SOURCE: Bell, Alexander. “The Royal Brut Interpolation.” Medium Aevum XXXII, no. 3 (1963): 190-202.

[In the following essay, Bell suggests that a manuscript of Wace's history of England, the Roman de Brut, called the Royal Brut contains an interpolation of some six thousand lines not written by Wace.]

Although the item commencing on f. 40^v of B.M. Royal 13 A xxi is introduced by the rubric:

Ci commence le brut ke maistre / Wace translata de latin en / franceis de tuz les reis ke / furent [en] bretaigne deske il / perdi son nun e fust apelé / engletere par la grant destruci / un ke daneis firent en la / terre,

it has long been known that from l. 52 onwards another translation of the Historia Regum Britannicæ has been substituted which continues for some six thousand lines, when the text of Wace's poem is taken up again. Already in his edition of the Roman de Brut (lxxii-iv), Le Roux de Lincy had noted the divergence and had printed a short specimen extract. Subsequently, in his Catalogue of Romances (I 264-5), H. D. L. Ward repeated this information and later still R. Imelmann in his Laȝamon: ein Versuch (Berlin, 1906) used this version, which he regarded as in some way representative of Gaimar's lost Estoire des Bretuns, to support his claim that a conflation of Wace and Gaimar was the source of the English poem and quoted a number of lines and short passages. More recently J. S. P. Tatlock, in his The Legendary History of Britain (Berkeley & Los Angeles, 1950) discussed this interpolation along with other minor French versions of the Historia, demolished Imelmann's claim, but acknowledged that he only knew our text from what had appeared in print. Thus the way seems open for a fuller study.

Our MS is a composite one in three parts, of which only the second one concerns us here. It is generally ascribed to the late thirteenth century and begins with the Imago Mundi which was at Hagneby Abbey (Li.) in the fifteenth century and may originally have been distinct from what follows—the ruling is different. Then comes on f. 40^v a ‘platte’, followed on f. 40^v in the second column, the first being blank apart from a couple of scribbles (? dates), by Wace's Brut which is interrupted on f. 41^v by our text and taken up again on f. 77^v; the Brut continues to f. 113^v and is followed by the Estoire des Engleis which concludes on f. 150^v. The section containing our text has been trimmed to fit its present binding; some of the headings and of the entries in the outer margins have been cropped and only one catchword survives. The ‘platte’ seems designed as an introduction to what follows; though more appropriate to Gaimar than to Wace, it would better fit the Description of England which actually follows the Brut in one MS (B.M. Addl. 32125). Nevertheless, it is not without interest. In a central circle are given the dimensions of Anglia, the width being reckoned from St. David's to Dover; round this circle are ranged seven others, each assigned to one kingdom of the Heptarchy; the whole is surrounded by a double circle in which are the names of the cardinal points, east at the top, and of the seven kingdoms. Beginning with Essex at the top, and proceeding clockwise, we have Kent, Sussex, Wessex, Mercia and Northumbria, but instead of the expected East Anglia we have Cantebrigge! In each of the seven circles are the names of counties and bishoprics of the kingdom. Under ‘Cambridge’ are Cambridge, with see at Ely, and Norfolk, with see at Norwich, formerly at Elmham or at Thetford, but no mention of Suffolk; under Wessex are Wiltshire and Berkshire, with see at Salisbury, formerly at Ramsbury or at Sherborne; under Northumbria is all the land beyond Humber usque ad mare Scoticum, with sees at York, Durham and Carlisle, called Carlionensis; the latter was founded in 1133, lapsed in 1156, and was revived in 1219. There is some connection, most probably indirect, between our ‘platte’ and an AN one at the College of Arms (Box 29, No. 10), described as No. 65 in the Catalogue of the Heralds' Commemorative Exhibition 1484-1934 (London, 1936) and illustrated there on Plate XXXIX. In the latter East Anglia is also omitted from the border, but Essex is duplicated; sees are usually not mentioned. Under Sussex both ‘plattes’ include Sussex, Hampshire and Cornwall! In the AN ‘platte’ Northumbria includes Northumberland as far as the mer d'Escoce; under Mercia our ‘platte’ includes, among others, Chester, Derby, Stafford, with sees at Coventry and Lichfield, the AN ‘platte’ includes, among others, Chester, Derby, Stafford, Coventry, Lichfield.

The pages of our MS are ruled in double columns, with at first 40 and then from 42 to 44 lines to the page, occasionally even more; f. 40^v has 40 lines of text in addition to the rubric. The sections of the poems are marked by two-line initials, in red and in blue, and subsections are indicated by a s in the margin; also in the margins are brief notes in Latin and in French indicating contents, though some on the inner margins do not show up on photostats owing to the tight binding; these, along with the headings are in the same hand as the text; hands in the margins here and there call attention to something deemed important; they may be due to the scribe as is the illustrative scheme of a legion in the outer margin of f. 65^v—VI m-VI c-LX-VI; at the lower outer corner of f. 77^v is a dragon, marking the birth of Arthur, but this is possibly a later addition; this is certainly true of one entry: on f. 58^v, where the text reads erroneously Austustus, is written August' in the margin. Our scribe had a written text in front of him; this is shown by self-made corrections: sometimes by expunctuation, e.g. f. 48^v—Ne ren fors sul sun cors sulement; sometimes by double accents to indicate a necessary transposition, e.g. f. 46^v—Geinz i vint sur Corineus; and once by the letters b a (f. 70^v) to show that two lines have been interverted. There are also a few breaks in the sense, sometimes coinciding with faulty rime-sequence, which point to omissions. The join at the beginning of the interpolation is also noticeable by the break in sense, though the same is not true of the other join; both occur in mid-column, so the insertion was not made to replace missing gatherings in our MS, though this may have been true of the one from which it is copied.

Our text, which I shall now refer to as Royal Brut (RB) is shorter than the corresponding section of Wace, so the question arises: is it a condensed version of that work? That there are verbal parallels between the two texts is undeniable, but the same is true of RB and the Munich Brut (MB) and in each case the explanation is the same: influence of the Latin original. Thus, when Wace has Brutus find in Greece Tute la lignee Eleni, Un des fiz al rei Priami (151-2) and RB calls them La lignee danz Elemi Ki fut fiz reis Priami (f. 41b), the proximity of the genitives in the Latin progenium Heleni filii Priami (I 3) in itself suggests the rime. Similarly, when MB tells us Iluec fu pris Antigonus Et ses cumpain Anacletus (591-2) and RB that Retenuz i est Antigonus E sis compainz Anacletus (f. 42c), the proximity of the names in the Latin again suggests the rime. The similarity between

Wace: Deus jors e une nuit sigelerent

(617)

          MB: Dous jurs cururent … E une nuit

(1123-4)

          RB: Duos jurs curent e une nuit

(f. 44b)

is amply accounted for by the Latin duobus diebus & nocte … concurrerunt (I 11). When, in the story of Lear, these three texts agree in their renderings of Et enim quantum habes tantum uales tantumque te diligo (II 11):

Wace: Tant as, tant vals e jo tant t'aim

(1742)

          MB: Tant as, tant vaus, e jo tant t'aim

(2880)

          RB: Tant as, tant vals e jo tant t'aim

(f. 48c),

it is clear that all three are making use of a current expression. These are typical examples and I do not recall one verbal parallel which compels us to assume the use of Wace by RB. Moreover, our author seems to adhere more closely to the text of the Historia than does Wace. To instance: Partholoim, referring to the duration of his voyage, says: Annus enim & dimidius iam emensus fuerat (III 12) which Wace renders by: An e demi avum erré (3287), but RB by An e demi ert ja passez (f. 52b); Androgeus after Cassibellaunus has entreated him to intercede with Julius Caesar, says of his uncle: Non est diligendus princeps qui in bello est mitis ut agnus in pace ferus ut leo (IV 9) which Wace renders by

Ne fait mie sire a preisier
Ki en pais se fait bald e fier,
Quant vient en guerre e en estur,
Si semble lievre de poür

(4735-8)

and RB by

ne fet a preiser cel seignur
Ki en pes est hardiz leons
E en bataille coard motuns

(f. 58a).

Further, there are indications that RB used a text of the Historia different in some details from that followed by Wace, but related in some way to that behind the Welsh version published in translation by Griscom in his edition of the Historia, from which I quote. To instance: of the organization of the church under Lucius we are told in

Historia: subiacuit metropolitano Eboracensi Deira & Albania

(IV 19)

RB: A Everwich tute Deirun Od tute Escoce e Bernicun

(f. 60c)

Welsh: to the Bishop-house of Kaer Efrawc belonged deifyr and brynaich

(pp. 329-30);

in answer to the summons of Asclepiodotus to the siege of London

Historia: Ad edictum itaque ipsius uenerunt Demeti & Venedoti Deiri &
                                        Albani

(V 4)

          Wace: Al comant Asclepiodot
                                        Vindrent Gualeis, vindrent Escot,
                                        De tutes pars Bretun i vindrent

(5527-9)

          RB: Donc i venent a grant fuison
                                        Li Gualeis e li Breton,
                                        La venent li Bernicun,
                                        Le Escot e li Deirun

(f. 61c)

          Welsh: men of S. Wales and men of N. Wales and the men of daifr and
                                        brynnaich and the men of Allan

(p. 335);

of the site of Vortigern's second castle we are told

Historia: in monte qui Cloartius nuncupatur

(VIII 2)

Wace: Ço fu desur Cloart, un munt

(7607) [two MSS give Doare Doar]

RB: Al mont Doard en est venu

(f. 72a);

after his repulse and retreat from the siege of York Uther Pendragun takes refuge on a mountain

Historia: ad montem Damen

(VIII 18)

          Wace: Danien li munz aveit nun

(8451)

                    RB: Icel mont a cel eé Diamned esteit apelee

(f. 76c)

          Welsh: mynydd Danned

(p. 420).

Finally, RB not infrequently has something which is in the Historia, but not in Wace. To instance: when Pandrasus after his defeat agrees to Brutus' terms, he says in RB: Mon frere des ore delivrez E en ostage mei retenez (f. 44a), which agrees with Historia: ut securiores sitis uobis quasi obses manebo (I 11), but the suggestion is not in Wace; describing the erection of Billingsgate RB says: Al pé del mur nefs arivouent Ki de tutes terres venouent (f. 52a), which echoes Historia: subtus ad pedem applicantibus nauibus idoneum (III 10), but this detail is not in Wace; in the time of Claudius St. Peter came to Rome and, says RB:

En Egipte si enveiat
Sen Marc qu'il mult amat
Ki seinte euuangelie escrit …

(f. 59d)

agreeing with Historia: misitque Marcum eunangelistam in Egyptum predicare euangelium quod scripserat (IV 15), but Wace does not mention this. All the facts given above suggest that we should regard RB as an independent translation of the Historia, not as a shortening of Wace's poem.

Next arises the question: is RB by Gaimar or derived from his lost Estoire des Bretuns? Already in my edition of his Estoire des Engleis (ANTS XIV-XVI, 1960) I expressed the opinion (p. lxxvi) that to ‘judge by the few short extracts quoted by Imelmann it cannot, in its present form, be attributed to Gaimar’. I have since had occasion to study the whole text and as a result I can go further and assert that it is not his work nor, probably, derived from it, and shall now adduce some grounds for that assertion.

(i) There are in RB a considerable number of inexact rimes, many without parallel in Gaimar; to avoid repetition I propose to discuss them later with other rimes when dealing with the date of our text;

(ii) Our author is fond of extended groups of lines on the same rime; those of 4 lines occur quite frequently and those of 6 lines are nearly as common, there are some half-dozen instances of groups of 8 lines, but only with the rime é: ié, there are two instances each of groups of 10 and of 12 lines, again in é: ié, and one of 16 lines in é; though 4 line groups may be found in Gaimar, he has nothing like the longer tirades; (iii) RB uses the demonstrative adjective, mostly in the plural, almost as a definite article, e.g. Si se mucent par ces guastines Par ces runces, par ces espines (f. 61a), which Gaimar never does; (iv) RB uses que … que in a distributive sense, e.g. Mult en i out que morz que pris (f. 52c), a use unknown to Gaimar; (v) Our two texts also differ in their employment of adverbs in -ment: not only does RB use more of them (54, of which 30 are not in Gaimar) than are found in the Estoire (38, of which 14 are not in RB), but they are used more frequently; indeed, only three—altrement (5: 1), malement (6: 3), veirement (11: 3)—occur more often in Gaimar than in RB, whereas there is an impressive list of more frequent occurrences in the latter text—corteisement (12: 0), ducement (10: 0), ferement (20: 4), forment (25: 10), gentement (13: 3), hastivement (33: 3), ignelement (18: 3), longement (17: 5), priveement (17: 4), vilment (13: 1), vivement (12: 0);

(vi) Our two texts differ again with respect to certain uses of the possessive pronoun—in the singular as ‘wealth, possessions’, in the plural as ‘men’ in a military context, as ‘people’ in relation to a ruler. Sporadically, and less clearly defined, these uses occur in the Estoire, the only examples I have noted being:

Le mien pernez, tuez ma gent

(5632);

Descumfit fud, mais bien guarid,
Ne gueres del suen n'i perdit

(1801-2);

Plusurs de lur venud i sunt

(2893);

Tut le lur orent guasted

(3394);

in RB we find li mien once or twice in these senses, li nostre, li noz occurs three times as ‘men, people’, li lur twice as ‘wealth’, but very different is it with li son which occurs four times as ‘wealth’, eleven times as ‘people’ and no less than forty-six times as ‘men’; examples of the latter are:

Brutus veit s'enseigne criant
E les sons ben amonestant

(f. 42a);

Julius Cesar dunc s'escriat,
Les sons entur sei reliat

(f. 55a);

Henges les sons donc apelat
En son language

(f. 72b);

in fact for the author of RB the term is an indispensable adjunct to a battle description;

(vii) In view of the approximately equal length of the two texts and the similarity of their subject matter, certain features of vocabulary are not without significance. To instance: ajuster is used eighteen times by RB, but once only by Gaimar; a la parfin is used nineteen times by RB, but only once, as en la parfin, by Gaimar; empur occurs seventeen times in RB, but not at all in the Estoire; herbergier is used eleven times by Gaimar with the meaning ‘shelter’ or ‘take shelter’, e.g. E Aveloc e sa muillier Vont a la vile herbergier (501-2), Cil herbergat tuz les seignurs (2595), but fourteen times in RB with the connotation of settlement and cultivation, e.g.

La terre vont dunc saisant,
Espessement la vont herbergant;
En bref terme l'ont si herbergé,
Cum tut dis i eussent esté

(f. 46d),

Issi fust Yrlande herbergee,
Einceis ert guaste e desherite

(f. 52b),

in addition to nine instances of the ‘shelter’ meanings.

The evidence already adduced would be sufficient, I submit, to cast doubt on the identity of authorship of the two texts even if they were contemporary. But were they? Before attempting to answer this question one feature of scribal usage in RB deserves mention: the large number of final dentals—as -t—surviving graphically in atonic 3rd singular endings. Altogether there are 27 instances, seven subjunctive presents, the remainder indicative presents; the proportion is rather high for a thirteenth-century MS and gains in significance when compared with the position in the Estoire, copied into this same MS by the same scribe, for of the 15 instances in that text he retains only two! This suggests lack of homogeneity in this respect in his source copy.

As RB is incomplete at beginning and end, we cannot know whether there was prologue and/or epilogue giving information about the author and even if there were, it does not follow that it would indicate the date of composition, as Wace did for his Brut. There are, however, three features in what has survived which point to a date c. 1200. (i) It is noticeable that when a king receives a missive, then in RB he reads it himself. To instance: when Brutus sends a letter to King Pandrasus, we are told of the latter that Li reis out lit le mandement (f. 41d) and again, when Julius Caesar sends messengers to Cassibellaunus, they deliver both an oral message and a missive and we are told that Li reis oit le mandement Ke Cesar fait tant ferement, Le bref lit oiant sa gent (f. 54a). Now in the Estoire (c. 1140) when William Rufus receives an urgent letter from the besieged garrison of Le Mans, Gaimar tells us that Li reis le prist, tost le fruissat, Randulf Flambart le bref baillat (5811-12) and in the Brut (1155) when Brutus sends the letter to king Pandrasus, Wace tells us that Li reis ad le bref esculté (253). Thus the author clearly takes it for granted that a king is literate and this was not generally true of English kings, whom he would naturally have in mind, until towards the end of the twelfth century. (ii) There is in RB a curious confusion with respect to Brutus' port of embarkation prior to his departure from Greece; neither the Historia nor Wace name it, but here it is said to be Mycenae and is equated with Messina! The text leaves us in no doubt:

A cel jur que fust nomé
Les galees sunt assemblé
A la cité de Miceine
Dunt jadis fut raine Heleine.
Cil qui meinent en Seizille
Entr'els l'apelent Meschine.
Iloc si curt un braz de mer
Ke Far l'ai oi numer

(f. 44a-b)

and a little later we are told:

Del rei de Grece unt pris cungé,
Enz en lur nefs sunt entré,
Amont le Far s'en sunt turné

(f. 44b).

At the time of the Third Crusade Messina was much in the news and the aftermath of that campaign seems eminently suitable for causing this confusion; again a pointer to the end of the twelfth century. (iii) Towards the end of RB we reach the fatal triangle—Gorlois, Igerne, Uther. Strange to say the first name is not to be found in our text; instead the duke of Cornwall whose wife attracts the king's attention is called Guerreheis, a name shown to be due to the author by its threefold occurrence in rime with reis. How the confusion arose we do not know, but that there is some reminiscence of the Arthurian Guerrehés seems certain and that was not likely to occur before the flourishing of the Matter of Britain towards the end of the twelfth century.

Further than this the external evidence does not take us, so we must now turn to the study of the rimes; though it is notoriously difficult to date an AN text on purely linguistic grounds, it may at least be possible to determine whether RB is contemporary with, or later than, the Estoire. First, however, a word must be said about the imperfect rimes in the two texts. They are not numerous in the Estoire and some, involving English names, e.g. evesque: Westsexe 1393, may be discounted, as Gaimar is clearly contenting himself with assonance. He is prepared on occasion to neglect an r in the rime, both before a consonant—dos: cors 5671, aprés: Waers 5715—and after—entrent: dementent 3261 and several rimes of -estre: -este; more frequently he rimes s and z [ts]—nefs: ez-2573, 2581, 5241, dis (dies): -iz 2963, dis (decem): fiz 4987, oscis: -iz 975, 1179, jurz: esturs 3027, sujurs: seignurs 6187, feiz: reis 921, purpens: adenz 179; and three times he neglects a final consonant after n—retint: -ing 1725; cinc: vint 4751, gent: -enc 5153.

When we turn to RB, we find the same types, but not the same distribution of occurrences; while rimes involving neglect of r before consonant are very common, there is only one instance of its neglect after—teste: destre (f. 73c); rimes of s and z [ts] abound, that of -is: iz occurring no less than 21 times; neglect of final consonant after n is found—estanc: grant (f. 71c), chamberleng: -ent (f. 53a, f. 69d) and champ rimes six times with -ant and once with sanc (f. 57d); there are 12 instances of -ins: -is, found only at 5967 in the Estoire, as well as fiz: Locrins (f. 47d),: Constantins (f. 62b), enginz: asailliz (f. 42d), vint: vit (f. 74a).

In so far as any conclusion can be drawn from the above comparison, we get the impression that RB is later than the Estoire. Other imperfect rimes in the former, without parallel in Gaimar, tend in the same direction. Thus we find rimes involving l and r—cruel: contreester (f. 51b), pel: eschaper (f. 70c), peus (=pels): trover (f. 55d), russelz: coverz (f. 54d); neglect of v before r—crere: decevre (f. 75c), envire: parsivre (f. 69d), fortune: aventure (f. 42d, f. 49a); and pudre: hure (f. 72b) on the one hand and pudre: fundre (f. 52a) on the other.

We can now proceed to discuss those rimes in RB which throw light on its date and for convenience I list them here.

• (i) É: IÉ. The confusion of these two sounds is absolute; in fact the spelling ie for the outcome of a + palatal and of free e is almost nonexistent.
• (ii) Ã: ˜E. This is general; dolant: enfant (f. 41b) and dolanz: enfanz (f. 51d), but four times dolent: -ent; talant: -ant occurs six times, but twelve times in rime with -ent.
• (iii) AI: EI. This is sporadic; moi: -ai (f. 63b), veneient: treient (f. 56a), Heleine: vileine (f. 62b), fait: esteit (f. 73c).
• (iv) AI: E. This is sporadic; fait: receit ‘refuge’ (f. 41c), repeire: terre (f. 52b) and three times faire: erre; the only similar rime for ei is aveir: guerreier (f. 52d) which could be due to suffix-substitution.
• (v) L. Effacement is suggested by vassals: Rududibras (f. 48a), chevals: dras (f. 49b), realme: blame (f. 69b) and vocalization by plout: vout (f. 67a), volt: desirout (f. 75d) and also, in view of pout: moniot (f. 67a) and sot(s.): moniot (f. 67c), by escult: mot (f. 71c).
• (vi) É: ÉE. There is no assured instance of this rime in our text, but a number of rimes between cité and past participles occur, where strict concord would require the latter to be feminine—Prof ert de une bone cité Ki Kaircradoc fust apelé (f. 74a); our author, however, does not always observe the rule and so we have, more probably, non-agreement, resulting in a rime é: é; the scribe, on the other hand, does accept é: ée, witness citee: apellee (f. 49d) and twice doné: contré ‘country’ (f. 69b).
• (vii) Analogical e in ind. pr. 1 of first conjugation verbs does not occur in RB; in rime we have aim: certain (f. 48c), otrei: mei (f. 68c), present: -ment (f. 41c); in the body of the text occur aim, cri, lo, pri, qui(d) and it is interesting to note that the scribe has nowhere thought it necessary to add a final e.
• (viii) In ind. imp. 3 and 6 of the first conjugation verbs the older endings are still in use; there is only one instance of the later analogical ending in rime—nomait: esteit (f. 68d); on the other hand there are two instances of the extension to verbs of other conjugations—enpleidout: destruout (f. 62b), arivouent: venouent (f. 52a).
• (ix) Instances of stressed terminations in normally feminine third persons plural are found—deisant: enfant (f. 71a), remuassent: quant (f. 74d), venissent: -ment (f. 75a), pousent: vengassent (f. 62d); sporadically in AN these stressed endings occur in the two forms (i) of -ant which seems to belong to the twelfth century and (ii) of -ent which seems rather to appear in the next century; in view of the confusion of ã and ˜e in RB it is most likely that we have really only one form, in spite of the two spellings; two other possible instances occur in our text: (a) Mult egrement les asaillent E tut dis plus e plus i perdent (f. 57c), (b) Pur iço la paene gent Cum berbiz vus en pernent (f. 65d) where the parallel couplet—Li paen les vont dechasçant, Cum berbiz les vont pernant (f. 66b)—suggests the correction vus en [vont] pernent.

The evidence here adduced is sufficient, I submit, to prove that RB is later than the Estoire, so that Gaimar cannot have been its author; it also warrants the belief that we have to do with a thirteenth-century text, of the first, rather than of the second, half of that century.

The text itself is not without interest; if no masterpiece, it is not entirely devoid of merit as narrative; the versification is far from regular, but here and there, it is to be suspected, a line has suffered at the hands of the scribe. For some reason, to us unknown, the author passes from the death of Cunedagius to the time of Donwal Molmutius, thus omitting all mention of the fratricidal strife of Ferrex and Porrex, and is content with the following summary:

Cum Cunedage fut tuez,
Le regne fust entremellez,
Departi fust entre cinc reis,
Chescon tint diverse leis.
Celes leis tindrent lur enfanz
Enprés lur tens plus de. canz.

(f. 49d);

there is a similar chronological indication referring to the period following the death of Elidurus:

Emprés lui furent. xxxv. reis,
Chescon d'els tint diverse leis.
Cinc cenz anz tindrent cil la terre,
Li uns od peis, li altre od guere

(f. 53b)

and then follows the long string of royal names taken from the Historia.

The two quotations last given show one feature of the author's style—repetition of phrases; we get the impression that once he has been attracted by a phrase, he is loth to let it go: thus he uses or e argent, a collocation which occurs in the Historia, fifteen times and a further five times with slight modification; the description of Brutus' activities in France—Citez asaillent, abatent turs, Pernent chastels, ardent burs (f. 45c)—is echoed in the description of Carausius' activities in Britain—Ardent viles e ardent burs, Pernent chastels, abatent turs (f. 61a); again, when Belinus learns of his brother's voyage to Norway, Si ad seisi ses citez, Ses chastels e fermetez (f. 50a) and this same collocation is used on five other occasions.

In his enumeration of combatants the author is, of course, influenced by contemporary political conditions; Goffarus Pictus of the Historia has become Gofforins li Peitevins (f. 45a), so it is not surprising to find among his followers li Peitevin, li Gascon e li Angevin (f. 45b), but when that monarch gathers his forces to attack Brutus and they comprise Franceis, Flamans, Alemans, Angevins, Manseis e Normans (f. 45d), we are inclined to suspect influence of the chansons de geste as well; this would certainly appear to be the case later, when Julius Caesar assembles his forces for a renewed invasion of Britain:

Il out od sei Moridiens
E Pincenaes e Indiens,
Il out od sei les Macedons
E d'Affrike les barons,
De Arabie e de Romenie,
E tuz icels de Hungrie,
Il out od si les Burgoignons
E tuz icels deça les mons

(ff. 55—56a).

One feature of the Historia, especially in its earlier sections, is the attention paid to the foundation of cities and to the identification of sites; these naturally reappear in the translation, but do not always agree in every detail with Geoffrey of Monmouth's account. In such cases it is almost impossible to distinguish between an original addition by the actual translator and one inherited by him from some earlier commentator. Thus when RB comes to the reign of Leil, the author records, in common with the Historia, the foundation of Kaerleil and adds:

Icist si fist Esecestre
E la cité de Porecestre,
Puis si fist citez plusurs
E les clost de riche murs

(f. 48a);

but on what authority? A little further on he comes to the cities founded by Rudhudibras; according to the Historia they were Canterbury, Winchester and Shaftesbury, but, says RB, Cil fist faire Kantorebire E Wincestre e Salesbire (f. 48a); Wace in his translation says: Cist fist Wincestre e Cantorbire E le chastel de Cestrebire (1613-14)—this latter name must be a scribal corruption of Schaftesbire—but one MS gives Salebiere; here RB would appear to be following a predecessor. This seems to be the case in another instance. We are told in the Historia that after a battle Constantine was crowned at Silcestria, but RB says:

Costentin unt puis honuré
En Cirecestre la cité,
Le realme lui unt duné,
De corune d'or l'ont coruné

(f. 66d);

Wace places the coronation at Cilcestre (6437), but three MSS give Cirecestre and the Harlech MS of the Historia reads Cerestriam.

This same difficulty arises in dealing with some of the identifications in RB. There Cassibelaunus gathers his forces to oppose Julius Caesar:

Idonc sunt tuit assemblé
En Dorobelle la grant cité,
Issi out nun a icel tens,
Dovre l'apelent noz parenz

(f. 54b),

which reads like an addition by the author, but when later he says with reference to Hengist's retreat to Conisborough:

A son chastel donc s'en turnat
Ke Kairconan apelat,
Conengesburc nus l'apelum
Ki enz el pais manum

(f. 72d),

he is simply echoing the Historia: petiuitque opidum Kaerconan quod nunc Cunungeburg appellabatur (VIII 5). Another identification in RB, no matter to whom it is due, is of interest: after Aurelius' disastrous attack on York he is driven ad montem Damen (VIII 18) of which RB tells us:

Icel mont a cel eé
Diamned esteit apelee,
Meis nus qui el pais manum
Windegates ore l'apelum.
El soverain est un coudrei
E el miliu un grant perrei,
Desuz si curt un russellet
Ke nus apelom Coquet

(f. 76c);

as I have already pointed out (p. 193) the ‘British’ name is closer to that in the Welsh version than to the Latin, and it looks Celtic, but the hazel thicket and the rocks come direct from the Historia. I am indebted to Professor J. C. Maxwell, of King's College, Newcastle, for confirming that there is actually still a Windygate Hill, not shown on the 1″ OS map, but found on others, near the source of the River Coquet; it is the peak (2034) on what is marked as Windy Gyle.

Whether the author of RB is responsible for the assertion concerning Bath that Bladud les bainz atemprat Od le suffre que mis i ad (f. 48a) we do not know, but perhaps he is to be credited with the following little touch anent Julius Caesar's first sight of Britain:

Com estut a Witsand sur rive de mer,
Vers Bretaigne prist a esgarder.
En Dovre vit blanc dras venteler,
As paisanz prist a demander:
‘Ore me dites, seignurs amis,
Quele terre ço est e quel pais
U jo vei ces dras blancheier
Tut dreit utre cele mer’

(f. 53d).

One last question remains for discussion: has any other part of this translation of the Historia survived? Already in his Legendary History of Britain Tatlock suggested that the Harley fragments published by Imelmann might be a later part of RB. In view of the small amount of text available (c. 250 lines) it is doubtful whether definite proof is possible, but there are indications which make his suggestion plausible. There is confusion in rime of é and ié—[he]rbergié: cité (II 11); of s and z—periz: oscis (III 11); there are rimes showing neglect of r before consonant—nefs: chevaliers (III 3), Escoz: esforz (IV 3); there is use of distributive que … que; the possessive pronoun is used in the sense ‘men’—Des soens i ad forment perdu (II 2); there is the line De corone de or le ad co[roné] (I 13), with which compare in RB: De corune d'or l'ont coruné (f. 66d) and four times repeated De corune d'or se corunat, but all after RB has told how Donwal Molmutius Corune de or fist aprester E si se fist reis coruner (f. 49d); and the promise given by the archbishop of Caerlion in his address before battle with the Saxons—tuit cil qui ci morunt Tut pur veir a Deu irrunt (IV 21)—is more reminiscent of archbishop Turpin in the Chanson de Roland: Se vos murez, esterez seinz martirs, Sieges averez el greignor pareïs (1134-5) than of the Historia: Si igitur aliquis uestrum in hoc bello subierit mortem, sit ei mors illa omnium delictorum suorum penitentia & ablutio (IX 4).

I would, however, go further and suggest that the Description of England, published in the Rolls' Series edition of Gaimar's Estoire des Engleis, could be another stray fragment of RB. It is based on Henry of Huntingdon's Historia Anglorum I §§ 4-7, but has clearly been designed to fit into a ‘Brut’, hence the references in it to Corineus qui les jeanz enchaça (104) and to Belin as the royal roadmaker, both of them additions to Henry's account. Once more the shortness of the text (260 lines) makes definite proof almost impossible, but, despite the contradictory identification of Dorobelle with Canterbury, the following indications support the suggestion. There is confusion in rime of é and ié—apeled: sied (17), of s and z—ocis: saisiz (215); there are rimes showing neglect of r before, and after, consonant—legistres: sistes (37), recetez: tierz (189); there is the use of distributive que … que; the rime in En Dorobelle la cité Que Cantuorbire est appellé (73-4) is ambiguous like the similar ones in RB; there is the line a son os lur onurs prist (56) with which compare le regne a son os prist (f. 61b, f. 62a) in RB; there is the use of herbergier ‘settle’—Cest Bade ot jadis altre nun Si cum dient li Saisun Qui primes la herbergerent (93-5); and there is the couplet—E furent saisi des citez, Des chastels e des fermetz (3-4), the collocation which is found no less than six times in RB.

It will be recalled that earlier (p. 190) I expressed the view that the ‘platte’ in our MS was more appropriate to the Description of England than to either Gaimar or Wace; if now that text is a part of RB, then it would be a legitimate conclusion that the ‘platte’ came into our MS from an earlier copy of RB rather than from one of Wace-Gaimar. Does by any chance the extended reference to Ely—again an addition—in the account of Cambridgeshire—

De Ely i est l'evesquié,
En un mareis siet la cité,
Cil qui la maint ad grant fuisun
Suventesfeiz de bon peissun
E volatille e veneisun,
Dedenz le mareis le prent l'um

(115-20)—

point in that direction?