bounds, a promise of which the addition of Oxfordshire may have been the fulfilment. This last view is incidentally confirmed in a singular manner by the way in which the town of Oxford is spoken of in Domesday (154). The duties payable to the Earl are described as paid to Ælfgar. Here of course, as in several other cases, the record describes a state of things existing "in the time of King Eadward," but not "on the day when King Eadward was quick and dead." A mention of Eadwine would have excluded Gyrth; a mention of Ælfgar does not exclude him. But it shows that Oxfordshire was at one time held by Ælfgar; it shows therefore that Gyrth did not receive Oxfordshire at the same time as Norfolk and Suffolk. The shire may have been taken from Elfgar at his second outlawry, or it may have been conferred on Gyrth after Ælfgar's death. But at all events, Gyrth became Earl of the East-Angles in 1057, only with a narrower jurisdiction than had been attached to that title when it was held by Harold, probably narrower than when it was held by Ælfgar. Harold had, together with the two strictly East-Anglian shires, held Huntingdonshire, Cambridgeshire, and Essex, probably including Middlesex. None of these, except perhaps Cambridgeshire, fell to the lot of Gyrth. He seemingly took the remote Oxfordshire in their stead. Of Huntingdonshire I have already spoken. The shires of Essex and Middlesex, together with that of Hertford, and probably Buckinghamshire (see above, p. 377), fell to the lot of Leofwine. Of Bedfordshire I cannot speak with any certainty.

We have no record of Leofwine's appointment as Earl, but one can hardly doubt that his investment with the large and important government which the writs set him before us as holding took place at the general distribution in 1057. But, as in the case of Gyrth, the question arises whether he had held a smaller government at an earlier time. There is a writ in Cod. Dipl. (iv. 191) addressed to Leofwine in Kent conjointly with Archbishop Eadsige, who died in 1050, and with Godwine, Bishop of Rochester, who died in 1046. If this document be genuine, it reveals the very curious fact that the young son of Earl Godwine, while still hardly beyond boyhood, held, under his father's immediate eye, the government of the shire which had been his father's first possession. If this be so, it may decide us as to the interpretation of the doubtful passage of the Biographer about Gyrth, and we shall have to look for some similar earlier endowment for Tostig. (Compare the application of the word "Dux" to him by the Biographer in 1051. See p. 87.) But, on the other hand, the Chroniclers, in recording the events of the years 1049-1052, while they carefully give the title of Earl to Godwine, Swegen, Harold, and Beorn, never give it to Tostig, Gyrth, or Leofwine. "Harold Eorl and Tostig his brocor," says the Peterborough Chronicler (1046). The early promotion of Tostig and Leofwine is therefore very doubtful; but of the extent of Leofwine's later government there is no doubt. It took in the shires of Essex, Middlesex, Hertford, Surrey, Kent, and probably Buckinghamshire. Writs are addressed to him for Surrey, jointly with Stigand (Cod. Dipl. iv. 205), for Essex (as he is coupled with Bishop William, iv. 213), for Middlesex jointly with William (iv. 214), for Hertfordshire, as we have seen, jointly with Wulfwig. "Men" of Earl Leofwine in Middlesex are also mentioned in Domesday, 130 b. But the general superiority of Harold, whether as elder brother or as elected Ætheling, seems shown by a writ addressed to him in Middlesex, jointly with Bishop William (iv. 211). It can hardly belong to the time between September 1052 and Easter 1053, between

THE LEGEND OF EMMA.

383

which dates it is just possible, and no more, that there may have been some moment at which Harold was Earl of the East-Angles and William also was in possession of the see of London (see pp. 229, 238). The Earldom of Leofwine thus nearly answered to the under-kingdom which formed an apanage under Ecgberht and Ethelwulf (see vol. i. p. 27), and it does not differ very widely from what Londoners sometimes speak of as the Home Counties. But the great city itself was not subject to the jurisdiction of any Earl. The King's writs for London are addressed to the Bishop, the Portreeve or Portreeves, to the Burgh-thegns, and sometimes to the whole people ("ealle e burhware"). See Cod. Dipl. iv. 212, 213, 214.

I have thus tried, as well as I could, to trace out these singular fluctuations in the boundaries of the great Earldoms. To make matters clear, I have endeavoured to represent them by a comparative map of England at two stages of the reign of Eadward. The idea of such an attempt was suggested by the map given by Sir Francis Palgrave in his History of the Anglo-Saxons, p. 327. Some points of course are conjectural, and I have not been able to express the various fluctuations which happened at dates between the two years which I have chosen for illustration. But I trust that the two maps between them fairly represent the state of things in the earlier and in the later days of Eadward.

NOTE H. p. 40.

THE LEGEND OF EMMA.

As the name of Godgifu is best known to the world in general through the legend of her riding naked through Coventry (besides the references in p. 31, see R. Wendover, i. 496), so the name of Emma is best known through the legend of her walking unhurt over the hot ploughshares. The tale appears to have grown out of the real history of her disgrace at this time, mixed up with other particulars from various quarters. And when a prince stands in such singular relations both to his mother and to his wife as those in which Eadward stood to Emma and Eadgyth, it is not wonderful that, in the process of legend-making, the two injured Ladies got confounded.

The tale may be seen in Bromton, X Scriptt. 941. He seems to place the event in 1050, when Robert was already Archbishop of Canterbury. He calls it indeed the fourth year of Eadward, but he places it immediately before the events of 1051. The Norman Primate persuades the King that Emma-forty-eight years after her first marriage, fifteen years after the death of her second husband-had been guilty of too close an intimacy ("nimia familiaritas ") with Elfwine, Bishop of Winchester. The choice of an episcopal lover was unlucky, as Elfwine had already been dead three years (see p. 61); a more ingenious romancer would have named Stigand. The Bishop is imprisoned; the Lady is spoiled of her goods and sent to Wherwell, a manifest confusion with Eadgyth's banishment thither in 1051. From her prison, where she was not very strictly kept ("laxius custodita"), Emma writes to those Bishops in whom she trusted, saying that she is far more shocked at the scandal against Elfwine than at the scandal against herself. She is even ready to submit to the ordeal of burning iron in order to prove the Bishop's innocence. The other Bishops advise the King to

allow the trial, but the Norman Archbishop uses very strong language indeed. Emma is "fera illa, non fœmina;" her daring went so far that "amasium suum lubricum Christum Domini nominavit," and so forth. She may make compurgation for the Bishop ("vult purgare pontificem"), but who will make compurgation for herself? She is still charged with complicity in the death of Elfred, and with having made ready a poisoned bowl for Eadward himself. Yet, if she will make a double purgation, if she will walk over four burning shares for herself and five for the Bishop, her innocence shall be allowed. By dint of prayer to Saint Swithhun, the ordeal is gone through successfully. The penitent King implores pardon, and receives stripes ("disciplinas recepit") both from his mother and from the Bishop; he restores their confiscated goods; and Robert, if not actually banished, finds it convenient to leave England. In honour of the deliverance of the Lady and the Bishop, each gives nine manors, one for each ploughshare, to the Church of Winchester.

The account in the Winchester Annals (p. 21 et seqq. Luard) is substantially the same, and it sometimes agrees in words with that in Bromton. Unless Bromton has simply abridged the Winchester story, both are borrowed from the same source. But the Winchester annalist is very much fuller, and, after his manner, he puts long speeches into the mouths of his actors, that made by the Norman Archbishop displaying a remarkable acquaintance with the less decent parts of the satires of Juvenal. The most important difference is the introduction of Godwine. The event is placed in 1043. Archbishop Robert—he is already Archbishop-persuades the King to banish Godwine and his sons, to send his mother to Wherwell, and to forbid Elfwine to come out of the city of Winchester. The tale then follows much as before, only, together with the restoration of Emma and flight of Robert, Godwine and his sons are restored at the petition of Emma. Also, it was after these doings that Eadward seems to have first taken to working miracles; "Rex Edwardus magnis post hæc cœpit coruscare miraculis etiam in vitâ suâ."

I suspect that this is the older version. This is the Winchester writer's only mention of the banishment and return of Godwine. Bromton, or whoever is represented by that name, knew that Godwine's banishment happened at quite another time and from quite other causes; he knew also that Robert was not Archbishop in 1043. He therefore left out all about Godwine, and moved the tale to the year 1050, when Robert was Archbishop. But he failed to mark that he thus brought in a chronological error as to the death of Ælfwine. On this last point the local Winchester writer is of course accurate.

I cannot help adding good Bishop Godwin's inimitable account of the charges brought by Robert against Emma. "He began therefore to beate into the King's head (that was a milde and soft natured gentleman) how hard a hand his mother had held upon him when he lived in Normandy; how likely it was that his brother came to his death by the practise of her and Earle Godwyn; and lastly that she used the company of Alwyn Bishop of Winchester, somewhat more familiarly than an honest woman needed."

I may add that M. de Bonnechose (" ut erat miræ simplicitatis et innocentiæ," as the Winchester writer says of Eadward) believes everything. All about Godgifu, all about Emma, the "cruelle épreuve " and the "tragique scène," will be found in his Quatre Conquêtes, ii. 81-88. In

THE APPOINTMENT OF BISHOPS AND ABBOTS.

385

short, his history gives us, as Sir Roger de Coverley says, "fine reading in the casualties of this reign." Mr. St. John exercises a sound judgement, and Thierry seems to hold his peace.

NOTE I. p. 43.

THE APPOINTMENT OF BISHOPS AND ABBOTS.

I HAVE already (see vol. i. p. 339) said something about the manner of appointing Bishops and Abbots in the eleventh century. It is clear that the appointment rested both practically and legally with the King and his Witan; still we hear ever and anon of capitular elections, and in one case at least we even see some traces of that primitive but almost forgotten practice by which the clergy and the people at large of the vacant diocese claimed a voice in filling the episcopal chair. It may be worth while to bring together some of the more striking instances of episcopal and abbatial nominations in a reign in which Bishops and Abbots were of even more importance than usual.

The joint action of the King and his Witan cannot be better set forth than it is in the account given by the Abingdon Chronicler (1050) of the acts of the Gemót of London in Midlent 1051 (see p. 77). The words are, “Þa hæfde Eadwerd cing witenagemot on Lunden to Midlencten, and sette Hrodberd to arcebiscop to Cantwarebyrig, and Sperhafoc abbud to Lunden, and geaf Rodulfe biscop his mæge þæt abbudrice on Abbandune." In the case of Spearhafcc the Peterborough Chronicler (1048) also distinctly speaks of the King's writ and seal as giving the formal conveyance of the Bishoprick to the person nominated. Spearhafoc goes to the Archbishop with a document of this kind ("mid þæs cynges gewrite and insegle") and demands consecration. Several writs of this kind are extant in the case of both Bishops and Abbots. The form always is that the King grants the Bishoprick or Abbacy to such a person. Thus in Cod. Dipl. iv. 195 we have the writ conferring the Bishoprick of Somersetshire on Gisa, and opposite to it is a Latin version of the writ conferring the Bishoprick of Hereford on Walter. Both follow the same form. The essential words are "ich kýðe eów dat ich habbe geunnen Gisan mînan préste des biscoprîche hér mid éow." The writ then goes on to confirm all the rights and possessions of the see as fully as they were held by his predecessor or by any other Bishop; "swó ful and swó forð swó Duduc biscop of any biscop hit firmest him tóforen hauede on allem þingan." The form is much the same in the case of an Abbot (Cod. Dipl. iv. 225); "Ic kide ihú dat ic habbe unnen Baldewine abbot de abbotriche intó seint Eâdmundes biri." So in the Chronicles the passages are endless in which the King is said to give a Bishoprick; as in Abingdon, 1047; Eadward cyncg geaf Hecan his preoste pat bisceoprice." So Worcester, 1051; "Se cyng sealde Rotbearde pan Freoncyscan þe ær was bisceop on Lundene pæt arcerice." And in Peterborough, 1048, we read how Spearhafoc "sæt on þam biscoprice pe se cyng him ær geunnan hæfde be his fulre leafe." William of Malmesbury (Ġest. Pont. Scriptt. p. Bed. 116) does not scruple to use still stronger language; "Rex Robertum, quem ex monacho Gemmeticensi Londoniæ fecerat Episcopum, Archiepiscopum creavit." It was from the King's hands also that the Bishop received his staff, the staff which in the case of Ulf (see p. 76) was so nearly broken by the Pope.

66

This point and all the points at issue are very fairly stated and discussed in Lingard's Anglo-Saxon Church, i. 93, 94, and investiture by the staff is implied in the famous legend of Saint Wulfstan at the tomb of Eadward. In one case also, that of Leofric of Exeter (see p. 55), we see something like a personal installation by the King. The charter in Cod. Dipl. iv. 118, purporting to be the act of union of the Bishopricks of Devonshire and Cornwall, is indeed of very doubtful genuineness, but even a forger would hardly describe such a ceremony if nothing of the sort had ever taken place either at Exeter or anywhere else. "Itaque hoc privilegium ego Eadwardus Rex manu meâ super altare Sancti Petri pono, et præsulem Leofricum per dexterum brachium ducens, meaque Regina Eadgyca per sinistrum, in cathedrâ episcopali consisto, præsentibus meis ducibus et consanguineis, nobilibus necnon capellanis, et affirmantibus laudantibusque archiepiscopis Eadsino et Ælfrico, cum cæteris aliis quorum nomina describuntur in metâ hujus cautionis.” These words would of course imply a meeting of the Witan at Exeter, which is quite possible. It is not of course to be inferred that such a personal installation was necessary or usual; the translation of the see was a special act which was likely to be accompanied by special ceremonies. The point is that it is the King who in those ceremonies is the personal actor.

But in this, as in other matters, the action of the King, even when he alone is spoken of, in no way shuts out the conjoint action of the Witan. We have seen in the Abingdon Chronicle the action of the Witan in the appointments of the year 1050. In the Worcester (1051) and Peterborough (1048) narratives of the same appointments the King only is mentioned. So again in Abingdon (1045) we read, “Eadward cyng geaf Heramanne his preoste pat bisceoprice;" but in Worcester (1046) we find man sette Hereman on his setle," a form of words which often implies the action of the Witan. So on the other hand it is plain that the deposition of Stigand from the East-Anglian Bishoprick in 1043 was also the act of the Witan.

At the same time, there can be little doubt that these appointments of Prelates were among the cases in which the Witan for the most part did little more than register the King's edicts. The King would always nominate, and it would no doubt be thought a strong measure to object to his nomination. The practical power of the Witan would certainly not be greater than that of the Senate of the United States with regard to nominations made by the President. Indeed such cases as those of Robert and Ulf show that it could hardly have been so great. In one case however, that of the appointment of Stigand in the Mickle Gemót of 1052 (see p. 226), it is plain that the Witan, or rather the nation, exercised the freest will of their own.

But the action of King and Witan did not shut out some sort of action on the part of the monks or canons of the cathedral church. In some cases at least it is plain that they made an election in canonical form, and then petitioned King and Witan to confer the Bishoprick on the person chosen. This is most plainly marked in the great case of Elfric and Robert in 1050 (see p. 77). The account given by Eadward's Biographer (399, 400) puts this in the clearest light;

66

Quem [Elfricum] tam totius ecclesiæ universales filii, quam ipsius monasterii monachi, in archipræsulem sibi exposcunt dari, huncque et affectu communi et petitione eligunt præesse regulari. Mittunt etiam ad