to be right about its duration--expire on 20th December, the day on which Roger says that Berkhamsted surrendered. The Biographer seems to imply that the second truce commenced immediately on the expiration of the first; and twenty days from 20th December bring us to 9th January. If, however, the first truce began on 6th December, it would end on 26th December; and this would bring the termination of the second truce to 15th January. These dates agree neither with the Barnwell annalist’s “octave of Epiphany” nor with the Merton Chronicler’s “octave of S. Hilary”; and what is of much more consequence, even the latest date alleged for the expiration of the second truce--that given by the Merton writer, 20th January--fails to account for the letter patent which shews that there was a truce not merely unexpired, but, seemingly, not even approaching expiration, as late as 19th January. There seems to be no way of overcoming this difficulty except by supposing that the second truce was followed by a third. My belief is that this was so, and that the key to the whole puzzle about the truces and the surrenders of castles in 1216–1217 is to be found in the words of the Barnwell annalist. This writer appears to me to deal with the various truces made between the end of November, 1216, and the end of February, 1217, not singly, but in a group. His account of the _treugae generales_ up to the meeting of the rival councils at Oxford and Cambridge includes, _explicitly_, what may be called the Biographer’s first truce (“reddito quod obsidebatur castello,” _i.e._, Berkhamsted--and Hertford--“pro treugarum impetratione” (cf. _Hist. G. le Mar._, ll. 15717–28); _implicitly_, the Biographer’s second truce (“Autre trieve de vint jors,” ending _approximately_ “ad octavas Epiphaniae”); and _implicitly_ also, I venture to think, a _third_ truce (“_durantibus adhuc treugis_ convocaverunt fautores suos ... tutores regii apud Oxoniam,” as we know from the Close Roll, _after_ the octave of Epiphany (see above, p. 19). After mentioning the two councils and the fruitless negotiations for peace, the annalist tells us that yet _another_ truce (seemingly the fourth) was proposed; and he winds up the whole subject by giving us, _not_ the conditions or the results of that particular proposal, but a general list of the castles--Hedingham, Orford, Norwich, Colchester--which “tunc temporis” (_i.e._ within the last five or six weeks) had, in consideration of the successive truces since the first, been surrendered to Louis, and of which the undisturbed possession was now secured to him for a further period of some two months or more, “pro treugis ad mensem post Pascha.” In a word, the Barnwell writer tells that these four castles were, at some time between the middle of December, 1216, and the middle of February, 1217, bartered for renewals of the truce which had begun with the surrender of Berkhamsted; but which particular castles were