Not only is the attitude of Wace and William towards Harold's action here virtually identical, but the mention of his exaction of homage seems special to them both.

The passages, however, on which I would specially rest my case are those in which these two writers describe the visit of Harold's spies to the Norman camp before the battle of Hastings. This legend is peculiar to William of Malmesbury and Wace, and though it may be suggested that they had heard it independently, the correspondence—it will, I think, be admitted—is too close to admit of that solution.

I print these passages side by side:

The story is just one of those that William of Malmesbury would have picked up, and Wace has simply, in metrical paraphrase, transferred it from his pages to his own.

Yet another story, on which Mr Freeman looked with some just suspicion, is common to these two writers, and virtually to them alone. It is that of 'the contrast between the way in which the night before the battle was spent by the Normans and the English' (iii. 760). Wace, says Mr Freeman, 'gives us the same account' as William 'in more detail', while William 'gives us a shorter account'. I here again append the passages side by side, insisting on the fact mentioned by Mr Freeman, that Wace expands the story 'in more detail':

This brings me to my destination, namely, § 241 of the 'Gesta Regum'. We may divide this section into three successive parts: (1) the description of the way in which the English spent the night—which is repeated, we have seen, by Wace; (2) the array of the English, with which I shall deal below; (3) the dismounting of Harold at the foot of the standard. I here subjoin the parallels for the third, calling special attention to the phrases, 'd'or e de pierres (auro et lapidibus)' and 'Guil. pois cele victoire Le fist porter a l'apostoire (post victorium papae misit Willelmus).'

The only part of § 241 which remains to be dealt with is the second. The two passages run thus:

Mr Freeman, of course, observed the parallel, but, oddly enough, missed the point. He first quoted the lines from Wace, and then immediately added, 'So William of Malmesbury' (iii. 764), thus reversing the natural order. The word that really gave me the clue was the escuz of Wace. It was obvious, I held, that, here as elsewhere,⁴³ it must mean 'shield'; and Mr Freeman consequently saw in the passage an undoubted description of the 'shield-wall' (iii. 763). Moreover, the phrase lever escuz is, in Wace, a familiar one, describing preparation for action, thus, for instance:

On the other hand, there are, in spite of Mr Freeman, undoubted difficulties in rendering the passage as a description of the 'shield-wall', just as there are in taking escuz to mean 'barricades' (iii. 471). The result was that, perhaps unconsciously, Mr Freeman gave the passage, in succession, two contradictory renderings (iii. 471, 763). Now, starting from the fact that the disputed passage supported, and also opposed both renderings, I arrived at the conclusion that it must represent some confusion of Wace's own. He had, evidently, himself no clear idea of what he was describing. But the whole confusion is at once accounted for if we admit him to have here also followed William of Malmesbury. His escuz—otherwise impossible to explain—faithfully renders the scuta of William, while the latter's testudo, though strictly accurate, clearly led him astray. The fact is that William of Malmesbury must have been quite familiar with the 'shield-wall', if indeed he had seen the fyrd actually forming it.⁴⁴ Wace, on the contrary, living later, and in Normandy instead of England, cannot have seen, or even understood, this famous formation, with which his cavalry fight of the twelfth century had nothing in common. It is natural therefore that his version should betray some confusion, though his Fait en orent deuant closture clearly renders William of Malmesbury's conserta ante se scutorum testudine. There is no question as to William's meaning, for a testudo of shields is excellent Latin for the shield-wall formed by the Romans against a flight of arrows. Moreover, the construction of William's Latin (conserta) accounts for that use by Wace of the pluperfect tense on which stress has been laid as proof that the passage must describe a 'barricade'.⁴⁵ That Wace could, occasionally, be led astray by misunderstanding his authority, is shown by his taking Harold to Abbeville, after his capture on the French coast, a statement which arose, in Mr Freeman's opinion, 'from a misconception of the words of William of Jumièges (iii. 224)'. No one, I think, can read dispassionately the extracts I have printed side by side, without accepting the explanation I offer of this disputed passage in Wace, namely, that it is nothing but a metrical, elaborate, and somewhat confused paraphrase of the words of William of Malmesbury.

Passing from William of Malmesbury to the Bayeux Tapestry, we find a general recognition of the difficulty of determining Wace's knowledge of it. I can only, like others, leave the point undecided. On the other hand, his narrative, as a whole, does not follow the Tapestry; on the other, it is hard to believe that the writer of ll. 8103-38 had not seen that famous work. His description of the scene is marvellously exact, and the Tapestry phrase, in which Odo confortat pueros—often a subject of discussion—is at once explained by his making the pueri whom Odo 'comforted' to be—

Of these varlets in charge of the 'harness' he had already spoken (ll. 7963-7). The difficulty of accounting for Wace, as a canon of Bayeux, being unacquainted with the Tapestry is, of course, obvious. But in any case he cannot have used it, as we do ourselves, among his foremost authorities.

In discussing his use of William of Jumièges, we stand on much surer ground. It certainly strikes one as strange that in mentioning the obvious error by which Wace makes Harold receive his wound in the eye early in the fight (l. 8185), before the great feigned flight, Mr Freeman does not suggest its derivation from William of Jumièges, though he proceeds to add (p. 771):

I need hardly stop to refute the strange mistake of William of Jumièges, followed by Orderic: 'Heraldus ipse in primo militum progressu ['Congressu', Ord.] vulneribus letaliter confossus occubuit'.

But a worse instance of the contradictions involved by the patchwork and secondary character of Wace's narrative is found in his statement as to Harold's arrival on the field of battle. 'Wace,' says Mr Freeman, 'makes the English reach Senlac on Thursday night' (p. 441). So he does, even adding that Harold

But Mr Freeman must have overlooked the very significant fact that when the battle is about to begin, Wace tells a different story, and makes Harold only occupy the battlefield on the Saturday morning:

I have little doubt that he here follows William of Jumièges: '[Heraldus] in campo belli apparuit mane', and that he was thus led to contradict himself.

Mr Freeman had a weakness for Wace, and did not conceal it: he insisted on the poet's 'honesty'. But 'honesty' is not knowledge; and in dealing with the battle, it is not allowable to slur over Wace's imperfect knowledge. Mr Freeman admits that 'probably he did not know the ground, and did not take in the distance between Hastings and Battle' (p. 762). But he charitably suggests that 'it is possible that when he says "en un tertre s'estut li dus" he meant the hill of Telham, only without any notion of its distance from Hastings'. But, in spite of this attempt to smooth over the discrepancy, it is impossible to reconcile Wace's narrative with that of Mr Freeman. The latter makes the duke deliver his speech at Hastings, and then march with his knights to Telham, and there arm. But Wace imagined that they armed in their quarters at Hastings ('Issi sunt as tentes ale'), and straightway fought. The events immediately preceding the battle are far more doubtful and difficult to determine than could be imagined from Mr Freeman's narrative, but I must confine myself to Wace's version. I have shown that his account is not consistent as to the movements of Harold, while as to the topography, 'his primary blunder', as Mr Freeman terms it, 'of reversing the geographical order, by making William land at Hastings and thence go to Pevensey', together with his obvious ignorance of the character and position of the battlefield, must, of course, lower our opinion of his accuracy, and of the value of the oral tradition at his disposal.

To rely 'mainly'⁴⁶ on such a writer, in preference to the original authorities he confused, or to follow him when, in Mr Freeman's words, he actually 'departs from contemporary authority, and merely sets down floating traditions nearly a hundred years after the latest events which he records'—betrays the absence of a critical faculty, or the consciousness of a hopeless cause.

¹ Dismissing ut supra the 'fosse' passage, which neither mentions nor implies it, together with the passage from Henry of Huntingdon.

² Norm. Conq., iii. 763-4. I have shown in the English Historical Review (ix. 225) that he meant here by the shield-wall 'exactly what he meant by it elsewhere', a shield-wall and nothing else.

³ Cont. Rev., 344.

⁴ English Historical Review, ix. 231-40.

⁵ English Historical Review, ix. 2.

⁶ Ibid., 260.

⁷ Norm. Conq., iii. 763-4.

⁸ Cont. Rev., p. 348.

⁹ English Historical Review, ix. 17-20.

¹⁰ I explained, in one of my replies to Mr Archer, that this statement applied only to its usage 'in Wace' (Academy, September 16, 1893), but, characteristically, he has not hesitated to suppress this explanation, and renew his sneers at my knowledge of 'Old French', on the ground of a statement which, I had explained, was not my meaning (English Historical Review, ix. 604). It is difficult to describe such devices as these.

Common as the word is in Wace, I have never found any other instance of its use (i.e. by him) in a metaphorical sense, nor, if there is one, has Mr Archer attempted to produce it.

¹¹ Infra, pp. 313-18.

¹² English Historical Review, ix. 260.

¹³ Norm. Conq., iii. 736-7.

¹⁴ The word 'fenestres', for instance, which Mr Archer first rendered 'ash', out of deference to Mr Freeman and his predecessors, but subsequently 'windows' (English Historical Review, ix. 18), is either a corruption or quite inexplicable. 'If it pleases Mr Archer,' as I wrote (Ibid., 236), 'to construct a barricade, of which "windows" are the chief ingredient, on an uninhabited Sussex down, in 1066, he is perfectly welcome to do so.' I may add that the rendering adopted by the two French scholars does not in the least alter my view as to the improbability, or rather absurdity, of the suggestion.

¹⁵ Ibid., ix. 244.

¹⁶ Q.R., July 1893, p. 95.

¹⁷ English Historical Review, ix. 251-3. I was careful to add that 'if it be claimed that his text is contradictory, this would but prove further how confused his mind really was as to the battle' (p. 252). Mr Archer, as I anticipated, now prints, as a conclusive reply (Ibid., ix. 603), words which look the other way, ignoring, as usual, the quotations on which I explicitly relied. He has thereby, as I said, only proved how confused, here as elsewhere, Mr Freeman's conception was.

¹⁸ Mr Archer now prefers to leave its details doubtful (English Historical Review, ix. 606).

¹⁹ As I have shown in Ibid., ix. 244-5.

²⁰ Cont. Rev., 344.

²¹ Ibid., 346.

²² I have shown (Academy, September 16, 1893) by reference to Godefroi and Michel that either Mr Archer or they must here have been ignorant of Old French. The former alternative seems to be accepted.

²³ Supra, pp. 269-70.

²⁴ The case of the battle of Varaville, in 1058, is precisely similar in this respect to that of the Battle of Hastings. Of the former Mr Freeman writes: 'Wace alone speaks, throughout his narrative, of a bridge. All the other writers speak only of a ford' (iii. 173). Now Wace's authority was better for this, the earlier battle, because, says Mr Freeman, he knew the ground. Yet the Professor did not hesitate to reject his 'bridge'. So again, in 'the campaign of Hastings', Mr Freeman rejects 'the falsehood of the story of William burning his ships, of which the first traces appear in Wace' (iii. 408). So much for placing our reliance upon Wace, when he stands alone.

²⁵ Q.R., July 1893, p. 96.

²⁶ Mr Archer's limit is 1066-1210.

²⁷ We have, I suspect, a similar instance, in Wace's gisarmes (ll. 7794, 7814, 8328, 8332, 8342, 8587, 8629, 8656). An excellent vindication of the Bayeux Tapestry—oddly enough overlooked by Mr Freeman—namely, M. Delauney's 'Origine de la Tapisserie de Bayeux prouvée par elle-même' (Caen, 1824)—discusses the weapons, the author observing: 'La hache d'armes ressemble à celle de nos sapeurs; celle des temps postèrieurs au xi^e siècle à, dans les monuments, une espèce de petite lance au-dessus de la douille du côté opposé au tranchant' (see Jubinal, La Tapisserie de Bayeux, p. 17). This exactly describes the true gisarme, a later introduction. So again, Wace makes the chevalier who has hurried from Hastings exclaim to Harold:

whereas bretasches of course were impossible at the time. One is reminded of the description, by Piramus, of the coming of the English, when 'over the broad sea Britain they sought':

²⁸ English Historical Review, ix. 66.

²⁹ Ibid., 31-7, 17-18, and throughout his paper.

³⁰ Ibid., ix. 32.

³¹ 'Al siege de Rouen le quidierent gaber' (l. 62).

³² 'Demn nicht etwa am Schlusse, sondern gleich zu Anfang des genannten Theiles' (l. 179) 'spricht er von den drei Königen Heinrich die er gesehen und gekannt' (p. xciv).

³³ 'Nimmt man das Jahr 1110 als Geburtsjahr des Dichters an', etc. (p. xciv).

³⁴ English Historical Review, ix. 33. It need scarcely be said that these 'old heroes' would be found rather in England than in Normandy.

³⁵ Ibid., ix. 17.

³⁶ Ibid., ix. 33.

³⁷ Compare his scornful rejection (iii. 469-71) of Wace's tales in ll. 7875-950.

³⁸ English Historical Review, ix. 34.

³⁹ Reprinted from Ibid., October 1893.

⁴⁰ Norm. Conq., iii. 783.

⁴¹ iii. 402, note 2.

⁴² iii. 782.

⁴³ I mean, as I explained above, elsewhere in Wace.

⁴⁴ He describes, as Mr Freeman observed, King Henry bidding the English 'meet the charge of the Norman knights by standing firm in the array of the ancient shield-wall' (William Rufus, ii. 411).

⁴⁵ Cont. Rev., March 1893, p. 351.

⁴⁶ 'It is upon Wace that we shall mainly rely.' Cont. Rev., p. 344.

NOTE ON THE PSEUDO-INGULF

I owe to my friend Mr Hubert Hall the suggestion that the great battle described by the Pseudo-Ingulf as taking place between the English and the Danes in 870—and all accepted as sober fact by Turner in his History of the Anglo-Saxons—may be a concoction based on the facts of the battle of Hastings. This is also the theory Mr Freeman advanced as to Snorro's story of the battle of Stamford Bridge. The coincidence is very striking. In both narratives the defending force is formed with 'the dense shield-wall';¹ in both it breaks at length that formation; in both it is, consequently, overwhelmed; and in both cases the attacking force consists of horsemen and archers. But the most curious coincidence is found in the principal weapon of the defending force. In Snorro's narrative, as Mr Freeman renders it, 'a dense wood of spears bristles in front of the circle to receive the charge of the English horsemen';² in the Pseudo-Ingulf the defending force 'contra violentiam equitum densissimam aciem lancearum prætendebant'.³ Such a defence savours of the days when the knight, fighting on foot with his lance,⁴ had replaced the housecarl with his battle-axe: it was not that of Harold's host, but one which we meet with in the twelfth century.

There are marks, however, in the Pseudo-Ingulf, of study, not merely of the Battle of Hastings, but of William of Malmesbury's account of it. From him, it would seem, are taken the words 'testudo' and 'tumulus'. The first parallel passages are these:

Again, after the disaster caused, in each case, by a feigned flight, we have the rally thus described:

The Pseudo-Ingulf alludes but briefly to the Battle of Hastings itself. Yet here again we have traces of William of Malmesbury's words in 'nec de toto exercitu, præter paucissimos eum aliquis concomitatur' and 'more gregarii militis manu ad manum congrediens', which phrases are applied to Harold.

¹ Norm. Conq., iii. 367.

² Ibid., p. 365.

³ Ed. 1684, p. 21.

⁴ Vide supra, p. 279. Cf. the fight at Jaffa, August 5, 1192.

REGENBALD, PRIEST AND CHANCELLOR

No better illustration could be given of the fact that valuable historical evidence may lurk, even in print, unknown, than the charters printed, from the Cirencester Cartulary, by Sir Thomas Phillips in Archæologia (1836).¹ One can imagine how highly prized they would have been by Mr Freeman, had he only known of their existence.

Regenbald, of whom Sir Thomas would seem never to have heard, was the first Chancellor of England.² Mr Freeman called him, I know not on what authority, 'the Norman chancellor of Eadward'. Whatever his nationality, it is well established that he was that king's chancellor. He occurs repeatedly in Domesday, where he is distinguished as 'Canceler', 'Presbyter', and 'de Cirencestre'. We learn also from its pages that he held land in at least three counties—Berkshire, Herefordshire, and Dorset T.R.E.—and that he seems to have received further grants from King William in his return.³

The three charters of which I treat are found in the Cirencester Cartulary and are in Anglo-Saxon. The first is one of King Edward's in favour of 'Reinbold min preost', and is a confirmation to him of soc and sac, toll and team, etc., as his predecessors had enjoyed it 'on Cnutes kinges daie'. The third is a notification from King William that 'ic hæbbe geunnen Regenbald minan preoste eall his lond' as 'he hit under Edƿearde hædde mine meie'. The chief points to be noticed here are that the land is granted de novo, not confirmed, and that the Conqueror speaks of Regenbald as 'minan preoste', implying that he has taken him into his service.

It is the second of these charters that is of quite extraordinary importance. I here append it in extenso as printed by Sir Thomas Phillips:

'Vyllelm king gret Hereman b. & Wulstan b. & Eustace eorl & Eadrich & Bristrich & ealle mine þegenes on Ƿyltoneshyre & on Glouc'shyre fronliche & ic cuþe eoƿ ic habbe geunnan Reinbold mina preost ꝥ land æt Esi & ꝥ land æt Latton & ealle þæra þinge ꝥ þar to lið binnan port & buten mið sace & mið socne sƿa full and sƿa forð sƿa his furmest on hondan stodan Harald kinge on ællan þingan on dæge & æfter to atheonne sƿa sƿa ealra lefest ys & ic nelle nenna men geþafian ꝥ him fram honda teo ænig þære þinga þæs þa ic him geunne habbe bi minan freonshype.'

The relevant entry in Domesday speaks for itself:

Reinbaldus presbyter tenet Latone et Aisi. Duo taini tenuerunt pro II. Maneriis T.R.E. Heraldus comes junxit in unum. Geldabat pro ix. hidis (68b).

If the charter were nothing more than a grant from the Conqueror to a private individual of lands duly entered in Domesday, it would, I believe, as such be unique. Historians have long and vainly sought for any genuine charter of the kind; and here it has been in print for nearly sixty years.

But the document, I hope to show, does far more for us than this: it opens a new chapter in the history of the Norman Conquest.

We first notice that the writ is addressed not to Norman, but to English authorities. The only exception is Count Eustace, who was, of course, not a Norman, and who was known in England before the Conquest as brother-in-law to Edward the Confessor. The obvious inference is that, at the time this writ was issued, Norman government had not yet been set up in the district. Urse d'Abetot, for instance, the dreaded sheriff of Worcestershire, would probably have been addressed in conjunction with Bishop Wulstan had he been then in power. But we know that he came into power soon after the Conquest, for he had time to be guilty of oppression and to be rebuked for it by Ealdred before that Primate's death in 1069. But as our writ is of this early date, it must be previous to the treason of Count Eustace in 1067. It must therefore belong to the beginning of that year, when William had only recently been crowned king.

We see then here, I think, the Conqueror, in his first days as an English king, addressing his subjects, in a part of the realm not yet under Norman sway, and doing so in their own tongue and in the forms to which they were accustomed. As King Edward in his charter to Regenbald had greeted bishops, earls, and sheriffs, so here his successor greets two bishops, 'Eustace Eorl', and two Englishmen representing the power of the sheriff. And so again in his charter to London he began by greeting the Bishop and the Portreeve.⁴

The writ, it will be seen, is addressed to the authorities of Gloucestershire and Wilts. The estate lay in the latter county, but the connection of Regenbald de 'Cirencestre' with Glo'stershire may account for the inclusion of that county. Can we identify 'Eadrich' and 'Bristrich' with any local magnates? With some confidence I boldly suggest that the latter was no other than the 'Bristricus' of the Exon Domesday, that famous Brihtric, the son of Ælfgar, who, to quote from the appendix Mr Freeman devotes to him, 'appears distinctly as a great landowner in most of the western shires', one from whose vast domains was carved out later the great Honour of Gloucester. Until now, all we have known of him has been derived from the Domesday entries of his estates T.R.E. and from the legend which associates his name with that of Queen Matilda. But this charter enables us to say that he was living and still holding his great position in the west in the early days of William's reign.⁵

From 'Bristric' I turn to 'Eadric', and ask if we may not here recognize 'Eadric the Wild' himself? This can only be matter of conjecture, but it is certain that these two Englishmen are here assigned the place that would be given to a sheriff, and that 'Eadric the Wild'—'quidam præpotens minister', as Florence terms him—was a magnate in the west (Herefordshire and Shropshire) at the time of the Conquest. Mr Freeman terms him 'a man about whom we should gladly know more'. It is stated by Orderic that he was one of those who came in and submitted to William at the outset. But Mr Freeman held it 'far more likely that he did not submit till a much later time', because Florence says of him in William's absence: 'se dedere Regi dedignabatur'. Orderic's statement, however, is not denied, and Florence's words seem to me quite explicable by the hypothesis that Eadric had refused the 'dangerous honour', as Mr Freeman terms it, of following William to Normandy in 1067 among 'his English attendants or hostages'. Harried, in consequence, by his Norman neighbours, he retaliated by ravaging Herefordshire in August of that year; while Count Eustace also threw off his allegiance and made his descent on Dover.

If the identity of 'Eadric' is matter of conjecture, that of 'Eustace eorl' is certain. But no one has known, or even suspected, that he held, at this period, high position in the west. It may be that, as I have already hinted, he was sent by William to a district, as yet only nominally subject, as being, from his previous connection with England, less obnoxious than a Norman was likely to prove. It would be refining overmuch to suggest that William might also intend to establish him as far as possible from his base of operations at Boulogne.

In any case, we have in this charter a welcome addition to our scanty knowledge of that obscure period when William, as it were, was feeling his feet as an English king. Nor is it its least important feature that it shows us William, contrary to what Mr Freeman held to be his fundamental rule, speaking of his predecessor as 'Harald kinge'.

Before taking leave of Regenbald, we may glance at one of the Domesday entries relating to his lands. Mr Freeman, in two distinct passages, wrote as follows:

Although these two passages are found in two different appendices, the entries thus diversely adduced, are, of course, one and the same. But, it will be seen, the 'tenuit' of Domesday is equated by the 'emit' of the Exon book. One of the two must be wrong. I should accept the Exon text because 'emit licentia regis' is the right Domesday phrase, because it makes better sense, and because it is a sound principle of textual criticism that the Exchequer scribe was more likely to write the usual 'tenuit' for the exceptional 'emit' than the Exon scribe to do the converse. I should then read the passage thus: 'emit de Rainboldo sacerdote—per licentiam regis, ut dicit—qui tenuit eam die', etc.

If my view be adopted, we here detect noteworthy error in our great and sacrosanct record.

The charter of Henry I to Cirencester Abbey—in which he had placed Canons Regular, and of which he claimed to be the founder—sets, as it were, the coping-stone on the story of Regenbald.⁷ In it we read:

Dedi et concessi ... totam tenuram Reimbaldi presbyteri in terris et ecclesiis, et ceteris omnibusquæ subscripta sunt....

De rebus autem predictis quæ fuerunt Rembaldi hec statuimus.

The details of Regenbald's possessions are given, and are of special value for collation with Domesday. They set him before us not only as a landowner in five different counties, but also as the first great pluralist. Sixteen churches, rich in tithes and glebe—one might really term them 'fat livings'—had passed into the hands of Regenbald 'the priest'. From the king's phrase, 'dedi et concessi', he would seem to have been not merely confirming an endowment by Regenbald, but granting lands which had escheated to himself.⁸

And this conclusion is confirmed by the fact that the king, while granting them, especially reserved the life interest of the Bishop of Salisbury and of two others—one of them, alas! a bishop's nephew—who must have acquired their rights since Regenbald's death.

This charter, apart from its contents, is of great interest from its mention of the place where and the time when it was granted, together with its list of witnesses. These were the two Archbishops, the Bishops of Salisbury, Winchester, Lincoln, Durham, Ely, Hereford, and Rochester: Robert 'de Sigillo', Robert de Ver, Miles of Gloucester, Robert d'Oilli, Hugh Bigot, Robert de Curci, Payne 'filius Johannis et Eustacio et Willelmo fratribus ejus, et Willelmo de Albini Britone'. The charter was granted 'apud Burnam in transfretatione mea anno incarnationis Domini MCXXXIII. regni vero mei XXXIII.'; and 'Burna', as I have elsewhere shown,⁹ was Westbourne in Sussex, on the border of Hampshire, then in the king's hands by forfeiture and near the coast. Here therefore we see the king, when leaving England for the last time, surrounded by his prelates and ministers, and are enabled to say positively who were with him. I would note the predominance of the official class represented by the Bishops of Salisbury, Lincoln, and Ely, by the late chancellor, the Bishop of Durham, and by laymen who are found specially entrusted with administrative work. A long list of witnesses such as this is specially characteristic of the closing period of the reign,¹⁰ and, of course, always possesses biographical value.¹¹

Another English writ of the Conqueror, which may be profitably compared with that we have discussed, is found in one of the cartularies of Bury St Edmund's.¹² Its address, as rendered in the transcript, runs:

William [sic] kyng gret Ægelmær Bischop and Raulf Eorl and Nordman and ealle myne thegnaes on Sudfolke frendliche.

This writ is obviously previous to the deposition of Bishop Æthelmær in April, 1070, but how far previous it is not easy to say. 'Nordman' is clearly the sheriff of Suffolk, who appears in Domesday as 'Normannus Vicecomes' (II. 438). His name affords presumption, though not proof, that he was of English birth;¹³ and as his Domesday holding consisted only of rights over two Ipswich burgesses (which he may have acquired during his shrievalty) he is hardly likely to have been one of the conquering race. Of the third official, Earl Ralf, we know a good deal. Mr Freeman was much puzzled by this 'somewhat mysterious person',¹⁴ but eventually came to the conclusion that 'there were two Ralfs in Norfolk, father and son, the younger being the son of a Breton mother: the elder was staller under Edward and Earl under William'. The younger was the Earl of Norfolk (or 'of the East Angles'), who rebelled and was forfeited in 1075; the elder was that 'Rawulf' who, in the words of the chronicle, 'wæs Englisc and wæs geboren on Norðfolce'. Putting our evidence together, I lean strongly to the view that we have here, as in the case of Regenbald, a writ addressed to English authorities before Norfolk had passed into the hands of Norman authorities. Mr Freeman held that a passage in Domesday (II. 194), to which he had given much attention, should be read—'Hanc terram habuit A[rfastus] episcopus in tempore utrorumque [Radulforum]', and that therefore 'the elder Ralph was living as late as 1070, in which year the episcopate of Erfast begins'. But the context clearly shows that we should read 'A[ilmarus] episcopus', and that, therefore, the elder Ralf died before Æthelmær was deposed. Moreover, Norwich, we are specially told, was entrusted by the Conqueror to William fitz Osbern before his departure from England in March 1067. William was placed, some two years later, in charge of York castle, and we read in Mr Freeman's work that 'the man who now (autumn, 1069) commanded at Norwich, and who was already, or soon afterwards, invested with the East-Anglian Earldom, was the renegade native of the shire, Ralf of Wader'.¹⁵ This, it will be seen, contradicts his own, and supports my reading of the Domesday passage quoted above. Everything therefore points to the 'Raulf Eorl' of our writ dying or being deposed shortly after the Conquest.

Before taking leave of this writ we may note that, dealing as it does with Suffolk, it is addressed to Earl Ralf as Earl, not merely of Norfolk, but of East Anglia. This is of some importance, because Mr Freeman wrote, speaking of the Regents appointed in 1067:

There was no longer to be an Earl of the West Saxons or an Earl of the East Angles.... Returning in this to earlier English practice, the Earl under William was to have the rule of a single shire only, or if two shires were ever set under one Earl they were at least not to be adjoining shires. The results of this change have been of the highest moment. (iv. 70.)

Yet on page 253, as we have seen, we read of 'the East Anglian Earldom', and on page 573 that the younger Ralph 'had received the Earldom of East Anglia'—Florence of Worcester distinctly terming him 'East-Anglorum comite'. Mr Freeman, indeed, was led by this passage to style him 'Earl of Norfolk or of the East Angles'.¹⁶ I believe this latter style to be perfectly correct, and, as I have shown in my Geoffrey de Mandeville (p. 191), to apply even to the Bigod earldom in the days of Stephen.

The curious English writ that has suggested these considerations ought to be compared with a Latin one, also in favour of St Edmund's, on which I lighted in examining the 'Registrum Album' of the Abbey. It is one of those exceedingly rare documents that find their correlatives in Domesday. The words of the writ are these:

W. rex Anglor' E. epo. B. Abbi W. Malet salm. sciatis vos mei fideles me concessisse servitium de Liuremere quam Werno hactenus de me tenuit sancto Ædmundo Et filia Guernonis in vita sua de Abbate B. tenuit.¹⁷

The last clause is clearly an addition by the cartulary scribe. Now this charter being addressed, like the other, to Æthelmær ('Ethelmerus'), Bishop of the East Angles, is, of course, previous to April 1070. I should, therefore, also place it previous to the capture of William Malet at York in September 1069. But this, unlike the other date, is matter of probability rather than of proof. Mr Freeman believed that William returned, and died 'in the marshes of Ely' (1071), but this is only a guess in which I cannot concur.¹⁸ In any case, we have evidence here of this well-known man having held a position in Suffolk (where he owned the great Honour of Eye) analogous to that of sheriff. He may have succeeded Northman in that office.

The relevant Domesday entry is as follows:

Hujus terram rex accepit de abbate et dedit Guernoni depeiz [de Peiz]. Postea licencia regis deveniens monachus reddidit terram. (363b.)

The charter records, I take it, the 'licencia regis' of Domesday.¹⁹

¹ Vol. xxvi., p. 256.

² Not counting Leofric, styled 'regis cancellarius' by Florence in 1046.

³ See my life of him in Dictionary of National Biography.

⁴ It might even be suggested that not only this charter but the Essex writ in favour of Deorman (addressed to Bishop William and Swegen the sheriff) belonged to the same early period. Compare, however, the Conqueror's Old English writ that I have discussed ('Londoners and the Chase') in the Athenæum of June 30, 1894.

⁵ It is a noteworthy coincidence that 'Brihtricus princeps' and 'Eadricus princeps' are among the witnesses to Harold's Waltham charter in 1062, which Regenbald himself also attests as Chancellor.

⁶ sic.

⁷ See Monast. Anglic., ii. 177.

⁸ It is possible, I think, that the only endowment entered to the church at Cirencester in Domesday, viz., two hides at Cirencester, had been originally given by Regenbald.

⁹ Henry I, at 'Burne' (English Historical Review, 1895).

¹⁰ As in the charters to Aubrey de Vere (Baronia Anglica, 158) and William Mauduit.

¹¹ Here, it would seem, is further proof of the Bishops of Ely and Durham assuming their styles before consecration (infra, pp. 366-7).

¹² Harl. MS., 743, fo. 8d.

¹³ Mr Freeman held him to be an Englishman.

¹⁴ Norm. Conq. 2nd Ed.), iii. 773. Cf. 1st Ed., iii. 752-3; iv. 277.

¹⁵ Ibid. (1st Ed.), iv. 252-3.

¹⁶ Ibid. (2nd Ed.), iii. 773.

¹⁷ Add. MS., 14,314, fo. 32b (pencil).

¹⁸ See my letter on 'the death of William Malet' in Academy of August 26, 1884.

¹⁹ Since this paper was written, there has appeared the valuable Bath Cartulary (Somerset Record Society) containing a most remarkable charter (p. 36), which should be closely compared with those to Regenbald. It is issued by William the King and William the Earl, and must undoubtedly be assigned to the former's absence from England, March-December 1067. It shows us therefore William fitz Osbern acting as Regent and anticipating the office of the later Great Justiciar by inserting in the document his own name. This charter, like that to Regenbald, is addressed to the still English authorities of an unconquered district.

THE CONQUEROR AT EXETER

'And y seide nay, and proved hit by Domesday.'¹

For a companion study to the Battle of Hastings, one could not select a better subject than the Siege of Exeter by William in 1068. It is so, because, in the tale of the Conquest, 'No city of England', in Mr Freeman's words, 'comes so distinctly to the front as Exeter':² and because, as editor of 'Historic Towns', he chose Exeter, out of all others, as the town to be reserved for himself.³ 'Its siege by William', we are told, 'is one of the most important events of his reign';⁴ but it was doubtless the alleged 'federal' character of Exeter's attitude at this crisis that gave its story for him an interest so unique. This episode, moreover, has many advantages: it is complete in itself; it is rich in suggestion; it is taken from the period in which the Professor described himself as 'most at home'; and its scene is laid within his own borders, his own West Saxon land. It presents an admirable test of Mr Freeman's work at the point where he was admittedly strongest, and his thoroughly typical treatment of it affords a perfect illustration of the method he employed.

The year 1067 was drawing to its close when the Conqueror, summoned back from Normandy by the tidings of pressing danger, returned to spend his Christmas at Westminster amidst 'the sea of troubles which still awaited him in his half-conquered island-kingdom'.⁵ Threatened at once by foes within and without the realm, he perceived the vital necessity of severing their forces by instant suppression of the 'rebellions' at home, swift suppression before the invaders were upon him, stern suppression before the movement spread. Let us bear in mind these twin motives, by which his policy must at this juncture have been shaped, the need for swiftness, with invasion in prospect, and the need for sternness as a warning to 'rebels'.

Of all the 'rebellious' movements on foot, that at Exeter, as Mr Freeman admits, was 'specially hateful in William's eyes'.⁶ It was against Exeter, therefore, that the Conqueror directed his first blow. In the depths of winter, in the early days of the new year, 'he fared to Devonshire'. Such is the brief statement of the English Chronicle.

We hear of William at Westminster; we next hear of him before the walls of Exeter: all that intervenes is a sheer blank. Of what happened on this long westward march not a single detail is preserved to us in the Chronicle, in Orderic or in Florence. Now it is precisely such a blank as this that, to Mr Freeman, was irresistible. We shall see below how, a few months later, we have, in William's march from Warwick to Nottingham, a blank exactly parallel.⁷ There also Mr Freeman succumbed to the temptation. He seized, in each case, on the empty canvas, and, by a few rapid and suggestive touches, he has boldly filled it in with the outlines of historical events, not merely events for which there is no sufficient evidence, but events which can be proved, by demonstration, to have had no foundation in fact.

The scene elaborated by Mr Freeman to enliven the void between the departure from London and the entrance into Devonshire is the resistance and the downfall of 'the Civic League'.⁸ This striking incident in the Exeter campaign I propose to analyse without further delay.

It must, in the first place, be pointed out that we have no proof whatever of this 'Civic League' having even existed. To apply Mr Freeman's words to his own narrative:

The story is perfectly possible. We only ask for the proof. Show us the proof;... then we will believe. Without such a proof we will not believe.⁹

For proof of its existence Mr Freeman relies on a solitary passage in Orderic.¹⁰ But Orderic, it will at once be seen, does not say that any such league was effected; he does not even say that the league which was contemplated was intended to be an exclusively Civic League. What he does say is that the men of Exeter sought for allies in the neighbouring coasts (plagæ)¹¹ and in other cities. The Dorset townlets, such as Bridport, with its 120 houses, would scarcely represent these 'cities'. Mr Freeman assumed, however, that 'the Civic League' was formed, assumed that the Dorset towns had 'doubtless' joined it, and finally assumed that they were 'no doubt' besieged by William in consequence.¹² These assumptions he boldly connected with the entries on the towns in Domesday, entries which we shall analyse below, and which are not only incorrectly rendered, but are directly opposed to the above assumptions.

What, then, is the inference to be drawn? Simply this. The 'Civic League' must share the fate of the 'palisade on Senlac'. The sieges which took place 'probably' never took place at all; the League never resisted; the League never fell; in short, there is not a scrap of evidence that there was ever such a League at all. The existence of such a League would be, unquestionably, a fact of great importance. But its very importance imperatively requires that its existence should be established by indisputable proof. Of such proof there is none. One can imagine how severely Mr Freeman would have handled such guesses from others. For he wrote of a deceased Somersetshire historian who boldly connects the story of Gisa with the banishment of Godwine:

One is inclined to ask with Henry II, 'Quære a rustico illo utrum hoc somniaverit?' But these things have their use. Every instance in the growth of a legend affords practice in the art of distinguishing legend from history.

It should, however, in justice be at once added that this story did not originate wholly with Mr Freeman himself. He refers us on the subject of the League to his predecessor, Sir Francis Palgrave. The brilliant imagination of that graceful writer was indeed led captive by the fascinating vision of 'the first Federal Commonwealth', yet he did not allow himself, when dealing with the facts, to deviate from the exact truth. His statement that Exeter 'attempted to form a defensive confederation' reproduces with scrupulous accuracy Orderic's words. And even when he passed from fact to conjecture, there was nothing in his conjecture at variance from fact. From him we have no suggestion that the Dorset towns resisted William or 'stood sieges'. It was left for Mr Freeman to carry into action Palgrave's line of thought, and, by forcing the evidence of the Domesday Survey into harmony with the story he had evolved, to show us, in his own words, 'the growth of a legend'. For, as he observed with perfect truth:

What we call the growth of a story is really the result of the action of a number of human wills. The convenient metaphor must not delude us into thinking that a story really grows of itself as a tree grows. In a crowd of cases ... the story comes of a state of mind which does not willingly sin against historical truth, but which has not yet learned that there is such a thing as historical truth.

Had Mr Freeman done so himself? Did he ever really learn to distinguish conjecture from fact? One asks this because within the covers of a single work, his English Towns and Districts, that Civic League which in the Norman Conquest is said to have existed 'no doubt', is in one place said to have existed 'perhaps', and in another is set forth as an undoubted historic fact:

Exeter stood forth for one moment ... the chief of a confederation of the lesser towns of the West.... A confederation of the western towns, with the great city of the district at their head, suddenly started into life to check the progress of the Conqueror.

Finally, in his 'Exeter' (1887), the same story again appears, without a word of caution, as absolute historic fact. Exeter, we read, was

the head of a gathering of smaller commonwealths around her; ... the towns of Dorset were in league with Exeter.... We have no record of the march, but it is plain that the towns of Dorset were fearfully harried.