Ælfgar, as a person of whose actions, if he ever existed, we have no knowledge, while of the other six brethren we know a good deal.

Among the daughters of Godwine, there is no need to prove the existence of Eadgyth the Lady. Another daughter, Gunhild, rests on the sure evidence of the Exon Domesday (pp. 96, 99, “ Gunnilla filia Comitis Godwini"). She also has a history, which will come in my fourth volume. The third daughter, Ælfgifu, also appears in Domesday (144 b), where land is held in Buckinghamshire by a "man" of hers, "homo Alvevæ soror Heraldi comitis." See Kelham, 153, and Ellis, i. 309, who both speak of her without any reference, and I have to thank Sir T. D. Hardy for pointing out the passage to me since the appearance of my second volume. This sister, though she is nowhere else directly spoken of, is of some historical importance. It is part of the story of Harold's oath (Sim. Dun. 1066 and elsewhere) that he promised to marry his sister to one of William's nobles. Obviously this cannot apply to Eadgyth, nor yet to Gunhild, who was devoted to a religious life. The sister intended must therefore have been Ælfgifu, and I shall, in my next volume, discuss the question whether she may not be the puzzling Ælfgyva of the Tapestry. See vol. iii. pp. 687, 690, 699.

As to the order of the sons there is no doubt. Swegen ("filius primogenitus Swanus," Fl. Wig. 1051) was the eldest. Harold came next. That Harold was older than Tostig is plain from the Biographer ("major natu Haroldus," 409), and indeed from the whole history. So even Saxo (207) speaks of "minores Godovini filii [which at least includes Tostig] majorem perosi." Orderic's notion (492 D) that Harold was younger than Tostig is simply a bit of the Norman legend, devised in order to represent Harold as depriving his elder brother, sometimes of the Earldom, sometimes of the Kingdom. Snorro's idea that Harold was the youngest of all is wilder still. The order of the several brothers is very plainly marked in the dates of their promotion to Earldoms; their order is Swegen, Harold, Tostig, Gyrth, Leofwine. Wulfnoth, who never held an Earldom, was doubtless the youngest.

The order in which the brothers sign charters is worth notice. Setting aside one impossible charter (Cod. Dipl. iv. 80-84), Swegen always signs before Harold, Harold always before Tostig, Tostig always before Gyrth and Leofwine. But Harold, Gyrth, and Leofwine do not observe so strict an order among themselves. May we

THE GREAT EARLDOMS UNDER EADWARD.

555

not infer from the recorded disposition and actions of Swegen and Tostig that a certain attention to ceremony was needed in their cases, while the other three brothers, who lived and died firm friends, could afford to dispense with it?

The order of the daughters among themselves must have been Eadgyth, Gunhild, Ælfgifu. A daughter of Godwine and Gytha who was talked of as an intended wife for any one in 1066, must have been the very youngest of the family.

The order of the sisters with regard to their brothers is more difficult to fix. It is hopeless to try to fix the place of Gunhild. But, as Ælfgifu must have been the youngest, there is some reason to believe that Eadgyth was the eldest of the family. The Biographer (p. 397) compares four children of Godwine, seemingly Eadgyth, Harold, Tostig, and Gyrth-he never mentions Swegen-to the four rivers of Paradise ;

"Felix prole piâ Dux, stirpe beatus avitâ,

His quatuor natis dans Anglia pignora pacis.
Prodit gemma prior, variæ probitatis amatrix,
In medio Regni, tanto Duce filia patre
Edgit digna suo, Regi condigna marito."

This looks as if Eadgyth was the eldest of all. Godwine and Gytha were married in 1019 (see vol. i. pp. 420, 723). Harold therefore, the second son, could not, even if Eadgyth was younger than himself, have been born before 1021, perhaps not till 1022 or later. He therefore could not have been above twenty-four when he became Earl, nor above forty-five at his death-he may of course have been younger. But none of Godwine's sons who held Earldoms could have been so young as William of Malmesbury fancied Gyrth to be in 1066, when he calls him (iii. 239) "plus puero adultus et magnæ ultra ætatem virtutis et scientiæ." He had then been Earl of the East-Angles for nine years.

NOTE G. p. 36.

THE GREAT EARLDOMS DURING THE REIGN OF EADWARD.

It is not always easy to trace the succession of the men who ruled the different Earldoms of England during the reign of Eadward. In several cases the Chronicles give us notices of the death, deposition, or translation of one Earl and of the appointment of his

successor. But these entries taken alone would not enable us to put together a perfect series of the Earls. For instance, Eadwine (1065), Gyrth (1066), Leofwine (1066), Waltheof (1066), are all spoken of as Earls without any account of their appointment, and, in the last three cases, without any hint as to the districts over which they ruled. To make out anything like a perfect list, we must go to various incidental notices in the royal writs and elsewhere. By their help we shall be able to recover, not indeed an absolutely complete account, but one much fuller than appears on the face of the history, and one which reveals to us a great number of anomalies which we should not have expected. The way in which several Earls held isolated shires detached from the main body of their Earldoms, and the way in which shires were transferred from the jurisdiction of one Earl to that of another, are both of them very remarkable.

For a complete view of these changes, indeed for any complete view of the general succession of the Earls, we must go back to the fourfold division of England by Cnut in 1017 (see vol. i. p. 404). Cnut then kept Wessex in his own hands, and appointed Eadric over Mercia, Thurkill over East-Anglia, Eric over Northumberland. In 1020 (see vol. i. p. 422), Wessex also became an Earldom under Godwine. Now in these four great governments we can trace the succession of Earls without difficulty, with the single exception of East-Anglia. We have no account of that Earldom from the banishment of Thurkill in 1021 (see vol. i. p. 425) to the appointment of Harold, seemingly in 1045 (see above, p. 37). As for Northumberland, I have already traced out the succession of its Earls (see vol. i. p. 520 et seqq.). There is no doubt that, at the accession of Eadward, Siward was in possession of both parts of the old Northern realm, and that he remained in possession of them till his death. The succession in Wessex is plainer still; Godwine was appointed in 1020, Harold succeeded him in 1053; there is no room for any question, except as to the disposal of the Earldom during the year of Godwine's banishment. And the mere succession in Mercia is equally plain. Leofwine succeeded Eadric in 1017; Leofric succeeded Leofwine some time between 1024 and 1032 (see vol. i. p. 718); Ælfgar succeeded Leofric in 1057; Eadwine, there can be no reasonable doubt, succeeded Ælfgar on his death, at some time between 1062 and 1065. Our difficulties are of other kinds. There

THE GREAT EARLDOMS UNDER EADWARD.

557

is, first, the great uncertainty as to the meaning of the name Mercia. There is the fact that various shires, especially in Mercia, are found in the hands of others among the great Earls than those to whom the fourfold division would seem to have committed them. There is the fact that we find mention of Earls holding Earldoms other than the four great ones, and seemingly formed by dismemberments of the four. Lastly, we find, especially under Cnut, the names of several Earls whom it is not easy to supply with Earldoms.

This last difficulty need not greatly trouble us. It does not follow that every Danish chief who signs a charter of Cnut with the title of Earl was actually established in an English Earldom. On the other hand, some one must have ruled in East-Anglia between 1021 and 1045, and it is a fair guess, though nothing more, that the successive husbands of Gunhild, Hakon and Harold (see vol. i. pp. 426, 427), who are spoken of as if they had some permanent connexion with England, were Earls of the East-Angles during some parts of that interval. The main difficulty springs from what seem to have been the constantly fluctuating arrangements of the Mercian shires. The old chaotic state of central England seems to revive. First, it is not always clear what we are to understand by the name Mercia. The name at this stage sometimes takes in, sometimes excludes, those parts of old Mercia which were ceded by Alfred to Guthrum. Secondly, we find various shires, Mercian in one or the other sense, which are not under the government of the person spoken of as the Earl of the Mercians.

Now when Wessex, Northumberland, East-Anglia, and Mercia are spoken of as an exhaustive division of England, as they are spoken of in the fourfold division made by Cnut, there can be no doubt that Mercia is to be taken in the widest sense, as meaning the whole land from Bristol on the Avon to Barton on the Humber. With this great government Eadric was invested. But it is equally plain that, at a somewhat later time, either Mercia in this sense was dismembered in favour of independent Earls, or else subordinate Earls were appointed under a superior Earl of the Mercians. I will now put together the evidence which we find on these heads.

The first hint which we come across of a dismemberment of this kind is in 1041, when we find Thuri or Thored, "Comes Mediterraneorum," and Rani or Hranig, "Comes Magesetensium," distinguished from Leofric, "Comes Merciorum." Of Thored we

also know that his Earldom took in Huntingdonshire. See vol. i. p. 515, where a writ of Harthacnut addressed to him is quoted. And one may suspect that we ought to substitute the same name for "Toli comes " who in a Huntingdon writ of Eadward (Cod. Dipl. iv. 243) is addressed along with Bishop Eadnoth, fixing the date of the writ to the years 1042-1050. (This Toli can hardly be Tolig who is elsewhere addressed in Suffolk, seemingly as Sheriff under the Earldom of Gyrth. Cod. Dipl. iv. 222, 223.) Of Ranig we know that he held the rank of Earl as early as 1023 (see vol. i. p. 515). We may therefore be inclined to suspect that Mercia was dismembered on the death of Eadric, and that, besides the Mercian Earldom held by Leofwine and Leofric, two fresh Earldoms, whether subordinate or independent, were formed within the limits of the old Mercian Kingdom. On the whole I am inclined to think that a certain superiority was always retained by Leofric, as chief Earl of the Mercians. He always fills a special place, alongside of Godwine and Siward, and we shall come across evidence to show that some of the dismembered shires did, in the end, revert to him or to his house.

As to this Earldom of the "Mediterranei" or Middle-Angles, held by Thored, we have no distinct account of its extent. But it is a probable guess that it took in the whole eastern part of Mercia, the part in which the Danish element was strongest. I am inclined to think that in this Earldom Thored was succeeded by Beorn. Our indications are certainly slight, but they look that way. We hear nothing of Thored in Eadward's time, except a signature of "Duri dux" in Cod. Dipl. iv. 131, which, as the deed is also signed by Mannig Abbot of Evesham, must be as late as 1044 (see p. 69). On the other hand it is plain (see p. 36) that Beorn held some Earldom from about the year 1045 till his murder. We know also that his Earldom took in Hertfordshire (Cod. Dipl. iv. 190). I infer then that in 1045 Beorn succeeded Thored as Earl of the Middle-Angles, of Eastern or Danish Mercia. I also infer that in that Earldom he had no one successor. No Earl is spoken of in the later days of Eadward who can show any claim to such a description, and several of the shires contained within the country which I conceive to have been held by Thored and Beorn seem to have remained in a sort of fluctuating state, ready to be attached to any of the great governments, as might be convenient.