^{page 313 note a} Medieval Military Architecture, i. viiGoogle Scholar.

^{page 313 note b} op cit. i 15.

^{page 314 note a} Medieval Military Architecture, i. 18Google Scholar.

^{page 314 note b} Ibid. 145.

^{page 314 note c} Ibid. 22.

^{page 314 note d} Ibid. 23.

^{page 315 note a} Norman Conquest, iv. 191Google Scholar.

^{page 315 note b} Ibid. v. 648. The first of the above quotations continues thus: “Most likely it stood in the void space between the mound, the gateway, and the later castle.” It is thus abundantly clear that, at Warwick, Mr. Freeman considered the castle of the Conqueror to have been something quite distinct and apart from the old English mound. On the other hand, at Wallingford, we find him writing: “One of the vast mounds which speak of later days of English victory under Eadward the Unconquered stood ready to become at William's bidding the kernel of a stronghold from which the new invader might hold Englishmen in bondage” (iii. [1875], 543).Google Scholar And, lastly, at Chester, under the heading “Restoration by Æthelflsed 907,” he writes without hesitation of “the mound on which the Norman castle was now to supplant her earlier stronghold” (iv. [1871] 313).Google Scholar Comparing this with the passage on Warwick quoted in the text from the same volume, we see, I think, clearly that he had not thought out the subject, and that his vague language leaves it in doubt what he really believed as to those mounds and their treatment by the Normans, if indeed he had then any definite belief on the subject.

^{page 315 note c} The Foundations of England, i. 272, 452Google Scholar; ii. 152.

^{page 316 note a} See below.

^{page 316 note b} Norman Conquest (1876), v. 648.Google Scholar

^{page 316 note c} Ibid. 649.

^{page 317 note a} Ibid. (1870), ii. 136. Compare the remarkable passage on the destruction of the castles at York by the English rebels in September, 1069: “Wherever a Norman castle had been reared it was the object of the bitterest of all hatred, as the living embodiment of the foreign yoke. We now look on those massive square keeps, wherever they are left to us, as among the most venerable and precious of the antiquities of our land…But when those towers were still newly built, when their square stones were still in their freshness, when the arches of their doors and windows were still sharp and newly cut, they were to our fathers the objects of a horror deeper even than that with which France in the moment of her uprising looked on the Bastille of her ancient kings. They were the very homes of the Conquest; within their impregnable walls the foe was sheltered…In the eyes of the men of those days the castle was an accursed thing, to be swept away from the earth by the stroke of righteous vengeance, as when liberated Syracuse swept away the citadel from which her Tyrants had held her in thraldom. On the very day on which the army reached York the two castles were broken down. We are not to suppose that the whole of the massive walls of two Norman keeps could be razed to the ground in the afternoon of a day of battle. But they were doubtless dismantled, wrecked, and left in a ruined state, so that they could not, for some while at least, be again used as places of defence” (Ibid. [1871], iv. 270). Yet of one of these “square Norman keeps,” with their “massive walls,” the writer himself had asserted, some pages back (p. 241), that it “must have been a hasty structure of wood or of the roughest kind of masonry. For we read that it was built during the king's sojourn of eight days” earlier in the year.

Compare with Mr. Freeman's dreams the plain words of Mr. Clark on the subject of these York castles. On the later of the two he wrote (op. cit. i. 86, ii. 545): “The other mound, the Bayle Hill, south of the river, was also fortified by William, but in haste and with timber only, which does not appear ever to have been replaced with masonry.” “That this was a mere stockade is clear from the fact that it was completed in eight days, before he left the city…The defences were familiar to the English, and would not strike them with the same terror as the stony and lofty keeps which the Normans had lately begun to build in Normandy, and which William had commenced in London.” Here I may incidentally observe that, according to Mr. Clark himself, the Tower of London was not commenced till some ten years later (i. 138; ii. 205), and that even this date is earlier than can be proved. But keeping to York, we find Mr. Clark pressing, very properly, the evidence as to its castles (i. 42-3), and concluding thus: “It is clear, from the time occupied by the whole sequence of events, that these castles were not of masonry. Moreover, the masonry of the present York keep contains nothing that can be attributed to the eleventh century, but much that is far too early to have replaced a really substantial keep or curtain of Norman date had such been built. Upon the great and artificial mound of Bayle Hill, the site of the second castle, there is neither trace nor tradition of any masonry at all.”

^{page 317 note b} Ibid. 605.

^{page 318 note a} The above plan has been kindly prepared by Mr. W. H. St. John Hope from the 25-inch Ordnance Map, and it may therefore be relied on as accurate. It differs seriously from the plan given in Mr. Clark's work, but his own description there facing it proves that the plan is erroneous.

When correctly depicted, as here, Richard's Castle presents the normal features of these fortresses, and may be taken as a type of the Conquest castle described in this paper. The appended base-court or bailey is “semi-lunar” in form, a shape produced in these strongholds by placing the moated mound with its proper ditch astride on the enceinte. The same arrangement will be observed in the plan of Pleshey beiow.

Mr. Clark observed that ‘Richard's Castle is one of a series of works common on the Welsh border and the Middle Marches. Such were…Kilpeok and Ewias Harold” (ii. 401). And of Ewias Harold lie wrote that “on the western margin is placed a conical table mound…which necessarily …reduces the eastern side to an open crescent-shaped platform” (ii. 40). This platform corresponds, it will be seen, with the base-courts of Richard's Castle and of Pleshey.

The immediat e proximity of Richard's Castle to the churchyard is another feature of common occurrence, and points to such castles having been used as the seats of local lords (mottes seigneuriales).

^{page 319 note a} Op. cit. 41.

^{page 319 note b} Ibid. 42.

^{page 319 note c} Ibid. 44.

^{page 319 note d} Ibid. 44-45.

^{page 320 note a} So too he wrote, in another place, of Richard's Castle that “as this castle was a great cause of offence, it probably was something different from the fortified timber houses of the English thanes, and may well have been of stone, after the rising Norman fashion” (i. 393).

^{page 320 note b} “A tolerably close examination has failed to discover, either at Arundel or elsewhere in England or in Normandy, any masonry of very early character, probably none that can safely be attributed to the eleventh century. The fact seems to be that the early timber structures, which are known to have been erected originally on the moated mounds, were found to be very defensible, and so were retained by the Norman lords until they were able to replace the timber by masonry” (i. 139). So too, “it may safely be said that, save a fragment of wall at Corfe, no military masonry older than that event (the Conquest) has as yet been discovered” (i. 37). This covers the remains of masonry at Richard's Castle, which he admitted looked later than 1052 (ii. 402). Indeed he elsewhere definitely stated that it has “no masonry of the age of the Confessor, nor is it probable that the keep was constructed before the age of Stephen, if so early” (i. 102).

^{page 320 note c} “I have already spoken of the extreme rarity of castles in England before the Norman Conquest, and of the general indignation which was awakened by the building of a single castle in Hereford-shire by a single Norman adventurerd…But now the age of castle-building fairly set in. It was William's policy to keep the conquered land in check by commanding' every town and every important point by one of these fortresses…And, as we follow his steps, we shall see that at every stage of the Conquest, the building of a castle in a conquered town was ever the first work of the Conqueror's foresight” (Norman Conquest [1871], iv. 66).Google Scholar And again he speaks of the houses overthrown “to make way for the building of the castles which were in English eyes the special homes of wrong and badges of bondage” (Ibid. [1876], v. 40).

^{page 321 note a} Ibid. [1870], ii. 605-7.

^{page 321 note b} Mr. Freeman, indeed, could not mistake the significance of the word as used by Lambert. We find him writing: “As the Lady of the Mercians had done a few years before, he raised a mighty mound–‘dunio,’ ‘firmissimus munitionis agger’—and girt it about with a twofold ditch. These words bring home to us that dunio, donjon, meant, first of all, the mota. the artificial dune or down, so characteristic of the earlier stage of fortification…The dunio, as we shall presently see, was primarily meant to carry only wooden buildings; at the outside it would never carry more than the shell-keep of Cardiff or Lincoln.” (Historical Essays, 4th S. 171.)

^{page 321 note c} See my Commune of London, 43-8.

^{page 321 note d} Ibid. 299-301.

^{page 322 note a} Social England, i. 328Google Scholar. Precisely the same view is taken by another writer in the same volume: “Conquered England was held down by dozens of castles, square keeps with walls of enormous thickness built of solid stone, and often relying solely on their own strength without being surrounded by any outworks. The English, in their revolts, could never storm one of the new castles. …In the eleventh century the art of defence had quite outgrown the art of attack” (p. 301).

^{page 322 note b} Art of War in the Middle Ages, i. 517Google Scholar.

^{page 323 note a} The legend on the tapestry is “jussit ut foderetur eastellum at Hestengaceastra.” The Anglo-Saxon Chronicle has: “Hi…worhton castel set Haestinga port” It s phrase for the Herefordshire castle of 1052 is “gewroht aenne castel on Herefordscire.”

^{page 325 note a} Orderic had said of the first castle, “in urbe ipsa munitionem firmavit,” and of the second “alterum presidium condidit.”

^{page 325 note b} See my Geoffrey de Mandeville, 168. Compare the use of “firmavit” by Lambert of Ardres in the passage quoted above.

^{page 326 note a} Ibid. 91.

^{page 326 note b} Ibid. 175.

^{page 326 note c} Berkeley Charters and Muniments, 2.

^{page 326 note d} Registrum monasterii de Winchecumla, 65. Ralf, we read, founded there, at the same time, church, and now “the meadow adjoining the church contains the earthworks of Ralfs castle” (Bristol and Gloucester Archæological Transactions, xxiii. 9Google Scholar). Mr. W. H. St. John Hope tells me that these are slight, so that the castle was probably destroyed as “adulterine” under Henry II., not being licensed to “stand and remain.” The proximity of church and castle should here again be observed.

^{page 326 note c} Vincent's Discovery of Brooke's Errors, 302-3. It is strongly suggested by the second clause that the castles contemplated would be rapidly raised and could not, therefore, be of masonry.

^{page 327 note a} A careful examination of the passage in the Chronicle, under 1051, will show that the grievance was the “harme and bismere” wrought to the king's men by the strangers rather than their raising “senne castel.” And what Godwine demanded was not the destruction of the castle, but the handing over of the Frenchmen themselves.

^{page 327 note b} The argument based on the introduction (or, more strictly, the fresh introduction) of the word “castel” is strengthened, I think, by the similar introduction of the word Tur into the Chronicle, under 1097, to denote a rectangular keep, the Tower of London.

^{page 327 note c} October, 1898 (lxiv. 209-238).

^{page 327 note d} Proceedings, xxxiv. 260–288Google Scholar.

^{page 328 note a} Mrs. Armitage, it seems, was not aware that the “castle of Wareham” was Corfe; and her inclusion of Berkeley, Hastings, Montgomery, Rochester, and Winchester among castles of the motte type known to have been built in the reigns of William I. or William II. is open to question. Nor is the destruction of houses in a town absolute proof in itself of the construction of a motte as she insists. It might well, conceivably, be due to the extension of a fortified area.

^{page 328 note b} “Some north to Rodbertes castele,” in the Chronicle.

^{page 328 note c} Transactions of the Essex Archæogical Society (N. S.), vii. 250Google Scholar.

^{page 329 note a} The above plan was prepared by Mr. I. C. Gould, to whom the Society is indebted for the loan of the block. It illustrates the addition to the moated mound and appended base-court, which were (as at Richard's Castle) the essential features of the type, of outer courts or baileys, which is not uncommon in the case of important strongholds.

^{page 330 note a} Shropshire, v. 225Google Scholar.

^{page 330 note b} History of the Norman Conquest, v. 808Google Scholar.

^{page 330 note c} In Domesday, for instance, the “Alre” of four entries in Devonshire becomes “Avra” in a fifth, just as with “Alretune” and “Avretone.”

^{page 330 note d} Medieval Military Architecture, i. 38, ii. 516Google Scholar; History of the Norman Conquest, v. 807Google Scholar.

^{page 330 note e} “Corfe appears in Domesday as the former holding of two Englishmen…and as the present possession of Robert son of Gerold. From this a local antiquary of high authority (Mr. Bond) has argued that the Domesday of Wareham Castle really refers to Corfe” (English Towns and Districts, 153). I need scarcely say that Mr. Bond's careful arguments were, of course, not based on an error which he would have known to be such. He specially explained that the “Corfe” of Domesday, held by Robert Fitz-Gerold, was not Corfe Castle, but Corfe Mullen (p. 90).

^{page 331 note a} History of the Norman Conquest (1871), iv. 284–5.Google Scholar

^{page 331 note b} See Medieval Military Architecture, ii. 376–7Google Scholar.

^{page 331 note c} Yorkshire Archæological Journal, iv. 142.Google Scholar

^{page 331 note d} Vol. ii. 3. The authority is recognised as a good one for that period, and it also represents local knowledge, for Abingdon was about half way between Oxford and Wallingford.

^{page 332 note a} Introduction to Domesday, i. 212, 223Google Scholar.

^{page 332 note b} History of the Norman Conquest (1871), iv. 66, 493Google Scholar; and William Bnfus (1882), i. 58, ii. 43Google Scholar.

^{page 332 note c} Medieval Military Architecture, i. 71, 195Google Scholar.

^{page 332 note d} It is on this account, doubtless, that Mrs. Armitage, I observe, has run her pen, in the copy I possess, through the perfectly sound observation that “It is certainly remarkable that Dover Castle, the only castle which has any claim to have been built by an Anglo-Saxon, has no motte” (p. 284).

^{page 332 note e} “Castellum Ewias” is similarly valued at £10 a year, and we read that 5 carucates of land were there (ibidem), and ploughs at work on them (fo. 186).

^{page 333 note a} Similarly the great moated mound with its surrounding courts at Castle Acre was a stronghold of William de Warenne, at the caput of whose Norman fief, Bellencombre, there is still to be seen “Château du xi^e siècle, sur une motte enorme entouree de fossés.”

^{page 333 note b} It is but fair to remember that the words of Orderic as to “castella” being “paucissima” in England before the Normans came are, though directly opposed to Mr. Clark's theory, antagonistic also to the view that they were previously quite unknown. These “castella,” however, need not have comprised mounds. It is admitted, for instance, that Dover Castle was without one.

^{page 333 note c} Mr. W. H. St. John Hope has drawn my attention to the striking case of Old and New Buckenham, in Norfolk. A priory was founded at Old Buckenharn by William d'Albini, earl of Chichester, not earlier than 1146, for Austin canons, to whom the founder gave inter alia “infra parcum cum bosco et piano cum sede castelli lxxx. acras, et castellum diruendum” (Dugdale, , Monasticon Anglicanum, vi. 419).Google Scholar The earl then built himself another castle at New Buckenham, where its great earthworks still remain. See the plan in Harrod's, H.Gleanings among the Castles and Convents of Norfolk (Norwich, 1857), 215Google Scholar.

^{page 334 note a} There is some reason to suppose that this may have been the case at Hastings also.

^{page 334 note b} See the cases of Hayles, etc. quoted above.

^{page 335 note a} A base court, if added, would have to depend on its own defences. The mound itself could not protect it.

^{page 335 note b} Exeter (Historic Town Series).

^{page 335 note c} Exeter, 21. This follows the words of the Chronicle, which here speaks of “pam fsestene” (I. 166-7). In Florence of Worcester its words are rendered, “quia civitatem jam intraverat, illam assequi non poterat.”

^{page 335 note d} Ibid. 21.

^{page 336 note a} Norman Conquest (1871), iv. 313,Google Scholar note. The Latin of “insequutus” is Mr. Freeman's.

^{page 336 note b} Exeter, 22.

^{page 336 note c} Norman Conquest, iv. 151–2Google Scholar.

^{page 337 note a} Norman Conquest, iv. 139Google Scholar.

^{page 337 note b} Ibid. 154.

^{page 337 note c} Ibid. 156. In Ms Exeter, 36-8, Mr. Freeman adopted precisely the same view.

^{page 337 note d} Vol. i. 309.

^{page 337 note e} Exeter, 26.

^{page 338 note a} Norman Conquest, i. 315–6Google Scholar, note.

^{page 338 note b} See my paper on “The Conqueror at Exeter,” in Feudal England, 431 et seq.

^{page 338 note c} Ibid. iv. 161. “To the right of the gate,” we read, on his arrival, “there rose the high ground which William's keen eye would at the first glance mark as the site of the future castle” ( P. 154).

^{page 338 note d} Ibid, iv. 272-3.

^{page 339 note a} Exeter, 39.

^{page 339 note b} Medieval Military Architecture, ii. 45Google Scholar.

^{page 339 note c} Ibid. i. 139.

^{page 339 note d} Exeter, 22.

^{page 339 note e} Orderic states that William, at Exeter, “locum iutra msenia ad exstruendum castellum delegit.” This implies a new construction.