542 APPENDIX Jordan". Muir accordingly puts the battle in a.d. 636.' But there seems to be no certaintj- as to the geographical position of Ajnadain, and it must therefore be regarded as possible that it lay east of the Jordan, and was the scene of a battle either shortly before or shortly after the battle of the Yermiik. The reader may like to have before him the order of e'ents in Tabari ; Mr. Stanley Lane-Poole has kindly supplied me with the references to the original text (ed. de Goeje) : — Abu Bekr sends troops into Syria (a.h. 13), i., 2079. Khalid brings up reinforcements in time for the Yermiik, i., 2089. Battle of the Yermiik, i., 2090 sqq. Battle of Ajnadain (end of July, 634), i., 2126-7. Battle of Fihl (Jan., Feb., 635), i., 2146. Capture of Damascus (Aug., Sept., 63.5), i., 2146. As to the date of the capture of Jerusalem, "Weil does not commit himself ; Muir places it at the end of a.d. 63G (so Tabari, followed by Abii-l-Fida, while other Arabic sources place it in the following year). Theophanes, under a.m. 6127, says : "In this year Omar made an expedition against Palestine ; he besieged the Holy City, and took it by capitulation at the end of two years", a.m. 6127 = March 634-635 ; but, as the Anni Mimdi are here a year late (see above), the presumption is that we must go hy the Anni Incarnationis and interpret the a.m. as March, 635- 636. In that case, the capitidation would have taken place at earliest in !March, 637 — if the two years were interpreted stricth' as twelve months. But Sterr) xp^fo" might be used for two military years, 635 and 636 ; so that the notice of Theophanes is quite consistent with Sir Wm. Muir's date. The same writer agrees with "Weil in setting the battle of Cadesia in a. u. 14, with Tabari, but sets it in Nov. 635, instead of near the beginning of the jear, Noldeke (in his article on Persian History in the Encyc. Brit. ) gives 636 or 637 for Cadesia. Muir's arrangement of the chronology is as follows : — A.D. 634. April, the opposing armies i)osted near the Yermiik. May and June, skirmishing on the Yermuk. August (23), battle of the Yermiik. ,, 635. Summer, Damascus capitulated ; battle of Fihl. November, battle of Cadesia. ,, 636. Spring, Eraesa taken. Other S^'rian towns, including Antioch, taken. Heraclius returns to Constantinople. Spring, battle of Ajnadain. End of the year, Jerusalem capitulates. Summer, siege of Madain begins. ,, 637. March, capture of Madain. ,, 638. Capture of Caesarea. Foundation of Basra and Kiifa. II. Conquest of Egypt. Our Greek authorities give us no help as to the date of the conquest of Egypt, and the capture of Alexandria; and the Arabic sources conflict. The matter, however, has been cleared up by Mr. E. W. Brooks (B^-z. Zeitschrift, iv. , p. 4.35 s(/q.). who has brought on the scene an earlier authority than Theophanes, Nicephorus and all the Arabic histories, — .John of Nikiu, a con- temporarj' of the event. (For his work see above. Appendix 1.) This chronicler implies (Mr. Brooks has shown) that Alexandria capitulated on October 17, a.d. 641 (towards the end of a. h. 20). This date agrees with the notice of Abii-l-Fida, who places the whole conquest within a.h. 20, and is presumably following Tabari (here abridged by the Persian translator) ; and it is borne out by a notice of the 9th centm-y historian Ibn Abd al Hakam (cp. Weil, i. p. 115, note). Along with the correct tradition that Alexandria fell after the death of Heraclius, there was concurrent an inconsistent tradition that it fell on the 1st of the first month of a.h. 20 (Dec. 21, a.d. 640); a confusion of the elder Heraclius with the younger (Hera- clonas) caused more errors (Brooks, loc. cit., p. 437); and there was yet another source of error in the confusion of the first capture of the city with its recapture, 3 It would thus have been fought in connexion with the capture of Ajnadain, which Taban places before the capture of Jerusalem (iii., p. 41c).
