Post by Peter G. M. Dale
Many thanks again Peter.
I set out below an extract from the website, 'Foundation for Medieval
Genealogy, Medieval Lands - A prosopography of medieval European noble
and royal families, Untitled English Nobility P-S', which provides certain
information on the identify of Eudo's wife Rose.
Does this assist in clarifying her identify?
"... m ROHESE, daughter of RICHARD FitzGilbert de Brionne & his wife
Rohese Giffard (-7 Jan 1121, bur Le Bec, Normandy [Domesday
Descendants, p. 400.]). "Eudo dapifer domini regis" founded Colchester
St John, for the souls of King Henry I, Queen Matilda "...uxore mea
Roaysia", by undated charter [Colchester St John, Vol. I, p. 1.].
Her parentage is confirmed by the undated charter under which "Rohais
uxor Eudonis dapiferi" donated "manerium de Halingberi sicut dominus
meus Eudo die qua vivus et mortuus fuit illud habebat" and land which
"Gelebertus frater meus" gave her, for the souls of "Eudonis dapiferi
mariti mei et Gilberti fratris mei" [Colchester St John, Vol. I, p.
48. ], which is corroborated by the undated charter under which
"Walterus filius Roberti" donated "terram de teia" to Colchester St.
John, for the souls of "patris mei Roberti filii Ricardi et matris mee
Matildis et...Rohaise amite mee que ecclesiam Sancti Johannis fundavit
et fratrum suorum", to Colchester St. John [Colchester St John,
Vol. I, p. 165.]. The History of the foundation of St John's abbey,
Colchester also names "Eudoni...major domus regiæ" and "Roasya uxor
eius...Gilbertum comes, Rohaisæ frater" [Dugdale Monasticon IV,
Colchester St John Abbey, Essex, I, Historia Fundationis, p. 607.].
Other sources suggest a different parentage for Rohese. According to
Guillaume de Jumièges and the Genealogia Fundatoris of Tintern Abbey,
she was Rohese, widow of Richard FitzGilbert de Brionne, daughter of
Gauthier Giffard & his wife Ermengarde (-after 1113, bur [Colchester]).
Guillaume de Jumièges names "Galterium Giffardum primum" as father of
"secundum Galterium Giffardum et filias plures" of whom "una...Rohais"
married "Richardo filio comitis Gisleberti" [Willelmi Gemmetensis
monachi Historiæ Normannorum, Du Chesne, A. (1619) Historiæ Normannorum
Scriptores Antiqui (Paris) ("Willelmi Gemmetencis Historiæ (Du Chesne,
1619)"), Liber VIII, XXXVII, p. 312.]. According to the Genealogia
Fundatoris of Tintern Abbey, Monmouthshire, "Rohesia" married secondly
"Eudoni dapifero Regis Normanniæ" after the death of "Ricardo filio
comitis Gisleberti" and that they were both buried "tempore Henrici
primi" in "castrum Clecestriæ...coenobio in honore sancti Johannis"
which Eudo constructed [Dugdale Monasticon V, Tintern Abbey,
Monmouthshire III, p. 269.]. The Complete Peerage says that this
parentage is "probably erroneous" [CP V 113-4.]. From a
chronological point of view, the connection would be tight, assuming
that the death date of Richard FitzGilbert is correctly estimated to
 and the birth of Rohese's granddaughter by her alleged second
marriage, Beatrix, is correctly assessed at . This supposed
different parentage is disproved by the three sources quoted above.
i) MARGUERITE ([1080/90]-). The Genealogia Fundatoris of
Tintern Abbey, Monmouthshire names "Margareta" as daughter of "Eudoni
dapifero Regis Normanniæ" and "Rohesia", adding that she married
"Willielmo de Mandavill" by whom she was mother of "Gaufridi filii
comitis Essexiæ et iure matris Normanniæ dapifer" [Dugdale
Monasticon V, Tintern Abbey, Monmouthshire III, p. 269.]. According
to the Complete Peerage, this genealogy is "probably erroneous" but it
does not explain the basis for the doubts [CP V 113-4.].
Marguerite's second marriage is suggested by the charter dated [1141/42]
under which Empress Matilda made various grants of property including a
grant to "Willelmo filio Otuel fratri...Comitis Gaufredi" [Round
(1892), p. 169.]. The only Ottiwell has been identified was the
illegitimate son of Hugh Earl of Chester. m firstly ([1100/05]) WILLIAM
de Mandeville, son of GEOFFREY de Mandeville & his first wife Adelisia
--- (-). m secondly ([1116/19]) OTTIWELL, [maybe OTTIWELL
FitzHugh, illegitimate son of HUGH Earl of Chester & his mistress ---]
(-drowned off Barfleur, Normandy 25 Nov 1120)."
The trouble with evidence from Colchester charters is that many of them are not reliable - the cartulary was compiled in the 13th century and some documents were substituted in the 14th. In 1911 Armitage Robinson concluded that ' the compiler or compilers of these forgeries must have had a number of genuine documents, which, though insufficient for the purposes contemplated, furnished the necessary historical setting ' (*Gilbert Crispin, Abbot of Westminster* p. 166). The Clare family connection may be true, but evidence independent of St John's abbey would be needed to establish the facts beyond question. Simply quoting snippets from dubious charters without context, as in the Medieval Lands database above, gives a false impression of certainty. The fact that the charter of Rohese does not call her brother Gilbert 'count' as in the foundation history, and the charter of Walter fitz Robert calling her his 'amita', suggest that these may be genuine documents, but these two sidelights are short of conclusive.
Post by Peter G. M. Dale
As per the consanguinity issue, I do not know enough to comment.
However, your statement that you have not seen similar cases of 2nd
cousins marrying in the 12th century (or dispensation provided
therefore) suggests that Geoffrey II de Mandeville is unlikely to be
the grandson of Rose fitz Richard de Clare, even if she was Eudo's
wife. For clarity, Rose fitz Richard de Clare, even if she was Eudo's
wife, is unlikely to be the mother of Margaret, dau. of Eudo.
The possibility of Geoffrey de Mandeville taking his rights to Eudo's estates and dapiferate from a collateral relationship rather than direct descent is very limited. Eudo had several brothers: of these, Adam is excluded as the maternal grandfather of Geoffrey since Eudo was his heir; Robert was a bishop; Hubert had male heirs of his own. The only one left who could have been Geoffrey's grandfather, giving him seniority as Eudo's heir, was Radulf, castellan of Nottingham - but in that case, why wasn't his castellany part of the inheritance that Geoffrey was entitled to?
Rohese de Vere's brother William, bishop of Hereford & chancellor, described their mother as 'Adeliza, filia Gilberti de Clare', so there is not much room for error there.
I think your caution about Geoffrey de Mandeville's maternal grandmother is warranted, and Keats-Rohan's unqualified assertion is too strong - without further evidence it is probably impossible to resolve this.