THE BEGINNING OF THE STONE CASTLE
[Illustration: Map of principal castles in north-east England]
In Domesday Book some fifty castles are mentioned by name or
implication; and the number was largely increased during the next
hundred years. In view, however, of the large number of temporary
private strongholds which came into being during the twelfth century,
it is difficult to estimate the number of permanent castles until, in
the later part of the century, Henry II. regulated and restrained the
efforts of private owners to guard their property with fortresses. The
castles included in Domesday do not represent the whole number which
existed at that period; and of such important castles as Colchester and
Exeter, which we know to have been founded before 1086, there is no
mention. To estimate the strategic plan which governed the foundation
of castles at its full value, we must therefore turn for a moment
to the later period at which the defence of England by a connected
system of these strongholds had been more thoroughly achieved. Here
also, it is not altogether easy, in view of the destruction of older
castles by Henry II., and the foundation of new ones at a later
epoch, to estimate the exact state of the castles of England at the
end of the twelfth century.[98] But, taking one special district, we
may at least gain an approximate notion of its lines of defence as
they existed about the year 1200. This is the north-eastern district
of England, containing the main strategic approach to Scotland, and
crossed by the rivers which descend eastwards to the sea. This was the
scene of the rebellion of the Mowbrays and the invasion of William
the Lion in 1174, in consequence of which four important castles
at least, those of Kinnard’s Ferry on the lower Trent, Thirsk and
Northallerton in the vale of Mowbray, and Kirkby Malzeard, on the
highlands above the right bank of the Ure, were demolished.[99] The
chief castles of this district will be found to guard the line of
the rivers. On the Trent were Nottingham, on the north bank, and the
bishop of Lincoln’s castle of Newark,[100] on the south; while the
greater part of the lower valley of the river was commanded at some
distance by the strongly-placed castles of Belvoir and Lincoln. On
the borders of Nottinghamshire and Yorkshire, Tickhill[101] stood in
advance of the Don, while the narrow passage of the Don, five miles
west of Doncaster, was guarded by Conisbrough. These castles defended
the approach from the high land on the west; the marshy country north
and east of Doncaster, towards the Humber, although it often proved
a refuge for freebooters, needed no permanent garrison. South of the
Calder, opposite Wakefield, stood Sandal. To the east, below the
junction of the Calder and the Aire, was Pontefract, in a position of
great strength and importance.[102] There was no great castle on the
Wharfe, although Harewood guarded the south bank of the river between
Otley and Tadcaster, and at Tadcaster itself there was a castle, of
which little is known; Cawood castle was simply a manor-house of the
archbishops of York.[103] On the Ouse, almost in the centre of the
shire, were the two castles of York, at the head of the tideway.
Knaresborough was west of York, on the north bank of the Nidd. Each of
the dales of the North Riding had its strong castle. In Wensleydale was
Middleham, south of the Ure. Richmond, from its cliff at the mouth of
Swaledale, commanded a vast tract of country, reaching to the Hambleton
hills and the forest of Galtres, north of York. Barnard Castle stood
in a strong position on the Durham bank of the Tees. The castles of
the eastern part of the North Riding were Skelton and Castleton, both
in Cleveland, and belonging to the house of Bruce. Helmsley stood at
the entrance of Ryedale; Pickering and Malton were on the Derwent,
and Scarborough guarded the coast. South of Scarborough, in the East
Riding, the one castle of importance at this date was Skipsea, on the
low coast-line of Holderness, between Flamborough and Spurn heads.
Returning to the border of Durham, and crossing the Tees, we find
Brancepeth and Durham on the Wear. On the south bank of the Tyne was
Prudhoe in Northumberland: north of the Tyne was the great fortress of
Newcastle. Most of the castles and small strongholds of Northumberland
were the growth of a later age. The principal castles at this period
were Mitford on the Wansbeck, Warkworth on the Coquet, Alnwick on the
Alne, Wark and Norham on the Tweed, and Bamburgh and Holy Island,
castles on the seaboard. This list might be extended, but the most
important fortresses east of the Pennine chain are included in it,
and from it the strategic geography of this important district can
be readily recognised. Of the thirty-four castles in this list, ten,
including the gateway-tower at Newark, had rectangular tower-keeps, of
which nine remain; Conisbrough and Barnard Castle (87) had cylindrical
tower-keeps. Of the rest, in most cases, as at Sandal (86), the mounts
remain, and in a few instances, as at Skipsea, there are remains of
a shell-keep. The shell-keeps of Lincoln and Pickering are still
excellent examples of their type. The masonry at York, Pontefract, and
Knaresborough belongs to a later period; and in almost all instances,
where masonry remains, it bears trace of substantial later additions.
[Illustration: Sandal Castle; Plan]
[Illustration: Barnard Castle; Plan]
It will be noticed that castles which guarded passes through hilly
districts were generally placed, like Middleham and Richmond, in
comparatively open country at the foot of the pass. This was the
case with Welsh castles like Brecon and Llandovery, or the tower of
Dolbadarn, below the pass of Llanberis. The isolated position at the
head of a pass was not easily victualled, nor was it so useful as the
situation on more open ground, from which, as at Brecon or Middleham, a
larger extent of mountain country could be commanded. Trecastle (44),
at the top of the pass between Brecon and Llandovery, has already been
mentioned as a site which was probably abandoned early: the tract which
it commanded is limited compared with that within reach of Brecon, the
point towards which all the valleys of the neighbourhood converge.
In places where a castle formed part of the defences of a walled town,
it was usually placed upon the line of the wall, so that the wall
formed for some distance part of its curtain. This can be well seen
at Lincoln, where the castle occupied the south-west angle of the
older Roman city. The castle of Ludlow is in the north-west angle of
the town, the wall of which joined it on its north face and at its
south-west corner. At Carlisle the castle filled up an angle of the
town, the town walls meeting its south curtain at either end. In such
cases, the castle, while defending the town, was also protected from
it by a ditch, across which a passage was furnished by a drawbridge.
The castle of Bristol stood upon the isthmus, east of the town, between
the streams of the Avon and Frome, and, in this strong position, was
joined by the city wall at either extremity of its west side. In 1313,
when the citizens were in rebellion, they cut off the castle from
the city by building a new wall on that side.[104] In the case of
Bristol, the building of the castle made some alterations in the town
wall necessary, as time went on, but, from the beginning, the castle
occupied its place in the regular _enceinte_. If, at York, the castles
were at first built, as seems to be the case, outside the defences
of the town, the circuit was soon extended to include the castle, at
any rate, upon the right bank of the river. Although the castle of
Southampton is almost entirely gone, the points of junction of its
curtains with the west wall of the town are quite clear. Similarly,
the position of castles such as Shrewsbury, Leicester, and Nottingham,
or close to the _enceinte_ of the town, can be traced, although little
is left of the walls. In foreign walled towns like Angers or Laval,
the castle formed, as in England, a portion of the outer defences. In
later castles like Carnarvon and Conway, the same relation to the plan
of the town was preserved. There are exceptions, of which the chief is
the Tower of London, within the Roman, but outside the medieval city
wall. Chepstow is also outside the town, between which and the castle
is a deep ravine: but in this case the town was of a growth subsequent
to the foundation of the castle. A distinction must be drawn between
castles founded in connection with fortified towns, in which the castle
formed part of a general scheme of defence, as at Bristol and Oxford,
and castles under the protection of which towns, like Chepstow, grew
up, and were subsequently fortified. A good example of this latter
class is Newcastle, in which the relations of town and castle are
exactly opposite to those at Chepstow. When the castle was founded by
the Conqueror, the place, once garrisoned by the Romans, and for a time
inhabited by a colony of English monks, was probably an inconsiderable
village. The town which grew up on the site took its name from the
castle, and was walled in the thirteenth and fourteenth centuries. The
walls, however, were brought down to the banks of the Tyne at some
distance east and west of the castle, which was thus contained entirely
within their circuit. At Norwich, a place of no great importance before
the Conquest, the castle is also entirely within the line of the old
city wall. One of the best examples of the connection between a walled
town and a castle is at Launceston, where the borough of Dunheved grew
up within a narrow area which was virtually the outer ward of the
strong hill-fortress.
The establishment of a castle upon a permanent site was followed,
sooner or later, by the building of stone fortifications. This work
was often very gradual. We have seen that even a castle so important
as that of York retained part of its timber stockade as late as 1324.
This, however, was an exceptional case. The walls and towers of
medieval castles show, as might be expected, a considerable variety of
masonry; but the epoch at which their fortification in stone became
general may be said to be the third quarter of the twelfth century. In
1155 Henry II. resumed castles and other royal property into his own
hands, and ordered the destruction of the unlicensed castles which had
risen during the civil wars of the previous reign.[105] This step was
followed unquestionably by much activity in strengthening the defences
of the castles which were left.
At the same time, there are many substantial remains of stone
buildings in castles earlier than this era. Stone donjons or keeps
were certainly exceptional in England before the reign of Henry II.,
although there are a few important examples of an earlier date. It
cannot be disputed, however, that a certain number of castles were
provided with a stone curtain-wall[106] and other stone buildings not
long after the Conquest. Curtain-walls thus built would follow the line
of the earthen bank surrounding the bailey, and take the place of the
timber stockade. They were at first of the simplest form. An edict of
the council of Lillebonne in 1080 laid down the rule, so far as the
Norman duchy was concerned, for constructing the defences of private
castles; and, although the details refer primarily to the ordinary
timber structure, they also have a bearing on the construction of early
curtains of stone. No ditch was to be deeper than the level from which
earth could be thrown by the digger, without other help, to the soil
above. The stockade was to follow a course of straight lines, and to
be without _propugnacula_ and _alatoria_—_i.e._, projecting towers and
battlements, and rampart-walks or galleries.[107]
The earliest type of curtain-wall would be strictly in accordance with
these rules—a strong wall of stone surrounding the bailey, and climbing
the sides of the mount to join the defences of the donjon. We read
of the destruction by Louis VI. of France of the stone fortification
with which the house of the lord of Maule was surrounded;[108] and the
edict already quoted applies to fortifications on level ground, and
includes, not merely castles, but strong private houses, which might
not necessarily follow the castle plan. The edict, however, proceeds
to forbid altogether the construction by private persons of castles
on rocks or islands. The reason of this is obvious. Such isolated
strongholds might become, in the hands of private owners, a centre
of rebellion against the suzerain. In 1083, Hubert of Maine held out
successfully against the Conqueror in his rock castle of Ste-Suzanne
(Mayenne) on the Erve, “inaccessible by reason of the rocks and the
thickness of the surrounding vineyards.”[109] William II. in 1095
besieged Robert Mowbray in his castle on the well-nigh impregnable
rock of Bamburgh, with considerably better fortune.[110] Such rocks
formed, as it were, natural mounts which made the construction of the
ordinary mount-and-bailey castle upon them unnecessary. The hardness
of the soil, moreover, made the construction of earthworks difficult
or impossible. The natural method of defence would be to raise a stone
wall which enclosed the stronghold.
[Illustration: BAMBURGH CASTLE: great tower]
[Illustration: BAMBURGH CASTLE]
[Illustration: Richmond; great tower]
Neither at Ste-Suzanne nor at Bamburgh (91) is there existing
stonework earlier than the twelfth century. Of the castle of
Saint-Céneri-le-Gérei (Orne), which we know to have been fortified
with stone walls before the end of the eleventh century,[111] only
indistinguishable masses of masonry remain to-day. On the other hand,
there are a certain number of castles on rocky and isolated sites, the
walls of which may be fairly attributed, in whole or part, to the later
half of the eleventh century. The most important example is Richmond
castle in Yorkshire (91), on a high promontory of rock above the Swale.
The shape of the enclosure is triangular. The most conspicuous feature
of the castle is the splendid square tower or donjon, which was
completed between 1170 and 1180, and stands on the north side of the
_enceinte_, at the head of the approach from the town. The curtain,
however, west of the donjon, contains “herring-bone” masonry,[112] and
is of a rough construction which affords the greatest contrast to the
regularly dressed and closely jointed masonry of the great tower. The
tower, on three sides, forms an outward projection from the curtain,
of great size and strength, and is a structure of one period from the
ground upwards. But, on entering the castle, it is at once obvious
that the lower part of the south wall of the tower is formed by part
of the earlier curtain. In the middle of this section of the work is a
wide doorway, with a round-headed arch of two unmoulded orders, which
now forms an entrance into the basement of the tower. The capitals
of the jamb-shafts of this archway are of an unmistakably eleventh
century character, with volutes at the upper angles, and a row of
acanthus leaves round the bell. This type of capital is seen in such
buildings as the two abbey churches at Caen, the nave of Christchurch
priory, the west front of Lincoln minster, and other fabrics completed
before 1100, and is a sure guide to the date of the work in which it
occurs. It would appear, then, that the masonry of this archway and
much of the curtain is the work of Alan of Brittany, earl of Richmond,
who certainly founded the castle, and died in 1088.[113] The castle
contains more work of his date, of which something will be said in the
sequel.
[Illustration: Ludlow; Entrance to inner ward]
When the great tower of Richmond was built, an entrance was made on
the first floor, from the rampart-walk of the curtain. It is quite
clear that, up to this time, the archway just described had been the
main entrance of the castle, and had probably been covered on the
side next the town by a rectangular building, which formed the lower
stage of a gateway-tower or gatehouse, lower than the present donjon.
This is borne out by a comparison with the keep at Ludlow, where it
is quite clear that an eleventh century gatehouse was converted at a
later period into a keep, by walling up the outer entrance (94). A
new entrance to the castle, as at Richmond, was made in the adjacent
curtain, where it could be easily commanded by the tower. The date of
the lowest stage of the donjon is revealed, as at Richmond, by the
details of capitals and shafts, which in this case belong to an arcade
in the east wall of the inner portion (95).[114]
[Illustration: Ludlow; Wall arcade in basement of great tower]
[Illustration: Ludlow; Plan]
The site of Ludlow (96), like that of Richmond, is a rocky peninsula,
where a stone curtain, for which material existed on the spot, formed
the obvious means of defence. There was no mount and no keep. Exeter,
again, is an early example of a stone-walled castle upon a rocky site,
where a gateway with a tower above formed the principal entrance.
Such sites were protected naturally by the fall of the ground on the
steeper sides; the side on which approach was possible was covered by
a ditch cut in the rock. The ditch would be crossed by a drawbridge,
let down from the inner edge, next the gateway. The gatehouse itself
would be a building of two or more stages; at Ludlow the upper stage,
as completed, was considerably loftier than the lower.[115] At Exeter
there were probably three stages. A single upper stage remains at
Tickhill. At Lewes and Porchester there is clear evidence of an upper
chamber. The gatehouse at Porchester, as at Ludlow, was the entrance
to an inner ward, divided by a ditch from the large outer ward,[116]
which, at Porchester, represented the greater part of the _enceinte_ of
the Roman station, and contained the priory church and buildings. In
these early gatehouses, the lower stage was closed at either end by a
heavy wooden door, and was covered by a flat ceiling of timber. There
was no arrangement for a portcullis. At Ludlow the lower stage appears
to have been divided into an outer porch and inner hall by a cross
wall, in which there must have been a door; but communication between
these parts was also obtained by a narrow barrel-vaulted passage in the
thickness of the east wall, which, opening from the outer division,
turned at right angles to itself in the direction of the length of the
wall, and, with another right-angled turn, opened into the inner hall
(95). This passage was guarded by doors, which opened inwards at either
end.[117] When the outer doorway of the gatehouse was blocked, the
lower stage was covered in with a pointed barrel-vault.[118]
[Illustration: Porchester; Plan]
As already indicated, the details of these gatehouses are very simple,
and it is only where an attempt is made at ornament that their date
can be fairly judged. Thus at Porchester, the entrance archway, masked
by defensive work of a more advanced period, consists of an unmoulded
ring of _voussoirs_, divided from the jambs by plain impost-blocks. The
outer bailey or base-court of the castle, which is still surrounded by
the Roman walls with their semicircular bastions, has two gatehouses.
These occupy the sites of the west and east gates of the Roman
_enceinte_, and the east or water-gate is in part Roman. The western
gatehouse was rebuilt at a date contemporary with the enclosure of the
castle proper within the north-west quarter of the Roman station, and
was much altered at a later period. The archways of the Norman building
remain, and show no attempt at ornament, the inner one alone having
impost-blocks below the arches. The work at Porchester is usually
attributed to the early part of the twelfth century, and the ashlar
facing of the side walls of the inner gateway appears to be of that
date. A similar severity of detail is seen in the parallel case of
Lewes, where the original gatehouse was also covered in the fourteenth
century by a barbican (98). The great gatehouse of three stages, at
Newark castle (99), was the work of Alexander, bishop of Lincoln from
1123 to 1148, whose uncle, Roger, bishop of Salisbury (1107-39),
appears to have built the gatehouse at Sherborne. The archways of the
lower stage at Newark are of great width, and are as simple in detail
as those at Porchester. The outer or northern wall of the tower is
faced with finely-jointed ashlar, and the archway on this side has a
hood-moulding, with billet ornament.[119]
[Illustration: Lewes; Barbican]
The position of the gatehouse in relation to the curtain varies. At
Richmond, Porchester, and Exeter the inner face of the gatehouse was
flush with the curtain. At Ludlow, Newark, and Tickhill it was partly
outside, but mostly inside the curtain. At Lewes the projection was
wholly internal. Its measurements also vary. Porchester was 23 feet
in length by 28 feet in breadth: Exeter and Lewes were about 30 feet,
Tickhill about 36 feet square: Ludlow was 31 feet broad, but was some
feet longer. The area of the gatehouse at Newark is larger than any,
and the general proportions and elevation were those of a rectangular
donjon rather than a mere gateway-tower.[120]
[Illustration: Newark; Gatehouse]
Stress has been laid on the occurrence of early stone fortifications
on rocky and precipitous sites, where the ordinary earthworks were at
once impracticable and unnecessary. It will be noted, however, that
the gatehouses which have been described are not found wholly in such
positions. Tickhill and Lewes were mount-and-bailey castles with strong
earthworks. Porchester is on level ground, open to Portsmouth harbour
on two sides, and defended by a ditch on the sides towards the land:
the site was already walled, but the rectangular keep appears to stand
upon the base of an earlier mount, which may have been thrown up so
as to enclose the Roman tower at the north-west angle of the station.
Newark (157) stands on a moderate height above the meeting of the Devon
and an arm of the Trent, with a deep ditch on the north and east sides
towards the town. There was no castle here before Alexander began to
build in or about 1130; and his work from the beginning consisted of
a rectangular enclosure without a mount, in which the gatehouse had
something of the importance of a keep. The necessity of defending the
entrance of the castle, whatever the nature of the site might be, led
to the construction of stone gatehouses at an early date; and, at
Tickhill or Lewes, the gate-towers were probably constructed at a time
when the mounts and embankments of the bailey were still defended by
timber.
Stone curtains which display “herring-bone” masonry may generally
be assumed to be early in date. It has been customary to look upon
“herring-bone” masonry as indicative of pre-Conquest work, and many
buildings have been described as “Saxon” on the strength of this detail
alone. On the other hand, it never occurs in direct association with
details which may be regarded as definite criteria of pre-Conquest
masonry; and the dimensions, apart from other features, of churches
in which it is found in any quantity, usually afford suspicion of its
post-Conquest origin.[121] Its use in castles, which, as has been
shown, were a Norman importation into England, demolishes its claim to
be regarded as a distinctive sign of Saxon work; and its employment
in Normandy, especially in the donjon of Falaise, where almost the
whole of the inner face of the walls shows “herring-bone” coursing,
may be set against any theory which would attribute it to English
masons after the Conquest.[122] It was used by Roman builders, and
much of it may be seen in the towers of the _enceinte_ at Porchester.
Saxon builders, however, did not copy Roman methods of walling, and
the surest criterion of Saxon work is the thin wall, wholly composed
of dressed stone, or of rag-work without facings. Norman builders,
coming from a country where the continuity of Roman influence was never
broken, used the ordinary Roman method of a compound wall, in which
a solid rubble core was faced with ashlar on one or both sides. It
is only natural that in early stone castles, which were constructed
as quickly as possible, the facing should be of a rough description,
of coursed rubble or of “herring-bone” courses laid in thick beds
of mortar. At a subsequent date, when masonry was added to already
fortified sites, the work could be pursued in a more leisurely manner.
The most striking example of “herring-bone” work in an English castle
is in the cross-wall of the great tower at Colchester (101), which is
unquestionably a building of the eleventh century. Here the work was
evidently hurried on, with the object of securing the greatest amount
of strength in the least possible time, and Roman tiles were re-used in
large quantities as bonding courses for the rubble walls, and for the
“herring-bone” coursing of the dividing wall. At Richmond, as has been
noted, there is a certain amount of “herring-bone” work in the curtain.
The castle was founded on an entirely new site by Alan of Brittany:
earthworks were out of the question, and the date of the older masonry
of the stone wall is beyond dispute.
[Illustration: Colchester; Cross-wall]
A very remarkable example of “herring-bone” walling is the curtain-wall
at Tamworth (48). The castle was founded by Robert Marmion after the
Conquest on the low ground at the meeting of the Tame and Anker, the
town, the fortified _burh_ of Æthelflæd, being on higher ground to the
north. Marmion’s fortress took the mount-and-bailey form. The bailey
was a triangular platform of earth, raised artificially above the level
of the river bank, with its apex towards the confluence of the streams.
The mount was on its west side, and was divided from it by a ditch. The
defences on the side next the town were of stone. Here the curtain-wall
remains in very perfect condition, crossing the ditch and climbing
the mount, with a sloping rampart-walk along the top. The inner face
is composed entirely of “herring-bone” courses, alternating with one,
two, and sometimes three, layers of thin horizontal stones. This
appearance of more than one horizontal course is very unusual.[123]
It is obvious that the site, being commanded by the town, would be
materially strengthened by a stone wall on that side: on the south
side, scarping and ditching would have been sufficient, and there is
no trace of stone-work of an early period here. The original entrance
was at the north-eastern angle of the enclosure, and probably took the
form of a stone gatehouse.[124] Other instances of “herring-bone” work
in curtain-walls that may be mentioned here are at Corfe, Hastings,
and Lincoln. Corfe was built on an isolated hill, which was scarped
and ditched, something after the manner of a “contour” fort of early
days: the portion of the curtain in which “herring-bone” coursing is
found follows the natural line of the edge of the hill. Hastings is
a fortress on a steep promontory: the mount, on the east side of the
enclosure, was defended by a deep ditch, and covered by a large outer
bailey with formidable earthworks. The curtain, on the east and north
sides of the inner ward, is chiefly of the thirteenth century; but part
of the north curtain, forming the north wall of the castle chapel, is
of “herring-bone” construction. Lincoln, as we have seen, was a large
mount-and-bailey fortress, surrounded by earthworks, which, on the west
side, enclosed portions of the wall of the Roman city. “Herring-bone”
masonry is seen here and there in the west and north curtains, which
have been raised on the top of the earthen banks.[125]
[Illustration: Chepstow; Hall]
The battlemented parapet with which the curtain-wall of a castle is
usually crowned, generally may be assigned, in its present state,
to a later repair and heightening of the curtain. This is the case
at Lincoln, where the parapet and upper part of the wall are of the
thirteenth century. It has been seen that the edict of Lillebonne
in 1080 forbade the defence of the curtain by flanking towers,[126]
rampart-walks, and other aids to defensive warfare; and, as a matter
of fact, the full development of the fortification of the _enceinte_
belongs to a later period. At the same time, towers projecting beyond
the line of the curtain are found in some of our early Norman castles
of stone. The line of the early curtain at Richmond is unbroken by
contemporary towers, and closely follows the edge of the rock on which
it is built. But at Ludlow (96) where the inner ward is the original
castle, founded probably by Roger de Lacy after 1085, the curtain is
flanked by four original towers in addition to the gatehouse, which has
been described. The shape of the ward is that of a triangle with convex
sides, the base of which, on the side of the outer ward and the town,
faces south and west. Some thirty feet to the east of the gatehouse, a
tower, in the basement of which an oven was inserted at a later date,
capped the south-west angle of the enclosure, projecting southwards as
far as the edge of the ditch. The west curtain continued in a line with
the west wall of this tower for some sixty feet, until it was broken
by a small postern tower. At the apex of the triangle, projecting to
the north-west, was another tower, the remaining tower being at the
north-east angle, with its north wall in a line with the north curtain.
All these towers are, roughly speaking, rectangular in shape, but the
outer angles of the north-east and north-west towers are chamfered.
The original openings were round-headed loops with wide inward splays.
Although the curtain was thus supplied with several projections, more
towers would be needed to flank it perfectly, and large portions of the
wall, particularly on the north and east sides, were left without more
protection than could be given by their own strength. Oxford castle is
another instance of early walling, where the tall rampart tower which
commanded the river and the castle mill still remains.[127]
[Illustration: Chepstow; Plan]
Of the stone buildings which existed within the enclosures of early
Norman castles, the traces which remain are comparatively few, and in
most cases work of an altogether later period has taken their place.
The great hall for the common life of the garrison, such as Robert
d’Oily built in Oxford castle in 1074, would be indispensable. At
Ludlow there can be little doubt that the original hall stood on the
site occupied by the present hall, much of the east wall of which is
apparently of the same date as the curtain. The two lower stages of
the oblong keep at Chepstow are the hall (103), with the cellar below,
founded by William, son of Osbern, before 1071. Although the upper
stage was transformed in the thirteenth century by the insertion of
traceried windows in the north wall, and of an arch between the daïs
and the body of the hall, the walls are of eleventh century masonry,
and the plain arcade which went round them is clearly visible on the
north and west sides. In the south wall of the cellar are the loops
which lighted it; these have lintel-heads with arch-shaped hollows
cut in the soffits. The hall and cellar at Richmond, which occupy the
south-east angle of the bailey, appear to be those built by Alan of
Brittany before 1088. A few additions took place here at the end of the
twelfth century,[128] but the windows in the north wall of the hall,
which are of two lights, with edge-rolls in the jambs, are clearly
of early date. When, for a time, the great stone tower became the
fashionable form of keep, a great hall formed part of its internal
arrangements; but this was the hall of the lord’s private dwelling, and
was used by the garrison only in time of siege. Domestic buildings,
including a great hall, may sometimes have been constructed of timber
within the bailey, and at the end of the twelfth century were probably
superseded by permanent buildings of stone, like the halls at Warkworth
or at Oakham. As at Richmond, such halls would be placed against or
close to the curtain, to leave the interior of the bailey as open as
possible. In case of siege, freedom of movement within the area of
the castle was essential, and the bailey formed the natural base of
operations. The hall at Chepstow was on the highest and narrowest part
of the rocky promontory on which the castle stands, at the head of the
bailey; its south wall formed part of the curtain overhanging the great
ditch between the castle and the town (106).[129]
[Illustration: LUDLOW CASTLE: inner bailey]
Norman castle-builders were careful to provide chapels within their
fortresses. In several cases, the chapel within the bailey appears to
have been the first building of stone raised inside the enclosure. The
small chapel at Richmond, in a tower of the east curtain, is almost
beyond doubt that which was granted by Alan of Brittany to St Mary’s
abbey at York about 1085. The details are very rude in character: there
is a plain wall arcade, supported on shafts, the capitals of which
have rough voluting and no abaci. The same type of capital is found in
the wall arcading of the original gatehouse at Ludlow (95), and also,
though with more finished ornament, in that of the circular nave of St
Mary Magdalene’s chapel (108) within the same castle. Certain details
in the chapel at Ludlow, especially the bands of chevron ornament round
the arches, seem to indicate that the nave is later than the eleventh
century. The arch which divided the nave from a rectangular chancel
ending in a half octagon, is of advanced twelfth century date; and it
is clear that the chancel must have been built or remodelled at a later
date than the building of the nave. The aisled chapel at Durham castle,
which now forms part of the basement of Bishop Pudsey’s building along
the north side of the bailey, has groined vaults, cylindrical columns,
and capitals with voluted crockets and square abaci, which may be
safely ascribed to 1075 or a little later. The capitals may be compared
with those of the original gateway arch at Richmond. The classical
spirit which is so noticeable in them, and is derived directly from
the contemporary work of Normandy, is also apparent in the capitals
of the crypt of the castle chapel at Oxford. Oxford castle was founded
in 1071, Durham in 1072. At Hastings, the first of the Conqueror’s
castles, there is, as has been said, much herring-bone work in the
north wall of the chapel nave and in the vice or turret-stair of
the central tower. Such definitely architectural detail as is left,
however, belongs to a rebuilding of the later part of the twelfth
century.
[Illustration: Ludlow; St Mary Magdalene’s Chapel]
The importance of the castle chapel in Norman times, and indeed
throughout the middle ages, deserves a note. Chapels were often
richly endowed, and, as at Hastings, Bridgnorth, and Leicester,
were sometimes founded as collegiate establishments, with a dean and
canons. The collegiate church of St Mary at Warwick, founded by Roger
of Newburgh, the second Norman earl, probably had its origin in a
castle chapel, removed to a new and enlarged site within the town. The
greatest of these collegiate chapels, although one of the youngest,
was St George’s at Windsor, founded by Edward III. The chapels at
Hastings, Bridgnorth, and Leicester were churches of some size and
importance; and their chapters, like those of the secular cathedrals,
usually consisted of royal clerks, generally non-resident, whose
duties were served by vicars. As royal chapels, they were exempt from
episcopal jurisdiction; and the term of “free chapel,” which was given
to them, became applied in course of time to chapels founded in private
castles and even upon manors.[130] In most cases a castle chapel was
served by a single priest, either the incumbent or his vicar. The
incumbent of the free chapel of St Michael in Shrewsbury castle,
usually a royal clerk holding his grant from the king, and inducted
by the sheriff as the king’s officer, held the church of St Julian in
Shrewsbury as parcel of his cure.[131] Where the Norman castle and
parish church stood side by side, as at Earls Barton or Higham Ferrers
in Northamptonshire, the lord of the castle and his household would
doubtless attend the church. But the foundation of a chapel within
the castle was a common thing, even when the church, as at Ludlow or
Warwick, was at no great distance; and in later years, when chantry
foundations became usual, castle chapels increased in number. Thus
at Ludlow, a second chapel, served by two chantry priests, was built
within the outer bailey about 1328;[132] and a college of eight chantry
priests was founded in 1308 by one of the Beauchamps in his castle of
Elmley in Worcestershire.[133]