CASTLE-BUILDING IN ENGLAND.

607

Chartres (see vol. i. p. 455) besieged the castle which Fulk of Anjou had built as an entireixioμós against Tours ("contra civitatem Turonicam firmaverat") and "turrem ligneam miræ altitudinis super domgionem ipsius castri erexit." The donjon itself was surely of stone. Stone was also at this time fast coming into use for domestic as well as for military and ecclesiastical buildings. Avesgaud, Bishop of Le Mans (994-1036), rebuilt in stone both the episcopal palace and also an hospital; before him they had been of wood; "quæ antea ligneæ fuerant petrinas. . . constituit" (Gest. Ep. Cenom. ap. Mabillon, Vetera Analecta, iii. 300*). We have also the remarkable description in William of Poitiers (81) of the fortified house at Brionne; "aulam lapideam arcis usum pugnantibus præbentem" (see p. 261). This was plainly something different from the ordinary donjon, though it was capable of being put to purposes of defence. It was probably what would in later days have been called a crenellated house, and it is doubtless distinguished as "lapidea" because an "aula" would often be of wood while "arces" were of stone. So we twice read in Domesday (184 b, 187) of "domus una defensabilis" in Herefordshire, which was seemingly something different from a castle.

The building of castles seems to be always mentioned in our Chronicles with some expression of horror. Thus we read in Chron. Wig. 1066; "And Oda biscop and Wyllelm eorl belifen her æfter, and worhton castelas wide geond pas peode, and earm folc swencte and á syan hit yflade swide." So in Chron. Petrib. 1087; "Castelas he lét wyrcean, and earme men swide swencean." The famous description of the castle-building in the year 1137 is familiar to readers even of the commonest English histories. A speaking witness to the impression which had been made on men's minds by the building of this particular Richard's Castle, probably the first of its class in England, is given by its being spoken of distinctively as "the Castle," even by the Worcester Chronicler (1052; see p. 309), who had not spoken of its building in his earlier narrative.

NOTE T. p. 172.

THE SURNAMES OF WILLIAM.

IT has been pointed out by more writers than one that a certain amount of confusion is involved in the familiar description of the great King-Duke as William the Conqueror. He is not often called Conquæstor" by writers of or near his own time. Moreover, Conquæstor" hardly means "Conqueror" in the common use of that word, but rather "Acquirer," or "Purchaser," in the wider legal sense of the word "purchase." A former colleague of mine in the Oxford Schools always made a point of describing him as “William the Purchaser." But the title of William the Conqueror, even as commonly understood, is so familiar, so true, and so convenient, that I have not the least wish to interfere with its use.

As far as I can see, he is known to his contemporaries as William the Bastard, and was, after his death, most commonly distinguished from his successor by the name of William the Great. The title of Bastard indeed stuck so close to him that some writers, who could hardly have known what it meant, seem almost to have taken it for his real name. Even Adam of Bremen, who certainly knew its meaning, uses it almost as a proper name. He introduces William (iii. 51) as "Willehelmus, cui pro obliquo sanguine cognomen est Bastardus," and goes on to speak of "Bastardus victor," and (c. 53) to say how "inter Suein et Bastardum perpetua contentio de Anglia fuit." So Marianus Scotus, a. 1089 (Pertz, v. 559), talks of "Willihelmus, qui et Bastart;" Lambert of Saint Omer (Pertz, v.65) says, "Terra Anglorum expugnata est a Willelmo Notho Bastart;" and most curiously of all, Lambert of Herzfeld, a. 1074 (Pertz, v. 216), calls him "Willhelmus, cognomento Bostar, Rex Anglorum." In our own Worcester Chronicle, a. 1066, he appears as “Wyllelm Bastard," and in Olaf Tryggvesson's Saga (p. 263), as “Vilialmur Bastardur Rudu Jarl." So in Orderic (663 C), "Guillelmus Nothus." So in the Annales Formoselenses (Pertz, v. 36), "Willelmus Bastardus invasit regnum Anglorum." One writer (Chron. Gaufredi Vosiensis, Labbé, iii. 284) for "Bastard" uses the equivalent word "Mamzer ". "Normannorum Ducis filius Mamzer Guillelmus."

THE SURNAMES OF WILLIAM.

609

It has been often said that William himself used the description in formal documents. This assertion rests on very slight authority. There is a charter in Gale's Registrum Honoris de Richmond, p. 225 (a reference for which I have to thank Professor Stubbs), beginning "Ego Willielmus, cognomine Bastardus, Rex Angliæ." It is given also in Selden's Titles of Honour, 535, with the corrupt modern spelling Gulielmus. It seems to me to be palpably spurious, and those who accept it allow it to be unique.

The other title may be seen growing from the vaguer form of "the great William" to the more distinct "William the Great." We read in a charter of William Rufus (Rymer, i. 5), " Ego Willelmus, Dei gratiâ, Rex Anglorum, filius magni Regis Willelmi." So Eadmer (lib. iii. 57, Selden), "quando ille magnus Willielmus hanc terram primo devicit : so William of Jumièges (vii. 16; cf. his description of Robert, vii. 1; see vol. i. p. 474), "Willelmus Dux magnus:" so the Ely History (ii. 41), "deditio Wilhelmi Regis magni." But we find more distinctly in Orderic (706 C), "Henricus Guillelmi Magni Regis Anglorum filius," and still more distinctly in William of Malmesbury (Prol. in lib. iv.), “Willelmus filius Willelmi Magni," and in Ethelred of Rievaux (X Scriptt. 393), "Vixit autem ad Willielmi Magni tempora."

66

The earliest instance, as far as I know, of "Conquæstor" is in Orderic (603 A), who joins it with "Magnus "— 'Guillelmus Magnus, id est Conquæstor, Rex Anglorum." In some manuscripts of Adam of Murimuth (56 ed. Hog) the propriety of the title is formally disputed; "Willelmus Rex improprie potest dici Conquestor, quia ipse fuit nepos et verus hæres beati Edwardi, quia non per judicium sed per potentiam devicit Haraldum, et jus suum virtute propriâ adquisivit." He is also called "Triumphator," which comes still nearer to the modern idea. This name is found twice in one of the foreign writers quoted above (Chron. Gaufredi Vosiensis, Labbé, iii. 293). William Rufus is "Guillelmus filius magni Triumphatoris Guillelmi;" and elsewhere (284) he speaks of "Triumphator ille Guillelmus Mamzer." We find also the same title in English writers. Osbert of Saint Clare (Ep. iii. p. 116), writing to William's grandson Henry of Blois, speaks of "avus Rex vester Willelmus, Angliæ Triumphator egregius," and at p. 121 King Stephen is again made to call him "Triumphator

Angliæ." So in the Vita Haroldi (Chron. Ang. Norm. ii. 208) he is called "Triumphator Willelmus," as if it was a familiar title.

NOTE U. p. 173.

THE BIRTH OF WILLIAM.

SEVERAL questions arise out of the narratives, historical and legendary, of the birth of the great William. No one doubts that he was the natural son of Duke Robert, or that he was born at Falaise; but there are several points open to doubt,—

Ist, As to the origin of his mother;

2nd, As to the exact date of his birth; 3rd, As to the exact place of his birth ;

4th, As to the number of his mother's other children. I will discuss these questions in order.

I. I have mentioned in the text, as a remarkable illustration of English feeling, the story which made William's mother a descendant of the royal house of England. It will be found at length, with some curious details, in the Winchester Annals of Thomas Rudborne, Anglia Sacra, i. 247. Rudborne professes to get the story from a book called "Chronica Danorum in Angliâ regnantium." As a piece of chronology and genealogy, the tale is strange enough. The tanner is called Richard, which looks rather as if he were a Norman, and he bears the surname of "Saburpyr," the meaning of which is far from clear. His wife is distinctly said to to be a daughter of Eadmund and Ealdgyth. Now Eadmund married Ealdgyth in 1015 (see vol. i. p. 372) and he died before the end of 1016. There is therefore hardly room for the birth of a daughter besides the apparently twin (see vol. i. p. 410) Æthelings, Eadmund and Eadward. Such a daughter must have eloped with the tanner at about the same time of life as Hermês when he stole the cows, and, as the mother of the mother of William, who was born at the latest in 1028, she must have been a grandmother at the age of twelve. William must also, besides being a distant cousin of Eadward, have been also a distant nephew, a fact nowhere else alluded to, unless in the extract from Adam of Murimuth quoted in the

last note. In this tale William's mother is called Helen, perhaps through some similarity of letters with Herleva.

The trade of Herleva's father seems to be agreed on at all hands. So the Chronicle of Saint Maxentius (Labbé, ii. 202); "Robertus Willelmum genuit ex eâ quæ fuit filia pelletarii burgensis." In the narrative of William of Jumièges, the bastardy of the Conqueror and the calling of his maternal grandfather dawn upon the reader by degrees. He first, when describing Robert's nomination of William as his successor, simply calls him "Willelmum filium suum, quem unicum apud Falesiam genuerat" (vi. 12). When he speaks of the indignation of the Norman nobles at William's accession, he is driven to mention his bastardy; "Willelmus enim, ex concubinâ Roberti Ducis, nomine Herlevâ, Fulberti cubicularii Ducis filia, natus, nobilibus indigenis, et maxime ex Richardorum prosapiâ natis, despectui erat utpote nothus" (vii. 3). The later dignity of the grandfather is here put forward as a sort of forlorn hope; but when it is necessary to explain the point of the insults offered to William at Alençon, the unsavoury trade of Fulbert at last unavoidably peeps out; "Parentes matris ejus pelliciarii exstiterant" (vii. 18).

He was a burgess of Falaise and a tanner.

It is possible that the word "indigense" in the second of the extracts just made may be taken to confirm the story according to which Fulbert was not only of a low occupation, but of foreign birth. Besides the English legend, which may possibly contain this small grain of truth, there is a tale in the Chronicle of Alberic "Trium Fontium" (A. 1035, Leibnitz, Accessiones, ii. 66), which is told with great glee by Sir Francis Palgrave (iii. 144). According to this version, Herbert, as he is called, was not a native of Falaise, but came with his wife Dodo or Duixa from some place in the Bishoprick of Lüttich, either Chaumont or Huy (Hoium). This tale however does not represent the tanner's daughter as the original object of the fancy of Robert. The Count sees the daughter of his provost or bailiff (præpositus) at Falaise dancing, and asks for her; but the lover is made the subject of a trick, and the daughter of the tanner takes the place of the daughter of the bailiff. Here is food for the Comparative Mythologists, as this tale is the same as the tale of Richard and Gunnor, and as one of the legends of our own Eadgar. See vol. i. p. 252.

II. The date of William's birth has been discussed by M. Deville