Odo de Carew

THE ORIGIN OF THE CAREWS It may have been observed by some that in the Fitzgerald pedigree which I gave at the close of my paper on the origin of that family,1 I followed the example set by Mr. Dimock, when editing Giraldus Cambrensis (Rolls Series), in omitting William Fitzgerald's eldest son Odo, who is claimed as the Carews' ancestor. This I did for two reasons. In the first place Giraldus himself, though naming Odo de Kerreu ' as his cousin (consobrinus)2, nowhere states distinctly who his father was, while he somewhat pointedly ignores him in the Expugnatio Hibernite ; in the second the charter of 1212, restoring Moulsford, Berks, 3 on which the origin of the family has long rested in pedigrees, proves that Odo was a grandson of Gerald Fitz Walter (de Windsor), but does not tell us which of Gerald's sons was his father, or indeed prove that he was not merely a maternal grandson. It was safer, therefore, to leave the descent open until it could be absolutely proved.4 Fortunately I have, since then, noted the missing link needed to complete the proof. In the Monasticon (vi. 837), among the endowments of the Hospitallers' Commandery at 'Walinton,' we find this entry : Willelmus filius Geraldi et Odo filius ejus dederunt totam villam de Rubard cum omnibus pertinentiis.5 The place is Redbarth, a parish adjoining that of Carew, and the entry is proof positive that Odo was a son, and indeed the heir, of William son of Gerald Fitz Walter.6

1 The Ancestor, ii. 98. 2 Gerald, who prided himself on his Latinity, may have used the term in a strict sense. 3 See p. 24 below. 4 Sir Harris Nicolas, in his Roll of Carlaverock, speaks of the house of Carew as supposed to have sprung from Otho de Windsor, the common ancestor of the illustrious families of Windsor and Fitz-Gerald '(p. 154), and Mr. G. T. Clark similarly treats the connexion as open to question (Medieval Military Architecture, i. 116). 5 In the confirmation charter by Anselm, Bishop of St. David's (1230-1), the place is given as ' Redeborth.' 6 Walinton,' the site of the Commandery, was East Walton, Pembrokeshire, some twelve miles north of Carew Castle.

With the help of this evidence it will now at length be possible to prove and illustrate the pedigree throughout the twelfth century, the darkest and most difficult period in genealogical research. The pivot on which the story turns is that singular district around Pembroke, that ' little England beyond Wales,' which was destined to form the stepping-stone between England and Ireland. Of this district, with its strange place-names still preserving the memory of Norman or Flemish knights, Pembroke was of course the head; and of Pembroke the constable was Gerald, the patriarch of a spreading race. The neck of the Pembroke peninsula was guarded by Carew on its northern coast and by Manorbier on its southern, and these castles came to be held by grandsons of the lord Gerald.1 A puzzling passage in the work of Giraldus, De rebus a se gestis, relates that, while he was at school (i.e. attending the university) at Paris as a young man, consanguine! ipsius, sc. Willelmus filius Hay,2 Odo de Kerreu,3 et Philippus de Barri,4 frater ejusdem decimas suas . . . longe ante contulerant (i. 28). This would at first sight suggest that Odo ' de Kerreu ' was a brother of Philip de Barri ; but the important genealogical passage two pages earlier clears up the matter. After mentioning that * Ricardus filius Tancardi * was a great man in Pembrokeshire, and that he hated Gerald himself and all his folk, Gerald continues: Odo de Kerreu consobrinus Giraldi et Philippus de Barri frater ejusdem Giraldi,5 qui viri probi et magni fuerunt in finibus illis, licet generi praedicti Ricardi, sc. filias suas habentes uxores, tamen acerbe dixerunt illi quod taceret et a stultiloquio temperaret ; quia non tanta vindicta sumpta fuit de alio Giraldo, fratre sc. Odonis primaevo, pro quo dudum a Rosensibus interempto ducenti vlri et plures de eisdem uno die corruerunt (i. 26-7).

In another of his works, the Itinerary of Wales Gerald recurs to this tragedy on the occasion of his visiting Camrose.6 Kamros, ubi, pro juvenis egregii, Giraldi scilicet filii Guillelmi, nece, multorum caedibus cruentam nimis et gravem, tempore Stephani regis, propinqui

1 The Barrys of Manorbier, descended from Gerald through his daughter, recur in connexion with their neighbours the Carews, both in Pembrokeshire and in Ireland, generations later. 2 Of Hay's Castle (?) 3 Of Carew. Of Manorbier. 5 This supplies the missing word in the previous quotation. 6 To the north-west of Haverfordwest.

et necessarii sui, quanquam minus in hoc necessarii, vindictam in Rosenses exercuerant.1 The death of William Fitz Gerald is placed by his nephew the historian in or about 1174 in a passage which, explaining that it recalled his son Reimund to Wales, would rather suggest, as I hinted at the outset, that Reimund was his heir.2 We now know, however, that this was not the case. Putting together the above evidence, it proves this pedigree

Here I may explain that, having thus satisfied myself of the pedigree, I received from Dr. Owen a copy of his Old Pembrokeshire Families, which he was good enough to send me. This work, which is described on the title-page as compiled in part from the Floyd MSS.,' is a valuable contribution to early genealogy. I glanced at Mr. Floyd's piles of notebooks when they were deposited at the London Library, and saw that they seemed to 'contain,' as Dr. Owen observes, 'a wealth of information as to the families of other counties.' That careful and laborious antiquary' adopted the excellent method of arranging his notes under the names of the families to which they referred. Whether Mr. Floyd was as critical as he was certainly industrious it would not be possible to say without study of his notes; but his collections, now deposited unfortunately for London genealogists in the College Library at Aberystwyth, deserve to be widely known. With regard to Odo de Carreu there are two points, I fear, on which I must differ from Dr. Owen. In the first place he gives as Odo's mother, Katherine, a daughter of Sir

1 Itinerarium Kambriee (Rolls ed.), vi. 99. 2 'Reimundus, ob patris quern audierat obitum, nobilis videlicet viri Guillelmi Giraldidae, remenso pelago in Kambriam recessisset' (v. 310).

Adam de Kingsley, in Cheshire.' This match, I venture to say, is obviously hall-marked as one of those I spoke of in my Studies in Peerage and Family History, where I observed that a Lane, 'under William Rufus, married into one of those "leading families" whose daughters have always been kept in stock at Her Majesty's College of Arms ' (p. xv.). In other words, the marriage must be taken from an old herald's pedigree. The good 'Sir Adam' would have lived about the year 1100, and maybe fitly compared with that Sir Richard Stackpole of whom Dr. Owen writes : There is in the writer's possession a print of one 'Sir Richard Stackpole of Pembrokeshire,' stated (although his looks belie it) to have been 'highly respected in the year 1091.' There is beneath the print a long and entirely inaccurate account of Sir Richard and of his descendants. Sir Richard had no existence. The second point on which I must differ is that Odo de Carreu is always styled 'Odo' (or ( 'Oddo'), and not as Dr. Owen makes it, 'Other,' which was the wholly distinct name of his first known ancestor. A good instance in point is afforded by a charter with which Dr. Owen seems to be unacquainted. This is the confirmation by Peter, Bishop of St. Davids, soon, I think, after 1176, of the dapiferatus of that see to William son of Maurice Fitz Gerald.1 Among the witnesses are : Ricardo de Haerford ; Tanchard filio ejus ; Oddone de Karreu . . .Philippe de Barry.

This is a specially interesting combination, for we here see the three men mentioned together by Giraldus in a passage quoted above, namely Richard (Fitz Tancard, constable) of Ha(v)erford as I should extend the name and his two sonsin- law Odo de Carreu (of Carew) and Philip de Barri (of Manorbier), together with his son 'Tanchard.' As Richard was a military tenant of the see he appears very fittingly among the witnesses to the charter. The above charter may be fittingly followed by that of Robert son of Elidir (of Stackpole) granting Trefduant (St. Edrin's) to St. Davids, for the lay witnesses to that charter are Odone de Carren (i.e. Carreu), Philippe de Barry, Adam de Rupe.'2 According to Dr. Owen's book * Other (sic) soon

1 Fourth Report Historical MSS. App. p. 583. 2 Late transcript in Harl. MS. 1249, fo. 28 (pencil). I am indebted to Dr. Owen's book for this reference. Adam 'de Rupe' was then of Roch Castle.

after his father's death ['1173'], got into trouble with the Welsh, who took from him his castle of Emlyn (sic), but he obtained from Henry II. the manor of Bampton, co. Oxon, so long as the Welsh held Emlyn' (p. 1 3). The reference given for this statement is 'Pipe Rolls, 2 Henry II. [i 156],' but there is not, and could not be, any such entry on the roll of that year. The entry required is on the Pipe Roll of 20 Henry II. (i 174) and refers, not to 'Bampton, co. Oxon,' but to Braunton, co. Devon, under which county it is found. It runs thus : Et Odoni filio Willelmi filii Geroldi (sic) c solidos in Branton' cum pertinentiis de quarta parte anni et amodo xx libras numero per annum in eadem villa in escambio castelli et terra de Emelin quamdiu Resus 1 filius Griffin ea habuerit (p. 89).

Apart from its genealogical value this entry proves that Odo was granted £20 a year from the royal manor of Braunton at Midsummer, 1174, while it implies that he had lost possession of Emlyn. The cantred of Emlyn was a district lying along the left (the south) bank of the Teify above Cardigan. It must have been held by Odo's father, for we find him confirming the gift of a church within it.'2 This leads me to suggest a bold emendation of the printed text of Giraldus, who is made to assert that of the seven cantreds obtained by the children of Nesta, William received Pembroke and 'Ginelin.'3 As there is no cantred of 'Ginelin,' I think we should read ' Emelin.'4 As for 'castellum de Emelin,' it must mean the castle of the district (?Kilgerran). Odo was still drawing from Braunton his £20 a year in 1189,5 but a curious entry on a plea roll of uncertain though later date reads as follows:

Willelmus Peche positus loco Odonis de Karliun (sic) venit in curiam et concessit Sibille de Sumeri c solidatas terre in Chause et faciet ei escambium de feodo i militis in Bramton pro feodo i militis in Emelin unde ipse Odo cepit escambium pro iiij marcis quas ipsa dedit ei.6

1 This was the celebrated Rhys, prince of south Wales, who was at this time on the side of Henry II. 2 In Emlyn ex dono Jordani de Cantitona et confirmatione Willelmi filii Geraldi ecclesiam de Castellan '(Fenton's Pembrokeshire, p. 64). 3 Ed. Rolls Series, i. 59. 4 in ' was easily misread for ' m ' so that only the initial letter requires alteration. 5 Pipe Roll, i Ric. I. 6 Rot. Cur. Reg. i. 374.

We hear again of his land at Braunton in 1201, where he still appears on the Pipe Roll as holding it.1 In this year on 5 January, John had granted to Robert de Secqueville all his rights in Braunton save the 'outhundred' and the land which 'Odo de Karun ' was holding.2 Robert is found, the following Michaelmas, accounting for 100 marcs due for the king's grant.3 Three years later (1204) the sheriff of Devon is ordered to give Robert, further, seisin of that portion of Braunton which Odo had held,4 a concession for which he is subsequently found paying the king 50 marcs and a palfrey.5

Odo was succeeded by his son William, who first appears in conjunction with his father in 1194." When we meet him again in 1207 he has succeeded his father.7 He paid in that year a large sum of money that he might not be impleaded for his land of 'Muleford' (i.e. Moulsford, Berks), which his predecessors, he said, had held since the days ofHenry I. The curtain rises again in 1212, when we find William restored to his 'house' and lands at Carew and to his manor of Moulsford by two documents which have not, I think, been brought together before. They were issued within four days of each

1 Rot. Cane. 3 John, p. 1 5 . It is interesting to observe that on this page he appears, as before, as * Odo filius Willelmi filius Geroldi,' though on p. 24 he is 'Odo de Carrio.' 2 Calendar of Charter Rolls, p. 83b. 3 'Robertas de Secchevilla reddit compotum de c marcis, pro habendo quicquid Rex habet in manerio de Branton' hereditarie excepto uthundredo et excepta terra quam Ode de Carrio tenet in eadem villa q[ue] potest devenire in man[um] Regis' (Rot. Cane. 3 John, p. 24). 4 * Rex . . . vicecomiti Devon' etc. Mandamus tibi quod facias habere Roberto de Sechevilla terram quam Odo de Carro tenuit de nobis in Branton' tenendam quamdiu nobis placuerit reddendo inde nobis annuatim xx libras. . . xix die Jan'(Rot. Lib. p. 77). 5 * Robertus de Sechevilla dat L marcas et I palefridum pro habenda terra que fuit Odonis de Carriou in Branton' reddendo domino Regi per annum xxx (sic) libras sterlingorum' (Fine Ro//[i2o6], p. 349). 6 * Odo de Karrio ponit Willelmum filium suum loco suo versus Gaufridum de Chausi de placito terre de Molesford ' (Berks, Mich. 6 Ric. I. ; Plac. Abbrev. p. I . Compare Rot. Cur. Reg. I. 20). The Chausi family gave name to Mapledurham 'Chausey' some seven miles lower down the Thames on the opposite bank. Emma (de) Chausi had given tithes at Moulsford to Wallingford Priory (Mon. Ang. iii. 280). 7 ' Willelmus de Carrou dat quadraginta uncias auri quod non implacitetur de terra sua de Muleford quam Gillebertus de Cause clamat versus eum per breve de recto, et quia antecessores ipsius Willelmi in pace tenuerunt a tempore Regis Henrici avi patris domini Regis et ipse usque modo, ut dicit' (Fine .&?//[ 1207], p. 414).

other, and are of great importance, for the first proves that William had been in possession of Carew in 1210, and the second establishes his pedigree from the days of Henry I., needing only the slight link which the grant of Redbarth supplies.1 (1). Rex dilecto et fideli suo Falkesio ballivo de Glanmorgan' etc. Sciatis quod reddidimus Willelmo de Carrio domum suum de Carrio cum terris quas habuit die quo venimus ultimo usque Penbroc' ad transfretandum in Hyberniam anno r. n. xij mo . . . xxj die Maii anno xiiij mo2 [21 May, 1212]. (2) Sciatis quod reddidimus et hac carta nostra confirmavimus manerium de Muleford' cum pertinentiis suis quod Henricus Rex avus Regis Henrici patris nostri dedit Geroldo filio Walteri avo Odonis patris predicti Willelmi de Carrio tenendum eidem Willelmo et heredibus suis de nobis et heredibus nostris per servicium I militis . . . xxv die Maii anno r. n. quarto decimo 3 [25 May, 1212].

It would seem that William de Carew obtained this reinstatement by paying a fine to the king, for although the record of it is now lost, it is referred to in that of the heavy fine by which a Somerset baron, William Fitz John de Harptree, seems to have obtained both these estates in the year following (17 Sept. 1213). The same baron is found ten or eleven years later (7 March, 1224) compounding with the king for the fine he had made with King John ( for having the wardship of the land and heir of William de Carew.' 5 Here then at least

1 See p. 19 above. 2 Calendar of Patent Rolls, i. (i) 92. 3 Calendar of Charter Rolls, i. (i) 186. 4 Willelmus films Johannis de Harpetre finem fecit eum domino Rege per quadraginta (sic) marcas et iiij palefridos pro habendis terris Willelmi de Karrio quas h[abe]t citra mare Hiberniae et pro jure quod predictus Willelmus de Karrio clamat de aliis terris citra mare Hiberniae pro quibus idem Willelmus de Karrio finem fecerat cum domino Rege per xl marcas reddendas domino Regi simul cum predictis iiij palefridis infra duos annos . . . et quod scire faciat per literas suas Vicecomiti Bercscir et Falkesio baillivo de Glamorgan in quorum bailliis predictus Willelmus terras suas h[abe]t etc.' (Fine Roll, 1 5 John, p. i, m. 3). The details of the fine prove the amount to have been not 40, but 400 marcs (266, 13s. 4d.). In the first volume of the Calendar of documents relating to Ireland, the reference is wrongly given as 'm. 6,' and 'citra' has been clearly misread as 'ultra,' the whole entry being thus made to relate to lands in Ireland! 5 Willelmus filius Johannis de Harpetre finem fecit cum domino Rege de cc et xx marcis quas domino Regi debet de fine quam fecit cum domino Johanne Rege pro habenda custodia terre et heredis Willelmi de Carrio '(Fine Roll, 8 Hen. III. p. 2, m. 8).

we have definite proof that William de Carew was dead before the close of John's reign. Having brought the pedigree clearly down to this point I might claim to have traced as far as needful of the origin of the Carews, for at this period the genealogist finds himself in smooth water with rolls of many kinds, fines and deeds yearly increasing in bulk. Oddly enough however it is here that difficulty begins, not indeed as to the descent, for of that there can be no question, but as to the details of the pedigree. It has been alleged that William de Carew was succeeded by his son Nicholas, then under age, in 15 Henry III. (I23O-I),1 but William we have seen was dead before Henry's reign, and as early as 1228 (11 July) we read on the Close Rolls: Dominus rex commisit Bertramo de Cryoil manerium de Molesford quod Nicholaus de Carrio de rege tenuit in capite ad se sustendandum in servicio domini regis quamdiu ei placuerit.2

This proves that Nicholas de Carew had before that date succeeded to Moulsford, and subsequent records make it certain that, as indeed the entry suggests, he was actually dead at its date. For the recently published Calendar of Patent Rolls (p. 203) contains the entry (20 Sept. 1228): Dominus rex concessit Bertram de Crioil manerium de Molesford cum pertinenciis quod Nicholaus de Karrio de rege tenuit in capite habendum sibi vel cui assignare voluerit usque ad etatem beredum ipsius Nicholai

And in 1230 (27 December) we read on the Fine Rolls: Johannes Marescallus finem fecit cum Rege per ducentas marcas pro habenda custodia terrarum et heredum (sic) Nicholai de Careho usque ad etatem eorundem heredum (sic) cum maritagio ipsorum heredum (sic). Et Mandatum est Vicecomiti Berk' quod, accepta ab eodem Johanne securitate de predictis cc marcis, de omnibus terris que fuerunt predicti Nicholai in ballia sua, etc.3

Four years later (21 Nov. 1234), John Marshal is excused the balance due from him for the wardship quam quidem custodiam postea reddi fecimus dilecto et fideli nostro Bertrano de Cryoil cui earn prius concesseramus per finem quam inde fecit nobiscum.4

Thus the wardship was restored to Bertram, who had secured Moulsford in right of it, we saw, in 1228. There

1 The Heralds' College Pedigree (C. I, 26), according to Maclean's Sir Peter Carew, App. i. 8 Calendar of Close Rolls, i. 6r. 3 Excerpta e rot. fin. i. 208. 4 Ibid. p. 296.

was clearly a long minority, for a return in the 'Testa de Neville which must be later by several years, contains the entry: Bertram' de Crioill tenet Mallesford ut wardam de ballio regis de hereditate fil' Eudon' (sic) de Karre nee facit scutagium (p. 125).

It is clear therefore that Nicholas de Carew, so far from succeeding his father in 1230/1, being under age at the time, was himself already dead in 1228, leaving, as the records put it, 'heirs,' a word which raises further doubts as to the correctness of the pedigree.1

Indeed, I am bound to point out that it is at present a blank until, some half a century later, we meet with Nicholas de Carew. So far as the Welsh inheritance is concerned, we certainly seem to be on sure ground with the mention of William de Carew's tenure of five fees in the division of the Marshal inheritance (i247),2 but unluckily, on verifying Dr. Owen's reference for the fact Clark, Earls of Pembroke p. 69'I can find nothing there about William or his five fees. Mr. Floyd however may have seen record evidence for the fact. And such evidence for a later date is duly found on the Close Roll of 1325, where we read that the purparty of Laurence de Hastings of Pembroke included 'five knights' fees in Carru, co. Pembroke, which John de Carru holds, of the yearly value of 100 marcs ; five in Maynerbier, in the same county, which John de Barry holds,' etc., etc.3 With Nicholas de Carew living at the close of the thirteenth century we are, in any case, on sure ground for both the English and the Welsh inheritance. This Nicholas, who is found as a witness in 1299 to a Pembrokeshire charter,4 was one of those who, in the parliament of Lincoln, affixed their seal to the letter to the pope, his style being given as 'Nicholas de Carru dominus de Muleford.' Moreover there is plenty of record evidence for his journeys to and from Ireland, where he held the barony of Idrone in the palatinate of Carlow

1 Dr. Owen, I observe, here gives a different pedigree from that which is accepted, and makes William (1212) father of Richard, who ' had an elder son, William Lord of Carew,' in 1 247. But he then proceeds to trace the descent through 'Richard's son, Sir Nicholas de Carew,' who 'died in 1311'(p.14). 2 See The Ancestor, i. 2479, for the value of such evidence. 3 Calendar of Close Rolls, 1323-7. The same record proves that the Stackpole holding was one of five fees. 4 Sloane Charter, xxxii. 14.

(to give it its modern name), 1 as '5 knights' fees, with the appurtenances in Odrone.' 2 Here we have further illustration of that system of the five-knight unit on which, as I have shown, in Ireland as in England, our knight service was based.3 From the kingdom of Jerusalem, indeed, in the east to the lordship of Ireland in the west the two realms which illustrate alike the flow and the ebb of conquest we trace the persistent presence of the five-knight unit. Pembroke and Carlow were both palatinates or at least quasi-palatinates and had both originally formed part of that vast inheritance of the Marshals of which Strongbow had laid the foundations. The Carews held in each of them a 'barony' of five fees, and, in my opinion, 'Baron of Carew,' a style they sometimes bore, was one of those interesting feudal tides which are found in such palatinates as those of Chester and Durham, though it seems to have been held that they were 'barons' also in virtue of their tenure of Idrone.4

At the famous siege of Carlaverock (July, 1300) Nicholas de Carew was present.

Un vaillant home et de grant los O lui Nichole de Kami Dont meintz foiz orent paru Li fait en couvert et en lande Sur la felloune gent dirlande Baniere ot jaune bien passable O trois passans lyons de sable.1

1 The Heralds' pedigree makes him acquire it by marrying the heiress of a mysterious 'Digon, Baron of Odrone,' whom I cannot identify.

2 Inquest of 11 Dec. 1306, in Calendar of Documents relating to Ireland, 1302-7, p. 179. Compare p. 173 for inquest of 8 April, 1307, in which he is said to hold 'one barony in Odrone.' After his death it was certified to the king by the escheator of Ireland, 20 May, 1317, that he had, by the king's licence, enfeoffed his son and heir John, in his life, of his lands in co. Carlow, with certain reservations to himself (Irish Close Roll, 10 Edw. II.). 3 Feudal England, p. 259-60. 4 'The country of Odrone, which was a barony, and parcel of the inheritance of the said Sir Peter, and sundry of whose ancestors had been barons of the same ... he was persuaded to have begun his suit against the Kavanaghs for the barony of Odrone, because the same was of his ancient inheritance, a great territory or country, and which carried the title of honour' (Hooker's Life of Sir Peter Carew [Ed. Maclean], pp. 73, 79). In Sir John Maclean's note (p. 2) Idrone itself is described wrongly as * a sort of palatinate formerly belonging to Margaret, Countess of Norfolk.' It was only a barony held of her palatinate or liberty, which was itself Catherlough (Carlow). Forth and St. Mullins were also baronies therein (see p. 48 below).

And half a century later his descendant Sir John de Carew was among the knights who fought in the king's division at Crecy.1 Against Irish felons,' against Scottish rebels, against the foemen of France, the black lions of Carew had gone forth in turn to war.