Friends mediate. The archbishop arranges (Many bishops and with John a conference other of the king’s for July 28. lieges mediate.) Longchamp consents, [440] and withdraws. Longchamp makes Description of agreement Brief summary peace as best he between John and of agreement (which could. Longchamp (wrongly Hovenden recites in dated April 25). full). Soon after, Longchamp hears that his mercenaries have landed, and repudiates the agreement. At length, however, they come to terms on a fresh footing. It is the contention of Dr. Stubbs that William of Newburgh, in the first of these columns, describes the first, or spring “campaign,” and that Richard and the ‘Gesta’ describe, in the other two, the second “campaign” later in the year. The difficulty I always felt, in accepting this conclusion, is the almost incredible coincidence of the sequence of events here described occurring twice over, in exactly the same order. But one would not be justified in questioning a view confidently enunciated by Dr. Stubbs, and accepted, it would seem, by every one else, on the ground merely of improbability, however extreme. Let us see, therefore, on what evidence the accepted view is based. In the first place, we are told that the above sequence was repeated twice over. The authorities, however, are all agreed in mentioning one such sequence, and one only.[441] Why, then, are we to convert it into two, in the face of all probability? The only definite reason I can find for so doing is that, according to William of Newburgh-- Longchamp’s proceedings against Lincoln took place early in the spring, before the death of pope Clement III. was known, _or the archbishop of Rouen landed_ [April 27];--[442] while the ‘Gesta’ distinctly state that Longchamp only set out against Lincoln “after Midsummer.” If this were so, the discrepancy would be obvious. But leaving aside, for the moment, the question of the Pope’s death, we find, on reference, that William of Newburgh, so far from placing the campaign, etc., _before_ the archbishop’s arrival, actually places it _after_ that event.[443] The one real discrepancy, therefore, is found to have no existence.[444]