In case any one should still be in doubt, I now bring up my reserves. The undisputed statement that Longchamp was at Winchester on March 24 was supported, we saw, by record evidence that he was there on March 28. Of more importance is the record evidence that he was at Lincoln on July 8,[448] for it strongly confirms the statement in the Gesta that he set out “after midsummer,” and, having rapidly reduced Wigmore, laid siege to Lincoln Castle. Although I have been trying for years to collect evidence of Longchamp’s movements in this eventful year, I have not been able to secure many fixed points. It is certain, however, that he was at Cambridge on April 21.[449] This affords welcome support to the crowning discovery I made, in a document preserved in France, that he was there on April 24.[450] It will, I presume, not be disputed that if the chancellor was at Cambridge on April 24, he cannot have devoted the following day to a conference with John at Winchester. * * * * * I have purposely refrained as yet from discussing a distinct question, namely, the terms of the agreement, or agreements, between Longchamp and John. For they do not affect the question of the sequence of historical events. We have (_a_) in Hoveden what purports to be an actual recital of the agreement made after the chancellor’s enforced withdrawal from Lincoln; (_b_) in Richard of Devizes a _résumé_ of such an agreement effected, according to him, at a conference on July 28, also, it would seem, consequent on the chancellor’s retreat.[451] Dr. Stubbs has argued as against Palgrave, and apparently with complete success, that two distinct agreements are in question. But this does not establish their date (or respective dates), nor even their right sequence. I have already disposed of the alleged conference on April 25, and both agreements, therefore, must be later than the Lincoln business in July. Now, it is singular that William of Newburgh distinctly speaks of two agreements, and implies that the second was the less unfavourable to the chancellor’s claims. This is, at first sight, in striking harmony with Dr. Stubbs’ conclusion that the agreement recited by Hoveden is the later of the two, and that in it “the chancellor gave way somewhat more than was wise, but less than he had done in April”[452] (_i.e._ in the agreement described by Richard of Devizes). But a more minute examination than Dr. Stubbs could give reveals a serious difficulty. According to him, the earlier agreement “engages the chancellor to support John’s claim to the crown in case of Richard’s death”;[453] while the later one contains no such provision. On this distinction he lays stress because “the succession of Arthur,” he holds, was a “main point” of Longchamp’s policy;[454] while the archbishop of Rouen also, he urges, would have “sacrificed other considerations to ... obtaining the omission of any terms which