Chapter VII: The Castles of the Normans in England (2)

CLITHEROE, Lancashire (Fig. 13).--There is no express mention of this castle in Domesday Book, but of two places in Yorkshire, Barnoldswick and Calton, it is said that they are in the _castellate_ of Roger the Poitevin.[363] A castellate implies a castle, and as there is no other castle in the Craven district (to which the words of the Survey relate) except Skipton, which did not form part of Roger's property, there is no reason to doubt that this castle was Clitheroe, which for centuries was the centre of the Honour of that name. The whole land between the Ribble and the Mersey had been given by William I. to this Roger, the third son of his trusted supporter, Earl Roger of Shrewsbury. One can understand why William gave important frontier posts to the energetic and unscrupulous young men of the house of Montgomeri, one of whom was the adviser and architect of William Rufus, another a notable warrior in North Wales, another the conqueror of Pembrokeshire. As it appears from the Survey that Roger's possessions stretched far beyond the Ribble into Yorkshire and Cumberland, it seems quite possible--though here we are in the region of conjecture--that just as his father and brothers had a free hand to conquer as they listed from the North and South Welsh, so Roger had a similar commission for the hilly districts still unconquered in the north-west of England. But fortune did not favour the Montgomeri family for long. They were exiled from England in 1102 for siding with Robert Curthose, and in the same year we find the castle of Clitheroe in the hands of Robert de Lacy, lord of the great Yorkshire fief of Pontefract.[364]

The castle of Clitheroe stands on a lofty motte of natural rock.[365] There are no earthworks on the summit, but a stout wall of limestone rubble without buttresses encloses a small court, on whose south-west side stands the keep. It is just possible that the outer wall may be the original work of Roger, as limestone rubble would be easier to get than earth on this rocky hill. The keep is small, rudely built of rubble, and has neither fireplace nor garde-robe, nor the slightest ornamental detail--not even a string course. But in spite of the entire absence of ornament, a decorative effect has been sought and obtained by making the quoins, voussoirs, and lintels of a dressed yellow sandstone. The care with which this has been done is inconsistent with the haste with which Roger must inevitably have constructed his first fortification, if we suppose, as is probable, that he received the first grant of his northern lands on William's return in 1070 from his third visit to the north, when he made that remarkable march through Lancashire to Chester which is described by Ordericus. It seems more likely that even if the outer wall or shell were the work of Roger, he had only wooden buildings inside its circuit. Dugdale attributes the building of the keep to the second Robert de Lacy, between 1187 and 1194, and it is probable that this date is correct.[366] The bailey of Clitheroe lay considerably below the keep, and is now overbuilt with a modern house, offices, and garden. It covers one acre. A Roman road up the valley of the Ribble passes near the foot of the rock.[367]

As the very name of Clitheroe is not mentioned in Domesday Book, it clearly was not an important centre in Saxon times. The value of Blackburn Hundred, in which Clitheroe is situated, had fallen between the Confessor's time and the time when Roger received it. It is quite possible that he never lived at Clitheroe, as he sub-infeoffed the manor and Hundred of Blackburn to Roger de Busli and Albert Greslet before 1086.[368]

COLCHESTER, Essex.--The remarkable keep of this castle has been the subject of antiquarian legend for many centuries, and Mr Clark has the merit of having proved its early Norman origin, by its plan and architecture. A charter of Henry I. is preserved in the cartulary of St John's Abbey at Colchester, which grants to Eudes the Dapifer "the city of Colchester, and the tower and the castle, and all the fortifications of the castle, just as my father had them and my brother and myself."[369] This proves that the keep and castle were in existence in the Conqueror's time; the Norman character of the architecture proves that the keep was not in existence earlier. We see, then, that the reason there is no motte at Colchester is that there was a stone keep built when first the castle was founded. As far as we are aware, Colchester, the Tower of London, and the recently discovered keep of Pevensey are the only certain instances of stone keeps of the 11th century in England.

That one of the most important of the Conqueror's castles, second only to the Tower of London, and actually exceeding it in the area it covers, should be found in Colchester, is not surprising, because the Eastern counties at the time of the Conquest were not only the wealthiest part of the kingdom (as Domesday Book clearly shows[370]), but they also needed special protection from the attacks of Scandinavian enemies. Mr Round has conjectured that the castle was built at the time of the invasion of St Cnut, between 1080 and 1085.[371]

The castle is built of Roman stones used over again, with rows of tiles introduced between the courses with much decorative effect.[372] The original doorway was on the first floor, as in most Norman keeps; but at some after time, probably in the reign of Henry I.,[373] the present doorway was inserted; and most likely the handsome stairway which now leads up from this basement entrance was added, as it shows clear marks of insertion. Henry II. was working on the walls of the castle in 1182, and it may be strongly suspected that the repairs in ashlar, and the casing of the buttresses with ashlar, were his work.[374] One item in the accounts of Henry II. is £50 "for making the bailey round the castle."[375] There were two baileys to the castle of Colchester--the inner one, which scarcely covered 2 acres, and the outer one, which contained about 11. The inner bailey was enclosed at first with an earthwork and stockade, the earthwork being thrown up over the remains of some Roman walls, whose line it does not follow. Afterwards a stone wall was built on the earthwork, the foundations of which can still be traced in the west rampart.[376] The outer bailey, which lay to the north, extended on two sides to the Roman walls of the town; on the west side it had a rampart and stockade. If the £50 spent by Henry II. represents the cost of a stone wall round the inner bailey, then the _palicium_ blown down by the wind in 1219 must have been the wooden stockade on the west side of the outer bailey.[377] The question is difficult to decide, but at any rate the entry proves that as late as Henry III.'s reign, some part of the outer defences of Colchester Castle was still of timber.

The position of Colchester Castle is exceptional in one respect, that the castle is almost in the middle of the town. But this very unusual position is explained by Mr Round's statement that the land forming the castle baileys, as well as that afterwards given to the Grey Friars on the east, was crown demesne before the Conquest, and consequently had been cultivated land, so that we do not hear of any houses in Colchester being destroyed for the site of the castle.[378] But by keeping this land as the inalienable appendage of the royal castle William secured that communication between the castle and the outside country which was so essential to the invaders.

The value of the city of Colchester had risen enormously at the date of the Survey.[379]

CORFE, Dorset (Fig. 13).--Mr Eyton has shown that for the _castellum Warham_ of Domesday Book we ought to read _Corfe_, because the castle was built in the manor of Kingston, four miles from Wareham.[380] And this is made clear by the _Testa de Nevill_, which says that the church of Gillingham was given to the nunnery of Shaftesbury in exchange for the land on which the castle of Corfe is placed.[381] Because King Edward the Martyr was murdered at Corfe, at some place where his stepmother Elfrida was residing, it has been inferred that there was a Saxon castle at Corfe; and because there is a building with some herring-bone work among the present ruins, it has been assumed that this building is the remains of that castle or palace. But the _Anglo-Saxon Chronicle_, the only contemporary authority for the event, says nothing of any castle at Corfe, but simply tells us that Edward was slain at Corfe Geat, a name which evidently alludes to a gap or passage through the chalk hills, such as there is at Corfe.[382] Nor is there any mention of Corfe as a fortress in Anglo-Saxon times; it is not named in the _Burghal Hidage_, and we do not hear of any sieges of it by the Danes. Nor is it likely that the Saxons would have had a fortress at Corfe, when they had a fortified town so near as Wareham.[383] Kingston, the manor in which Corfe is situated, was not an important place, as it had no dependent soke. The language of Domesday absolutely upsets the idea of any Saxon castle or palace at Corfe, as it tells us that William obtained the land for his castle from the nuns of Shaftesbury, and we may be quite sure they had no castle there.[384]

Corfe Castle stands on a natural hill, which has been so scarped artificially that the highest part now forms a large motte. Three wards exist--the eastern or motte ward, the western, and the southern. The two former probably formed the original castle. On the motte (which possibly is not artificial, but formed by scarping) stands the lofty keep, of splendid workmanship, probably of the time of Henry I. In the ward pertaining to it are buildings of the time of John and Henry III.[385] The western ward has towers of the 13th century, but it also contains the interesting remains of an early Norman building, probably a hall or chapel, built largely of herring-bone work; this is the building which has been so positively asserted to be a Saxon palace. But herring-bone masonry, which used to be thought an infallible sign of Saxon work, is now found to be more often Norman.[386] The building is certainly an ancient one, and may possibly have been contemporary with the first Norman castle; its details are unmistakably Norman. But very likely it was the only Norman masonry of the 11th century at Corfe Castle.[387] It is clear that the stone wall which at present surrounds the western bailey did not exist when the hall (or chapel) was built, as it blocks up its southern windows. Probably there was a palisade at first on the edge of the scarp. Palisades still formed part of the defences of the castle in the time of Henry III., when 62_l._ was paid "for making two good walls in place of the palisades at Corfe between the old bailey of the said castle and the middle bailey towards the west, and between the keep of the said castle and the outer bailey towards the south."[388] This shows that the present wing-walls down from the motte were previously represented by stockades. The ditch between the keep and the southern bailey has been attributed to King John, on the strength of an entry in the _Close Rolls_ which orders fifteen miners and stone-masons to work on the banks of the ditch in 1214.[389] But we may be quite certain that this ditch below the motte belonged to the original plan of the castle; John's work would be either to line it with masonry, or to enlarge it. It is not without significance for the early history of the castle that Durandus the _carpenter_ held the manor of Mouldham near Corfe, by the service of finding a carpenter to work at the keep whenever required.[390]

The area of Corfe Castle, if we include the large southern bailey, is 3-3/4 acres; without it, 1-1/2 acres. This bailey was certainly in existence in the reign of Henry III. (as the extract from the _Close Rolls_ proves) before the towers of superb masonry were added to it by Edward I.

The value of Kingston Manor had considerably increased at the date of the Survey. After the Count of Mortain forfeited his lands (in 1105), the castle of Corfe was kept in the hands of the crown, and this increases the probability that the keep was built by Henry I.

About 400 yards S.W. of Corfe Castle is an earthwork which might be called a "Ring and Bailey." Instead of the usual motte there is a circular enclosure, defended by a bank and ditch of about the same height as those of its bailey, but having in addition an interior platform or berm. This work is probably the remains of a camp thrown up by Stephen during his unsuccessful siege of Corfe Castle in 1139.

DOVER, Kent (Fig. 14).--The Norman historian, William of Poitiers, tells us that the castrum of Dover was built by Harold at his own expense.[391] This comes from the celebrated story of the oath of Harold to William, a story of which Mr Freeman says that there is no portion of our history more entangled in the mazes of contradictory and often impossible statements.[392] But let us assume the statement about the _castrum_ to be true; the question then to be answered is this: of what nature was that castrum? We never are told by English chroniclers that Harold built any castles, though we do hear of his fortifying towns. The present writer would answer this question, tentatively indeed, and under correction, by the theory that the castrum constructed or repaired by Harold was the present outer rampart of Dover Castle, which encloses an area of about 34 acres, and may have enclosed more, if it was formerly complete on the side towards the sea.[393] The evidence in support of this theory is as follows:--

1. There certainly was a _burh_ on the top of the cliff at Dover in Saxon times, as the _Anglo-Saxon Chronicle_ tells us that in 1048 Eustace of Boulogne, after coming to Dover, and slaying householders there, went _up to the burh_, and slew people both within and without, but was repulsed by the burh-men.[394] There was then a burh, and valiant burh-men on the cliff at Dover in Edward the Confessor's reign. But the whole analogy of the word burh makes it certain that by the time of Edward it meant a fortified town.[395]

2. That the burh at Dover was of the nature of a town, with houses in it, is confirmed by the poem of Guy of Amiens, who says that when King William entered the _castrum_, he ordered the English to evacuate their houses.[396] William of Poitiers also states that there was an innumerable multitude of people in the castle,[397] though he may refer to a multitude gathered there for safety.

3. Though the whole of the outer enceinte is generally credited to Hubert de Burgh in Henry III.'s reign, the truth probably is that he built the first stone walls and towers on the outer rampart; but the existence of this earthen rampart shows that there was a wooden wall upon it previously. It is not improbable that it was for the repair of this wooden wall that so much timber was sent to Dover in the reigns of Richard I. and John.[398] Dering, who was lieutenant of the castle in 1629, records the tradition that the tower in the outer enceinte, called Canons' Gate, dates from Saxon times (of course this could only be true of a wooden predecessor of the stone tower), and that Godwin's Tower, on the east side of the outer vallum, existed as a postern before the Conquest.[399] Nearly all the towers on this wall were supported by certain manors held on the tenure of castle-guard, and eight of them still retain the names of eight knights to whom William is said to have given lands on this tenure. Mr Round has shown that the _Warda Constabularii_ of Dover Castle can be traced back to the Conquest, and that it is a mere legend that it was given as a fief to a Fienes. He remarks that the nine wards of the castle named in the Red Book of the Exchequer are all reproduced in the names still attached to the towers. "This coincidence of testimony leads us to believe that the names must have been attached at a very early period; and looking at the history of the families named, it cannot have been later than that of Henry II."[400] May it not have been even earlier? Eight of these names are attached to towers on the outer circuit,[401] and five of them are found as landholders in Kent in Domesday Book.

4. William of Poitiers further tells us that when the duke had taken the castle, he remained there eight days, _to add the fortifications which were wanting_.[402] What was wanting to a Norman eye in Anglo-Saxon fortifications, as far as we know them, was a citadel; and without laying too much stress on the chronicler's eight days, we may assume that the short time spent by William at Dover was just enough for the construction of a motte and bailey, inside the _castrum_ of Harold, but crowned by wooden buildings only.

Taking these things together, we venture to assume that the inner court in which the keep of Dover stands, represents an original motte, or at any rate an original citadel, added to the castle by William I. Whether what now remains of this motte is in part artificial, we do not pretend to say; it may be that it was formed simply by digging a deep ditch round the highest knoll of ground within the ancient ramparts.[403] Anyhow, it is still in effect a motte, and a large one, containing not only the magnificent keep, but a small ward as well. That this keep was the work of Henry II. there can be no manner of doubt; the _Pipe Rolls_ show that he spent more than £2000 on the _turris_ or keep of Dover Castle between the years 1181 and 1187, and Benedict of Peterborough mentions the building of the keep at this date.[404] The curtain around the motte may also be reckoned to be his work originally, as the _cingulum_ is spoken of along with the _turris_ in the accounts. Modern alterations have left little of Norman character in this curtain which shows at a glance, and the gateways (one of which remains) belong to a later period.

Attached to this keep ward is another ward, whose rampart is generally attributed to Saxon times. We are not in a position positively to deny that the Saxons had an inner earthwork on the highest part of the ground within their _burh_. But considering that small citadels are unusual in Saxon earthworks: considering also that this bailey is attached to the motte in the usual manner of a Norman bailey, and that its size corresponds to the usual size of an original Norman bailey in an important place, it does not seem unreasonable to suppose that this was the original bailey attached to the Conqueror's motte. Its shape is singular, part of it being nearly square, while at the S.E. corner a large oval loop is thrown out, so as to enclose the Roman Pharos and the Saxon church. The outline of the bailey certainly suggests that it was built after the Pharos and the church, and was built with reference primarily to the keep or motte ward. The nature of the ground, and the necessity of enclosing the church and the Roman tower within the immediate bailey of the castle, which would otherwise have been commanded by them, were the other factors which decided the unusual shape of the bailey.

On this earthwork the foundations of a rubble wall were formerly to be traced,[405] probably built by Henry II., as considerable sums for "the wall of the castle" are mentioned in his accounts.[406] Whether there are still any remains of this curtain we are unable to say, but so many of the features of the middle ward have been swept away by modern alterations, and the difficulty of examining what remains, owing to military restrictions, is so great, that little can be said about it, and we find that most authorities observe a judicious silence on the subject. But as the carriage of stone is expressly mentioned in Henry II.'s accounts, we may with great probability assign to him the transformation of the original wooden castle of William into a castle of stone; while the transformation of the Anglo-Saxon borough into a stone enceinte was the work of Henry III.'s reign.

We think the evidence suggests that this _burh_ or outer rampart was in existence when the Conqueror came to Dover, crowned in all probability with a stockade and towers of wood. It may possibly have been a British or even a Roman earthwork originally (though its outline does not suggest Roman work); or it may have been built by Harold as a city of refuge for the inhabitants of the port.[407] The Saxon church which it encloses, and which has long been attributed to the earliest days of Saxon Christianity, is now pronounced by the best authorities to be comparatively late in the style.[408]

The size of the inner castle of Dover appears to be about 6 acres, reckoning the keep ward at 2, and the bailey at about 4.

The value of the town of Dover had trebled at the time of the Survey, in spite of the burning of the town at William's first advent.[409]

DUDLEY, Staffordshire (Fig. 15).--William Fitz Ansculf held Dudley at the time of the Survey, "and there is his castle."[410] Mr Clark appears to accept the dubious tradition of a Saxon Dodda, who first built this castle in the 8th century, since he speaks of Dudley as "a great English residence."[411] This tradition, however, is not supported by Domesday Book, which shows Dudley to have been only a small and unimportant manor before the Conquest. The strong position of the hill was no doubt the reason why the Norman placed his castle there. There is no Norman masonry in the present ruins. The earliest work is that of the keep on the motte, a rectangular tower with round corner turrets, attributed by Mr W. St John Hope to about 1320. The first castle was demolished by Henry II. in 1175,[412] and an attempt to restore it in 1218 was stringently countermanded.[413] The case of Dudley is one of those which proves that Henry II. destroyed some lawful castles in 1175 as well as the unlawful ones. In 1264 a license to restore it was granted to Roger de Somery, in consideration of his devotion to the king's cause in the Barons' War.[414] The whole area of the castle, including the motte, but not including the works at the base of the hill on which it stands, is 1-3/4 acres. The bailey is an irregular oval, following the hill top. Dudley is an instance in which the value of the manor has gone down instead of up since the erection of the castle; this may perhaps be laid to the account of the devastation caused through the Staffordshire insurrection of 1069.

DUNSTER, Somerset (Fig. 15).--Called Torre in Domesday Book. "There William de Moion has his castle."[415] The motte here appears to be a natural rock or _tor_, whose summit has been levelled and its sides scarped by art. About 80 feet below the top is a (roughly) half-moon bailey, itself a shelf on the side of the hill; there is another and much smaller shelf at the opposite end.[416] Some foundations found in the S.W. corner of the upper ward appear to indicate a former stone keep.[417] Dunster was only a small manor of half a hide before the Conquest, but afterwards its value tripled. There was a borough as well as a castle.[418] The castle became the _caput baroniæ_ of the De Moions, to whom the Conqueror gave fifty-six manors in different parts of the county. There is not the slightest reason to suppose that the site was fortified before the Conquest. Mr Clark remarks that "it is remarkable that no mouldings or fragments of Norman ornament have been dug up in or about the site, although there is original Norman work in the parish church." The simple explanation, probably, is that the first castle of De Moion was of wood, although on a site where it would have been possible to build in stone from the first, as it does not appear that any part of the motte is artificial. The area of the bailey is 1-3/4 acres. The value of Dunster had risen at the date of Domesday.[419]

DURHAM (Fig. 16).--The castle here was first built by the Conqueror, on his return from his expedition against Scotland in 1072.[420] It was intended as a strong residence for the bishop, through whom William hoped to govern this turbulent part of the country. He placed it on the neck of the lofty peninsula on which the cathedral stands. The motte of the Conqueror still remains, and so does the chapel[421] which he built in the bailey; probably the present court of the castle, though crowded now with buildings, represents the outline of the original bailey.[422] The present shell keep on the motte was built by Bishop Hatfield in Edward III.'s reign,[423] but has been extensively modernised. There can be little doubt that up to 1345 there were only wooden buildings on the motte, as the writer was informed by Canon Greenwell that no remains of older stone-work than the 14th century had been found there. It is so seldom that we get any contemporary description of a castle of this kind, that it seems worth while to translate the bombastic verse in which Laurence, Prior of Durham, described that of Durham in Stephen's reign:[424]

"Not far hence [from the north road into the city] a tumulus of rising earth explains the flatness of the excavated summit, explains the narrow field on the flattened vertex, which the apex of the castle occupies with very pleasing art. On this open space the castle is seated like a queen; from its threatening height, it holds all that it sees as its own. From its gate, the stubborn wall rises with the rising mound,[425] and rising still further, makes towards the comfort (amæna) of the keep. But the keep, compacted together, rises again into thin air, strong within and without, well fitted for its work, for within the ground rises higher by three cubits than without--ground made sound by solid earth. Above this, a stalwart house[426] springs yet higher than the [shell] keep, glittering with splendid beauty in every part; _four posts are plain, on which it rests, one post at each strong corner_.[427] Each face is girded by a beautiful gallery, which is fixed into the warlike wall.[428] A bridge, rising from the chapel [in the bailey] gives a ready ascent to the ramparts, easy to climb; starting from them, a broad way makes the round of the top of the wall, and this is the usual way to the top of the citadel.... The bridge is divided into easy steps, no headlong drop, but an easy slope from the top to the bottom. Near the [head of the] bridge, a wall descends from the citadel, turning its face westward towards the river.[429] From the river's lofty bank it turns away in a broad curve to meet the field [_i.e._, Palace Green]. It is no bare plot empty of buildings that this high wall surrounds with its sweep, but one containing goodly habitations.[430] There you will find two vast palaces built with porches, the skill of whose builders the building well reveals. There, too, the chapel stands out beautifully raised on six pillars, not over vast, but fair enough to view. Here chambers are joined to chambers, house to house, each suited to the purpose that it serves.... There is a building in the middle of the castle which has a deep well of abundant water.... The frowning gate faces the rainy south, a gate that is strong, high-reaching, easily held by the hand of a weakling or a woman. The bridge is let down for egress,[431] and thus the way goes across the broad moat. It goes to the plain which is protected on all sides by a wall, where the youth often held their joyous games. Thus the castellan, and the castle artfully placed on the high ridge, defend the northern side of the cathedral. And from this castle a strong wall goes down southwards, continued to the end of the church."[432]

The original bailey of this castle covers 1 acre.

ELY, Cambridgeshire (Fig. 17).--This castle was built by William I. in 1070, when he was repressing the last struggle of the English under the heroic Hereward. The monks of Ely felt it a sore grievance that he placed the castle within their own bounds.[433] Both this castle and the one built by William at Aldreth, to defend the passage into the Isle of Ely, had a continuous existence, as they were both refortified by Nigel, Bishop of Ely in Stephen's reign, and Ely Castle was besieged and taken by Stephen.[434] The earthworks of this castle still exist, to the south of the Minster. There is a fine motte with an oval bailey, of which the banks and ditches are traceable in parts. The area of the bailey is 2-1/2 acres. Of Aldreth or Aldrey there appear to be no remains.

The value of the manor of Ely was £33 in the Confessor's reign; it fell to £20 after the devastations of the Conquest, but had risen again to £30 at the time of the Survey.[435]

EWIAS, Herefordshire (Fig. 17).--The brief notice of this castle in Domesday Book throws some light on the general theory of castle-building in England.[436] William FitzOsbern, as the king's vicegerent, rebuilt this march castle, and committed it to the keeping of another Norman noble, and the king confirmed the arrangement. But in theory the castle would always be the king's. This is the only case in the Survey where we hear of a castle being _rebuilt_ by the Normans. We naturally look to one of King Edward's Norman favourites as the first founder, for they alone are said by history to have built castles on the Welsh marches before the Conquest. Dr Round conjectures that Ewias was the "Pentecost's castle" spoken of in the (Peterborough) _Anglo-Saxon Chronicle_ in 1052.[437] No masonry is now to be seen on the motte at Ewias, but Mr Clark states that the outline of a circular or polygonal shell keep is shown by a trench out of which the foundations have been removed. The bailey is roughly of half-moon shape and the mound oval. The whole area of the castle, including the motte and banks, is 2-1/3 acres.

EXETER.--This castle is not mentioned in Domesday Book, but Ordericus tells us that William _chose_ a site for the castle within the walls, and left Baldwin de Molis, son of Count Gilbert, and other distinguished knights, to finish the work, and remain as a garrison.[438] In spite of this clear indication that the castle was a new thing, it has been obstinately held that it only occupied the site of some former castle, Roman or Saxon.[439] Exeter, of course, was a Roman castrum, and its walls had been restored by Athelstan. In this case William placed his castle inside instead of outside the city walls, because, owing to the natural situation of Exeter, he found in the north-west corner a site which commanded the whole city. Although Domesday Book is silent about the castle, it tells us that forty-eight houses in Exeter had been destroyed since William came to England,[440] and Freeman remarks that "we may assume that these houses were destroyed to make room for the castle, though it is not expressly said that they were."[441]

Exeter Castle stands on a natural knoll, occupying the north-west corner of the city, which has been converted into a sort of square motte by digging a great ditch round the two sides of its base towards the town.[442] That this ditch is no pre-Roman work is shown by the fact that it stops short at the Roman wall, and begins again on the outside of it, where, however, the greater part has been levelled to form the promenade of the _Northernhay_ or north rampart of the city. On top of this hill, banks 30 feet high were thrown up, which still remain, and give to the courtyard which they enclose the appearance of a pit.[443] On top of these banks there are now stone walls; but these were certainly no part of the work of Baldwin de Molis, who must have placed a wooden stockade on the banks which he constructed. One piece of stonework he probably did set up, the gatehouse, which by its triangle-headed windows and its long-and-short work is almost certainly of the 11th century. It has frequently been called Saxon, but more careful critics now regard it as "work that must have been done, if not by Norman hands, at Norman bidding and on Norman design."[444] It was no uncommon thing at this early period to have gatehouses of stone to walls of earth and wood. Of these gatehouses Exeter is the most perfect and the most clearly stamped with antiquity.

One thing we look for in vain at Exeter, and that is a citadel. There is no keep, and there is no record that there ever was one, though a chapel, hall, and other houses are mentioned in ancient accounts. Mr Clark says that probably the Normans regarded the whole court as a shell keep. It certainly was, in effect, a motte; but it was altogether exceptional among Norman castles of importance if it had no bailey. And in fact a bailey is mentioned in the _Pipe Roll_ of 1 Richard I., where there is an entry for the cost of making a gaol in the bailey of the castle.[445] Now Norden, who published a plan of Exeter in 1619, says that the prison which formerly existed at the bottom of Castle Lane (on the south or city front of the present castle) was "built upon Castle grounde," and he states that the buildings and gardens which have been made on this ground are intrusions on the king's rights.[446] The remarkably full account of the siege of Exeter in the _Gesta Stephani_ speaks of an outer _promurale_ which was taken by Stephen, as well as the inner bridge leading from the town to the castle, before the attack on the castle itself. Unfortunately the word _promurale_ has the same uncertainty about it that attaches to so many mediæval terms, and the description given of it would apply either to the banks of a bailey, or to the _heriçon_ on the counterscarp of the ditch of the motte. We must, therefore, leave it to the reader's judgment whether the evidence given above is sufficient to establish the former existence of a bailey at Exeter, and to place Exeter among the castles of the motte-and-bailey type.

The description of the castle given by the writer of the Gesta has many points of interest.[447] He describes the castle as standing on a very high mound (_editissimo aggere_) hedged in by an insurmountable wall, which was defended by "Cæsarian" towers built with the very hardest mortar. This must refer to Roman towers which may have existed on the Roman part of the wall. Whether there was a stone wall on the other two sides, facing the city, may be doubted, as the expenditure entered to Henry II. in the _Pipe Rolls_ suggests that he was the first to put stone walls on the banks, and the two ancient towers which still exist appear to be of his time.[448] The chronicler goes on to say that after Stephen had taken the _promurale_ and broken down the bridge, there were several days and nights of fighting before he could win the castle, which was eventually forced to surrender by the drying-up of the wells. The mining operations which he describes were no doubt undertaken with the view of shaking down the Roman wall at the angle where it joins the artificial bank of Baldwin de Molis. Possibly the chamber in the rock with the mysterious passages leading from it, which is still to be seen in the garden of Miss Owthwaite, at the point where the ditch ends, is the work of Stephen's miners.[449] The description of his soldiers scrambling up the _agger_ on their hands and knees (_quadrupede incessu_) will be well understood by those who have seen the castle bank as it still rises from that ditch.

The present ward of Exeter Castle, which is rudely square in plan, covers an area of 2 acres, which is as large as the whole area of many of the smaller Norman castles. The castle was allowed to fall into decay as early as 1549,[450] and since then it has been devastated by the building of a Sessions House and a gaol. No plan has been preserved of the former buildings in this court, though the site of the chapel is known.

There is no statement in Domesday Book as to the value of Exeter.

EYE, Suffolk (Fig. 17).--This castle was built by William Malet, one of the companions of the Conqueror, who is described as having been half Norman and half English.[451] Eye, as its name implies, seems to have been an island in a marsh in Norman times, and therefore a naturally defensible situation. The references in the _Pipe Rolls_ to the _palicium_ and the _bretasches_ of Eye Castle show that the outer defences of the castle at any rate were of wood in the days of Henry II.[452] That there were works in masonry at some subsequent period is shown by a solitary vestige of a wing wall of flints which runs up the motte. A modern tower now occupies the summit. The bailey of the castle, the outline of which can still be traced, though the area is covered with buildings and gardens, was oval in shape, and covered 2 acres.

The value of the manor of Eye had gone up since the Conquest from £15 to £21. This must have been due to the castle and to the market which Robert Malet or his son William established close to the castle; for the stock on the manor and the number of ploughs had actually decreased.[453] A proof that there is no deliberate register of castles in Domesday Book is furnished by the very careful inventory of the manor of Eye, where there is no mention of a castle, though it is noticed that there are now a park and a market; and it is only in the account of the lands of the bishop of Thetford, in mentioning the injury which William Malet's market at Eye had done to the bishop's market at Hoxne, that the castle of Eye is named.

GLOUCESTER.--"There were sixteen houses where the castle sits, but now they are gone, and fourteen have been destroyed in the _burgus_ of the city," says Domesday Book.[454] Gloucester was undoubtedly a Roman _chester_, and Roman pavements have been found there.[455] The description in the Survey would lead us to think that the castle was outside the ancient walls,[456] though Speed's map places it on the line of the wall of his time, which may have been a mediæval extension. The castle of Gloucester is now entirely destroyed, but there is sufficient evidence to show that it was of the usual Norman type. There was a motte, which was standing in 1819, and which was then called the Barbican Hill;[457] it appears to have been utilised as part of the works of the barbican. This motte must originally have supported a wooden keep, and Henry I. must have been the builder of the stone keep which Leland saw "in the middle of the area;"[458] for in 1100 Henry gave lands to Gloucester Abbey "in exchange for the site where now the keep of Gloucester stands."[459] The bailey had previously been enlarged by William Rufus.[460] Possibly the _framea turris_ or framework tower spoken of in Henry II.'s reign may refer to the wooden keep which had been left standing on the motte.[461] The walls of Gloucester Castle were frequently repaired by Henry II.,[462] but the word _murus_ by no means implies always a stone wall, and it is certain that the castle was at that time surrounded by a wooden stockade, as a writ of a much later period (1225) says that the stockade which is around our castle of Gloucester has been blown down and broken by the wind, and must be repaired.[463] Wooden bretasches on the walls are spoken of in the _Pipe Rolls_ of 1193, and even as late as 1222.[464]

The value of the city of Gloucester had apparently risen at the time of the Survey, though the entry being largely in kind, T. R. E., it is not easy to calculate.

HASTINGS, Sussex (Fig. 18).--In this case we have positive contemporary evidence that the earthen mound of the castle was thrown up by the Normans at the time of the Conquest, for there is a picture in the Bayeux Tapestry which shows them doing it. A number of men with spades are at work raising a circular mound, on the top of which, with the usual all-inclusiveness of mediæval picturing, a stockade is already erected. A man with a pick seems to be working at the ditch. The inscription attached is: "He commands that a castle be dug at Hestengaceastra."[465] There is no need to comment on the significance of this drawing and its inscription for the history of early Norman castles; what is extraordinary is that it should have been entirely overlooked for so long. In no case is our information more complete than about Hastings. Not only does Domesday Book mention the _castellaria_ of Hastings,[466] but the _Anglo-Saxon Chronicle_ also tells us that William built a castle there, while the chronicle of Battle Abbey makes the evidence complete by telling us that "having taken possession of a suitable site, he built a _wooden castle_ there."[467] This of course means the stockade on top of the motte, with the wooden tower or towers which would certainly be added to it. Wace states that this wooden castle was brought over in pieces in the ships of the Count of Eu.[468]

The masonry now existing at the castle is probably none of it older than the reign of Henry II. at the earliest, and most of it is certainly much later.[469] The _Pipe Rolls_ show that Henry II. spent £235 on the castle of Hastings between the years 1160 and 1181, and it is indicated that some of this money was for stone, and some was for a keep (_turrim_).[470] There is no tower large enough for a keep at Hastings now, nor have any stone foundations been found on the motte, and Mr Harold Sands, who has paid particular attention to this castle, concludes that Henry II.'s keep has been carried away by the sea, which has probably torn away at least 2 acres from the area of the castle.[471] The beautiful fragment of the Chapel of St Mary is probably of Henry II.'s reign; the walls and towers on the east side of the castle appear to be of the 13th century. The ditch does not run round the motte, but is cut through the peninsular rock on which the castle stands, the motte and its ward being thus isolated. The form of this bailey is now triangular, but it may have been square originally. Beyond the ditch is another bailey, defended by earthen banks and by a second ditch cut through the peninsula.[472] No exact estimate can be given of the original area of the castle, as so much of the cliff has been carried away by the sea.

Hastings itself had been a fortified town before the Norman Conquest, and is one of those mentioned in the _Burghal Hidage_. The name Hæstingaceaster, given to it in the _Anglo-Saxon Chronicle_ (1050), is a proof that the Saxons used the name _chester_ for constructions of their own, as no Roman remains have been found at Hastings. But the Norman castle is outside the town, on a cliff which overlooks it. As in the case of the other ports of Sussex, the castle was committed to an important noble, in this case the Count of Eu.

The manor of Bexley, in which Hastings Castle stood, had been laid waste at the Conquest; at the date of the Survey it was again rising in value, though it had not reached the figure of King Edward's days.[473]

HEREFORD.--There can be little doubt that the castle of Hereford was built by the Norman Ralph, Earl of Hereford, Edward the Confessor's nephew, about the year 1048.[474] It was burnt by the Welsh in 1055, after which Harold fortified the town with a dyke and ditch; but as Mr Freeman remarks, it is not said that he restored the castle.[475] The motte of Earl Ralph is now completely levelled, but it is mentioned several times in documents of the 12th century,[476] and is described in a survey of 1652, from which it appears that it had a stone keep tower, as well as a stone breastwork enclosing a small ward.[477] It stood outside the N.W. corner of the bailey, surrounded by its own ditch; the site is still called Castle Hill. If the castle was not restored before the Norman Conquest it was certainly restored afterwards, as in 1067 we find the "men of the castle" fighting with Edric Child and the Welsh. The castle appears to have had stone walls by the time of Henry II., as the mention of a kiln for their repair proves.[478] But these walls had wooden towers.[479] The timber ordered in 1213 "ad hordiandum castellum nostrum de Hereford"[480] refers to the wooden _alures_ or machicolations which were placed on the tops of walls for the purpose of defending the bases.

Though Hereford was a private castle in the Confessor's reign, it was claimed for the crown by Archbishop Hubert, the Justiciary, in 1197, and continued to be a royal castle throughout the 13th century.[481]

The bailey of Hereford Castle still exists, with its fine banks; it is kite-shaped and encloses 5-1/2 acres. The castle stood within the city walls, in the south-east angle.

The value of Hereford appears to have greatly increased at the date of the Survey.[482]

HUNTINGDON (Fig. 18).--"There were twenty houses on the site of the castle, which are now gone."[483] Ordericus tells us that the castle of Huntingdon was built by William on his return from his second visit to York in 1068.[484] Huntingdon had been a walled town in Anglo-Saxon times, and was very likely first fortified by the Danes, but was repaired by Edward the Elder. As in the case of so many other towns, the houses outside the walls had to pay geld along with those of the city, and it was some of the former which were displaced by the new Norman castle. Huntingdon was part of the patrimony of Earl Waltheof, and came to the Norman, Simon de Senlis, through his marriage with Waltheof's daughter and heiress. The line of Senlis ended in another heiress, who married David, afterwards the famous king of Scotland; David thus became Earl of Huntingdon. In the insurrection of the younger Henry in 1174, William the Lion, grandson of David, took sides with the young king, and consequently his castle was besieged and taken by the forces of Henry II.,[485] and the king ordered it to be destroyed. The _Pipe Rolls_ show that this order was carried out, as they contain a bill for "hooks for pulling down the stockade of Huntingdon Castle," and "for the work of the new castle at Huntingdon, and for hiring carpenters, and crooks, and axes."[486] We learn from these entries that the original castle of the Conquest had just been replaced by a new one, very likely a new fortification of the old mounds by William, in anticipation of the insurrection. We also learn that the new castle was a wooden one; for a castle which has to be pulled down by carpenters with hooks and axes is certainly not of stone. It does not appear that the castle was ever restored, though "the chapel of the castle" is spoken of as late as the reign of Henry III.[487]

The motte of Huntingdon still exists, and has not the slightest sign of masonry. The bailey is roughly square, with the usual rounded corners; the motte was inside this enclosure, but had its own ditch. The whole area was 2-1/2 acres, but another bailey was subsequently added.

The value of Huntingdon appears to have been stationary at the time of the Survey, the loss of the twenty houses causing a diminution of revenue which must have been made up from the new feudal dues of the castle.

LAUNCESTON, or Dunheved,[488] Cornwall (Fig. 19).--There, says Domesday Book, is the castle of the Earl of Mortain.[489] In another place it tells us that the earl gave two manors to the bishop of Exeter "for the exchange of the castle of Cornwall," another name for Dunheved Castle. We have already had occasion to note that the "exchange of the castle," in Domesday language, is an abbreviation for the exchange of the site of the castle. The fact that the land was obtained from the church is a proof that the castle was new, for it was not the custom of Saxon prelates thus to fortify themselves. The motte of Launceston is a knoll of natural rock, which has been scarped and heightened by art. This motte now carries a circular keep, which cannot be earlier than the 13th century.[490] There is no early Norman work whatever about the masonry of the castle, and the remarkably elaborate fortifications on the motte belong to a much later period.[491] The motte rises in one corner of a roughly rectangular bailey, which covers 3 acres. It stands outside the town walls, which still exist, and join those of the castle, as at Totnes. Launceston was only a small manor of ten ploughs in the time of the Confessor. In spite of the building of the castle, the value of the manor had greatly gone down in William's time.[492] The ten ploughs had been reduced to five.

LEWES, Sussex (Fig. 19).--The castle of Lewes is not mentioned in its proper place in Sussex by Domesday Book, and this is another proof that the Survey contains no inventory of castles; for that the castle was existing at that date is rendered certain by the numerous allusions in the Norfolk portion to "the exchange of the castle of Lewes."[493] It is clear that at some period, possibly during the revolt of Robert Curthose in 1079, William I. gave large estates in Norfolk to his trusty servant, William de Warenne, in exchange for the important castle of Lewes, which he may have preferred to keep in his own hands at that critical period. This bargain cannot have held long, at least as regards the castle, which continued to belong to the Warenne family for many generations. We cannot even guess now how the matter was settled, but the lands in Norfolk certainly remained in the hands of the Warennes.

Lewes is one of the very few castles in England which have two mottes.[494] They were placed at each end of an oval bailey, each surrounded by its own ditch, and each projecting about three-fourths beyond the line of the bailey. On the northern motte only the foundations of a wall round the top remain; on the other, part of the wall which enclosed a small ward, and two mural towers. These towers have signs of the early Perpendicular period, and are very likely of the reign of Edward III., when the castle passed into the hands of the Fitz Alans. The bailey, which enclosed an area of about 3 acres, is now covered with houses and gardens, but parts of the curtain wall on the S.E. and E. stand on banks, bearing witness to the original wooden fortifications. The great interest of this bailey is its ancient Norman gateway. The entrance was regarded by mediæval architects as the weakest part of the fortress, and we frequently find that it was the first part to receive stone defences.[495] It is not surprising that at such an important place as Lewes, which was then a port leading to Normandy, and at the castle of so powerful a noble, we should find an early case of stone architecture supplementing the wooden defences. But the two artificial mottes have no masonry that can be called early Norman.