Historical Events

# The Date of the Battle of Mu'tah in Light of Theophanes and the Reconstruction of the Arab Solar-Lunar Calendar ## Abstract This paper investigates the date of the Battle of Mu'tah through the Byzantine testimony of Theophanes the Confessor and the reconstruction of the Arab solar-lunar calendar. Theophanes places the battle of Mothous under AM 6123 and records four Arab commanders, the death of three, the survival of Khalid, and Khalid's title "the Sword of God". These details correspond closely with the Islamic account of Mu'tah. The chronological importance of the testimony is that Theophanes places Mu'tah in the same context in which he records Persian internal conflict and the sending of congratulatory gifts by the king of India to Heraclius for his victory over the Persians. This context can be compared with independently dated events of AD 630, especially Heraclius' presence in Jerusalem and the restoration of the Cross on 21 March 630, and the decisive Sasanian disturbances of June 630, including the killing of Shahrbaraz. The conclusion adopted here is that Mu'tah occurred in Jumada I of year 8, but that this date should be converted according to the reconstructed Arab solar-lunar calendar employing intercalation rather than the later purely lunar Hijri calendar. Under that reconstruction, Jumada I of year 8 begins on 17 June 630. ## 1. Introduction The commonly transmitted Islamic date of Mu'tah is Jumada I, year 8 AH. Converting that date by the later purely lunar Hijri calendar normally places the event in AD 629. That conversion, however, assumes that the Prophetic-period Arab calendar operated exactly like the later Islamic calendar after the abolition of intercalation. The present investigation instead considers the possibility that the Arab calendar was solar-lunar and employed nasi' (intercalation), keeping lunar months connected with the seasons through the insertion of a leap month. The central question is therefore where Jumada I of year 8 falls under that reconstructed system. ## 2. Theophanes' Chronological Placement Theophanes places the account of Mu'tah under AM 6123, the year 6123 from Creation. At the head of this chronological section he gives Heraclius in his twenty-second year and Abu Bakr as leader of the Arabs in his first year. The Mango and Scott edition renders AM 6123 as AD 630/1, although other reconstructions produce AD 631/632 because of chronological problems in Theophanes. This does not change the fact that Theophanes himself places Mothous under AM 6123. ## 3. The Greek Account > Τούτῳ τῷ ἔτει ἐπανέστησαν Πέρσαι κατ' ἀλλήλων καὶ ἐπολέμησαν ἑαυτούς. ἐν αὐτῷ δὲ τῷ καιρῷ ὁ βασιλεὺς τῶν Ἰνδῶν πέμπει συγχαρίκια τῷ Ἡρακλείῳ ἐπὶ τῇ τῶν Περσῶν νίκῃ, μαργαρίτας καὶ λίθους τιμίους ἱκανούς. ἦν δὲ προτελευτήσας ὁ Μουάμεδ, ὃς ἦν στήσας τέσσαρας ἀμηραίους τοῦ πολεμεῖν τοὺς ἐξ Ἀράβων γένους Χριστιανούς· καὶ ἦλθον κατέναντι Μουχέων κώμης λεγομένης, ἐν ᾗ ὑπῆρχε Θεόδωρος ὁ βικάριος, θέλοντες ἐπιρρίψαι κατὰ τῶν Ἀράβων τῇ ἡμέρᾳ τῆς εἰδωλοθυσίας αὐτῶν. μαθὼν δὲ τοῦτο ὁ βικάριος παρά τινος κορασηνοῦ, Κουταβᾶ λεγομένου καὶ μισθίου αὐτοῦ γενομένου, συνάγει πάντας τοὺς στρατιώτας τῶν παραφυλάκων τῆς ἐρήμου, καὶ ἀκριβωσάμενος παρὰ τοῦ Σαρακηνοῦ τὴν ἡμέραν καὶ τὴν ὥραν, ἐν ᾗ ἤμελλον ἐπιρρίπτειν αὐτοῖς, αὐτὸς ἐπιρρίψας αὐτοῖς ἐν χωρίῳ ἐπιλεγομένῳ Μόθους ἀποκτέννει τρεῖς ἀμηραίους καὶ τὸ πλῆθος τοῦ λαοῦ. ἐξῆλθε δὲ εἷς ἀμηρᾶς ὁ Χάλεδος, ὃν λέγουσι τὴν μάχαιραν τοῦ θεοῦ. ἦσαν δέ τινες τῶν πλησίον Ἀράβων λαμβάνοντες παρὰ τῶν βασιλέων ῥόγας μικρὰς πρὸς τὸ φυλάξαι τὰ στόμια τῆς ἐρήμου. ἐν αὐτῷ δὲ τῷ καιρῷ ἦλθέ τις εὐνοῦχος δίδων τὰς ῥόγας τῶν στρατιωτῶν, καὶ ἐλθόντες οἱ Ἄραβες κατὰ τὸ ἔθος λαβεῖν τὴν ῥόγαν αὐτῶν, ὁ εὐνοῦχος ἀπεδίωξεν αὐτούς, λέγων ὅτι «ὁ δεσπότης μόγις τοῖς στρατιώταις δίδωσι ῥόγας, πόσῳ μᾶλλον τοῖς κυσὶ τούτοις;» θλιβέντες οὖν οἱ Ἄραβες ἀπῆλθον πρὸς τοὺς ὁμοφύλους, καὶ αὐτοὶ ὡδήγησαν αὐτοὺς ἐπὶ τὴν χώραν Γάζης στομίου οὔσης τῆς ἐρήμου κατὰ τὸ Σιναῖον ὄρος πλουσίας σφόδρα. ## 4. Translation "In this year the Persians rose against one another and fought among themselves. At the same time the king of India sent congratulations to Heraclius on his victory over the Persians, sending pearls and a considerable number of precious stones. Muhammad had already died, and he had appointed four commanders to fight those of the Arab race who were Christians. They came opposite a village called Mocheon, where Theodore the vicar was stationed, intending to attack the Arabs on the day of their idolatrous sacrifice. When the vicar learned this from a Qurayshite named Qutaba, he gathered the soldiers of the desert outposts. Having ascertained from the Saracen the day and the hour on which they intended to attack, he attacked them at a village called Mothous, killing three commanders and the greater part of the army. One commander, Khaled, escaped; he is called the Sword of God. Some neighboring Arabs received small rations from the emperors for guarding the entrances of the desert. At that time a eunuch came to distribute the soldiers' rations. When the Arabs came, according to custom, to receive theirs, he drove them away, saying that the master scarcely gave rations to the soldiers, much less to these dogs. The Arabs were distressed and went to their fellow tribesmen, who led them into the land of Gaza, an entrance to the desert toward Mount Sinai and a very rich land." ## 5. Correspondence with Mu'tah The correspondence is substantial: four commanders are mentioned; three are killed; one named Khalid survives; Khalid is called "the Sword of God"; and the place is called Mothous. These elements correspond to Zayd ibn Harithah, Ja'far ibn Abi Talib, Abd Allah ibn Rawahah and Khalid ibn al-Walid in the Islamic account of Mu'tah. The identification is therefore based on the combination of the place-name, the number and fate of the commanders, Khalid's name and his distinctive title, rather than on phonetic similarity alone. ## 6. The Chronological Sequence Theophanes' sequence is crucial. He first records Persian internal conflict: > ἐπανέστησαν Πέρσαι κατ' ἀλλήλων "The Persians rose against one another." He then says: > ἐν αὐτῷ δὲ τῷ καιρῷ "At the same time / in that period." He records the Indian king's gifts to Heraclius: > ὁ βασιλεὺς τῶν Ἰνδῶν πέμπει συγχαρίκια τῷ Ἡρακλείῳ "The king of India sent congratulations to Heraclius." The reason is: > ἐπὶ τῇ τῶν Περσῶν νίκῃ "On account of his victory over the Persians." Immediately afterward Theophanes begins the account of Muhammad's death and then narrates Mothous. The sequence is therefore: **Persian internal conflict → Heraclius' victory over Persia → Indian royal gifts → Mu'tah.** ## 7. Heraclius in Jerusalem An independent chronological anchor is Heraclius' presence in Jerusalem and the restoration of the True Cross. The date most commonly favored in the reconstruction used here is **21 March 630**. This places Heraclius in Palestine in the spring of 630, within the same broad geographical and political setting as the later Mu'tah narrative. ## 8. The Sasanian Conflict After the fall of Khosrow II, the Sasanian Empire entered a period of severe instability. In AD 630 Shahrbaraz became king briefly and was subsequently killed; his death is commonly dated to **9 June 630**. The succession struggle continued and included the accession of Boran. Theophanes' statement that the Persians rose against one another therefore fits the Sasanian civil turmoil of 630, particularly its decisive phase in June. ## 9. The Indian Gifts Theophanes specifically says that the king of India sent Heraclius congratulatory gifts consisting of pearls and precious stones because of his victory over Persia. The account does not preserve the exact day or month of the embassy's arrival, so the embassy alone cannot establish a precise date for Mu'tah. Its value is that it confirms the chronological context in which Theophanes places Mu'tah. ## 10. The Calendar Problem The common conversion of Jumada I of year 8 to AD 629 assumes the later purely lunar Hijri calendar. If the Arab calendar of the Prophetic period employed intercalation, however, it must instead be reconstructed as a solar-lunar calendar in which a leap month maintained the relation between lunar months and the seasons. Therefore, Jumada I of year 8 need not correspond to the date obtained from the later purely lunar calendar. ## 11. Reconstruction of Year 8 Under the calendrical reconstruction adopted in this investigation: > **Jumada I of year 8 begins on 17 June 630.** The chronological correspondence is: | Event | Date | |---|---:| | Heraclius in Jerusalem and restoration of the Cross | **21 March 630** | | Sasanian internal conflict | **630** | | Killing of Shahrbaraz | **9 June 630** | | Beginning of Jumada I, year 8, under the reconstructed solar-lunar calendar | **17 June 630** | | Battle of Mu'tah | **Jumada I, year 8** | This is a convergence of the Byzantine political context, the Sasanian chronology, the chronology of Heraclius, and the reconstructed Arab calendar. ## 12. "The Day and the Hour" Theophanes states that Theodore learned: > ἀκριβωσάμενος ... τὴν ἡμέραν καὶ τὴν ὥραν "the day and the hour." The surviving account therefore preserves evidence that the planned attack had a known day and hour, although the actual name of the day and the hour are not preserved. ## 13. The Ritual Day Theophanes further says that the intended attack was to occur: > τῇ ἡμέρᾳ τῆς εἰδωλοθυσίας αὐτῶν "on the day of their idolatrous sacrifice." The account thus preserves a ritual chronological marker in addition to the year and the day-and-hour reference, although it does not name the month. ## 14. Why the Byzantine Evidence Need Not Be Rejected The apparent conflict between Mu'tah in year 8 and AD 629 arises if year 8 is converted through the later purely lunar calendar. Under the reconstructed solar-lunar calendar employing intercalation, year 8 extends into AD 630. Theophanes' AM 6123 can therefore be treated as an external chronological witness compatible with this reconstruction rather than simply as an erroneous date. ## 15. Evaluation of the Evidence The evidence may be summarized in five points: 1. **Identification:** Mothous, four commanders, three deaths, Khalid's survival, and the title "Sword of God" correspond closely to Mu'tah. 2. **Geography:** the account is situated in the Levant. 3. **Political context:** Theophanes places the event amid the Persian internal conflict and after Heraclius' victory. 4. **Heraclius:** his presence in Jerusalem and restoration of the Cross on 21 March 630 provides an independent chronological anchor. 5. **Sasanian chronology:** the severe internal disturbances of 630, including Shahrbaraz's death on 9 June in the commonly accepted reconstruction, fit the context recorded by Theophanes. ## 16. Conclusion On the combined evidence, the date favored by this investigation is: > ## **Jumada I, year 8 AH** This date is not converted according to the later purely lunar Hijri calendar, but according to the reconstructed Arab solar-lunar calendar employing intercalation and a leap month. Under that reconstruction: > ## **Jumada I of year 8 begins on 17 June 630.** Theophanes provides an important external witness: he places Mothous, whose principal features correspond to Mu'tah, under AM 6123, in the same chronological context in which he records Persian internal conflict and the Indian king's gifts to Heraclius following his victory over Persia, immediately before narrating the battle. This framework is consistent with the historical setting of AD 630, especially Heraclius' presence in Jerusalem in March and the decisive Sasanian disturbances of June. ## 17. Final Statement > **The Battle of Mu'tah occurred in Jumada I of the eighth year of the Prophetic period. The common conversion to AD 629 results from projecting the later purely lunar Hijri calendar backward onto the Prophetic period. If the earlier Arab calendar is reconstructed as a solar-lunar calendar employing intercalation, Jumada I of year 8 falls in AD 630 and begins, under the reconstruction adopted here, on 17 June 630.** > > **The testimony of Theophanes provides an external Byzantine witness to this chronological framework. He places the battle of Mothous, whose principal features correspond to Mu'tah, under AM 6123, in the same context in which he records Persian internal conflict and the Indian king's gifts to Heraclius following his victory over Persia, and then immediately narrates the battle. This framework corresponds with Heraclius' presence in Jerusalem in March 630 and with the decisive Sasanian disturbances of June 630.** > > **Accordingly, Theophanes does not explicitly name the month of Mu'tah, but his testimony provides an important external chronological indication consistent with AD 630, which corresponds to Jumada I of the eighth year under the reconstructed Arab solar-lunar calendar employing intercalation.** ## 18. Methodological Note A distinction must be maintained between: **"Theophanes stated that Mu'tah occurred on 17 June 630."** — This is not correct; Theophanes did not state that date. and: **"The reconstruction of the Arab solar-lunar calendar places the beginning of Jumada I of year 8 on 17 June 630, while Theophanes places Mu'tah within a chronological framework consistent with AD 630."** — This is the conclusion supported by the evidence assembled in this investigation. ## References - Theophanes Confessor, *Chronographia*, AM 6123, ed. Carl de Boor, I, 335–336. - Cyril Mango & Roger Scott, *The Chronicle of Theophanes*. - Robert Schick, *The Christian Communities of Palestine from Byzantine to Islamic Rule*. - Marie Louise Chaumont, "Bōrān," *Encyclopaedia Iranica*. - Modern studies concerning the chronology of Theophanes and the identification of Mothous with Mu'tah.

The Testimony of Agapius of Manbij on the Death of Abu Bakr al-Siddiq and the Accession of Umar ibn al-Khattab A Study of "Year Thirteen of the Arabs" and What It Reveals About the Arab Solar–Lunar Calendar Abstract This study examines a single, specific testimony from the chronicle of the Arab Christian historian Agapius of Manbij: his notice of the death of Abu Bakr al-Siddiq and the accession of Umar ibn al-Khattab to the caliphate after him. Agapius links this event to an explicit double synchronism: year 946 of the Alexandrian (Seleucid, Anno Graecorum) calendar, and "the thirteenth year of the Arabs." The study compares this testimony with the Arabic tradition, which places the same event in the thirteenth year of the Hijra, then tests where this year falls on the Julian timeline under two possible models of the early Arab calendar: an intercalated solar–lunar model, and the purely lunar (non-intercalated) model of reckoning that later became standard in Islamic jurisprudence. While a year-level Julian test alone does not settle the dispute between the two models (because Alexandrian year 946 touches the edges of both frameworks), the study proceeds to a finer test based on Arabic reports that specify the very month of Abu Bakr's death (Jumada al-Akhira) and the weekday on which it occurred. This test reveals that, under the traditional tabular lunar Hijri calendar, this month falls entirely within August 634 CE — a full two months before the start of year 946 (fixed at 1 October 634 CE) — placing it entirely outside AG 946 and contradicting Agapius's own statement. Under the solar–lunar model, by contrast, the same month falls in June–July 635 CE, comfortably inside AG 946. This is corroborated by an independent test of the weekdays mentioned in the reports (Thursday, then Friday). Taken together — the explicit methodological testimony, the numerical match, the AG-946 boundary test, and the astronomical match of weekdays — the study concludes that the chronological framework reflected in Agapius's testimony for the death of Abu Bakr and the accession of Umar is "year 13" of an Arab solar–lunar reckoning, not "year 13" of the purely lunar Hijri reckoning devised later — a conclusion consistent with the dating method used in non-Arabic sources contemporary with the event. 1. Introduction: The Importance of Agapius of Manbij's Testimony Agapius ibn Constantine of Manbij, known in Arabic sources as Mahbub ibn Constantine, bishop of Manbij, was an Arab Christian historian who lived in the first half of the tenth century CE. He authored the work known as Kitab al-'Unwan (Historia Universalis), relying in large part on Syriac and Greek sources older than his own time. The value of this source for the study of the early Arab calendar lies in the fact that it simultaneously uses a double dating system linking "the years of the Arabs" to the Alexandrian calendar (the Seleucid-era calendar, a fixed solar calendar whose nature and epoch are not in dispute). This duality makes Agapius an independent witness against whom the years of the Arabs can be tested, rather than relying solely on the modern retrospective calculation of the lunar Hijri calendar. This study is limited to a single notice among Agapius's reports — the death of Abu Bakr al-Siddiq and the accession of Umar ibn al-Khattab to the caliphate — placing it in direct comparison with the Arabic account of the same event, then testing the result against the two possible models of the Arab calendar. The study is deliberately confined to this one event, to the exclusion of other events Agapius records with similar synchronisms, in the interest of focus and precision. 2. The Source Report: The Death of Abu Bakr and Umar's Accession According to Agapius Agapius writes, at the point where he records the transfer of the caliphate: "Abu-Bekr died; and after him reigned Omar, son of Khattab, for twelve years, from the 946 year of Alexander and the thirteenth year of the Arabs." — Agapius of Manbij, Kitab al-'Unwan. ("Abu Bakr died, and after him Umar ibn al-Khattab ruled for twelve years, beginning from year 946 of Alexander and the thirteenth year of the Arabs.") This is an extremely significant formulation, because it does not merely state that Umar succeeded Abu Bakr, but offers an explicit double synchronism between two independent calendars: Year 946 of Alexander (a fixed solar calendar) = the thirteenth year of the Arabs. It should also be noted that Agapius states Umar's reign lasted twelve years — a figure worth recording faithfully, since the well-known Islamic accounts place Umar's caliphate at roughly ten years and several months (from year 13 AH to Dhu al-Hijjah of year 23 AH). This discrepancy between the two traditions is a point that merits independent investigation, which this study does not undertake in detail; it is noted here in the interest of scholarly honesty, without building any unsupported conclusion upon it. 3. The Corresponding Arabic Account: Umar's Accession in Year 13 AH The Arabic sources agree that Umar ibn al-Khattab's accession to the caliphate, following the death of Abu Bakr al-Siddiq, occurred in the thirteenth year of the Hijra. The sources differ on the precise day and month, but the year number itself is not disputed in the historical and biographical literature: Death of Abu Bakr and accession of Umar = year 13 AH. This produces a direct numerical match between the two sources. However, the number alone — as in any sound historical comparison — is not sufficient by itself to establish the nature of the calendar behind it, since a year bearing the number "13" can fall at entirely different points in time depending on the type of calendar used to count it. This requires testing the precise Julian position of this year, which is the subject of the following section. 4. Testing the Chronological Position: Year 13 Under the Two Calendar Models This section tests where "the thirteenth year of the Arabs" falls on the Julian timeline under two possible models of the Arab calendar in this early period: - The intercalated solar–lunar model: a calendar with lunar months but subject to a periodic intercalary correction that keeps it close to the solar year — the model this study proposes. - The traditional lunar Hijri model: the purely lunar, non-intercalated calendar that became standard in Islamic reckoning after the prohibition of nasi' (intercalation), and which is used today to convert Hijri dates. Under the solar–lunar model under study: Year 13 = 25 January 635 – 12 February 636 (Julian). Under the traditional lunar Hijri reckoning: Year 13 = 6 March 634 – 23 February 635 (Julian). Note that the difference between the two models here is not a matter of a few days, but a difference of nearly a full year in where the year falls within the Christian calendar. As for the Alexandrian year Agapius cites — 946 — it falls, by the epoch of the Seleucid calendar, within the range: AG 946 ≈ 634/635 CE. 5. Detailed Arabic Reports on the Day of Abu Bakr's Death Within the Month of Jumada al-Akhira Some Arabic sources go beyond citing the year alone, specifying the month, the weekday, and the day's position within the month. This level of detail allows for a much finer test than a general test at the level of the whole year, because it allows the event to be pinned to a specific point in time rather than a broad range. First report: Ibn Ishaq (via Usd al-Ghaba) "Abu Bakr, may God be pleased with him, died on Friday, with seven nights remaining of Jumada al-Akhira, in year thirteen, and Umar ibn al-Khattab led the funeral prayer over him... Others said he died on Monday evening. It was also said: on Tuesday night. And it was said: on Tuesday evening, with eight nights remaining of Jumada al-Akhira." — Ibn al-Athir al-Jazari, Usd al-Ghaba fi Ma'rifat al-Sahaba, vol. 3, p. 333 (chapter on his death). Second report: Ibn Ishaq (via al-Bad' wa'l-Tarikh) "Ibn Ishaq said: he died on Friday, with seven nights remaining of Jumada al-Akhira, in the thirteenth year of the Hijra. Abu al-Yaqzan said: he died on Monday. They disagreed on the cause of his death: some said he was poisoned and died; others said he bathed on a cold day, caught a fever, and died, may God be pleased with him." — al-Mutahhar ibn Tahir al-Maqdisi, al-Bad' wa'l-Tarikh, vol. 5, p. 79 (the death of Abu Bakr, may God be pleased with him). It is worth noting that the report "Friday, with seven nights remaining of Jumada al-Akhira, in year thirteen" is preserved from Ibn Ishaq himself in two independent sources (Usd al-Ghaba and al-Bad' wa'l-Tarikh) — without thereby entirely dismissing the other, differing reports; they are recorded here faithfully as transmitted. Third report: Futuh al-Sham — an independent witness outside the context of the death "The day I wrote you this letter was a Thursday, with two nights having passed of Jumada al-Akhira, and we are returning to Damascus, God willing." — Kitab Futuh al-Sham (attributed to al-Waqidi), p. 61 (the report of the battle of Ajnadayn). The importance of this third report is that it does not concern Abu Bakr's death at all, but rather a letter written in the context of the battle of Ajnadayn, which mentions in passing that the day it was written was "Thursday, with two nights having passed of Jumada al-Akhira." This makes it an anchor entirely independent of the death reports, fixing another point in time within the same month (Jumada al-Akhira), which can be used to test the internal consistency of the calendar assumed by both reports together. 6. The Decisive Test: The Boundaries of Alexandrian Year 946 and the Position of Jumada al-Akhira Within It It is well established regarding the Seleucid calendar (the calendar "of the two-horned one" or "of Alexander"), in the Syrian form adopted by Syriac writers, including Agapius, that its year begins on the first of October each year, not at the start of the Christian calendar year. Based on this well-known and agreed-upon principle of Syriac chronology: Year 946 of Alexander = 1 October 634 CE – 30 September 635 CE. This gives the year an explicit, unambiguous boundary: 1 October 634 CE. It now becomes possible to test the position of the month of Jumada al-Akhira, year 13 — the month the Arabic sources specify for Abu Bakr's death — against this particular boundary, rather than against the year as a whole range. Under the traditional tabular lunar Hijri calendar (the fixed arithmetical reckoning scholars use to convert early Hijri dates retrospectively, independent of actual lunar sighting), the month of Jumada al-Akhira, year 13 AH, falls within the range: Jumada al-Akhira 13 AH (tabular lunar Hijri) = 1–29 August 634 CE (Julian). Under the solar–lunar model proposed in this study, the same month falls within the range: Jumada al-Akhira 13 (solar–lunar) = 21 June – 20 July 635 CE (Julian). The result here is qualitatively different from the year-level test. There, the test operated at the level of the whole year, and year 946 touched the edges of both models, settling nothing between them. Here, the test operates against a single explicit boundary (the start of AG 946 on 1 October), and the result is decisive: calculating Jumada al-Akhira by the tabular lunar Hijri calendar places the entire month roughly two months before the start of AG 946 — meaning that, under this model, it does not belong to AG 946 at all, but to the preceding year (AG 945, running from 1 October 633 to 30 September 634). This contradicts Agapius's own statement, which ties the event to year 946, not year 945. Calculating the same month under the solar–lunar model, by contrast, places it squarely within year 946, with no contradiction. 7. Testing the Weekdays: The Match of Thursday and Friday The reports cited above provide an additional test entirely independent of any calendrical model, since the day of the week is an astronomical fact unaffected by the type of calendar used — the seven-day week cycle has run uninterrupted for centuries. If the precise Julian date of two days mentioned in the reports can be determined, it becomes possible to check whether they match the stated weekdays (Thursday and Friday) through a purely astronomical correspondence. The two relevant reports are: "Thursday, with two nights having passed of Jumada al-Akhira" (Futuh al-Sham, an independent witness outside the context of the death), and "Friday, with seven nights remaining of Jumada al-Akhira, year 13" (Ibn Ishaq). Under the lunar Hijri calculation, the second day of the month of Jumada al-Akhira, year 13, falls on 2 August 634, a Tuesday, and the day on which "seven nights remained" falls on 23 August 634, also a Tuesday. Under the solar–lunar model, for Jumada al-Akhira, year 13, the day on which "two nights had passed" falls on 22 June 635, which is, by the Julian calendar, a Thursday; and the day on which "seven nights remained" (i.e., the twenty-fourth day of a thirty-day month) falls on 14 July 635, which is, by the Julian calendar, a Friday — the day immediately following Thursday, with the correct one-week interval. 8. Why the Solar–Lunar Interpretation Is Preferred This study's preference for interpreting "year 13 of the Arabs" as a year of a solar–lunar calendar, rather than a purely lunar Hijri calendar, rests on interconnected points: First point: The numerical match between the two years "Year 13 of the Arabs" in Agapius corresponds exactly, numerically, to "year 13 AH" in the Arabic accounts, indicating that the very year number preserved by the Arabic tradition for Umar's accession is the same number Agapius records within the solar–lunar calendar whose mechanism he himself explains. Second point: The month-level test of the AG 946 boundary Calculating the month of Jumada al-Akhira, year 13, by the lunar Hijri calendar places it a full two months before the start of AG 946 — outside the year Agapius explicitly ties the event to — whereas the solar–lunar model places it entirely within AG 946. This is a boundary test admitting no middle ground, and it directly favors the solar–lunar model. Third point: The independent astronomical match of the weekdays "Thursday" and "Friday," as given in two separate reports, do not match under the lunar Hijri calendar, while they match exactly under the solar–lunar model. This is a test independent of any prior calendrical assumption, since it relies solely on the uninterrupted continuity of the seven-day week — an established astronomical fact not in dispute. Taken together, these points make the reading most consistent with all the texts — Agapius's methodical statement, the report of Umar's accession, and the detailed Arabic reports on the month and day of the death — that "the thirteenth year of the Arabs," in which Abu Bakr died and Umar acceded, is a year of an intercalated Arab solar–lunar calendar, not a year of the purely lunar Hijri calendar that later became standard in Islamic jurisprudence after the prohibition of nasi'. 9. Conclusion and Final Result Based on the totality of the evidence presented, this study concludes the following: The year in which Abu Bakr al-Siddiq died and Umar ibn al-Khattab acceded to the caliphate — which Agapius of Manbij calls "the thirteenth year of the Arabs" and ties to year 946 of the Alexandrian calendar — is a year belonging to an intercalated Arab solar–lunar calendar, not a year of the purely lunar Hijri calendar that later became standard in Islamic reckoning after the prohibition of nasi'. This conclusion rests on three mutually reinforcing levels of evidence: First, the exact numerical match between "year 13 of the Arabs" in Agapius and "year 13 AH" in the Arabic accounts. Second, the test of the AG 946 boundary (1 October 634) at the level of the specific month of Jumada al-Akhira, which showed that the lunar Hijri calculation places the month outside AG 946, two months before its start, while the solar–lunar model places it inside without difficulty. Third, the independent astronomical test of the weekdays, which showed that the descriptions "Thursday" and "Friday," found in two sources, do not match within Jumada al-Akhira under the traditional lunar Hijri calculation, but are naturally satisfied within the window of the solar–lunar model. The best scholarly formulation of this result, therefore, is that Agapius of Manbij's testimony, read together with the detailed Arabic reports on the month and day of the death, confirms that the year of Abu Bakr's death and Umar's accession was "the thirteenth year" by an Arab solar–lunar reckoning — consistent with the dating method reflected in non-Arabic sources contemporary with the event — rather than "the thirteenth year" by a purely lunar Hijri reckoning that was devised at a later stage of Islamic historiography. Sources - Agapius of Manbij (Mahbub ibn Constantine), Kitab al-'Unwan (Historia Universalis), Arabic text with French and English translation in the Patrologia Orientalis series, ed. A. Vasiliev. - Ibn al-Athir al-Jazari, Usd al-Ghaba fi Ma'rifat al-Sahaba, Sha'b edition, vol. 3. - al-Mutahhar ibn Tahir al-Maqdisi, al-Bad' wa'l-Tarikh, vol. 5. - Kitab Futuh al-Sham (attributed to al-Waqidi). Agapius's English text was translated into Arabic directly and faithfully, with the original English retained at each point for review and verification. The other Arabic texts are reproduced verbatim from their cited sources.

The death of Heraclius offers a calendrical witness to the divergence between the early Arab calendar and the later purely lunar Hijri calendar. Agapius of Manbij records that Heraclius died on Sunday, February 7, in year 952 of Alexander, year 19 of the Arabs, and year 7 of Umar. Syriac sources (Michael the Syrian and the Chronicle of 1234) preserve the same synchronism: AG 952 + year 19 of the Arabs + year 7 of Umar. This is corroborated by two explicit Arabic reports in Ibn Abd al-Hakam's Futuh Misr wa-l-Maghrib, both placing Heraclius's death in "year nineteen." Combining the day (February 7), the Julian/Gregorian year (641 CE), and the Arab year number (19) explicitly preserved in the sources makes the witness calendrically testable: under the luni-solar model examined here, year 19 spans 18 January 641 to 5 February 642 (Julian), placing 7 February 641 inside it, matching the testimony. Under the traditional purely lunar Hijri reckoning, year 19 spans 31 December 639 to 19 December 640 (Julian), placing 7 February 641 outside it, in year 20, contradicting the testimony. This synchronism therefore suggests that "year 19 of the Arabs" in this historical material tracked a luni-solar cycle via intercalation, rather than the later purely lunar Hijri calendar. -------- Research paper: "The Death of Heraclius in Year Nineteen of the Arabs — A Witness to the Divergence of the Early Arab Calendar from the Later Lunar Hijri Calendar"