^{page 225 note 1} For the Alpine routes that radiated from north Italy, see Baker, J. N. L., ‘Medieval Trade Routes’, Social Life in England, ed. Barraclough, G., London 1960, 224–46Google Scholar, esp. 231–4. It should be remembered that the St. Gotthard Pass was not opened until the thirteenth century.

^{page 226 note 1} In his study of the church of Sion, J. Gremaud gives the dates of the death of Ermenfrid's predecessor, Aimo, as 13 July 1054, and of Ermenfrid's succession as 1055: Mémoires et documents publiés par la société d'histoire de la Suisse Romande, xviii (1863), 496Google Scholar; but he offers no evidence in support. Ermenfrid was concerned with Norman affairs at about this time, and the possibility that he may have succeeded Aimo in 1054 cannot be ruled out. It is uncertain by whom he was promoted to the see, although both before and after his time it was controlled by the counts of Savoy: Previté-Orton, C. W., The Early History of the Counts of Savoy, Cambridge 1912, 93–4.Google Scholar

^{page 226 note 2} See Jacob, L., ‘Le royaume de Bourgogne sous les empereurs franconiens (1038–1125)’, Bulletin de la société de statistique du départment de l'Isère, 4th ser., x (1908), 57–8Google Scholar. Aimo was a son of count Humbert Whitehands of Savoy: Previté-Orton, op. cit., 29, 63.

^{page 226 note 3} See Cibrario, L. and Promis, D. C., Documenti, sigilli e monete appartenenti alla storia della monarchia di Savoia, Turin 1833Google Scholar, Documenti, 34–5. Ermenfrid witnessed as ‘Ermenfredus sedunensis episcopus et s. Mauritii canonicus’.

^{page 226 note 4} On 6 May 1065, he witnessed at Rome a privilege of pope Alexander II in favour of St. Denis: Migne, P.L., cxlvi. 1306–9, and in 1072 a privilege of the same pope for the church of Châlons-sur-Saâne: P.L., cxlvi. 1377–9. If Ermenfrid was one of the unnamed papal agents whom, according to William of Malmesbury, pope Nicholas II dispatched from Rome in 1061 with archbishop Ealdred of York, this would establish a third visit: see below, 229, n.1.

^{page 227 note 1} See esp. Schieffer, T., Die päpstlichen Legaten in Frankreich vom Vertrage von Meersen (870) bis zum Schisma von 1130, Historische Studien, cclxiii, Berlin, 53–5Google Scholar. The date of Ermenfrid's legatine mission to Normandy depends on that of the council of Lisieux at which archbishop Mauger was deposed. The Norman chronicles are unanimous in giving the date as 1055: e.g., Annales Uticenses, appended to Ordericus Vitalis, Historiae ecclesiasticae, ed. le Prévost, A., Paris 1838–55, v. 157Google Scholar; Chronicon sancti Stephani Cadomi, Recueil des historiens des Gaules et de la France, ed. Bouquet, M. (= R.H.F.), xi (1767), 379Google Scholar. But Professor Douglas has powerfully argued that 1054 is the more probable date: Douglas, D. C., William the Conqueror, London 1964, 69Google Scholar. The testimony of the Acta archiepiscoporum Rotomagensium that Mauger was deposed ‘auctore papa Leone’ suggests that the chain of events which ended with his deposition may have begun, at least, before Leo IX died on 19 April 1054: P.L., cxlvii. 277–8. It may be added that the ducal charter of 1055 referred to by Schieffer which purports still to name Mauger as archbishop, and which appears in Gallia Christiana (= G.C.) xi, 2nd ed., Paris 1874Google Scholar, instrumenta, 13, is a late forgery which is unlikely to depend on anything save the chronicles: Recueil des actes des dues de Normandie de 911 à 1066, ed. Fauroux, M., Mémoires de la société des antiquaires de Normandie, xxxvi, Caen 1961, no. 136, 311–12.Google Scholar

^{page 227 note 2} Douglas, op. cit., 38–43.

^{page 227 note 3} Ibid., 75–80.

^{page 227 note 4} For Mauger's character, see, e.g., Ordericus Vitalis, Hist. eccles., v. 9, ed. le Prévost, ii. 367; William, of Poitiers, , Histoire de Guillaume le Conquérant, ed. Foreville, R., Paris 1952, 130–2Google Scholar; Acta archiep. Rotomag., loc. cit.

^{page 227 note 5} Chronicon Sithiense, R.H.F., xi. 382; cf. William, of Malmesbury, , Gesta regum Anglorum, iii. 267, ed. Stubbs, W., ii, R.S., London 1889, 327.Google Scholar

^{page 227 note 6} For Maurilius, see Douglas, op. cit., 120–2.

^{page 227 note 7} The canons have not survived, but they are referred to in the account of the council of Rouen (1072) in Ordericus Vitalis, Hist. Eccles., iv. 9, ed. le Prévost, ii. 240–2. The fullest account of the council is in G.C., xi. 28–31.

^{page 228 note 1} See Schieffer, op. cit., 60. For a description of events, see G.C., x (1751), instrumenta, 22–4.

^{page 228 note 2} See Tillmann, H., Die päpstlichen Legaten in England bis zur Beendigung der Legation Gualas (1218), Bonn 1926, 11–12.Google Scholar

^{page 228 note 3} See the professions of faith to Lanfranc by bishops Remigius of Lincoln: Geraldi Cambrensis Opera, vii, ed. Dimock, J. F., R.S., London 1877, Appendix A, 151–2Google Scholar, and Wulfstan, of Worcester, : The Vita Wulfstani of William of Malmesbury, ed. Darlington, R. R., Camden Soc., 3rd Ser., xl (1928)Google Scholar, Appendix, 190.

^{page 228 note 4} See Douglas, op. cit., 168–70; Stenton, F. M., Anglo-Saxon England, 2nd ed., Oxford 1947, 458–62.Google Scholar

^{page 228 note 5} Florence, of Worcester, , Chronicon ex chronicis, ed. Thorpe, B., i, London 1848, 220.Google Scholar

^{page 229 note 1} In his Vita Wulfstani, William of Malmesbury offers another account of events at Worcester: ed. Darlington, i. 10–11, 16–18. Pope Nicholas II had forbidden bishop Ealdred of Worcester to retain the see when he was translated to York in 1061. During a visit to Rome, Ealdred agreed to surrender it if he might be succeeded by the worthiest candidate from his own diocese. To implement this agreement, Ealdred was accompanied to England by papal agents whom William calls ‘cardinales’ and does not name. In Lent, 1062, after attending to much other business in England, they came to Worcester and met Wulfstan, whom, at the Easter court, they warmly recommended to be bishop. Nicholas II died in Rome on ?22 July 1061 and Alexander II was elected there on 29 or 30 September. Either pope may, therefore, have initiated the sending of the legates.

^{page 229 note 2} Tillmann, op. cit., 12–15.

^{page 229 note 3} Hist. Eccles., iv. 6, ed. le Prévost, ii. 199. The deposition of Stigand is discussed by Chibnall, M., in her new edition of Ordericus: The Ecclesiastical History of Orderic Vitalis, ii, Oxford 1969, xxxii–iv.Google Scholar

^{page 229 note 4} Crispin, Milo, Vita Lanfranci, 12, Opera Lanfranci, ed. Giles, J. A., i, Oxford 1844, 291.Google Scholar

^{page 229 note 5} Ordericus Vitalis completed his history between 1123 and 1141 in the Norman monastery of St. Evroul; Milo Crispin, precentor of Bee, also wrote during the first half of the twelfth century. They used a common source: Chibnall, loc. cit.

^{page 229 note 6} Hist. Eccles., loc. cit.

^{page 229 note 7} Vita Lanf., 15; ed. Giles, 292. Cf. also an addition to the Durham abridgement of William of Malmesbury's Vita Wulfstani, 12: Vita Wulfstani, ed. Darlington, 77. But, apart from the lateness of the addition, the councils of Winchester and Windsor are confused under the name of Westminster, and no new evidence is provided.

^{page 229 note 8} Hist. eccles., loc. cit., but cf. his retrospective phrase, iv. 6, ii. 212.

^{page 230 note 1} Vita Wulfstani, ed. Darlington, Appendix, 189–90. Bishop Wulfstan and the abbots of his diocese were summoned to attend ‘tercia die post proximum pascha’. It is, of course, possible that the delay until the following Sunday was occasioned by the late arrival of Ermenfrid.

^{page 230 note 2} Chron., ed. Thorpe, ii (1849), 5–6. Whoever its author was, this chronicle preserves an early tradition which reflects the outlook of English men of substance around Worcester, where it was compiled. The early departure of the cardinals is implicitly confirmed by Lanfranc, Epistolae, 3; Op. Lanf., ed. Giles, i. 19–21.

^{page 230 note 3} For Leofwine, cf. Lanfranc, Ep., 4; Op. Lanf., i. 22–3.

^{page 230 note 4} Fl. Worc., Chron., loc. cit.

^{page 230 note 5} Ermenfrid was certainly assisted at Windsor by another legate, if not by more than one: Milo Crispin, Vita Lanf., 15; Op. Lanf., ed. Giles, i. 293; Lanfranc, Ep., 9; ibid., 31. When he went to Normandy to persuade Lanfranc to accept the see, he was accompanied there by Hubert, cardinal of the Roman church: Lanfranc, Ep., 3; ibid., 19. Hubert may well also have been at Windsor. He was repeatedly employed as a legate to the Anglo-Norman kingdom between 1070 and 1085: see Tillmann, op. cit., 15–17.

^{page 230 note 6} Lanfranc, Ep., 9; ibid., 31; cf. Fl. Worc., Chron., ii. 6.

^{page 230 note 7} Fl. Worc., Chron., ii. 7.

^{page 230 note 8} Loc. cit.; Ord. Vit., Hist. eccles., iv. 6, ed. le Prévost, ii. 200. Archbishop Ealdred of York had died on 11 September 1069. According to ‘Florence of Worcester’, Walkelin was consecrated ‘iussu regis’ by Ermenfrid of Sion on 30 May: Chron., ed. Thorpe, ii. 7. William probably wished to delay the consecration of Thomas of York until it could be performed by an archbishop of Canterbury.

^{page 231 note 1} Fl. Worc., Chron., ii. 6–7.

^{page 231 note 2} Milo Crispin, Vita Lanf., 15; Op. Lanf., ed. Giles, i. 293; where, after confusion with regard to the Eastertide council between Winchester and Windsor, Lanfranc's nomination is associated with a later, and unnamed, council.

^{page 231 note 3} Loc. cit.; Lanfranc, Ep., 3; Op. Lanf., ed. Giles, i. 19–20; Ord. Vit., Hist Eccles., iv. 6, ed. le Prévost, ii. 212–3.

^{page 231 note 4} Eadmer, , Historia novorum in Anglia, i, ed. Rule, M., R.S., London 1884, 10.Google Scholar

^{page 231 note 5} Although on 2 March 1072, he was present as an ordinary bishop at the council of Châlons-sur-Saône which met under the presidency of cardinal Gerald of Ostia to settle a dispute between the bishops of Valence and Romans: Martène, E. and Durand, U., Thesaurus novus anecdotorum, iv, Paris 1717, 97–8.Google Scholar

^{page 231 note 6} Before 1070, only one piece of evidence records Ermenfrid's presence within the boundaries of the empire. In October 1057 he was present at Tribur or Speyer when the new bishop of Speyer, Gundechar, received his pastoral staff, and he was also at Speyer when Gundechar was enthroned: Gundechar, Liber pontificalis Eichstetensis, ed. L. C. Bethmann, Monumenta Germaniae Historica (= M.G.H.), Scriptorum, vii. 245–56.

^{page 232 note 1} G.C., iii (2nd ed., 1876), instrumenta, 151. For the matter at issue, see Triumphus sancti Remacli de Malmundariensi coenobio, ed. D. W. Wattenbach, i. 22, ii. 1–38, M.G.H., Scr., xi. 449–61.

^{page 232 note 2} For Ermenfrid's presence at Mainz and for an account of the council, see Udalrici codex, 37, Monumenta Bambergensa, ed. Jaffé, P., Bibliotheca rerum Germanicarum, v, Berlin 1869, 70–7.Google Scholar

^{page 232 note 3} ‘… orantes et deprecantes, ut de regno suo eliminet omnem apostasiam et omnia scandala, nec ultra sathan eiusque apostolos in sua sancta regnare sinat ecclesia; sed in omnibus consiliis et operibus eorum dignetur servos suos consolari et clementer implere, quod ecclesiae suae se promisit affuturum omnibus diebus usque ad consummationem seculi’: ibid., 76.

^{page 232 note 4} The forged diploma of Henry IV which purports to have been issued at Worms on 27 March 1076 to confirm the foundation of the Cluniac priory of Rüggisberg (cant. Berne) ‘ob interventu … Sedunensis episcopi Ermenfredi’ is too unreliable to be used as evidence without further support: Heinrici IV diplomata, ed. D. von Gladiss, no. 281, M.G.H., Diplomata, vi, pt. 1, 362–5.

^{page 232 note 5} Ibid., no. 321, M.G.H., Dipl., vi, pt. 2, 421–2. Ermenfrid is described as ‘fidelis nostrorum charorum fidelium defunctorum apud Dominum intercessor’.

^{page 233 note 1} S. Guichenon, Histoire généalogique de la royale maison de Savoie, vi, Lyons 1660, 19. For the dispute, see Previté-Orton, op. cit., 233.

^{page 233 note 2} This office had been vacant since the death in 1066 of Ermenfrid's partner as legate in 1059, archbishop Hugh of Besançon.

^{page 233 note 3} Hein. IV. dipl., no. 344, M.G.H., Dipl., vi, pt. 2, 454–6.

^{page 233 note 4} Ibid., no. 397, 525–6. The date and place raise problems, but the date is unlikely to be earlier than 1087.

^{page 233 note 5} In the Necrology of the cathedral of Sion, Ermenfrid was remembered on 10 December, but there is no indication of the year of his death. Gremaud proposed the year 1088, but without stating his authority: Mém. et doc., xviii (1863), 289, 496.Google Scholar

^{page 233 note 6} The ordinance itself bears no date, and the customary assigning of it to 1070 depends entirely on the fact that Ermenfrid's presence in Normandy in that year is otherwise well attested in connexion with the election of Lanfranc to Canterbury. Another visit during the previous three years cannot be ruled out. Even if the date 1070 is correct, Ermenfrid may have confirmed a document which the bishops had drawn up at a considerably earlier time. I owe to the Rev. C. Morris some very cogent arguments for thinking that this may have been so. 1070 was a very long time after the battle to wait for the issue of regulations, and it is odd that the Norman bishops should then have been making regulations for warriors, many of whom were, by this time, virtually residents of England. Moreover, the distinction of periods (Hastings, before the coronation, and after) suggests immediacy. It may well be that the ordinance was first issued in 1067, on William's return to the duchy, e.g., at the Easter court at Fécamp: William of Poitiers, Hist. de Guill. le Conq., ed. Foreville, 260.

^{page 234 note 1} For the text, see Spelman, H., Concilia, decreta, leges, constitution's in re ecclesiarum orbis Britanniae, ii, London 1664, 12–13Google Scholar. Spelman states that he took it ‘from an old Saxon book formerly belonging to the church of Worcester’. To this, J. Marsham, in his Propylaion to Dugdale's Monasticon, adds the information that the book, which had been seen by Gerard Langbaine the elder (1609–58), provost of Queen's College, Oxford, appeared to have been written in a Saxon hand about the time of William the Conqueror, by a Saxon scribe at Worcester, named Wulfgeat: Dugdale, W., Monasticon Anglicanum, i, 2nd ed., London 1817, xivGoogle Scholar. The MS. has been lost since the seventeenth century. There is no evidence to suggest that the slightly emended version in Wilkins, D., Conciliae Magnae Britanniae et Hiberniae, i, London 1737, 366Google Scholar, is anything more than a copy of Spelman's text, emended by Wilkins's own unsupported conjecture. The versions which were published on the continent depend on Spelman's printed text and have no independent MS. support: Spelman's text was copied by Labbe, P. and Cossart, G., Sacrosancta concilia ad regiam editionem exacta, x, Paris 1671, 352–3Google Scholar, and thence by Bessin, G., Concilia Rotomagensis provinciae, Rouen 1717, 50–1Google Scholar. For Spelman's text, see the Appendix to this paper.

^{page 234 note 2} Thus, in current standard books, its authenticity is fully accepted by Sir Frank Stenton, Anglo-Saxon England, 653–4. D. C. Douglas, however, expresses reservations: William the Conqueror, 191–2.

^{page 234 note 3} Bishop Odo of Bayeux's part in the battle of Hastings is familiar from the Bayeux tapestry. Cf. number 12 of the thirteen canons which are probably to be assigned to the council of Winchester in 1072: ‘Ut nullus clericus secularia arma ferat’. These canons are also noteworthy, in relation to the Norman ordinance, as calling on the clergy to summon laymen to do penance: see no. 7. ‘Ut episcopi et sacerdotes laicos invitent ad poenitentiam’. They are printed by Spelman, op. cit., ii. 12, and Wilkins, op. cit., i. 365.

^{page 235 note 1} Usually seven years: see, e.g., the provisions of Burchard of Worms and Fulbert of Chartres, below, 240.

^{page 235 note 2} The precise phrase bellum publicum is characteristic of canonist and penitential sources: see below, 239–40; rather than of feudal sources. But for bellum commune, or nominatum, as designations of the legitimate warfare of a feudal superior, see Mitteis, H., Lehnrecht and Staatsgewalt, Weimar 1948 (= 1933), 594Google Scholar. The canonists' ideas are well discussed by Keen, M. H., The Laws of War in the Late Middle Ages, London 1965, 63–72.Google Scholar

^{page 235 note 3} Cf. the firm rules that the Conqueror himself is said to have made for his own forces in England, ‘frenans ut populos armis, ita legibus arma’: William of Poitiers, Hist. de Guill. le Conq., ed. Foreville, 232; Ord. Vit., Hist. eccles., iv. 1, ed. le Prévost, ii. 164–5.

^{page 236 note 1} This ordinance was printed from a MS. preserved at Troyes by du Chesne, A., Historiae Francorum scriptores, ii, Paris 1636, 588.Google Scholar

^{page 237 note 1} For discussions, see esp. Müller, K., ‘Der Umschwung in der Lehre von der Busse während des 12. Jahrhunderts’, Theologische Abhandlungen Carl von Weizsäcker gewidmet, Freiburg-im-Breisgau 1892, 287–320Google Scholar; Poschmann, B., Die abendländische Kirchenbusse im frühen Mittelalter, Breslau 1930, esp. 116–66Google Scholar, and Der Ablass im Licht der Bussgeschichte, Bonn 1948, 1–63Google Scholar; Amann, É., L'époque carolingienne (Histoire de l'église, ed. Fliche, A. and Martin, V., vi), Paris 1947, 346–52Google Scholar. Mayer, H. E., Geschichte der Kreuzzüge, Stuttgart 1965, 31–4Google Scholar, gives a useful summary of recent work.

^{page 237 note 2} Liber Gomorrhianus, x; P.L., cxlv. 169.

^{page 237 note 3} P.L., cxl. 949–1014.

^{page 237 note 4} P.L., cxli. 339–40.

^{page 238 note 1} Freeman, E. A., The History of the Norman Conquest of England, iv, 2nd ed., Oxford 1876, 810–11Google Scholar, expressed doubts about the credibility of the penitential ordinance because, by its terms, the king, to whom it made no reference of any kind, would presumably have incurred a penance of two thousand years for the two thousand Englishmen whom he was reputed to have slain at Hastings. The hyperbole of the figure apart, it may first be commented that the king was not necessarily in worse case than any of his followers who had performed comparable prodigies of slaughter. In practice, as the ordinance following the battle of Soissons (see above, 236, n. 1) and clause 3 of the Norman ordinance both suggest, penances were commonly commuted for acts of almsgiving. The Chronicle of Battle Abbey indicates that, amongst the Conqueror's motives for its foundation, the need to atone for the gravest sins of blood-shedding at Hastings had its place. Before the battle, William vowed that ‘coenobium … pro omnium, illorumque nominatim qui in eodem bello occumberent, salute construeret, qui locus refugii et auxilii omnibus esset, quatinus iugi bonorum operum instantia commissa illic effusi cruoris redimerentur’: Chronicon monasterii de Bello, Anglia Christiana Society, London 1846, 22. Thus, the founding of Battle may have represented the discharge of William's personal debt. However, the Chronicle does not explicitly relate it to his own killings and, apart from such exceptional cases as the emperor Henry IV at Canossa and king Henry II of England after Becket's murder, little is known about the penance of kings. In the eleventh century, they may have enjoyed a privileged position on account of the contemporary understanding of their special duty, as expressed in the Christian liturgy, to defend the Church and Christendom. For the long Christian reluctance to look with favour upon the profession of arms, for the special position of kings and for the changing attitudes of die eleventh century, see Erdmann, C., Die Entstehung des Kreuzzugsgedankens, Stuttgart 1935, 1–85Google Scholar; Hirsch, H., ‘Der mittelalterliche Kaisergedanke in den liturgischen Gebeten’, Mitteilungen des österreichischen Instituts für Geschichtsforschung, xliv (1930), 1–20Google Scholar; and Erdmann, ‘Der Heidenkrieg in der Liturgie und die Kaiserkrönung Ottos I’, ibid., xlvi (1932), 129–42; It may also be suggested, as a speculation, that William's prohibition of capital punishment may have had, as its background, the canonists' attitude to homicide: ‘Laws of William the Conqueror’, 10, Select Charters, ed. Stubbs, W., 9th ed., Oxford 1913, 99.Google Scholar

^{page 238 note 2} E.g., in Abbo of Fleury's Collectio canonum (c. 988–96), 50–1, P.L., cxxxix. 506. Abbo quotes with approval the pseudo-Augustinian dictum that ‘militare non est delictum; sed propter praedam militare peccatum est’, and comments that proper wages are assigned to those who fight in order to leave no need for spoils. In the light of Luke iii. 14, a knight should not seek more than the ‘stipendia sibi publice decreta’. Mercenary service is thus recognised as being fully right and honourable.

^{page 238 note 3} P.L., xl. 1113–30. This manual, falsely ascribed to St. Augustine, was probably compiled in the eleventh century.

^{page 238 note 4} Ibid., 1125.

^{page 238 note 5} See the decrees of the Lateran council of November 1078, Gregorii VII Registrum, vi. 5b, ed. Caspar, E., M.G.H., Epistolae selectae, ii, Berlin 1923, 404Google Scholar; his letter to his standing legate Amatus of Oloron of November 1079, Reg., vii. 10, 471–2; and the decrees of the Lateran council of Lent, 1080, Reg., vii. 14a, 481–2.

^{page 239 note 1} vi. 23, P.L., cxl. 770–1.

^{page 239 note 2} iv, P.L., cx. 471–3. The date of its compilation is 841.

^{page 239 note 3} Deut. xix. 4–10.

^{page 239 note 4} Exod. xxi. 14; cf. Deut. xix. 11–12.

^{page 239 note 5} William of Poitiers's statement that knights were drawn to join William from many parts of France, partly by the well-known liberality of the duke and partly by confidence in the justice of his cause, indicates the mixture of motives among his warriors: Hist. de Guill. le Conq., ed. Foreville, 150.

^{page 240 note 1} P.L., clx. 951–6.

^{page 240 note 2} P.L., cxli. 339–40.

^{page 240 note 3} If this interpretation of the penitential ordinance is correct, it cannot be used as evidence that the Conqueror's army at Hastings consisted partly of men who fought from feudal duty and partly of men who fought for pay (a conclusion for which there is, however, plenty of other support): see esp. Prestwich, J. O., ‘War and Finance in the Anglo-Norman State’, Transactions of the Royal Historical Society, 5th ser., iv (1954), 24–5Google Scholar; Hollister, C. W., The Military Organization of Norman England, Oxford 1965, 167Google Scholar. Neither Mr. Prestwich nor Professor Hollister examines the text in detail, but two sections are relevant, (i) The introduction. With or without Wilkins's emendation, which should almost certainly be dismissed as conjectural, the text cannot be confidently interpreted. However, the words ‘suo iussu, et qui ante (?absque) iussu sui erant’ recall Burchard of Worms's contrast between homicides in war which were and which were not committed ‘iussu legitimi principis’: see above, 240; on this analogy, they might well relate to those who killed men under due orders or not (cf. clauses 10 and 11), rather than to those originally armed, or not armed, at the duke's orders—a consideration which, from the bishops' point of view, would have been irrelevant. The words ‘ex debito ei militiam debebant’ might be the remains of an antithesis between feudal service for duty and mercenary service for money, but they might also refer to all forms of duty, however they arose. Perhaps the most economical sense of this hopelessly corrupt passage is that it refers to whose whom duke William led and who owed him military service as a duty, and who killed or wounded others, whether by his command or not. In any case, no satisfactory conclusion about the composition of the duke's army can be based upon it. (ii) Clauses 6 and 7. Because of its purpose as a penitential, the ordinance is primarily concerned with the motives of those who fought, that is, with their subjective moral dispositions. Hence the very severe treatment in clause 6 of those whose motive was merely gain and who were, therefore, treated as though they were guilty of common homicide. This treatment would be altogether disproportionate if directed towards those who were obliged to fight because they received wages rather than because they held a fief. It is, however, intelligible in the light of Burchard's concern with the differing motives of those who commit homicide in war: the bishops insist that all who fought, no matter whether, objectively viewed, they did so for money or from feudal duty, must consider whether their subjective motive was nothing better than some form of gain to themselves (tantum praemio), or whether it was service in a just war against the perjured Harold (in publico bello); they should then do penance accordingly. It may be concluded that the penitential ordinance carries little weight when it is used by Mr. Hollister as evidence to support the conclusion that the Conqueror's army comprised ‘those who owed him service as a duty, and those who fought for pay’, although it is perfectly consistent with it. Mr. Prestwich, perhaps, places undue reliance on Wilkins's emendation when he goes so far as to conclude that the ordinance ‘distinguishes between those whom William had armed on his own orders, those armed not on his orders, those who owed him service as a matter of duty and those who fought for hire’.

^{page 241 note 1} Wilkins emends this manifestly corrupt passage as follows: ‘… suo iussu armavit, et qui absque iussu suo erant armati, et ex debito ei militiam debebant’. In the text, Spelman's excessive punctuation is retained.

^{page 241 note 2} Wilkins omits in.

^{page 241 note 3} Wilkins adds numerum.

^{page 242 note 1} Spelman has Poenitentiam … statuantur.

^{page 242 note 2} Wilkins adds et.

^{page 242 note 3} Wilkins adds poeniteant.

^{page 242 note 4} Spelman has poeniteat.