1. The Angevin chronicles never mention the marriage at all. The Gesta Cons., Will. Jumièges and several other writers mention it without any kind of date. The English Chronicle, Sim. Durh., Will. Malm. and Hen. Hunt. give no distinct date, but imply that the proposal was immediately followed by the wedding. They speak as if Robert and Brian had taken Matilda over sea and married her to Geoffrey without more ado.
2. Orderic mentions the marriage in two places. In the first (Duchesne, Hist. Norm. Scriptt., p. 763) he gives no clue to the date; in the second (ib. p. 889) he dates it 1129.
3. The Chron. Fiscannense (Rer. Gall. Scriptt., vol. xii. p. 778) dates it 1127.
4. A charter of agreement between the bishop of Séez and the convent of Marmoutier (printed in Gilles Bry’s Hist. de Perche, p. 106) has “signum Henrici Regis quando dedit filiam suam Gaufredo comiti Andegavensi juniori.” It is dated “anno ab Inc. Dom. 1127, Indictione VI.”
5. The last witness is John of Marmoutier, the author of the Historia Gaufredi Ducis. From him we might have expected a distinct and authentic statement; but he does not mention the year at all. He says that Geoffrey was knighted on Whit-Sunday and married on its octave, and that he was then fifteen years of age (Hist. Gaufr. Ducis, Marchegay, Comtes, pp. 236, 233). Afterwards, in speaking of the birth of Henry Fitz-Empress, he says that it took place in the fourth year of his parents’ marriage (ib. pp. 277, 278). Henry was born on Mid-Lent Sunday, March 5, 1133; if therefore the writer reckoned backwards from the Whitsuntide of that year, his words ought to mean that the marriage was in 1129. But as he goes on to state that Matilda’s third son was born in the sixth year of her marriage, and that Henry I. died “anno eodem, ab Incarnatione videlicet Domini 1137,” it is impossible to say what he did mean. Whether he is collecting the traditions of the ancient counts or writing the life of his own contemporary sovereign, John’s chronology is pursued by the same fate; whenever he mentions a date by the year, he is almost certain to make it wrong. But that he should have done the like in his reckoning of days, or even of his hero’s age, by no means follows. To consider the latter point first: Geoffrey the Handsome was born on August 24, 1113 (Chron. S. Albin. ad ann., Marchegay, Eglises, p. 32). Therefore, if John meant that he was past fifteen at his marriage, it must have been in 1129. But if he only meant “in his fifteenth year,” it would be 1128. In that year the octave of Pentecost fell on June 17; Geoffrey then lacked but two months to the completion of his fifteenth year; and considering Matilda’s age, it is no wonder that the panegyrist tried to make her husband out as old as possible. It is in fact plain that such was his intention, for though he places Geoffrey’s death in the right year, 1151, he gives his age as forty-one instead of thirty-eight (Hist. Gaufr. Ducis, Marchegay, Comtes, p. 292).
The most important matter, however, is John’s statement that the wedding took place on the octave of Pentecost. The date in this case is not one casually slipped in by the writer in passing; it comes in a detailed account of the festivities at Rouen on the occasion of Geoffrey’s knighting, which is expressly said to have occurred at Pentecost, and to have been followed by his marriage on the octave. Now this leaves us on the horns of a dilemma fatal alike to the date in the Chron. Fiscann., 1127, and to that of Orderic, 1129. For, on the one hand, Will. Malm. (Hist. Nov., l. i. c. 3, Hardy, p. 692) says that Matilda did not go to Normandy till after Whitsuntide [1127]; and Hen. Hunt., l. vii. c. 37 (Arnold, p. 247), adds that the king followed her in August (Sim. Durh., ed. Arnold, vol. ii. pp. 281, 282, really witnesses to the same effect; for his chronology of the whole story is a year in advance). Consequently, as Mrs. Everett Green remarks, “the union could not have taken place before the spring of the following year, 1128” (Princesses of England, vol. i. pp. 107, 108). On the other hand, it is plain that Fulk was present at his son’s wedding; but before Whitsuntide 1129 Fulk was himself married to the princess of Jerusalem (Will. Tyr., l. xiii. c. 24).
From all this it results: 1. If Geoffrey and Matilda were married in 1127, it cannot have been earlier than September, i.e. at least three months after Whitsuntide. 2. If they were married in 1129, it must have been quite at the beginning of the year, and Orderic must, on this occasion at least, have made his year begin in English fashion, at Christmas. 3. If they were married at Whitsuntide, it can only have been in 1128.
We have in short to choose one out of three authorities: the Chronicle of Fécamp, Orderic and John of Marmoutier—for the Séez charter, as Mrs. Everett Green remarks (Princesses, vol. i. p. 108), proves nothing more than that the betrothal had taken place in 1127. Of these three, the first is certainly of least account. Orderic, on the other hand, is on most other subjects a far better authority than John. But his chronology is very little better than John’s, at any rate towards the close of his work; his whole account of Henry’s later years is sketchy and confused; while John is Geoffrey Plantagenet’s own special biographer, writing within sixty years of the event, from materials furnished by personal followers of his hero. I cannot but regard him as our primary authority on this subject, and believe on his testimony that the real wedding-day of Geoffrey and Matilda was the octave of Pentecost, June 17, 1128.