[707] _Gesta_, _l.c._; _Est._, ll. 5054, 5055. [708] _Gesta_, _l.c._; _Est._, ll. 5056, 5057, 5062, 5063; _Itin._, 235. Geoffrey only “held” his fiefs in the sense that he was legally seised of them; they were Joppa, Caesarea, and Ascalon, all in the enemy’s hands. Conrad’s were Tyre, Sidon, and Beyrout. [709] _Gesta_, 184. [710] R. Diceto, ii. 95. [711] _Est._, ll. 5305-28; _Itin._, 238; _Gesta_, _l.c._; in this last authority the clause about forty days’ notice after Richard’s return is omitted, and the date of the oath is given, July 29. [712] _Gesta_, _l.c._ [713] Rigord, 118. [714] _Est._, ll. 5333-4; _Itin._, 239; _Gesta_, 185. The latter make the date July 31; the _Itinerarium_ makes it August 1. [715] Rigord, 117; date from R. Howden, iii. 126. [716] Otto of S. Blaise, Pertz, xx, 323. [717] _Gesta_, 185, 186. [718] Bohadin, 238-40. [719] _Gesta_, 180. [720] _Itin._, 232. [721] A certain number of the captive Christians of rank were to be chosen by name by the two kings. See Note II at end. [722] Bohadin, 240. [723] From 17 Jomada II (= July 12) to 18 Rajab. [724] The writer of the _Gesta_, 187, gives the appointed day as August 9; no doubt imagining the “month” to mean four weeks. [725] Rigord, 116. [726] _Gesta_, 185. If the prisoners were really as numerous as our authorities represent, the whole of Philip’s share could hardly have gone, with him and his suite, in two galleys. Probably he took the picked ones only. [727] _Ib._, 186, 187; _Est._, ll. 5414-86; _Itin._, 242, 243. [728] _Gesta_, 187. [729] Bohadin, 241, 242. [730] The writer of the _Gesta_, 187, who gives the date for the original first term as August 9, says it was on that day postponed “in diem undecimum post illum.” It is, however, clear from Bohadin that the postponement cannot have been agreed upon till after the 11th; and it is equally clear from the sequel that the term as ultimately fixed cannot have been later than the 20th. This would be the fortieth day from the surrender—which is what the writer of the _Gesta_ asserts in p. 179 to have been the term originally fixed for payment of the whole ransom. Evidently he is correct in his implied date, and wrong only in his mode of arriving at it. [731] _Gesta_, 188, 189. “Sui cum eo” in p. 189 must surely be an error for either “sui cum me,” or, much more probably, “mei cum eo.” [732] The _Est._, ll. 5613-46. and _Itin._, 245, place Richard’s encampment outside the walls and the skirmish or skirmishes which followed it after the slaughter of the garrison, _i. e._ after August 20. But the whole narrative of the surrender of Acre and the proceedings there is in the _Gesta_ arranged with such minute chronological order that it can hardly fail to be founded on documentary authority so far as its dates are concerned, while the chronology of both _Estoire_ and _Itinerarium_, just at this period, is vague and confused in the extreme.