APPENDIX. (continued 4)

The story goes on to say that Hubert did not come with William into England because disturbances were looked for in Maine, which he was sent to quell or to hinder (“veniente Willielmo in Angliam accipere sibi regnum, quoniam a Cenomanicâ regione suspicabatur tumultus, Hubertus, quia erat promptus manu et consilio bonus, missus est illic prætendere et servare pacem”). I did not venture, without 684better authority, to mention this in my text at p. 384, but it is worth comparing with the seeming disloyalty of the Cenomannian knight at Senlac mentioned in p. 486.

Now, after going through all these accounts, what is the real state of the evidence with regard to the alleged promise of Eadward to William? A death-bed bequest, as we have seen, was not alleged by William or by his contemporaries. Such a bequest would be inconsistent with any of the versions of the story of Harold’s oath, all of which conceive William as asserting some right to the succession before Eadward’s last sickness. A promise made in earlier times, before Eadward’s accession, is possible, but it is by no means likely, and such a promise could not be of any legal force. An act of the King and his Witan in William’s favour is impossible in itself and is confirmed by no kind of evidence. But that there was some promise I think cannot be doubted. When I believe that promise to have been made I have already said (see vol. ii. p. 303). Here is another fact which looks the same way. In one of the manuscripts quoted by M. Francisque Michel (Benoît, iii. 164), William tells us how he reigned, “devicto Haroldo Rege cum suis complicibus, qui mihi regnum prudentiâ Domini destinatum et beneficio concessionis domini et cognati mei gloriosi Regis Edwardi concessum conati sunt auferre.” This is vague enough, but it suggests one hint. The feudal language employed, the words “beneficium” and “dominus”—the latter of which is applied by William to Eadward in other documents (see above, pp. 249, 555, and cf. ii. 18)—might suggest that, when the promise was made, William did homage to King Eadward as his lord and adopted father. There is but one time when this could have happened. We have here another confirmation of the view, supported by no direct evidence, but the only view which is not upset by opposing evidence, that the promise was made by Eadward, and that the homage, if there was any, was performed by William, at the time of William’s visit to England in 1051.

III. The next point is as to the time, the place, and the matter of Harold’s oath. As to the time, William of Poitiers (108) distinctly places the oath, as well as the knighthood, before the expedition into Britanny. Such seems also to be the account of Orderic (492 A-B), though his way of telling the story in the pluperfect 685tense takes off somewhat from the clearness of his narrative; “Heraldus ... sacramentum fecerat, ... tunc etiam Dux eumdem Heraldum in expeditione secum contra Conanum Comitem Britonum duxerat, armisque fulgentibus et equis aliisque insigniis cum commilitonibus suis spectabiliter ornaverat.” The other accounts seem to place both the knighthood and the oath after that expedition. In the Tapestry (pl. 6) we read immediately after the taking of Dinan, “Hic Willelm dedit Haroldo arma,” on this follow the words, “Hic Willelm venit Bagias, ubi Harold sacramentum fecit Willelmo Duci.” So William of Jumièges (vii. 31) and Wace (Roman de Rou, 10816) place the oath immediately after the return from Britanny, though Wace (10812; see above, pp. 228, 240) seems to place the knighthood before the expedition. As to the place, there is yet more difference than as to the time. The Tapestry clearly means to place the event in its own city, for “Bagias,” curious as the form is, can mean no place but Bayeux. So does Wace, a Canon of Bayeux (Roman de Rou, 10826). But William of Poitiers (108) places it at Bonneville (“coadunato ad Bonamvillam concilio”), in which he is followed by Benoît (36594);

“Si josta li dux son concile,

Ce sui lisant, à Bone-Vile.

Là fu li serremenz jurez,

Que Heraut meisme a devisez.”

But Orderic (492 A) places the scene of the oath at Rouen (“ipse Heraldus apud Rotomagum Willelmo Duci coram optimatibus Normanniæ sacramentum fecerat”). It is also placed at Rouen by the romantic Biographer of Harold (Chron. Angl. Norm. ii. 184 et seqq.), who adds that the oath was sworn under a very large and aged oak, which, as soon as the oath was broken, immediately lost its bark, which, as he truly adds, was “dictu mirum” and “res digna spectaculo.” He adds that he himself saw it an hundred and forty years after the event, which fixes his own date to about the year 1205.

As to the form of the oath, William of Poitiers (108) simply says, “Heraldus ei fidelitatem sancto ritu Christianorum juravit.” But, according to his manner, he implies in a later passage (131; see p. 463, note 4) that the oath was made upon relics. So Orderic (492 A); “Homo ejus factus, omnia quæ ab illo requisita fuerant super sanctissimas reliquias juraverat.” The “phylactery called ‘the bull’s-eye’” is found in the Brevis Relatio (4); “Ei, sicut multi dicunt, super filacterium quod vocabant oculum bovis quod 686ei fidem et promissionem quam ei faciebat bene custodiret.” The Hyde writer (290) tells us why it was called the bull’s-eye; “Infinitam sanctarum multitudinem reliquiarum deferri jussit, superque eas filacterium gloriosi martyris Pancratii, quod oculum bovis vocant, eo quod gemmam tam speciosam quam spatiosam in medio sui contineat, collocavit, certissimè sciens tantum martyrem nullâ temeritate posse deludi.” In the Tapestry (pl. 6) Harold is shown swearing between two chests or phylacteries (see Ducange in voce) of different shapes, one of which has something on the top which might fairly pass for the “gemma speciosa et spatiosa” of the Hyde writer. But in none of these accounts do we find anything about the trick played upon Harold by William. Whatever Harold swears upon, it is not at all implied that he was otherwise than fully aware of what he was doing. The tale according to which Harold is made unwittingly to take an oath of a more solemn kind than he supposed comes from Wace;

“Toz li corz sainz fist demander,

Et un liu tuz asembler;

Tut une cuve en fist emplir,

Pois d’un paele les fist covrir,

Ke Heraut ne sont ne ne vit,

Ne ne li fust mostré ne dit

De suz out une filatire,

Tut li meillor k’il pout eslire,

E li plus chier k’l pout trover:

Oil de boef l’ai oï nomer.”

(vv. 10828–10837.)

So again, after the oath has been taken;

“Quant Heraut out li sainz beisiez

Et il fu suz levez en piez,

Verz la cuve li Dus le trait,

E lez la cuve ester le fait:

De la cuve a le paesle osté,

Ki tut aveit acoveté;

A Heraut a dedenz monstré,

Sor kels cors sainz il a juré.

Heraut forment s’espoanta

Des relikes k’il li monstra.”

(vv. 10850–10859.)

I leave the reader to judge which of these two versions is the older. It is quite possible that the striking circumstance of the hidden relics, which is peculiar to Wace, may be merely an instance of the usual growth of legend. But it is no less possible that Wace, a very honest writer, may have preserved a fact which, as tending to the discredit of William and in some measure lessening the crime of Harold, was left out by two violent partizans. I leave the point in the same uncertainty in which I must leave all the other details of the story.

We now come to the matter of the oath said to have been taken 687by Harold. The fullest accounts of its terms are those which are given by William of Poitiers (108) and by Eadmer (5). I have in the text (p. 241) given its terms according to the accounts of these two writers. The oath, as given by William of Poitiers, runs as follows;

“Se in curiâ domini sui Edwardi Regis, quamdiu superesset, Ducis Willelmi vicarium fore; enisurum, quanto consilio valeret aut opibus, ut Anglica monarchia post Edwardi decessum in ejus manu confirmaretur; traditurum interim ipsius militum custodiæ castrum Doveram, studio atque sumptu suo communitum; item per diversa loca illius terræ alia castra, ubi voluntas Ducis ea firmari juberet; abunde quoque alimonias daturum custodibus.”

The terms in Eadmer take the form of a speech addressed by William to Harold. William mentions Eadward’s promise of the Crown to him, and goes on thus;

“Tu quoque, si mihi te in hoc ipso [the succession to the Crown] adminiculaturum spoponderis, et insuper castellum Dofris cum puteo aquæ ad opus meum te facturum, sororemque tuam quam uni de principibus meis dein in uxorem te ad me, tempore quo nobis conveniet, destinaturum, necne [necnon?] filiam meam te in conjugem accepturum fore promiseris, tunc et modò nepotem tuum, et quum in Angliam regnaturus venero, fratrem tuum incolumem recipies.”

In both these versions Harold binds himself in different terms to do all that he can to bring about William’s succession to the Crown; but William of Poitiers makes no mention of the engagements as to the two marriages, the marriage of Harold to a daughter of William and the marriage of Harold’s sister to one of William’s nobles. In a later passage however (p. 145), which I shall quote in another note (see Appendix U), he takes for granted both the marriage of Harold with William’s daughter, and also another stipulation of which we hear more distinctly in William of Jumièges (vii. 31). This is the important addition that Harold should have half the Kingdom as the portion of Adeliza (“deinde Dux postquam Heraldus fidelitatem sibi de regno pluribus sacramentis firmavit, Adelizam filiam suam cum medietate Anglici regni se daturum eidem spopondit”). William of Malmesbury (ii. 228) gives much the same account as Eadmer, but with the strange addition that Harold offered the oath of his own accord. His 688words are; “Ibi Haroldus, et ingenio et manu probatus, Normannum in sui amorem convertit; atque, ut se magis commendaret, ultro illi tunc quidem castellum Doroberniæ quod ad jus suum pertineret, et post mortem Edwardi Regnum Anglicum, sacramento firmavit: quare et filiæ, adhuc impubis, desponsione, et totius patrimonii amplitudine donatus, familiarium partium habebatur.” The words in Italics would seem to refer to some sort of commendation of Harold’s lands to William, and something of the same kind seems implied in the words with which William of Poitiers goes on with his story (109); “Dux ei [Heraldo] jam satelliti”—a common equivalent for “miles” or vassal—“suo accepto per manus ante jusjurandum terras ejus cunctumque potentatum dedit petenti.”

Wace says nothing about the castle of Dover, still less does he stoop to Eadmer’s detail about the well, but he mentions the two engagements to marry William’s daughter and to make over the Kingdom to him at Eadward’s death. He gives these terms twice;

“Entretant a li Dus parlé

Tant ke Heraut li a graé,

Ke Engleterre li liverra

Tres ke li Reis Ewart morra;

Et à moillier s’il velt prendra

Ele une fille ke il a:

Co se li plaist li jurera,

Et Willame le graanta.”

(vv. 10816–10823.)

And again in describing the actual taking of the oath;

“Poiz a juré et a prami

Si come home ki eschari:

Ele, la fille al Duc prendra,

Et Engleterre al Duc rendra;

De ço il fera son poeir

Sulunc sa force e son saveir

Emprès la mort Ewart, s’il vit;

Si veirement Dex li aït,

E li corz sainz ki iloc sont.”

(vv. 10840–10848.)

I have discussed in the text the chief points connected with the terms of the oath, and I have put forth the view which, on the whole, seems to me the least improbable. I will now get together several passages in which the promise to marry William’s daughter becomes the chief, or even the only, engagement taken by Harold. They come from writers who had not such good means of information as William of Poitiers, but who were much less likely to misrepresent or colour the story. They show what aspect of the tale most struck those who were not immediately interested in 689the matter. Let us begin with Snorro. The only engagement he knows of on Harold’s part is his engagement to marry William’s daughter. The proposal, as far as William is concerned, comes from Harold himself, but seemingly from Harold talked over by Matilda. The Duke does not altogether approve of the long evening talks between Harold and his wife. Matilda tells Harold that her husband is anxious to know the subject of their discourse, so it is agreed that Harold shall tell the Duke the next day. Harold accordingly asks William for his daughter in marriage, saying that he has fully talked the matter over with the Duchess, who favours his suit (“þat er at segia ydr, Jarl, atfleira býrr í hingat-komo minni, enn þat er ec hefir enn upp borit fyrir ydr. Ec vill bidia dóttr þinnar til eigin-kono mer, hefir oc þetta máll rætt fyri módor hennar optliga, oc hefir hon mer því heitit, at lidfinna þetta mál vid ydr.” Johnstone, 191; Laing, iii. 76). The suit is accepted; Harold is betrothed to the princess, but, on account of her youth, the marriage is to be put off for a time. Harold then goes to England, and comes back no more to Welshland to marry the girl (“oc kom eige sídan til Vallandz, at vitia þessa ráds”). When Harold is elected King, William indeed bethinks himself that his kindred with Eadward gives him a better right to the Crown than Harold (“enn Vilhiálmr þóttiz betr tilkominn ríkis í Englandi, en Haralldr, fyrír frændsemis sakir þeírra Játvardar Konungs.” Johnstone, 216; Laing, iii. 94), but it is not made to appear that there was any breach of faith on this score on Harold’s part. William invades England, partly to assert his hereditary right to the Crown, partly to punish Harold for not marrying his daughter (“þat var oc med, at hann þóttiz eiga at giallda Haralldi svívirding, er hann hafdi slitit festamálom vid dóttur hans”).

So in the Cartulary of Saint Bertin, in a passage already quoted, we read (p. 197) how “Willelmus, Comes Nortmanniæ.... Angliam petiit, ac Haroldo ipsius terræ [Rege?] occiso, eo quod filiam ipsius Wilhelmi in uxorem recipere recusaverit, Anglos gravi prœlio, multorum sanguine fuso, sibi subjugavit, et utrique populo Anglico et Nortmannico regnavit.” Here the only reason given for William’s attacking England is Harold’s neglect to marry his daughter. No other was known to the author of the Chronicle of Saint Andrew’s at Cambray, who wrote in 1133. Here (Pertz, vii. 537) William is described as invading England without any 690apparent reason, till we reach the words “Rex Anglorum Heroldus, olim contra prædictum Comitem Willelmum perjurus, nam filiam ejus se accepturum juraverat.” The Waltham writer “De Inventione” (cap. 20) is in the like case; “Insidiantibus ei perfidis Normannorum versutiis, quia filiam Willelmi Ducis Normannorum nuptui traditam contempsit.” These “versutiæ” exactly describe the sort of constraint under which I conceive Harold to have made the whole engagement.

These accounts mention no ground at all for the invasion except Harold’s refusal to marry William’s daughter. Other accounts, without going so far as this, put the question of marriage forward in a very remarkable way, as if everything else was incidental. This is the case in the version of the messages between William and Harold which is given by Eadmer and Simeon. William’s main object is to demand Harold’s sister and to require Harold to marry his daughter. Other matters are quite secondary. “Venit nuntius in Angliam a præfato Willelmo directus, expetens sororem Haroldi, juxta quod convenerat Willelmo et illi. Alia etiam quæ, violato sacramento, servata non erant, calumniatus est” (Eadmer, 5). And afterwards, “iterum ei amicâ familiaritate mandavit quatenus, aliis omissis, servatâ fidei sponsione, saltem filiam suam uxorem duceret.” (Simeon reads, “ut quamvis violatâ fide cætera non servâsset, si tamen filiam suam duceret uxorem, leviter ferret.”) Harold, in his answer, does undoubtedly speak of the Kingdom and of the castle of Dover, but these subjects are thrust in between his answers about the two marriages, how his sister is dead—does the Duke wish for her corpse?—and how he cannot marry a foreign wife without the consent of the Witan. Throughout this story William is made much more anxious to find a husband for his daughter than to find a Kingdom for himself. This must surely come from some account like that of Snorro and the other writers quoted above, which spoke of the marriage only, and which has been mixed up with the account given by William of Poitiers.

It will be remembered that, in the wild account of Eadward’s death-bed given in the French Life (see above, p. 589), everything is made to turn on an expected marriage between Harold and William’s daughter. It is on that marriage that he is made to ground his hopes of the Crown. In the account in Matthew Paris (p. 1) again, the marriage comes first. Harold, carried to Ponthieu 691by accident, and handed over by Guy to William, pretends that he is come secretly into Normandy to make a league with the Duke and to marry his daughter (“Haraldus asserebat se hæc omnia sponte fecisse, ut clam veniens in Normanniam confœderaretur Duci, filiam ejus in sponsam accepturus”). This he swears to do on the relics of the saints (“quod et juravit super sanctorum multorum reliquias se fideliter ad quemdam terminum completurum”). Harold and William, hitherto enemies, now become great friends, all the more so because Harold has come secretly (“tantò igitur majori honore susceptus est quantò secretiùs adventâsset; fuerant enim ante inimici ad invicem”). Then, as something quite secondary, comes the oath about the Kingdom; “Juravit insuper se post mortem Regis Edwardi, qui jam senuit sine liberis, regnum Angliæ Duci, qui in regnum jus habuit, fideliter conservaturum.” There is no explanation of the alleged right of William, no mention of any bequest by Eadward. Matthew had surely read some account which mentioned the promise of marriage only, and he added the bit about the Kingdom in deference to the more usual statement.

Lastly, we have that most singular account in the Hyde writer (288–90), to which I have already (see above, p. 686, and vol. ii. p. 519) had to refer. According to this version, hostages, one of them a son of Godwine, were given to William for the safety of Eadward, when the English embassy came to offer him the Crown on the death of Harthacnut. If any one chooses to accept this statement, here is a ready way of accounting for the other stories about the hostages. And though I do not think it likely that a son of Godwine was given as a hostage to William at that time, it is distinctly less unlikely than that he should have been so given at the time of Godwine’s return. But in this account the hostages are not spoken of again. Harold, sailing to some place not mentioned (“quibusdam caussis navem ingressus”), is driven by adverse winds to Ponthieu; he is imprisoned by Guy, and set free at the prayer (“precibus”) of Duke William. The Earl and the Duke, according to this account, appear, not as the old enemies which Matthew Paris calls them, but as old friends (“ab eodem [Willelmo] optimè cognitus [Haroldus] in multis familiariter est habitus”). They agree at last that, as William’s cousin King Eadward has no heir, Harold shall receive the Kingdom of England on condition of 692marrying William’s daughter and, it would seem, of holding the Kingdom as a fief of his father-in-law. Such seems to be the meaning of the words, “Ad hoc inter eos sermo progressus est, ut quia Edwardus Rex Anglorum, consobrinus Comitis Willelmi, hæredem non habebat, regnum Anglorum Willelmus Haroldo concederet eo tenore, ut filiam ipsius matrimonio acciperet, eique per omnia fideliter [fidelis?] exsisteret.” Harold agrees so readily that William, who did not easily trust Englishmen, becomes suspicious, and binds him by an oath (“quod quum promptissimè annueret, ab eodem Anglorum fidem suspectam habente ad districta sacramenta est coactus”). The phylactery of Saint Pancras, already spoken of, is accordingly brought, and Harold swears on it “se omnia, scilicet sicut fuerat postulatus, constantissimè se servaturum et Normannis fidelem affuturum.”

Now it is singular when, after this, we read in the same account (see above, p. 595) that Eadward on his death-bed left the Crown to William, and that Harold usurped it. This hardly fits in with an agreement between Harold and William that Harold should have the Crown on certain conditions. But the account which this writer (291) gives of the messages between the two princes exactly fits in with his account of the oath. William calls on Harold to do what he has promised to do (“ut sacramentorum in Normanniâ gestorum reminiscens, fœdus quod juraverat persolveret”), that is, doubtless, to marry his daughter and do homage for the Kingdom. Harold’s answer is that he has no need of any kind of connexion or intercourse with the Normans, and will therefore not perform his oath (“remandat Haroldus Normannorum societatem non esse Anglis necessariam, et ideò nullum ei juramentum persolvere, nullum cum eo consortium habere”). Of resigning the Kingdom there is not a word.

The reader will by this time know the peculiar position of the Hyde writer. His accounts are often strange and incredible, and contradicted by better authorities. But they are always independent. The writer is no servile copyist. He followed independent traditions and exercised an independent judgement of his own. His statements therefore have a certain value. And in this case they have more value than usual, as, though his prejudices are strongly Norman, his story points to a tradition quite different from the received Norman version.

693Putting then together the hints of these various writers, combined with the probability of the case, I am led to the view which I have suggested in the text. Harold’s oath was primarily an oath to marry William’s daughter, and this oath, or perhaps his knighthood, was accompanied by an act of homage to William. He became his “satelles” in the language of William of Poitiers, his “homo” in the language of Orderic. I have shown in the text how vague and lax were the obligations thereby incurred. But such an act obviously lay open to many interpretations. Harold would doubtless hold that whatever obligations were imposed either by gratitude or by formal vassalage would be amply fulfilled by his services in the Breton war and by continued friendly relations with Normandy. But nothing would be easier than for William to magnify the simple obligation of homage into a promise to hold the English Crown in fief or to make it over to William himself. Out of such a groundwork as this the elaborate Norman story could easily grow up. But if Harold expressly promised all that William of Poitiers makes him promise, it is not easy to see how the engagement to marry William’s daughter could ever, in so many versions of the tale, including some distinctly favourable to the Norman side, have usurped the place which would have properly belonged to the more important engagements about the Kingdom.

I will now, by way of relaxation, wind up this part of my subject with the most wonderful tale of all, that with which Gervase of Tilbury amused the Imperial leisure of Otto the Fourth. I will give only a summary, referring the reader to the original of the “Otia Imperialia,” Decisio Secunda, cap. xx., which will be found in Leibnitz’ collection of Brunswick historians, vol. i. p. 945.

The holy King Eadward left as his heir his nephew Harold, a prince whose virtues are set forth in the most glowing terms. In his boyhood he had been sent by his uncle for education to Normandy, according to the use of the English nobles, who commonly sent their sons into France, to learn the use of arms, and to lay aside the barbarism of their native tongue. The reigning Duke had a bastard son named William, who was brought up as his heir. The two lads formed a boyish friendship, and promised to marry each other’s sisters. King Eadward and the unnamed Norman Duke 694die about the same time; Harold succeeds his uncle in England, and William, after some opposition, succeeds his father in Normandy. The new Duke sends a message to the new King, calling on him to fulfil his promise by an exchange of sisters. Harold has now no mind for either marriage, but his own marriage with William’s sister is brought about by an accident. The King of the English sets out with a few companions on a yachting expedition. Stress of weather drives him to Flanders—we should surely read that, intending to go to Flanders, he was driven to Normandy—there he pretends that he is come to carry out his promise as to both marriages. He does marry the Duke’s sister, and takes her with him to England, and again promises to send his own sister to the Duke. Once at home again, he neglects his promise, and, puffed up with pride, he invades Scotland and defeats the Scots in a stoutly contested battle. Directly after this, he hears that William has landed in the south of England, and hastens to attack him. The Normans, few in number and fearing Harold’s prowess, defend themselves with a dyke. But God, who can conquer by few as well as by many, overthrows the proud English, whose King is either killed or escapes by flight. William then marches on London, marries Harold’s sister, and reigns over England by virtue of the marriage.

This is altogether the wildest of all the tales that I have come across. I need not stop to point out all its blunders, anachronisms, and confusions. The most amusing perhaps is the story of Harold’s youthful sojourn in Normandy. This is evidently a confusion between Harold and Eadward, and the reason which is given must be one which belonged to Gervase’s own day rather than to Eadward’s. But even this amazing fable is worth a moment’s thought, as it shows how strong the tradition was that some question about a marriage of some kind or other was the primary ground of quarrel between Harold and William.

IV. I have given in the text as the probable date of Harold’s visit to Normandy the only date which seems possible, namely 1064. Mr. St. John (ii. 226 et seqq.), who rejects the story altogether, enlarges on the varying statements as to the date. The earliest accounts give no date; some seem to place it very soon after the death of Godwine, others, sometimes indeed the same, place it very 695soon before the death of Eadward. The writers seem not to have stopped to think that these two events were thirteen years apart. Mr. St. John says, very truly, though with a curious confusion of the respective dates and values of his authors,

“Bromton, p. 947, places it in the fourteenth year of Edward, that is, A.D. 1056; Matthew of Westminster and Roger of Wendover in A.D. 1059; Henry of Huntingdon A.D. 1063; Ranulph Higden, iii. 283, in 1064; Hoveden, Malmesbury, Hemingford, Wace, Simeon of Durham, run through the whole gamut of chronology from the period immediately succeeding the death of Godwin to the period immediately preceding the death of Edward, so little possible did they find it to give any stability or coherence to their fable.”

He then goes on;

“Modern historians, discovering insuperable objections to all the earlier dates, imagine there are fewer obstacles in the way of adjudging the voyage to the last year of Edward’s reign. The selection seems unfortunate. At the time of the expedition against Conan the corn is said to have been almost ripe in the fields, which in Bretagne is never the case till towards the end of August or the beginning of September. Now from the most unimpeachable of all testimonies, we know that Harold was in Wales during the summer of A.D. 1065, overlooking the erection of the hunting-palace which he undertook to build for the pleasure of his brother-in-law. We may infer, though it is not stated, that Harold left Wales sometime before the end of August, because on the 24th of that month Caradoc, son of the murdered king Griffith, whose widow Harold had married, exterminated the earl’s workmen, and put a period to the construction of the palace. Immediately after this, that is, early in September, the insurrection took place in Northumbria, when Harold was at hand, ready at the King’s request to negotiate with the rebels at Northampton.

“From this view of the occurrences of A.D. 1065, it seems difficult, if not impossible, to imagine an unoccupied interval lying between midsummer and autumn long enough to admit of our crowding into it all the events which are said to have occurred during Harold’s imprisonment at Ponthieu and forced detention in Normandy.”

These arguments are, I think, decisive against 1065. Mr. St. John places the Northumbrian revolt in September instead of in October; 696still the English events of the autumn of 1065 hardly leave time for Harold’s captivity at Beaurain, his sojourn at Rouen, and his warfare in Britanny. But Higden’s date of 1064 seems open to no objection. Florence (see vol. ii. p. 474) certainly carries the Welsh war into that year; but the Chronicles leave it an absolute blank. At the same time, as I have already said, I do not commit myself to the date or to anything else. Mr. St. John’s arguments, though often expressed with needless violence, have throughout great weight as against the details of the story. But I still think that the story must have had some groundwork of truth, and I have tried to show what that groundwork may have been.

V. It would be an interesting question, how far Harold’s obligations to William, whatever they were, were known in England either before or after Eadward’s death. But this is a point on which we have absolutely nothing to guide us. I know of no writer who has anything to say on the subject, except the romantic Biographer of Harold, whom I have quoted in p. 601.

I have thus done what I could to throw light on the most perplexing question in my history, one of the most perplexing questions in all history. I shall not be surprised if I am thought to have only made what was before dark darker still. But no one can conceive how thick the darkness really is except by groping in it, as I have done, in his own person.

NOTE S. p. 221. The Ælfgyva of the Bayeux Tapestry.

There is no representation in the whole of the Tapestry which is more thoroughly puzzling than the one referred to in the text (pl. 4), with its legend “Ubi unus clericus et Ælfgyva.” Who is the lady, bearing a purely English name, who is thus suddenly introduced, seemingly at the gate of William’s palace, with no apparent reference to anything before or after? One would naturally look for the figure of William’s wife or daughter in such a position, rather than for that of any other woman. Harold’s promise to marry William’s daughter, which is so prominently dwelt upon in every other version of the story, is not once alluded to in the Tapestry, 697unless this place has reference to it. But how could William’s wife or William’s daughter be described by the familiar English proper name Ælfgifu? On the other hand, what chance is there that any Englishwoman, really bearing the name of Ælfgifu, could be present in Duke William’s palace at such a moment? And if any such Ælfgifu really was there, what bearing had her presence on the general course of the story, so as to account for the prominent position thus given to her?

Some of these difficulties naturally struck the very earliest commentators on the Tapestry, and from their days to ours a series of the wildest conjectures have been poured forth with regard to the “Ælfgyva” in question. The matter is treated of by Lancelot (Mémoires de l’Académie, viii. 612), by De la Rue and his translator Mr. Douce (Archæologia, xvii. 100), by Mr. Amyot (Archæologia, xix. 199), by De la Rue again in his Appendix of 1824 (Recherches sur la Tapisserie, p. 53), by Mr. Bolton Corney (p. 19), by Dr. Bruce (p. 53), and lastly by Mr. Planché (Journal of Archæological Association, 1867, p. 142). The strange thing is that several of these writers seem not to have understood that Ælfgifu is simply a very common English name, but to have fancied that it was a sort of title, meaning Queen or Princess. Their stumbling-block was the double name of Eadward’s mother, “Ælfgifu-Emma,” in which formula Lancelot argued that Ælfgifu was equivalent to Hlæfdige. Any one who turns to the passages which I have referred to will find a great number of guesses, some of which refute themselves, while others are refuted by other writers in the dispute. “Ælfgyva” has been identified with the Duchess Matilda, with her daughter Adeliza, with Harold’s sister Eadgyth and his wife Ealdgyth, while some have taken the trouble to show that she cannot be either Ælfgifu-Emma or “the other Ælfgifu” (see vol. i. p. 453). What it is that Ælfgyva and the clerk are doing no one seems to know for certain, neither can I throw any light on the matter. Out of all this mass I will only, by way of relaxation, quote Mr. Bolton Corney’s remarks, as at once the most curious and the least generally accessible.

“William promised to bestow one of his daughters on Harold. She is represented beneath the inscription ÆLFGYVA—but Elfgiva was not her name. Emma, daughter of Richard I. of Normandy, and mother of Edward the Confessor, is sometimes 698called by the Saxon annalists Elfgiva Emma. Elfgiva therefore, whatever we read in Florence of Worcester, seems to have been an appellation of honour, and may have been understood as such by the Saxons Bayeusains. If so, why was the name of the betrothed omitted? Could it not be ascertained, or was it deemed superfluous? I apprehend the latter to have been the case; she was the DAME par excellence—she was buried and was annually commemorated at Bayeux.”

We may infer then, First, that the Saxon language was spoken at Bayeux in the thirteenth century, the date to which Mr. Corney assigns the Tapestry; Secondly, that in the Saxon language of Bayeux Ælfgyva meant “Lady;” Thirdly, that one particular daughter of William was known, distinctively and familiarly, as “the Ælfgyva;” Fourthly, that Mr. Bolton Corney understood Old-English better than Florence of Worcester.

Now leaving all wild conjectures, let us try and see what really suggests itself about this obscure matter. The Tapestry represents a woman named Ælfgifu as being in Duke William’s palace at the moment of Harold’s coming thither. Who was she? We may put aside Matilda and all other women who never were, or could have been, called Ælfgifu. We may put aside all those women who were named Ælfgifu but who were dead and buried at the time. But of all the women named Ælfgifu who were living at the time, which could have been in William’s palace at that particular moment? Several guesses have occurred to me at different times. They are mere guesses, of no more value than the guesses of other writers. They are all, I allow, improbable guesses, but I think that they have the advantage over some other guesses of not being absolutely impossible.

1. In my second volume (p. 629) I threw out, half in jest, the suggestion that Ælfgifu, the name assumed by Emma on her marriage with Æthelred, was the name usually assumed by foreign women who married English husbands. Is it possible that there is really something in this? Is it possible that William’s daughter, if she had married Harold, would have had to change her name to Ælfgifu? Is it possible that she is here called Ælfgifu proleptically, perhaps sarcastically? This is, I grant, very far-fetched and unlikely, but it is perhaps not absolutely impossible. We should 699certainly expect the Tapestry to contain some reference to the intended marriage between Harold and William’s daughter. We should certainly expect to find William’s daughter, rather than any other girl or woman, represented where we find Ælfgifu represented. And here is a way, however far-fetched, in which it is just possible she might be called Ælfgifu.

2. Ælfgifu was (see vol. ii. p. 629) the name of the widow of Ælfgar, the mother of Harold’s wife Ealdgyth. According to some accounts, she was of Norman birth. Could she have been living or visiting in Normandy at this time? And can her introduction have any reference to Harold’s marriage with her daughter?

3. I have mentioned in my second volume (p. 554) the probability that Harold had a sister of the name of Ælfgifu, and that she must have been the sister whom Harold (Eadmer, p. 5; Sim. Dun. 1066) promised, as part of his oath, to give in marriage to one of William’s nobles. Is it possible that she was in Normandy at this time? If Harold’s voyage really was, as I believe it to have been, a mere yachting expedition, he may very well have been accompanied by his sister, as well as by his brother and his nephew. If it should be asked how Ælfgifu came to be in William’s palace while her brother was still a captive at Beaurain, it may be answered that even Guy may not have pressed his right of wreck so far as to imprison a woman, and that it is certain that one or more of Harold’s party escaped Guy’s clutches, if only to carry the news of his imprisonment to William (see above, p. 224). If therefore Harold was accompanied by his sister, it is quite possible she might find her way to Rouen before he did. I throw this out as a mere conjecture, and it certainly has its difficulties about it, but every explanation of this puzzling group must be mere conjecture, and it certainly strikes me that this conjecture has less of difficulty about it than some of the others.

Whomever we fix upon as the Ælfgifu of the Tapestry, it is still by no means clear what is happening between her and the clerk, or why the incident should receive so prominent a place in the pictured story. Like the introduction of Turold, Vital, and Wadard, there is evidently an allusion to some fact which was perfectly well known at the time, but of which no other record has been preserved. As such, it is another witness to the contemporary date, and thereby to the authority, of the Tapestry.

700

NOTE T. p. 229. The Breton Campaign of William and Harold.

Our only detailed accounts of this campaign come from William of Poitiers and the Tapestry. Between these two both Lord Lyttelton (i. 354) and Mr. Planché (145) see a distinct contradiction; only Lord Lyttelton assumes that the Tapestry must be wrong because it contradicts William of Poitiers, while Mr. Planché assumes that William of Poitiers must be wrong because he contradicts the Tapestry. But there is really no distinct contradiction between the two authorities; their accounts may easily be reconciled, if we only suppose a somewhat remarkable omission on the part of William of Poitiers.

William tells us that the object of the expedition was to deliver Dol, which was held on Duke William’s behalf by Rhiwallon, and that Conan fled on the approach of the Norman army. He gives no details of any further progress of William and Harold in Britanny, though he has a good deal to tell us as to what passed between William and Rhiwallon. He makes no mention of Rennes or of Dinan, the two other places represented in the Tapestry.

There is nothing in the Tapestry (pl. 5) which at all contradicts this account of what happened at Dol. William’s approach to the city is clearly not a hostile approach. The Duke himself and those immediately about him are not even in armour. There are no defenders on the walls, such as we presently see at Dinan. The legend is simply “Venerunt ad Dol,” while in the other case it is “Pugnant contra Dinantes.” But on the other side Conan and his host are running away—“Conan fugâ vertit.” I do not profess to explain the intentions of the man who is letting himself down by a rope from the wall.

Thus far the Tapestry and the Archdeacon exactly agree. The Tapestry also does not, as Mr. Planché understands it, make William pursue Conan to Rennes. Rennes—“Rednes”—is indeed introduced in the Tapestry, but it forms no part of the story; the city is represented on a smaller scale than the other subjects, and certainly no one is made to pursue the flying Bretons. Rennes is simply brought in as a kind of indication of the point to which Conan fled.

701The story of the siege of Dinan belongs wholly to the Tapestry. It is certainly strange that William of Poitiers should leave out all mention of so considerable an exploit, but the difficulty is a good deal lessened if we accept Wace’s statement (see p. 230) that Harold accompanied William on more than one raid into Britanny. That a siege of Dinan really is intended, and not a siege of Rennes, I have no doubt at all.

NOTE U. p. 259. The Embassies exchanged between William and Harold.

The different statements as to Harold’s oath naturally lead to statements equally different as to the messages which, after Harold’s election, were exchanged between him and William. Whatever Harold had promised, that William demanded, and whatever William demanded, that Harold refused. We thus get a set of statements differing exactly as the other set of statements differ. I have already had occasion to refer incidentally to several of them. I will now try to put them together in order.

Of the purely Norman authorities, the Tapestry is silent. William of Poitiers does not directly speak of any messages between the rivals till after William’s landing in England (128–131; see above, p. 430). But there is a passage further on which seems to imply an earlier message. This is when the Archdeacon breaks forth into that wonderful panegyric on his master (145) which follows his account of William’s coronation;

“Hic [Willelmus] ne Heraldum vellet occubuisse. Immò voluit patris Godwini potentiam illi ampliare, et natam suam, Imperatoris thalamo dignissimam, in matrimonium, ut fuerat pollicitus, tradere.”

This clearly implies that an offer of the Earldom of the West-Saxons, perhaps of something greater, together with the hand of William’s daughter, had been twice made by William to Harold. The first time is of course at the taking of the oath; the second time must be in some message sent before the expedition. For the messages exchanged after William’s landing are given at length, and they contain no such terms. And an offer of William’s 702daughter could then at least have been nothing but sheer mockery. By that time, at all events, Harold was married to Ealdgyth.

William of Jumièges (vii. 31) has an account which exactly agrees with the implied narrative of William of Poitiers. The Duke, as soon as he hears the news of Harold’s election, sends messengers to remind him to keep his oath. The terms of the oath, according to him, were that William should be King, Harold having half the Kingdom and William’s daughter. His words are,

“Ad quem [Heraldum] Dux protinùs legatos direxit, hortans ut ab hâc insaniâ resipisceret, et fidem quam juramento spoponderat condignâ subjectione servaret. At ille non solùm hoc audire contempsit, verùm omnem Anglorum gentem ab illo infideliter avertit.”

He then goes on with the account of Gruffydd and Ealdgyth which I quoted in vol. ii. p. 630.

Orderic does not mention the message. The Roman de Rou (vv. 11066–11075) speaks of several messages;

“Willame li manda sovent,

K’il li tenist sun serement,

E Heraut li manda vilment,

K’il ne fereit por li néient,

Ne il sa fille ne prendreit,

Ne il terre ne li rendreit,

Et Willéalme le desfia,

E desfiance li manda;

E Heraut toz tems reponeit

Ke nule rien mez ne creimeit.”

Here the terms of William’s messages are not given, but they may be inferred from the terms of Harold’s answer. These terms, and also the mention of more messages than one, seem on the whole to fall in with the account given by Eadmer and Simeon, which I have already mentioned (see above, p. 671). Harold there makes this answer, of which I do not fully understand the clause about the castle and well of Dover;

“Soror mea, quam juxta condictum expetis, mortua est. Quod si corpus ejus, quale nunc est, vult Comes habere, mittam, ne judicer sacramentum violâsse quod feci. Castellum Dofris et in eo puteum aquæ, licet nesciam cui, ut vobis convenit, explevi.”

He then goes on with the passages which I have already quoted (pp. 262, 264), in which Harold declares his inability either to decline the Kingdom which was offered to him or to marry a foreign woman without the consent of the Witan. William then sends a second message (see above, p. 260), of which the tone is greatly 703lowered. He now only asks that, if Harold will do nothing else that he has promised to do, he will at least marry his daughter. “Alioquin,” he continues, “se promissam regni successionem armis sibi vindicaturum proculdubiò sciret. At ipse [Haroldus] nec illud quidem se facere velle, nec hoc formidare respondit.”

It will be easily seen how completely this agrees with the account in the Roman de Rou, except in the prominence given to the engagement about Harold’s sister, which Wace does not mention at all. The statement about her death seems to me to bear the stamp of genuineness, as it is not a thing which anybody would take the trouble to invent. And it has sometimes struck me that this explains the statement of William of Malmesbury, where he makes Harold say that the daughter of William whom he had promised to marry was dead (see above, p. 660). In this statement he stands quite alone, and it seems not unlikely that he confounded the two marriage engagements, and that the death of William’s daughter is really a reproduction of the death of Harold’s sister. The rest of William of Malmesbury’s story, and the constitutional doctrine put into the mouth of Harold, I have mentioned already in the text and notes. See above, pp. 262–265.

The version of the Hyde writer I have already given at p. 692.

In Benoît’s account (36732–36757) the Kingdom alone is dwelt upon. He had mentioned the marriage (36622 et seqq.), but he seems to look on the marriage of Adeliza as something volunteered by William after the oath, not as part of the oath itself.

I do not know that there is any need to quote the accounts of any later writer.

NOTE W. p. 286. William’s Councils and Negotiations.

Two points of some difficulty meet us here. We have no trustworthy guide to the chronology of the different embassies and assemblies which William used to put forward his claim; doubts may also be raised as to the nature of the Norman Assemblies which he consulted. To fix the exact chronology seems hopeless; 704as I said in the text, several negotiations were most likely going on at the same time. And the writer who ought to be our best authority, William of Poitiers, is now, as usual, very careless about the order of events, arranging them not so much according to the almanac as according to any arrangement which may best suit his rhetoric. Two distinct questions however arise. First, Did the Assembly which William of Malmesbury places at Lillebonne meet before or after the gift of the consecrated banner by the Pope? Secondly, Was the English expedition discussed in one or in two Norman Assemblies?

On the first point, William of Malmesbury (iii. 238) distinctly puts the Lillebonne Assembly after the receipt of the banner; “Perpensis apud se utrimque partibus, Papa vexillum in omen regni Willelmo contradidit; quo ille accepto, conventum magnatum apud Lillebona fecit, super negotio singulorum sententias sciscitatus.” Wace, on the other hand (11436), places the embassy to the Pope last of all, after all the other negotiations.

On the second point, Wace (11120 et seqq.) is the only writer who distinctly marks the two assemblies, first, a small body of select counsellors who recommend the gathering of a larger assembly, and secondly, the larger assembly of the whole baronage of Normandy which is gathered according to their recommendation. The words of William of Malmesbury would seem to imply one assembly only.

Let us see whether anything can be got out of William of Poitiers to strengthen either view. Having recorded (p. 121) Harold’s accession after his own fashion, he tells us of a council held by William, in which many of his chief men dissuaded him from the undertaking as being one too great for the power of Normandy (“Dux Willelmus, habitâ cum suis consultatione, armis injuriam ulcisci, armis hæreditatem reposcere decrevit; tametsi complures majorum id ingeniosè dissuaderent, ut rem nimis arduam, Normanniæ viribus longè majorem”). He then enlarges on the number of wise and illustrious men whom Normandy then contained, and gives a list of them, which differs in one or two names only from the list which Wace gives of the former and smaller council. Then comes an account of the preparation of the ships, the pouring in of foreign soldiers, the delay at the mouth of the Dive. Then, and not before, we read of the embassies to the 705German King, the Pope, and the King of the Danes. Then comes the tale about Harold’s spies, which is followed by a great number of arguments against the expedition, which read very much like Wace’s account of the assembly of Lillebonne, and which I have not scrupled (see p. 294) to transfer to that gathering. One of them is that the ships, though seemingly already collected at the mouth of the Dive, could not be got together or supplied with crews within the appointed space of a year. The Duke then makes a speech in answer, and the fleet, which had been waiting in vain for the south wind at the Dive, sails for Saint Valery.

In a tale told in this way there is clearly no regard paid to chronology. The facts are much the same as the facts in Wace, but as to their order nothing can be made out. It will be remembered that several other incidents in the history, as the comet and some points in the engagements between William and Harold, are (see pp. 643, 687, 701) recorded or alluded to by William of Poitiers in places still more distant from their chronological order. The short account in William of Jumièges (vii. 31–34) proves hardly anything. The events which he records come in this order; the message to Harold; the coming of Tostig; the death of Conan; the preparation and voyage of William. Orderic (492–494) is fuller. He begins with the coming of Tostig, to whose suggestion he seems to attribute the gathering of the Assembly (see p. 303). His list of persons consulted is the list in William of Poitiers, with a few names added. Then comes the embassy to Rome and the gift of the banner. Then comes the unsuccessful enterprise of Tostig against England, followed by William’s own preparations. Orderic then, after his custom, leaves the subject to talk about quite other matters. He comes back to it after some while (499, 500), to tell of the Norwegian invasion of England, of the delay of the Norman ships at the Dive, of William’s final voyage and all that followed it.

The short and inadequate account in Henry of Huntingdon (M. H. B. 761, 762) is of some importance, because part of it is clearly drawn from the same sources as the account in Wace. He tells us that William held an Assembly, and that his fleet afterwards met at Saint Valery. He says nothing about any embassies at all. But he is the only writer besides Wace who makes any special mention of William Fitz-Osbern, though the way in which 706he tells the story is not exactly the same as that in which it is told in the Roman de Rou. See p. 297.

The evidence being in this state, I thought that I might safely follow Wace. His account of the two assemblies is full and clear, and it derives a sort of incidental support from William of Poitiers and Henry of Huntingdon. The former seems in a manner to imply that there were two assemblies, though he puts the second out of its place. Following Wace thus far, I have also followed him in placing the Assembly at Lillebonne before the receipt of the banner from Rome, though William of Malmesbury asserts the exact contrary. The reader must judge for himself which order he thinks the more probable. The approval of the Pope would of course be likely to tell with great effect on the mind of the Assembly. On the other hand, it would be a great point for William to be sure of the support of his own Duchy before he applied to the Pope. It would be hardly like the wisdom of William to ask, or like the wisdom of Hildebrand to grant, a blessing on an undertaking which might easily fall through altogether. And the religious excitement would probably tell less on William’s own subjects than on the mixed multitude of Crusaders who flocked to him from all parts. But the question—not one of any great importance—must be left an open one.

As to the embassies to foreign powers, the application to the Pope is mentioned in every account which tells the story in any detail. It is in fact the centre and soul of the whole business. All accounts again agree as to the influx of foreign soldiers of all kinds. But as to the particular potentates applied to, our two fullest accounts differ. William of Poitiers speaks of the applications to Germany and Denmark, Wace of those to France and Flanders. There is here no contradiction. The object of William of Poitiers was to exalt his hero, and his own mind was deeply impressed with Imperial ideas. The distant embassies, above all the embassy to the future Emperor, were those which tended most to set forth the greatness of the Norman Duke; they were those therefore on which the loyal Archdeacon was most inclined to dwell. It was equally natural that Wace, writing at a later time, should think most of those nearer embassies which ordinary Norman tradition would be most likely to understand and to remember. His account of the 707dealings of William with the King of the French is straightforward and probable enough, except in the omission of all mention of Baldwin of Flanders in his character of Regent. I have therefore followed it in the text. But of the intercourse between William and Baldwin in his character of sovereign of Flanders, Wace has (11390–11432) a tale which strikes me as so purely legendary that I did not venture to introduce it into the text. The Duke sends to Baldwin, who is spoken of as his brother-in-law, not as his father-in-law (“cum od serorge”—sororius—“et od ami” in v. 11392; and again, “vostre seror et vos nevoz” in vv. 11424, 11425), and asks his help. Baldwin answers that he must first know what share of England he is to have for his pains. William, who seems to be conceived as talking with Baldwin face to face, says that he must go home and consult his Barons, and that he will then send word by letter what the result is.

“E li Dus dist k’il s’en ireit,

A sez Barunz en parlereit,

Et a els s’en cunseillereit,

E ço ke l’en li loerreit

Par son bref li remandereit.”

(vv. 11399–11403.)

No consultation with the Barons seems to follow, but the Duke does a thing which nobody had ever done before (“poiz fist ço, ke ainz ne fist nus”). He takes a small piece of parchment on which nothing is written, then seals it up with wax, and causes to be written on the outside that the Count shall have such part of England as is stated in the inside of the letter.

“De parchemin prist un petit

K’il ni out leitre ne escrit,

Tot voi le séela en cire,

Et en la coe fist escrire,

Ke d’Engleterre tant areit

Comme li brief dedenz diseit.”

(vv. 11406–11411.)

The packet is sent to Baldwin by a cunning varlet (“vaslet enloçonez”); the Count breaks the seal, looks inside—seemingly he could read—finds nothing, and shows it to the varlet. The varlet then says that there is nothing there and that Baldwin shall have nothing. The honours which the Duke was seeking would belong to Baldwin’s sister and nephews. If Baldwin had joined in the enterprise no one would have gained more by it than himself. As it was, William would with God’s help conquer England for himself without help from Baldwin. Wace adds, as he so often does, that he does not know what answer the Count made.

708The sort of practical joke described in this story would be quite in keeping with one side of William’s character (see p. 162) if one could only see the point of the joke. But that point is, to say the least, not very obvious, and the whole story seems quite inconsistent with the real relations between William and Baldwin. See the note of Prevost, ii. 137, and Taylor, 110.

NOTE X. p. 300. The Movements of Tostig after his Banishment.

The presence of Tostig in Normandy is asserted by Orderic in the passages which I have quoted in the text, and it is implied in the short narrative of William of Jumièges (vii. 32), who simply says, after describing Harold’s accession, “Porro Dux Tosticum Comitem in Angliam misit, sed militia Heraldi mare servans eum armis abegit” (see p. 304). William of Poitiers and Wace make no mention of any visit of Tostig to Normandy; they do not speak of him at all till they come to their almost incidental mention of Harold Hardrada’s invasion of England. The English writers also make no mention of any visit to Normandy on Tostig’s part. He goes to Flanders and comes on his vain expedition to England in May “from beyond sea” (see p. 324), which of course leaves the matter open between Normandy and Flanders. Snorro again knows nothing of a visit to Normandy, but takes Tostig (see p. 329) from Flanders to Denmark by way of Friesland. I do not however wholly reject the account of Orderic. The omission of Tostig’s visit both by William of Poitiers and the English writers is not at all hard to understand, while it is not easy to see why William of Jumièges or Orderic should have invented it. Its omission by Snorro is still less surprising; it is hard indeed to reconcile his version with that in the English Chronicles even on those points which are the very life of his story.

Orderic’s story however cannot be accepted in its details. First of all, it involves (see vol. ii. p. 626) the central mistake of making Tostig’s banishment follow the accession of Harold. Secondly, as a necessary consequence of this mistake, it greatly hurries Tostig’s movements, making him go to Flanders, leave his wife there, and 709hasten to Normandy, whereas we know that he left England on November 1st, 1065, and stayed with Baldwin the whole winter (see vol. ii. p. 501). Thirdly, it represents Tostig’s first expedition as not reaching England at all, whereas we know that he landed and harried the coast at several places (see p. 326). But, except the first mistake, which we know how to account for and to correct, these are not very fatal difficulties. Tostig did reach England, but he did nothing there of any moment, and a foreign writer might easily speak of an expedition which so utterly missed its mark, as if it had never touched English ground at all. And, small as the point is, it is worth noticing as an undesigned coincidence that Orderic makes Tostig set sail from the Côtentin, while the English Chronicler makes him land in the Isle of Wight. I have therefore not hesitated to accept the main outline of Orderic’s story, and to represent Tostig’s first expedition, in May, 1066, as undertaken with at least the connivance of William. Tostig must have had some force, Norman or Flemish, adhering to him during the whole time. The presence of Flemings at Stamfordbridge is beyond doubt. See p. 371.

What follows next in Orderic and William of Jumièges is far harder to reconcile with our own Chronicles. They both make Tostig, after struggling in vain against many winds (“Zephyro Notoque aliisque ventis alternatim impellentibus angores multos pertulit,” Orel. Vit. 593 C), make his way at last to Harold Hardrada in Norway. “Post plurimos labores ad Heraldum Regem Nortwigenarum, qui Harafagh cognominabatur, accessit,” says Orderic. So William of Jumièges (vii. 32); “At ille non valens salubriter Angliam introïre, neque Normanniam, quia ventus obstabat, redire, Heraldum Herfagam, Northwegæ Regem, adiit.” There is nothing about his going to Scotland or to Denmark. Wace indeed takes him to Denmark, but seemingly only through confounding Denmark and Norway.

“Tosti, ki mult s’en coroça,

En Danemarche trespassa,

Daneiz è Norreiz amena,

Deverz Euroïc ariva.”

(vv. 11803.)

So Benoît, who carries Tostig to Norway in v. 36842 (“vers Northwege l’estut sigler Par merveilles orrible mer”), calls the force which he brought into England “ceus de Norwege” in v. 37065, and “Daneis” in 37103. None of these writers know anything of 710the double negotiation, first with Swend, then with Harold, which is so prominent in Snorro. Again, the Norman accounts take Tostig and Harold Hardrada straight from Norway to Yorkshire. The “erroneus exsul” pleads his cause before the tyrant, as Orderic (493 D) himself calls Harold, though Tostig is made to address him by many respectful titles (“Sublimitatem vestram, magnifice Rex”—see vol. i. pp. 63, 286—“supplex adeo, et me servitiumque meum Majestati vestræ fideliter offero”). He tells how his younger brother has risen against him and deprived him of the honours—whether the Earldom of the West-Saxons or the Empire of Britain—which were due to him by hereditary right (“ut possim restitui per vestrum suffragium honori ex paternâ successione debito. Nam Heraldus frater meus, qui jure mihi, utpote primogenito, debuisset parere, fraudulenter insurrexit contra me”). He goes on to enlarge on his brother’s perjury (“regnum Angliæ perjuriis præsumpsit usurpare”), an argument which, however telling with William, sounds strangely out of place when addressed to Harold Hardrada. He proposes to Harold of Norway to overthrow Harold of England, and to occupy his Kingdom, granting half of it to Tostig in fief (“proterviam perfidi fratris bello proterite, medietatem Angliæ vobis retinete, aliamque mihi, qui vobis inde fideliter serviam dum advixero, retinete”). Harold consents, he occupies six months in preparations, and sets sail in August. Orderic therefore conceived that Tostig’s banishment, his journeys to Flanders and Normandy, his attempt on England, and his escape to Norway, all took place in the course of January and February. In August then Harold and Tostig set sail together and sail straight for Yorkshire (“mense Augusto Heraldus Noricorum Rex cum Tostico et ingenti classe immensum pelagus intravit, et, Aparciate seu Boreâ flante, ad Angliam applicuit et Eboracensem provinciam primitùs invasit,” 499 D, 500 A).

It is clear that this account differs in some respects from that of Snorro, which I have given in the text. Snorro makes Tostig leave Harold in the spring (see p. 333) and go and collect English and Flemish troops in Flanders, with whom he meets Harold at some point of his voyage. The stoppage in Orkney again (see p. 344) is left out by Orderic, which is not very wonderful.

When we turn to our own Chronicles, we see that whatever amount of truth the stories of Orderic and Snorro may contain, 711their chronology at least is wholly wrong. Tostig’s first expedition to England, which Snorro leaves out, and which Orderic places at some time in January or February, did not happen till May (see above, p. 708). After its failure, Tostig went, not to Denmark or Norway, but to Scotland, and stayed there all the summer (see p. 327). The question now comes, which I left doubtful in the text, whether he ever went to Norway at all. Let us look at the different accounts.

The Abingdon Chronicle, which is followed by Florence, having mentioned the summer sojourn of Tostig in Scotland, the preparations of Harold of England in the south, and the return of his fleet to London (see p. 338), goes on; “þa ða scipu ham coman, þa com Harold cyning of Norwegan norð into Tínan on unwaran ... and Tostig eorl him com to mid eallum þam þe he begiten hæfde, eallswa hy ær gesprecen hæfdon” (“ut priùs condixerant,” Fl. Wig.).

The Worcester and Peterborough Chronicles record the flight of Tostig to Scotland, and at once connect it with the Norwegian invasion. “And he [Tostig] for to Scotlande mid xii. snaccum, and hine gemette [hine gemette þær. Wig.] Harold se Norrena cyng [cyng of Norwegon. Wig.] mid ccc. scipum; and Tostig him to beah [him to beah and his man wearð. Wig.], and hi bægen foran into Humbran oð þæt hi coman to Eoferwíc.”

This version is followed by William of Malmesbury (ii. 228); “Tostinus ... versus Scotiam vela convertit: ibi Regi Noricorum, Haroldo Harvagre, obvio manus dedit, qui cum ccc. navibus Angliam aggredi meditabatur. Ambo ergo consertis umbonibus terram Transhumbranam populabantur.”

These two accounts are clearly independent. Abingdon puts the meeting of Harold and Tostig in the Tyne. Worcester and Peterborough seem to put it in Scotland. But their words might be taken as leaving the place uncertain. Nor is there any contradiction in the more important question whether the expedition had been already planned between Harold and Tostig. This is directly asserted by the Abingdon Chronicle. The two others would certainly, taken by themselves, be understood to imply that Harold Hardrada had already set out on his own account, and that his meeting with Tostig was not the result of any agreement. This appearance becomes still stronger in William of Malmesbury, from his use of 712the verb “meditabatur.” But there is no word exactly answering to this in the Chronicles, and the difference in the two accounts may be simply one of omission. The words “eallswa hy ær gesprecen hæfdon” may in themselves mean anything, from a mere agreement to join forces in the Tyne made after an unexpected meeting in Scotland up to the voyage to Norway spoken of by Snorro and Orderic. The question is whether the chronology will admit of that voyage. Snorro and Orderic, as we have seen, place it early in the year. If we give up the story of Tostig’s journey to Normandy and his concert with William, it would be easy to take Tostig into Norway before his first attack on England, at any time between January and May. But it is plain that he did not go to Norway immediately on the failure of that expedition, for he went to Scotland and stayed there all the summer. And it is hard to find room between the end of summer and September 25th for a voyage of Tostig to Norway, for the preparations of Harold Hardrada, and for his voyage. A communication by messengers is more possible, as that may be spread over all the time from May to September. But, if we make Tostig visit Norway so late in the year, it could not have been his visit which first suggested to Harold to invade England. Indeed the invasion could hardly have been first suggested by letters or messages sent so late as May.

We must then either reject the partnership between William and Tostig in the May expedition and carry Tostig into Norway early in the year, or else we must suppose that Harold Hardrada was already planning an attack on England, and that all that any visit or message of Tostig did was to strengthen and hasten a purpose already formed. On the whole I incline to this last view, because I cannot see why any Norman writer should have invented a connexion between Tostig and William, if it never happened. But if we reject the Norman story, we can accept nearly the whole tale in Snorro. Tostig goes to Denmark and Norway early in the year; he then returns to Flanders and collects a force. Snorro does not say what he did with it till he met Harold in the autumn. The gap will thus be filled by the May voyage to England and the summer in Scotland. I am quite content to leave the point open, but I would remark that it is at least as hard to reconcile the Norman and Norwegian accounts with one another as to reconcile either with the English account.

713The Hyde writer, as usual, gives (292) an independent account which is worth a passing notice.

“Frater ejus [Haroldi], nomine Tostius, genere et animo temerarius, mox ut fratrem regium honorem usurpâsse vidit, nutu Domini, qui perjurium undique persequebatur, quâdam simultate inter eos ortâ, Angliâ discessit, atque cum multâ manu et thesaurorum suorum copiâ multitudinem Flandriam ad Sanctum Odmarum, ut aiunt, devenit, ubi commendatâ thesaurorum suorum copiâ navem ascendens insulam Norweiam, quondam Scanciam dictam, et ut Gothorum historia testatur, multarum gentium officinam, petivit; quam ingressus, Regem ejus Haroldum, cognomento Hervard [Harfragi], id est, crine formosum, erat enim et staturâ corporis et formâ decorus, tum precibus tum promissionibus ita illexit, ut congregato exercitu et classe paratâ, cum eodem Angliam ad debellandum Regem Haroldum festinus adveniret. Denique Angliam ingressi, Eboracam opulentissimam civitatem et archiepiscopatûs sedem obsidione cingunt.”

We have already (see vol. ii. p. 231) seen this writer talking about the “Isle of Norway,” when, by the way, he really meant Denmark. We here see where he found it, namely in the opening chapters of Jornandes, especially the well-known passage in c. iv. (Muratori, i. 193); “Ex hâc igitur Scanziâ insulâ, quasi officinâ gentium, aut certè velut vaginâ nationum, cum Rege suo nomine Berig, Gothi quondam memorantur egressi.”

NOTE Y. p. 313. Arnold of Ardres.

I get the account of these adventurers from the “Historia Comitum Ardensium” in Bouquet, xi. 305. The office held is described as the “villicatura sive præpositura Sancti Bertini in terrâ Ghisnensi,” which had been held “hæreditario jure” for some generations. Arnold came “opitulante Boloniensi Comite Eustachio,” and Geoffrey seems to follow him at William’s own summons—“vocatus ab eodem Rege Willelmo.” Their rewards are thus described; “Servientes igitur ambo fratres, Arnoldus videlicet et Gaufridus, jam dicto Regi, tantam ejus adepti sunt gratiam quod, præter quotidiana stippendia et munuscula, quæ ipsis contulit 714innumerabilia, contulit etiam eis et in perpetuitatis concessit feodum, Stebintoniam et pertinentias ejus, Dokeswordiam, Tropintoniam, Ledefordiam, Toleshondiam, et Hoilandiam.” All these, as far as I can identify them in Domesday, are possessions of Eustace, but in two cases only do I find Arnold as under tenant. These are at “Dokeswordia,” in Domesday “Dochesuuorde” (196), and Trumpington (ib.), both in Cambridgeshire. In the former we read, “hanc terram tenet Hernulfus de comite Eustachio,” and under Trumpington we have still more distinctly, “tenet Ernulfus de Ardâ sub Comite.” Stebingtonia, in Domesday (205) Stebintune, is held of Count Eustace by Lunen, and the lands in Essex, Selefordia, Hoilandia, and Toleshondia, in Domesday Toleshunta, now Tolleshunt, are held of Eustace (Domesday, ii. 32, 33) by Adelolfus. This may be an English Æthelwulf; but, as the name also occurs in Flanders (see Cart. S. Bertin, 142, 153 et al.), he may equally well be a foreign follower of Eustace. Arnold however appears again as Ernulfus de Arde, as a tenant under Count Eustace of various places in Bedfordshire (211), and one might almost guess that he is the same as Arnulfus de Hesding (no doubt Hesdin in Ponthieu) who holds (205 b) Upford in Huntingdonshire of the King. Arnold and Arnulf are, I need not say, names whose various forms are often and easily confounded. This is a fair specimen of the way in which men came to William from all parts of the world for whatever they could get.

NOTE Z. p. 315. The Death of Conan.

The only place that I know of where William is directly charged with poisoning Conan is in the speech (Ord. Vit. 534 B) at the bride-ale of 1076, where the charge is coupled with the kindred charge of poisoning Walter and Biota (see p. 207). The revellers are made to say, “Conanum quoque, strenuissimum Consulem, veneno infecit, quem mortuum Britannia tota pro ingenti probitate ineffabili luctu deflevit.”

The text of William of Jumièges (vii. 33) is as follows;

“Tempore quo Willelmus Dux disponebat Angliam adire, et armis eam sibi vendicare, audax Chunanus Comes Britanniæ nisus est eum, 715missâ legatione hujusmodi, terrere: Audio te, inquit, nunc velle trans mare proficisci, et Angliæ tibi regnum nancisci. Inde multum gaudeo, sed ut mihi Normanniam reddas obsecro. Robertas autem Dux Normannorum, quem tu fingis esse patrem tuum, iturus in Hierusalem, Alanno patri meo, consobrino scilicet suo, commendavit omnem suam hæreditatem. Tu autem cum complicibus tuis Alannum patrem meum apud Winmusterium in Normanniâ veneno peremisti, et terram ejus, quam ego quia puer eram possidere nequibam, invasisti; et contra fas, cum sis nothus, hucusque tenuisti. Nunc igitur aut mihi debitam redde Normanniam, aut ego tibi totis viribus bellum inferam. His auditis, Willelmus Dux aliquantulum territus est. Sed mox eum Deus, frustratis inimici minis, eripere dignatus est. Unus enim ex proceribus Britonum, qui utrique Comiti juraverat fidelitatem, et hujusmodi legationem inter eos ferebat, lituum Chuningi, et habenas, atque chirothecas intrinsecus livit veneno. Erat quippe cubicularius Chuningi. Tunc idem Comes Britonum in Andegavensi Comitatu Castellum-Guntherii obsederat, et oppidanis militibus sese illi dedentibus suos intromittebat. Interea Chuningus chirothecas suas incautè induit, tactisque habenis, manum ad os levavit. Cujus tactu veneno infectus est, et paullò post omnibus suis lugentibus defunctus est. Hic multum sagax fuit et probus, ac amator justitiæ. Qui si diu vixisset, multa bona ut fertur fecisset, ac ad regendum honorem utilis fuisset. Proditor autem conscius sui reatûs, mox de expeditione aufugit, et mortem Chuningi Willelmo Duci mandavit.”

This story, in the way in which it is brought in, looks very like an interpolation, and the message of Conan sounds very like a romance. And it is certainly most remarkable that it seems to be a purely Norman story. At least I have not been able to find it in such Breton and Angevin chronicles as I know anything of. The Breton and Angevin writers record Conan’s war with Anjou and also his death, and they place both in 1066. But they say nothing which at all lays his death to the charge of William. Of three Breton chronicles in the collection of Morice (Mémoires pour servir de Preuves à l’Histoire de Bretagne), the first in the collection says merely, “1066. Cometa apparuit. Obiit Conanus Dux Britanniæ filius Alani. Normanni Angliam ceperunt.” Another, the Chronicle of Saint Brieuc (p. 36), mentions the war with Anjou. Conan “quum territorium Andegavense devastâsset, in eodem territorio, paullò ante destructionem 716Heraldi Anglorum Regis, sine liberis morte præventus est anno Domini 1066.” A third Chronicle (p. 102) tells us how, in “1066, Comes Britannorum Conanus juvenis et malitiosus, Andegavorum terram adorsus, superbæ pervasioni suæ in ipsâ Andegavorum terrâ morte subitâ præreptus est.” The Angevin Chronicle which I have quoted at p. 642, after its account of the Comet and of the invasion of England, mentions the death of Conan in nearly the same words as the third Breton Chronicle just quoted. These accounts connect the death of Conan in some way, if only by way of coincidence, with the Conquest of England, yet not one of them breathes the least suspicion against William. Conan’s death was sudden; but a sudden death need not be a death by poison, and a death by poison need not have been brought about by the devices of William. I cannot think that we have evidence enough to charge the great Duke with so infamous a crime.

NOTE AA. p. 338. The Operations of the English Fleet in 1066.

Did the English fleet, or any part of it, ever encounter the Norman fleet or any part of it? The general run of our narrative would lead us to say, No; but there are some passages which look the other way. Thus the Peterborough Chronicler, immediately after describing Harold’s election, says, “And þy ylcan geare þe he cyng wæs, he for út mid sciphere togeanes Willelme, and þá hwile com Tostig eorl into Humbran míd lx. scipum.” And this seems to be followed by Henry of Huntingdon (M. H. B. 762 A); “Quod audiens Rex Haraldus, vir bellis acerrimus, cum navali exercitu contra Willielmum Ducem in mare profectus est.” These expressions do not necessarily imply a battle, but they seem to point to some operations beyond merely watching the coast.

Of the entries in Domesday referred to in the text, one, that about Æthelric of Kelvedon, distinctly asserts a battle, but without mentioning its date. It occurs under Essex, ii. 14 b. Æthelric held Kelvedon T. R. E. The Survey adds, “Hic supradictus Ailricus abiit in navale prœlium contra Willelmum Regem.” This 717seems to imply an actual engagement, though of course it need not have been a general engagement between the whole of the two fleets, which seems quite impossible. The entry goes on to say that Æthelric on his return—when the fleet returned to London in September?—fell sick, perhaps from a wound, and left his lands at Kelvedon to Saint Peter at Westminster. “Quum rediit, cecidit in infirmitate, tunc dedit Sancto Petro istud manerium.” The words which follow in Domesday have an importance of another kind, which I shall discuss in a future volume.

The other Domesday entry referred to is less distinct than that of Æthelric, but it looks the same way. In Norfolk, ii. 200, we find mention of one Eadric, described as “Rector navis Regis Edwardi.” On William’s accession he was outlawed and fled to Denmark (“postquam Rex W. venit in Angliam fuit iste Edricus exlex in Daciam”). One may guess that Eadric commanded in the engagement or skirmish in which Æthelric was concerned. And in connexion with this East-Anglian entry, we may take the statement of John of Oxenedes (293), about Ælfwold, Abbot of Saint Benet’s; “Huic a Rege Haraldo marina committebatur custodia.”

There are some other passages which might seem to imply naval operations at a later time. Of the losses of William’s fleet during the voyage, and of the affair of Romney, whenever it happened, I have spoken in the text, pp. 410, 534. A deed in the Cartulary of the Holy Trinity at Rouen (pp. 453, 454) looks the same way. One Roger, the son of Turold, who was going to join William’s expedition (“ultra mare cum Willelmo Comite navigaturus”), gave lands to the monastery, but died on the voyage before the gift was complete (“in eâdem navigatione morte præventus, hoc confirmare non valuit”). We must also remember the account given by William of Poitiers (131), where he describes Harold’s march into Sussex, and says that an English fleet of seven hundred ships was sent somewhere or other to cut off the Norman retreat (“ne perfugio abirent, classe armatâ ad septingentas naves in mari opposuerat insidias”). Guy of Amiens (319) puts nearly the same statement into the mouth of one of the messengers between William and Harold;

“Per mare, per terram, prœlia magna parat.

In mare quingentas fertur misisse carinas,

Ut nostri reditûs præpediatur iter.”

718But there is no sign of this great fleet doing anything, and William and Guy are so careless of chronology that they are quite capable of meaning the fleet which went back to London in September. As for the affair of Romney, it is not likely that Thegns from Norfolk and Essex would be concerned in an action so purely local. And the story of Æthelric would seem to imply that he died before the end of Harold’s reign.

NOTE BB. p. 357. The March of Harold to York.

The words of some of the Chroniclers, taken literally, might imply that the first news of the landing of the Northmen was brought to Harold of England after the Battle of Fulford. The Peterborough Chronicle describes that battle, and adds “and se Norrena cyng ahte siges geweald.” It then goes on, “And man cydde Harolde cyng hu hit wæs þær gedón and geworden, and he com mid mycclum here Engliscra manna, and gemette hine æt Stængfordes brycge.” So too the Worcester Chronicle describes the battle, and adds “ac þa Normen ahte sige.” It then goes on in the same way; “Man cyðde þa Harolde Engla cynge þæt þis wæs þus gefaren; and þis gefeoht wæs on Vigilia Sc̃i. Mathei. Ða com Harold ure cyng on unwær on þa Normenn, and hytte hi begeondan Eoforwic æt Steinford brygge, mid micclan here Englisces folces.”

The literal meaning of these accounts would certainly be that the news of the defeat of Fulford was the first news of the Norwegian invasion which reached Harold, and that he started for his Northern march on hearing it. But this is simply impossible. The Battle of Fulford was fought on Wednesday, and the Battle of Stamfordbridge was fought on the Monday following. As the Worcester Chronicler emphatically says, “þas twa folcgefeoht wæron gefremmede binnan fif nihtan.” Now for news of a battle fought close to York to reach London, for an army, including men from distant shires, to be collected and to march to Stamfordbridge, and all in the space of five days, would need an age of railways and telegraphs. The news must have been brought, and the 719march must have begun, before the Battle of Fulford was fought. In fact our remaining Chronicle gives us a hint that that battle was fought while Harold was already on his march. The Abingdon narrative runs thus;

“And foran þa begen [Tostig and Harold Hardrada] mid eallum þam liðe andlang Use up to Eoferwic ward. Ða cydde man Harolde cynge be suðan, þa he of scipe cumen wæs, þæt Harold cyng on Norwegan and Tostig eorl wæron up cumene neh Eoferwic, þa for he norðweard dæges and nilites, swa hraðe swa he his fyrde gegaderian mihte. Þa ær þam þe se cyning Harold þyder cuman mihte, þa gegaderode Eadwine eorl and Morkere eorl of heora eorldome swa mycel werod swa hi begitan mihton.”

Then follows the Battle of Fulford. The same account is followed by Florence;

“In loco qui Richale dicitur applicuerunt. Quod ubi Regi Haroldo innotuit, versus Northhymbriam expeditionem properè movit. Sed priusquam Rex illuc veniret, duo germani Comites,” &c.

Here it is not the news of the Battle of Fulford, but the news of the landing at Riccall, on hearing of which Harold sets forth. This is at least possible, as we have no distinct statement how long a time passed between the landing and the battle. But the story certainly reads as if the battle followed very fast upon the landing, and as if Harold must have been on his march, not only before the battle but before the landing. And indeed some news must have reached him of the approach of the Norwegian fleet, of the muster in the Tyne, and of the ravage of the Yorkshire coast. One cannot help thinking that the Chroniclers, even the Abingdon Chronicler, have fallen into a certain inaccuracy of expression, and that Harold must have known of the approach of the Northern enemy at a much earlier time than their words would imply. And yet, after all, the inaccuracy is hardly a literal one. News both of the landing and of the battle would reach the King on his march and would stir up him and his army to still greater exertions. The great march of Harold is in itself one of the most wonderful things in our wonderful history. But, if we take the words of the Worcester and Peterborough Chronicles in their literal and grammatical sense, the march becomes not only wonderful but miraculous.

720

NOTE CC. p. 363. The Details of the Battle of Stamfordbridge.

Years ago, when I first began these studies, one of my greatest difficulties was that Harold of England is described, in all the ordinary histories, as furnished with a strong force of cavalry and archers at Stamfordbridge, while he was utterly destitute of both arms at Senlac. The haste of his march southwards would not of itself account for the difference; for if he could have collected cavalry and archers for the one campaign, he could doubtless have collected them again for the other. I soon saw that the only authority for the usual description of the fight at Stamfordbridge was the Saga of Harold Hardrada, and I gradually came to the conclusion that that part of the Saga was mythical. The whole conception of the English army was clearly taken from an English army of Snorro’s own age, of which horsemen and archers undoubtedly formed the most important part. That is to say, the English had by that time adopted the Norman tactics. Indeed some of the incidents in Snorro’s account of Stamfordbridge seem very much as if they had been transferred thither from Senlac. The defeat of the Norwegian army, just like that of the English army at Senlac, is owing to their breaking the line of the shield-wall, and Harold the son of Sigurd is killed by the chance shot of an arrow, just like Harold the son of Godwine. But though the account of the Battle of Stamfordbridge is clearly mythical, the like is not the case with the whole of the story. The writer shows a knowledge of the Yorkshire coast, and his narrative seems quite trustworthy up to the Battle of Fulford. His account of that battle quite agrees with the appearance of the site (see p. 350). He seemingly confounds Waltheof with Eadwine (see p. 351), but then he utterly confounds all English genealogies and personalities (see vol. ii. p. 553). It is only when he gets to Stamfordbridge that he begins wholly to break down. For instance (see p. 354) he fancies that place to be close under the walls of York, whereas it really is eight miles off. May not the difference in value between these two parts of his narrative be explained by the circumstances of the case? A division of 721the army remained at Riccall (see p. 348), and never went to Stamfordbridge at all. All the men of this division (see p. 375) went back safe to Norway, while hardly any fugitives escaped from Stamfordbridge. It was therefore only natural that the story of the former part of the campaign, taking in the Battle of Fulford, should be much better known in Norway than the details of the greater battle itself. The Saga-maker therefore had trustworthy tradition to go by for one part of his story, while for the other he had to draw largely on his imagination.

I have drawn my own ideas of this, as of other battles, from an examination of the ground compared with the accounts of the original writers. I have twice visited the field of Stamfordbridge, in July and in December 1867, the former time in company with Archdeacon Jones and Mr. J. R. Green, when I also examined the site of Fulford with the Archdeacon. In December I also visited Aldby and the ground between Aldby and Stamfordbridge. I have compared the impressions thus formed with the short accounts in the Chronicles, and with those in Henry of Huntingdon and William of Malmesbury. The narrative of Henry has a special value. As soon as he reaches the actual fight, his narrative, hitherto meagre and inaccurate, suddenly lights up, and becomes minute, poetical, and evidently founded on an accurate knowledge of the spot. That is to say, his description here, as in so many other cases, is founded on contemporary and local ballads, of whose words distinct traces may be recognized in his narrative. An English ballad of Stamfordbridge must have been absolutely contemporary with the event—made perhaps to be sung at King Harold’s feast of victory. Such a relic, did we possess it in full, would be almost more precious than the songs of Brunanburh and Maldon.

The accounts in Lappenberg, Thierry, and St. John seem to have been written without any knowledge of the ground. Lappenberg puts the single-handed defence of the bridge quite at the end of the battle. This is perhaps merely an attempt to patch on this incident to the account in the Saga, or it is perhaps because the story comes at the end of the Abingdon Chronicle. But it comes at the end of the Chronicle, simply because it is an addition by another hand, and the words in which the anecdote is told show that it did not take place at the end of the battle. Any one who compares the ground with the account in Henry of 722Huntingdon will easily see that the defence of the bridge was more than a mere incident after the battle was decided. It is, I feel sure, the central point of the whole fight. Henry’s account begins (M. H. B. 762 B);

“Pugna igitur incepta est, quâ gravior non fuerat. Coeuntes namque a summo mane usque ad meridiem, quum horribiliter ruentes utrimque perseverarent, maximus numerus Anglorum Norwagenses cedere sed non fugere compulit. Ultra flumen igitur repulsi, vivis super mortuos transeuntibus, magnanimiter restiterunt.”

Then follows the defence of the bridge, as I have quoted the account in p. 370, and the narrative goes on thus;

“Transeuntes igitur Angli Haroldum Regem et Tosti occiderunt, et totam Norwagensium aciem vel armis straverunt vel igne deprehensos combusserunt.”

Here are two distinct acts of the battle, fought on opposite sides of the Derwent. As the English were coming from York, the first act must have been on the right bank of the river, and the second on the left. The crossing of the Derwent is the main point, and the crossing of the Derwent was for a while hindered by the valour of this single Northman. And the most hard-fought and decisive part of the battle is clearly placed on the left bank, after the defender of the bridge was slain. Up to that time the Northmen, in Henry’s words, “gave way, but did not fly.” It was after the crossing of the river that the great slaughter took place; but this was not a mere slaughter of flying men, but the most stoutly contested part of the battle. For it was then that Tostig and Harold Hardrada were slain, and they assuredly were not slain flying. There was fighting then on both sides of the river, but the great and decisive struggle took place on the left side, the further side from York. We must here take into account the statements of so many writers, from the Chroniclers onwards, that the English came on the Northmen “unawares,” that the Northmen were without their breast-plates, and the like (see p. 369). This has to be reconciled with the fact of the hard fight which was kept on for so many hours. When we once understand the topography, the explanation is easy. The events naturally arrange themselves as I have given them in the text. The Norwegian army is spread abroad on both sides of the river, Harold Hardrada and Tostig with their main strength being on the further side. The English 723come “unawares” on the unprepared division on the right side, and drive them back. The single hero defends the bridge, and gives time for the main army beyond it to form. Then comes the great fight in which the two leaders are slain and the whole Norwegian army is at last cut to pieces.

The only point on which I have any doubt is as to the hours of the day. Henry of Huntingdon makes the battle begin in the early morning, “summo mane;” the first fight lasts till noon; the defender of the bridge withstands the passage of the English till three in the afternoon. Then comes the second fight, and, after all this day’s work, Harold of England is back at York the same evening, and the same evening hears at dinner—or rather supper—of the landing of William (see p. 376). The reckoning of time is in itself suspicious, and it is clearly not without a reference to the sacred hours of the Church. The first and last reckonings can be shown to be wrong. It is possible that Harold may have returned to York on the night of the battle. But he certainly did not hear that evening of William’s landing, because William had not yet landed (see p. 400). Neither could the battle have begun very early in the morning. The English army got no further than Tadcaster on the Sunday evening (“com Harold ... on ðone Sunnandæg to Táða” Chron. Ab.). On Monday morning they marched from Tadcaster, through York, to Stamfordbridge (“fór þa on Monandæg þurh ut Eoferwic”). A march of about seventeen miles, with doubtless some, though not a very long, halt in the city, could not be accomplished so as to make the battle begin very early on a September morning. Then it is quite inconceivable that the resistance on the right side lasted from early morning till noon, and the defence of the bridge from noon till three o’clock. A three hours’ defence of the bridge by one man is impossible. The affair was probably an affair of minutes, though at such a moment minutes would seem like hours. Most likely the hours have got into the wrong places. From noon till three would be a very likely amount of time for the whole of the actual fighting. The march and the pursuit have to be added at each end to make up the whole day’s work.

But this mistake as to the mere reckoning of hours need not throw any doubt as to the main facts of the one intelligible and consistent account of the battle. Henry’s account also exactly 724agrees with the two accounts in the Abingdon Chronicle. These two accounts must not be read as a consecutive narrative. They come from two different hands (see p. 369), and they contain an apparent, though not a real, contradiction. The first tells the story in general terms;

“Ða com Harold Engla cyning heom ongean on unwaran begeondan þære brycge, and hi þær togædere fengon, and swyðe heardlice lange on ðæg feohtende wæron. And þær wæs Harold cyning of Norwegan and Tostig eorl ofslagen, and ungerim folces mid heom, ægðer ge Normana ge Englisca, and þa Normen flugon þa Englisca.”

This is a complete story by itself; then comes the supplement;

“Ða wæs þer an of Norwegan þe widstod þet Englisce folc, þet hí ne micte þa brigge oferstigan ne sige gerechen. Þa seite an Englisce mid anre flan, ac hít nactes ne widstod. And þa com an oþer under þere brigge, end hine þurðstang en únder þere brunie; þa com Harold Engla chinge ofer þere brigge, and hys furde forð mid hine, and þere michel wel geslogon, ge Norweis ge Flæming, and þes cyninges sunu Hetmundus let Harold faran ham to Norwéie mid alle þá scípe.”