[20] Latouche shows that the treaty must be later than the election of Vougrin, bishop of Le Mans, 31 August 1055, and earlier than the death of Geoffrey Martel, 1060. He thinks it probably later than the battle of Varaville, 1058. Maine, p. 32, n. 5.
[21] William of Poitiers, in H. F., xi, pp. 85, 86; Ordericus, ii, pp. 102, 252. The two sources are not in complete accord. Except at one point I have preferred the former as being the more strictly contemporary. William of Poitiers represents the betrothal of William and Margaret not as a part of the original treaty, but as a later arrangement made by Duke William after Herbert’s death in order to forestall a possible controversy as to Norman rights in Maine. But this marriage alliance looms so large in the narrative of Ordericus Vitalis that it seems hardly likely that it was a mere afterthought on Duke William’s part. Ordericus represents it as the fundamental provision of the treaty. According to his view it was through Margaret that Norman rights in Maine arose. He does not seem to realize that upon such reasoning they would also terminate with her death. For William of Poitiers, on the other hand, the fundamental provision of the treaty was the agreement that Duke William should be Count Herbert’s heir. This would give the duke permanent rights after Herbert’s death. It seems not unlikely that both provisions were included in the treaty and that Duke William regarded them both as important. At times he dealt with Maine as if of his own absolute right; at other times he put forward his son as bearer of the Norman rights.
[22] Ordericus, ii, p. 104; William of Poitiers, in H. F., xi, p. 86.
[23] Latouche, Maine, p. 146, nos. 32, 33.
[24] Ibid., p. 33.
[25] Latouche has shown that the date of the revolt falls between 9 March 1062 and 14 March 1063. Maine, p. 33, n. 4. The account of Ordericus Vitalis is confused, and the date (1064) which he gives is impossible. Ordericus, ii, pp. 101-103. The suit held before the ducal curia at Domfront, “cum Guillelmus, Normanniae comes, Cenomannicam urbem haberet adquisitam,” should probably be assigned to 1063 rather than to 1064. Bertrand de Broussillon, Maison de Laval, i, p. 41, no. 28.
[26] Herbert Éveille-Chien was grandfather of Herbert II. Biota, therefore, was aunt of Margaret, Robert Curthose’s fiancée. The genealogy of the counts of Maine in the eleventh century has at last been disentangled by Latouche. Maine, pp. 113-115, appendix iii. F. M. Stenton, William the Conqueror (New York, 1908), pp. 129 ff., and appendix, table d, is inaccurate.
[27] Halphen, Anjou, pp. 137, 293-294, no. 171. Cf. Latouche, Maine, pp. 33-34.
[28] Ordericus, ii, pp. 103, 259. William of Poitiers makes no mention of the poisoning. Halphen (Anjou, p. 179) and Latouche (Maine, p. 34, and n. 6) accept the account of Ordericus as true, the latter explaining that William of Poitiers, as a panegyrist, naturally passes over such an act in silence. Freeman, on the other hand, holds the story to be an unsubstantiated rumor, inconsistent with the character of William the Conqueror. Norman Conquest, iii, p. 208.
[29] Cf. Latouche, Maine, pp. 34-35. The primary authorities are William of Poitiers, in H. F., xi, pp. 85-86, and Ordericus, ii, pp. 101-104.