A multi-period site encompassing part of a Roman legionary annexe, part of a Roman colony, a classical temple known as the Temple of Claudius, a late Anglo-Saxon or Norman chapel and associated buildings, and a Norman hall-keep castle known as Colchester Castle.
Reasons for Designation
Colchester Castle and the Temple of Claudius, a multi-period site that encompasses part of a Roman legionary fortress annexe, part of a Roman colony, a classical temple, a late Anglo-Saxon or Norman chapel and associated buildings, and a Norman hall-keep castle, is scheduled for the following principal reasons:
* Rarity: as the rare remains of a classical temple and a Norman hall-keep castle;
* Historic interest: the site of the first Roman legionary fortress, earliest colony and first provincial capital. The Temple of Claudius served as the centre of the Roman imperial cult in Britain and a primary target of the Boudiccan revolt (AD 60-61). It was occupied in the Anglo-Saxon period and then provided the foundation for a Norman hall-keep castle, which served as a physical manifestation of the wealth and power of England’s new king, William the Conqueror;
* Architectural interest: The Temple of Claudius is the largest classical temple in Britain, upon which was built the largest Norman keep in existence; a royal castle at the forefront of medieval military architecture;
* Survival: Roman, Anglo-Saxon and medieval archaeological remains survive to over 3m deep beneath Castle Park, including well preserved charred organic remains relating to the sacking of Colchester during the Boudiccan revolt. Most of the Norman hall-keep remains upstanding, including considerable architectural detail of the entrance, chapel, corner turrets and outer walls;
* Diversity: the multi-period remains demonstrate occupation of the site from the Roman period onwards and will enhance our understanding of Colchester, a major urban centre for nearly 2000 years;
* Potential: a large proportion of the site still remains unexcavated and there is a high degree of potential for further investigation, which will provide a particular insight into the relationship between the Romans and the native population, as well as the manner in which the legionary fortress was converted into a Roman colony;
* Documentation: the site is well documented in both historical and archaeological terms, including references to the Roman temple in Tacitus’ Annals and the Norman hall-keep in the medieval Colchester Chronicles;
* Group value: with the designated sites of late Iron-Age Camulodunon; the Roman city walls and gateways; Roman circus; the Anglo-Saxon Holy Trinity Church, and the many listed medieval buildings in Colchester, as well as the registered C19 park, Castle Park. Altogether these form an impressive and valuable ensemble that well illustrates the development of Colchester.
Colchester is situated on a low plateau between the river Colne to the north and the Roman River to the south. It was the location of a territorial oppidum known as Camulodunon in the late Iron Age; a tribal capital of the Trinovantes and a defended settlement. Established by c50 BC, the oppidum occupied an area of c28 square miles. It was partly bounded by dykes at the west, effectively linking the two rivers and creating a defended perimeter. There were two main centres of agricultural and industrial activity within it; at Sheepen and Gosbecks, respectively. The latter is considered to have been the residence of King Cunobelin and his descendants, who controlled a substantial portion of south-eastern Britain prior to the Roman conquest. The name Camulodunon appears on early coinage (as ‘CAM’) and in Roman literary sources (as ‘Camulodunum’), and has hence been heralded as Britain’s oldest recorded town (Gascoyne and Raford 2013: xv). It formed a major strategic objective for the Roman invasion force in AD 43. The Emperor Claudius is recorded as having led the advance on Camulodunon himself, where he received the submission of several British tribes. He subsequently established Britain’s first legionary fortress within the native oppidum but separate to the two existing main tribal centres. The fortress included a small defended annexe to the east. The remains of the Roman legionary fortress annexe extend beneath Castle Park; part of the annexe defences survive below-ground within the park.
In AD 49 the Roman legion (probably Legio XX) was relocated, the defences of the fortress were slighted and a Roman colony (a settlement for retired legionary soldiers) was established; Colonia Victricensis (loosely meaning ‘City of Victory’). It was the first of four colonies and served as the first capital of the province, Britannia. A new street grid was laid out, some of the military buildings were re-used as dwellings, and the fortress annexe was extended to accommodate public buildings. Among these was the Temple of Claudius; the centre of the Imperial cult in Britain. The temple was erected in cAD 54 and may have been incomplete when Boudicca, Queen of the Iceni, led a revolt against the Romans in cAD 60-1. Her primary target was the colony, which was sacked and largely destroyed. The Roman historian Tacitus (cAD 56-c120) records that the citizens of Colchester made a last stand in the temple, holding out for two days before they died. Following the revolt, the Roman colony was re-built and a town wall constructed. It enclosed a rectangular area of about 108 acres. Much of the wall’s circuit is still upstanding and passes east to west across the northern boundary of Castle Park (Scheduled Monument No.1003772). The area now covered by the park is situated upon six street blocks or insulae of the Roman colony, including the Roman temple.
The Temple of Claudius was dedicated to the deified Claudius after his death, and served as a venue for religious rituals and public ceremonies. It is the largest known classical temple in Britain. The podium (raised platform) of the temple survives as the base of the Norman keep. Although the superstructure of the temple does not survive, the size and position of the podium allows it to be determined. The building was approached via steps at the south. It comprised the podium supporting the cella (inner chamber) of the temple and the pronaos; a colonnaded entrance façade formed of eight columns (octastyle) supporting a pediment. The cella was flanked with columns to each side but had a solid wall at the rear. The distance between the columns was two and a quarter times their diameter at the base (eustyle). In front of the temple was an altar. The whole complex was enclosed to form a sacred precinct or temenos. This appears to have originally been defined by timber structures before a masonry enclosure was built in cAD 80-100. The latter was perhaps formed by ranges with open porticos and rooms on three sides, and a monumental arcade with a central arched entrance on the fourth (south) side. It has been suggested, based on tentative evidence, that the temple may have been rebuilt in the C4 with a basilical hall on the podium, flanked on the south by a long narrow entrance hall with an apsidal east end. However, this remains uncertain, and such a conversion is unlikely. The precinct appears to have been occupied for domestic purposes in the late C4 or early C5, possibly as a defended residence.
The transition of Colchester from the Roman to Anglo-Saxon period is not well understood. Evidence indicates it was preceded by a decline and contraction in the settlement. Excavations have shown that Anglo-Saxon Grubenhäuser (sunken-featured buildings) were constructed from about the early C5. In the middle (c660-899) and late (c899-1066) Saxon periods Colchester was a defended burh (fortified settlement). A middle Saxon cemetery developed within the former temple precinct; eight burials have been recorded through partial excavation. In c916 King Edward the Elder captured Colchester from Danish occupants and subsequently re-founded it, repairing the fortifications. It subsequently hosted meetings of the Witenagemot (royal council). In the C10 the former precinct may have been part of a villa regalis (royal manor or estate). The ranges of the temple precinct remained upstanding and there is tentative evidence that the south range was in use; partial excavation has recorded late Saxon slots dug between the piers of the arcade. The temple (or basilical hall) may also have been in use at this time. A wooden chapel was built just to the south in the late Anglo-Saxon period or slightly later, and was decorated with a wall painting of the Virgin and Child. Structural remains belonging to two or more buildings have also been identified nearby. The wooden chapel was subsequently succeeded by a masonry building. At a later date, but possibly before the construction of the Norman keep, a domestic stone hall was built to the south. It formed the western block of a larger double pile building.
After the Norman Conquest, a royal castle was built on the site. In about the late 1060s or 1070s construction of a Norman keep was initiated by the steward Eudo Dapifer on behalf of William the Conqueror. The temple site was utilised probably because the remains provided an immediate source of building material and the podium an appropriate platform. However it may also have been symbolic; the Norman rulers were perhaps asserting their status as the inheritors of Rome and a continuity of authority from the Anglo-Saxon villa regalis. The north, east and west walls of the Norman keep were butted directly against the temple podium (rather than upon it), whilst the southern wall extended beyond it. This extension contained the castle well, which would have otherwise been difficult to sink through the podium.
The Norman keep was built in at least two main phases. In the first phase a single-storey keep was constructed. It stood 9m high with an entrance in the south-west corner reached by a staircase and a crenellated parapet wall. Soon afterwards its corner towers were raised by one storey. A bailey was created to the north of the keep, consisting of a substantial bank and ditch, probably with a timber palisade. The walls of the former temple precinct were upstanding at this time and substantial parts of it were buried beneath the ramparts. In the second phase the keep was raised by at least one storey, and a fore-building was added on the south side to protect the main entrance. The earlier crenellations were retained and incorporated into the outer walls. A secondary entrance was added to the keep at first-floor level. The bailey was also extended southwards and the chapel there rebuilt. The completed keep was the largest (in area) known to have been built by the Normans. It was a fortress palace which shared similarities in design to the White Tower, London, (built from c1075-9) particularly an apsidal projection for a chapel at the south-east corner, and Norwich Castle (c1100). The apsidal projection may have been influenced by an earlier castle at Ivry-la-Bataille, Normandy, of c1000. The number of storeys and layout of the finished keep is uncertain, largely due to damage caused in the C17 (see below), and have been widely debated; the most recent expounded view is that it was only ever two-storeys high (Berridge 2016).
Colchester Castle and the town were granted to Eudo Dapifer in 1101. He held them until his death in 1120 when they reverted back to the crown. Substantial sums were spent on repairs in the late C12 and early C13. A new bailey was added shortly before the rebellion of 1173-4. This was probably the lower (or ‘nether’) bailey located between the northern defences of the upper bailey and the town wall. By 1182 a stone wall had been erected on the south and west sides of the upper bailey, probably on top of the bank, with a main gate at the south-west corner. During the First Barons’ War (1215-17), the castle was occupied by the French and garrisoned in support of the barons against King John. It was recovered but suffered damage during the siege and numerous repair works were carried out in the following years. A barbican with two D-shaped towers was added to strengthen the entrance in the C13. From the C14 onwards the military value of the castle appears to have declined. Its main function was now as a prison (later the county gaol), a role it had fulfilled since at least 1226. It occasionally hosted trials, including trial by combat, which was last recorded in 1375.
The castle was allowed to decay in the C15 and C16 and the upper bailey buildings, along with much of the bailey wall, had collapsed by 1622. It was alienated by the Crown in 1629. At about this time several prisoners appear to have been buried outside the south entrance of the keep; partial excavation has recorded several inhumations. From 1645 the ‘Witchfinder General’ Matthew Hopkins used Colchester Castle as a base to interrogate women suspected of witchcraft; he is considered to be responsible for at least 100 deaths. During the English Civil War, Parliamentarians laid siege to Colchester. Royalists attempted to make the castle defensible in 1648 and it probably suffered damage before the town fell. At the end of the siege two Royalists; Sir Charles Lucas and Sir George Lisle, were held in the room now known as the ‘Lucas Vault’ within the castle before being executed by firing squad on 28th August 1648. The upper bailey was divided into yards and gardens in around the mid-C17, and houses began to encroach onto the site of the bailey ditch. The bailey walls were dismantled and the stone sold in 1649. House plots were laid out over the bailey ditch in 1680-3. The keep was sold to John Wheeley for building materials in 1683. He attempted to dismantle the towers and walls using gunpowder and excavated the sandy soil beneath the arched foundations in search of treasure, creating the vaults that remain today. This enterprise eventually proved unprofitable and was abandoned. The castle again reverted to a county prison or House of Correction, a function it continued until 1835. The fittings of the prison cells survive in the ground floor of the keep.
In 1727 the castle came into the possession of Charles Gray, antiquarian and MP for Colchester, who resided at nearby Holytrees house. He employed the architect James Deane to carry out repairs and alterations to the keep, including new segmental-headed windows, a study above the north-east tower in 1746, and a dome above the great staircase in 1760. The area around the castle was landscaped as pleasure grounds. A walk was created along the upper bailey from a rotunda to a summerhouse, and a parterre and canal garden were laid out to the north. The remainder of the land was divided into fields. In 1782 the property passed to the Round Family. The crypt of the keep was opened as a museum holding archaeological collections in 1860. In 1892-6 the surrounding grounds were altered to form a public park, now known as Castle Park (a Grade II registered park, List entry No: 1000208). The keep was gifted to Colchester Borough Council in 1919 to commemorate the First World War; a memorial was erected by the south entrance. The vaults beneath it were reinforced in 1931 by the Office of Works, and in 1934-5 the keep was roofed in steel over a concrete frame with a bridge built to the entrance. The whole building was thereafter occupied by the museum, a function continued to the present day (2017). During the Second World War, the Roman vaults were used as an air raid shelter whilst the Norman keep displayed exhibitions in support of the war effort. Conservation work to the historic fabric was carried out in the 1980s and in 2013-14.
The first significant excavations on the site were carried out by Revd J T Round in 1845. Seven years later the Essex Archaeological Society held its inaugural excavation in Holytrees Meadow, which was overseen by Dr P M Duncan. In 1892 Dr Henry Laver directed excavations during the laying out of Castle Park. Mortimer Wheeler excavated a series of Roman town houses in the park in 1920. He also determined that the Norman keep had been built over the Temple of Claudius. The RCHME carried out a survey of the upstanding remains, which was published in 1922. Dr Philip Laver continued in the footsteps of his father Henry, undertaking excavations at the keep in 1922 and immediately to the south in 1931-33. In 1926 and 1954 Rex Hull directed excavations in Holytrees Meadow, identifying the remains now known as the Roman waterworks. The bailey bank was excavated by M Alwyn Cotton in 1950, followed by further works on the defences by Ros Niblett in the 1960s. Max Hebditch uncovered part of the south side of the Roman temple precinct in 1964. The castle well was examined by the North East Essex Aqua Club in 1972, followed by excavations in and around the keep by P J Drury and E J Rodwell in 1977. A roof-top excavation of the keep chapel was undertaken in 1988. In 1992 a lift-shaft was excavated within the building by the Colchester Archaeological Trust. An architectural survey of the keep was carried out by Historic Royal Palaces in 1999. Watching briefs or evaluations have been undertaken on the site of the temple precinct and castle by the Colchester Archaeological Trust in 2001, 2002, 2006, 2010 and 2012. An excavation by the Trust prior to the erection of an apartment block; One Castle Park, revealed part of the monumental arcade on the south side of the temple precinct in 2015-16. This is considered to be the largest Roman arcade in Britain, including a monumental arch or gateway at the centre.
A multi-period site encompassing part of a Roman legionary annexe, part of a Roman colony, a classical temple known as the Temple of Claudius, a late Anglo-Saxon or Norman chapel and associated buildings, and a Norman hall-keep castle known as Colchester Castle.
ROMAN LEGIONARY FORTRESS ANNEXE AND COLONY
The Roman legionary fortress annexe laid out in AD 44 was located immediately east of the fortress and covered a large area, part of which is beneath Castle Park. A V-shaped ditch and rampart, which served as the defences enclosing the legionary annexe, survives as buried remains just to the east of the north bank of the upper bailey of the medieval castle (TQ9996125429). It marks the north-east corner of the Roman annexe defences, which continue to the west and the south from this point.
The Roman colony (Colonia Victricensis) founded in AD 49 included 46 street blocks or insulae (Gascoyne and Radford 2013: 105). Five of these street blocks survive as buried remains beneath Castle Park, including Roman street and road surfaces, drains, foundations, walls, floors and tessellated pavements. The insulae are typically numbered as part of the wider street grid and the area covered by Castle Park is two street blocks wide. Hence the street blocks beneath it include from north to south: insula 6 and 7, insula 14 and 15, and insula 22 (the Roman temple). There are also partial remains of insula 23 at the south-east corner of the park. The north wall of the colony survives as upstanding remains and is a separate scheduling (List entry No.1003772) that passes east to west across the north boundary of Castle Park. An intervallum road runs inside it and survives below-ground.
At the north-west of Castle Park are the buried remains of insula 6; a rectangular street block extending c128m east-west by c91m north-south. It is delimited by the intervallum road at the north and three streets on the south, east and west. Four houses have been partially excavated fronting onto the street to the south (see Hull 1958 for ground plans, 81-85). Three of these buildings were uncovered in 1920. The trenches were subsequently backfilled and they survive as buried remains beneath the gardens and bandstand, c120m north of the Norman keep (TL9987125466). They include a corridor house and two courtyard houses built in the mid-C2 but with later alterations. The courtyard houses are positioned adjacent to each other, at the south-east corner of the insula. They probably formed a single residence at one time, c46m wide and 70m long. Each house is rectangular in plan with four ranges opening onto corridors surrounding an open courtyard or garden. The walls are built of septaria and tile courses, and the floors are largely tessellated pavements. The remains of a hypocaust survive in the north range of the eastern house. Fragments of painted plaster indicate that the rooms were decorated with wall paintings. The corridor house is separated from these two buildings by an open passage at the west. It is a long rectangular range orientated north-south, c7.5m wide and over 32m long. The building comprises at least seven rooms opening off a west corridor, with a further corridor or open passage to the south running parallel to the street. Most of the rooms have tessellated floors, and three have traces of mosaics with geometrical and guilloche patterns. Immediately to the south are the buried remains of a Roman drain and street that serviced the buildings. The street is a cambered metalled surface, formed of pebble, gravel and tile fragments. It is c7.5m wide with a septaria retaining wall on the north-side. The street continues to the east to form the southern boundary of insula 7. At the south-west corner of insula 6 are the exposed upstanding remains of a house uncovered in 1892 during the laying out of Castle Park. It is over 13m long by 6m wide and includes four rooms with red tessellated pavements divided by modern York stone slabs capping the original walls.
The buried remains of insula 7 survive at the north-east of Castle Park, beneath the area known as Hollytrees Meadow. It is a square block extending c99m east-west by c99m north-south, delimited by roads on each side (Hull 1958, 85-91). A tessellated pavement was recorded within the insula in 1848, although its exact location is uncertain. On the east side, near the boundary of the park, are the remains of a building; two septaria stone walls were recorded in 1927-8. At the south-east of the insula are the buried remains of two further buildings with red tessellated pavements; one possibly forming a corridor running north-south c1.5m wide by 10m long, and another forming a room c4m wide and over 5m long. An arched drain runs beneath a street on the east side of the insula, near the east boundary of Castle Park. The street is a cambered metalled surface, c8.5m wide and up to 0.8m thick.
Immediately south of insula 6 is insula 14 (centre west of Castle Park). It is c134m wide and 36m long (north-south), delimited by a road on each side. The ditch of the medieval upper bailey has encroached upon a large part of it. However, when the steps of a rose garden were built in 1929 several Roman finds were recovered indicating that the Roman levels are partly preserved in this area.
A Roman waterworks, a shrine and several houses, have been recorded through partial excavation within insula 15 beneath the east side of Castle Park (Hull 1958, 106-113). This insula forms a trapezoidal area c94m wide at the north but c88m wide at the south and c100m long. It now covers the southern half of Hollytrees Meadow. The waterworks is situated near the south-east corner (TM0000325376) and is constructed of septaria and tile with opus signinum floors. It is an L-shaped building of at least two phases; originally a rectangular range, c30m long and c8m wide, orientated north-south but with an additional wing, c7m by c5m, built onto the south end of the east wall at a later date. There are six rooms within it. At the centre of the main range is a large sunken room, c12m long by 6m wide, with walls up to 1.8m high and nearly 1m thick, faced internally with tiles. There are several channels in the floor, as well as a sump-hole 1m square and 0.5m deep. A spring rises near the centre of the building. There are slots within it, either for timber partitions or to hold machinery. Partial excavation in 1927-8 recovered iron shackles, indicating that slaves worked on the site. Surrounding the building is a rectangular walled enclosure formed by double walls at the south, east and west, and a single wall at the north. An arched drain extends from the north-east corner of the main sunken room. It takes a sinuous course north-east for c21m, where it meets the north-south road, before continuing 168m north on a straighter course beneath the road, through insula 7, out of the north gateway and into the outer ditch of the Roman colony. It is built of tile upon a blue lias stone base. The drain increases in size along its length and varies from c0.3m-0.6m wide and 0.7m-1.4m high at the outlet near the Roman gateway. Two sections of the drain are partially exposed beneath metal grates within Castle Park.
There are the remains of several other buildings within insula 15. Just to the west of the waterworks are the foundations of a small rectangular building, c7m long and 5m wide, with a porch attached to the west side. It is orientated east-west and may be a shrine. There are also remains of several houses. A tessellated pavement was identified at the north-west corner of the insula in 1852-3 (TL9994225440). Septaria walls and tessellated floors were partially uncovered on the west side of the insula in 1927-8. The largest was c6m long and 2m wide. On the north side workmen excavating a goalpost uncovered a red tessellated pavement around 4m square and a hypocaust c3.5m square.
THE TEMPLE OF CLAUDIUS
The Temple of Claudius occupied insula 22 of the Roman colony, and survives as upstanding and buried remains beneath the south-west of Castle Park. It comprises a walled rectangular precinct or temenos within which are located centrally the temple podium and altar. The temple podium survives as upstanding and buried remains that form the base of the Norman keep. The superstructure of the temple no longer survives; its place being occupied by the keep. However the size and position of the podium indicate that it comprised a cella (inner chamber) at the north flanked by columns and a pronaos (colonnaded entrance façade) at the south formed of eight columns (octastyle) supporting a pediment.
The temple podium is a rectangular platform orientated north-south built of ragstone rubble faced with septaria and brick. It is c32m long, 23.5m wide and 4m high. The platform is formed of two types of foundations; solid, load-bearing, stone foundations that originally supported the outer walls and columns, and arched foundations that originally supported the floor. The solid outer foundations are up to c6m thick at the east and west, and c2.5m thick at the north and south. Within these are the arched foundations, formed of four irregular arches separated by cross walls up to 1.8m thick. These arches were originally constructed by smoothing the sandy soil into curves, over which wooden shuttering was placed and then stone and mortar built on top. The wooden shuttering subsequently rotted away but marks remain in the mortar on the underside of the arches. The soil beneath the arches was removed in the C17, thereby forming four arched voids or vaults that survive under the Norman keep. They comprise two long vaults, c18.5m long by c6.5m wide, underlying the position of the cella, and two short vaults, c8.5m long and c6.5m wide, under the pronaos. There are also several later subdivisions and reinforcing walls within the vaults; those built in brick are probably C18 but those in concrete were added in 1931. On the south side of the podium are the foundations of Roman steps that led to the temple entrance.
The altar of the temple survives as buried remains, c18m south of the Norman keep. It is formed of stone foundations c10m square. A vaulted drain 0.5m wide has been recorded c4m north and west of the altar, and probably follows a square course around it. The drain is built of tile and stone and lined with plaster. Two pedestal bases for statues survive north and west of the altar. These may be part of a grouping that originally surrounded it. They are constructed of coursed tile and are c1.5m wide and over 3m long.
The surrounding precinct or temenos survives largely as buried remains. It was originally c164m long and 150m wide, with ranges on the north, east and west sides, formed of open porticos and enclosed rooms. The walls are constructed of dressed septaria and tile with a rubble core. The north range is partly located beneath the north bank of the upper bailey and comprised an inner and outer wall c8m apart. Partial excavation has shown that these walls survive up to 3m high beneath the bailey bank. Between the walls are the remains of partition walls and floor surfaces. There are two exposed sections of the outer wall of the north range upstanding within the park (at TL9981325388). These were uncovered in 1892 during the laying out of a path across the upper bailey and are Grade II listed (List entry No. 1123675). They are built of septaria, capped by modern concrete slabs and septaria, and measure c3m by 1m and c0.5m by 1m. The west range survives partly as buried remains close to the west boundary of Castle Park. Partial excavation has uncovered a vaulted drain that ran underneath a metalled road outside the north-west corner of this range. It is partly exposed within modern railings adjacent to Ryegate Road. The south side of the precinct was formed by a monumental arcade resting on a foundation platform c4.5m wide and 1.6m deep. This is considered to be the largest Roman arcade in Britain. The footings of the arcade survive as buried remains, immediately to the south of Castle Park, some of which are exposed beneath a glass covering within the floor of the apartment block One Castle Park. The arcade was formed of stone piers with half-engaged columns, around 3.5m long by 2m wide, which supported the arches. It has the footings of a monumental entrance arch or gateway at the centre. The west side of the gateway survives to approximately 1.8m above foundation level.
LATE ANGLO-SAXON OR NORMAN CHAPEL AND ASSOCIATED BUILDINGS
In the C10 the former temple precinct may have been occupied by a villa regalis (royal manor or estate). The ranges of the temple precinct remained upstanding and there is tentative evidence that the south side was in use; partial excavation has recorded late Saxon slots dug between the piers of the arcade. Immediately to the south of the temple podium are the buried remains of a chapel built in the late Anglo-Saxon period or slightly later. It has a concrete floor and foundations for timber-framed walls of plaster infill over wattles. Fragments of wall plaster found in association with the building include part of an interior wall painting of the Virgin and Child. The chapel was subsequently rebuilt in masonry. To the south and west are the buried masonry sills and concrete floors of two or more timber-framed buildings. In about the late C11 or early C12 the chapel was rebuilt a second time. This structure is a single-celled apsidal building c15m long by 7m wide. The walls are 0.8-0.9m thick and constructed of septaria with some Kentish ragstone and Roman tile. In the early C13 the chapel was reconstructed for a third time when a straight wall was built to replace the apse at the east end. The walls survive as low footings immediately to the south of the later keep.
A domestic stone hall was built to the south of the chapel, possibly prior to the construction of the Norman keep. Partial excavation on the southern boundary of Castle Park has recorded the buried remains of a rectangular range orientated north-south. It is c16m long by 6m wide and formed the western end of a block extending eastwards. The walls are built of septaria to c0.9m thick with quoins of Roman tile and survive up to 1.1m high. There are two doorways and a semi-circular fireplace, inserted at a later date, in the west wall.
THE NORMAN CASTLE
The Norman keep, built in late 1060s or 1070s, is rectangular in plan with square turrets projecting from three corners and an apse projecting east from the south-east angle. It is 46m by 33.5m wide, with the longer axis running north-south, and c27m high. The building survives to two storeys high. The walls are nearly 4m wide at the base with foundations c7.5m deep. They are constructed of coursed rubble, consisting of septaria, Roman tile and ragstone, with dressings of ashlar (Barnack, Caen and other freestones) and Roman tile. Much of this is reused Roman material. The north, east and west walls abut the Roman temple podium but the south side is placed beyond it. At the base is a plain chamfered string-course of Barnack stone, below which the walls batter outwards into a sloping plinth c5m high. The keep has flat pilaster buttresses that divide the north wall into two bays, the east and west walls into three bays and the apse into five bays. There are two main phases of construction visible between the lower and upper storey. The first phase, of finer construction, is marked by ashlar quoins and the upright joints of a temporary crenellation that can be seen in the east and west walls, and the south-west turret. The second phase is marked by the use of Roman tile quoins. Numerous putlog holes (for medieval scaffolding) can be seen. There are narrow loopholes to the ground and first floor. Large segmental-headed windows, with brick mullions and leaded lights, were inserted into the keep in the mid-C18. There are six in the south wall and several in the apse, north-east turret and south-west turret. The keep is largely spanned by a C20 felt and glass roof but there are shallow-pitched tiled roofs to the south side, apse, south-west turret and north-west turret. The latter were added in the mid-C18 when the keep was mistakenly thought to be a Roman building; ‘Mediterranean style’ tiles were therefore considered to be in-keeping. The conjectured site of a chapel on top of the keep is covered by a timber and glass roof structure added in 1988-9.
A 1930s footbridge (refurbished in 2013-14) now provides access to the main entrance of the keep, which is situated just above the plinth, at the west end of the south side. Beneath the bridge are the low walls of a narrow rectangular forebuilding, entered from the east. This abutted the south wall and projected out in front of the entrance. The walls are nearly 1.5m wide and built of septaria with tile quoins. The entrance was further strengthened in the C13 when the forebuilding was incorporated within a larger barbican, low walls of which also survive. These abut the western half of the south side of the keep, projecting about 15m from the main doorway. The barbican is entered between D-shaped turrets at the south. The walls are 1.5m wide and constructed of courses of septaria and tile. There are several loopholes in each side. Internally the building is divided into two compartments by a wall with a doorway at its north end.
The main doorway into the keep is through a round-headed arch of three moulded orders, above which is a hoodmould enriched by a double semi-circular motif. It rests on imposts supported by two orders of columns of which only the capitals and one base survive; the inner order have cushion capitals and the outer order have capitals decorated with volutes and leaves. Behind the arch are slots for a portcullis. The ground floor of the keep comprises three small rooms at the south, divided by thick walls, and two large rooms or halls at the north. There are no fireplaces or garderobes on this floor and only the embrasures for loopholes in the walls, although a few have been widened in the C18 to form segmental-headed windows. The south doorway leads through to a small vestibule, which contains a small alcove probably for a guard. This is the first of the three rooms at the south of the keep. On the left is a round-headed doorway leading to the Great Stair in the south-west turret; a stone spiral staircase nearly 5m wide with a rising tunnel vault. It is one of the largest Norman staircases in England. On the right, sub-divided off by a mid-C18 arcade, is the castle well, which is about 15m deep. It was damaged in the 1680s and restored in brick in 1787. Next to the well is a staircase leading to the ‘vaults’ or Roman temple foundations under the keep (see the temple description above). Further east are two small chambers entered from the north; a rectangular barrel-vaulted room known as the ‘Lucas Vault’ and a crypt within the apse. The latter was sub-divided with wooden-boarded partitions to form prisons cells in 1727. These have iron grills, studded and boarded doors and wooden floors. It is covered by two vaults; a cross-vault at the west end and a barrel vault with an apsidal half-domed end at the east. The space immediately outside the prison cells includes several timber posts with arched braces supporting the floor above; these were possibly added when a museum exhibit was inserted into the building in the 1930s. To the north, occupying the rest of the ground floor, are two large rooms; a wide western space and a narrow eastern one. These contain several late C20 and early C21 partition walls and floors for museum services, meeting rooms, offices and toilets, as well as numerous museum display cases, a lift, and modern staircases to the first floor. The western space was originally subdivided by a longitudinal wall or arcade but only traces of it remain. The north-west and north-east turrets are solid at this level.
On the first floor the space was originally divided in a similar way to the ground floor. It is provided with fireplaces, latrines, and a greater number of embrasures for loopholes, indicating that it was a principal residential floor. Beam slots in the walls indicate the position of the original timber floor. There were originally two large rooms at the north; a wide western space (c29m by 18m) that may have been the Great Hall and a narrow eastern one (c28m by 6m) that may have served as apartments, or, more specifically, a bed chamber and audience chamber. However there is now a C20 gallery occupying the Great Hall at this level containing museum displays. An aisle may have run along the east side of the Great Hall; the possible scar of an arcade is visible in the north wall. There are four large fireplaces; two each in the east and west walls, set in line with the buttresses on the exterior, with rounded arches and double flues. In the north-west turret are two garderobes, one of which is blocked, and a spiral staircase. Next to it, in the north wall, is a blocked round-headed doorway that provided a postern (secondary entrance) directly into the Great Hall of the keep. It opens out onto a landing, which was originally accessed via a flight of timber steps. The north-east turret contains a barrel-vaulted room. Next to it is the narrow eastern space that probably served as apartments, separated off from the Great Hall by a partition wall of herringbone brickwork. These were serviced by a garderobe with a small barrel-vaulted lobby set into the thickness of the east wall of the keep. Within the apsidal south-east angle is a chapel or, less likely, a sub-chapel; a low vaulted room with two pairs of apsidal side chambers covered by groined cross vaults. It is lit by two loopholes and three large windows inserted in 1754-5; two in the south wall and one at the east. There appear to have been two original entrances at the north-west but a further entrance has since been inserted into the north wall. Immediately west of the chapel is a library, added in 1754-5 for Charles Gray, which now serves as a museum learning space; ‘The Charles Gray Room’. This may originally have been the location of an antechamber. It is lit by two large windows either side of an (ex-situ) late C15 or early C16 fireplace with an elaborately carved early C17 overmantel. On the north side of the library is an arcaded passage of brick and stone. The south-west turret contains a small barrel-vaulted chamber in addition to the Great Stair at this level.
At the top of the keep are further built remains. A short stretch of the west wall next to the north-west tower survives to approximately 3m higher than the rest of the wall and may indicate the original height of the battlements. At the south-east corner the apsidal wall survives to nearly 2m high and has four inner pilasters in line with those on the exterior. It has been suggested that there was a chapel in this location but this has recently been disputed (Berridge 2016: 61); a 1988-89 timber and glass roof structure is placed in the former conjectured position of the chapel. A chamber is built within the thickness of the buttress of the south wall and, although previously thought to be part of the chapel, may instead be part of an enlarged corner tower. Further west there are the remains of an inner wall for a south passage that probably formed a walkway between the original roof edge and outer wall. There are two mid-C18 additions at this level; a 1760 extension of the south-west turret covered by a brick dome with a tile roof and a 1746 extension of the north-east turret covered by a pyramidal tiled roof. The latter formed a study. The traces of a passage along the west and north walls of the keep are considered to have been a walkway constructed under Charles Gray in the C18, providing a protected route to the study.
The Norman keep was set in the middle of an upper bailey with a lower (or ‘nether’) bailey covering the ground between the upper bailey and the town wall to the north. The defences of the upper bailey comprised a substantial bank and ditch, originally topped by a palisade. The northern and eastern arms of the defences survive as landscaped earthworks within Castle Park. The bank is about 4m high and 28.5m wide. South of the castle the bailey earthworks have been levelled but survive as buried remains. Partial excavation indicates that the ditch is about 22m wide and over 5m deep. The High Street appears to have been diverted to skirt the outer edge of it. Immediately to the north of the upper bailey are the earthworks and buried remains of the lower bailey, which was also enclosed by a bank and ditch. The southern end of the eastern arm of the defences are visible as a landscaped ditch, whilst the western arm is marked by a slight rise near the course of Ryegate Road. Partial excavation has recorded the denuded remains of the eastern defences surviving below-ground. The bank is c5.5m wide and just over 0.5m high whilst the ditch is c10m wide and 1m deep.
The scheduling excludes all memorial stones, the Park Café, the park lodge and Ranger’s Station, the Grade II-listed bandstand, Grade II-listed summer house, Grade II-listed classical archway of 1747, the late-C20 beacon, all modern buildings, sheds and pavilions, the green house, the children’s playground equipment, the surfaces of all modern roads and pathways, the tarmacadam surfaces of all car parks, cycle stands, the boundary walls and railings of Castle Park, the concrete footbridge to the keep, all modern garden furniture and fountains, all flagpoles, telephone boxes, flood lights, benches and tables, litter bins, bollards, lamp posts, noticeboards, signs and sign posts, fences and fence posts, gates and gate posts. However the ground beneath all these features is included. The lift, modern doors, modern museum displays and fittings, electric lighting and water pipes in the keep are also excluded. The Grade II-listed Hollytrees Museum (but not the adjoining public toilets) has a basement and is completely excluded. The north wall of the Roman colony forms the south boundary of Castle Park and is a separate scheduling (List entry No. 1003772). Immediately south of Castle Park, the car park and buildings known as: Nos 1-17, The View; One Castle Park; Nos 6, 8, 10-12 Museum Street, Nos 93-96 High Street; and The Castle public house are excluded but the ground beneath them is included.