[56] The point is of some importance in its bearing on the right of the individual to assess himself, which is held in this case to have been exercised. 'The assessment,' writes Dr Stubbs, 'of the individual depended very much on his own report, which the exchequer had little means of checking.'—servitia, i. 585.
[57] By one of those slips so marvellously rare in his writings Dr Stubbs writes that 'the Bishop of Durham's service for his demesne land was that of ten knights, but it was not cut up into fees' (i. 263). What the bishop said was that he owed no service for his demesne, because there were already over seventy fees created on his fief, though he only owed ten.
[58] This is one of the points on which Madox is completely at sea. He quotes the case of the Bishop of Durham (1168) as an instance of 'Doubts about the number of knights' fees' (Baronia Anglica, p. 122); and he writes, of the above uniform formula: 'This uncertainty about the number of the fees frequently happened in the case of ecclesiastical persons, Bishops, and Abbots.'—Exchequer, i. 647.
[59] servitia, i. 264.
[60] See my papers on 'The House of Lords; the Transition from Tenure to Writ' (Antiquary, October and December 1884, April 1885).
[61] See, for instance, the language used in the carta of Ralf de Worcester (p. 441): 'Teneo de vobis in capite de veteri fefamento feodum i. militis, unde debeo vobis facere servitium i. militis. Et de eodem feodo Jordanus Hairum debet mihi facere iiii.am. partem servitii,' etc. In Normandy (1172), the phrase ran: 'quot milites unusquisque baronum deberet ad servicium regis, et quot haberet ad suum proprium servicium'.
[62] Sometimes Exeter pays on 15½ (14, 33, Hen. II), but 17½ (2, 5, 7, 18 Hen. II) is the normal amount. The explanation of this odd number is found in the Testa de Nevill (p. 226) where ('Veredictum militum de Rapo de Arundel') we read: 'Episcopus Exoniensis tenet de Domino Rege de Capellaria de Boseham vii. feoda militum et dimidium.' The Bosham estate (as belonging to Osbern) had formed part of the episcopal fief in Domesday, but (the bishops having founded a church there) we find it assessed and paying separately as 7½ fees.
[63] I have found a case bearing upon this point and reported at great length (Thorpe's Registrum Roffense, pp. 70 et seq.). It arose from an attempt of the Archbishop of Canterbury, in 1253, to distrain the Bishop of Rochester for the 'auxilium ad filium regis primogenitum militem faciendum'. The bishop 'posuit se super recordum rotulorum de Scaccario, per quos rotulos poterit et illa quam rex contra episcopum et etiam illa quam archiepiscopus contra episcopum movit questio diffiniri. Didicerat enim episcopus per unum fidelem amicum quem in scaccario tunc habebat quod nunquam tempore alicujus regis pro aliquo feodo episcopatus aliquod fuit regi factum servicium vel datum scutagium.... Unde consulebat quod audaciter poneret se episcopus super recordum rotulorum de Scaccario, nichil enim tenet episcopus per baroniam de rege, sed per puram elemosinam, quod non est dicendum de aliquo episcopatu Anglie, nec de Archiepiscopatu, nisi dumtaxat de Karleolen. Cumque cum audacia institisset episcopus, quod decideretur per rotulos de Scaccario quibus creditur in omnibus illis sicut sancto evangelio', etc., etc. The barons of the exchequer examined the rolls, 'a tempore primi conquestus' (?) and reported: 'nusquam invenerunt episcopum Roffensem solvisse aut dedisse aliquod servicium regibus temporale'. But the dispute was not finally decided till 1259. The clue to the matter is found in the Canterbury 'Domesday Monachorum' (8th Report Hist. MSS. i. 316), where a list of the archbishop's knights, perhaps coeval with Domesday (vide infra, p. 236), is headed by 'Episcopus Roffensis' with a servitium of ten knights to the Primate.
[64] Cerne had to provide 'ten' knights ad wardam at Corfe Castle, or 'two' ad exercitum (vide cartam).
[65] This indeed is proved by an extract quoted by Madox (Exchequer) from the Roll of 22 Hen. II (rot. 10a).