The servitium debitum, therefore, was a standing institution in Normandy, and 'to the mass of his (William's) followers', as Mr Freeman frankly admits,¹⁰³ a 'feudal tenure, a military tenure, must have seemed the natural and universal way of holding land'. When we find them and their descendants holding their fiefs in England, as they had been held in Normandy, by the service of a round number of knights, what is the simple and obvious inference but that, just as Henry II granted out the provinces of Ireland to be held as fiefs by the familiar service of a round number of knights,¹⁰⁴ so Duke William granted out the fiefs he formed in England?

If to escape from this conclusion the suggestion be made that these servitia debita were compositions effected by English antecessores, it need only be answered that the fiefs acquired were wholly new creations, constructed from the scattered fragments of Anglo-Saxon estates. And though in the case of the church fiefs this objection might not apply, yet we have evidence, as I shall show, to prove that their servitia also were determined by the conqueror's will, as indeed might be inferred from their close correspondence with those of the lay barons.

But if the lands of the conquered realm were so granted to be held by a servitium debitum of knights, the key of the position is won, and the defenders of the existing view must retire along the whole line; for, as Mr Freeman himself observed, 'Let it be once established that land is held as a fief from the crown on condition of yielding certain services to the crown, and the whole of the feudal incidents follow naturally.'¹⁰⁵

I am anxious to make absolutely clear the point that between the accepted view and the view which I advance, no compromise is possible. The two are radically opposed. As against the theory that the military obligation of the Anglo-Norman tenant-in-chief was determined by the assessment of his holding, whether in hidage or in value, I maintain that the extent of that obligation was not determined by his holding, but was fixed in relation to, and expressed in terms of, the constabularia of ten knights, the unit of the feudal host. And I, consequently, hold that his military service was in no way derived or developed from that of the Anglo-Saxons, but was arbitrarily fixed by the king, from whom he received his fief, irrespectively both of its size and of all pre-existent arrangements. Such propositions, of course, utterly and directly traverse the view which these passages best summarize:

The belief that William I divided the English landed property into military fees is erroneous.... According to the extent and the nature of the productive property it could be computed how many shields were to be furnished by each estate, according to the gradually fixed proportion of a £20 ground-rent.¹⁰⁶

There is no ground for thinking that William directly or systematically introduced any new kind of tenure into the holding of English lands. There is nothing to suggest any such belief, either in the chronicles of his reign, in the Survey, which is his greatest monument, in the genuine or even in the spurious remains of his legislation.... As I have had to point out over and over again, the grantee of William, whether the old owner or a new one, held his land as it had been held in the days of King Edward.¹⁰⁷

There can be no doubt that the military tenure ... was itself introduced by the same gradual process which we have assumed in the case of the feudal usages in general. We have no light on the point from any original grant made by the Conqueror to a lay follower; but ... we cannot suppose it probable that such gifts were made on any expressed condition, or accepted with a distinct pledge to provide a certain contingent of knights for the king's service.¹⁰⁸

If my own conclusions be accepted, they will not only prove destructive of this view, but will restore, in its simplicity, a theory which removes all difficulties, and which paves the way to a reconsideration of other kindred problems, and to the study of that aspect of Anglo-Norman institutions in which they represent the feudal spirit developed on feudal lines.

III. SCUTAGE, AID, AND 'DONUM'

Precious for our purpose as are the cartae of 1166, their evidence, as it stands, is incomplete. It needs to be supplemented by the early Pipe-Rolls of Henry II's reign. By collating these two authorities we obtain information which, singly, neither the one nor the other could afford. All those entries on the rolls which relate to scutagia, auxilia or dona require to be extracted and classified before we can form our conclusions. Hitherto, historians have remained content with repeating Swereford's obiter dicta, as extracted from the Liber Rubeus by Madox, without checking these statements by the evidence of the rolls themselves.

The question of Swereford's authority is one which it is absolutely necessary to deal with, because his statements have been freely accepted by successive historical writers, and have formed, indeed, the basis on which their conclusions rest. Now the presumption is naturally in favour of Swereford's knowledge of his subject. His introduction to the Liber Rubeus is dated 1230, and he tells us that he had been at work among the records in the days of King John, under William of Ely¹⁰⁹ himself: he wrote with the actual rolls before him; he had been intimate with the leading officials of the exchequer, and enjoyed full knowledge of its practice and its traditions. I cannot wonder that, this being so, his positive assertions should have been readily believed, or that Mr Hall, when, for a short time, I was associated with him in preparing the Red Book for the press, should, with a kindly bias in favour of so venerable an authority, have shrunk from my drastic criticism of his famous introduction to that volume.

On the other hand we have Swereford's own admission that he worked from the rolls alone.¹¹⁰ These rolls are, for all purposes, as accessible to us as they were to him, while we possess the advantage of having, in contemporary chronicles, sources of information which he did not use, and with which, indeed, he shows no sign of being even conversant. We must go, therefore, behind Swereford and examine for ourselves the materials from which he worked.

Passing, for the present, over minor points, I would fix on the 'Great Scutage', or 'Scutage of Toulouse', as the test by which Swereford's knowledge and accuracy must stand or fall. If he is in error on this matter, his error is so grievous and so far-reaching that it must throw the gravest doubt on all his similar assertions. The date of the expedition against Toulouse was June 1159 (the host having been summoned at Mid-Lent): from the chroniclers we learn that, to provide the means for it, and especially to pay an army of mercenaries, a great levy was made in England and beyond sea. The roll of the following Michaelmas records precisely such a levy, and the payments so recorded must have been made for the expenses of this campaign. But we can go further still; we can actually prove from internal evidence that sums accounted for on the roll of 1159 were levied expressly for the Toulouse campaign.¹¹¹ Yet we are confidently informed by Swereford that this levy was for a Welsh war, and that the scutage of Toulouse is represented by the levies which figure on the rolls of 1161 and 1162. He appears to have evolved out of his inner consciousness the rule that a scutage, though fixed and even paid in any given year, was never accounted for on the rolls till the year after.¹¹² But as even this rule will not apply to his calculation here, one can only suggest that he was absolutely ignorant of the date of the Toulouse campaign.¹¹³ The value of Swereford's calculations is so seriously affected by this cardinal error, that one may reject with less hesitation his statement that the scutage of 1156 was taken for a Welsh war, and not, as there is evidence to imply, for a campaign against the king's brother. Swereford, again, may be pardoned for his ignorance of the fact that scutage existed under Henry I,¹¹⁴ but when he unhesitatingly assigns the Domesday Survey to the fourteenth year of the Conqueror (1079-80), he shows us that the precision of his statements is no proof of their accuracy. On both these points he has misled subsequent writers.¹¹⁵

The incredible ignorance and credulity even of officials at the time are illustrated by the fact that the Conqueror was generally believed to have created 32,000 knights' fees in England, and that Swereford plumed himself on his independence in doubting so general a belief.¹¹⁶ His less sceptical contemporary, Segrave, continued to believe it, and even Madox hesitates to reject it.

The persistent assertion that the Cartae Baronum were connected with, and preliminary to, the auxilium ad filiam maritandam of 1168 is undoubtedly to be traced to Swereford's ipse dixit to that effect. He distinctly asserts that the aid was fixed (assisum) in the thirteenth year (1167), that the returns (cartae) were made in the same year (1167), and that the aid was paid and accounted for in the fourteenth year (1168).¹¹⁷ Modern research, however, has shown that the returns were made quite early in 1166, while the youthful Matilda, we know, was not married till October 1168. This throws an instructive light on Swereford's modus operandi. Finding from the rolls that the payments made in 1168 were based on the returns in the cartae, and not being acquainted with the date of the latter, he jumped to the conclusion that they must have been made in 1167, it being his (quite unsupported) thesis that all levies were fixed in the year preceding that in which they were accounted for on the rolls.

Proceeding further, we find him explaining (p. 9) that he omits the aid of 1165, 'quoniam probata summa auxilii propter hoc non probatur numerus militum'. And yet this aid, the last to be taken before the returns of 1166, is of special value and importance for the very purpose he speaks of. It is, indeed, an essential element in the evidence on which I build; and this compels me to discuss the point in some detail.

Those who contributed towards this aid either (1) gave arbitrary sums for the payment of servientes—whose number was almost invariably some multiple of five—or (2) paid a marc on every fee of their servitium debitum. We are only here concerned with those who adopted the latter course. Now let us take the case of those who adopted this alternative in the counties of Notts and Derby, and compare their payments with their servitium debitum as known to us from other sources.

In this case there is no doubt as to the servitium debitum, for it is ascertained from the cartae themselves. Having then proved, by this test, the exact correspondence of the payments, I turn to the case of Devonshire.

Here we are supplied by this roll with four important servitia which would otherwise be absolutely unknown to us. And they happen to be of special interest. For while the carta of William de Braose returns twenty-eight fees, and that of Oliver de Traci twenty-three and a half (though he pays on thirty and a half),¹¹⁹ their payments in 1165, by revealing their servitium debitum, show us that their fiefs represent the two halves of the Honour of Barnstaple (which, therefore, was assessed at 50 knights) then in their respective hands. Again, William de Traci returns his fees in his carta as twenty-five and three-quarters, and says nothing about any balance on his dominium, as he should have done. Hence we should not have known his servitium but for the roll of 1165.

Swereford's extraordinary failure to understand this roll aright is possibly due to the fact that most of the relevant payments are entered without mention of their object. He seems to have been very dependent upon the rolls explaining themselves, and to have worked in the spirit of a copying clerk rather than of an intelligent student.

One more example of his errors will suffice. In his abstracts from the aid 'ad maritandam primogenitam filiam regis' (1168), we read:

Abbas Gloucestriæ de promissione, sed non numeratur quid; sed in rotulo praecedenti dicitur:—Abbas Gloucestriæ debet xxxviij. l. ij. s. vj. d. de veteri scutagio Walliae.

Now (1) the amount of the abbot's contribution is duly entered on the roll ('xl. marcas de promissione de eodem auxilio'), and it is not paid in respect of fees, but is a voluntary proffer; (2) the phrase in the preceding roll is not 'de veteri scutagio', but 'de veteri exercitu'; (3) the payment there recorded represents a contribution of fifty servientes, and had nothing to do with scutage, for the abbot (as Swereford should have known) did not hold by military service, and ought not, therefore, to figure in his lists at all.¹²⁰

Let us turn, therefore, to the rolls themselves. Now, although the language of the exchequer was not so precise as we could wish, it is possible, more or less, to distinguish and classify these levies. Thus, we have of course a typical 'aid' in the levy for the marriage of the king's daughter (1168), while, on the other hand, we have an equally typical 'scutage' in 1156, in the payments made by the church tenants in lieu of military service.

On the institution of 'scutage' there has been much misconception. It is placed by our historians among the great innovations wrought by Henry II, who is supposed by them to have introduced it in 1156.¹²¹ Here we see, once again, the danger of seeking our information on such points secondhand, instead of going straight to the fountainhead for ourselves.

John of Salisbury implies that scutage was no novelty in 1156 when he writes, not that the king imposed it, but that he 'could not remit it'. This inference is at once confirmed by the appearance of scutage eo nomine in the reign of Henry I.

The following charter is found in the (MS.) Liber Eliensis (Lib. III), No. xxi, and in the Cottonian MS. Nero A. 15:

H. rex Anglorum Archiepiscopis, Episcopis, Abbatibus, Comitibus, etc. Salutem. Sciatis me condonasse Ecclesiæ S. Ætheldredæ de Ely pro Dei amore et anima Patris et Matris meae et pro redemptione peccatorum meorum, et petitione Hervei ejusdem Ecclesie Episcopi 40 libras de illis 100 libris quas predicta Ecclesia solebat dare de Scutagio quando Scutagium currebat¹²² per terram meam Anglie: ita quod Ecclesia amodo inperpetuum non dabit inde nisi 60 libras quando Scutagium per terram evenerit, et ita inperpetuum sit de predictis libris Ecclesia predicta quieta. T. Rogero Episcopo Saresberiensi, Gaufrido Cancellario meo et Roberto de Sigillo et Willelmo de Tancarvilla et Willelmo de Albineio Pincerna et Radulfo Basset et Gaufrido de Clintona et Willelmo de Pondelarche. Apud Eilinges in transitu meo.

This is followed by (No. xxii) a grant of Chatteris Abbey to the church of Ely;¹²³ and this again is followed, in a register of Chatteris Abbey,¹²⁴ by a remission of 6s 7d Wardpenny hitherto paid by that abbey. The first and third charters receive singular confirmation, being thus accounted for in the Pipe-Roll of Henry I:

Et idem Episcopus debet ccxl. li. ut rex clamet eum quietum de superplus militum Episcopatus, et ut Abbatia de Cateriz sit quieta de Warpenna (p. 44).

This entry, moreover, connects the scutagium with the system of knight-service (superplus militum).

It is delicious to learn, on comparing the records, that the virtuous king who made these grants for the weal of his parents' souls and the remission of his own sins, extorted from the church, for making them, an equivalent in hard cash.¹²⁵

Again, the (MS.) Cartulary of St Evroul contains a confirmation by Randulf, Earl of Chester (1121-29) of his predecessor (d. 1120) Earl Richard's benefaction, 'liberam et quietam ab escuagio', etc., etc. The list of the Abbot of Peterborough's knights (see p. 131) is a further illustration of knight-service temp. Henry I, while the entry as to Vivian, who was enfeoffed by Abbot Turold: 'servit pro milite cum auxilio' (Chron. Petrob., p. 175), must refer to the somewhat obscure 'auxilium militum' of the period. So also, it would seem, must the curious charter of Eustace, Count of Boulogne,¹²⁶ in which he speaks of his knights serving: 'sive in nummis, sive in exercitu, sive in guarda', under Henry I. Most important of all, however, is a passage on which I have lighted since this essay first appeared. In reading through the letters of Herbert (Losinga), Bishop of Norwich (d. 1119), I found this appeal to the Bishop of Salisbury, in the king's absence from England:

In terris meis exiguntur quinquaginta libræ pro placitis, cum earundem terrarum mei homines nec in responsionem nec in facto peccaverint.¹²⁷ Item pro militibus sexaginta libræ quos [? quas] tanto difficilius cogor reddere, quanto annis præteritis mea substantia gravius attenuata est (Ed. Giles, p. 51).

The sum is that to which the Ely contribution is reduced by the above charter, and the death of the writer in 1119 proves the early date of the payment.

Indeed, a little consideration will show that payment in lieu of military service, which was the essential principle of scutage, could be no new thing. The two forms which this payment might assume—payment to a substitute, or payment to the crown—both appear in Domesday as applicable to the fyrd; the former is found in the 'Customs' of Berkshire, the latter in other passages. From the very commencement of knight service, the principle must have prevailed; for the 'baron' who had not enfeoffed knights enough to discharge his servitium debitum, must always have hired substitutes to the amount of the balance. Nor is this a matter of supposition: we know as a fact, from the Abingdon Chronicle and the Ely History, that under William I knights were so hired.¹²⁸ Here it should be noted, as a suggestive fact, that the 'forty days' of military service, though bearing no direct proportion either to the week or to the month, do so to the marc and to the pound. The former represents 4d, and the latter 6d, for each day of the military service.¹²⁹ It may fairly be assumed that this normal 'scutage' would be based on the estimated cost of substitutes paid direct. Thus the only change involved would be that the tenant would make his payments not to substitutes, but to the crown instead.

There is a valuable entry bearing on this point in the roll of 8 Henry II (p. 53). We there read:

Et in liberatione vii. militum soldariorum de toto anno quater xx. et iiii. li. et xviii. s. et viii. d. Et in liberatione xx. servientium de toto anno xxx. li. et vi. s. et viii. d. Et in liberatione viii. Arbalist' viii. li. et xvi. sol. Et in liberatione v. vigilum et i. Portarii vi. li. et xvi. d.

This represents 8d a day to each of the seven knights for a year of 364 days, which, be it observed, corresponds precisely with the statements in the Dialogus: 'Duo milites bajuli clavium quisque in die viii. [den.] ratione militiae; asserunt enim quod equis necessariis et armis instructi fore teneantur', etc. (i. 3). And so, we see, a scutage of two marcs, such as that which was raised for the expedition of Toulouse (1159), would represent, with singular accuracy, 8d a day for the forty days of feudal service, or exactly a knight's pay. Again the pay of the serviens, recorded in this passage, works out at a penny a day for a year of 364 days, which has an important bearing, we shall find, on the roll of three years later (11 Henry II). A similar calculation shows that the porter received 2d a day, and the vigil 1d—the very pay assigned him in the Dialogus (i. 3). There is another similar passage in the roll of 14 Henry II (p. 124):

Et in liberatione i. militis et ii. Portariorum, et ii. vigilum de Blancmost' xviii. li. et v. sol. Et in liberatione xl. servientum de Blancmust' de xxix. septimanis xxxiii. li. et xvi. s. et viii. d. Et xx. servientibus qui remanserunt xxiii. septimanas xiii. li. et viii. s. et iiii. d.

Here again the knight's pay works out at 8d a day, while the porters, the watchmen, and the servientes received 1d. Specially valuable, however, are the entries (to which no one, I think, has drawn attention) relating to the small standing guards kept up in the summer months at 'Walton' and Dover.¹³⁰ Eventually the payments to these guards were made from the central treasury ('exitus de thesauro'), and are therefore appended, on the rolls, to the list of combustiones where no one would think of looking for them.

On the roll of 10 Henry II we find: 'Liberatio iiii. militum et ii. servientum de Waletone a festo Ap. Phil' et Jac' usque ad festum S. Luce xxiiii. li. et xx. d.' This works out at exactly 8d a day for the miles, and 1d for the serviens. On the roll of the next year the five knights at Dover are paid £25 for 150 days' service, or exactly 8d a day each. So too on the roll of the thirteenth year we read: 'Liberatio iiii. militum de Waletone xxiii. li. et ix. s. et iiii. d. de clxxvi. diebus.... Et ii. servientibus de clxxvi. diebus xxix. sol. et iiii. d.' Here again the miles gets 8d, the serviens 1d a day. It is needless to multiply instances, but it may be added that similar calculations show the sailors of Richard's crusading fleet to have received 2d and their boatswains 4d a day.

It is, perhaps, possible to trace a complete change of policy in this matter by the crown. The Conqueror, we may gather from divers hints, was anxious to push forward the process of sub-infeudation, that as many knights as possible might be actually available for service. As the chief danger lay, at first, in the prospect of English revolt it was clearly his policy to strengthen to the utmost that 'Norman garrison', as we may term it, which the feudal system enabled him to quarter on the conquered land.¹³¹ But as the two races slowly coalesced, the nature of the danger changed: it was no longer a question of Norman versus Englishman, but of danger to the crown from war abroad and feudal revolt at home. Thenceforth its policy would be no longer to encourage personal service, but rather payment in lieu thereof, which would provide the means of hiring mercenaries, a more trustworthy and useful force. Clearly the accession of the Angevin house would, and did, give to this new policy a great impetus.

The first levy to which the rolls bear witness is that of 1156. As this was only raised from the church fiefs, Henry II was, as yet, confining himself strictly to the precedent set him, as we know, in his grandfather's reign. This levy was at the rate of one pound on the fee, and was made on the old assessment (servitium debitum).

I have already shown that the levy in question was not, as alleged, an innovation. Dr Stubbs writes: 'The peculiar measure of the second year was the collection of scutage from the knights' fees holding of ecclesiastical superiors,¹³² a measure which met with much opposition from Archbishop Theobald at the time';¹³³ and speaking of William of Newburgh, he suggests that 'possibly in William's estimation the consent of St Thomas took from the scutage on church fees its sacrilegious character'.¹³⁴ But if the institution was fully recognized under Henry I, how was it 'sacrilegious'? Theobald's 'opposition' in 1156 can only be inferred from the king's reply explaining the necessity for the levy,¹³⁵ and was clearly directed, not against the principle, but by way of appeal against the necessity in that instance. Miss Norgate holds that 'no resentment seems to have been provoked by the measure', although she sees in it 'the origin of the great institution of scutage'.¹³⁶ Then there is the question of the object for which the levy was made. Swereford says 'pro exercitu Walliæ',¹³⁷ and this misled, through Madox, Dr Stubbs (who wrote 'the scutage of 1156 was also for the war in Wales',¹³⁸) and Gneist.¹³⁹ The former writer, however, has elsewhere¹⁴⁰ pointed out that 'its object was to enable Henry to make war on his brother'; and Miss Norgate gives the same explanation.¹⁴¹ Swereford's error, I believe, can undoubtedly be traced to an entry on the Pipe-Roll of the third year (1157) recording the payment by the Abbot of Abbotsbury of two marcs 'de exercitu Walie'.¹⁴² But this must refer to the Welsh campaign of that year, not to the foreign trouble of the year before.¹⁴³

The next levy was 'the scutage of Toulouse' in 1159. This, 'the great scutage' of Miss Norgate,¹⁴⁴ is, strange as it may seem, on the Pipe-Roll itself almost uniformly styled not a scutage, but a donum. The explanation given by Swereford is wholly inadequate, and is this: 'Intitulaturque illud scutagium De Dono ea quidem, ut credo, ratione quod non solum prelati qui tenentur ad servitia militaria sed etiam alii abbates, de Bello et de Salopesbiria et alii tunc temporis dederunt auxilium'.¹⁴⁵

Miss Norgate, adopting this explanation, writes:

The reason doubtless is that they were assessed, as the historians tell us, and as the roll itself shows, not only upon those estates from which services of the shield were explicitly due, but also upon all lands held in chief of the crown, and all church lands without distinction of tenure; the basis of assessment in all cases being the knight's fee, in its secondary sense of a parcel of land worth twenty pounds a year. Whatever the laity might think of this arrangement, the indignation of the clergy was bitter and deep. The wrong inflicted on them by the scutage of 1156 was as nothing compared with this, which set at nought all ancient precedents of ecclesiastical immunity, and actually wrung from the church lands even more than from the lay fiefs.¹⁴⁶

I am obliged to quote the passage in extenso, because, in this case, the accomplished writer betrays a singular confusion of ideas, and misrepresents not only the levy, but also the point at issue. The whole passage is conceived in error, error the more strange because Miss Norgate enjoyed over her predecessors the advantage of writing with the printed roll before her. The lay estates were not, as implied ('all lands held in chief of the crown'), in any way exceptionally assessed: in no case was the basis of assessment the unit alleged by the writer; and as to the 'church lands', a reference to the roll will show that all over England there were only eight cases in which those not owing 'services of the shield' contributed (and that in no way as an assessment on imaginary knights' fees) to this levy, while in six out of the eight their contributions were so insignificant that their collective amount barely exceeded £50.¹⁴⁷

The true explanation is probably to be found in the fact that only a portion of the tax was raised by way of scutage. As this great levy has been wrongly supposed to have consisted of a scutage alone,¹⁴⁸ and as it played an important part in the development of direct taxation, I propose to set forth, for the first time, the various methods by which the money was raised. These were eight in number:

Of these, the first was strictly regular, being merely a repetition of the scutage of 1156, at the rate of two marcs instead of twenty shillings. The second presents some difficulty. Subject to correction, there are some fifteen cases in which the payment is made separately by fiefs, and in which the rate is clearly two marcs, while there are twenty-two in which the milites of the county pay as a group through the sheriff, and in which, therefore, we cannot actually test the rate of the levy or the manner of raising it. Swereford's ipse dixit as to the rate in these latter cases was probably based on analogy, here our only guide.

With the third and fourth divisions we return to sure ground. To them I invite particular attention, because it is to them (and especially to the third) that apply the complaints of the church chroniclers, and not (as has always, but erroneously, been supposed) to the perfectly legitimate levy of two marcs on the fee. It is necessary to emphasize the fact that the matter has been wholly misunderstood. The bitter complaint of John of Salisbury that Henry, on this occasion, 'omnibus (contra antiquum morem et debitam libertatem) indixit ecclesiis ut pro arbitrio ejus satraparum suorum conferrunt in censum', would have been without meaning had it referred (as alleged) to the latter levy (or even to the insignificant sums contributed ut supra by eight foundations); but when we learn that, over and above this legitimate levy, a far larger sum was arbitrarily wrung from the church, the truth and justice of the protest are at once made evident. I here give two tables illustrative of this exaction. Each is divided into three columns. In the first column I give the number of the knights due from each bishopric and each religious house. In the second column I give the marcs due, and paid on this occasion, on the old assessment (servitium debitum). In the third will be found the exaction complained of, namely, the dona extorted from the spiritual 'barons' themselves.

We thus obtain a grand total of 1,101 marcs raised from the church by legitimate scutage, and 4,442½ (or, adding the dona from non-feudal houses, 4,700) marcs by special imposition.¹⁴⁹ This distinction at once explains the real extortion of which churchmen complained;¹⁵⁰ and shows that it had nothing to do with scutage, but was a special imposition on the church fees from which the lay ones were exempt.¹⁵¹ The idea of the impost was not improbably the adjustment of inequalities in cases where the knight-service was a quite inadequate assessment; the precedent created was not forgotten, and it proved in later days a welcome source of revenue.

The discovery of this exaction identifies, it will be seen, in spite of Swereford's error, the levy accounted for on the roll with the famous 'scutage of Toulouse'. And if even further proof were needed, it is found in an incidental allusion which clinches the argument. Giraldus Cambrensis (iii. 357) refers to Bishop Henry of Winchester assembling all the priests of his diocese 'tanquam ad auxilium postulandum (dederat enim paulo ante quingentas marcas regi Henrico ad expeditionem Tholosanam)'. The sum here named is that which he paid in 1159, as my table shows. Its destination is thus established, as also, it may be noted, the means by which he was expected to recoup himself.

As to the scutage on the lay fiefs, the general impression, broadly speaking, is that Henry replaced his English feudal host by an army of mercenaries paid from the proceeds of a scutage of two marcs per fee on all lands held by military service.¹⁵² But is that impression confirmed by the evidence of the rolls? Without setting forth the evidence in detail, I may sum it up as amounting to this: that the grouped payments found under twenty-two counties¹⁵³ present, I think, a total of 1,895 marcs, while those of the fiefs which paid separately amounted to 666. This gives us a grand total of 2,561 marcs, representing, of course, 1,280 knights. Now although the amount of knight service due to the crown from its English realm has been, as we shall see, absurdly exaggerated, the above number, I need scarcely say, must represent a minority of the knights due from the lay fiefs. This sets the matter in quite another aspect. In spite of the passage in Robert de Monte, on which the accepted view is based,¹⁵⁴ the roll presents proof to the contrary, and indeed the words of Robert show that he knew so little of the levy in England as to believe that it was wholly arbitrary. There are, perhaps, indications that the fiefs which, on this occasion, paid scutage, were largely those in the king's hands,¹⁵⁵ and if we add to these the escheated honours, of which the scutage would be paid through the sheriffs, we must conclude that the great bulk of the tenants who had a choice in the matter served abroad with their contingents and did not pay scutage.

Before taking leave of 'the great scutage', another point demands notice. Gervase of Canterbury sets forth its proceeds in terms of great precision:

Hoc anno rex Henricus scotagium sive scutagium de Anglia accepit, cujus summa fuit centum millia et quater viginti millia librarum argenti (i. 167).

Quite desperate attempts have been made to reconcile this statement with the actual sums raised. In his preface to the Gesta Henrici Regis, Dr Stubbs suggests that Gervase included in his total the scutage of two years later (1161), but adds that, if so, the rolls are very incomplete. In his Constitutional History he speaks of 'this [scutage] and a very large accumulation of treasure from other sources, amounting, according to the contemporary writers, to £180,000' (i. 457), but admits, in a footnote, that 'the sum is impossible', and throws out as probable a different explanation. Miss Norgate writes that 'the proceeds, with those of a similar tax levied upon Henry's other dominions, amounted to some £180,000'.¹⁵⁶ But Gervase distinctly states that this sum was raised from England. Now the actual sum raised, by scutage, in England (1159) was £2,440 in all, as I reckon it, while the special clerical impost produced some £3,130 in addition. Consequently, no ingenuity can save the credit of Gervase. He was not, after all, worse than his fellows. We shall find that when mediæval chroniclers endeavour to foist on us these absurd sums they require much bolder handling than they have ever yet received.

Pass we now to the third levy, that of 1161. For this the rate was again two marcs on the fee according to Swereford (followed, of course, by subsequent writers), though the study of the roll (7 Henry II) reveals that in many cases, on the lay fiefs at least, the rate was one marc. Both this and the levy of the following year are most difficult to deal with in every way. We have seen that an entry on the roll of 1163 led Swereford to believe that the levy of 1161 was made for the Toulouse campaign, and Dr Stubbs has made the suggestion that it might have been raised to defray 'debts' incurred on that occasion;¹⁵⁷ but the difficulties in the way of accepting this view seem insuperable.¹⁵⁸

The fourth levy, which is that of 1162 (8 Henry II), was at the rate of one marc, and is recorded by Swereford, but not by Dr Stubbs.¹⁵⁹ Though richer in names than that of 1161, it is even less useful for our purpose, as the sums entered are most irregular, perhaps owing to the adoption of a new method of collection.¹⁶⁰ Neither of these levies affords, in the absence of corroboration, trustworthy evidence on the servitium of any lay fief.

The fifth levy, on the other hand, in 1165 (11 Henry II), affords most valuable evidence, although it is ignored by Swereford and by those who have followed him. It is, however, of a singular character. The money was raised, we gather from the roll, on two different systems:

(I) By a fixed payment at the rate of one marc on the fee (old assessment).

(II) By an arbitrary payment of certain mysterious sums, which prove to be multiples of the unit 15s 3d. But there is no fixed proportion to be traced between the amount paid and the number of servitia due. Numerous instances are found of a single knight's fee being charged with a sum equivalent to five of these mysterious units. Magnates, again, are found paying apparently strange sums, which prove on dissection to represent 50, 100, 200 and even 300 of these units. The clue to the mystery is found in an entry on the Pipe-Roll of the following year (12 Henry II), which proves that this unit was the pecuniary equivalent of a serviens, and that the various payers had 'promised' the king so many servientes for the war in Wales.¹⁶¹ Such 'promises' were evidently offers, made independently of the actual service due from the 'promising' party. Following up this clue, we see that the Abbot of Abingdon must, like the Bishop of Hereford, have promised 100 'serjeants',¹⁶² that the Abbot of St Alban's must have done likewise,¹⁶³ while the Bishop of London must have promised 150, in addition, be it noted, to paying a scutage of a marc on each knight's fee (20) of his servitium debitum.¹⁶⁴ For the rolls of 1162 and 1163 prove that he had duly paid the scutage of the former year, and that this was a further payment. The varying form of these entries should be observed, for it was evidently quite immaterial to the clerks whether they wrote '5 serjeants' or their equivalent—76 shillings and 3 pence.¹⁶⁵ Taking the pay of the serviens at 1d a day, the unit in question would represent six months' pay (for a year of 366 days).

But, for our present purpose, we must confine ourselves to the scutage proper. The passage on which I would specially dwell is the entry on the roll in which the custos of the archbishopric of Canterbury 'reddit compotum de cxiii. li. de Militibus de Archiepiscopatu de ii. Exercitibus' (p. 109).¹⁶⁶ In the first place, we have here, surely, witness to the two Welsh campaigns of this year, which Mr Eyton adopts, following Mr Bridgeman,¹⁶⁷ but which Miss Norgate rejects.¹⁶⁸ Secondly, this sum resolves itself, on analysis, into two constituents of 84¾ marcs each. Now the return for the archbishopric the following year is: 'Archiepiscopus habet iiij^xx. et iiij^or. et dimidium et quartam partem feffatos.'¹⁶⁹ Having set forth this exact corroboration, I will briefly trace the servitium of the See. In 1156 and 1159 it pays no scutage when the other church fiefs do, but within six months of Theobald's death it pays to the scutage of 1161 on a servitium of sixty knights, being then in the hands of the crown. Under Becket, in 1162, it is once more omitted; but in 1165 it again pays, as we have seen, and now not on sixty knights but on 84¾. In 1168 it contributes, on the same amount, to the auxilium, and in 1172, but the latter year is the first in which the recognoscit formula is employed, enabling us to determine that, as in 1161, the servitium debitum was sixty knights.

The typical difference between these sixty knights and the 84¾ actually enfeoffed will serve to illustrate the point on which I insist throughout. Had the fee been held by its tenant, he would have raised 84¾ marcs, paid sixty to the crown, and kept 24¾ for himself.¹⁷⁰ But when a custos held the fief, he could keep nothing back, and therefore paid over the whole. We have, I think, an illustration of the same kind in the payment (p. 202, note 76) by the custos of the Romare fief, 'de noviter feffatis' (noviter, be it observed not yet de novo).

Having brought the levies down to 1165, I hope it has now been made clear that the officials of the exchequer were well aware of the amount of servitium debitum from every fief, the levies being always based on the said amount. Swereford, therefore, was quite mistaken in the inference he drew from the inquest of 1166:¹⁷¹ indeed, his words prove that he completely misunderstood the problem.

This was the last levy raised previous to the making of the returns (cartae) in 1166. These returns were followed in 1168 by the first levy on the new assessment. I have already dealt with the changes which this new assessment involved, but I would here again insist upon the fact that the church and the lay fiefs were not dealt with alike, the latter being assessed wholly de novo, while the former retained their old assessments, while accounting separately, and under protest, for the fees in excess of their servitium debitum. So far as the lay fiefs were concerned, their servitia, congenital with Norman rule, were now swept away. Here, from the single county of Northumberland, are three cases in point:

The change thus made by the restless king was permanent in its effect, and thenceforth the only assessment recognized was that based upon the fees, which, by 1166, had been created de veteri and de novo.¹⁷⁴

Before leaving the subject of this levy, there is one point on which I would touch. When we find, as we often do, that the sum paid in 1168 in respect of a fief does not tally with the number of fees recorded in the cartae, we must remember that in the Liber Niger and Liber Rubeus we have not the original cartae, but only transcripts liable to clerical error. Checking the cartae by these payments, we constantly find cases in which the number of fees should be slightly greater than is recorded in the carta.¹⁷⁵ I suspect that the transcriber, in these cases, has omitted entries in the original carta, and this suspicion is strongly confirmed by the fact that where the original return enables us to test the transcript, we find in the great carta for the honour of Clare that the original transcriber has omitted half a fee of William de Hastinges, has left out altogether the entry 'Reginaldus de Cruce, dimidium militem', and has changed the quarter fee of Geoffrey fitz Piers into half a fee; while in that of the Bishop of Chichester, Robert de Denton's half fee is converted into a whole one. The later (Red Book) transcriber has made a further omission.

Another source of discrepancy may be found in the dangerous resemblance of formulae. Thus the carta of Ranulf fitz Walter records three and three-quarter fees duly accounted for. Yet his payment in 1168 is not £2 10s but £2 4s 5d. The explanation is that the holding was really three and one-third fees,¹⁷⁶ but the transcriber read 'iij[^a.] pars' (one-third) as 'iij. partes' (three-quarters).

How easily such errors arose may be seen in the elaborate entries on Simon de Beauchamp's fief. Here the formula 'decem denarios quando Rex accipit marcam de milite', correctly reproduced in the Black Book, becomes 'x. denarius', etc., in the Red Book. The former expression means 'tenpence in the marc' (i.e. one-sixteenth of a fee); whereas the latter is equivalent to 'the tenth penny in the marc' (i.e. one-tenth of a fee), and upsets the whole reckoning. The correct formula is a not uncommon one and should be compared with the 'de xx. solidis viii. denarios' (eightpence in the pound) which is given as the holding of two knights of the honour of Clare, and represents the thirtieth of a fee.¹⁷⁷

Lastly, I think that, on further examination, there are three fiefs of which the servitia debita, though at first sight irregular,¹⁷⁸ may fairly be brought into line as multiples of the constabularia. That of Bohun, though implied by the carta to be thirty and a half knights, paid in the fifth and eighth years on exactly thirty; that of Malet, though similarly given as twenty and one-sixth in the carta, is returned in the Testa de Nevill as exactly twenty;¹⁷⁹ that of Beauchamp of Hacche, though distinctly given as seventeen in the carta, will be found, on careful collation of the rolls for 7 and 8 Hen. II, to be claimed by the exchequer as 17 + 3, i.e. 20.

Here also, perhaps, it may be allowable to glance at the foreign parallels to fiefs of sixty fees and smaller multiples of five. There is a charter of Charles the Fair (1322-28) 'qua Alphonsum de Hispania "Baronem et Ricum Hominem" Navarræ creat; et, ut Baronis et Rici Hominis statum manu tenere possit, eidem de gratia speciali 60 militias [knight's fees] in regno sua Navarræ concedit modo consueto tenendos et possidendos',¹⁸⁰ while an edict of earlier date proclaims: 'De Vasvassore [i.e. baron] qui quinque milites habet, per mortem [? pro morte] ejus, emendetur 60 unciæ auri cocti, et per plagam [? pro plaga] 30, et si plures habuerit milites, crescat compositio sicut numerus militum.'¹⁸¹

IV. THE TOTAL NUMBER OF KNIGHTS DUE

'Ad hoc solicitius animum direxi ut per regna Angliæ debita Regi servitia militaria quatinus potui plenissime percunctarer.'¹⁸² So writes Swereford, who proceeds to explain that neither the famous Bishop Nigel himself, nor his successor, Bishop Richard, nor William of Ely (ut supra) had left any certain information on the subject; while he (Swereford) could not accept the common belief that the Conqueror had created servitia of knights to the amount of 32,000.¹⁸³ The cause of his failure is found in the fact that he confused two different things: (1) the debita Regi servitia, which formed the only assessment of fiefs down to 1166; (2) the assessment based on the cartae of 1166, which superseded the debita servitia, and is not evidence of their amount.¹⁸⁴ But then, as I have already explained above, the exchequer official was concerned only with the actual claims of the crown; for him the original 'service due' had a merely academic interest.

There are two estimates for the total of which we are in search. One is 32,000 knights; the other 60,000.

'Stephen Segrave,' Dr Stubbs reminds us, 'the minister of Henry III, reckoned 32,000 as the number' (which confirms Swereford's statement); but he himself wisely declines to hazard 'a conjectural estimate',¹⁸⁵ adding that 'the official computation, on which the scutage was levied, reckoned in the middle of the thirteenth century 32,000 knights' fees, but the amount of money actually raised by Henry II on this account, in any single year, was very far from commensurate'. Gneist repeats this figure, but holds that 'as far as we may conjecture by reference to later statements, the number of shields may be fixed at about 30,000'.¹⁸⁶

On the wondrous estimate of 60,000 I have more to say. Started by Ordericus,¹⁸⁷ this venerable fable has been handed down by Higden and others, till in the Short History of the English People it has attained a world-wide circulation.¹⁸⁸ Dr Stubbs has rightly dismissed the statement 'as one of the many numerical exaggerations of the early historians';¹⁸⁹ but neither he nor any other writer has detected, so far as I know, the peculiar interest of the sum. What that interest is will be seen at once when I say that Ordericus, who asserts that the Conqueror had so apportioned the knight-service 'ut Angliæ regnum lx. millia militum indesinenter haberet' (iv. 7), also alleges that the number present at the famous Salisbury assembly (1086) was 60,000. It is very instructive to compare this 'body whose numbers were handed down by tradition as no less than sixty thousand',¹⁹⁰ with the 'sixty thousand horsemen'¹⁹¹—'ut ferunt sexaginta millia equitum'—of thirteen years earlier, and with the number of the Norman invaders, 'commonly given at sixty thousand',¹⁹² of seven years earlier still. It is Ordericus, too, who states that the treasure in Normandy at the death of Henry I was £60,000. His father seems to have left behind him the same sum at Winchester, for, though the chronicle left the amount in doubt, 'Henry of Huntingdon,' Mr Freeman observed, with a touch of just sarcasm, 'knew the exact amount of the silver, sixty thousand pounds, one doubtless for each knight's fee'.¹⁹³ He also reminds us, as to the crusade of William of Aquitaine, that 'Orderic allows only thirty thousand. In William of Malmesbury they have grown into sixty thousand. Figures of this kind, whether greater or smaller, are always multiples of one another'.¹⁹⁴

Pursuing the subject, we learn from Giraldus that the Conqueror's annual income was 60,000 marcs.¹⁹⁵ Fantosme speaks of marshalled knights as

Meins de seisante mile, e plus de seisante treis,

and the author of the Anglo-Norman poem on the conquest of Ireland gives the strength of the Irish host, in 1171, as 60,000 men. Even 'Sir Bevis', if I remember right, slew in the streets of London 60,000 men; and Fitz Stephen asserts that, in Stephen's reign, London was able to turn out 60,000 foot.¹⁹⁶ It may, also, not be without significance that 60,000 Moors are said to have been slain at Navas de Tolosa, and that William of Sicily was said to have bequeathed to Henry II three distinct sums of 60,000 each.¹⁹⁷

The fact is that 'sixty thousand' was a favourite phrase for a great number, and that 'sixty' was used in this sense just as the Romans¹⁹⁸ had used it in classical times and just as Russian peasants (I think I have read) use it to this day. The 'twice six hundred thousand men', who were burning to fight for England,¹⁹⁹ and the £180,000 (60,000 × 3) of Gervase (1159), are traceable, doubtless, to the same source.

How strangely different from these wild figures are the sober facts of the case! The whole of the church fiefs, as we have seen, were only liable to find 784 knights, a number which, small as it was, just exceeded the entire knight service of Normandy as returned in 1171. As to the lay fiefs it is not possible to speak with equal confidence. I have ventured to fix the approximate quota of 104 (more or less), of which ninety-two are in favour of my theory: forty-eight fiefs, of five knights and upwards, remain undetermined.²⁰⁰ If the average of knights to a fief were the same in the latter as in the former class, the total contingents of the lay barons would amount, apparently, to 3,534 knights; but, as the latter one includes such enormous fiefs as those of Gloucester and of Clare, with such important honours as those of Peverel and Eye, we must increase our estimate accordingly, and must also make allowance for fiefs omitted and for those owing less than five knights (which are comparatively unimportant).