It is generally believed that Scrivelsby and the rest of the Honour of Dispenser came to the Marmions through the marriage of Roger Marmion's grandson,[31] Robert Marmion, who was the husband of Matilda de Beauchamp, the grand-daughter of Urso de Abitot, and grand-niece of Robert Dispenser. But the Roll proves that Roger Marmion was the immediate heir of Robert Dispenser (p. 14).
I know of no such general belief. Stapleton, to whom one would naturally turn, had pointed out long before, in his 'Rolls of the Norman Exchequer', that this survey proves Roger Marmion to have held the Lincolnshire fief of Robert 'Dispensator',[32] while those who have identified the latter magnate with Robert 'Marmion' have traced the descent of Scrivelsby in the Marmions even from the Conquest.[33]
In any case, as I wrote in my Ancient Charters (1888) of a document there published:
The succession of Urse [de Abetot] to this [Lincolnshire] fief is a genealogical discovery which throws a wholly new light on the very difficult problem of the relation of Marmion to Despenser, and is fatal to the assertion of Mr Chester Waters that 'Roger Marmion was the immediate heir of Robert Dispenser'.
Moreover, in the Leicestershire Survey,[34] and still more in that of Worcestershire,[35] we have evidence that Robert's inheritance was shared between Beauchamp and Marmion which points there also to descent through Urse de Abetot. In my Geoffrey de Mandeville (pp. 313-5) I have suggested that in their rivalry for Tamworth,[36] the Marmions embraced the cause of Stephen, and the Beauchamps that of Maud, their variance being terminated under Henry II by a matrimonial alliance. Such a compromise was common enough. It was agreed on in the case of Grantmesnil; it was carried out at this very period in that of Fitzharding and Berkeley; it was again resorted to at a later stage in the history of the house of Berkeley; it was arranged in the case of Hastings; and it was repeated in that of Boleyn, where the Butler inheritance was at stake.[37]
[1] English Historical Review, v. 96.
[2] I have discussed above (pp. 69-72) the bearing of its evidence on the problem of Domesday assessment, so need not recur to the subject here.
[3] See note 31 below.
[4] Vol. II. p. xcvi.