THE GREAT EARLDOMS UNDER EADWARD.

559

But

Thus Huntingdonshire was within the Earldom of Thored. in 1051 (Flor. Wig. in anno) we find it, together with Cambridgeshire, a shire still so closely connected with it as to have a common Sheriff, detached altogether from Mercia, and forming part of the East-Anglian Earldom of Harold. "Men" of Harold's in Huntingdonshire accordingly occur in Domesday (p. 208). But Huntingdonshire was afterwards separated from East-Anglia, perhaps on Harold's translation to Wessex in 1053. It then became, strange to say, an outlying portion of the Earldom of Northumberland. It does not however appear that Cambridgeshire followed it in this last migration. That Huntingdonshire was held by Siward is shown by a writ (Cod. Dipl. iv. 239) coming between 1053 and 1055. It is certain that it was afterwards held by Waltheof. Domesday also (208) implies the succession of Siward, Tostig, and Waltheof, by speaking of "men" and rights belonging first to Tostig and afterwards to Waltheof. It might be worth considering whether some confused tradition of these transfers of the shire formed an element in the legend of Tostig, Earl of Huntingdon, slain by Siward. See vol. i. pp. 461, 587.

as

Northamptonshire, like Huntingdonshire, was separated from Mercia and attached to Northumberland. This is distinctly shown by a royal writ addressed to Tostig as its Earl (Cod. Dipl. iv. 240). The only other Northamptonshire writ that I know (iv. 216) is addressed to Bishop Wulfwig without any Earl's name. But, to Northamptonshire, another question might arise. The singular description of the daughter of the Northumbrian Earl Ælfhelm as Elfgifu of Northampton (see vol. i. p. 453) may possibly point to an earlier connexion between the two districts. This last is a mere guess, but the connexion between Northumberland and Northamptonshire during part of the reign of Eadward is quite certain. But Northamptonshire and Huntingdonshire were afterwards again detached from Northumberland, and held as a separate Earldom by Waltheof. On this point the evidence seems quite plain; the only question is as to the exact date. Waltheof held some Earldom at the end of the year 1066, when he is spoken of as an Earl with Eadwine and Morkere (Chron. Wig. 1066). Under William, besides his great Northumbrian government, he was certainly Earl of Northamptonshire (Ord. Vit. 522 C) and of Huntingdonshire (Will. Gem. viii. 37). We

may therefore infer that these fragments of his father's government formed the Earldom which he had held under Harold. The false Ingulf (Gale, i. 66) makes him receive both these shires on his father's death in 1057, Tostig receiving Northumberland. The Chronicle of John of Peterborough, which, though not contemporary, has some authority as being a local record, distinctly makes Waltheof succeed to Northamptonshire on his father's death in 1055; Siwardus Dux Northanhumbrorum obiit; . . . cujus filius Waldevus, postea martyr sanctus, factus est Comes Northhamptoniæ; comitatus autem Northanhumbrorum datus est Tostio fratri Haroldi" (Giles, p. 50). But this is shown to be incorrect by the charter just quoted, which shows that Tostig was Earl in Northamptonshire. And the course taken by the Northumbrian rebels in 1065 (see p. 489) seems to point to a still abiding connexion between that shire and Northumberland. We can therefore hardly doubt that both Northamptonshire and Huntingdonshire were obtained by Waltheof as a result of the Northumbrian revolt in 1065.

About Nottinghamshire I do not feel quite certain. It appears from Domesday (280) that Tostig had certain rights in the town of Nottingham; but he is not distinctly spoken of as Earl of the shire. But the connexion between this shire and the Northumbrian Primate makes a connexion with the Northumbrian Earl far from unlikely.

In

Hertfordshire formed part of the Earldom of Beorn. We have no further account of it till after the redistribution in 1057 (see above, pp. 418, 419), when it appears in the hands of Leofwine. Two writs (Cod. Dipl. iv. 217, 218) are addressed to him as Earl, conjointly with Wulfwig, Bishop of Dorchester-the Prelate of the Middle-Angles-whose episcopacy ranges from 1053 to 1067. Domesday also (132) eighteen burghers in the town of Hertford are described as being "homines Heraldi Comitis et Lewini Comitis," perhaps a sign of the superiority exercised by Harold over the Earldoms of Gyrth and Leofwine. Men of Leofwine occur also in the town of Buckingham (143) and in other parts of that shire (144, 145), suggesting that Buckinghamshire also made part of his Earldom. Of Bedfordshire we seem to have no distinct account. Waltheof (Domesday, 210 b) held lands there, but it need not have been in his Earldom.

THE GREAT EARLDOMS UNDER EADWARD.

561

Oxfordshire appears in 1015 (Flor. Wig. in anno) as part of the Earldom of Swegen. (See above, p. 36.) After 1057 it appears as an outlying appendage of the East-Anglian Earldom of Gyrth. Two writs for Oxfordshire are addressed to him conjointly with Bishop Wulfwig (Cod. Dipl. iv. 215, 217). The former is the well known grant of Islip to the church of Westminster.

Of the other East-Mercian shires we have no account. But I am inclined to believe that they must have reverted to Leofric, perhaps on the death of Beorn. I am led to this belief by the almost certain fact that Lincolnshire did. All history and tradition connects Leofric and his house with that shire; one of the great objects of his bounty, the minster of Stow, is within its borders, and it is plain that, in 1066 (Flor. Wig. in anno), Lindsey formed part of the Earldom of his grandson Eadwine.

The shiftings of the East-Mercian shires are thus frequent and perplexing, but those of West-Mercia are equally so. That the north-western shires remained constantly under Leofric and his house there can be no reasonable doubt. Our one writ in those parts (Cod. Dipl. iv. 201) is addressed to Eadwine in Staffordshire, and the entries of property held in that shire and in Cheshire by him and his father are endless. The same may be said of Shropshire, but as soon as we get south of that limit, we are at once in the region of fluctuations. We have seen that Ranig was Earl of the Magesætas or of Herefordshire in 1041. It is impossible to say whether his government extended beyond that limit. One can hardly doubt that Ranig was succeeded by Swegen, whose Mercian possessions (Flor. Wig. 1051) consisted of the shires of Hereford, Gloucester, and Oxford. It is therefore not unlikely that Ranig's government was of the same extent, but we cannot be certain. But it is quite certain that Herefordshire was detached from the government of Leofric and his successors during the whole reign of Eadward. It is not clear what became of the shire during Swegen's first banishment. Something belonging to Swegen, either his Earldom or his private estate, was (see pp. 89, 101) divided during his absence between Harold and Beorn. It is therefore quite possible that one or other of them may have governed Herefordshire from 1046 to 1050. But it is equally possible that the shire was, during that interval, held by Ralph of Mantes, Ralph the Timid, the son of Walter and Godgifu.

Indeed this last view becomes the more likely of the two, when we remember the firm root which the Normans had taken in Herefordshire before 1051 (see p. 138), which looks very much as if they had been specially favoured in these parts. That Ralph succeeded Swegen on his final banishment in 1051 I have no doubt at all. Sir Francis Palgrave (English Commonwealth, ii. ccxc.) calls this fact in question on the grounds that, at the time when William of Malmesbury (ii. 199) calls him "Comes Herefordensis," Herefordshire was under the government of Swegen, and that, when Florence (1055) speaks of his doings in the Herefordshire campaign, he does not formally describe him as Earl of the shire. But surely, when a certain shire is invaded, and a certain Earl goes forth to defend it, the presumption, in the absence of some distinct evidence the other way, is that the Earl who so acts is the Earl in charge of the shire. The passage of William of Malmesbury is simply one of his usual confusions of chronology. Speaking of Eustace of Boulogne and his visit to England in 1051, he mentions his marriage with Godgifu, and goes on thus, quæ ex altero viro, Waltero Medantino, filium tulerat Radulfum, qui eo tempore erat Comes Herefordensis, ignavus et timidus, qui Walensibus pugnâ cesserit, comitatumque suum, et urbem cum episcopo, ignibus eorum consumendum reliquerit; cujus rei infamiam maturè veniens Haroldus virtutibus suis abstersit. Eustachius ergo.. Regem adiit." Undoubtedly, according to strict grammatical construction, "eo tempore" ought to mean in 1051, but William so jumbles together the events of 1051 and of 1055 that it is hardly safe to argue from this expression that he meant distinctly to assert that Ralph was Earl of Herefordshire in 1051. He may just as well have meant that he was so when he waged his unfortunate campaign with the Welsh, and certainly no one who got up his facts from William of Malmesbury only would ever find out that that campaign happened four years after the visit of Eustace.

Ralph then, I hold, was certainly Earl of Herefordshire in 1055, and the natural inference is that he succeeded Swegen in 1051, and that, as Swegen never came back, he was allowed to retain his Earldom in 1052. That Ralph was succeeded by Harold in 1057 there can be no doubt. But Harold's Herefordshire Earldom is so important as a piece of national policy, and it is connected with so many points in Harold's character, that I have spoken

THE GREAT EARLDOMS UNDER EADWARD.

563

of it somewhat largely in the text. See pp. 395, 417, and, for writs addressed to Harold in Herefordshire, see p. 547.

But we have also the fact that Ralph certainly held the rank of Earl in the year 1051, while Swegen was still acting as Earl of the Magesætas (see p. 141). We have also his signatures as Earl as early as 1050 (see p. 111). Sir Francis Palgrave is therefore very possibly right in quartering him in Worcestershire. That shire, he is inclined to think, was in Cnut's time held by Hakon the doughty Earl, the first husband of Gunhild. This view he rests on a writ of Cnut's (Cod. Dipl. iv. 56) addressed to him as Earl in Worcestershire. The writ is clearly spurious, but it is perhaps one of those cases in which a spurious document proves something. Would a forger insert a name so little known as that of Hakon in a spurious writ, unless he had seen it in a genuine writ? Again, it is rather remarkable that in two Worcestershire documents (see a deed of Bishop Ealdred, Cod. Dipl. iv. 137, evidently passed in a Worcestershire Scirgemót, and another, iv. 262) there is mention of Danish Thegns ("ealla da yldestan þegnas on Wigeraceastrescire, Denisce and Englisce") as a distinct class in Worcestershire. This is what we should hardly have looked for so far west, and it may possibly be taken in connexion with the complaints about Danish spoilers of the Church of Worcester, which we have seen in pp. 544, 560. This prevalence of Danes in the shire looks of itself like the effect of the administration of a Danish Earl, and we find also what seems to be a distinct mention of a Hakon as holding a prominent position in the shire. In a document of Bishop Æthelstan of Hereford in Cod. Dipl. iv. 234 we find, joined together in a transaction of the time of Cnut, "Leofwine Ealdorman and Hacc . . and Leofric, and eal seo scír." In Mr. Thorpe's Diplomatarium, p. 376, the name is supplied in full, "Hacun," which one might almost have ventured to do without manuscript authority. Hakon is thus placed between Ealdorman Leofwine and his son and successor Leofric. This looks very much as if Hakon were a subordinate Earl of Worcestershire under Leofwine as superior Earl of the Mercians. If so, he may, or may not, have been removed from Worcestershire to the greater government of the East Angles. But, if we admit Hakon, we still have no means of bridging over the interval between his death in 1030 and Ralph's appearance in 1041. Ralph, I suspect,