The last charter which I propose to adduce, as attested by "Robert, the king's son," is one which, in all probability, may be assigned to this same occasion, for the whole of its thirteen witnesses had attested the previous charter, with the exception of two bishops, whose presence can be otherwise accounted for,[1245] and of William de Warenne (Earl of Surrey).
The importance of this charter is not so great as that of those adduced above, for it is known to us only from the Rymer Collectanea (Add. MSS., 4573), of which an abstract is appended to the Fœdera.[1246] Moreover, in one minute detail its accuracy may be fairly impugned, for "Willielmo de Warennâ" clearly stands for "Willielmo Comite de Warennâ," Nor, indeed, is its evidence needed, the proof being complete without it. Yet, as the charter (quantum valeat) has been assigned, I think, to a wrong date, the point may be worth glancing at. In the Rymer Collectanea the date is fixed as "1115" (or "16 Henry I.") on the ground that it belongs to the same date as a charter of Henry I. to Bardney, which was granted "Apud Wynton' xvj. anno postquam rex recepit regnum Angliæ."[1247] Mr. Eyton also, in a late addition to his MS. Itinerary of Henry I.,[1248] wrote that the presence of three of the bishops (Lincoln, Salisbury, and St David's) suggested "the latter part of 1115." But we must remember that the Bardney charter is known to us only from a late Inspeximus,[1249] and that the dating clause is somewhat suspicious. Yet even if the version were entirely genuine, the fact remains that the list of witnesses has only four names[1250] in common with that in the charter I am discussing, which has, on the contrary, no less than ten in common with those in the original charter of 1121.[1251] I cannot, therefore, but fix on 1121 as a far more probable date for its grant than 1115-1116.
This, however, as I said, is but a small matter. The really important fact is this: that we have a continuous chain of evidence, proving that "Robert, the king's son," was not yet Earl of Gloucester, at least as late as April-May, 1121.
Against this weight of accumulated evidence what is there? Absolutely nothing but that Tewkesbury charter, which is quoted from Dugdale's Monasticon, where it is quoted from a mere Inspeximus of the 10th Henry IV. (1408-9), some three centuries after its alleged date![1252] I need scarcely say that this miserable evidence for the assertion that Robert was Earl of Gloucester, at Easter, 1116, is simply annihilated and crumpled up by the proof afforded by original charters that he had not yet received the earldom even five years later on (1121).
It is, however, satisfactory to be able to add that, even independent of this rebutting evidence, the charter itself, on its own face, bears witness of its spurious character. Mr. Eyton, indeed, was slightly uneasy about two of the witnesses, it being, he thought, as unusually early for an attestation of Brian fitz Count, as it was late for that of Hamo Dapifer.[1253] Yet he was not, on that account, led to reject it; indeed, he not only accepted, but unfortunately built upon its evidence. He never, however, we must remember, committed his conclusions to print, so that it may be urged with perfect justice that he might have reconsidered and changed his views before he made them public. Not so with Mr. Chester Waters. Announcing the discovery which Mr. Eyton had so strangely anticipated, he wrote—
"We know that the earldom [of Gloucester] was conferred on him [Robert] before Easter, 1116, for he attested as earl the royal charter in favour of Tewkesbury Abbey which was executed at Winchester, on the eve of the king's embarkation for Normandy (Monasticon, vol. ii. p. 66)."[1254]
When Mr. Waters thus wrote, had he observed that in this charter the king's style appears as "Henr' dei gratia Rex Angl' et dux Norm'"? And if he had done so, if he had glanced at the charter on which he based his case, is it possible that he was so unfamiliar with the charters and the writs of Henry I., as not to be aware that such a style, of itself, throws doubt upon the charter?[1255] To those who remember that he confessed (in reply to certain criticisms of my own) to having "carelessly repeated a statement which comes from a discredited authority,"[1256] and that he announced a discovery as to the meeting of Henry I. and Robert of Normandy, in 1101,[1257] which, as I proved, was based only on his own failure to read a charter of this reign aright,[1258] such a correction as this will come as no surprise.
Having now shown that Robert fitz Roy was not yet Earl of Gloucester in April-May, 1121, I proceed to show that he was earl in June, 1123.
The charter by which I prove this is granted "apud Portesmudam in transfretatione meâ."[1259] It is dated in the thirty-first Report of the Deputy Keeper of the Records (in the calendar of these charters drawn up by the late Sir William Hardy) as "1115-1123." Its exact date can, however, be determined, and is 3-10 June, 1123. This I prove thus. The parties addressed are Theowulf, Bishop of Worcester (who died October 20, 1123), and Robert, Earl of Gloucester (who was not yet earl in April-May, 1121). These being the limits of date, the only occasion within these limits on which the king "transfretavit" was in June, 1123. And we learn from the Anglo-Saxon Chronicle that the king, on that occasion, was at Portsmouth, waiting to cross, all Pentecost week (June 3-10). This is conclusive.
It is certain, therefore, that Robert fitz Roy received the earldom of Gloucester between April-May, 1121, and June, 1123. We may even reduce this limit if we can trust a charter in the Register of St. Osmund (i. 382) which is absurdly assigned in the Rolls edition to circ. 1109. The occurrence of Robert, Earl of Leicester, proves that it must be subsequent to his father's death in 1118, and consequently (as the charter is tested at Westminster) to the king's return in 1120. Again, as Bishop Robert of Lincoln witnesses the charter, it must be previous to his death, January 10, 1123, But as the king had not been at Westminster for some time before that, it cannot be placed later than 1122. Now, we have seen that in April-May, 1121, Robert was not yet Earl of Gloucester; consequently, this charter must belong to the period between that date and the close of 1122. It is, therefore, the earliest mention, as yet known to me, of Robert as Earl of Gloucester. As we increase our knowledge of the charters of this reign we shall doubtless be able to narrow further the limit I have thus ascertained.