Domesday Survey: Introduction I

DOMESDAY SURVEY

I. The Assessment—the Hundreds—manorialization of the shire, pp. 80–88. II. Ploughs and teamlands—ploughs on demesne and ploughs among the men—holders of the men's teams, pp. 88–90. III. Classes of peasantry in 1086—francigene, milites and Anglici—villani—bordarii and cotarii—servi, pp. 90–94. IV. Meadow—mills—eels and fishponds—woodland—herbage—vineyards, pp. 95–98. V. Royal demesne before and after 1066—secular estates in 1066—ecclesiastical estates—secular estates in 1086, pp. 98–118.

I

The Domesday Survey of Middlesex, (fn. 1) a small county, occupies only 4½ folios of the larger of the two volumes (fn. 2) which contain the final abstract of the greater part of the information obtained at the inquiry. London is not included, which is surprising. It is possible that the two nearly blank folios which precede the Middlesex entry were to have contained the London Survey. (fn. 3) The available evidence strongly suggests that the material collected by the Domesday commissioners was first set down according to hundreds, and that afterwards the returns for each shire were rearranged on the basis of fiefs, and abridged. For the purposes of the visitation Middlesex appears to have been grouped with the counties of Hertfordshire and Buckinghamshire, and it is possible that Cambridgeshire and Bedfordshire belong to the same circuit (fn. 4) for in all of these counties the method of estimating woodland in terms of swine is employed instead of the more usual reckoning by length and breadth or area measurement. The Middlesex Survey contains only two allusions to the taking of evidence by the commissioners. Both are references, common enough in Domesday, to the testimony of the hundred. The first relates to the seizure of 53 acres in Stepney (85) by Hugh de Berneres. (fn. 5) In the other it is stated that the whole hundred of Spelthorne supported the assertion of Geoffrey de Mandeville's men that he had been possessed of a small estate (98) which Alveve, the wife of Wateman, held in 1086. (fn. 6)

Parts of the Middlesex Survey are not very clearly written and some readings are uncertain, but it has obviously been compiled with care. The hundred to which the estate belongs is almost always given in a rubric and the entries, with very few exceptions, conform to the same orderly pattern throughout. First is given the assessment in hides and the number of ploughlands in the estate (pro a hidis se defendit, terra est b carucarum), followed by the hides and the plough-teams (ibi sunt c caruce) in demesne. The number of plough-teams among the men is then stated, after which come details of the men's holdings. Information concerning woodland, pasture, weirs, mills, and other matters is then recorded, followed by a statement concerning the value of the property. This is invariably given for the periods mentioned in the king's writ preserved in the Inquisitio Eliensis, (fn. 7) namely the value in 1086, at the date when the present holder received it, and in the time of King Edward. (fn. 8) Finally are recorded pre-Conquest holders of the land, by name or status, and sometimes other details relating to ownership. In statements concerning the pre-Conquest ownership, instead of the more usual expression T(empore) R(egis) E(duuardi), in a few instances the formulae die qua rex Eduuardus fuit vivus et mortuus (fn. 9) and, once only (13), die qua rex Eduuardus obiit are used. In entries concerning ecclesiastical holdings, where there is continuous occupation by the same church, the expressions fuit et est in dominio ecclesie (fn. 10) and iacuit et iacet in ecclesia (fn. 11) frequently occur. The compilers of the Middlesex Survey are also careful to record, in the case of homines and sokemen, whether the holder of the land was free or not free to take his land to another lord.

The calculation of the total assessment resting on a county as small as Middlesex presents little difficulty. There are at the most three instances of lands in dispute and there appear to be no duplicate entries. Although it is not specifically stated in the entry cited above (fn. 12) (98) that Geoffrey de Mandeville claimed ²/3 hide, a dispute seems to be implied. The other two examples both relate to Stepney, the greater part of which was held by the Bishop of London. The bishop claimed Robert Fafiton's 4 hides (85) there which had belonged in 1066 to Sired, Canon of St. Paul's, and 3½ hides there held by Robert fitz Roscelin (87). Included in Robert Fafiton's manor, it may be noted, are the 53 acres which, the hundred declared, had been taken from the Canons of St. Paul's and added to it by Hugh de Berneres. (fn. 13) This entry suggests that Hugh, who occurs as one of the bishop's tenants at Stepney (7), had possessed the whole of Fafiton's land here at some date between 1066 and 1086. The holdings of Robert Fafiton and Robert fitz Roscelin are not reckoned in the bishop's 32 hides at Stepney. In the survey of the Count of Mortain's fief Bedfont is said to belong to the count's manor of Feltham but there is no reason to suppose that Bedfont's 2 hides are included in the 12 hides at which Feltham is assessed. (fn. 14)

It may be assumed that Middlesex was originally assessed at a round number of hides, but there is no earlier estimate with which the Domesday figure can be compared, for this shire is not included in the County Hidage. (fn. 15) The total assessment in Domesday is 880¼ hides, (fn. 16) and the totals for the hundreds, discussed below, (fn. 17) do not indicate whether this irregular sum represents an original assessment at 800, 850, or 900 hides. Although the shire and the hundred totals of Domesday obscure the artificiality of the ancient scheme of assessment in hides, the number of Domesday villages (fn. 18) assessed at 5 hides or a multiple of 5 is very large and the Middlesex evidence is cited by Round in his well-known paper on the Five-hide Unit. (fn. 19) Of the 61 places named in the Survey 32 have assessments which are exact multiples of 5 hides. Twenty-four of these villages were held as undivided units in 1086, but the eight others were divided into a number of holdings. Of these divided villages 'Ticheham' (58, 70, 86) (fn. 20) is perhaps the best illustration. In 1086 it was divided into three holdings which were held respectively by under-tenants of Earl Roger, 9½ hides (58), Geoffrey de Mandeville, 3½ hides (70), and by Robert Fafiton, 2 hides (86). Before the Conquest it was divided into seven parts.

As in other counties the hidages of neighbouring villages which may have been assessed together at some earlier date, form multiples of 5 when added together. Groupings of this kind may also be suggested in Middlesex. Sunbury, assessed at 7 hides (41), and Shepperton, 8 hides (42), form together a 15-hide unit. (fn. 21) The three Bedfonts, (61, 77, 78) form a 20-hide unit, and Ashford, 1 hide (60), with Staines, 19 hides (40), form another 20-hide unit. Round assumes that Ashford and Staines together form one entity, (fn. 22) and since Domesday specifically states that the soke of Ashford belonged to Staines there is no reason to doubt this.

The six Domesday Hundreds of Middlesex, Speletorne [Spelthorne], Helet(h)orne [Elthorne], Gara or Gare [Gore], Honeslauu [Hounslow], (A)delmetone [Edmonton], Osuluestan(e) [Ossulstone], agree to a remarkable extent with their modern counterparts. The careful rubrication of the text makes it possible to assign most holdings to their appropriate hundreds. Hayes (3) and Harmondsworth (48) are both entered without hundred rubrics; so also are Draitone (35) which follows entries assigned to Ossulstone Hundred, and Kingsbury (75) which follows an entry relating to Ruislip in Elthorne Hundred. These four manors may, however, be assigned to the hundreds in which they are found later. Draitone is proved by its connexion with St. Paul's to be West Drayton in Elthorne Hundred. Hayes and Harmondsworth may also be assigned to Elthorne and Kingsbury to Gore Hundred. The Harmondsworth and Kingsbury entries relate to parts of divided villages and other holdings in the same villages (fn. 23) are known from the rubrics to belong to the hundreds to which the unrubricated entries are here assigned. Confirmation is to be found in a 12th-century document called the Hidagium Comitatus Totius Middlesexe, printed by Round, (fn. 24) which, although corrupt and in part fragmentary, generally agrees with the Middlesex Domesday. (fn. 25) In this document a total assessment of 10 hides is given for Kingsbury in Gore Hundred, which corresponds to the total for the two Domesday holdings. The Hidagium lacks the part relating to Elthorne but the total number of hides in that hundred is given as 224 which is only ½ hide short of the hundred's hidage in Domesday if the unrubricated entries for Hayes, Harmondsworth, and West Drayton are included.

Hounslow as a hundred name did not survive long after 1086, for in subsequent documents (fn. 26) the hundred is referred to as Isleworth, after the manor of that name. Towards the end of the 12th century or possibly the beginning of the 13th this hundred suffered the loss of Hampton (81) which became part of Spelthorne. (fn. 27) The name of Gore Hundred survived in Gore Farm, in Kingsbury, which was demolished in 1937 and it is probable that the hundred met nearby. (fn. 28) Although a 17th-century map (fn. 29) shows the two divisions, Finsbury and Wenlocksbarn Hundred and Ossulstone Hundred within the earlier hundred, the boundaries of the original Ossulstone remained unaltered until the 19th century when parts of it were absorbed into the administrative county of London. (fn. 30)

An analysis of the Domesday entries gives the following assessments for the hundreds: (fn. 31)

A certain symmetry may be noticed in these figures. Spelthorne and Hounslow each have approximately 100 hides, while Elthorne and Ossulstone are also roughly equal and appear to be double hundreds. The hundreds of Gore and Edmonton together are almost equal in hidage to Ossulstone or Elthorne. In the later Hidagium (fn. 32) Edmonton is described as the 'half Hundred of Mimms' and Gore is presumably to be regarded as 1½ hundred.

A reconstruction of pre-Conquest Middlesex from Domesday shows that there are relatively few examples of villages divided among several lords. What has been described as 'the free, the lordless, village' (fn. 33) may perhaps be traced in the two western hundreds of Elthorne and Spelthorne. (fn. 34) In 1066 Middlesex was already a county of large estates and no less than seven-eighths of the whole shire formed part of the estates of ten great landowners, ecclesiastical and lay, and their dependants. The three great ecclesiastical estates, those of the Archbishop of Canterbury, the Bishop of London and the Canons of St. Paul's, and the Abbot of Westminster, like the estates of other ancient ecclesiastical bodies, continued almost unchanged into the Norman period. The greatest of the lay estates are those of Earls Loefwine and Ælfgar and Ansgar the Staller. Lesser, although by no means negligible, estates were held by Earl Harold, Wlward 'White', Azor the housecarl, and Wigot (of Wallingford). To each of the great estates were attached men who were either free or not free to take themselves and their land to another lord. In some instances men had commended themselves to lords who did not hold estates in Middlesex. Manorialization had reached an advanced stage in Middlesex before 1066. (fn. 35)

The Survey shows that the manorial arrangements had altered little since the coming of the Conqueror. In several instances a French lord had been given an Englishman's estate. For instance the two manors of Isleworth (80) and Hampton (81), recorded as held by Earl Ælfgar, passed to Walter of St. Valery; and the three manors of Northolt (71), Edmonton (72), and Enfield (73), formerly held by Ansgar the Staller, passed into the possession of Geoffrey de Mandeville. Consolidation of estates took place on the fief of Earl Roger where four west Middlesex holdings (fn. 36) became associated with Colham (55), which had been held by Wigot (of Wallingford). (fn. 37) An estate of 1 hide in Harmondsworth (53) and another of 3 hides at Dawley (57) were each said to pertain to Colham (modo iacet in Coleham). In both cases the previous holder had been a 'man of Wigot', and, although each had held his land freely and had been able to do with it as he pleased, this relationship may have been used as a reason for the consolidation of the estates with Colham in 1086. The other two estates which were bound to Colham were an estate of 9½ hides at 'Ticheham' (58) (iacet modo in Coleham) and one of 1½ hide at Hatton (51) (modo apposita est in Coleham). The consolidation of these estates probably took place in response to economic and managerial requirements. Signs of depression, particularly of sokemen into the class of villani, and signs of a general tightening up of the relationship between lord and man are not wanting, (fn. 38) but the pattern of landholding did not as yet show any radical change. 'Ticheham' supplies a good example of continuity in manorial arrangements.

There seems to be little doubt that the two Englishmen holding of Geoffrey de Mandeville were the two sokemen who held the land in 1066, and it is reasonable to assume that of those holding of Earl Roger the Englishman at least was one of those who held this land in 1066. 'Ticheham' was divided among three tenants-in-chief in 1086 and to that extent was more consolidated, but the number of small estates was the same as before.

One of the special points of interest is the detail which is given concerning the men's holdings. Whereas in most counties Domesday gives only the hidage of the manor and the hides in demesne, in Middlesex the men are graded according to their relative importance and to the assessment of their holdings, which is given in hides, fractions of hides, virgates, or acres. There are four virgates to the hide and in entries where holdings are given in acres there are 120 acres to the hide. Although the Domesday hide did not everywhere comprise 120 acres, the details relating to Staines (40) strongly suggest that it did in Middlesex. On the basis of 120 acres to the hide the total holding of the men at Staines together with the eleven hides in demesne falls short of the total assessment of 19 hides by one acre, and, in view of the large number of men's holdings involved, this can hardly be a coincidence.

The words dominium or dominicum, translated above as 'demesne', have two meanings in Domesday. (fn. 39) The first refers to that part of a fief which is retained by the lord and is not subject to subinfeudation. In Middlesex this usage occurs most frequently in connexion with the ecclesiastical fiefs. Thus William de Ver is recorded as holding one hide of the Bishop of London (10), and it is stated that William, Bishop of London, held this land in dominio cum suo manerio Stibenhede T.R.E., and similarly a number of estates held by the Abbot of Westminster are stated to 'belong to' the demesne of St. Peter. (fn. 40) In 1086 most estates in Middlesex were held in demesne, and, where subinfeudation has taken place, it is usually found in the smaller holdings. Notable exceptions occur on the fiefs of Earl Roger who retained as demesne four estates totalling 14½ hides out of his 42-hide fief, and Walter fitz Other who retained a single manor of 15 hides out of an estate of nearly 35 hides. The whole of Robert Blund's 8-hide manor and William fitz Ansculf's manor of 5 hides were held by under-tenants. Geoffrey de Mandeville retained 90 out of 99 hides, Ernulf of Hesdin retained 30 out of 37½ hides, and Robert Gernon held 2 out of 4 hides. The remaining lay lands were all held in demesne. Among the ecclesiastical estates the Archbishop of Canterbury retained his two vast manors of Hayes and Harrow and his sole feudatory was Geoffrey de Mandeville with a small 2-hide tenement. The Abbot of Westminster held 93 hides in demesne leaving only 6½ hides in the hands of three under-tenants. Among the lands of St. Paul's there is a noticeable increase in the number of under-tenancies, but even so the Bishop of London retained 32 hides out of 52 in Stepney and 40 out of 50 hides in Fulham. Among the fees held by the canons a number were farmed out to villani, the largest of these being the 15-hide manor of Willesden (22). Despite this increased subinfeudation the amount of St. Paul's estates held in demesne, measured in terms of assessment, was twothirds of the whole. The total proportion of all Middlesex estates held in demesne in 1086 was approximately four-fifths. It is possible that the scribe responsible for the Middlesex Survey intended to distinguish demesne holdings from others by a marginal M. The symbol is so used on the fiefs of Westminster Abbey and of Geoffrey de Mandeville but elsewhere the scribe's practice is not consistent. (fn. 41)

The second meaning of dominium is that part of the estate which is not held by dependent villani or other men but is reserved for the use of the lord whether he is a tenant-in-chief or an under-tenant. The word is much more commonly used in this sense as almost every entry in the Middlesex Survey shows. Although the Survey of this county supplies a wealth of detail concerning the hidage of the various holdings, the information as a whole is often incomplete. In a minority of entries, mostly relating to very small estates, the assessment alone is given and since no distinction is made between the lord's land and the men's it must be assumed that the lord was responsible for the whole of the geld. Sometimes here, as in other counties, the assessment on the estate and the hides in demesne are given, from which the hides among the men may be inferred. Surveys of the larger estates set out the individual holdings of the men in hides, virgates, or acres, generally after the demesne has been recorded. On some estates the aggregate of the demesne and the men's holdings agreed with the total assessment, in others there was disparity, and sometimes serious disparity between them. The manor of Staines, analysed above, (fn. 42) is an example of a holding in which hides in demesne and among men agreed with the total assessment. The manor of Edmonton (72), analysed below, is one in which the totals do not agree; 16 of the total of 35 hides were demesne.

The disparity varies considerably. In some instances it is negligible, in others it is so great as to raise the question whether the hides of the village assessment and those of the men's holdings are the same units. Shepperton (42) has a total assessment of 8 hides, but the demesne and the men's hides amount to 7 hides 3 virgates and 24 acres, a difference of only 6 acres. On the manor of Edmonton (Table 3) the difference is 6 hides and 20 acres. At Hendon (47) there is a difference of 4 hides. The difference is greatest on some of the largest estates, which appear to have suffered the least disruption at the time of the Conquest. Within the Archbishop of Canterbury's 100-hide manor of Harrow (4), for example, the demesne and the men's holdings account for only 60 hides 2 virgates 13 acres. On the 70-hide manor of Isleworth (80), there is, a deficiency of 40 hides 3 virgates. At Fulham (17), which is assessed at 40 hides there is, on the other hand, a surplus of 9 hides. A complete list of entries, hundred by hundred, is given in Appendix I. (fn. 43) In this analysis are shown separately instances in which the assessment on the whole estate agrees with the sum of hides distributed among demesne and men, instances in which the assessment and the sum of the details do not agree, and the estates for which the information given is incomplete. The tables seem to support Vinogradoff's view that 'the county was, as a rule, assessed in a much heavier manner than the occupation of the soil at the time of Domesday would have warranted'. (fn. 44) In order to explain the disparities Vinogradoff held that although the half-virgates, virgates, and acres at which the men's holdings were assessed 'were intended to be sub-divisions of the hides at which the manors are rated for the geld', (fn. 45) it is necessary 'to take the estimate of the tenants' land to refer not only to artificial units, but to field holdings, the hides, virgates and acres of agrarian occupation'. (fn. 46) This may be the correct explanation but if, as he suggests, 'a geld-hide may have corresponded to one hide and a half or two hides distributed among the rustics', (fn. 47) it is difficult to understand why Domesday should give estimates based on a unit which sometimes corresponded to and sometimes bore no relation to the geld hide. If the instances in which the sum of the demesne and the peasant holdings exceeds the total assessment were numerous it might be necessary to adopt Vinogradoff's theory, although the estimates of the peasant holdings would be deprived thereby of much of their value. These instances are, however, in a minority. (fn. 48) More frequently the total of the individual holdings falls short of the assessment on the whole estate and the deficiency may be accounted for by assuming that some of the tenements to which part of the geld assessment had been allotted had ceased to exist, or rather had reverted to the lord although they had not been incorporated in the demesne. It may be suggested that their omission in Domesday is connected with the exemption of the demesne from geld.

It is clear from the geld accounts or 'Geld Rolls' relating to the south-western counties and preserved in the Exon Domesday, that when geld was levied at 6s. on the hide in 1084 the demesnes of tenants-in-chief were exempt. How long such exemption was enjoyed is unknown, but presumably it was assumed in 1086 that such exemption would or might be enjoyed in the future and the information in Domesday ought to enable the collectors of geld to distinguish between exempt land and that which paid geld. The exempt land was the tenant-in-chief's demesne in the narrow sense (fn. 49) and on many estates there were two sorts of geld-paying land: that which had been subinfeudated and that which the men, villani and others, held. There is good reason to suppose that the demesne could vary in extent since in a variety of circumstances lands granted to knights might revert to the lord and be reunited with the demesne. It is not certain, however, whether the lord was free to add any part or all of the men's land to his desmesne. There are passages in Domesday which suggest that the lord was encroaching on the land of the villani if he used it to endow knights and servants. (fn. 50) It is possible that terra villanorum of this kind could not be added to the exempt demesne but remained liable for geld even when taken into the lord's hands. If this were so, the deficiency mentioned above might be land of this kind. We should have to assume that at Harrow, although the archbishop paid no geld on the 30 hides of demesne, he had in his hands nearly 40 hides on which the geld was payable just as it was on the men's holdings, amounting to approximately 30 hides.