The Domesday survey: Introduction

Domesday Survey

The county boundaries of the Anglo-Saxon period possessed a significance which does not belong to their modern representatives. In the age immediately preceding the Norman Conquest, when the central government was weak, and its action spasmodic, responsibility for the good government of a shire lay in the first instance on its thegns assembled in its court. The court of the shire was the custodian and exponent of its local customs, such as those which are set out in an appendix to the description of the king's land in the Oxfordshire Domesday, and its judgement, based on the continuity of its traditions, was usually decisive in pleas relating to land within its limits. In the organization of public finance, although it was in the lesser court of the hundred that the individual tax-payer met the king's officers, the assessment of each village represented the share assigned to it in an original large, round assessment laid on the shire as a whole. At the beginning of the 11th century, and probably for a hundred years before this time, the king's financial officers assumed that Oxfordshire contained exactly 2,400 of those ancient family lands, known in English as 'hides', which formed the basis of all early taxation in England. (fn. 1) Under conditions of which nothing is now known, these 2,400 hides were distributed in large blocks among the several hundreds of the shire, and the number of hides assigned to a particular place by Domesday Book is the result of a further process by which the men of each hundred distributed its share of the county assessment among the several villages within its territory.

Naturally, they worked by round numbers, and their habit of mind is seen in the assessment of innumerable villages at precisely 5 hides, (fn. 2) or some multiple of that amount. Assessments of this type are distributed over the whole of the county. Ardley, Islip, Elsfield, Checkendon, Cogges, Hanwell, Stratton Audley, Swerford, Cuxham, and Harpsden are each assessed at 5 hides; Headington, Glympton, Mapledurham, Kingham, Ambrosden, Emmington, North Leigh, Taynton, Charlton-on-Otmoor, Newnham Murren, and Wigginton, at 10; Sydenham and Great Rollright, at 15; Goring, Broughton, Sarsden, Caversham, Dunsden, and Tadmarton, at 20; Hooknorton and Witney, at 30; Pyrton with its hamlets, at 40. There are many village assessments which do not seem to fit into this system, such as those of Wroxton (17 hides), Bicester (15½ hides), and Broadwell (24¼ hides). But it has been shown in regard to some other counties that an irregular assessment recorded for a single place is often part of a round number of hides laid on two or more villages jointly. Few combinations of this kind have been traced in Oxfordshire, but it cannot be through chance that the irregular figures given in the three entries which relate to Bloxham and Adderbury (34½ hides, 14½ hides, and 1 hide) amount together to a neat total of 50 hides. In any case, the 5-hide unit of assessment appears so clearly on the surface of the Oxfordshire survey that it is hardly necessary to search for further illustrations of the system to which it belongs.

For each of the counties which adjoin Oxfordshire, Domesday supplies sufficient information for a reconstruction of the hundreds existing in 1086. But the Oxfordshire Domesday rarely indicates the hundred within which a particular village was situated, and although it seems clear that at least 22 divisions known as hundreds or half-hundreds existed within the county in the 11th century, most of their names are unknown, (fn. 3) and it is only from later evidence that any section of their boundaries can be drawn. The Domesday description of the king's land states that 4½ hundreds were annexed to the manor of Bensington, 2 hundreds to Headington, 2½ to Kirtlington, 3 to Wootton, 3 to Shipton-under-Wychwood, 2 to Bampton, and 2 to Bloxham. In addition to the hundreds represented in this list, it is probable that in the 11th century, as later, the ancient estates of the Bishop of Lincoln within the county formed three separate hundreds: one consisting of the manor of Dorchester, assessed at 90 hides, another of the manors of Banbury and Cropredy, each assessed at 50 hides, and the third of the manors of Thame and Milton, assessed respectively at 60 and 40 hides. (fn. 4) The system by which groups of hundreds were annexed to single royal manors survived in Oxfordshire far into the Middle Ages. (fn. 5) The 'three hundreds of Wootton' are mentioned in the reign of Henry II; (fn. 6) the manor of Headington with two hundreds, that of Bullingdon and 'the hundred outside the North Gate' of Oxford, was given by King John to Thomas Basset; (fn. 7) and the Hundred Rolls of the reign of Edward I show that the hundreds of Lewknor, Pyrton, Langtree, and Binfield, with the half-hundred of Ewelme, were still attached to Bensington. (fn. 8) But the system as described in Domesday Book was itself descended from a form of local organization, older than either shire or hundred, by which wide regions were regarded as belonging to the royal estates where the king's food-rent from the dependent territory was collected. In the Middle Ages, fragments of this archaic system can be identified in many parts of England, but there are few counties where its influence was so permanent as in Oxfordshire.

The seven great royal manors of Bensington, Headington, Wootton, Shipton-under-Wychwood, Bampton, Kirtlington, and Bloxham each yielded a yearly rent in money to the king. These rents, which ranged between £85 from Bensington and £18 from Wootton, represented a revenue drawn from many different sources. The profits of justice done in the courts of annexed hundreds must have formed an important part of it, although the Survey gives no indication of their amount. At Bensington and Headington a small sum was received under the name of 'church scot': an ancient charge, originally payable to the 'old minsters' which had been the first centres of parochial organization, but frequently secularized before the end of the 11th century. At Headington the king received 30s. as the commutation of a duty known as 'helueuuecha', a compound word of which the second element is clearly the Old English wæcce, 'watch', but the first is uncertain. At Bampton the manorial rent covered 50s. from the miscellaneous sources described as 'pannage, salt-pans at Droitwich, and the other customs of the men'. (fn. 9) But the most important of all these payments was the considerable sum taken within each manor, year by year, as a commutation of ancient corn-rents, once paid in kind. These sums, which amounted to £30 at Bensington and £28. 10s. at Bloxham, were only exacted from the seven great manors to which hundreds were annexed, and they may be regarded as distinctive of the ancient demesnes of the Crown in this county. Little is known of the conditions of these estates in King Edward's time. Domesday incidentally refers to a period when Bloxham had been held by Tostig, Earl of Northumbria, King Edward's friend, and states that Edwin, Earl of Mercia, held the manor in 1066. But there is no evidence that any other of these royal demesnes had been alienated in the years before the Conquest. Kirtlington had certainly belonged to Edward the Confessor, (fn. 10) for he gave Launton, which lay within the ambit of its jurisdiction, to his new foundation of Westminster. His grant of land at Marston to the same abbey (fn. 11) suggests that he was in possession of Headington, for it is probable that in the 11th, as in the 13th, century Marston was a member of that manor. The whole tenor of the survey implies that Bensington, Wootton, Bampton, and Shipton-underWychwood had also been in his hands.

Only a brief description is given of the forests which for more than a century after the date of Domesday Book continually attracted the late Norman and the early Angevin kings to Oxfordshire. A concise entry states that the king's demesne forests of Shotover, Stowood, Woodstock, Cornbury, and Wychwood were 9 leagues in length and in breadth, that six villeins and eight bordars had 3½ ploughs on 4½ hides within these forests, and that a certain Rainald paid the king £10 each year for all the profits arising out of them. These meagre details conceal rather than express the fact that in the 11th century the open country to the north of Oxford between the lower courses of the Evenlode and the Cherwell formed the only considerable interruption to a thick belt of woodland extending from the hills above Burford to the forest of Bernwood in western Buckinghamshire. Royal interest in the forests of Oxfordshire is carried back more than fifty years beyond the Conquest by the record of a council held by Æthelred II 'at Woodstock in the land of the Mercians', (fn. 12) which anticipates the numerous assemblies of magnates held at the king's hunting-quarters in that place during the 12th century. The statement in the Domesday description of Bampton that Henry de Ferrars holds a certain wood which Bundi the forester had held makes an appreciable addition to the little that is known of the organization of the Old English royal forest. The succession of Henry de Ferrars to Bundi the forester identifies Bundi with the important thegn of that name, whose lands in several counties, as at Dean and Chalford in Oxfordshire, had passed to Henry by 1086. As Bundi is described elsewhere in Domesday Book as 'staller' and 'constabularius', it is clear that he was one of Edward the Confessor's household officers, (fn. 13) and the employment of such a man as a royal forester curiously resembles the Norman practice of using persons of high position as guardians of the king's sporting privileges.

The description of the king's forests is followed by a list of miscellaneous payments due to the king. Two sums, of £20 by weight 'from the borough', and of £20 in pennies at 20 to the ounce 'from the mint', clearly represent charges laid on the county town in addition to its own 'farm' of £60. The payment from the mint, to which there are many parallels, presents no difficulty, but it is not easy to form a clear impression of the dues in respect of which the £20 'from the borough' was paid. It is perhaps most probable that it represented the commutation of a custom which had once associated the town with the county in the support of the king by the payment of a 'farm', or food-rent. The first entry in this list of payments states that the shire rendered £150 as the 'farm of three nights', the second states that it paid £25 by weight 'by way of increment', clearly an addition to the original £150, and the position of the third entry, which records the payment of £20 by weight from the borough, suggests that it relates to the same charge. The remaining entries in the list, as a whole, are less obscure. The county paid 4s. for weapons, 100s. by tale as a gift to the queen, £10 for a hawk, 20s. for a sumpter-horse, £23 in pennies of 20 to the ounce for hounds, 6 sestiers of honey, and 15d. 'by way of custom', a phrase which must be left without explanation. The adjacent county of Northampton was charged with a similar, but much less onerous series of payments; the 'three nights' farm' which it rendered only amounted to £30 by weight, without any increment such as was demanded from Oxfordshire, and the difference was not offset by larger payments for the queen's gift and for hounds, and by two additional sums of 20s. each, exacted 'for alms' and 'for a huntsman's horse'. Although some of its details are obscure, the series as a whole gives an impression of high antiquity, with its insistence on the duty of the shires to provide hawks and hounds for the king's sport, and its hint of a time when the king had been maintained by his subjects as he passed over the country.

The account of the king's Oxfordshire lands and revenues is completed by a short list of local customs from which profit might come to him at any time. In the 11th century breach of the special peace given to an individual under the king's hand or seal was among the gravest of crimes. If the slayer of a man enjoying this protection were taken, his life and limbs were at the king's mercy. If he escaped, he became an outlaw, and the man who could kill him received his spoils. The housebreaker who killed, wounded, or assaulted any one as a result of an unlawful entry could make terms with the king by the payment of 100s., but the man who sought out another in his house in order to kill him forfeited his life, and all his property except his wife's dower passed to the king. (fn. 14) Neglect of a special summons to the host could be emended in Oxfordshire by a payment of 100s., in contrast to the contemporary Berkshire rule which visited this offence with the total forfeiture of the offender's land. A rule which seems to stand alone in Domesday Book provided that if a stranger who had chosen to live in Oxford and had a house there died without kinsfolk, all that he had escheated to the king. It is evident that these passages, which are often described as a 'custumal' of Oxfordshire, are really a casual collection of rules, set down because of the king's financial interest in their provisions.

In 1086 six bishops of English sees held land directly from the king in Oxfordshire. Most of their estates had come to them from Anglo-Saxon predecessors. The Bishop of Worcester's 10-hide manor of Spelsbury, included anomalously in the survey of Warwickshire, had been granted to his church by Berhtwulf king of the Mercians, between 838 and 845. (fn. 15) It can safely be identified with the 10 hides in Wychwood Forest which Berhtwulf gave to Bishop Heahberht of Worcester by a charter dated on Christmas Day 841, that is, 840 by modern reckoning. (fn. 16) The Archbishop of Canterbury's estate at Newington near Dorchester had been obtained for the monks of Christ Church by Emma, mother of Edward the Confessor, from her second husband, King Cnut, to whom the land had fallen through the forfeiture of a certain thegn. (fn. 17) The Bishop of Winchester held Witney in virtue of a gift by the Confessor to Ælfwine, the predecessor in that see of the pluralist Archbishop Stigand. (fn. 18) Adderbury, the smaller of his two Oxfordshire manors, had come to his church under the will of the Ætheling Æthelstan, the eldest son of Æthelred II. (fn. 19) The Bishop of Salisbury's single estate of Dunsden was only separated by the Thames from his vast manor of Sonning, to which it was afterwards annexed. Nothing is known of the conditions under which either Sonning or Dunsden came into religious hands, but the extent and continuity of the estate formed by the two manors, which stretched for nearly 20 miles from the summit of the Chilterns to the Hampshire border at Sandhurst, suggests that it represents a royal gift of the 7th or at latest the early 8th century. (fn. 20) The early history of the Bishop of Exeter's manor of Bampton, though clearer, is by no means free from uncertainty. Domesday itself records that the manor was held by Leofric, Bishop of Exeter, before the Conquest, and it is natural to assume that it was an ancient possession of his see. But in 1069 the Conqueror made a charter for the express purpose of allowing Bishop Leofric to grant this land, with a smaller property at Holcombe in Dawlish, to St. Peter's Church at Exeter, (fn. 21) and a list of the bishop's gifts to his cathedral states that the land at Bampton had formed part of his own property. (fn. 22) In the Conqueror's charter, the Bampton estate, described in an endorsement as 6 hides at Bampton, Aston, and Chimney, is called 'the land which king Eadwig gave to the saint and the community at Bampton', a phrase which strongly suggests that it had formed the endowment of a minster at that place before into came in Leofric's possession.

The Oxfordshire estates of the Bishop of Lincoln fall into two main divisions. The first comprises the ancient endowments of the see of Dorchester, which were organized in relation to the bishop's demesnes at Dorchester itself, Thame, Great Milton, Banbury, and Cropredy. Although Domesday incidentally refers to the bishop's 'manors' of Thame, Banbury, and Cropredy, these estates, like the remainder of the group to which they belong, were far older than the appearance of any manorial system in England. Dorchester had probably remained in ecclesiastical hands since the days of Birinus, the first bishop of the West Saxons. Each of these so-called manors contained a number of distinct villages, and each of them probably represents the exemption of a large number of peasant households from royal dues for the benefit of one or other of the early bishops who sat at Dorchester, or elsewhere in this country. Charlbury, which with its hamlets formed a detached portion of the bishop's estate of Banbury, is connected with the conversion of the midlands by the fact that the relics of Diuma, the first bishop of the Mercians, were preserved there in the 10th century. (fn. 23) Like the king's manor of Bensington and the Bishop of Salisbury's manor of Sonning, the Bishop of Lincoln's Oxfordshire manors are survivals from a time when lordship, secular or ecclesiastical, was maintained by food-rents brought to the lord's house from the villages and farms of a wide region.

In 1086 each of these manors fell into two parts, the first consisting of the portion of the estate which yielded a regular revenue to the bishop, the second, of the portion held of him by tenants of more than peasant rank. The distinction is not maintained with absolute precision. On the manor of Dorchester the large holding of an Englishwoman named 'Bristeva' (fn. 24) is reckoned as part of the bishop's share, and the 2 hides held by a certain Robert at Wickham within the manor of Banbury are expressly described as belonging to the bishop's 'inland', or demesne. The significance of the distinction is made clear by the fact that on each of the five manors except Banbury, the tenants are collectively described as knights. Here again the distinction is not maintained rigorously; a group of English free men held 3½ hides of the knights' land of Dorchester. But it seems clear that most, if not all, of the tenants who are named were knights in the technical sense of the word, and some of them can be identified with men of the bishop who appear in later records. (fn. 25) The bishop had been charged with the heavy burden of finding 60 knights for the king's service, and although he had received extensive estates in other counties, which he used for their enfeoffment, the Oxfordshire Domesday shows that he had been compelled to sacrifice for this purpose a very considerable portion of the ancient demesne of his see.

The second main division of the bishop's land consists of a group of manors which belonged to the abbey of Eynsham. There had been an important church at Eynsham in the ninth century, (fn. 26) but nothing is known of its later history, and the abbey mentioned in Domesday Book was a new foundation of the reign of Æthelred II. (fn. 27) It had been deserted by its monks in the disorders which followed the Norman invasion, and its re-establishment was the work of Bishop Remigius of Lincoln, (fn. 28) to whom the king seems to have granted a part of its pre-Conquest endowment. Of its original lands there remained to it in Oxfordshire Eynsham itself, 3 hides in Shifford, and 9½ hides in Yarnton, but 5 hides in Little Rollright are also said to 'belong to the church', a phrase which implies that they had been given to it before the Conquest. In 1086 Roger d'Ivry, one of the greatest barons of Oxfordshire, was the bishop's tenant of the land at Yarnton, but the rest of the property was held under the bishop by Columbanus, the first abbot of the reconstituted monastery.

Of the seven bishops who ruled the Norman Church in 1086, three were tenants in chief in Oxfordshire. Gilbert Maminot, Bishop of Lisieux, the Conqueror's physician, held an estate in Dunstew, Dunthrop, and Westcot Barton which had belonged to an Englishman named Leofwine. Geoffrey, Bishop of Coutances, does not appear in the list of Oxfordshire landowners prefixed to the description of the county, nor are his possessions described in its text, but the sum of them amounted to nearly 50 hides, and they formed an important property in the centre and east of the shire, extending from Glympton to Finmere on the Buckinghamshire border. For some reason, which remains a mystery, the entire holding was entered in the survey of Northamptonshire. On most of his fee the bishop was the successor of Wulfward the White, a thegn of Queen Edith, who had survived until within a few years of 1086. (fn. 29) The whole fee was sublet to under-tenants, but none of them have yet been identified with men occurring in other records, and in particular the William who held Glympton of the bishop has not yet been connected with the family which derived the name of Clinton from its lordship of that place.

Odo of Bayeux, the third Norman bishop, had been under arrest since 1082, but the manors assigned to him in the Oxfordshire Domesday show that he was the greatest landowner in the county at the time of his fall. His holding, like that of the Bishops of Lisieux and Coutances, was a personal grant from the king, and on, if not before, his death in 1097 it was disintegrated, some of his tenants' fees becoming separate baronies, while others were annexed to the existing fiefs of different tenants in chief. Historically, the most interesting of his tenants is the Wadard who held of him in Oxfordshire at Tythrop, Fringford, Wilcote, Brighthampton, Cassington, South Newington, Barton Ede, Balscot, Little Tew, Ludwell in Wootton, and Cogges. These, with Wadard's lands in other counties, afterwards formed the barony of Arsic, of which Cogges was the head, but Wadard's chief distinction is that he is represented on the Bayeux Tapestry as a mounted knight, armed with shield and lance, under the legend 'Hic est Wadard'. Ilbert de Lacy, an important tenant in chief in Yorkshire, Lincolnshire, and Nottinghamshire, held of the bishop's fee in Tythrop, Stanton St. John, Shipton-on-Cherwell, Bampton, Chastleton, Lyneham, Ascot Earl, Little Tew, and Cassington. Before 1090, and probably before 1086, he gave two-thirds of his demesne tithes at Stanton St. John and Lyneham to the church of St. Clement which he founded in his castle of Pontefract. (fn. 30) In the 13th century the honour of Pontefract, of which most of Ilbert's Oxfordshire lands formed part, included also the manors of Little Haseley, Thomley, Britwell Baldwin, and Warpsgrove, held by a certain Hervey of Bishop Odo in 1086. The fact that all these manors were held in the 13th century by the Yorkshire baron William of Skelbrook identifies Hervey, Bishop Odo's tenant, with Hervey de Saio, who held Skelbrook of Ilbert in 1086, and gave tithe there to St. Clement's Church in Pontefract Castle. (fn. 31) Among the bishop's other tenants the Adam who held at Sandford St. Martin and Barton Ede has been definitely identified (fn. 32) with Adam, son of Hubert de Ryes, and brother of Eudo the king's dapifer, who had recently acted as one of the commissioners for the making of the Domesday inquisition in Worcestershire. It is safe to assume his identity with the bishop's tenant of the same name at South Newington, Baldon, Over Worton, and 'Saxintone'. Under Steeple Aston, where he had enfeoffed a sub-tenant on his holding, he is carefully described as Adam son of Hubert. Manorial descent proves that the Roger of Yarnton, Forest Hill, Woodperry, and Cowley was Roger d'Ivry. Robert d'Oilly, who is often associated with Roger elsewhere, held in Barford St. John and Baldon. The highest in rank of all the bishop's tenants, Aubrey de Coucy, who had recently resigned the earldom of Northumbria, only held of the bishop the single manor of Burford.

The four valuable manors of Combe, Deddington, Stanton Harcourt, and Great Tew were held by the bishop in demesne. At Tew, and probably also at Stanton Harcourt, he had succeeded one of the few English thegns who enter into the general history of the time. Alnot Chentisc, his predecessor there and at many places in other counties, is known to have been one of the English nobles who accompanied the Conqueror, as hostages rather than guests, on his first visit to Normandy after his coronation. (fn. 33) The history of Alnot's manor of Tew in the 11th century is known in unusual detail. In 1003 or 1004 it was bequeathed by Ælfric, Archbishop of Canterbury, to the abbey of St. Albans, (fn. 34) and in the 13th century it was still remembered there that the abbey's estate in Tew had passed to Odo of Bayeux. (fn. 35) Early in the Confessor's reign, the property, then known as 'Cyrictiwa', was the subject of an agreement between the abbot and a woman named Tova by which she and her son were to keep the land for their lives, so that afterwards it should come back to the abbey. (fn. 36) Among the witnesses to this document, Leofwine of Bærtune can safely be identified with the Leuuin who had preceded the Bishop of Lisieux at Westcot Barton. There is no evidence to connect either Tova or her son with Alnod Chentisc, but the agreement is interesting as an illustration of the complex histories which may lie behind a simple statement of Domesday Book about ownership in King Edward's time.

In 1086 three English and two French abbeys held land directly from the king in Oxfordshire. The abbey of Abingdon, which in Berkshire was the greatest landowner after the king, possessed few Oxfordshire manors. Lewknor, Cuddesdon, and the greater part of Tadmarton were kept in demesne. Among the tenants who held the remainder of the monastic property were the prominent barons Robert d'Oilly and Roger d'Ivry; a certain Gilbert, whom an early list of the abbot's knights (fn. 37) describes as Gilbert the Marshall; and an Englishman named Sueting, who reappears in the same list as 'Matthew's grandfather'. It is highly probable that the tenant named Wenric, who held of the abbot at Sandford-on-Thames in Oxfordshire and Chilton in Berkshire, was identical with Gueres de Palences, (fn. 38) whose name comes first in the list of the abbot's knights as the holder of four knights' fees in Sandford, Chilton, and other places. It is also probable that the 'son of Wadard' who held of Roger d'Ivry at Barford Cheney on the abbey's land and at Thrup on Roger's own honour was identical with the 'Walchelin Waard' who occurs on the Oxfordshire Pipe Roll of 1130, and whose daughter brought Barford Cheney by marriage to the family from which it derived its distinctive name. (fn. 39)

Of the remaining monastic properties of Oxfordshire the most interesting is the manor of Taynton, held by the abbey of St. Denis near Paris. Something of its history is told in a remarkable charter issued by William I in a council held at Winchester on 14 April 1069. (fn. 40) According to this document Edward the Confessor gave the church of Deerhust in Gloucestershire to a monk of St. Denis named Baldwin, who afterwards became Abbot of Bury St. Edmunds. (fn. 41) Taynton, according to the charter, was a separate gift, made by the Confessor directly to St. Denis, without reference to Baldwin. In Domesday Book, among the possessions of ecclesiastical landowners in Oxfordshire and Gloucestershire, the church of Deerhurst with its appurtenant lands, and the manor of Taynton, are each assigned to St. Denis, and no mention is made of any interest retained by Baldwin in either the church or the manor. But in the list of persons holding property in the borough of Oxford, 'the Abbot of St. Edmunds', that is, Baldwin himself, appears in possession of a dwelling belonging to Taynton. It would therefore seem that Taynton, like the church of Deerhurst, had originally been granted to Baldwin, and that after his election as Abbot of Bury he had annexed the property in Oxford which had belonged to Taynton to the possessions of the house of which he had become the head. (fn. 42)

Between the monastic landowners and the lay barons of Oxfordshire Domesday places under a single heading the canons of St. Frideswide's Minster, and four other clerks unconnected with that house. Three of these clerks seem to have been Englishmen; the fourth was Rannulf Flambard, afterwards Bishop of Durham, who in 1086 was, at most, an undistinguished member of the clerical staff in the king's service. The canons of St. Frideswide's were holding only a part of the land in Oxfordshire which had belonged to their predecessors earlier in the century. In 1004, King Æthelred II had confirmed to their church 10 hides at Winchendon in Buckinghamshire, which still remained to the minster, 3 hides at Whitehill, 3 hides at Cowley, 2 or 3 hides at Cutslow, and the tithe of Headington. (fn. 43) By 1086 the land at Whitehill had been lost, and the 3 hides at Cowley can only be traced among the possessions of the minster on the assumption that they were included among the 4 hides 'near to Oxford', which formed its chief possession within the county at that date. The land at Cutslow, held 'of the canons' by an Englishman named Seward, was probably a portion of their estate which had come to form a separate prebend for one of their number.

In Oxfordshire, as in other counties, the Domesday clerks gave precedence among lay barons to the men bearing the rank of a count in France or an earl in England. Among the Oxfordshire tenants in chief of this class, Earl Hugh of Chester held by far the most valuable estate. Most of his predecessors had been men of consequence; he had followed Archbishop Stigand at Pyrton, Earl Harold at Churchill, and Hugh, King Edward's chamberlain, at Tackley. Three, at least, of his tenants were greater men than would be gathered from the bare entry of their names in Domesday Book. The Walter who held Churchill was an important tenant of the earl in Cheshire, and brother of William de Vernon, one of the chief barons of the Chester fee. The Robert who held of the earl in South Weston and Ardley was Robert d'Oilly, and the William who held Pyrton was William fitz Nigel, constable of Chester, who gave the church of Pyrton to the canons of his own establishment at Runcorn. The fact that the Drogo who held of Robert d'Oilly at Ardley on the earl's fee is known to be identical with Drogo de Andelei, the tenant of Sherburn and Hardwick on Robert's own fee, is a good illustration of the intricacies of tenure which often connected different honours in the 11th century. (fn. 44)

Among the fiefs of Oxfordshire barons below the rank of earl three are peculiarly connected with the history of the county. The lands of Robert d'Oilly, Roger d'Ivry, and Milo Crispin occupy more than a fifth of the space given by the Domesday clerks to the whole shire. Each of these barons held much land in other counties, but it was in Oxfordshire that the feudal influence of each was chiefly concentrated. Each of the three was sufficiently in the king's confidence to be entrusted with an important military command. Milo Crispin, who belonged to the ancient aristocracy of Normandy, (fn. 45) was set in charge of Wallingford Castle. Robert d'Oilly, who was a royal constable, and had been much employed by the king in administrative business, (fn. 46) was the first castellan of Oxford. Roger d'Ivry, who held the office of butler in the king's court, is not associated with any English castle, but the tower of Rouen, in some ways the most important of Norman fortresses, is known to have been in his custody in 1078. (fn. 47) Apart from the familiarity which must have connected any group of barons thus trusted by the king, there are traces of a more special connexion between these three magnates. In Oxfordshire, as elsewhere, Robert d'Oilly and Milo Crispin shared the lands which in 1066 had belonged to Wigod of Wallingford, an important thegn whom Edward the Confessor recognized as a kinsman, and whose son had died in the Conqueror's service at the battle of Gerberoi. In 1183 a jury of Oxfordshire knights committed itself to the statement that Milo Crispin married Maud daughter of Robert d'Oilly. (fn. 48) It is probable that this marriage, which there is no reason to question, explains the transference of a considerable portion of Robert's fee to the honour of Wallingford, which otherwise represented the fief of Milo Crispin. There is evidence of a still closer connexion between Robert d'Oilly and Roger d'Ivry. According to a tradition preserved in Oseney Abbey (fn. 49) Robert and Roger were 'sworn brothers, united by faith and oath'. There seems to be no evidence that the relationship of sworn brotherhood was formed between any other of the Conqueror's barons, but the tradition was recorded before the history of the Norman Conquest had become romanticized, and it is pointless as an invention. That Robert d'Oilly and Roger d'Ivry were especially intimate is proved by the remarkable fact that Roger joined with Robert in the foundation of the collegiate church of St. George within Robert's castle of Oxford. (fn. 50) There is no need to doubt the tradition that they had confirmed their friendship by a formal ceremony.

Apart from the history of their fees there is little to be said about the individual tenants of these three lords. Peter, who held Wheatfield and a manor in Lewknor of Robert d'Oilly, is of interest as the ancestor in the male line of Robert de Witefeld, whom Henry II frequently employed as a judge and administrator. (fn. 51) It is safe to identify the Euruin who held a manor in Little Tew of Robert d'Oilly with the man of that name who held other land in that place of Robert of Stafford. The identification gives a good example of the tenure of land by the same individual on different honours, a practice which complicated even the earliest phases of English military feudalism. (fn. 52) The tenure of many manors on the same honour by a single person is well illustrated by the identification of the William who held of Milo Crispin at Chesterton, Adwell, and Henton near Chinnor in Oxfordshire with the William who held in Sulham and Betterton in Berkshire and Bradwell in Buckinghamshire of the same lord. (fn. 53) A tenant holding on this scale can certainly be regarded as a forerunner of the 'barons of the court' of Wallingford, who are mentioned collectively before the end of the 12th century. (fn. 54) The frequent identification of groups of manors held in 1086 by the same tenant with groups of knights' fees held together in the 12th or 13th century proves that most of the tenants mentioned in the Survey were holding their lands by military service. But on most of the largest honours there were manors held by Englishmen who are very unlikely to have been knights themselves, and whose contribution to the military burden laid on the lord of the honour probably amounted, at most, to a money payment towards the occasional expenses of a knight's equipment. Some Englishmen were holding of Norman lords land which had been their own property before the Conquest. On the Oxfordshire fief of Milo Crispin a certain Toli was holding land at Cowley which he had held freely in 1066, and at Berwick Salome and Gangsdown an Englishman named Orgar was holding land of Milo which the local jurors considered should have been held of the king, 'for Orgar and his father and his uncle had held it freely in King Edward's time'.

According to the list of Oxfordshire landowners prefixed to the county survey and followed in its text, forty-three lay tenants in chief of some importance held land within the shire. One deduction should perhaps be made from this list, for 'Hunesworde', the only Oxfordshire manor assigned to William fitz Ansculf, has generally been identified with Handsworth in Staffordshire. (fn. 55) On the other hand, the manors in Cottisford, Sibford Gower, Charlton-on-Otmoor, and Shipton-on-Cherwell, held in chief by Hugh de Grentemaisnil, are described under Northamptonshire, and the manor in Sibford Gower, similarly held by William son of Corbucion, is described under Staffordshire. No reference is made to either of these barons in the Oxfordshire survey. The existence of so many tenancies in chief in a county of moderate size meant that individually most of them were small, and with few exceptions they formed isolated portions of honours centred in other shires. In later records, the Oxfordshire lands of Henry de Ferrars, Walter Giffard, Rannulf Peverel, Guy de Reinbuedcurth, Ralf de Mortimer, and Hascolf Musard appear as fragments of fiefs of which the respective capita were Tutbury in Staffordshire, Long Crendon in Buckinghamshire, Hatfield Peverel in Essex, Chipping Warden in Northamptonshire, Wigmore in Herefordshire, and Staveley in Derbyshire. Berenger de Todeni, who held a small fief in north Oxfordshire, was the son of Robert de Todeni, the founder of Belvoir castle; (fn. 56) Gilbert de Gand, whose fief included land at Handborough and Ewelme, was among the greatest magnates of Lincolnshire; and William de Warenne, holding only two Oxfordshire manors in chief, was lord of the rape of Lewes, of a vast fee in Norfolk, and of a very considerable estate in south Yorkshire. The interests of the smaller tenants in chief were dispersed in the same way. William son of Manne, who held part of Arncot in Oxfordshire, held single manors of the king in Hampshire and Buckinghamshire, and a manor in Sussex of William de Braose. (fn. 57) Naturally there is little record of any personal connexion between the lord of one of these scattered fiefs and the individual estates which he kept in demesne. Ernulf de Hesdin seems to be the only Oxfordshire tenant in chief who can be seen in residence at one of his manors within the county. It was at his house at Chipping Norton, in the presence of his wife and daughter, his sons and the knights of his household, that he confirmed to the priory of St. George at Hesdin all that he held in that place of Ingelram, the local count. (fn. 58)

It is always difficult to obtain contemporary information about the tenants holding the various manors of which an 11th-century honour was composed. Most of them appear in Domesday Book under their Christian names alone, and very few of them can be identified in other documents. Now and then an unusual combination of references tells something about an under-tenant's position in feudal society. The Robert who held in Ewelme of Gilbert de Gand is obviously identical with the Robert 'Armenteres' who appears in the Berkshire Domesday as the owner of a house in Wallingford belonging to this manor, (fn. 59) and with the Robert de Armenteres who attested a charter of Gilbert de Gand in favour of Abingdon abbey. (fn. 60) He is the first recorded member of a family which in the 12th century was very important in Lincolnshire and Northamptonshire. (fn. 61) But in general, the position of the under-tenants of 1086 can only be inferred from the facts which can be discovered about the descent of their holdings, and continuous evidence for the descent of a manor rarely begins before the middle of the 12th century. At present the detailed history of most of the great 11thcentury honours is still unwritten. Nevertheless, it is already clear that the undertenants named in Domesday Book cannot be regarded as the members of a homogeneous social class. Even if men of English birth are left out of the reckoning, the under-tenants of 1086 ranged in social position between the knight who had received a single fee or less from his lord and the baron who held a large fief from the king. The Hugh who held seven out of the ten manors of Walter Giffard's fee in Oxfordshire was Hugh de Bolbec, himself a tenant in chief in three counties, of which Oxfordshire was one. In the aggregate, the number of under-tenancies held by men of this class was very considerable. (fn. 62) More obscure, but of great significance in the structure of feudal society, were the innumerable under-tenancies in the hands of men who held no land in chief of the king, but ranked as barons in the social scale by virtue of the position which they held at their lord's court and in his honour. The honour of Wallingford, which has recently been analysed in detail, (fn. 63) included many tenants of this type. Anschetil de Grai, who held seven manors on the honour which had belonged to Earl William of Hereford, was clearly much more than a simple knight. The tenant named Oidelard, who held Idbury under Ralf de Mortimer, held under the same lord at Preston Candover, Knowle, and Hordle in Hampshire, Peasemore and Hodcott in Berkshire, Tockenham in Wiltshire, Downton Castle in Herefordshire, and Buckton in Shropshire. (fn. 64) Men of this type are the predecessors of the honorial barons who appear in 12th-century records, and their lands were rarely concentrated in any single part of their lord's fief.

It is not surprising that it should be impossible to generalize about the status of under-tenants, for the tenants in chief of the Crown were themselves a highly miscellaneous company. There is little that can be said in general terms even about the tenants in chief whose lands receive a separate rubrication in Domesday Book. Most of them were territorial magnates, holding their lands by military service. Some of them, such as Guy, brother of Robert d'Oilly, and Ilbod, brother of Ernulf de Hesdin, clearly owed their position to their relationship to greater men. (fn. 65) In all counties, the compilers of Domesday Book tend to give the first place in the county survey to the greatest of the men with whose lands they were dealing. But they paid little attention to questions of precedence among tenants in chief below the rank of earl, and it is always dangerous to draw conclusions as to a tenant's importance merely from the place which he receives in the schedule of landowners within a shire. Turchil, the fifty-seventh among the numbered tenants in chief in Oxfordshire, was Thurkill of Arden, whose great estates in Warwickshire formed a substantial part of the fee with which the earldom of Warwick was endowed by William II. Many of the lesser tenants in chief whose fees are described separately can be shown to have held some specific office in the royal household. Hugh d'Ivry and Richard de Courcy, whose estates in England were small, were officers in the Conqueror's Norman court. (fn. 66) Reinald, whose Oxfordshire holding consisted merely of a hide at Boycot, had been the chief among the Confessor's clerks, and is once described in Domesday Book as his chancellor. But there is no trace in the Oxfordshire Domesday of any feeling that a separate rubrication should be reserved for the lands of the greatest household officers. From the subsequent history of their fees it would seem that Benzelin, who was counted separately as a tenant in chief in virtue of a manor in Lillingstone Lovel, was an usher at court, and that Robert, son of Murdrac, who was similarly distinguished as the holder of Broughton Poggs, was one of the king's falconers. Neither of these men can have been as important in the circle of household officers as Richard Ingania, who only appears in the Oxfordshire Domesday in a miscellaneous group of royal ministri, but, as one of the king's chief huntsmen, held lands in Northamptonshire (fn. 67) and elsewhere, which were regarded as a barony in the 13th century.

Three ladies of high rank follow Benzelin in the list of Oxfordshire tenants in chief. The Countess Judith, Earl Waltheof's widow, was daughter of Lambert, Count of Lens, and Adelaide, daughter of Arlette of Falaise, King William's mother. The lady who is described as 'Cristina' was of even higher rank. She was sister of Edgar the Ætheling and grand-daughter of King Edmund Ironside; in 1086 she was living as a nun at Romsey. (fn. 68) The third lady, who appears in Domesday Book merely as 'the wife of Roger d'Ivry', was Adelina, daughter of Hugh de Grentemaisnil. (fn. 69) Her estate, which consisted of the manors of Islip and Oddington, was said to be held of the king in commendatione. The word commendatio frequently occurs in the surveys of the eastern counties, where it generally denotes the relationship between a powerful lord and a smaller man who has placed himself under his protection. It cannot bear its usual sense in this passage, but it clearly denoted some form of tenure less permanent than that which connected the normal tenant in chief with the king. Its nature can only be conjectured, but it is at least certain that the land did not remain with the descendants of Roger d'Ivry or of his wife. The land at Islip had been granted by Edward the Confessor to Westminster Abbey, and it afterwards appears, together with the land at Oddington, in the possession of the abbey. As the Confessor's grant of Islip was made at the dedication of the abbey church on 28 December 1065, the abbey cannot well have been in seisin of the estate when the Confessor died on 5 January 1066. (fn. 70) In strictness of law, the land was therefore at King William's disposal, and it is on the whole probable that Adelina's tenure in commendatione meant that she was regarded as its temporary holder, pending a further investigation of the claims of Westminster Abbey. (fn. 71)

In most counties the compilers of Domesday Book brought together under one heading the lands of a number of persons owing some form of personal service to the king. In Oxfordshire the list is headed Terra Ricardi et aliorum Ministrorum Regis. Now and then Domesday Book itself specifies the form of service which the tenant rendered. The Hervey who held at Ibstone and Bix Brand is described as Hervey 'legatus' in the Buckinghamshire survey, and can therefore be identified as a king's messenger. Rainald 'the archer' appears at Ibstone and Chadlington, and Siward 'the huntsman' at Chadlington alone. Siward had held the same land before the Conquest, and there can be little doubt that he had served Edward the Confessor in the same capacity. Theodoric the goldsmith, who held at Brize Norton, Weld near Bampton, and Bensington, is another link with the group of men retained in King Edward's service. He held land in Berkshire and Surrey as well as in Oxfordshire, and his estate at Kennington, in Surrey, had belonged to him in 1066. In some cases the position held by a minister at court, though not defined in Domesday Book, can be inferred from the subsequent association of his lands with a particular household office. Robert son of Thurstan, of Great Rollright and Ludwell in Wootton, founded a family of which the successive heads, for two centuries, held Rollright by the service of acting as a dispenser at court. (fn. 72) The descent of Great Rollright is unusually clear, but there are other cases in which the service rendered for a manor in the 13th century gives a clue as to the place of its 11th-century holder among the king's servants. Under Henry III, Alvescote and Middle Aston formed part of a group of manors held by the service of being usher at the door of the king's hall. (fn. 73) In 1086, all but one of these manors were held by a minister, apparently of English birth, named Saric. In 1236, four yardlands in Lew, which clearly represent a part of the manor held there by Aretius the king's minister in 1086, were held by the serjeanty of keeping a falcon in season. (fn. 74) It is often difficult to trace the serjeanties recorded in the 13th century back into the Norman period. Many ancient serjeanties seem to have been commuted into other forms of tenure before the great feodaries of the reigns of John and Henry III were compiled. But there is no doubt of the frequent persistence of arrangements which the Conqueror had made for the organization of his household or the maintenance of his sport.

The last section of the Oxfordshire survey deals with the lands within the county which had been held by William fitz Osbern, Earl of Hereford. The earl had played a very important part in the reduction and settlement of the country in the critical period following King William's coronation. But early in 1071 he was killed in Flanders, and his lands passed to Roger, his second son, who lost them in consequence of an irresponsible rebellion in 1075. In most parts of England, the Fitz Osbern fee was incorporated into the royal demesne, (fn. 75) or granted out to different tenants in chief, and it is only in Oxfordshire that it is described under a separate rubrication. Most of the tenants holding of the fee in 1086 were men of considerable importance. Anschetil de Grai has already been mentioned. Gilbert de Breteuile, who heads the list, derived his name from the famous bourg of Breteuil, the head of Earl William's fief in Normandy. In Oxfordshire he only held the manors of Bolney and Sydenham in farm from the king, but he was recognized as a tenant in chief in Hampshire, Berkshire, and Wiltshire, and his lands in these counties formed an important honour. (fn. 76) Roger de Laci, who held at Begbrook, Kiddington, Mongewell, and Salford, was the greatest tenant in chief in Herefordshire. The Robert who held of the earl's fee in Ducklington was clearly Robert d'Oilly, for the remainder of the village formed part of Robert's own barony, and the whole village was afterwards held by his representatives. Roger d'Ivry is expressly named as the holder of 2 hides in Black Bourton. But in some ways the most interesting of the tenants on Earl William's fee is the Aluui who held a hide at Miltonunder-Wychwood under a certain Roger who does not seem to have been identified. Towards the year 1200, (fn. 77) Robert of Astrop granted to the monks of Bruern all the land of Milton which belonged to Alewi his grandfather, the son of Eilsi of Faringdon. In the Oxfordshire Domesday, Eilsi appears as the holder of land in Langford and Shipton-under-Wychwood at farm from the king, and land at Rycot and Shipton-under-Wychwood as a king's minister. But he clearly belonged to the generation which was disappearing at the time when Domesday Book was compiled; his name connects him with the royal manor of Faringdon in Berkshire, and he can safely be regarded as an Englishman who had been in the king's service before the Conquest. (fn. 78) There are very few cases in the whole of England in which an estate can be shown to have passed by direct descent from a man of this class to a landowner of some local consequence in the 13th century.

In the 11th century, the normal borough, whatever its other functions in the life of the state, was essentially a centre to which goods were brought for distribution over the surrounding country. Already before the Conquest, the desirability of access to the borough market had led many country landowners to acquire houses within the borough. These houses were often annexed to rural manors, and passed with them to their new French lords. In most mid land counties there was only one borough, and, as a rule, the manors to which burghal houses belonged were all situated within the county of which economically it was the capital. But in the valley of the middle Thames, the twin boroughs of Oxford and Wallingford, 14 miles apart, attracted lines of traffic which disregarded county boundaries, so that many Oxfordshire landholders held property in Wallingford, (fn. 79) and some, if not many, Berkshire landholders can be shown to have held property in Oxford. In 1086 a single haga, or messuage, in Oxford was attached to Geoffrey de Mandeville's Berkshire manor of Streatley; in 1066 thirteen hagae in Oxford had been attached to Earl Harold's manor of Steventon near Didcot; and some thirty years before the Conquest, a certain Æthelwine bequeathed to Abingdon abbey the village of Lyford in Berkshire with the haga in Oxford in which he himself dwelt. (fn. 80)

There is no doubt that before the Conquest Oxford had been a place of more than local importance. National assemblies had been held there in 1018, 1036, and 1065, and the number of moneyers who had worked there under Edward the Confessor indicates a considerable population of traders. (fn. 81) At the middle of the 12th century it seems to have been the most prosperous town in the whole of the Midlands. (fn. 82) But in 1086 it was in decay. The description of the city begins with a general statement to the effect that out of 721 houses which then stood within and without its walls, 478 were in such ruin that nothing could be obtained from them when the king took a geld. There is no record of any catastrophe sufficient to account for this decline. No siege of the town is mentioned in any authentic account of the wars by which the Norman rule in England was secured, (fn. 83) and the building of the castle, to which some of the decay should doubtless be attributed, affected only a small portion of one of the four quarters into which Oxford was divided by its four principal streets. At Wallingford the building of a castle larger than that of Oxford, only obliterated eight hagae. (fn. 84)

The detailed description of the city falls into clearly marked sections. The first section assigns 217 mansiones, or messuages, to twenty-nine owners, ranging downwards from the king to a person called the son of Manasse, who possessed a single mansio attached to an estate at Bletchington. (fn. 85) Most of them were important tenants in chief, and all of them are known to have held, or at least possessed, a claim to land elsewhere in the shire. At the end of the list there is inserted a statement to the effect that all the mansiones enumerated in it were 'free', propter reparationem muri, and this statement is followed by a sentence asserting that in King Edward's time all the mansiones which were called 'mural' were free from every custom except the duty of repairing the wall. The 'customs' from which these mansiones were exempt are not defined, but Dr. Salter has recently shown (fn. 86) good reason for believing that the tenements which repaired the wall did not contribute to the gelds laid on the city. If so, it follows that the 'mural mansions' of Oxford were not included among the 721 geldable tenements which have previously been mentioned. In any case, it is at least clear that at Oxford the burdens which fell upon the men of the city were differentiated after a fashion for which there is little evidence elsewhere, and that responsibility for the repair of the wall lay only upon a part, if a large part, of the tenements which it inclosed.

The second section relates to the holdings of men who, apart from the priests of St. Michael's and the canons of St. Frideswide's, may be described collectively as burgesses of Oxford. It includes 80 tenements in all, of which 67 are described as mansiones and 13 as domus. In such a context mansio usually means an inclosed plot of land within a borough, (fn. 87) in contrast to a mere house, but it is very doubtful whether the clerks who wrote the description of Oxford paid any regard to this distinction. Towards the end of the list they note that certain of these tenements were free from custom because they contributed to the charge of repairing the wall, and they finally remark that if the holder of such a 'house' failed to repair the wall when repair was needed, he paid 40s. to the king or else forfeited the house. It is possible, though it cannot be proved, that all the tenements in this as in the former section were subject to this burden, and that the detailed list of tenements in the Oxford Domesday is in reality a list of what the clerks called mansiones murales. (fn. 88) If so, the total number of these mural mansions will amount to 297, and the total number of tenements assigned by Domesday Book to Oxford will amount to 1,018–a figure slightly in excess of the number of houses standing within and without the walls in 1279. In any case, it is certain that the detailed enumeration of properties within the city is incomplete. It contains no reference to the haga in Oxford attached to the Berkshire manor of Streatley, nor to the 13 hagae attached to the king's manor of Steventon in that county. It is more remarkable that this enumeration omits no fewer than 42 inhabited houses within and without the wall which belonged to 'the benefice of St. Peter', and are entered under the fee of Robert d'Oilly. These omissions show that the description of Oxford in Domesday Book cannot be regarded as in any sense an exhaustive survey of the borough, and to some extent, though by no means conclusively, they support the opinion that the tenements enumerated in detail were only those which contributed to the repair of the borough wall.

Like most boroughs, Oxford in 1086 was carrying a fiscal burden much heavier than that which had lain on it before the Conquest. In King Edward's time it had produced a yearly 'farm' of £20 and 6 sestiers of honey to the king, and £10 to Earl Ælfgar of Mercia. By 1086 the farm had been increased to £60 by tale in pennies of twenty to the ounce of silver. (fn. 89) The pre-Conquest farm is described as a commutation for toll and rent and all other customs. Whatever may be the explanation of the decay shown by the Domesday description of the borough, it seems clear that the sources of revenue in respect of which the farm was levied must have been shrinking in the years during which its amount was increased. The 'customs' from which the borough obtained quittance by its farm are not described. Apart from the reiterated assertion of the responsibility of certain tenements for the repair of the wall, the only reference in Domesday to the ancient public duties of the burgesses of Oxford is a statement that when the king went to war, they used to send 20 of their number with him, or else pay £20 in order that all of them might be exempt. There is no trace of any service to the king in time of peace such as was rendered by the burgesses of Wallingford, who provided horses and boats when he needed them. In view of the close resemblance between Oxford and Wallingford in the 11th century, it is not unlikely that duties of this kind, formerly demanded from the burgesses of Oxford, were among the consuetudines covered by the farm which they rendered to the king.

In Oxfordshire, as in most, if not all, other counties, Domesday Book records a number of place-names which cannot now be identified. Owing to the work of Dr. Salter on the early medieval charters of Oxfordshire, the number is small. Robert of Stafford's manor of Bumerescote and Pismanescote is known to have been situated in the parish of Alvescot, (fn. 90) but neither of these names seems to have come down to modern times. The most interesting of these vanished names is Adingeham, which covers a small manor on the fee of Earl William fitz Osbern. If, as its spelling suggests, this name represented an Old English Addingaham, its disappearance would be remarkable, for most names of this type denoted ancient and important places. It is far more probable that the first syllable of the name stood for the Old English preposition aet, and that the real name of the place was 'Ingham'. There are two references to a place of this name in early Oxfordshire documents. In the 10th century boundaries of an estate described as lying in Britwell and Watlington, (fn. 91) the boundary line is traced to Cuces hæma gemære, thæt to Incghæma gemære, a description which indicates that Ingham lay to the south of Cuxham. The second document is the agreement, already quoted, (fn. 92) between St. Albans Abbey and Tova the widow about Great Tew, which is witnessed by Alwinus de Ingham. The name is not found on any map, but within recent years an estate at Watlington was still known as the manor of Ingham and Watcombe.

No very difficult problems arise in regard to the various orders of society mentioned in the Oxfordshire survey. On manor after manor, villeins, bordars, and serfs occur in varying, but rarely abnormal, proportions. A group of 17 'boors' (buri) on the king's manor of Bampton is interesting as a survival of pre-Conquest manorial organization. The Old English gebur was a tenant who had received both stock and land from his lord, and was burdened with heavier services than were demanded from the higher ranks of the manorial peasantry. In Domesday Book men of this class are generally included in the heterogeneous body of villani, and their appearance at Bampton is probably due to the conservatism of the local jurors who represented this ancient royal estate. At the other end of the peasant scale 23 liberi homines are mentioned in the Oxfordshire Domesday; of whom 15 lived at Aston Rowant, 4 at Pyrton, and 4 at Church Enstone. No hint as to their condition is given by the Survey, but they should probably be regarded as ceorls who had maintained their preConquest independence with sufficient tenacity to escape absorption into the manorial routine. It is also probable that men of this type were present on many estates where no separate record was made of their numbers. A solitary burgensis who appears on Roger d'Ivry's manor of Whitehill was probably a burgess of Oxford whose rent was regarded as part of the profit for which the manorial officer in charge of Whitehill accounted each year to his lord.

As a source of statistical information, the Oxfordshire Domesday is an unsatisfactory text. It presents the difficulties which are common to every portion of the Survey, such as the frequent uncertainty whether the details given in an appendix to the description of a manor are a mere expansion of its information or relate to a separate property. But the uncertain points in the Oxfordshire Domesday are unusually numerous. It is not easy to explain the inclusion of many Oxfordshire villages in the surveys of other counties. Most of them were undoubtedly part of Oxfordshire in 1086, and were described under other counties because of some confusion in the distribution of the original returns to the Domesday Inquest among the clerks who wrote Domesday Book. But some manors, such as Caversfield, which were situated in Oxfordshire, seem to have belonged administratively to the counties under which they were surveyed, and it is possible that even the extraordinary division of Mollington between Oxfordshire, Northamptonshire, and Warwickshire represents a complicated piece of administrative geography. As the Oxfordshire hundreds of the 11th century are unknown in detail, it is impossible to trace the lines along which villages were combined for purposes of assessment. It is a more serious difficulty that the section devoted to the ancient estates of the Bishop of Lincoln arranges information relating to a large number of scattered villages under no more than five manorial headings. (fn. 93) Any attempt to reconstruct the distribution of the Oxfordshire population in 1086 is attended by such complications as the probability that Charlbury and its hamlets were described under Banbury, and that Epwell was described under Dorchester. (fn. 94) There are, moreover, villages of great age which Domesday Book seems to ignore. Culham, which had been in the possession of Abingdon Abbey since the 10th, if not since the 8th century, is unsurveyed, and no manor in the Oxfordshire or any other portion of the Survey has yet been identified with Souldern. Under these conditions, the figures which follow are only offered as a tentative indication of the general relationship between assessment, (fn. 95) agricultural equipment, and population within the county as a whole.

Manorial Values: (fn. 96) In 1066, £1,983. 18s.

In 1086, £2,434. 8s. 6d.

The Oxfordshire Domesday is distinguished from the surveys of neighbouring counties by frequent reference to the 'inland' of a manor. Already in Edgar's laws the word occurs in a technical sense in a passage which orders that tithe should be paid 'both from the thegn's inland and from geneatland'. (fn. 97) A twelfth-century translator of this passage renders thegnes inland by 'dominium liberalis hominis', and geneatland by 'terra villanorum'. From many entries in the Oxfordshire Domesday, as from Edgar's law, it is clear that originally inland meant demesne, and there are passages which show that sometimes the terms could be used indifferently. In the description of Robert d'Oilly's fee, the account of Watlington begins Ibi sunt viii hidae . . . De hac terra iii hidae sunt inland, and that of Bicester, Ibi sunt xv hidae et dimidia . . . De hac terra iii hidae sunt in dominio. There are, however, entries in which inland and demesne are clearly distinguished. At Deddington, for example, it is stated that there were 11½ hides in demesne 'apart from the inland'. In such cases it seems evident that the inland was the ancient desmesne of the manor, and that the additional land held in dominio had been annexed by the lord at some later date from the holdings of the manorial peasantry. In the description of Brize Norton there is a definite reference to this kind of annexation. Fulk, the tenant of Roger d'Ivry, is said to have 5 hides in dominio, de terra villanorum. The care with which the extension of the inland was recorded on manors such as Deddington was due to the fact that before the Conquest the lands which formed a thegn's inland had normally, if not universally, been exempt from national taxes such as the Danegeld. (fn. 98) According to the political theory of the time, a thegn's service in war was considered to acquit the lands of his own cultivation from these public burdens. But the enlargement of the inland at the expense of the local peasant holdings meant, or at least threatened, the reduction of the area from which gelds could be taken, and the Domesday clerks, in noting cases where this enlargement had occurred, were recording manorial arrangements which directly affected the king's financial interests.

In contrast to inland, the lord's demesne, exempt from geld, early Norman records sometimes use the word warland to denote the peasant holdings on which this burden fell. The word, which is one of many compounds formed from the Old English waru, 'defence', is much rarer than inland, and few examples of it occur in Domesday Book. In the Oxfordshire survey it is used in the description of a manor in Cowley held by the king's English minister Lewin. The passage runs: Ibi sunt iiii hidae et dimidia. Terra x carucis. Ibi i hida de warland in dominio. It is clear that there is nothing but a difference of phrasing between this passage and words which state less concisely that a lord has taken so much land into demesne de terra villanorum. In either case, the point which the Domesday clerks wished to make was the possible reduction of the geldable area of a manor by the addition of peasant holdings to an exempt demesne.

In connexion with a piece of Oxfordshire inland, Domesday Book makes one of its rare explicit references to the open-field system of agriculture. In 1086, Abingdon Abbey possessed 10 hides in Garsington, and one of them is described as a hide of inland, which never paid geld, lying in parcels (particulatim) among the king's land. The passage is of especial interest as showing a piece of monastic demesne intermingled, not with the lands of manorial tenants, but with the lands of the king himself. The royal estate to which these lands belonged cannot be identified in Domesday Book, but in 1122 Henry I gave demesne tithes in Garsington to the canons of St. Frideswide, (fn. 99) and it is possible that the manor of Headington included outlying property in that village.

In some counties, such as Worcestershire and Berkshire, it is often possible to penetrate behind the impersonal uniformity of Domesday Book to the clash of interests out of which the tenurial conditions of 1086 arose. In the Oxfordshire survey little is said about pleas relating to land within the shire or about the personal history of its leading men. Now and then some other record shows that a simple entry in Domesday Book represents a mere phase in a complicated story. The Survey states, for example, that Richard de Courcy was holding a manor at Nuneham Courtenay which Hacon had held in 1066. Nothing is said to suggest that there was anything unusual in the history of the manor between the Confessor's death and the taking of the Domesday Inquest. But the 12th-century history of Abingdon Abbey relates that an English nobleman named Leofwine sold the manor, which was part of his inheritance, to Abbot Athelhelm at a time when the king was out of England. The Bishop of Bayeux, who was then governing the country, confirmed the sale. But when the bishop had fallen into disgrace, the king revoked the confirmation as if it had been obtained out of disrespect to himself, and gave the land to another person, who is unnamed, but can only have been Richard de Courcy. (fn. 100) There are other cases in which it is clear that the Domesday clerks have ignored contentious matter. A 12th-century list of the lands of which the Bishop of Bayeux had deprived Evesham Abbey (fn. 101) includes 5 hides at Salford, 3 hides at Cornwell, 6 hides at Chastleton, 4 hides at Dornford, and 5 hides at Shipton-on-Cherwell. According to the list, the bishop gave all these lands to a knight named Nigel. There is no hint in the Oxfordshire survey that Evesham Abbey had any claim to land within the county. Land at Salford, Chastleton, and Shipton-on-Cherwell is assigned to the bishop's fee, though no tenant named Nigel held any part of it, Dornford is not mentioned by name in Domesday, and the only manor in Cornwell of which the Survey takes notice belonged, not to the bishop, but to the fee of Earl William fitz Osbern. That the monks of Evesham exaggerated their wrongs is highly probable, but there is some evidence to support their tale of losses in other counties, and the list of the estates which the bishop had annexed in Oxfordshire is too precise to be a mere invention.

In spite of their somewhat narrow view of what was relevant to the purpose of the Domesday Inquest, the clerks responsible for the Oxfordshire survey preserved two pieces of unusually interesting information. The statement that the king gave the manor of Ludwell in Wootton to Robert d'Oilly at the siege of St. Suzanne is one of the very few references in the whole of Domesday Book to the obscure and often unsuccessful wars of the Conqueror's last years. It suggests that the creation of an honour by the king for a tenant in chief may often have been a very gradual process. Robert d'Oilly was in attendance at court, (fn. 102) and had probably received the nucleus of his fief, as early as 1067, but it was not until 1083 that the king besieged Hubert de Beaumont in St. Suzanne. (fn. 103) The second of these exceptional passages relates to the confused period between King William's coronation and the outbreak of the last English revolt in 1069. Itruns: 'From the time of Earl Tosti, a thegn named Saiet dwelt in Bloxham, and did service like a free man. Earl Edwin gave him to Ralf d'Oilly, and "R." d'Oilly drew him back into the king's demesne.' The story is interesting because it reflects the brief period in which English magnates and individual followers of the Conqueror were on friendly, or at least peaceful, terms. The transference of a thegn's service from one lord to another was a normal, and probably a common, incident in Old English social life. The outstanding feature of the Bloxham case is the surrender of a thegn by the last earl of the Mercians to a man bearing one of the most famous surnames in English feudal history. (fn. 104)