Chapter XLII: Appendix (11)

We shall as we go on come across many passages in which the housecarls both of the King and of the great Earls are spoken of. Among the charters of Eadward are several (Cod. Dipl. iv. 201, 204, 221, cf. 200) containing grants to the King’s housecarls. The three grantees spoken of are called Thurstan, Urk, and Wulfnoth—the last at all events being an Englishman, perhaps a kinsman of Godwine. The two latter writs are addressed to Earl Harold. In the oldest (201), a Middlesex writ, addressed to Bishop Robert, Osgod Clapa, and Ulf the Sheriff, Thurstan is described in the English copy as “Ðurstan min huskarll,” in the Latin as “præfectus meus palatinus Ðurstanus.” As Mr. Kemble says (ii. 123), such a description could not apply to every man in so large a body; so we may infer that Thurstan stood high in the service. There is also the will of Wulfwig, Bishop of Dorchester (Cod. Dipl. iv. 290), which is witnessed by a crowd of people, great and small, from the King and the Lady downwards, including some signatures of large bodies of men; “On eallra ðæs kynges húscarlan and on his mæsse-préostan ... and on eallra ðára burhwara gewitnesse on Lincolne and on eallra ðæra manna ðe seceað gearmorkett tó Stowe.” This immediately follows on the signatures of the Stallers Ansgar, Ralph, and Lyfing, from which Mr. Kemble (ii. 122) infers that the Stallers were the special commanders of the force. Housecarls are also mentioned several times in Domesday (see Ellis, i. 91; ii. 151; Kelham, 238), and in Simeon of Durham (Gest. Regg. 1071; see vol. iv. pp. 304, 513) we find a housecarl not only reckoned among the “principales viri” of Northumberland, but high in personal favour with William; “Eilaf huscarl apud regem præpollens honore.”

NOTE LLL. p. 448.
CNUT’S RELATIONS WITH SCOTLAND.

I. The authorities for the Battle of Carham are the Melrose Chronicle (in anno), and two entries of Simeon of Durham, one in his general History (Gest. Regg. in anno), the other in his History of the Church of Durham (iii. 5; ap. X Scriptt. 30). The Melrose writer (p. 155) simply says, “Ingens bellum inter Anglos et Scottos apud Carham geritur.” This entry seems an abridgement of that in Simeon’s Annals; “Ingens bellum apud Carrum gestum est inter Scottos et Anglos, inter Huctredum filium Waldef comitem Northymbrorum, et Malcolmum filium Cyneth regem Scottorum, cum quo fuit in bello Eugenius Calvus rex Lutinensium.” From neither of these accounts should we learn which side was victorious. But in the Durham History Simeon becomes explicit, if not exaggerated; “Universus a flumine Tesa usque Twedam populus, dum contra infinitam Scottorum multitudinem apud Carrum dimicaret, pene totus cum natu majoribus suis interiit.” In the Durham Annals (Pertz, xviii. 507) there is a further notice; “Fuit apud Carrum illud famosum bellum inter Northanhymbros et Scottos, ubi pene totus sancti Cuthberti populos interiit, inter quos etiam xviii sacerdotes, qui inconsulte se intermiscuerant bello; quo audito præscriptus episcopus dolorem et vitam morte finivit.” It is not clear whether this is the event referred to by Fordun (iv. 40), where he tells us that Duncan was sent by Malcolm to meet the Danes and Northumbrians (“qui tunc velut una gens coierant”), who were on their march to ravage Cumberland. He met them and defeated them with great slaughter. Fordun seems to place this before the death of Æthelred; in so confused a writer the chronological difficulty is of no great consequence; it is of more importance that a Northumbrian army, marching to invade any part of Cumberland, would hardly pass by Carham.

There are several points to be noticed here. First, the event of 1018, like the event of 1066, was ushered in by a comet which, though it is not mentioned by our national Chroniclers, seems to have deeply affected local imaginations. “Northanimbrorum populis,” says Simeon in his local work, “per xxx. noctes cometa apparuit, quæ terribili præsagio futuram provinciæ cladem præmonstravit. Siquidem paullo post, id est post triginta dies,” &c. Then follows the account of the battle.

Secondly, Simeon, accurate as he commonly is, has gone wrong—who could feel certain of not going wrong?—among the Earls of his own land. His Uhtred ought (see above, p. 587) to be Eadwulf. It was he, “ignavus valde et timidus,” who now, according to one view (see above, p. 585), ceded Lothian to the victorious Malcolm.

Thirdly, for “Lutinensium” in Simeon we should, according to Mr. Robertson (i. 99), read “Clutinensium.”

Fourthly, the extent of the district from which the English army came should be noticed. It came from the land between Tees and Tweed, that is from old Bernicia, without Lothian. This suggests the question why Lothian, if it was not ceded for the first time till after the battle, did not take a part in the war as well as the rest of the earldom.

Fifthly, the “natu majores” of Simeon are doubtless the “yldestan” of our English Chronicles. See p. 591, and below, p. 777. On this slaughter of the nobility, compare the same result at Assandun, p. 392.

II. With regard to Cnut’s later relations with Scotland, our own Chronicles contain no entries on Scottish affairs earlier than the great submission of 1031. So far as the sagas can be relied upon, they certainly represent Cnut as exercising lordship in Scotland at an earlier time. In Snorro’s Saga of Olaf Haraldsson (Laing, ii. 195) we read how Cnut “reigned over Denmark and England and had conquered for himself a great part of Scotland.” And again we read (Laing, ii. 196; Johnstone, 148), both in prose and verse, how two Kings came to Cnut from Scotland out of Fife, and how he received them to favour, restored their lands, and gave them fresh gifts (“til hans komo tveir konungar nordan af Skotlandi, af Fifi, oc gaf hann þeim upp reídi sína, oc lönd þau öll, er þeir höfdo ádr átt, oc þar med stórar vingiafir”). This is placed while Cnut is still only intriguing, and not yet fighting, against Olaf, that is, at some time before the battle of the Helga in 1025. This story may be merely a transfer to a wrong date of the submission of 1031, or it may be a record of some earlier submission. If the sagas are extremely confused in their chronology, our Chronicles are during this reign extremely meagre in their entries. The Knytlinga Saga also (c. 17; Johnstone, 144) not only makes Cnut subdue a large part of Scotland, but sets his son Harold over it as Under-king, as Swegen was in Norway and Harthacnut in Denmark (see below, p. 775). This seems to be put before the Roman pilgrimage, but the chronology is very confused. The Roman pilgrimage seems to be put after the conquest of Norway. And of a reign of Harold in Scotland nothing, as far as I know, is mentioned elsewhere.

There is also the account in Fordun (iv. 41) of Cnut’s relations with Cumberland, to which I have referred in the text (see p. 449). This story may be true in itself; but the prominence which is given to it certainly looks like an attempt to evade the fact of the submission of Scotland itself. Fordun places the Cumbrian expedition after the Roman pilgrimage, and that he places (iv. 40) in the eighth year of Conrad, meaning seemingly 1032. The refusal of Duncan to do homage is thus described; “Non enim hactenus Anglorum regi Cnutoni, quia regnum invaserat, pro Cumbria Duncanus, quamquam iterum et iterum ab eo submonitus, homagium fecerat, quia non inde sibi de jure, sed regibus Angligenis fidem deberi rex rescripsit.” Cnut then marches against him; that it was with the intention of incorporating Cumberland with the English kingdom, of dealing with the dominions of the recusant as being, in feudal language, a forfeited fief, I infer from the words “Cumbriam suo subdendam dominio pedetentim advenit.” The terms on which peace was finally made are thus described; “Ut regis [Scottorum sc.] nepos Duncanus Cumbriæ dominio libere, sicut predecessorum aliquis liberius tenuit, de cetero gaudeat in futurum, dum tamen ipse futurorumque regum hæredes qui pro tempore fuerint regi Cnutoni ceterisque suis successoribus Anglorum regibus fidem consuetam faciant.” There is nothing unlikely in all this, except perhaps in the extreme loyalty towards the house of Cerdic which is attributed to the Cumbrian Under-king; but we must always remember the strong tendency of Scottish writers to make too much rather than too little of the vassalage of Cumberland to England.

III. We now come to the undoubted submission of Scotland to Cnut in 1031, as recorded in our own Chronicles. I do not understand Mr. Burton (i. 368), when, after quoting Mr. Thorpe’s translation (ii. 128), which is certainly made up confusedly from the Worcester and Peterborough Chronicles, he says that “in only one of the four accepted versions of the original is there anything resembling this.” The Abingdon Chronicle is certainly silent, but Worcester and Peterborough both record the submission, though in different words, and Canterbury follows Peterborough. The Worcester entry runs thus; “Þa fór he [Cnut] to Scotlande, and Scotta cyng eode him on hand and wearð his mann; ac he þæt lytle hwile heold.” Peterborough says, “he [Cnut] for to Scotlande, and Scotta cyng him to beah Mælcolm, and twegen oðre cyningas, Mælbæþe and Iehmarc.” Mr. Robertson (i. 97, ii. 400) seems unable to identify Jehmarc. Mr. Skene (Celtic Scotland, i. 397) makes him the same as Imergi, from whom Somerled, who was killed in 1166, was fourth in descent, and places him in Airergaidhel or Argyll, the old Scottish land of Dalriada. His companion, “Mælbæþa,” “Mealbæaðe,” must be the same as the “Macbeoðe” of the Worcester Chronicle in 1054, that is the Machabæus of Fordun, the Macbeth of Shakespere. The words of the Worcester Chronicler, “ac he þæt lytle hwile heold,” may refer to Malcolm’s death soon after in 1033. Scotland soon fell into confusion, and before long England also.

The submission recorded by our two Chroniclers is not to be doubted; but I confess that I am not quite clear about the date. Both Chroniclers pointedly connect the Scottish expedition with Cnut’s return from Rome (“sona swa he ham com þa fór he to Scotlande,” Wig. “Þy ilcan geare he for to Scotlande,” Petrib.); so it is possible that the true date may be 1027 or 1028 instead of 1031.

Lastly, there is the curious account of Rudulf Glaber (ii. 2) which I have referred to in the text (see p. 450), and which comes in a passage which I shall have to refer to again. In the year 996 a whale as big as an island came out of the North towards Gaul and portended the troubles which were to come upon Gaul and Britain. In Britain especially there was frightful confusion; various Kings were striving and wasting the land, till in the end one got the better of them all; “Viso ... Oceani portento exorsus est bellicus tumultus in universali occidentali orbis plaga, videlicet tam in regionibus Galliarum quam in transmarinis Oceani insulis, Anglorum videlicet atque Brittonum necnon et Scotorum. Siquidem, ut plerumque solet contingere, propter delicta infimi populi, versi in dissensionem illorum reges ac cæteri principes, statimque exardescentes in subjectæ plebis depopulationem scilicet usque dum perducuntur ad suimet sanguinis effusionem. Quod videlicet tamdiu patratum est in prædictis insulis, quousque unus regum earumdem vi solus potiretur regiminis ceterarum.” This lucky King of course is Cnut, who is conceived to be King of the West-Saxons. He seizes the kingdom of Æthelred, who is conceived, it would seem, to be King of one of the British islands called Denmark. “Denique mortuo rege Adalrado, in regno scilicet illorum qui Danimarches cognominantur, qui etiam duxerat uxorem sororem Ricardi Rotomagorum ducis, invasit regnum illius rex videlicet Canuc Occidentalium Anglorum, qui etiam post crebra bellorum molimina ac patriæ depopulationes, pactum cum Ricardo stabiliens ejusque germanam, Adalradi videlicet uxorem, in matrimonium ducens, utriusque regni tenuit monarchiam.” It might be refining too much to hint that this wonderful turning about of the dominions of Cnut and Æthelred had anything to do with the strangely reversed state of geographical parties in 1015–1016 (see p. 377). Then follows the account of the Scottish expedition, as follows;

“Post hæc quoque idem Canuc cum plurimo exercitu egressus, ut subjugaret sibi gentem Scotorum, quorum videlicet rex Melculo vocabatur, viribus et armis validus et, quod potissimum erat, fide atque opere Christianissimus. Ut autem cognovit quoniam Canuc audacter illius quæreret invadere regnum, congregans omnem sui gentis exercitum, potenter ei ne valeret restitit. Ac diu multumque talibus procaciter Canuc inserviens jurgiis, ad postremum tantum prædicti Ricardi Rotomagorum ducis ejusque sororis persuasionibus, pro Dei amore, omni prorsus deposita feritate, mitis effectus, in pace deguit. Insuper etiam et Scotorum regem amicitiæ gratia diligens, illiusque filium de sacro baptismatis fonte excepit.” One does not quite see why either Emma or Duke Richard or Rudolf Glaber should be seized with such a sudden fit of interest in the affairs of Scotland. Still Rudolf’s account is less wonderful than that of a contemporary German writer, Ekkehard the historian of Saint Gallen, who boldly carries Cnut back into the tenth century, and sends Otto the Great over into England to fight against him (Pertz, ii. 119); “Ottone apud Anglos cum _Adaltage_, rege ipsorum, socero suo, aliquamdiu agente, ut junctis viribus Chnutonem Danorum debellaret regem.” Yet Ekkehard was born in 980 and died in 1036.

NOTE MMM. p. 454.
THE BATTLE AT THE HELGA.

This battle is not mentioned in the Abingdon and Worcester Chronicles. Peterborough, followed by Canterbury, places it in 1025. No enemies but Swedes are spoken of, and their commanders are called Ulf and Eglaf. “Þær comon ongean Ulf and Eglaf and swiðe mycel here, ægðer ge landhere ge sciphere of Swaðeode.” Many of Cnut’s men, both Danish and English, are killed, “and þa Sweon hæfdon weallstowe geweald.” As for the place of the battle, Cnut is said to go “to Denmearcon mid scipon to þam Holme æt ea þære halgan.” See Earle, Parallel Chronicles, p. 342; only I do not understand how the “Helge-Aa” could be “then the boundary between Sweden and the Danish possessions,” as the old frontier of Sweden and Scania lies some way to the north of that river.

Ulf and Eglaf are doubtless the sons of Rognvald and Ingebiorg, of whom Snorro speaks in the Saga of Saint Olaf, c. 95 (Laing, ii. 119). At any rate the Ulf spoken of cannot be Ulf the son of Thorgils and brother of Gytha (see above, Note FFF), nor can Eglaf be the Eglaf whom we have already heard of (see p. 447). But both Snorro in c. 159 (Laing, ii. 246) and Saxo (194, 195) agree in making no mention of the sons of Rognvald, and in making Cnut fight the battle against the two Kings Olaf of Norway and Omund of Sweden. They also agree in bringing in Ulf the son of Thorgils; only they bring him in in quite different characters. Saxo makes him a traitor who has invited the combined Swedish and Norwegian invasion, while Snorro makes him redeem former misdeeds by saving Cnut when in great danger. In the Annales Islandorum (Langebek, iii. 40) the date is given as 1027 and the death of Ulf Thorgilsson is placed in the same year.

We can hardly be wrong in accepting the presence of Omund and Olaf on the combined witness of all the Scandinavian writers. But the two Ulfs and Eglaf are puzzling. It has sometimes struck me that “Ulf and Eglaf” in our Chroniclers may be a mistake for “Ulf and Olaf,” taking of course Saxo’s view of the conduct of Ulf Thorgilsson. The Peterborough writer might very easily get wrong in his Ulfs, but he was hardly likely to mistake Saint Olaf, whose history he knew very well, for a man of such small renown as Eglaf Rognvaldsson.

It must not be forgotten that it is to this battle that William of Malmesbury and other writers have, with an utter misconception of the result of the battle, transferred Godwine’s exploit of 1019. See above, p. 743. Henry of Huntingdon translates the Peterborough Chronicle. Florence, following Abingdon and Worcester, is silent.

NOTE NNN. p. 455.
CNUT’S RELATIONS WITH THE EMPIRE.

Cnut, according to Saxo (196), was lord of six kingdoms; “sex præpollentium regnorum possessor effectus.” But he does not give their names. His commentator Stephanius (p. 212) says, “nempe Daniæ, Sveciæ, Norvegiæ, Angliæ, Sclaviæ, Sembiæ” [Semba or Samland in Eastern Prussia?]. The Encomiast (ii. 19) says, “Quum rex Cnuto solum imprimis Danorum obtineret regimen, quinque regnorum, scilicet Danomarchiæ, Angliæ, Britanniæ, Scotiæ, Nordwegæ, vendicato dominio, _Imperator_ exstitit.” Swegen Aggesson (c. 5; Lang. i. 54) outdoes them all. Cnut’s empire extends over the adjoining realms (“circumjacentia regna suo aggregavit _Imperio_”) from Thule to the Byzantine frontier (“ab ultima Thyle usque ad Græcorum ferme Imperium”), taking in, seemingly, not five or six, but ten kingdoms; “quippe Hyberniam, Angliam, Galliam, Italiam, Longobardiam, Teotoniam, Norwagiam, Sclaviam, cum Sambia sibi subjugavit.” Swegen clearly believes in three Empires, Greek, German, and Scandinavian. His exaggerations may be compared with the exaggerations of Dudo and Rudolf Glaber with regard to the Norman Dukes. On the other hand, Prior Godfrey (Satirical Poets, ii. 148) allows Cnut only three kingdoms, the three most obvious,

“Sic insigne caput trino diademate cingit,
Dum Danos, Anglos, Northigenasque regit.”

The Danish writers thus paint Cnut as at least the equal of Conrad; but I am not quite sure that Wipo, in a passage already quoted (see p. 752), where he describes Conrad at his Imperial coronation as walking between the two Kings Cnut and Rudolf, has not a lurking wish to imply that Cnut stood in much the same relation to Conrad that Rudolf did. And the circumstances of the visit, the sight of Pope and Cæsar in all their glory in the old home of both, would be very likely to impress the mind of the still newly-converted lord of Northern Europe, and to make him feel somewhat less Imperial than he felt either at Winchester or at Roskild. But even in Wipo’s account there is nothing to make us think that Cnut did more than yield to Conrad the formal precedence to which he was certainly entitled, and above all at such a moment.

As to the marriage of Gunhild to King Henry there is no kind of doubt; but the plain fact has been clouded over with many fables. That the betrothal took place during the reign of Cnut I infer from the account of Adam of Bremen (ii. 54), who after talking largely of Cnut, Archbishop Unwan of Hamburg, and the Emperor Conrad, goes on to say; “Cum rege Danorum sive Anglorum, mediante archiepiscopo [Unwan], fecit pacem. Cujus etiam filiam Imperator filio suo deposcens uxorem, dedit ei civitatem Sliaswig cum marcha quæ trans Egdoram est, in fœdus amicitiæ; et ex eo tempore fuit regum Daniæ.” But there is no doubt that the marriage was not celebrated till 1036, when Cnut was dead. See Wipo, c. 35, who calls the bride Cunehildis, and the Hildesheim Annals in anno (Pertz, iii. 100), where we read that “regina Cunihild nomine ... in natali Apostolorum regalem coronam accepit et mutato nomine in benedictione Cunigund dicta est.” See also Hepidanni Annales in anno (ap. Duchesne, Rer. Franc. Scriptt. iii. 479). William of Malmesbury (ii. 188) is so far accurate as to place the marriage after Cnut’s death, but he tells the story with great confusion. He grows specially eloquent on the splendour of the bride’s progress, just as Roger of Wendover (iv. 332 et seqq.) does over the marriage of Isabella, daughter of John, with Frederick the Second; but William makes Harthacnut send his sister from England, though Harthacnut certainly was not there in 1036, and he seems to place the marriage after the trial and acquittal of Godwine in 1041. It was probably this confusion which led him to speak of Henry as “Imperator Alemannorum,” for though Henry did not become Emperor till 1046, yet his father died in 1039, leaving to Henry, as Wipo (c. 39) says, “regni rem, Imperii autem spem, bene locatam.” Wace also (Roman de Rou, 6552) tells us;

“Gunnil fu à Rome menée,
Et à Rome fu mariée;
Fame fu à l’Emperéor;
Ne pout aveir plus halt seignor.”

Besides that Henry was not yet Emperor, the marriage was (see the Hildesheim Annals, u. s.) not celebrated at Rome but at Nimwegen. Gunhild died July 18th, 1038, “quasi in limine vitæ,” as Wipo (c. 37) says, before the death of Conrad. There is another inaccurate account of the marriage in Heming’s Worcester Cartulary, 267 (Monasticon, i. 596), where the bridegroom is described as “Imperator Cono” and Brihtheah, Bishop of Worcester, appears as one of the bride’s suite; “Idem vero episcopus Brihtegus quodam in tempore ad Saxoniam Gunnildæ, Cnuti regis filiæ, ductor exstitit, quum eam Imperator Cono uxorem duceret, et quemdam ministrum sibi valde carum, Hearlewinum nomine, socium itineris secus habuit.” But the mistakes of all these writers seem pardonable when we turn to the wonderful romance which some of the Scandinavian writers have devised by rolling together the Roman pilgrimage of Cnut, the marriage of Gunhild, and seemingly also the Italian expedition of Conrad and Henry, which happened (see Wipo, c. 35) soon after Henry’s marriage. Saxo (196) is comparatively brief. After the description of Cnut as lord of six kingdoms, he tells us how he married Gunhild to Henry and then went and restored the authority of his son-in-law over certain rebels in Italy. “Canutus ... eximio sui fulgore etiam Romanum illustravit Imperium. Enimvero ejus principi Henrico filiam Gunnildam nuptum tradidit, eumdemque paullo post Italica consternatione perculsum auxilio prosequutus, pristinæ fortunæ, pressa rebellium conspiratione, restituit.” Swegen Aggesson (c. 5; Langebek, i. 54) is much fuller. Henry, already Emperor, marries Gunhild; he is driven from Rome by a sedition, and comes to crave help of his father-in-law (“quem quum Romani tumultuaria seditione a regio pepulissent solio, socerum adiens ejus auxilium imploravit”). Cnut, seemingly glad of the chance (“nactus occasionem illustris ille præcluisque Kanutus”), sets out to avenge his wrongs. On the road, seemingly by way of pastime, he ravages Gaul (“assumpto exercitu suo, primo Galliam depopulando invasit”); he then harries Lombardy and Italy, which, it will be remembered, Swegen had already reckoned as separate kingdoms, and compels the Romans to receive their Emperor back again (“multimoda virtute compulit Romanos civitatem sibi resignare, tandemque Imperatorem et generum throno suo restituit”). He then goes to France, which is seemingly looked on as something different from Gaul; yet most certainly Latin and not Teutonic _Francia_ is intended, for Cnut goes to Tours (“cum ingenti tripudio ad Franciam usque commeavit, Turonisque profectus,” &c.) and carries off thence (“potenter secum asportavit”) the relics of Saint Martin, which he translates to Rouen, on account of his great love for that city; “eo quod illam [Rothomagum] præ ceteris specialem diligeret.” This wild talk about Rouen must be compared with another equally wild tale which I shall have to mention presently about Cnut dying before Rouen.

It is no wonder that Swegen’s editor says, “Mirum est Suenonem et in hoc et in plurimis historiæ Canuti M. momentis adeo hallucinatum esse.” Swegen wrote about 1186, in the days of Frederick Barbarossa and Henry the Sixth, and it is worth noting how thoroughly both he and his elder contemporary Wace look on the Roman Emperor as the local sovereign of Rome, in opposition to William of Malmesbury’s slipshod talk about “Imperator Alemannorum.”

About Gunhild, William of Malmesbury has a legend which is the same as that of Sir Aldingar and Queen Eleanor in Percy’s Reliques. The King’s name in both tales is Henry. Gunhild left a daughter, Beatrice, Abbess of Quedlingburg (see Struvius, i. 355). The only English princess, Matilda daughter of Henry the Second, who was the mother of an Emperor, was not the wife of an Emperor or even of a King.

On the cession of Sleswick, Adam, as quoted in the text, seems quite explicit. On the Eider as the boundary of the Carolingian Empire, see the Annals of Einhard, 808, 811, 815, 828, and the Annals of Fulda (Pertz, i. 355 et seqq.), 811, 857, 873. Nothing can be plainer than the last passage, “fluvium nomine Egidoram, qui illos [Danos] et Saxones dirimit.” In saying that it remained the boundary till 1866, I should perhaps except the time of confusion, 1806–1814, when the Roman Empire had been dissolved and when the German Confederation had not yet been founded. During these years Holstein, the “Transalbiana Saxonia” of Einhard, was united to the kingdom of Denmark by an act as regular as any act of that irregular time.

NOTE OOO. p. 469.
ÆLFRED THE GIANT.

Ælfred is a name so purely English that the presumption in favour of the English birth of any one bearing it in this generation is extremely strong. There is no doubt that Ælfred is the name intended. The giant is “Alvredus cognomento gigas” in William of Jumièges, and “Alvredus” is the name by which he calls the English Ætheling Ælfred. In the Roman de Rou he is “Auvere,” “Alverei,” “Alvere;” the Ætheling is “Auvered” and “Alvred.” So in Mary of France (see Roquefort, ii. 34 and vol. iii. p. 572, iv. p. 796, v. p. 594) Ælfred appears as “Auvert,” “Auvres,” “Alurez,” “Affrus.” The only chance against Ælfred being an Englishman is the chance—a somewhat faint one, I think—that the name may also have been in use among the Saxons of Bayeux. M. Pluquet (Roman de Rou, ii. 17) says that the name is still common in the district, seemingly under the form of “Auvray.” But “Auvray” may be “Alberic;” and we shall find that Ælfred and Eadward were just the two English names which we shall find that a later generation of Normans did adopt.

I have a note, but I cannot lay my hand on the reference, of a charter of Hugh Capet in 967 signed by “Alfredus monachus;” and “Alfridus abbas Sancti Vulmari” signs in 1026 (Chron. Sithiense, p. 175) a charter of Baldwin, Bishop of Terouanne. These two can hardly be the same man, but both may be Englishmen. It is more singular to find the name in Italy. Yet we read in Donizo’s Life of the Countess Matilda (Murat. v. 372),

“Ac Mons Alfredi capitur certamine freni.”

Was the mount called from any English pilgrim, the great King himself perhaps, or did any cognate name exist among Goths or Lombards? The elfish names are mainly English; yet Elberich is said to be the same as Ælfric, and Alboin as Ælfwine. See Miss Yonge’s Christian Names, ii. 346, 347.

Our Ælfred signs two charters with the title of “vicecomes,” one in 1025 and one in 1027. He afterwards became a monk at Cerisy. Roman de Rou, 8717 et seqq. He seems (see Neustria Pia, 660) to have left a son William and a daughter who bears the odd-sounding name of _Athselinoc_. Can this be a corruption of any English name beginning with _æðel_?

NOTE PPP. p. 473.
CNUT’S RELATIONS WITH NORMANDY.

The Norman and English writers do not mention the marriage of Robert and Estrith. It is asserted by Saxo, Adam of Bremen, and Rudolf Glaber. But the two former tell the story with much confusion, making Estrith marry, not Robert, but Richard. They both connect this marriage with Cnut’s own marriage with Emma. Saxo’s words (p. 193) are; “Quum Anglorum rebus obtentis nectendam cum finitimis amicitiam decrevisset, Normanniæ præfecti [an odd title] _Roberti_ filiam Immam matrimonio duxit, _ejusque fratri Rikardo_ sororem Estritham conjugio potiendam permisit.” The utter confusion of Saxo’s ideas about the Norman Dukes is manifest. Adam (ii. 52) says; “Chnud regnum Adalradi accepit uxoremque ejus Immam nomine, quæ fuit soror comitis Nortmannorum Rikardi. Cui rex Danorum suam dedit germanam Margaretam pro fœdere. Quam deinde Chnut, repudiatam a comite, Wolf duci Angliæ dedit.... Et Rikardus quidem comes, declinans iram Chnut, Jherosolimam profectus, ibidem obiit, relinquens filium in Nortmannia nomine Rodbertum, cujus filius est iste Willelmus quem Franci Bastardum vocant.” Here we get a little light. The marriages of Richard the Good with Judith and Papia are well ascertained, and there is no room left for a marriage with Estrith. But, as Lappenberg remarks (479. Eng. Tr. ii. 217), Adam’s mention of the pilgrimage to Jerusalem shows that Robert is the person really meant among all this confusion. Lastly, Rudolf Glaber, a better authority on such a point than Saxo or even than Adam, steps in to settle the matter. He describes (iv. 6. p. 47) Robert’s pilgrimage to Jerusalem and his death without lawful issue, “quamlibet sororem Anglorum regis Canuc manifestum est duxisse uxorem, quam odiendo divortium fecerat.” This seems to put the fact of a marriage between Robert and Estrith on firm ground. Among the Danish annalists, the Esrom Annals (Lang. i. 236) simply copy Adam of Bremen; those of Roskild (Lang. i. 377) tell the same tale in different words; “Kanutus victor exsistens, ipsam Ymmam duxit uxorem, genuitque ex ea filium Hartheknud. Kanutus Ricardo suam dedit sororem nomine Estrid. Quæ ab illo repudiata duci Ulf _sine fratris consensu_ [cf. Saxo’s tale quoted in p. 746] est conjuncta.” The name Margaret given by Adam to the Danish princess is remarkable. Estrith might possibly, like Emma and Eadgyth the daughter of Malcolm, have been required to take a Norman name on her marriage. But the name of Margaret, which became popular only through Eadgyth’s mother, is rare throughout the century, and this would perhaps be the first instance of it in the West.

As for the date of the marriage, see Lappenberg, ii. 217, and Pertz’s note to Adam, ii. 52. A dispute between Robert and Cnut which could be connected, even mythically, with Cnut’s death and Robert’s pilgrimage must be placed quite late in their reigns. And as the offender is always looked on as the reigning Duke, 1028, or (if we take the reckoning of Florence under 1026 and the Peterborough Chronicle under 1024) 1026, is the earliest year to which the transaction can be referred. Ulf was killed in 1025. William the Bastard was born in 1027 or 1028. As for Estrith’s dowry, Saxo tells us that Cnut, before her marriage with Ulf, “sororem Sialandiæ redditam regiarum partium functione donavit” (p. 194). After Ulf’s death, execution, or murder, “Canutus violatæ necessitudinis injuriam, ac sororis viduitatem, duarum provinciarum attributione pensavit” (p. 197). He adds that she gave them to the Church of Roskild. The Roskild Annals (Lang. i. 377) makes her rebuild the church with stone, it having before been of wood; “Honorifice sepelivit, ecclesiamque lapideam in loco ligneæ construxit, quam multis modis ditavit.”

I need hardly say that Cnut’s expedition to Normandy is quite mythical. We have already seen (see above, p. 768) a legendary account of a campaign of Cnut in Gaul, including a visit to Rouen, which seems to have grown out of his Roman pilgrimage. The present legend seems further to mingle up with this the pilgrimage of Robert to Jerusalem and the beginning of the Norman exploits in Sicily and Apulia. Saxo, so far as anything can be made out of his chronology, seems to make two Norman expeditions on the part of Cnut. The first (p. 194) seems to be early in his reign; “Rikardum, acerrimum uxoris osorem effectum, patria exegit.” Afterwards (pp. 200, 201) we have the story of his great expedition and death before Rouen. Richard is still Duke, but, for fear of Cnut, he flees to Sicily; “Cujus [Canuti] impetum Richardus Siciliam petens, fuga præcurrere maturavit.” The mention of Sicily is of course suggested by the exploits of the Normans in those regions. Adam, as we have seen, makes Richard flee to Jerusalem. His scholiast adds that the conquest of Apulia was begun by forty of his comrades on their return. The source of confusion is obvious.

This wild story of Cnut’s death before Rouen seems peculiar to Saxo. Several of the other Danish writers distinctly assert his death in England. Chron. Esrom. ap. Langebek, i. 236 (which makes him die in 1037); Chron. Rosk. i. 377. The attempt of Robert against England is described by William of Jumièges (vi. 10, 11) and Wace (Roman de Rou, 7897 et seqq.). I have followed their account in the text. Only two English writers mention it, William of Malmesbury (ii. 180) and John of Wallingford (Gale, 549–550). William mentions only the intended invasion, and says nothing of the embassies before and after it. John of Wallingford tells the story much as William of Jumièges does, only, with the usual confusion, he talks of Richard instead of Robert. But it is plain from the two Williams that Robert was the Duke concerned, so that John of Wallingford is clearly wrong when he places the story in the first years of Cnut—“in primordiis regni sui.”

William of Jumièges (vi. 10) thus describes the message sent by Robert to Cnut; “Mandavit Chunuto regi ut jamjamque satiatus eorum [the Æthelings] exterminio illis parceret, et _sua eis vel sero_ pro sui amoris obtentu redderet.” So John of Wallingford; “Venerunt legati a Normannia ... qui cum Cnutone de regni jure disceptantes juvenibus prædictis regnum postulabant.”

William of Jumièges, it should be mentioned, distinctly implies the personal presence of Robert on board the fleet, but says nothing of that of the Ætheling. Wace (7941) speaks of both Robert and Eadward.

NOTE QQQ. p. 480.
THE DIVISION OF CNUT’S DOMINIONS.

That Cnut, like Charles, established a system of under-kingdoms, to be held by his sons in subordination to his own Imperial authority, is distinctly asserted by Saxo (196). “Inde [from Rome, see p. 768] reversus, Haraldum natu majorem Angliæ, Daniæ Canutum, Norvagiæ Suenonem, quem ex Alvina sustulerat, absque ulla majestatis suæ diminutione præfecit. Nam etsi tres provincias totidem filiorum regimini tradidit, nihilominus commune sibi trium _imperium_ reservavit, neque summam penes alium consistere voluit. Præterea teneris adhuc ducibus in officiorum tutelam fortissimorum præsidia sociavit.” The Knytlinga Saga (c. 17; Johnstone, 144) gives a similar account, only instead of England, it makes Harold Under-king over part of Scotland (see above, p. 761); “Knutr konungr hafdi oc til forrada mikinn hlut af Skotlandi, oc setti hann þar Haralld son sinn konung ysir: enn þo var Knutr konungr ysir-konungr [overkonge] allra þeirra.” Now this statement that Cnut established his sons as Under-kings under the guardianship of some of his chief men falls in exactly with the statement in our own Chronicles that Thurkill was established in Denmark as guardian to one of Cnut’s sons (see p. 429). The words of the passage (1023) are, “And he betæhte Þurcille Denemearcan and his sunu to healdenne;” but the details of this arrangement, as described both in Saxo and in the saga, seem open to much doubt. There is not a shadow of evidence that Harold ever reigned as Under-king in England, and the statement that he reigned in Scotland, though very remarkable, is hardly to be accepted without better authority than that of the Knytlinga Saga. The further question arises, who was the son whom Cnut left in Denmark? Not Harthacnut, who succeeded him there, for that kingly bairn was with his mother in England (see Chron. Wig. in anno). It must have been one of the two doubtful sons, Swegen and Harold, whom it may have been convenient to remove from England, together with their mother, “the other Ælfgifu.” She and Swegen, it is well known, were afterwards quartered in Norway, and this looks as if Harold were now, in the like sort, quartered in Denmark. This would prove a change of purpose on Cnut’s part as to the succession of his children, as it was Harthacnut who actually succeeded him in Denmark.

On Swegen’s reign in Norway under the guardianship of his mother, see Saxo, 196; Snorro, c. 252; Laing, ii. 344. I suspect that Saxo conceived the three sons as having been Under-kings in the several kingdoms to which they actually succeeded; but if it be true, as seems likely, that Harold was first quartered as Under-king in Denmark and afterwards displaced to make way for Harthacnut, the fact becomes of importance with reference to the disputed election which followed his death.

As to the division on Cnut’s death there seems no doubt at all. The account given by Adam (ii. 72) runs thus; “Post cujus mortem, _ut ipse disposuit_, succedunt in regnum filii ejus, Haroldus in Angliam, Svein in Nortmanniam, Hardechnut autem in Daniam.... Suein et Harold a concubina geniti erant; qui, ut mos est barbaris, æquam tunc inter liberos Chnut sortiti sunt partem hæreditatis.” This is copied by the Esrom Chronicle, Lang. i. 237; cf. Chron. Rosk. p. 377; Chron. Erici, p. 159. As Harold actually succeeded in England, foreign writers seem to have taken for granted that his succession was in accordance with Cnut’s will; but it is evident that Cnut latterly intended England for Harthacnut.

On the expulsion of Swegen and Ælfgifu from Norway, see Snorro, Saga of Magnus, c. 4; Laing, ii. 363.

NOTE RRR. p. 482.
THE CANDIDATURE OF HAROLD AND HARTHACNUT.

I have gathered my account of this disputed election wholly from our own Chronicles, which are the only trustworthy guides. The cause of all the difficulties and contradictions with which the subject is involved, is the fact that the division of the kingdom between Harold and Harthacnut proved a mere ephemeral arrangement, and was set aside within two years. It seems therefore to have quite passed out of mind, except with the very few writers with whom minute accuracy was really an object. No one would find out the fact from Adam of Bremen, from the Encomiast, or even from William of Malmesbury. Of the Danish writers it is needless to speak. The Encomiast (iii. 1 et seqq.) sees, so does William of Malmesbury (ii. 188) still more plainly, that a strong opposition was made to the election of Harold; they do not see that that opposition was so far successful that a temporary sovereignty over a part of the kingdom was secured to Harthacnut. Even Florence, seemingly hesitating, as he sometimes does, between two versions of a story, tells the tale with some confusion. But on comparing the Abingdon, Worcester, and Peterborough Chronicles, the matter becomes much clearer. The Peterborough Chronicle is the primary authority for the division of parties in the Witenagemót, for the division of the kingdom between the two competitors, for the regency of Emma and Godwine on behalf of Harthacnut. Its statements are copied, with more or less of confusion and misconception, by the Canterbury Chronicle, Florence, and William of Malmesbury. The Abingdon and Worcester Chronicles do not distinctly mention the division of the kingdom under the year 1035; but they imply it under 1037, in the words, “Hér man geceas Harald ofer eall to cinge, and forsoc Harðacnut,” which, unless Harthacnut had before possessed part of the kingdom, would be meaningless. Oddly enough, the Peterborough Chronicler does not distinctly mention this second election of Harold, though he perhaps alludes to it in the words, “And he [Harold] wæs þæh full cyng ofer eall Englaland.” Thus the two accounts in the Chronicles fill up gaps in each other, and between the two we get a full and consistent narrative.

I believe the controversy to have lain wholly between the two sons of Cnut, Harold and Harthacnut. That there was a party in favour of one of the sons of Æthelred (see p. 476) is asserted by William of Malmesbury (ii. 188); “Angli diu obstiterunt, magis unum ex filiis Ethelredi, qui in Normannia morabantur, vel Hardecnutum filium Cnutonis ex Emma, qui tunc in Danemarchia erat, regem habere volentes.” But in the Chronicles, where the proceedings in the Witenagemót are described, we hear nothing of any voices being raised on the side of the Æthelings, and William himself says (u. s.) of a time a little later; “Filii Ethelredi jam fere omnibus despectui erant, magis propter paternæ socordiæ memoriam, quam propter Danorum potentiam.” These last words are at least a witness to the freedom of election on this occasion.

The geographical division of parties is clearly marked in the Peterborough Chronicle, which is also the only one which notices the share taken by London in the election. We now hear only of the “liðsmen,” not, as in the election of 1016, of the “burhwaru.” The proposal for a division I understand to come from Harold’s supporters, most likely from Leofric, the natural mediator between the two extreme parties. I do not see what else can be the meaning of the expression in the Peterborough Chronicle that Leofric and others chose Harold _and Harthacnut_ (“Leofric eorl and mæst ealle þa þegenas be norðan Temese and þa liðsmen on Lunden gecuron Harold to healdes Englelandes; _him and his broðer Hardacnute_, þe wæs on Denemearcon”). This proposal—namely the division—Godwine and the West-Saxons resist (“and Godwine eorl and ealle þa yldestan menn on West-Seaxon lagon ongean, swa hi lengost mihton; ac hi ne mihton nan þing ongean wealcan”); that is, they claim the whole kingdom for Harthacnut. At last they are obliged to consent to the division and the regency (“and man gerædde þa þæt Ælfgifu Hardacnutes modor sæte on Winceastre mid þæs cynges huscarlum hyra suna, and heoldan ealle West-Seaxan him to handa; and Godwine eorl was heora healdest man”). Henry of Huntingdon (M. H. B. 758 B) translates this account, but he was evidently puzzled by the words about electing Harold _and Harthacnut_, as he says, “elegerunt Haraldum, ut conservaret regnum fratri suo Hardecnut”—a most unlikely story. The last clause he translates; “Consilium ergo inierunt quod Emma regina cum regis defuncti familia [huscarlum?] conservaret Westsexe apud Wincestre in opus filii sui, Godwinus vero consul dux eis esset in re militari.” Henry says nothing of the second election of Harold in 1037. William of Malmesbury (ii. 188), though telling the story in a most confused way, seems quite to take in the position of Godwine; “Maximus tum justitiæ propugnator fuit Godwinus comes, qui etiam pupillorum [his notion about the sons of Æthelred, as well as Harthacnut, here comes in] se tutorem professus, reginam Emmam et regias gazas custodiens, resistentes _umbone nominis sui_ aliquamdiu dispulit; sed tandem, vi et numero impar, cessit violentiæ.” Mr. St. John (Four Conquests, ii. 106, 107) makes Godwine first assert the rights of the Æthelings, which I suppose is his interpretation of the words of William, and then himself propose the compromise in favour of Harthacnut. For this he refers us to Simeon; but Simeon (X Scriptt. 179) only copies the narrative of Florence, and that narrative, as well as that of the “Saxon Chronicle” [Abingdon as opposed to Peterborough?], Mr. St. John had just before cast aside as “confusing the whole subject.”

I see that the idea of the Imperial supremacy being reserved to Harold has also occurred to Mr. St. John (ii. 110). It was suggested to me by the words of the Peterborough Chronicle (evidently misunderstood by the Canterbury Chronicler), “And he [Harold] wæs þæh full cyng ofer eall Englaland.” This however, as I remarked just above, may perhaps refer to Harold’s second election in 1037. The same idea might also lurk in the other words of the Peterborough Chronicler, quoted in the last page, “gecuron Harold to healdes ealles Englelandes; him and his broðer Hardacnute,” &c. But an Imperial supremacy on the part of Harold seems quite consistent with the general tenor of events, and such a supposition may perhaps render the account of the fate of the Ætheling Ælfred one degree less obscure.

The story of Æthelnoth’s refusal to crown Harold comes wholly from the Encomium Emmæ, iii. 1. But it is possible that the tale, if true, may belong to the second election of Harold in 1037, and may have been thrust back in the confused chronology of the Encomiast. A coronation, sooner or later, seems quite certain. It is asserted by Ralph de Diceto, ii. 238, ed. Stubbs; “Haroldus filius Cnutonis regnavit annis iii. consecratus ab Ethelnodo Dorobernensi archiepiscopo apud Lundonias.” So Roger of Wendover (i. 473); “Prævaluit pars Haroldi, et regni Angliæ illum diademate insignivit.” According to Bromton (X Scriptt. 932), Harold was “ab Ethelnodo Dorobernensi archiepiscopo apud Lundonias consecratus.” But the higher authority of the list of coronations in Rishanger (427) places it at Oxford, which seems to have been Harold’s capital. Believers in the false Ingulf may also entertain themselves with a story about Harold’s coronation robe, and a great deal more about which authentic history is silent. See St. John, ii. 107–110.

NOTE SSS. p. 489.
THE DEATH OF THE ÆTHELING ÆLFRED.

I have stated in the text the chief versions as to the death of Ælfred. The different statements may be grouped under two main heads, those which put the event at its right date under the reign of Harold, and those which move it to some other time. It is the former class whose statements we must weigh against other; the latter are useful mainly as illustrating particular points, and as examples of the way in which legends grow.

The earliest English account is that which is found, in different shapes, in the Abingdon and Worcester Chronicles. Peterborough is silent about the whole matter. The story, except a few lines at the beginning, takes the form of a ballad, as it appears in Mr. Earle’s Parallel Chronicles. It is astonishing that Mr. Thorpe should have printed it as plain prose, when it plainly is, not only, like the songs of Brunanburh and Maldon, in rhythm, but actually in rime. This was seen long ago by Dr. Ingram, who not only printed it as verse, but attempted a riming version of his own in modern English. I have in the text analysed the account thus given. The remarkable point is that the Abingdon Chronicle distinctly accuses Godwine, while the Worcester version leaves out his name. In the prose introduction, Ælfred the innocent (“unsceððiga”) Ætheling lands and wishes to go to his mother, who sat at Winchester. Then says Abingdon, “Ac hit him ne geþafode _Godwine eorl ne éc oþre men_ þe mycel mihton wealdan: forþan hit hleoðrode þa, swiðe toward Haraldes, þeh hit unriht wære.” But in Worcester it stands, “Ac þæt ne geþafodon þa þe micel weoldon on þisan lande; forþan hit hleoþrade þa swiðe to Harolde, þeah hit unriht wære.” So the beginning of the ballad stands in Abingdon,

“Ac Godwine hine þa gelette,
And hine on hæft sette;”

while in the Worcester version it runs,

“Ða let he [Harold] hine on hæft settan.”

There can be no doubt that here the Abingdon version is the original, and that the Worcester text, which destroys rhythm and rime, was altered by an admirer of Godwine. But as to the prose introduction the case is far less clear; the words “Godwine eorl ne éc oþre men” might just as well be an interpolation. So in Florence the mention of Godwine comes in very awkwardly; “Quod indigne graviterque ferebant potentes nonnulli, quia, licet injustum esset, Haroldo multo devotiores exstitere quam illis, _maxime, ut fertur, comes Godwinus_.”

Florence’s version is made up by modifying the account in the Chronicles, with some touches from other quarters. He makes both brothers come, changing the words “Ælfred se unsceððiga æþeling” into “innocentes clitones Ælfredus et Eadwardus.” While in the Chronicles Ælfred simply wishes to go to his mother (“wolde to his moder þe on Wincestre sæt”), and is hindered by certain men, Godwine or others, in this account both the Æthelings actually visit their mother (“ad suæ matris colloquium, quæ morabatur Wintoniæ, venere”), and Godwine and the other powerful men are simply displeased at their coming (“indigne graviterque ferebant,” as above). Then comes the strangest part of his statement; that Godwine seized and imprisoned Ælfred is simply translated from the ballad, but Florence now introduces the almost incomprehensible assertion that Ælfred, when he was seized, was going to London for a conference with Harold; “Hic quidem [Godwinus] Ælfredum, quum versus Lundoniam, ad regis Haroldi colloquium, ut mandarat, properaret, retinuit, et arcta in custodia posuit [‘hine þa gelette and hine on hæft sette’].” The companions of Ælfred, to the number of six hundred, are sold, killed, or tortured at Guildford; the place is not mentioned in the Chronicles. Emma then sends back her son Eadward, who had stayed with her (“qui secum remansit”) and had not set out with his brother, with all haste to Normandy. Then, at the bidding of Godwine and certain others (“Godwini et quorumdam aliorum jussione”), Ælfred is taken to Ely, and the rest of the story follows as in the ballad.

It is plain that Florence in writing this had one or both of the Chronicles before him, and tried to work in details from other sources which were really inconsistent with the account which the Chronicles gave. One change is of special importance. The ballad simply mentions the companions (“geferan”) of Ælfred without any account of their number or who they were. Florence makes them six hundred, and adds the very important statement that they were Norman knights or soldiers. The Æthelings come, “multis Normannicis militibus secum assumptis, in Angliam paucis transvecti navibus.” This touch clearly comes from the Norman version, which represents the first attempt of the Æthelings as an actual invasion, an idea which the Chronicles do not suggest. It is also plainly from the same source that he got the idea that Eadward had any share in the business.

The ballad in the two Chronicles has about it something of that vagueness which is natural in a poem which is rather a pious lamentation than a narrative. The Norman account, true or false, is at least fuller and clearer. It first appears in William of Poitiers, the Conqueror’s chaplain, the extant part of whose narrative begins at this point. He is followed by William of Jumièges (vii. 8, 9), who is followed by Wace (Roman de Rou, 9759 et seqq.). I have given the substance of their story in the text, and I do not know that there is anything to remark, except that at the end of his tale William of Poitiers turns round and reviles Godwine in an address in the second person, much as at a later stage of his narrative he reviles Harold.

We now come to the version of the Encomiast. He is a perplexing writer to deal with; one knows not what to make of an historian who was either so easily imposed upon or else so utterly reckless as to truth. A contemporary writer who wipes out Emma’s marriage with Æthelred, who looks on the Æthelings as sons of Cnut, who is ignorant that his heroine was actually Queen-regent over Wessex, is really somewhat of a curiosity. His astounding statement (ii. 18) that Eadward and Ælfred were the sons of Cnut I have already spoken of in Note BBB. In his present account (iii. 1) Emma remained in England after the death of Cnut, grieving for the death of her husband and the absence of her sons (“sollicita pro filiorum absentia”). He then goes on; “Namque unus eorum, Hardecnuto scilicet, quem pater regem Danorum constituit, suo morabatur in regno; duo vero alii in Normanniæ finibus ad nutriendum traditi, cum propinquo suo degebant Rotberto.” These last are the sons of whom one, Ælfred, the younger of the two (“Alfridus minor natu,” iii. 4), is the victim of the present story. It is plain therefore that the “liberales filii” of Cnut, spoken of in the former passage, are meant to be the same as the Æthelings. All three brothers being absent, “factum est,” he goes on to say, “ut quidam Anglorum, pietatem regis sui jam defuncti obliti, mallent regnum suum dedecorare quam ornare, relinquentes nobiles filios insignis reginæ Emmæ, et eligentes sibi in regem quemdam Haroldum,” &c., &c. (see above, p. 775). He then goes on with Æthelnoth’s refusal to crown Harold (see above, p. 487), and with Harold’s ungodly manner of life (see p. 504). Then comes the forged letter and the rest of the story, as I have told it in the text. The piece of detail most worthy of notice is the writer’s remarks (iii. 5) on the _decimation_ of Ælfred’s companions (on the alleged decimation at Canterbury in 1012, see above, p. 674);

“Traditi sunt carnificibus, quibus etiam jussum ut nemini parcerent nisi quem sors decima offerret. Tunc tortores vinctos ordinatim sedere fecerunt, satis supraque eis insultantes, illius interfectoris Thebææ legionis exemplo usi sunt, qui decimavit primum innocentes multo his mitius. Ille enim rex paganissimus Christianorum novem pepercit, occiso decimo: at hi profanissimi falsissimique Christiani bonorum Christianorum novem perimerunt, decimo dimisso.”

Now when a writer, whether through ignorance or through design, goes so utterly wrong about the birth of his hero, about the position of his heroine and the general condition of the kingdom, one hardly knows how to accept anything that he tells us. Yet his account, if used with caution, seems to supply some useful hints. His account is the only one which, while consistent with Godwine’s innocence, explains the origin of the belief as to his guilt. If we accept his account of what happened between Godwine and Ælfred, the various statements become intelligible; we see how the opposite stories could arise, which in any other way it is hard to see. The tale of the forged letter has a very odd sound, and the details may easily be mythical. Yet something of the kind would fill up the gap in the Chronicles, in which Ælfred comes over to England without any particular reason for his coming, better than William of Poitiers’ wild tale of a Norman invasion, which is most likely a mere repetition of the attempt of Robert.