interpretations are possible based on how Josephus phrased the account, but only two are plausible chronologically and only one is strictly grammatical. Josephus relates that on the night before a Jewish fast, High Priest Matthias recused himself from officiating that fast and that same night there was an eclipse of the moon. Ambiguity enters when Jospehus also relates how that High Priest was deposed by Herod who also had others burned alived as punishment for rebellion, and because Josephus' writing style is occasionally terse and disjoint, and lacking a reliable chronological context, misinterpretation is understandable.
Jewish Fasts proximate to Herod's Eclipse
Of the three possible total lunar eclipses, Jewish fasts are known only for the eclipses of 5 B.C. because the eclipse of 1 B.C. occurs in January and there are no Jewish fasts in evidence for that time of year.
The entire chronology of Herod's life from his being appointed governor of Galilee in 47 B.C. to his death in 4 B.C. followed by the reigns of his sons, and Roman consulships, and the birth and crucifixion of Jesus Christ, all reconcile consistently and accurrately with straight forward interpretation of historical texts and coins.
Unreconciled and unresolved are an absence of historical evidence in 5 B.C. (the year of Jesus' birth) for:
This last event is evidenced in 7 or 6 B.C. (not 5 B.C.) as explained by the Actium era coinage of Antioch bearing Varus' name.
The historical reports as compiled by Josephus and other historians (e.g. Dio, Appian, Julius Caesar) presumed different calendars: e.g. Roman and Greek sources typically used A.U.C., Olympic, Actian, and Julian calendars, while Jewish sources used either Tishri or Nisan calendars.
The purpose of constructing comparative chronologies (linked below) is to illustrate which method of reckoning and dating is most consistent with Josephus' reports.
The dates in the Herodian chronologies are slightly hypothetical for the B.C. portion because the intercalation method used by the Hebrews prior to 5 B.C. is hypothetical. Yet, Jewish reigns, festivals, and dates (and hence synchronisms) are reported by Josephus, and the Jewish Sanhedrin intercalated their calendar by which those events were observed, and so some accommodation to intercalation, even hypothetical, is necessary to approach a realistic chronology. Further, the Romans intercalated haphazardly until they adopted the Julian calendar in A.U.C. 709 (45 B.C.) Accordingly, the dates and durations noted pre-B.C. may have a 1 month error. Regardless, in fact the chronology largely confirms Josephus' reports, which suggests the intercalation formula used by the ancient Hebrews and the Fourmilab converter are nearly identical.
But can the death of Herod the Great be dated reliably?
Yes. The entirety of Josephus' reports impacting on the life and death of Herod the Great and his sons, and evidence from coins can all be reconciled with Herod's death in March of 4 B.C. It can further be shown that Josephus' same reports cannot be reconciled with a 1 B.C. date for Herod's death.
Three studies were undertaken to develop detailed chronologies of Herod's life including all of Josephus' reports that were directly or indirectly related to events in Herod's life and his sons. Three chronological tables were prepared: extending from 168 B.C. to 72 A.D. (Antiochus IV desecration of Solomon's Temple to Titus' destruction of Herod's Temple) with the Julian, Olympic, A.U.C, Actian, and A.M. calendars (including intercalary months) plus the regnal years of the Herods, Jewish priests and kings, and various durations relative to Jerusalem and the Temple, all juxtaposed in their proper inter-relationships. In all three chronologies, fixed events such as Antiochus' desecration of the Temple, Pompey's capture of Jerusalem, the Battle of Actium, lunar eclipses, Roman emperors and consulships, etc., were specified by their equivalent dates on all four calendars. Roman intercalary months were determined by Chris Bennet at Roman Dates, and Roman consulships were excerpted from E.J. Bickerman, Chronology of the Ancient World (Thames & Hudson; rev. 1980). Lastly, Josephus' reports of the reigns of priests and kings and construction projects, etc., were arranged to reconcile with those fixed dates. Historical and biblical synchronisms and relative dates of reigns and periods were then highlighted.
The Julian based chronology was further revised in two variants reflecting different schedules in Herod's final months, assuming Herod's eclipse was either March 23 or September 15 of 5 B.C. Construction of a similar chronology predicated on Herod's death in 1 B.C. was abandoned as unmanageable due the significant irreconcilable dates and reigns. Rather an analysis showing the synchronisms which fix Herod's death to 4 B.C. was manageable and more informative.
The differences between the chronologies are several months confined to the reigns of Hyrcanus II, Aristobulus II and Antigonus, because only their reigns are sensitive to Hebrew regnal dating by either Nisan or Tishri calendar years. Unfortunately, because Josephus doesn't report months for the reigns of the high priests (including Simon, Matthias, Joazar, and Eleazar who are all mentioned in the context of Herod's death), their reigns don't help discriminate between either lunar eclipse of 5 B.C.
Since Herod's death was some months after a lunar eclipse, the events of Herod's death can also be made to reconcile with either total lunar eclipse of 5 B.C. And because Herod died just before Passover, the total lunar eclipse on 1 B.C. January 9 also plausibly fits the winter-time events surrounding Herod's death, but as will be shown, other larger constraints preclude Herod's death being dated to 1 B.C.
Rather than analyze the myriad ways that Josephus' reports about Herod won't reconcile with accession reckoning, it is more instructive to compare Nisan, Tishri and Julian non-accession reckoning of Josephus dates and durations using four different calendar systems. These chronologies of the events in the lives of the Herod and his sons are:
- Herodian Chronology by Julian non-accession reckoning with Herod's death related to the lunar eclipse of 5 B.C. March 23
- Herodian Chronology by Julian non-accession reckoning with Herod's death related to the lunar eclipse of 5 B.C. September 15
- Herodian Chronology by Nisan non-accession reckoning
- Herodian Chronology by Tishri non-accession reckoning
These chronologies are based on non-accession reckoning, rather than accession reckoning, because:
The problems with Josephus using either Nisan or Tishri accession year dating of any Herodian are significant and immediate, literally. Either Nisan or Tishri accession reckoning of the reign of Hyrcanus II minimally delays the reign of Aristobulus II by 1 year and further delays the 2nd reign Hyrcanus II by 2 years. The problem being that the transition from the reign of Aristobulus II to the 2nd reign of Hyrcanus II must coincide with Pompey's capture of Jerusalem in 63 B.C.; and the end of Hyrcanus II 2nd reign must coincide with Parthians making Antigonus king in 40 B.C.; and Antigonus' reign must end with Herod's capture of Jerusalem in 37 B.C.; which must also coincide with both Pompey's capture 27 years earlier and ending 126 years of Hasmonean rule, and then Titus destroying the Temple 107 years later.
None of which reconciles with accession reckoning which essentially postpones the start of any reign until the following year, and those postponements collectively extend any chronology by many years. But these reigns can be neither postponed nor the chronologies extended because they must synchronize with historical events whose dates Josephus also reported and are corroborated by non-Josephean sources (e.g. 1Maccabees, Appian, Dio, Caesar, etc.).
Because Josephus was a Jew writing about Jewish kings and events, it is often assumed he used either Nisan or Tishri accession reckoning. While the chroniclers of the Old Testament kings used some form of regnal reckoning, it would seem either Josephus himself, or more likely his sources, did not. However, Josephus' reports do reconcile closely with historical events when non-accession inclusive reckoning is used, especially in Julian dates. Having studied this, Harold Hoehner concludes:
In conclusion, it seems reasonable that in computing Herod's reign one needs to consider that it was reckoned in accordance with the Julian calendar, the nonaccession-year system, and inclusive reckoning. [p. 105]
Thus, a careful look at the dates of the succession of Herod the Great's sons and grandson, one can only conclude that the 4 B.C. date is the best date for Herod's death. [p. 107]
Harold W. Hoehner, "The Date of the Death of Herod the Great",
Chronos, Kairos, Christos I (1989) p. 101-111Of the events related to the life of Herod and his sons, most of Josephus reports are dated or elapsed in whole years. A few are also specific to a number of months (e.g. the reigns of Aristobulus II and Antigonus being 3 years and 3 or 6 months), or based on a season or festival which can be approximated to a particular month (e.g. Herod's siege and capture of Jerusalem 5 or 6 months after winter ends, and Herod's death before Passover).
Josephus is our only source for the reigns of the High Priests, and except for Aristobulus II and Antigonus, Josephus only reported their reigns in whole years. Consequently, with a few exceptions, reconciling Josephus' reports to actual dates and durations was limited in resolution to whole years. In developing the chronologies, the starting and ending years of Hyrcanus II and Aristobulus II and the high priests (e.g. Matthias) were chosen to fit fixed date and duration constraints, but the specific months chosen were somewhat arbitrary and might be adjusted a couple months earlier or later. However, adjusting the reigns of Hyrcanus II and Aristobulus II are greatly constrained (if not fixed as shown) because they must also synchronize with Pompey's capture of Jerusalem, which in turn Josephus relates to Herod's capture of Jerusalem, which in turn is fixed in relation to other historical events.
What reckoning system(s) did Josephus use?
To ascertain how credible is any Herodian Chronology based on Josephus' reports, it must be understood what calendar(s) and reckoning system(s) did Josephus use, and accordingly how accurate and consistent are Josephus' reports.
Upon comparison, it would seem based on chronologies which best fit Josephus' reports and actual history, that from the Hasmonean revolt in 164 B.C. until 45 B.C., Josephus or his primary sources likely used Nisan non-accession reckoning by the Anno Mundi or Olympic calendars, and then after 45 B.C., when the Julian calendar was established, Josephus or his primary sources used a Julian non-accession reckoning.
To perform that comparison, for each of the three chronologies (Julian, Nisan and Tishri reckoned) comparative tables (shown below) were constructed using excerpts of critical historical events from each of the three Herodian chronologies identified above. Excerpted were the reigns of key persons and events in the life of Herod the Great, events which could also be correlated with other independent historical sources (synchronisms), which were then compared against Josephus' reports.
The durations in the "Josephus" column are as reported by Josephus. The Julian, Nisan, and Tishri "reign" columns are the actual count of years (and months if known) from a start-event through to and including the end-event year or month, from the Herodian Chronological tables.
For each key reign or event, these summaries compare what Josephus reported ("Josephus" column) against the inclusive year counts ("reign" column) and date ranges ("non-accession exclusive range" columns) imposed on each chronology by each calendrical reckoning system. The standard (against which Josephus' accuracy was compared) is the dated, tabularized calendar underlying each Herodian chronology:
- The years counted from the Herodian chronological tables is presumed to be the correct standard and shown in the "Julian reign" for the Julian comparison table, the "Nisan reign" for the Nisan comparison table, and the "Tishri reign" for the Tishri comparison table.
- The Julian, Nisan, and Tishri "exclusive range" columns are the difference between the pairs of ending and beginning years (for each calendar system) as extracted from the Herodian chronological tables for Julian, Nisan, and Tishri reckoning (respectively).
All three tables below compare "non-accession exclusive ranges" (what would normally be obtained by subtracting the start date from the end date) using years dated in Hebrew A.M. (anno mundi "in the year of the world"), Varronian A.U.C. (ab urbe condita "from the foundation of the city" [i.e. Rome]), Julian, and Olympic systems. Because subtraction excludes the length of start year, that difference is an "exclusive" reign. Consequently, if Josephus' reckoned inclusively, then correct differences will always be 1 year less than what Josephus reported, and any other difference is incorrect.
The "Factual (inclusive reckoning)" columns are speculative and portray durations resulting from different lengths of solar years vs lunar calendar years in A.M. dating. These columns did not contribute to the conclusions of this analysis.
Here are a few examples of what the chronology comparisons show:
Consider the first row of each table. Since Josephus reports 126 years of Hasmonean rule ended when Herod captured Jerusalem, then a manual count of the years from the Hasmonean revolt against Antiochus IV to Herod's capture of Jerusalem should equal 126, in one or more of the three chronologies, by some calendar. Accordingly, the 1st row of each table below compares this period of Hasmonean rule. But in all 3 tables none of the calendar dates corroborate this period. In actuality, examination of the chronological tables show the period lasting 127 years and some months. While this is not a significant error in Josephus' accuracy, it does demonstrate that whatever were Josephus' sources, they were not reckoned in non-accession years by any of the presumed calendar systems. So these entries ("Josephus" column and the conflicting "range" columns) in the table are all highlighted in yellow to illustrate their non-corroboration.
Since Josephus reports Herod's reign as 34 years from killing Antigonus and 37 years from being appointed king by Rome, then a count of the years from each of those events to the year of Herod's death should equal 34 and 37 respectively, by some calendar. The tables confirm this for Julian and Tishri reckoning, but the Nisan reckoned chronology shows the actual count of years ("Nisan reign" column) to be 33 and 36 years respectively, i.e. if Josephus used Nisan non-accession reckoning, then his report of 34 and 37 years conflicts with the count of actual Nisan non-accession reckoned years. Because in the Nisan comparison table the "Nisan reign" values are the presumed standard, the "Josephus" column values are highlighted in yellow to illustrate their non-corroboration as are "range" columns because neither could Josephus have calculated Nisan reckoned inclusive reigns from any of these ranges. Yes, they reconcile with what Josephus reported (same as in the Julian and Tishri comparison tables as well), but incorrect if Josephus intended to report Nisan reckoned reigns.
Consider the last row of each table. Josephus reports the period from Herod's rebuilding of the Temple to its destruction by Titus as 107 years, but an actual count of the intervening years shows it was 106 years and maybe 3 months in all three chronologies, which in non-accession reckoning is 107 years. 9 months short of 107 years, mathematically, is not a significant error in Josephus' accuracy, but it demonstrates that whatever were Josephus' sources, they did reckon in non-accession years by the A.M., A.U.C. and Julian calendars with only the Olympic period being 1 year too long (and highlighted in yellow).
Consider the second row of each table. Josephus reports the 1st reign of Hyrcanus II as 9 years. The arrangement of the reigns of Hyrcanus II and Aristobulus II are somewhat arbitrary in the chronologies because Josephus has not reported any specific dates, and they were positioned to reconcile with Pompey's capture of Jerusalem ending the reign of Aristobulus II and reinstating Hyrcanus II to his 2nd reign. Regardless, an actual count of the years of Hyrcanus II 1st reign is 9 (as reported by Josephus) for the Julian and Nisan reckoned chronologies, but 10 in the Tishri reckoned chronology. Further, only in the Tishri chronology could the reign have been mathematically computed from the A.M. "range".
Lastly, consider the period between Pompey's capture of Jerusalem and Herod's capture of Jerusalem. Josephus reports the period as beginning and ending on the same fast of the same month 27 years apart. This period is corroborated by Nisan, Tishri, and Julian non-accession reckoning in all four calendar systems, the difference between the end and start dates always equaling 26 (1 less being consistent with 27 year inclusive non-accession reckoning). Accordingly, none of those entries conflict.
Examination of the three comparative tables for the fewest inconsistencies shows that Tishri reckoned reigns and dates exhibit most irreconcilable conflicts, and seem unlikely to have been used by Josephus' sources.
Note that the age of Herod the Great at death cannot be precisely reconciled in any system. Josephus records it as 70 years, but that conflicts with Josephus also reporting Herod's age as 15 in 47 B.C., which most scholars presume to be an error and consequently they emend the age as 25 to conform more closely with an age of 70 at death.
Note also that only in Julian non-accession reckoning does the reign of Antigonus perfectly reconcile with other reigns and events.
Ignoring the imprecision inherent in the reigns of the high priests and irreconcilability of Herod's age, Josephus' chronological reports regarding the life of Herod the Great are correct within a few months, by non-accession reckoning in A.M. Nisan years before 37 B.C. and Julian years after 40 B.C., except for the periods relative to Jerusalem or the Temple. A consistent explanation presents itself:
Events in 2nd century B.C. hellenized Israel could have used Olympic dates, and certainly would use A.M. dates on the Nisan calendar, but would not have been dated in A.U.C. (relative to founding the city of Rome) and the Julian calendar didn't exist. Consequently Josephus' source(s) for 2nd century B.C. events likely used A.M. dates reckoned on the Nisan calendar (and possibly Olympic dates as well). Later, after Rome adopted the Julian calendar in 45 B.C., Josephus' source(s) for Herod's siege and capture of Antigonus (40-37 B.C.) transitioned from Nisan to Julian dating. In such a transition, either Josephus changed sources, or the source(s) changed dating systems, or both.;
However, durations relative to Jerusalem or the Temple would not be considered "regnal" (which is for royalty, e.g. rulers, kings, emperors, etc.) and instead would be reckoned factually, inclusively, for which A.M. date ranges reckon consistently, except for Josephus' reported period of Hasmonean rule. While this is not the regnal accession reckoning of the Old Testament, it is a natural consequence of someone (a source of Josephus or an assistant to him?) simply subtracting the difference of years from the beginning to the end and adding 1 year, except neglecting to do so for the period of Hasmonean rule. Because the durations are lengthy, it is also plausible that beginning events were dated A.M. (prior to the adoption of the Julian calendar) and ending events were dated Julian thereafter, and math errors were committed when converting date systems.
During a hypothetical transition from A.M. Nisan dated sources to Julian dated sources, Nisan non-accession reckoning reconciles exactly with Josephus for Herod's siege of Antigonus (37 B.C.) but Julian non-accession reckoning reconciles exactly with Antigonus' reign (3 years and 2-3 months from 40-37 B.C.). Thereafter Josephus' source(s) reconcile with non-accession reckoning on the Julian calendar which was by then well established.
To summarize, it would seem Josephus' source(s) used:
- A.M. inclusive reckoning for periods relative to Jerusalem and the Temple and during 2nd century B.C.
- A.M. Nisan non-accession reckoning as late as 37 B.C., transitioning to Julian
- Julian non-accession reckoning as early as 40 B.C. and thereafter, transitioning from A.M. Nisan
Note that 4th, 3rd and 2nd to last entries in each table are not actual "reigns" reported by Josephus but rather what this analysis inferred from Josephus' reports of when Antipas and Philip acceded and then were deposed or died.
The biblical quote in the 2nd to last entry, is obviously not a Josephus reference either, but is included to show the synchronism with Herodian chronology. In John 2:20 the Pharisees exclaim "It took forty-six years to build this temple [ναός] …" which occurred in Nisan while Jesus was attending the 1st Passover recorded by John, which Passover occurred soon after Jesus was baptized "at about 30 years of age" (Luk 3:23) "in Tiberius' 15th year, when Pilate was governor of Judea" (Luk 3:1). Having shown Jesus was likely born mid-August 5 B.C., then Jesus' 30th birthday would fall proximate to the Hebrew month of Elul, which plausibly puts Luke 3 in September (~Tishri) of A.D. 26 and John 2:20 in April (~Nisan) of A.D. 27.
The highlighting in following tables shows:
corroboration of Josephus' reports with inclusive reckoning (no highlighting)
Josephus' reports that are not 1 greater than the range difference
durations relative to Jerusalem or the Temple that reconcile with A.M. date ranges
reigns of the high priests ambiguously dated
Josephus' reports compared to Julian non-accession reckoning:
Reign, Period or Event Hasmonean rule3 Hyrcanus II reign 1st* Aristobulus II reign* 4 Hyrcanus II reign 2nd* Jerusalem from Pompey to Herod Herod's siege of Antigonus5 Antigonus' reign5 Battle of Actium in Herod's 7th year Herod aids Egypt in his 14th year Herod begins Temple rebuild in his 18th year Herod completes Cesarea in 10 years in his 28th year in 192nd Olympiad Herod's reign from kingship by Rome
35.49
36.52 Herod's reign from killing Antigonus Herod's age at death** Archelaus' reign Philip's reign Antipas' reign to Gaius' 2nd year6 "46 years to build this Temple" Temple from Herod to Titus
Josephus' reports compared to Nisan non-accession reckoning:
Reign, Period or Event Hasmonean rule3 Hyrcanus II reign 1st* Aristobulus II reign* 4 Hyrcanus II reign 2nd* Jerusalem from Pompey to Herod Herod's siege of Antigonus5 Antigonus' reign5 Battle of Actium in Herod's 7th year Herod aids Egypt in his 14th year Herod begins Temple rebuild in his 18th year Herod completes Cesarea in 10 years in his 28th year in 192nd Olympiad Herod's reign from kingship by Rome
35.49
36.52 Herod's reign from killing Antigonus Herod's age at death** Archelaus' reign Philip's reign Antipas' reign to Gaius' 2nd year6 "46 years to build this Temple" Temple from Herod to Titus
Josephus' reports compared to Tishri non-accession reckoning:
Reign, Period or Event Hasmonean rule3 Hyrcanus II reign 1st* Aristobulus II reign* 4 Hyrcanus II reign 2nd* Jerusalem from Pompey to Herod Herod's siege of Antigonus5 Antigonus' reign5 Battle of Actium in Herod's 7th year Herod aids Egypt in his 14th year Herod begins Temple rebuild in his 18th year Herod completes Cesarea in 10 years in his 28th year in 192nd Olympiad Herod's reign from kingship by Rome
35.49
36.52 Herod's reign from killing Antigonus Herod's age at death** Archelaus' reign Philip's reign Antipas' reign to Gaius' 2nd year6 "46 years to build this Temple" Temple from Herod to Titus * Precise regnal year reckoning is impossible for many high priests as their beginning month is unknown to history and not recorded by Josephus. Often, Josephus is our only source for reigns of Jewish High Priests, and Josephus only reported the duration and sequence of their reigns, but not their accession and termination dates. Consequently, when actual reign dates are unknown, the starting month in the chronology is arbitrarily adjusted for best-fit reconciliation with Josephean chronology, endeavoring to demonstrate where their reigns would occur assuming Josephus' reports are correct.
** Herod's age can not be analyzed exactly as there is no date for his birth, Josephus is the only source, and Josephus gives Herod's age as "15" when appointed governor of Galilee (~47 B.C.) and about "70" near death (~5B.C.). Both numbers can't be correct because 42 years elapsed between 47 and 5 B.C., and 15 years of age + 42 years yields an age of 57 near death, not 70. Most scholars therefore assume the "70" year age near death is better attested and thus the "15" year age when appointed governor is an error, which is emended by assuming 28 years (=70-42) is approximately correct and then round down to "25" years of age when Herod was appointed governor of Galilee.
1 365.25-day factual reckoning is computed by converting the Hebrew range (last day of the end month – first day of the start month) to Julian day range and dividing by 365.25 to yield the number of solar years in the range.
2 355-day factual reckoning is computed by converting the Hebrew range (last day of the end month – first day of the start month) to Julian day range and dividing by 355 to yield the number of average-length Hebrew calendar years in the range. Note: a more exact computation would be to count the number of actual deficient, regular and complete common and embolismic years.
3 Josephus reports Herod's capture of Jerusalem and killing of Antigonus as ending 126 years of Hasmonean rule (Ant. 14.16.4), presumably reckoned from Judah Maccabee's rededication of the Temple (1Macc 4:52; Ant. 12.7.6). This computation is correct by the Hebrew (Anno Mundi) and Olympic calendars but is 127 years measured by A.U.C and Julian calendars.
4 Josephus reports two different durations for the reign of Aristobulus II, who deposed Hyrcanus II and was then himself deposed by Pompey who restored Hyrcanus II, all of which must synchronize with Pompey's siege and capture of the Temple in 63 B.C. Josephus further reports Pompey entered the Temple 'on the 3rd month on the day of the fast' and further synchronizes Herod's capture of the Temple from Antigonus 27 years later again 'on the 3rd month on the fast' which ended 126 years of Hasmonean rule. The only date which reconciles Pompey and Herod's respective captures of the Temple is Sivan 23, Sivan being the 3rd month of the Hebrew sacred (Nisan-reckoned) calendar, of 37 B.C. So, using 63 B.C. Sivan 23 as the end of Aristobulus' II reign, and assuming it lasted:
- 3 years and 6 months yields Teveth of 66 B.C. (further assuming Julian reckoning), but leaves a 3 month gap between the end of Hyrcanus' II 1st reign and Aristobulus' II when using Tishri reckoning, or when using Nisan reckoning shortens Aristobulus' reign to just 3 regnal years (4 regnal years are needed to accommodate a reign of 3 years and 3 or 6 months).
- 3 years and 3 months yields Nisan of 66 B.C. (further assuming Nisan reckoning), but would leave a larger 6 month gap between the end of Hyrcanus' II 1st reign and Aristobulus' II when using Tishri reckoning or a 3 month gap when using Julian reckoning.
5 Josephus reports Antigonus reigned 3 years and 3 months when Herod besieged him (in Adar as winter ended). But it wasn't until 4 months later in Sivan that Herod captured Jerusalem and killed Antigonus, marking the beginning of Herod's reign reckoned from replacing Antigonus and becoming king in Jerusalem.
Though Josephus gives exact months, Antigonus' reign cannot be reconciled to the exact month from Josephus' self-contradictory reports:
- 40 B.C. After the festival of Pentecost (in the month of Sivan), Lysanias succeeded his father Ptolemy of Chalcis and the Parthians deposed Hyrcanus II and made Antigonus king.
- 37 B.C. Herod besieged Jerusalem as winter ended (Adar or Nisan) at which time Antigonus had reigned for 3 years and 3 months.
- 37 B.C. when Agrippa and Gallus were consuls, Herod captured Jerusalem 27 years after Pompey (63 B.C.), to the same fast day, in the "3rd month". As Tishri cannot be the 3rd month of the year, and the siege began as winter ended (6 or 7 months before Tishri), the "3rd month" can only be Sivan (as reckoned by the Hebrew sacred calendar from Nisan) in which the 23rd is an ancient fast day (commemorating Jeroboam blockading the pilgrimage to Jerusalem).
- But there are exactly 3 years from Sivan (~ Pentecost) of 40 B.C. to Sivan (~23rd) of 37 B.C. and Antigonus could not have reigned 3 years and 3 months when Herod began his siege 3 months earlier.
- Josephus also reports Herod besieged Antigonus in Herod's 3rd year, but this synchronism is only possible in a Nisan non-accession reckoning. Both a Tishri and Julian reckoning cause Herod's siege to be in Herod's 4th year. While accession Tishri or Julian reckoning would synchronize Herod's siege of Antigonus, it breaks other synchronisms with Herod's death.
6 Josephus reports Antipas was deposed by Roman emperor Gaius (Caligula) in Gaius' 2nd year. Antipas' last known coin was dated ΜΓ ("43") and minted in his 43rd year:
Yet on the coins struck in "year 43" the name and title appear in the nominative form: ΗΡΩΔΗC ΤΕΤΡΑΡΧΗC. This latest series also depicts, in the dative, the name and title of Gaius Caligula, which suggests that the coins were symbolically struck for and in honor of the emperor.26 The inscription appearing on both sides of the year 43 coins may be regarded as a tribute to the emperor, translated as: "Herod the Tetrarch to Gaius Caesar Germanicus". [p.39-41]
26This interesting variation in declensions, which has not previously been analyzed, increases our understanding of the coins dated ΜΓ as well as justifies their association with the circumstances surrounding the appointment of Agrippa I to the kingship. [p. 205]
Ya'akov Meshorer, Ancient Jewish Coinage vol 2,
Amphora books (1982)No later coins of Antipas are known. Only one reference to a coin alleged to have been minted by Antipas and dated ΜΔ ("44") was published by J. Eckhel, Doctrina Numorum Veterum, iii (Vienna, 1794), pp. 486-489 based on a report from Vaillant and Galland. However, the reading of that "44th year" date has been questioned by A. Reifenberg, Ancient Jewish Coins (1973) p. 19 who cites Schürer, and Schürer contends:
… 2. Of the coins of Antipas of A.D. 43 (ΜΓ) there are at present three known examples (Madden, Coins of the Jews, 1881, p. 121 sq.; two according to Lenormant, Trésor de Numismatique, p. 125, pl. LIX. n. 19 and 20; one according to de Sauley, Mélanges de Numismatique, t. ii. 1877, p. 92). Their existence is thus put beyond question. But great difficulties are caused by a coin with the supposed date of A.D. 44 (ΜΔ). It has been described not only by the little to be depended upon Vaillant, but also in a manuscript account of travels by Galand, who found it at Jericho in A.D. 1674 (communicated by Fréret, Mémoires de l'Académie des inscr. et belles-lettres, t. xxi. 1754, pp. 292 sq.). Sanclemente, pp. 315-319, and Eckhel, Doctrina Num. iii. 487, have both occupied themselves with this discussion. Both conjecture that the date has been incorrectly read (it may rather have been ΛΔ = 34). Compare pro and contra also : Ideler, p. 391; Madden, History of Jewish Coinage, p. 99; Coins of the Jews, p. 122; Riess, 1880, pp. 55-57; Sattler, Das Jahr. 749; Memain, p. 448 f.; Kellner, p. 176. Eckhel's reasons are very convincing; he points especially to this, that the coin described by Galand, according to its other peculiarities, corresponds to those of the year 34, but not to those of the year 43. The only point of difficulty arises from the statement made by Fréret at p. 293 in reference to Galand's description: "les lettres de l'époque ΜΔ sont très-nettement figurées dans son manuscrit et absolument séparées l'une de l'autre." The matter therefore continues undecided. But even if the date 44 be the correct reading, still the death of Herod cannot in any case be placed earlier than A.U.C. 750. [p. 466]
Emil Schürer, History of the Jewish people in the time of Jesus Christ, div1, vol1,
Translated by Rev. John MacPherson, Charles Scribner's Sons (1891)… and his coins, just like those of old Herod, have upon them no image.16
16 On the coins of Herod Antipas, compare Eckhel, iii. 486-490; Mionnet, v. 5G6; Lenormant, Trésor de Numismatique, p. 125, pi. lix. n. 16-20; Cavedoni, Bibliiche Numismatik, i. 53, 58-60; Levy, Geschichte der jüd. Münzen, p. 80; Madden, History of Jewish Coinnge, pp. 95-99; De Sauley, Numismatic Chronicle, 1871, p. 254; Madden, Numismatic Chronicle, 1875, pp. 47-49; De Sauley, Melanges de Numismatique, t. ii. 1877, p. 92; Madden, Coins of the Jews, 1881, pp. 118-122 (this gives the most complete list).—The coins fall into two classes: 1. The one class has the inscription ΗΡΩΔΟΥ ΤΕΤΡΑΡΧΟΥ, with the number of the year, 33, 34, 37, 38; on the other side the name of the city, ΤΙΒΕΡΙΑΣ. 2. The other class has the inscription ΗΡΩΔΗΣ ΤΕΤΡΑΡΧΗΣ; on the other side, ΓΑΙΩ ΚΑΙΟΑΡΙ ΓΕΡΜΑΝΙΚΩ. Of this second class there are only three examples which can be with certainty identified, all with the year number ΜΓ or 43 = A.D. 39-40. Since this was most probably the last year of Herod Antipas, the existence of the year number 44, which some prefer to read, is extremely questionable. One of the two who contend for this date, Vaillant, is generally not to be depended on; the other, Freret, describes a coin (in the Memoires de I'Academic des Inscriptions et Belles-Lettres, alte serie, t. xxi. 1754, p. 293, according to a manuscript by Erland) which had on one side the inscription ΗΡΩΔΟΥ ΤΕΤΡΑΡΧΟΥ (sic), while the inscription on the other side was quite illegible. The coin seems therefore to have belonged to the first class, and it may be reasonably conjectured that instead of ΜΔ, 44, should be read ΛΔ, 34. Compare, however, what is said in vol. i. p. 465.—The coins of Antipas, with the name of the emperor, without his image, occupy a middle position between those of Herod the Great, which have neither name nor image of the emperor, and those of Philip, which have both. [pp. 20-21]
Seeing then that the deposition of Antipas was closely connected with the appearance of Agrippa, it would seem that it must have taken place, if not in A.D. 38, at least in A.D. 39. In fact, it can be proved that it actually occurred not earlier and not later than the summer of A.D. 39. Not earlier : for the forty - third year of Antipas, of which we have coins extant, only began with 1st Nisan 792 A.U.C, A.D. 39. But also not later. Caligula was absent from Rome from autumn A.D. 39 till 31st August A.D. 40 on an expedition to Gaul, Germany, and Britain (Dio Cassius, lix. 21-25 ; Suetonius, Caligula, 17, 43-49 : his entry into Rome "natali suo," i.e. 31st August, see Suetonius, Caligula, 8). Seeing then that the deposition of Antipas took place while Caligula was at Baiae, and seeing also, according to Josephus, Antiq. xix.8.2, that it cannot have occurred after the German campaign, it must have happened before that campaign, i.e. before autumn A.D. 39. [p.37]
—On the coins of Herod bearing what is supposed to be the year number 44, which would require an extension of his reign down to A.D. 40, see above, vol. i. p. 466, and the present vol. pp. 20, 21. Were the existence of this coin well established, we should be obliged, with Lewin, to assign the deposition of Antipas, not to the period of Caligula's residence at Baiae, but to the period of his Gallic campaign, and so to assume a serious error in Josephus. [p. 38]
Emil Schürer, History of the Jewish people in the time of Jesus Christ, div1 vol2,
Translated by Rev. John MacPherson, Charles Scribner's Sons (1891)
Josephus, in Wars 1.33.8 and Ant. 17.8.1, essentially double-dates the reign of Herod the Great upon death, relative to two different historical reference events; 1) being appointed King and 2) defeating Antigonus:
Wars 1.33.8. So Herod, having survived the slaughter of his son five days, died, having reigned thirty-four years since he had caused Antigonus to be slain, and obtained his kingdom; but thirty-seven years since he had been made king by the Romans.
Wars 1.33.8
William Whiston, The Works of Josephus, trans. (1736)Ant. 17.8.1 When he had done these things, he died, the fifth day after he had caused Antipater to be slain; having reigned, since he had procured Antigonus to be slain, thirty-four years; but since he had been declared king by the Romans, thirty-seven.
Ant. 17.8.1
William Whiston, The Works of Josephus, trans. (1736)
Herod the Great died in March of 4 B.C.: 37 years since being appointed King by Antony and Octavian in 40 B.C. and 34 years since Jerusalem was retaken and Antigonus was slain in 37 B.C.
Josephus' accurate dating of these two reference events, Herod being appointed King in 40 B.C. and defeating Antigonus in 37 B.C. will be demonstrated, in chronological order.
Herod was appointed king of Judea in 40 B.C. by Mark Antony and Caesar Octavian.
There are five constraints in reconciling Josephus' reports to date Herod's appointment as king in Rome:
- Appian's accounts place Antony and Octavian apart and away from Rome until after the Treaty of Brundisium in early October of 40 B.C.
- Josephus' accounts place Herod in Jerusalem prior to Pentecost in early Sivan (May) when the Parthians attack.
- Josephus' accounts of violent wintery storms on Herod's voyage to Rhodes implies the winter season, but it would have been late spring and early summer after Herod fled Jerusalem in May.
- Since May, Herod would only have had a few months to reach Rhodes and build a "three-decked ship" and sail it to Brundisium and then Rome by October.
- Antioch coins also date the Parthian invasion of Syria (and hence Judea) to 41/40 B.C.
The analysis of Herod the Great's 40 B.C. voyage to Rome and return demonstrates that:
Josephus' only error was in dating Herod's appointment as King to the 184th Olympiad. All other details of Josephus' account reconcile with other historical reports and circumstances.
Herod the Great fled Jerusalem in early May of 40 B.C. to escape the Parthian attack. He then marched to Masada, Petra, Rhinocolura and Pelusium, and then sailed to Alexandria (per Josephus Ant. 14.13.1-14.14.2 and Wars 1.13.1-1.14.2) where he missed meeting Mark Antony who had wintered there (41/40 B.C.) with Cleopatra but who had left earlier in the spring for Tyre (per Appian, The Civil Wars 5.1.10-5.6.52).
Herod then sailed for Rome to catch up with Antony, but a not-uncommon summer storm on the Mediterranean detoured him to the island of Rhodes. While on Rhodes for 86-90 days, Herod "refitted and provisioned a three-decked ship", possibly a 120-oared, three-level, trihemiolia (which were well known at Rhodes) or a 170-oared trireme and also hired a captain and crew, and then sailed for Brundisium (per Josephus Ant. 14.14.3, and Wars 1.14.3) likely along the southern coast of Crete and then the western coast of Greece. This voyage would be windward, against prevailing winds, and likely took 32-42 days.
Herod arrived at Brundisium, Italy just after Antony and Octavian had returned to Rome having just agreed to the Treaty of Brundisium (dated variously late September to 12 October of 40 B.C. per the Concordia Casinum Inscription and Virgil's 4th "Eclogue"). Herod then proceeded to Rome, likely overland via the faster Roman post road, while he sent his ship around the longer coastal voyage to meet him at Rome.
Antony, Octavian, and the Roman Senate appointed Herod the Great as King of Judea after the Treaty of Brundisium sometime in early to mid-October of 40 B.C., and within 7 days thereafter Herod sailed back to Ptolemais (Acco) on the coast of Judea. This would be a short 10-15 day leeward (downwind) voyage arriving in late October or early November of 40 B.C.
Upon Herod's arrival in Ptolemais, it was late October or early November of 40 B.C.; in time for Herod to gather an army and begin a campaign against Antigonus and the Parthians (per Josephus Ant. 14.14.4-14.15.1 and Wars 1.14.4-1.15.4), ultimately capturing Antigonus and Jerusalem in 37 B.C., as is demonstrated next.
Herod captured Antigonus and Jerusalem in 37 B.C.
Josephus accounts for all three years between Herod's appointment as King in 40 B.C. and his capture of Jerusalem in 37 B.C. Jospehus describes Herod gathering an army through Galilee immediately following Herod's return from Rome (implicitly during the winter of 40/39 B.C.), and further explicitly refers to two "winter" seasons (39/38 B.C. and 38/37 B.C.) prior to Herod's capture of Jerusalem and Antigonus:
Season and Year Wars Ant. Event 40 B.C. 1.15.2 14.15.1 Ventidus assigned to fight the Parthians in Syria. fall 40 / 39 B.C. * 1.15.3 Herod returns from Rome as King and begins campaign in Galilee. 39 B.C. 1.15.4 Herod captures Joppa, recovers relatives from Masada & Resa, marches to Jerusalem 39 B.C. 1.15.5 14.15.2 Herod defeated at Jerusalem, Silo's men demand winter quarters
(after treaty of Misenum (at Puteoli), Antony dispatched lieutenants while not yet in winter quarters, to Herod in Judea to collect tribute; Appian Civil Wars 5.8.75)winter 39 / 38 B.C. ** 1.15.6 14.15.3 Herod and Roman army get supplies at Jericho, Roman army into winter quarters 1.16.2 14.15.4 Herod captures Sepphoris in snow, then 40 days later comes to Arbela 1.16.3 40 days after capture of Sepphoris & Arbela Herod sends Galilean army into winter quarters 38 B.C. 1.16.4 14.15.5 Herod subdues robbers in caves 38 B.C. 1.16.5 14.15.6 Herod destroys the siegeworks against multiple fortresses in Galilee 1.16.6 14.15.7 Ventidius has defeated Pacorus and driven Parthians from Syria summer 38 B.C. 1.17.1 14.15.10 corn at Jericho was seized (in the midst of summer; Wars 1.17.1) 38 B.C. 1.17.4 14.15.11 Herod marched to Jericho in haste winter 38 / 37 B.C. *** 1.17.6 14.15.12 Herod's march on Jerusalem hindered by depth of winter 37 B.C. 1.17.8 14.15.14 end of winter, siege of Jerusalem begins, 3rd year since Herod made King at Rome summer 37 B.C. 1.18.2 14.16.2 Herod with Sosius captures Jerusalem and Antigonus after 5 month siege (6 months, Wars 5.9.4) 14.16.4 M. Vipsanius L.f. Agrippa; L. Caninius L.f. Gallus were consuls in 37 B.C.; 185.3 Olympiad; Sivan is 3rd month by Nisan reckoned calendar; 27 years (inclusive) after Pompey captured Jerusalem in 63 B.C. * Herod did not take winter quarters upon his return to Ptolemais. Instead, Herod marched through Galilee increasing his army every day (Ant. 14.15.1; Wars 1.15.3), so Herod initially (in Nov and Dec) had no sizable army in need of winter quartering. Further, Herod's priority was to rescue his family left at Masada, but he had to first capture Joppa. Having done so, Herod then went on to Jerusalem (Ant. 14.15.2; Wars 1.15.4). But Herod couldn't breach the walls and instead was forced to go gather winter provisions at Jericho for Silo's soldiers, and Herod (with Roman soldiers) then had to defend those provisions at Jericho from attack by Antigonus' troops. Beginning November of 40 B.C. and marching through Galilee building an army, then taking Joppa, then rescuing his family at Masada, then proceeding to Jerusalem and attacking there, only to be defeated, all would take more than a month or two. It would take several months into 39 B.C., so it would seem Herod did not "winter over" in 40 / 39 B.C.
** Herod's building an army, capture of Joppa, rescue of his family at Masada, being repelled at the walls of Jerusalem, and subsequently defending from Antigonus' attack at Jericho those winter provisions gathered for Silo's soldiers, ostensibly ends in late 39 B.C., and so the winter quartering of Ant. 14.15.3-4 and Wars 1.15.6-16.3 would be for the winter of 39 / 38 B.C.
*** During 38 B.C., Herod campaigned throughout Samaria, and went to assist Antony at Samosata. After Antiochus surrendered to Antony, Antony sent Sosius and troops back with Herod who then campaigned against Antigonus thoughout Galilee, Samaria, and Judea, capturing five cities (Ant. 14.15.5-12) and would have gone on to Jerusalem "had it not been for the depth of winter" (Ant. 14.15.12); "he had been hindered by the depth of winter's [coming on]" (Wars 1 17.6). When winter ended Herod moved to Jerusalem and besieged it in 37 B.C.
See the highlighted "winter" entries at 40 - 37 B.C. in the chronologies. This analysis here adds some detail, but William Whiston originally noted as much in 1736 in a footnote to Ant. 14.16.4:
(30) Note here, that Josephus fully and frequently assures us that there passed above three years between Herod's first obtaining the kingdom at Rome, and his second obtaining it upon the taking of Jerusalem and death of Antigonus. The present history of this interval twice mentions the army going into winter quarters, which perhaps belonged to two several winters, ch. 15. sect. 3, 4; and though Josephus says nothing how long they lay in those quarters, yet does he give such an account of the long and studied delays of Ventidius, Silo, and Macheras, who were to see Herod settled in his new kingdom, but seem not to have had sufficient forces for that purpose, and were for certain all corrupted by Antigonus to make the longest delays possible, and gives us such particular accounts of the many great actions of Herod during the same interval, as fairly imply that interval, before Herod went to Samosata, to have been very considerable. However, what is wanting in Josephus, is fully supplied by Moses Chorenensis, the Armenian historian, in his history of that interval, B. II ch. 18., where he directly assures us that Tigranes, then king of Armenia, and the principal manager of this Parthian war, reigned two years after Herod was made king at Rome, and yet Antony did not hear of his death, in that very neighborhood, at Samosata, till he was come thither to besiege it; after which Herod brought him an army, which was three hundred and forty miles' march, and through a difficult country, full of enemies also, and joined with him in the siege of Samosata till that city was taken; then Herod and Sosins marched back with their large armies the same number of three hundred and forty miles; and when, in a little time, they sat down to besiege Jerusalem, they were not able to take it but by a siege of five months. All which put together, fully supplies what is wanting in Josephus, and secures the entire chronology of these times beyond contradiction.
Josephus, Antiquities of the Jews, Book 14, Chapter 16, para. 4, note 30
William Whiston, trans. (1736)Lastly, Josephus reports that Herod persuaded Antony to have Antigonus killed:
Wars 1.18.3 Hereupon Sosius dedicated a crown of gold to God, and then went away from Jerusalem, leading Antigonus away in bonds to Antony; then did the axe bring him to his end, who still had a fond desire of life, and some frigid hopes of it to the last, but by his cowardly behavior well deserved to die by it.
Josephus, Wars of the Jews, 1.18.3
William Whiston, trans. (1736)Ant. 14.16.4 Out of Herod's fear of this it was that he, by giving Antony a great deal of money, endeavored to persuade him to have Antigonus slain, which if it were once done, he should be free from that fear. And thus did the government of the Asamoneans cease, a hundred twenty and six years after it was first set up.
Josephus, Antiquities of the Jews, 14.16.4
William Whiston, trans. (1736)Ignoring Josephus' error of dating Herod's appointment as King to the 184th Olympiad, all other historical events and reports reconcile with Herod being made King of Judea in 40 B.C. and capturing Jerusalem and having Antigonus slain in 37 B.C., from which two events Josephus double-dated Herod's death in 4 B.C.
Herod captured Jerusalem from Antigonus in Sivan of 37 B.C.
In 37 B.C., Herod, with Sosius' support, besieged Jerusalem for 5 months (Wars 1.18.2) or 6 months (Wars 5.9.4) in the summer of a sabbatical year, afterwhich Antigonus surrendered to Sosius (Ant. 14.16.2). Herod captured both Jerusalem and the temple on the 3rd month, on the fast, ending 126 years of Hasmonean rule, on the same day as Pompey 27 years earlier (Ant. 14.16.4). Here are the pertinent passages from Josephus:
2 … Indeed, though they had so great an army lying round about them, they bore a siege of five months, till some of Herod's chosen men ventured to get upon the wall, and fell into the city, as did Sosius's centurions after them; … Then it was that Antigonus, without any regard to his former or to his present fortune, came down from the citadel, and fell at Sosius's feet, who without pitying him at all, upon the change of his condition, laughed at him beyond measure, and called him Antigona.
Wars 1.18.2
William Whiston, The Works of Josephus, trans. (1736)4. … When Herod, the son of Antipater, brought upon us Sosius, and Sosius brought upon us the Roman army, they were then encompassed and besieged for six months, till, as a punishment for their sins, they were taken, and the city was plundered by the enemy. …
Wars 5.9.4, ibid.
2. Now the Jews that were enclosed within the walls of the city fought against Herod with great alacrity and zeal (for the whole nation was gathered together); … for it was summer time,… so they brought their engines to bear, and shook the walls of the city, … for this happened to be a Sabbatic year. … for the first wall was taken in forty days, and the second in fifteen more, … And when the outer court of the temple and the lower city were taken, the Jews fled into the inner court of the temple, and into the upper city; … Herod … made an assault upon the city, and took it by storm; and now all parts were full of those that were slain, by the rage of the Romans at the long duration of the siege, and by the zeal of the Jews that were on Herod's side, who were not willing to leave one of their adversaries alive; … and then Antigonus, without regard to either his past or present circumstances, came down from the citadel, and fell down at the feet of Sosius, …
Ant. 14.16.2, ibid.
4. This destruction befell the city of Jerusalem when Marcus Agrippa and Caninius Gallus were consuls of Rome on the hundred eighty and fifth olympiad, on the third month, on the solemnity of the fast, as if a periodical revolution of calamities had returned since that which befell the Jews under Pompey; for the Jews were taken by him on the same day, and this was after twenty-seven years' time. … And thus did the government of the Asamoneans cease, a hundred twenty and six years after it was first set up. …
Ant. 14.16.4, ibid.
In 63 B.C., Pompey had received Jerusalem and the king's palace under surrender (Wars 1.7.2, Ant. 14.4.2). Within Jerusalem, Pompey began a siege of the temple (Wars 1.7.3), and battered the temple wherein Aristobulous' party had retreated (Ant. 14.4.2). During this siege the priests within the temple continued to twice daily offer sacrifices (Ant. 14.4.3). After 3 months, Pompey entered the temple (Wars 1.7.4) and the city (actually should read the temple) was taken on the 3rd month, day of the fast, and Pompey cut the throats of those within (Ant 14.4.3). The next day, Pompey ordered the temple cleansed and reinstated Hyrcanus II as high priest (Wars 1.7.6). Here are the pertinent passages from Josephus:
2. … So Aristobulus's party was worsted, and retired into the temple, … but as the others had received the Romans into the city, and had delivered up the palace to him, Pompey sent Piso, one of his great officers, into that palace with an army, who distributed a garrison about the city, because he could not persuade any one of those that had fled to the temple to come to terms of accommodation; he then disposed all things that were round about them so as might favor their attacks, as having Hyrcanus's party very ready to afford them both counsel and assistance.
Wars 1.7.2, ibid.
3. But Pompey himself filled up the ditch that was on the north side of the temple, and the entire valley also, the army itself being obliged to carry the materials for that purpose. …
Wars 1.7.3, ibid.
4. … Nor indeed when the temple was actually taken, and they were every day slain about the altar, did they leave off the instances of their Divine worship that were appointed by their law; for it was in the third month of the siege before the Romans could even with great difficulty overthrow one of the towers, and get into the temple. …
Wars 1.7.4, ibid.
6. … Moreover, he made Hyrcanus high priest, as one that not only in other respects had showed great alacrity, on his side, during the siege, but as he had been the means of hindering the multitude that was in the country from fighting for Aristobulus, which they were otherwise very ready to have done; …
Wars 1.7.6, ibid.
2. Now there was a sedition of the men that were within the city, who did not agree what was to be done in their present circumstances, while some thought it best to deliver up the city to Pompey; … Now these prevented the others, and seized upon the temple, … and prepared themselves to abide a siege; but the others admitted Pompey's army in, and delivered up both the city and the king's palace to him. … and fortified the houses that joined to the temple, … but Pompey pitched his camp within [the wall], on the north part of the temple, where it was most practicable; … he battered the temple with the stones that were thrown against it. …
Ant. 14.4.2, ibid.
3. … since the priests were not at all hindered from their sacred ministrations by their fear during this siege, but did still twice a-day, in the morning and about the ninth hour, offer their sacrifices on the altar; … for although the city was taken on the third month, on the day of the fast, (6) upon the hundred and seventy-ninth olympiad, when Caius Antonius and Marcus Tullius Cicero were consuls, and the enemy then fell upon them, and cut the throats of those that were in the temple; …
(6) That is, on the 23rd of Sivan, the annual fast for the defection and idolatry of Jeroboam, "who made Israel to sin;" or possibly some other fast might fall into that month, before and in the days of Josephus.
Ant. 14.4.3, ibid.
Note above that Ant. 14.4.3 records the city was taken by Pompey on the 3rd month, day of the fast, as well as the temple (Pompey cut the throats of those within in the temple). However, that contradicts the accounts of Wars 1.7.3 and Ant. 14.4.2 (wherein the city had surrendered to Pompey) and of Wars 1.7.2-6 which further describes the details of Pompey's 3-month siege of the temple. Further, during the siege, sacrifices were still being offered in the temple, ending on the 3rd month, day of the fast, when 'those within the temple had their throats cut'. The only narrative consistent with these details is that Jerusalem had earlier surrendered to Pompey who had then besieged the temple for 3 months, which was then "taken" on the 3rd month, day of the fast.
Accordingly, it would seem Josephus erred in the detail of Ant. 14.4.3 that Pompey took the temple (not the city) on the third month, on the day of the fast, following a 3-month siege of the temple, which then reconciles with Ant. 14.16.4 that 27 years to the same day (3rd month, day of the fast) following a 5-month siege of Jerusalem, Herod with Sosius captured the temple and Jerusalem together:
- Pompey took the temple in the 3rd month of the year on the day of the fast, after a 3-month siege of the temple (Jerusalem having previously surrendered)
- Herod took the temple (and Jerusalem) in the 3rd month of the year on the day of the fast, after a 5-month siege of Jerusalem
Josephus is drawing a parallel between Herod's capture of Jerusalem and the temple on the 3rd month, day of the fast, in 37 B.C. with the same day that Pompey captured the temple (i.e. 3rd month, day of the fast) in 63 B.C., 27 years earlier (26 full years and 3 months of the 27th year).
Because Josephus reports these captures as being on the "day of the fast" in the "3rd month" of the year, the 3rd month of a Tishri-reckoned year is Kislev which has neither fasts nor feasts, whereas the 3rd month of a Nisan-reckoned year is Sivan which has the feast/festival of Weeks or Shavuoth (Pentecost) on Sivan 6th, and a fast on the 23rd to commemorate the time when the Jews delayed in bringing the firstfruits to Jerusalem during the reign of king Jeroboam. Whiston also noted this fast on 23 Sivan in his footnote (6) to Ant. 14.4.3 (above).
With Herod's capture of Jerusalem fixed to the month of Sivan in 37 B.C., Josephus also synchronizes that to:
- the beginning of Hasmonean rule 126 years earlier (Ant. 14.16.4) which began with the Maccabee's defeat of Antiochus IV and the rededication of the Temple in year 148 of the Seleucid Era (1Macc 4:52) which equals 164 B.C.;
- Pompey's capture of the Temple on the same month (shown to be Sivan) 26 years earlier in 63 B.C.;
- Antigonus reign of 3 years and 3 months from his installation as king by the Parthians in 40 B.C., from which Herod fled to Rome and was appointed king;
- and Titus' destruction of the Herodian Temple 107 years later in A.D. 70. (Ant. 20.10.1, Wars 6.4.8);
What the table below demonstrates is the interlocked relationship between Josephus' chronology and historically dated events in the life of Herod the Great and his predecessors and successors. Josephus' chronology is accurate to within a couple months and intersects with fixed dates of events known to us from non-Josephean sources (e.g. Caesar, Dio Cassius, Appian, Seutonius, Plutarch, 1st Maccabees, etc.). Consequently, the date of Herod's death can not be moved more than a couple months without needing to also 're-write history or re-date the rest of Josephus'.
Judas Maccabeus rededicates Temple
127Pompey reinstates Hyrcanus II
27
24Rome makes Herod king
37Parthians install Antigonus over Hyrcanus II Herod kills Antigonus, captures Jerusalem
7th
14th
18th
34
107Battle of Actium in Herod's 7th year Herod aids governor of Egypt in 14th year Herod begins Temple in his 18th year
46Total lunar eclipses of Mar 23 and Sep 15 Herod's death Antipas reigned until Gaius' 2nd year
43Philip reigned until Tiberius' 22nd year
37Archelaus reigned 10 years
10Total lunar eclipse January 9 A.D. 6 Archelaus banished in 10th year A.D. 27 Pharisees claim 46 years to build Temple A.D. 34 Tiberius 20th/22nd year in Vitellius consulship A.D. 39 Gaius Caligula's 2nd year (latter half) A.D. 70 Titus destroys Temple Consider, for example, the constraints shown in the table above that prevent Herod's death from synchronizing in 1 B.C.:
Herod's reign, doubled-dated from killing Antigonus (col B) and appointed by Rome (col C) cannot be lengthened by 3 years:
- Archelaus' banishment (dated independently to A.D. 6) would be in his 7th year instead of 10th (col A)
- Philip reigned 37 years and would have died while Proculus and Nigrinus were consuls, instead of Vitellius (col B)
- Antipas reigned 43 years (dated by coins) and would have been deposed after Caligula's death instead of Caligula's 2nd year (col C)
Herod's reign, doubled-dated from killing Antigonus (col B) and appointed by Rome (col C) cannot be postponed by 3 years.
- Pompey's capture of Jerusalem, independently dated, would be 30 years, instead of 27, before Herod's capture of Jerusalem (col D)
- Herod's capture of Jersusalem would be 130 years after Maccabeus rededicated the Temple (dated 1st Macc.), instead of 127 (col A)
- Battle of Actium, which is independently dated, would then be in Herod's 4th year, instead of 7th (col B).
- Aelius Gallus governor of Egypt campaign against Sabos king of Arabia Felix, dated by Dio to Augustus' 10th Consulship, would be in Herod's 11th year, instead of 14th (col B).
- Herod's start on the Temple would be in his 15th year, instead of 18th (col B)
Note also the assertion by the Pharisees in John 2:20 that the [Herodian] Temple took 46 years to build (col D) reconciles with Herod beginning construction in his 18th year in 20 B.C. (col B) and Jesus attending Passover in A.D. 27 (following his baptism at "about age 30", early in Tiberius' 15th year which was A.D. 26).
Josephus' report in Wars. 1.21.1 that Herod began Temple reconstruction in his 15th year conflicts with Ant. 15.11.1 wherein Herod began reconstruction in his 18th year. Hoehner notes "15th year" is likely a scribal error or preparation for reconstruction.
- Titus' destruction of the Temple, independently dated, would only be 104 years after Herod's capture of Jersusalem, not 107 (col F).
Josephus' dating of Herod's death is not speculative and adjustable, but is fixed in relationship to verified history.
Josephus is our only source for the reigns of the high priests connected to Herod the Great, but these reigns lack dates, they are just sequences and durations of whole years, consequently they are little help in fixing the date of Herod's death.
As there are two eclipse candidates in 5 B.C., the best fit for identifying Herod's eclipse would seem to be the total lunar eclipse of 5 B.C. March 23, which preceeded by 1 day the Ta'anit Bechorim (Fast of the Firstborn) on 5 B.C. Mar 24 (Nisan 14, 3756 A.M.). A year later, Herod (who was already ill) falls gravely ill, pares and eats an apple, kills Antipater and dies 5 days later in early 4 B.C. before Passover. That dating allows ample time for all the events associated with Herod's discovery of Antipater's conspiracy, the messages exchanged with Rome, and then Herod's faltering health and death.
The greatest anomaly in the final months of Herod's life is: if Herod died prior to Nisan 14th in 4 B.C. (or even in 1 B.C. the same issue remains) but he pared and ate an apple that had ripened in Elul of 5 B.C. which apple was harvested in Tishri or blossomed late and was harvested not later than Heshvan of 5 B.C., then apples were somehow kept sufficiently firm for 5-7 months (without refrigeration or preservation) to be peeled with a knife by Herod just prior to his death in late WeAdar (2nd Adar) or early Nisan. But American pioneers are known to have had cool, dry "root cellars" in which properly selected apples (not too ripe, not harvested too early, ideally firm fleshed, tart and thick skinned), carefully placed so as to avoid bruising and crushing, will keep for a few months. Herod died at his winter palace in Jericho, but as a winter palace it was warmer in Jericho than generally elsewhere in Israel. Possibly, apples that ripened and were harvested late (and keep well) were carefully stored in a "root cellar" in colder climes over the winter, and were periodically brought to Jericho so that Herod might have a small, steady supply of firm apples into early spring.
Josephus on Herod's Eclipse
It remains to explain Josephus' phrasing in his report of a lunar eclipse prior to the death of Herod the Great.
For this analysis, the translation by Ralph Marcus from the Loeb Classical Library will be cited because the grammar and inferred punctuation is important, and Daniel Schwartz also cites Marcus' translation on some aspects to be emphasized.
Josephus sets the context in Ant. 17.6.2-3 that Judas, the son of Saripheus, and Matthias, the son of Margalothus incite youths to tear down the golden eagle that Herod had erected over a Temple. They are arrested and Herod sent them to be imprisoned in Jericho.
Then in Ant. 17.6.4, Josephus relates what Herod did to the rebels and the High Priest Matthias whom Herod seemed to think was complicit in the rebellion (below, explanatory notes [in square brackets] have been inserted into Marcus' translation):
Because of his [Herod's] savage state and out of fear that in his fury he might avenge himself upon them, those present [the principal men among the Jews whom Herod called to the theater in Jerusalem (Ant. 17.6.3)] said that these things had been done without their consent, and it seemed to them that the perpetrators should not be exempted from punishment. Herod therefore dealt rather mildly with these others but removed the high priest Matthias from his priestly office as being partly to blame for what had happened, and in his stead he [Herod] appointed his [Matthias'] wife’s brother Joazar as high priest. Now it happened during this Matthias’ term as high priest that another high priest was appointed for a single day - that which the Jews observe as a fast - for the following reason. While serving as priest during the night preceding the day on which the fast occurred, Matthias seemed in a dream to have intercourse with a woman, and since he was unable to serve as a priest because of that experience, a relative of his, Joseph, the son of Ellemus, served as priest in his place. Herod then deposed Matthias from the high priesthood. As for the other Matthias [son of Margalothus], who had stirred up the sedition, he [Herod] burnt him alive along with some of his companions. And on that same night there was an eclipse of the moon.
Josephus, Antiquities of the Jews 17.164-167,
Loeb Classical Library (1963) trans. Ralph MarcusFive different chronologies are possible depending on how the grammar of Josephus' account of Herod's eclipse is interpreted, whereupon the:
- night before the fast is the night of the eclipse, but Herod both deposed High Priest Matthias (son of Theophilus) and burned alive Matthias son of Margalothus on later events.
- night of the fast is the night of the eclipse, after Herod both deposed High Priest Matthias (son of Theophilus) and burned alive Matthias son of Margalothus earlier that same day.
- day of the fast, Herod deposed High Priest Matthias (son of Theophilus) and the eclipse was that same night, but Herod burned alive Matthias son of Margalothus on a later date.
- Unassociated with the fast and deposing high priest Matthias, the night of the eclipse followed when Herod burned alive Matthias son of Margalothus.
- Unassociated with the fast, the night of the eclipse followed when Herod both deposed High Priest Matthias son of Theophilus and also burned alive Matthias son of Margalothus.
However, the reconciliation of Josephus' Herodian chronologies (reckoned by Julian, Nisan and Tishri calendars) evidences three constraints which eliminate 3 of the 4 grammatically plausible interpretations, those constraints are:
- Josephus' chronology is reconciled with Herod's death in 4 B.C. but cannot be reconciled with Herod's death in 1 B.C.
- When Herod burned alive Matthias son of Margalothus and his co-conspirators, it was at Jericho while Herod was there in winter residence. This constraint eliminates any interpretation where Matthias son of Margalothus and his co-conspirators are burned alive followed by either lunar eclipse of 5 B.C. because neither of those eclipses occurred in winter.
- There are no Jewish fasts proximate to January when the lunar eclipse of 1 B.C. occurred. This constraint eliminates any interpretation where the lunar eclipse occurs on the night of a Jewish fast in January.
Following are Josephus' description of the events associated with "Herod's Eclipse", with the five different grammatical interpretations portrayed visually, with:
- ancillary accounts indented as parenthetical expressions; and,
- eclipse coincident events highlighted in blue:
- The eclipse occurs on the night before the fast, apart from Herod's actions:
Now it happened during this Matthias’ term as high priest that another high priest was appointed for a single day - that which the Jews observe as a fast - for the following reason. While serving as priest during the night preceding the day on which the fast occurred, Matthias seemed in a dream to have intercourse with a woman, and since he was unable to serve as a priest because of that experience, a relative of his, Joseph, the son of Ellemus, served as priest in his place.
(Herod then deposed Matthias from the high priesthood. As for the other Matthias [son of Margalothus], who had stirred up the sedition, he [Herod] burnt him alive along with some of his companions.)And on that same night there was an eclipse of the moon.Josephus, Antiquities of the Jews 17.165-167,
Loeb Classical Library (1963) trans. Ralph MarcusThis is the grammatical interpretation wherein Josephus relates Herod's punishment of the two Matthias's seemingly as if it were a parenthetical explanation not contemporaneous with either the fast or eclipse (note, the night of the eclipse and the night of the dream are the same night and before the fast).
Above, the 'night before the fast and the eclipse' occur as part of the same event. Herod's deposing of Matthias (son of Theophilus) and burning the other Matthias are separate incidents (essentially, in parenthetical notes which are merely ancillary to Josephus' reporting of Matthias (son of Theophilus) recusing himself from High Priestly duties and the eclipse omen which are Josephus' main subject).
This interpretation reconciles both grammatically and chronologically.
Grammatically, this interpretation is supported by Brann, Schwartz, and Vanderkam.
… However, Schwartz correctly notes that the eclipse is connected, not with Herods death, but with the high priest Matthias’s nocturnal defilement and is thus the one that took place in September 15-16 of 5 bce. The Day of Atonement would have fallen on September 12-13 of that year and is therefore the fast mentioned by Josephus (Daniel R. Schwartz, “Joseph ben Illem and the Date of Herod’s Death,” in his Studies in the Jewish Background of Christianity; WUNT 60 [Tubingen: Mohr/Siebeck, 1992], 157-66).James VanderKam, "29. Matthias Son of Theophilus (5-4 BCE)", n. p. 410
From Joshua to Caiaphas: High Priests after the Exile, Fortress (2004)Daniel Schwartz explains the grammatical antecedence that "same night" refers to "night" before the fast:
[Ant. 17.6.4] is usually taken to mean that there was a lunar eclipse on the night following the firing of Matthias ben Theophilus and the execution of Matthias the sage and his companions. As the insurrection occurred only when the king was quite ill (Ant. 17.148,150,161,168 ff.), this understanding of B leads to the conclusion that the eclipse occurred not long before the king died, that is, in March and not in the preceding September. However, it appears preferable to assume that Josephus means there was an eclipse the night Matthias ben Theophilus dreamed his fateful dream - as was argued over a century ago in a little-noticed Latin dissertation by Marcus Brann. For:
1. The words “that night” refer most naturally to the only antecedent “night” in this story - the one when Matthias dreamed.
2. In the parallel narrative in BJ 1.648-2.9 we find virtually all of what we listed in sections A, C, D, E, and the burning of the insurgents is also described, in BJ 1.665, very similarly to its description at the beginning and end of section B (§§ 164, 167). That is, only the matters pertaining to the high priest (his removal from office and the erotic anecdote) and the notice concerning the eclipse “that same night” have been added. Now we know, from the parallels in rabbinic literature, that the erotic anecdote circulated independently, and it is reasonable to assume that Josephus received it from some such source, probably a written one. His extract from this source is introduced with the conventional language, “for the following reason,” but its end is unclear. From the methodological point of view, it is much more reasonable to assume that the source on the high priesthood supplied the notice of the eclipse as well, than to assume input from two new sources. But if the eclipse notice was found with the dream story, then it applies to it.
3. Josephus never says that the firing of the high priest and the execution of the insurgents occurred on the same day, but if he meant the eclipse occurred during the night following those events the reader will have to make that assumption. Otherwise, he will have to ask “Which night?” But that assumption asks quite a lot of the bare words “that night.” As a rule, Josephus, as other writers, must be assumed to have told us enough to be able to make sense of his narrative.
4. Indeed, if one applies the words “that night” to the various punitive actions reported in this paragraph, then he should also infer that they were carried out at night - yet another detail we would be expected to infer from between the lines. (Usually, of course, it is supposed that the reference is to the night following the punitive acts. But when Josephus refers to the night following an aforementioned day [English “that night”], he normally uses either the plain nux [“night”] or else specifies by using ekeine [“that night”] or even epiousa [“the succeeding night”]; but the insistent use of autē, as here, seems rather to mean “in the same [= aforementioned] night.” Never in Antiquities, and only once anywhere [BJ 4.640], did I find Josephus using autē with regard to the night following an aforementioned day.)
We conclude, therefore, on the basis of these four considerations, that Josephus mentioned the eclipse only apropos of the story of Matthias’ dream, which, in turn, he told only as an interesting episode which occurred sometime during that high-priest’s tenure. This is, basically, an aspect of Josephus’ usual practice, of reporting sundry secondary matters before leaving his account of a given individual.
Daniel R. Schwartz, “Joseph ben Illem and the Date of Herod’s Death”,
Studies in the Jewish Background of Christianity; J.C.B. Mohr (1992) pp. 161-163VanderKam, Schwartz, and Brann also offer plausible explanations as to why the date of the eclipse may differ a few days from the date of the fast, due to how the Jewish festival dates were determined. Both VanderKam and Schwartz assume, as is customary and reasonable, that the fast to which Josephus referred was the Day of Atonement (Yom Kippur) on Tishri 10.
However, chronologically, the eclipse of March 23, 5 B.C. can be shown to occur 1 day before the Ta’anit Bechorim (Fast of the Firstborn) on Nisan 14, assuming the Jews determined Nisan 1 in 5 B.C. by observation of the new moon (weather conditions permitting). If instead, Nisan 1 in 5 B.C. was determined by intercalary computation, then the fast would have been observed 1 day before the eclipse.
Jewish fasts that are historically attested and observed, and proximate to the eclipses are:
- Ta’anit Bechorim fast on Nisan 14 (by new moon observation) was 1 day after the total eclipse of March 23, 5 BC
- Ta’anit Bechorim fast on Nisan 14 (by intercalated calendar) was 1 day before the total eclipse of March 23, 5 BC
- Yom Kippur fast on Tishri 10 (by new moon observation) was 2 days before the total eclipse of September 15, 5 BC.
- Yom Kippur fast on Tishri 10 (by intercalated calendar) was 4 days before the total eclipse of September 15, 5 BC.
Regardless, a minor discrepancy with both 5 B.C. eclipses is that their durations are early enough in the evening that it seems unlikely Matthias was actually asleep and dreaming before an eclipse occurred:
- Mar 23, 5 B.C. eclipse from 6:30 P.M. - 10:12 P.M. Jerusalem time
- Sep 15, 5 B.C. eclipse from 8:23 P.M. - midnight Jerusalem time
While the detailed cause-and-effect sequence of Matthias' defilement and recusement, and the lunar eclipse are unknown, it remains plausible that the March 23, 5 B.C. eclipse occurred 1 day before the Ta’anit Bechorim fast was observed on Nisan 14, 5 B.C.
Though incorrectly sequenced, but still plausibly close depending on how Nisan 1 was determined, the eclipse of September 23, 5 B.C. falls 2-4 days after the Yom Kippur fast.
Lastly, both 5 B.C. eclipses in this interpretation reconcile well with the Chronology of the Herod's in which Herod the Great discovers and prosecutes Antipater's conspiracies, punishes the rebels who pulled down the golden eagle, falls ill, and dies in March of 4 B.C. just before Passover.
While the eclipse of Jan 9, 1 B.C. is precluded from this interpretation because there are no attested fasts proximate to Jan 9, since the nearest would be 1 month earlier (the Asarah B’Tevet) or 1 month later (except the Ta’anit Esther is not attested prior to the medieval period), the duration of that eclipse was 11:22 P.M. – 2:56 A.M., which is easily at a time when Matthias would have been asleep and dreaming.
It is further plausible in that the eclipse is dissociated from Herod's deposing of Matthias (seemingly months later) and killing of the others (which took place in winter at Jericho).
- The eclipse occurs on the night of the fast, following Herod punishing both Matthias':
Now it happened during this Matthias’ term as high priest that another high priest was appointed for a single day - that which the Jews observe as a fast - for the following reason. While serving as priest during the night preceding the day on which the fast occurred, Matthias seemed in a dream to have intercourse with a woman, and since he was unable to serve as a priest because of that experience, a relative of his, Joseph, the son of Ellemus, served as priest in his place.(Herod then deposed Matthias from the high priesthood. As for the other Matthias [son of Margalothus], who had stirred up the sedition, he [Herod] burnt him alive along with some of his companions. And on that same night there was an eclipse of the moon.)Josephus, Antiquities of the Jews 17.165-167,
Loeb Classical Library (1963) trans. Ralph MarcusThis is the grammatical interpretation wherein Josephus relates Herod's punishment of both Matthias's and conspirators seemingly as if it were a parenthetical explanation, but everything happening on the day of the fast followed by the eclipse that same night.
This interpretation is not plausible for any of the total lunar eclipses:
- the eclipse of Jan 9, 1 B.C. is precluded because there are no proximate attested fasts.
- both eclipses of 5 B.C. are precluded because they occurred in spring and summer (respectively) during which time Herod was unlikely to be at his winter palace in Jericho where the rebels were held and ostensibly burned alive.
- The eclipse occurs on the night of the fast, following Herod deposing Matthias:
Now it happened during this Matthias’ term as high priest that another high priest was appointed for a single day - that which the Jews observe as a fast - for the following reason. While serving as priest during the night preceding the day on which the fast occurred, Matthias seemed in a dream to have intercourse with a woman, and since he was unable to serve as a priest because of that experience, a relative of his, Joseph, the son of Ellemus, served as priest in his place. Herod then deposed Matthias from the high priesthood.(As for the other Matthias [son of Margalothus], who had stirred up the sedition, he [Herod] burnt him alive along with some of his companions.)And on that same night there was an eclipse of the moon.
ibid.
This is the grammatical interpretation wherein Josephus relates Herod's deposing high priest Matthias (son of Theophilus) on the day of the fast with the eclipse following that night, and seemingly as if it were a parenthetical explanation an ancillary account of burning alive Matthias son of Margalothus and co-conspirators sometime later (note, only Herod deposing high priest Matthias is grammatically antecedent to the night of the eclipse).
As with interpretation 1) this interpretation is chronologically possible only for the lunar eclipses of 5 B.C. which both occur proximate to an attested fast. It essentially is no different than interpretation 1) except that it also presumes that Herod deposed the high priest on the same day as the fast, which action would seem to conflict with the solemnity the Jews expect whom Herod ostensibly would not want to unnecessarily aggravate.
This interpretation is impossible for the eclipse of Jan 9, 1 B.C. as there are no attested fasts proximate to the month of January.
- The eclipse occurs apart from the fast and deposition, following Herod burning the rebels:
Now it happened during this Matthias’ term as high priest that another high priest was appointed for a single day - that which the Jews observe as a fast - for the following reason. While serving as priest during the night preceding the day on which the fast occurred, Matthias seemed in a dream to have intercourse with a woman, and since he was unable to serve as a priest because of that experience, a relative of his, Joseph, the son of Ellemus, served as priest in his place. Herod then deposed Matthias from the high priesthood.
(As for the other Matthias [son of Margalothus], who had stirred up the sedition, he [Herod] burnt him alive along with some of his companion. And on that same night there was an eclipse of the moon.)ibid.
This is the grammatical interpretation wherein the 'night before the fast', the fast itself, and Herod's deposing of high priest Matthias are earlier separate events from Josephus (seemingly as a parenthetical expression) relating an ancillary account of an eclipse following Herod's burning alive of Matthias [son of Margalothus] and his rebel companions (which are antecedent to the night of the eclipse).
This is the interpretation presumed by most scholars, including Steinmann. For the purposes of this analysis, treating Herod's deposing of high priest Matthias as an isolated intermediate event gains some chronological flexibility but otherwise does not clarify when the eclipse occurred.
While this interpretation is chronologically plausible for the limited event of the eclipse of Jan 9, 1 B.C. because it occurred in winter when Herod would have been at his winter palace in Jericho, where the rebels were held and ostensibly burned alive. Regardless, Herod's death following this eclipse in 1 B.C. cannot be reconciled with the rest of the Chronology of the Herod's, because to date Herod's death in 1 B.C. requires dismissing of all of Josephus' dates for the lives of the Herods. While Josephus may have gotten a few minor details wrong (such as Herod's age when he was made governor of Galilee), the chronology demonstrates that Josephus recorded everything consistently with his own Herodian and Jewish histories and consistently with the historical dates of others.
Further, this interpretation is not plausible:
- grammatically for any eclipse, because the only antecedent "night" for "that same night of the eclipse" is the night Matthias dreamed before a fast.
- chronologically for either of the total lunar eclipses of 5 B.C., because they occurred in spring and summer (respectively) during which time Herod was unlikely to be at his winter palace in Jericho (where the rebels were held and ostensibly burned alive).
- The eclipse occurs apart from the fast, following Herod punishing both Matthias':
Now it happened during this Matthias’ term as high priest that another high priest was appointed for a single day - that which the Jews observe as a fast - for the following reason. While serving as priest during the night preceding the day on which the fast occurred, Matthias seemed in a dream to have intercourse with a woman, and since he was unable to serve as a priest because of that experience, a relative of his, Joseph, the son of Ellemus, served as priest in his place.(Herod then deposed Matthias from the high priesthood. As for the other Matthias [son of Margalothus], who had stirred up the sedition, he [Herod] burnt him alive along with some of his companions. And on that same night there was an eclipse of the moon.)ibid.
This is the grammatical interpretation wherein the 'night before the fast' and the fast itself are an earlier separate event from Josephus (seemingly as a parenthetical expression) relating an ancillary account of an eclipse following Herod's deposing Matthias' [son of Theophilus] and burning alive of Matthias [son of Margalothus] and his rebel companions (which are antecedent to the night of the eclipse).
Herod could depose high priest Matthias at any time (other than on fast or sabbath days) from any location, and accordingly the timing of Matthias' deposition doesn't help distinguish among eclipses. Consequently, including the deposition of Matthias on the same night as burning the rebels alive yields the same problems as scenario 4).
As with interpretation 4) this interpretation is chronologically plausible only for the limited event of the eclipse of Jan 9, 1 B.C. because it occurred in winter when Herod would have been at his winter palace in Jericho, where the rebels were held and ostensibly burned alive. Regardless, Herod's death following this eclipse in 1 B.C. cannot be reconciled with the rest of the Chronology of the Herod's, because to date Herod's death in 1 B.C. requires dismissing of all of Josephus' dates for the lives of the Herods. While Josephus may have gotten a few minor details wrong (such as Herod's age when he was made governor of Galilee), the chronology demonstrates that Josephus recorded everything consistently with his own Herodian and Jewish histories and consistently with the historical dates of others.
Lasttly, neither is this interpretation plausible:
- grammatically for any eclipse, because the only antecedent "night" for "that same night of the eclipse" is the night Matthias dreamed before a fast.
- chronologically for either of the total lunar eclipses of 5 B.C., because they occurred in spring and summer (respectively) during which time Herod was unlikely to be at his winter palace in Jericho (where the rebels were held and ostensibly burned alive).
Here is a table that summarizes the relative strengths, weaknesses, and conflicts of each of the eclipse dates:
Ta'anit Bechorim Yom Kippur none* same night by new moon 2-4 day mismatch 5 BC Mar 22 = Nisan 14 5 BC Sep 11 = Tishri 10 none* 5 BC Mar 24 = Nisan 14 5 BC Sep 13 = Tishri 10 none* doesn't matter doesn't matter doesn't matter later in winter at Jericho** later in winter at Jericho** coincident with eclipse** 363 187 68*** 4 BC Mar 21 4 BC Mar 21 1 BC Mar 17 21 21 21 4 BC Apr 11 4 BC Apr 11 1 BC Apr 7 yes yes no * The Fast of Esther on Adar 13 would fall on 1 B.C. Feb 7, but Daniel Schwartz reports this fast was "not in evidence prior to the medieval period"
Daniel R. Schwartz, “Joseph ben Illem and the Date of Herod’s Death”,
Studies in the Jewish Background of Christianity; J.C.B. Mohr (1992) pp. 161-163** The 'burning alive of the rebellion leaders' occurred at Jericho where Herod had earlier sent them following their capture (Ant. 17.6.5), which burning occurred after Herod had moved to his winter palace at Jericho. However, during the winter months of 5/4 B.C. there were no lunar eclipses. 1 B.C. Jan 9 is the only "winter" total lunar eclipse for the 10 years between 5 B.C. and A.D. 5.
*** Steinman (and most scholars) assume the eclipse is associated with Herod deposing Matthias and burning the rebels. However, Josephus reports that subsequent to the eclipse, Herod discovers Antipater's conspiracy and requests permission from Rome to execute Antipater, which request and response would take a minimum of 98-112 days between the eclipse and Herod's death.
**** assuming Herod died approximately 21 days before Passover as estimated by Steinmann and also assuming Herod died following the 1 B.C. Jan 9 eclipse.
Andrew E. Steinmann, "When did Herod the Great reign?", p. 16
Novum testamentum, vol 51 (2009) pp. 1-29Moreover, Hebrew year 3757 A.M. (approximately 4 B.C.) was an embolismic complete year. An extra month, Adar II, was added at the end of the Hebrew calendar for year 3757 A.M. which effectively pushed the next Passover out another month giving additional time in early 4 B.C. (because the intercalated month fell between February and April on the Julian calendar) for Herod to die in early 4 B.C., before the Passover at which Archelaus had to quell an uprising consequent to Herod's death, also as reported by Josephus (Wars 2.1.1-3; Ant. 17.9.3)
Some researchers argue that 363 days (or 187 days following the eclipse of 5 B.C. Sep 15) is too long an interval for the final events of Herod's life. For example, Andrew Steinmann argued against the 4 B.C. view of Herod's death, advocating instead a 1 B.C. date (already shown above it can't be synchronized):
Table 2. Tally of Days Elapsed Between the Eclipse and Passover
....
Thus, at least forty-one days are needed for the events between the eclipse and the Passover. A more likely tally that would take into account reasonable medical practice by Herod’s physicians would add at least about three weeks, bringing the tally to sixty-two days minimum. This fits well within the ninety-two days between the 1 BCE eclipse and the Passover allowing for a less compressed schedule than the absolute minimum assumed above. But even at the very rushed pace assumed for the events in the tally of forty-one days, the twenty-nine days between the 4 BCE eclipse and the Passover is inadequate. [pp. 15-16]Andrew E. Steinmann, "When did Herod the Great reign?"
Novum testamentum vol 51 (2009) pp. 1-29But Steinmann errs conceptually on two points:
- In arguing that "twenty-nine days between the 4 BCE eclipse and the Passover is inadequate", Steinmann is reaffirming the "Schürer consensus which holds the eclipse to be the one of March 13, 4 bce" [ibid. p. 13] having dismissed the validity of the 5 B.C. March 23 eclipse on the basis of an erroneous dating of Passover falling 29 days later. However Passover in 5 B.C. falls on March 22 or 24 (depending on how the Jewish Temple priests determined its date). Because the 1 or 2 days following March 23 is obviously too brief for all the events of Herod's final months, the Passover of 5 B.C. is disqualified as following Herod's death. Accordingly, the Passover on 4 B.C. April 11 is the next closest that qualifies, before which are 363 days in which all the final events of Herod's life will fit.
- In determining that at least 41 days (or or more likely 62) are needed between the eclipse and Passover, Steinmann has overlooked the turnaround delays in the ongoing communications between Herod and Caesar Augustus regarding the conspiracy of Antipater. The permission Herod receives to execute Antipater was requested by Herod after the eclipse, and the courier travel to Rome and back would take a minimum of 98-112 days. That correspondence alone exceeds the 92 days Steinmann allows between the 1 B.C. Jan 9 eclipse and Passover of 1 B.C. Apr 7. [ibid. p. 12]
Jewish Fasts proximate to Herod's Eclipse
There are Seven Fast Days of the Jewish Year. Here is a table showing their proximity to the selected visible lunar eclipses:
Ta’anit Esther 3756 Adar 13 -5 February 21 -5 February 21 Ta’anit Bechorim 3756 Nisan 14 -5 March 22 -5 March 22 Tzom Tammuz 3756 Tammuz 17 -5 June 22 -5 June 22 Tishah B’Av 3756 Av 9 -5 July 13 -5 July 13 Tzom Gedaliah 3757 Tishri 3 -5 September 4 -5 September 4 Yom Kippur 3757 Tishri 10 -5 September 11 -5 September 11 Asarah B’Tevet 3757 Tevet 10 -5 December 10 -5 December 10 Ta’anit Esther 3757 Adar I 13 -4 February 10 Ta’anit Bechorim 3757 Nisan 14 -4 April 11 Tzom Tammuz 3757 Tammuz 17 -4 July 12 Tishah B’Av 3757 Av 9 -4 August 2 Julian dates in the above table were converted from the Jewish fast dates using the Fourmilab Calendar Converter. Note that while Jewish fasts are normally ordered by the Hebrew sacred calendar (running from Nisan to Adar) the converter increments the Hebrew year at Tishri because it computes dates on the Hebrew civil calendar running from Tishri to Elul).
As was noted above, Schwartz reports the Fast of Esther on Adar 13 is disqualified from consideration because it was "not in evidence prior to the medieval period"
Daniel R. Schwartz, “Joseph ben Illem and the Date of Herod’s Death”,
Studies in the Jewish Background of Christianity; J.C.B. Mohr (1992) pp. 161-163Consequently, the only two qualified fasts in consideration are the Ta’anit Bechorim and Yom Kippur.
Following are two tables that demonstrate the chronological relationship mentioned by Josephus in Ant. 17.6.4 of a Jewish fast preceeded by a total lunar eclipse, sometimes known as "Herod's eclipse". Only lunar eclipses viewed from Jerusalem which are "total" (because of their notoriety) and are proximate to a Jewish fast (because that is what Josephus described) are of interest.
To ensure that all such eclipses are considered, a lunar eclipse calculator is used to generate a list of candidate eclipses. Using the Javascript Lunar Eclipse Explorer and selecting the city "Jerusalem, ISRAEL", "all eclipses" and clicking on century button "-0099 to -0000" generates a table of all lunar eclipses visible to Jerusalem from 100 to 1 B.C., and from that list, lunar eclipses in the last 10 years of the millenium (during which Herod died) were excerpted:
Lunar Eclipses from -0099 to 0000 Jerusalem, ISRAEL Latitude: 31° 46' 00" N Longitude: 35° 14' 00" E Altitude: 808.9m Time Zone: 02:00 E ---------- Date --------- --------------------------- Eclipse ---------------------------- Julian Hebrew Type Partial Alt Total Alt Mid. Alt Total Alt Partial Alt B.C. A.M. yt Begins Begins Ends Ends 10 Jan-19 3751 EC She 15 N - - - - 02:33 +53 - - - - 10 Jun-15 3751 EC Siv 14 N - - - - 23:57 +35 - - - - 10 Dec-10 3752 CR Kis 15 P 05:25 +12 - - 06:47 -03 - - 08:09 -17 9 Jun-03 3752 CR Siv 14 T 23:15 +36 00:39 +33 01:01 +32 01:24 +29 02:48 +19 9 Nov-28 3753 CC Kis 14 T 19:02 +30 20:01 +42 20:51 +52 21:41 +63 22:39 +74 8 Nov-18 3754 ED Kis 15 P 03:50 +29 - - 05:04 +14 - - 06:19 +00 7 May-13 3754 ED Iya 14 N - - - - 21:15 +29 - - - - 7 Nov-07 3755 CR Hes 15 N - - - - 06:15 -02 - - - - 6 Apr-04 3755 CR Nis 16 P 05:41 -03 - - 06:58 -19 - - 08:15 -34 5 Mar-23 3756 CC Nis 15 T 18:30 +09 19:30 +21 20:21 +31 21:12 +40 22:12 +49 5 Sep-15 3757 EC Tis 14 T 20:23 +32 21:22 +42 22:12 +49 23:01 +53 00:00 +54 4 Mar-13 3757 EC Ad2 14 P 01:32 +52 - - 02:41 +40 - - 03:50 +26 3 Mar-02 3758 CD Ada 15 N - - - - 03:06 +38 - - - - 3 Jul-27 3758 CD Av 15 N - - - - 20:45 +21 - - - - 3 Aug-26 3758 CD Elu 15 N - - - - 05:33 -06 - - - - 2 Jul-17 3759 CR Av 16 P 05:38 -13 - - 07:14 -31 - - 08:49 -51 1 Jan-09 3760 EC She 14 T 23:22 +79 00:20 +78 01:09 +70 01:58 +60 02:56 +48 1 Dec-29 3761 CC Tev 13 P 15:17 -16 - - 16:32 -03 - - 17:48 +11Fred Espenak, "Javascript Lunar Eclipse Explorer"
NASA Goddard Space Flight Center, 2007 May 31The entries above from that generated table have modified from what Espenak's generator produces:
- columns for penumbral eclipse start and end times have been removed (because penumbral eclipses were not visible to ancient observers without telescopes), and;
- Julian dates have been adjusted from negative years to years B.C., and;
- using the Fourmilab Calendar Converter, those Julian dates have been converted to Hebrew dates and inserted in their own column.
- lastly, eclipse rows for years earlier than 10 B.C. were removed, and still showing all eclipses for 10 B.C. to 1 B.C.
The two total lunar eclipses on 5 B.C. are highlighted above and analyzed in the tables below. The one total lunar eclipse on 1 B.C. is also highlighted above, but an analysis of its proximity to a Jewish fast has not been done because no Jewish fast is attested in the months of Shevat and Adar prior to the medieval period.
The two lunar eclipses that best fit Josephus' criteria are those of:
- March 23, 5 B.C. occurring the night after the fast of Ta’anit Bechorim on Nisan 14th (the date fits closest but not the characteristics of the fast)
- September 15, 5 B.C. occurring four nights after the fast of Yom Kippur on Tishri 10th (the characteristics of the fast fits closest but not the date)
Both eclipses are total, entirely visible from Jerusalem, and would be unmistakable assuming clouds did not obscure the night sky. Both eclipses are in 5 B.C. allow several months for Herod the Great to grow ill, kill his son Antipater, die, and then for his other son Archelaus to be named king and proclaim Herod's mourning, all prior to the next Passover on April 12, 4 B.C.
During this time, there was considerable political strife among the various Jewish sects (Pharisees, Saducees, and Essenes) over control of the calendar and festival observance, and dating methods varied depending on which group controlled the Temple and the Sanhedrin. While it is known that the Jews used a computed, intercalated calender by A.D. 66, it is not known for certain that they used it regularly in 5 B.C., though there is some evidence. The Temple priests and the Sanhedrin could have determined when was the first day of a month by either by calendrical calculation or by observation of the new moon lunar crescent. They would do this every month, but only the months of Nisan and Tishri of 5 B.C. are analyzed here because those are months in which both a total lunar eclipse and a Jewish fast occurred.
As determined above in interpretation 1) Josephus reported that a lunar eclipse occurred the night before a Jewish fast was observed. The date when the fast was observed must be determined and compared to the date of the eclipse. The date of the fast depends on the date of the first day of the month, which depends on sighting the crescent of the new moon, assuming the computed calendar was either not known or was ignored (for political reasons).
Below, are the "astronomical new moon" dates (determined by Fred Espenak of NASA) when the moon is perfectly aligned in the Earth's shadow and thus too dark to see without any reflected sunlight striking it. Within 1-2 days, the as the new moon ages, it begins to move out of the Earth's shadow and sunlight can strike its edge creating the very slim "crescent" appearance. For an experienced observer without any equipment, crescent new moon visibility is between one and two days after the "astronomical new moon", assuming skies are cloudless.
In each table below, for the Ta’anit Bechorim and Yom Kippur fasts respectively, the analysis tabulates the Julian date of the "astronomical new moon" as computed by NASA. That Julian date is converted to Hebrew and tabulated, to which two days are added (to account for the delay before the crescent new moon is visible) and that date is assumed to be the first of the month by obervation and tabulated. Then relative to that first day, the day of the fast is dated and tabulated and is designated "fast by observation of visible new moon". The dates of the first of the month and the fast as determined by the computed calendar are tabulated as well.
Ultimately, while it can't be ensured which method or on which day the Jews determined to observe either Ta’anit Bechorim (Fast of the Firstborn) or the Yom Kippur fast, the possibilities can be shown, and the Ta’anit Bechorim in 5 B.C., when dated by new moon observation, falls exactly as Josephus described in Antiquities 17.6.4 one day before the fast.
In the two tables below, the Hebrew days are numbered as if they used their computed calendar followed by a "/" and the day of the month by new moon observation. Also highlighted are the:
- astronomical new moon and the total lunar eclipse
- 1st day of the month and the fast, where dated by the computed calendar
- new moon visibility (which could be the 1st day of the month) and the fast, where dated by observation
- possible dates when High Priest Matthias would have recused himself from officiating the fast (Ant. 17.6.4)
The following table shows the proximity of the lunar eclipse of 5 B.C. March 23, to the Ta’anit Bechorim (Fast of the Firstborn) which is observed on Nisan 14. Note that if Nisan 1 was determined by observation of the new moon, then Nisan 14 and the fast would actually fall on Mar 24 which is the day after the lunar eclipse, as Josephus described in Ant. 17.6.4.
Comment March 5 BC 1 Adar 3756 22 2 23 3 24 4 25 5 26 6 27 7 28 8 Wed 29 Astronomical new moon 4:44 UT/GMT 9 Thu Nisan 3756 1 Nisan 1 (by computed calendar) 10 2 new moon visible 11 3 / 1 Nisan 1 by observation of new moon 12 4 / 2 13 5 / 3 14 6 / 4 15 7 / 5 16 8 / 6 17 9 / 7 18 10 / 8 19 11 / 9 20 12 / 10 21 13 / 11 Matthias recuses himself from officiating the fast by computed calendar (Ant. 17.6.4) 22 Wed 14 / 12 Ta'anit Bechorim (Fast of the Firstborn) - Tractate Soferim 21:3 (by computed calendar) 23 Thu 15 / 13 1st of Unleavened Bread by computed calendar
Total lunar eclipse & full moon;
Matthias recuses himself from officiating the fast by observation of visible new moon (Ant. 17.6.4)
24 16 / 14 Ta'anit Bechorim (Fast of the Firstborn) - Tractate Soferim 21:3 (by observation of visible new moon) 25 17 26 18 27 19 28 20 29 21 30 22 31 23 April 5 BC 1 24 2 25 3 26 4 27 5 28 6 29 7 30 8 Iyar 3756 1 The following table shows the proximity of the lunar eclipse of 5 B.C. September 15, to Yom Kippur (Day of Atonement fast) which is observed on Tishri 10. Note below that if the new moon were observed on 5 B.C. Septermber 1, then Yom Kippur still would be 2 days before the eclipse (not 1 day after), and though the possible dates on which High Priest Matthias would have recused himself are highlighted, the astronomy, calendrics, and Jewish festivals don't reconcile as well as the earlier eclipse in the table above.
Comment August 5 BC 30 Elul 3756 27 31 28 September 5 BC 1 Fri 29 Astronomical new moon 18:09 UT/GMT 2 Sat Tishri 3757 1 Tishri 1 and Rosh Hashanah (by computed calendar) 3 2 new moon visible 4 3 / 1 Tishri 1 and Rosh Hashanah (by observation of visible new moon) 5 4 / 2 6 5 / 3 7 6 / 4 8 7 / 5 9 Sat 8 / 6 10 Sun 9 / 7 Matthias recuses himself from officiating the fast (Ant. 17.6.4) (by computed calendar) 11 Mon 10 / 8 Yom Kippur (by computed calendar) 12 Tues 11 / 9 Matthias recuses himself from officiating the fast (Ant. 17.6.4) (by observation of visible new moon) 13 Wed 12 / 10 Yom Kippur (by observation of visible new moon) 14 Thu 13 15 Fri 14 Total lunar eclipse & full moon (Ant. 17.6.4) 16 Sat 15 17 16 18 17 19 18 20 19 21 20 22 21 23 22 24 23 25 24 26 25 27 26 28 27 29 28 30 29 October 5 BC 1 30 2 Heshvan 3757 1
Astronomical new moons in 5 B.C.
If the Jewish Temple priests were attempting to determine the start of the month by observation of the new moon, they would be watching the evening skies for the first visible crescent of the new moon. Fred Espenak at NASA has computed a catalog of the phases of the moon. The dates when new moons would have occurred in 5 B.C.are shown below and highlighted for the months of March and September :
Phases of the Moon: -4 to 0 ( 5 to 1 BCE)
Universal Time (UT)Year New Moon First Quarter Full Moon Last Quarter ΔT -4 Jan 2 15:42 02h57m Jan 9 06:55 Jan 16 08:21 Jan 24 13:06 Feb 1 02:03 Feb 7 17:32 Feb 15 03:27 Feb 23 05:09 Mar 1 09:23 Mar 8 04:44 P Mar 15 22:45 Mar 23 18:16 t Mar 30 14:54 Apr 6 16:51 P Apr 14 16:57 Apr 22 04:31 Apr 28 20:07 May 6 05:57 May 14 09:16 May 21 12:29 May 28 02:29 Jun 4 20:00 Jun 12 23:11 Jun 19 19:13 Jun 26 11:13 Jul 4 10:55 Jul 12 10:34 Jul 19 01:58 Jul 25 23:05 Aug 3 02:28 Aug 10 19:40 Aug 17 09:56 Aug 24 14:26 Sep 1 18:09 P Sep 9 03:12 Sep 15 20:05 t Sep 23 09:02 Oct 1 09:15 P Oct 8 10:10 Oct 15 08:54 Oct 23 05:58 Oct 30 23:08 Nov 6 17:49 Nov 14 00:21 Nov 22 03:28 Nov 29 11:32 Dec 6 03:09 Dec 13 17:59 Dec 21 23:19 Dec 28 22:39Fred Espenak, Phases of the Moon: -99 to 0 (100 to 1 BCE)
NASA Goddard Space Flight Center, 2007 March 27The Julian dates March 8 and September 1, astronomical year -4 (5 B.C.) convert to the Hillel II Hebrew calendar (using the Fourmilab Calendar Converter) as:
- -5 March 8 (Julian) => 3756 A.M. Adar 29
- -5 September 1 (Julian) => 3756 A.M. Elul 29
It is certain that year-after-year for hundreds and thousands of years, the Hebrews somehow kept their lunar calendar aligned with the annual solar cycle. It is certain that the Hebrews intercalated their lunar calendar, because without some periodic insertion of "leap months", after about 20 years a lunar calendar would be opposite of the actual seasons; i.e. an unadjusted lunar calendar would declare a "summer" month but the actual season would literally be winter, and consequently Passover would not be observed when barley buds (Exo 34:18, Deu 16:1) in the month of Abib (Nisan). Getting the sacred festival months correct, so the festivals would be observed as the Bible commanded was critical to the Hebrews. They must have either adjusted their lunar calendar by inserting months every 2-3 years to keep it aligned with the solar year as in Hebrew and Babylonian Calendar Intercalation, or they observed the vernal equinox and realigned the calendar every year, similar to how the Essenes determined Passover Dates, or by 5 B.C. they were already using their computed calendar (which was definitely in use by A.D. 66 though its calculations weren't publically disclosed until A.D. 359 by Rabbi Hillel II).
The point being, because the Hebrews somehow kept their lunar calendar aligned with the solar year, the conversions done by the Fourmilab converter (regardless of the precise day) accurately compute which Hebrew month was about to begin, and hence when the Temple priests would have been watching for a new moon crescent by which to set the first day of that month. Consequently, it is certain that proximate to the astronomical new moon of 5 B.C. March 8 the Temple priests would have been watching the moon to see what day to declare Nisan 1, and proximate to 5 B.C. Septermber 1, they would have been watching the moon to see what day to declare Tishri 1.
It can be concluded that when the astronomical new moons occurred on 5 B.C. March 8 and Septermber 1, the Temple priests would have been watching the evening skies to detect the new moon lunar crescent, and subsequently declare the date to be Nisan 1 and Tishri 1 (respectively). This observation of new moon crescents would be done every month, every year, but the analysis here is only concerned with those particular months in which total lunar eclipses also occurred.
But the new moon crescent is not visible immediately:
In the first two days after New Moon, the young crescent Moon appears very low in the western sky after sunset, and must be viewed through bright twilight. It sets shortly after sunset. The sighting of the lunar crescent within one day of New Moon is usually difficult. The crescent at this time is quite thin, has a low surface brightness, and can easily be lost in the twilight. Generally, the lunar crescent will become visible to suitably-located, experienced observers with good sky conditions about one day after New Moon. However, the time that the crescent actually becomes visible varies quite a bit from one month to another. The record for an early sighting of a lunar crescent, with a telescope, is 12.1 hours after New Moon; for naked-eye sightings, the record is 15.5 hours from New Moon. These are exceptional observations and crescent sightings this early in the lunar month should not be expected as the norm.
Crescent Moon Visibility
U.S. Naval Observatory, 14 June 2011
Actium era coinage of Antioch bearing Varus' name
Coins minted in Antioch and dated in years 25, 26, and 27 of the Actian era indicate P. Quinctilius Varus was in Syria approximately September of 7 B.C. through September of 4 B.C. Exemplars of these coins held in the British Museum (BMC 57, BMC 58, BMC 59) are described by Cecil Torr:
Josephus repeatedly mentions Varus as the legate of Syria at the end of Herod's reign and the beginning of the reign of his successor, Archelaus15. And the fact is established by the coins of Varus, struck in Syria, with the dates of 25 and 26 and 27 in the era of Actium, which represent the three years running from the 6th of September in 7 to the 6th of September in 4 B.C. Josephus says that Varus was the successor of Saturninus16.
15 Ant. xvii. 5. 2-7; 9. 3; 10. i, 9, 10.
16 Ant. xvii. 5. 2
FlG. 3. THREE BRONZE COINS IN THE BRITISH MUSEUM. Actual size. Obverse. Head of Zeus, to right. Reverse. The statue (by Eutychides) of Antioch on the Orontes. Inscribed Ἀντιοχἑων and ἐπἰ Ὀυάρου or ἐπἰ Ὀάρου with the dates εκ and Ϛκ and ζκ respectively.
[p. 9]
Cecil Torr, On Portraits of Christ in the British Museum,
C.J. Clay and Sons (1898)Similar coins excavated at Antioch between 1932 and 1939, are further described by Dorothy Waage (note her more accurate use of UNCIAL letters) as:
AUGUSTUS, 27 B.C.-14 A.D.
MUNICIPAL ISSUESDATED BY ACTIAN ERA
DATED ΕΚ, OCT. 7-OCT. 6 B.C.300. ΑΝΤΙΟΧΕΩΝ ΕΠΙΟVΑΡΟV clockwise from lower l.
Tyche of Antioch seated r. with river-god Orontes at her feet. In r. field, ΕΚ.
Obv. Head of Zeus r., laur.
SINGLE DENOMINATION, 19.5-21 mm.
N.C., 4th ser., IV (1904), pp. 106-109; cf. Grant, I. to A., pp. 397-398.DATED
, OCT. 6-OCT. 5 B.C.
301. Similar, but in r. field
Obv. Similar.
SINGLE DENOMINATION, 20 mm.
Same references.DATED ΖΚ, OCT. 5-OCT. 4 B.C.
302. ΑΝΤΙΟΧΕΩΝ ΕΠΙ ΟVΑΡΟV clockwise from lower 1. Similar, but in r. field, ΖΚ.
Obv. Similar.
SINGLE DENOMINATION, 19-20 mm.
Same references.
[p. 29]Dorothy B. Waage,
Antioch on the Orontes vol IV part 2 - Greek, Roman, Byzantine and Crusaders' Coins
Princeton University Press (1952)There is additional and earlier discussion of these coins by:
Emil Schürer, Géza Vermès, Fergus Millar, Matthew Black, pp. 256-261 of History of the Jewish People in the Age of Jesus Christ vol 1,
Continuum International Publishing Group (1973)George MacDonald, pp. 106-109 of "The pseudo-autonomous coinage of Antioch."
The Numismatic Chronicle, Series IV, vol. 4 (1904) pp. 105-135Warwick William Wroth, pp. 158-159, nos., 57-59 of Catalogue of the Greek Coins in the British Museum vol. 20: Catalogue of the Greek Coins of Galatia, Cappadocia, and Syria, British Museum (1899)
Emil Schürer, p. 351 of History of the Jewish people in the time of Jesus Christ div1 vol1,
Scribner (1891)Edward Greswell, p. 534 of Dissertations on principles and arrangement of a harmony of the Gospels v1
Oxford University Press (1837) 2nd ed.Théodore Edme Mionnet, p. 156 of Description de médailles antiques v5
de L'imprimerie de Testu (1811)Joseph Hilarius von Eckhel, p. 275 of Doctrina numorum veterum p1 v3,
Svmptibvs Iosephi Camesina (Vienna 1794)The inscriptions "ΑΝΤΙΟΧΕΩΝ ΕΠΙΟVΑΡΟV" (#300 & BMC 57; #301 & BMC 58) and "ΑΝΤΙΟΧΕΩΝ ΕΠΙ ΟVΑΡΟV" (#302 & BMC 59) mean "Antioch [ΑΝΤΙΟΧΕΩΝ] in the time of [ΕΠΙ] Varus [ΟVΑΡΟV]":
The Greek alphabet in 1st century B.C. did not have the letter "V" and instead used the combination "ΟY" (capital omicron, capital upsilon). That is why the leftmost letters of Varus' name are stamped on the coin as "ΟV". But the rightmost letter also appears to be a "V", but in Greek it is actually a "Y" (capital upsilon). The artisans who cut the dies to strike coins at Antioch used the symbol "V" instead of the symbol "Y" ostensibly because the "V" is nearly identical (as far as their accuracy permitted) and was less difficult to shape. So the name Ὀυάρου in Greek is the masculine, singular, genitive case, 2nd declension of "Οὔαρος" ("Varus") which was then actually stamped on the coin as "ΟVΑΡΟV".
Then, as per Liddell and Scott, when the Greek word ΕΠΙ is used with a genitive case ( Ὀυάρου in this instance), it means "in the time of":
ἘΠΙ᾿, Prep. with gen., dat., and acc.: Radic. signf. upon.
A. With Gen.:
I. of Place,
II. of Time, in the time of, ἐπὶ προτέρων ἀνθρώπων Iliad; ἐπὶ Κύρου Herodotus; ἐπ᾿ ἐμοῦ in my time, Id.
III. in various Causal senses:An Intermediate Greek-English Lexicon: Founded Upon the Seventh Edition of Liddell and Scott's Greek-English Lexicon.
The dates on the coins (visible next to Tyche's knee, just to the right of center on the reverse sides) are ΕΚ (epsilon kappa), ϚΚ (stigma kappa), and ΖΚ (zeta kappa). Note the middle coin (BMC 58) is dated ϚΚ (stigma kappa), where the "stigma" is flattened and elongated, ostensibly because it is a difficult symbol to carve on a small coin die. In the ancient Greek system of number-date equivalences:
K kappa = 20; E epsilon = 5; digamma Ϝ (replaced by stigma Ϛ) = 6; Z zeta = 7; Η eta = 8
Consequently, the three coins (e.g. BMC 57, 58, 59) indicate "Antioch in the time of Varus", for Actian era years 25, 26, and 27 which correspond to September of 7 B.C. thru September of 4 B.C.
Michael Grant further argues that the presence of names on these particular coins (including Varus') indicated control and authority:
It is true that the use of ΕΠΙ with a name in the Genitive is, at a later date, merely eponymous and chronographical; but its contexts in this period suggest strongly that, contrary to the general view, they imply, not only mere eponymy, but a practical control over the city-mint or some of its mintages. … Since ΕΠΙ does not differentiate ‘ free ’ from stipendiary cities, it must indicate actual interference in the administrative processes which produced the coinages. [p. 397]
Other evidence also can be adduced in support of the view that these formulas are expressive, rather than a meaningless record of eponymy. … The entirely indiscriminate appearance of the Genitive and Dative cases after ΕΠΙ, in recording these officials’ names, indicates that the former usage, as well as the latter, was held to imply a measure of control. It was not merely eponymous and chronographic, but bore witness to the financial supervision of the proconsul, who was, indeed, as elsewhere, ultimately responsible for the κοινόν. [common good - CDD] [p. 399]
Michael Grant, From Imperium to Auctoritas
Cambridge University Press (1946)But Josephus said Varus was in Jerusalem to succeed Saturninus as president of Syria, at the time Herod tried Antipater for conspiracy:
... Now Varus, the president of Syria, happened to be in the palace [at this juncture]; so Antipater went in to his father, and, putting on a bold face, he came near to salute him. But Herod Stretched out his hands, and turned his head away from him, and cried out, "Even this is an indication of a parricide, to be desirous to get me into his arms, when he is under such heinous accusations. God confound thee, thou vile wretch; do not thou touch me, till thou hast cleared thyself of these crimes that are charged upon thee. I appoint thee a court where thou art to be judged, and this Varus, who is very seasonably here, to be thy judge; ... "
Josephus, Wars of the Jews, Book 1, Chapter 31, para. 5
William Whiston, trans. (1736)Now Quintilius Varus was at this time at Jerusalem, being sent to succeed Saturninus as president of Syria, and was come as an assessor to Herod, who had desired his advice in his present affairs; and as they were sitting together, Antipater came upon them, without knowing any thing of the matter; so he came into the palace clothed in purple. The porters indeed received him in, but excluded his friends. And now he was in great disorder, and presently understood the condition he was in, while, upon his going to salute his father, he was repulsed by him, who called him a murderer of his brethren, and a plotter of destruction against himself, and told him that Varus should be his auditor and his judge the very next day; ...
Josephus, Antiquities of the Jews, Book 17, Chapter 5, para. 2
William Whiston, trans. (1736)In the Herodian chronologies based on Josephus' reports, Saturninus sends the conspirators Corinthus and Fabatus to Rome (Ant. 17.3.2) in late 6 B.C. and then Varus arrives to replace him in mid 5 B.C. (Wars 1.31.5) and also judge Antipater (Ant. 17.5.2), whereas the coins would have Varus succeed Saturninus not sooner than mid 7 B.C. and not later than mid 6 B.C., one to two years earlier.
Just prior to his death, Herod receives permission from Augustus to execute Antipater, permission which was requested a few months earlier following Antipater's trial a few months earlier still, a trial at which Varus presided while in Jerusalem to succeed Saturninus, Saturninus having sent conspirators to Rome several months before Varus' arrival.
Through a series of interrelated and overlapping events, the arrival of Varus can't be shifted one year earlier (so as to reconcile with "his" Antioch coins) without also shifting one year earlier Antipater's trial and execution and Herod's death with all the related correspondence with Rome. But there are no total or even partial lunar eclipses in 6 or 7 B.C., there is only a partial eclipse in 8 B.C. and then two total lunar eclipses in 9 B.C. Though imperfect and imprecise, the Herodian chronologies based on Josephus' reports are not 4 years too long, rather, aside from these coins they reconcile well within a couple months of other historical synchronisms.
Obviously, neither can these "Antioch in the time of Varus" coins be reconciled with a death of Herod in 1 B.C., because as described above, Herod's death is (interlinked through a series of events) several months to a year after Varus succeeds Saturninus, which succession the coins seem to date as not later than mid 6 B.C.
Two possibilities for reconciliation to investigate are:
- If Josephus or his source for Ant. 17.5.2 meant instead of "Varus was in Jerusalem to succeed Saturninus" but rather "Varus (who was Saturninus' successor) was in Jerusalem", i.e. without synchronizing the succession to being in Jerusalem as implied in Wars 1.31.5.
- If the Antiochian coins reckoned the era slightly differently than from Augustus' victory at Actium, e.g. perhaps from being named "Augustus".
