This is supposed to be another inaccuracy of Master Wace's, and we are told by M. le Prévost that we should read Roger instead of William, the Norman poet having substituted the name of the son for that of the father. That William, the son of Roger de Roumare, was not at Hastings I readily admit, but Wace does not say he was. He simply mentions a "Dam Willame de Romare," and unless we could clearly show there was no such person then existing, it is hardly fair to tax an almost contemporaneous author with even unintentional misrepresentation. The pedigree of the family of Roumare is one of the most puzzling in the whole catalogue of Norman nobility. The diligent study of forty years has not enabled me to penetrate its mysteries. Edward of Salisbury, one of its most important members, has still to be satisfactorily affiliated, and the Roger de Roumare suggested to be substituted for the William of Waee is equally difficult to identify.

It is almost impossible to move a step in these directions without acknowledging our obligations to the late Mr. Stapleton, who has done so much to elucidate the descent of our Anglo-Norman ancestors.

To him we are indebted for the information that previous to the Conquest there lived a certain G-erald, who had two wives, Albreda and Emicia, and a son probably by the first, who is presumed to be the Robert Fitz Gerald of Domesday, and the brother of Roger Fitz Gerald, father of William de Roumare, created Earl of Lincoln by King Stephen.

In my paper on "The Family and Connections of Robert Fitz Gerald," the Domesday holder of Corfe, in the county of Dorset (Congress of the British Archaeological Association, at Weymouth, 1872), I exposed the absurd story, stereotyped in English History, of the three husbands of Lucia, Countess of Chester, which had been first doubted by the Rev. Mr. Bowles in his "History of Laycock Abbey;" but with the particular object of that Paper I have at present nothing to do.

All that we know of Roger Fitz Gerald, also called De Roumare, or De Romara, is that he was the father of the William de Roumare, first of that name, Earl of Lincoln, by a lady named Lucia, who, through the neglect of verifying dates, has been confounded probably with her mother, married to her father before she was born, set down as the sister-in-law of her own son, and thus innocently made the cause of considerable trouble to the learned and curious in history and genealogy. The first fact we are in possession of respecting Roger Fitz Gerald is his appearance as Lord of Spalding in the county of Lincoln, before the death of Eufus in 1100. The date of his marriage is unknown, but his son William must have been of full age in 1122, as in that year he claimed of King Henry 1 certain lands which his step-father, Ranulf de Briquessart, had surrendered to the King for the earldom of Chester. It is clear, therefore, that Roger was dead and William twenty-one and upwards in 1122, so that the latter could not possibly have fought at Senlac, seeing that he was not born till at least thirty years after it.

It is a question, indeed, whether his father Roger de Roumare was present at Hastings, as we find him Lord of Spalding thirty-four years afterwards, and are informed that he was a young man newly married at that period, and I am not aware of any reliable evidence to the contrary.

But, as I have already observed, there is nothing in what we do know to disprove the statement of Wace, that there was a William de Roumare in the ranks of the Norman army of invasion. Without relying on the statement of Peter de Blois, that Roger Fitz Gerald had an elder brother named William, by whom Lucia was honourably received on her marriage, and whom the writer inaccurately styles Earl of Lincoln, there is every probability that such was the fact.

Gerold de Roumare, the presumed father of Roger, had two wives — Albreda and Emicia; but we have no information whatever that can be relied on respecting the number of his offspring, or, with the exception of Robert, of which of his wives they were the issue.

The above little but important fact is derived from a charter printed in Pommeraye's "Histoire de l'Abbaye de St. Amand de Rouen," fol. 1662, in which a knight named Gerold gives to the Abbey of St. Amand the Church of Roumare for the sake of his own soul and that of his wife Albreda, with the assent of his son and heir Robert, and the attestation of Ralph, brother of Gerold.

The son Robert is supposed to be the Robert fitz Gerald of "Domesday," and the brother Ralph the Chamberlain of Tankerville, of whom I shall have to speak presently. Roger is not mentioned, nor any William; but if there was a William de Roumare, an elder brother, he would at the time of the Conquest be "Dom (Dominus) William de Romare," and dying unmarried before the compilation of "Domesday," no traces might have been left of him. At all events I have found nothing to justify the rejection of Wace's statement, and therefore leave the name of William at the head of this chapter as a companion of the Conqueror, convinced that there might be a Robert, but certainly not a Roger, Fitz Gerald in the host at Hastings.

Added to this site through the courtesy of Fred L. Curry, who provided a photocopy of the section.