Late Iron Age shrine and Roman temple within palisaded (and later walled) precinct, 'Great Chesterford Roman Temple', East of Walden Road (EHER 4978)

The buried remains of a Romano-British temple complex located within an arable field on the sloping hillside on the east side of the River Cam. The central building was discovered in 1847 and excavated under the direction of the Hon. R.C. Neville (also known as Lord Braybrooke), who incorrectly interpreted the remains as part of a Roman villa. He exposed the foundations of a square cella about 7 metres in width, surrounded by an ambulatory. The remains of two elaborate mosaic floors were found within this structure. The re-excavation of the temple in 1978 provided far greater details of the date and evolution of the structure. Evidence was found of a small Late Iron Age ditched enclosure and structure, similar in outline to the temple and probably a precursor to the Romanised building which was erected on the same spot in the period AD 60-90. The temple, with walls of mortared flint and tiled roof, remained in use throughout the later first and early second centuries. The building is thought to have been abandoned in the mid-second century and allowed to decay until major refurbishment took place between AD 280-370. The restored structure remained in use until the late fourth century. Pits and hollows of various sizes were found throughout the excavated areas, but a small excavation in the south western corner of the precinct revealed a distinct concentration of pits containing accumulations of ash, animal bone, oyster shells and pottery. These are thought to have been used for the disposal of waste from religious feasts which appear, from the evidence of animal remains, to have coincided with the culling of lambs in the spring and autumn. Other ritual activities are indicated by the large number of votive offerings including coins, brooches and other items of personal adornment. The most spectacular object related to the religious nature of the site is a silver mask with Celtic-type lentoid eyes and moustache, discovered during the 1978 excavations. Scheduled.

The following is a brief description of the work which has been undertaken at the site, ordered chronologically. 1847. Initial discovery and archaeological excavation. The below ground remains of a masonry structure were encountered by a labourer whilst ploughing and the landowner, Mr Wakefield, contacted Lord Braybrooke to excavate. Braybrooke had the entirety of the temple’s masonry foundations stripped within two days and incorrectly interpreted the 'square-within-square' design of the remains as a two room villa. <1> The site of the temple has been identified by a spread of building material at TL 5143 4360. The site was first excavated by the Hon. R.C. Neville (also known as Lord Braybrooke) in 1847, who uncovered the footings and floors of a Romano-British temple of conventional type, planned as a square shrine with surrounding portico. At the time, Neville incorrectly interpreted the feature as a villa. Hull later reinterpeted it (in source <2>) as a rare example of a temple where the cella walls were slighter than those of the portico. The building was 40ft square and the portico was paved in red brick tesserae. The cella was 18ft square internally and also had a red tesserae mosaic floor. Neville noted the centre of both mosaics had been defaced. The cella walls were plastered and painted with fruit and flowers. Two coins, one illegible, the other of Constantius, were found. Other coins have been presented to Saffron Walden Museum. <2> - <6> May 1950. Field survey. Scatter of tegulae fragments, fragments of imbrices and Roman brick, coarse red tesserae and a few oyster shells centred at TL 5140 4360. The area is quite small with a radius of 10.0 to 15.0 metres and the situation fits Neville's description of the temple site quite well. An old man told me that years ago the field belonged to the Wakefield family who owned the ground on which Neville made his excavations in 1847. A small square cropmark can be seen at this position on RAF air photograph 106G/UK/1635 3405. <7> <8> Square Romano-Celtic country temple, covering an area of about 2116 square feet and built about 150 AD. A circular blank space in the centre of the mosaic floor inside the cells showed that an alter stood there. The site was deliberately destroyed in the 4th century when the mosaics were smashed. <9> January 1977. Field survey. The site at TL 5143 4360 is marked by a dense scatter of Roman tile, brick, tesserae, and pottery which has been spread by the plough over an area of 50.0m east-west by 34.0m north-south. It lies on a gentle south-east facing slope above a small stream. Scatter surveyed at 1:2500. <10> 1978. Archaeological excavation. Excavations by Great Chesterford Archaeology Group (CGAG) were begun in 1978. A report (<11>) lists Braybrooke's more important finds from the earlier excavation: a Samian cup, sherds of a dark vessel, a massive bronze toga pin, a small stylus, a coin of Constantine and another, illegible coin. Other coins have been found about the site and presented by the owner to the Walden Museum. In the GCAG excavations three main periods were discerned: Period 1 (before AD 60, probably Pre-Roman Iron Age) features probably represented 'activity prior or during construction of the Period 2 temple'. Period 1 features were a hard level floor of flints and a post hole (possibly containing a ritual deposit). Provisionally this period was dated to the early 1st century AD. Period 2, divided into three phases, dated AD 60-90, AD 75-105 and AD 105-270 respectively. During this period the Romano-Celtic temple with a 'cella' and an ambulatory was constructed. Pits, a possible hearth, a possible mixing pit for mortar and possible ritual pits were found. There was a porch to the Period 2 temple, two post holes may be connected with it. During a late stage of Period 2 the temple was renovated, but 'must have been approaching a ruinous state' by the start of Period 3. Period 3 (post AD 280, up to about AD 390) saw 'a major renovation programme' carried out. The walls were strengthened by adding a second foundation and the walls were widened. The front porch platform was widened. It seems that this was open leading 'to a probably grander front entrance'. A seat within the ambulatory was found, also evidence for a niche. A wall at right angles to the southern ambulatory wall at the south east corner was also probably of this period. Many certainly votive objects were found and others may have been votive. The most important votive is the silver mask plaque. A second silver plaque was also found as well as bronze feathers; part of a bronze letter 'M'; silver and bronze bindings (possibly from a sceptre); a hoard of 18 bronze 'minimi' and a miniature bronze arrowhead. Other finds include: shale and bronze bracelets, rings, brooches, nail cleaners, tweezers, pins an ear scoop, iron door furniture (?), coins, a ligula, a bronze needle, silver and bronze scraps, an iron collar, a pruning hook, and a bronze ring. Large amounts of pottery were found, samian being rare. Glass was found in small quantities (not window glass). Large amounts of bone and shell also found. Neville found painted plaster; also found by GCAG, both internal and external plaster (the former being the larger part) with various painted panels being represented. The temple walls were of flint, sandstone and rock chalk. Much roof tile was found, also brick quoins. Tesserae were mostly brick cubes and mostly brick cubes and mosaic cubes were also found. Neville suggested that damage to the mosaics had been caused by Christians. Collins thought this unlikely - the damage could have been due to falling stone. 'Evidence for a pre-Roman shrine is also lacking'. Coins showed the building was still in use after AD 370. The building seems to have fallen slowly into decay and was used for stone quarrying in Saxon and later times. In 1978 a cropmark of a road was observed heading away from the site. Excavations were to continue. <11> In a report on the silver mask Liversidge suggests it may have been stuck on wood or metal for display. At the top of the face two locks of hair (?) may indicate a horned god or the mask could be a male Medusa as at the Bath Sulis -Minerva. Perhaps it is easier 'to see here simply the face of the deity who watched over Great Chesterford. <12> A report on the building materials has a plan and section of the site and more details on wall plasters, mortars, building tiles, etc. <13> Photo of silver mask is in the EHER. <14> TL 51434360. The site of the temple was re-excavated in 1978 in advance of further plough damage. The walls and mosaics recorded by Neville had been removed by ploughing and stone robbing, though external features survived. The first porch of wooden columns on a chalk pediment had been replaced by an open podium perhaps leading to a grander entrance. Two circular pits with funnel-like openings on one side are seen as libation pits. Two phases of temple construction were noted. The earlier with a massive collar is dated to AD 60 to 90, whilst a rebuild was constructed after AD 280 continuing in use up to about AD 370. Three further masonry structures have been recorded in the vicinity whilst continuity with the Iron Age is suggested. Finds include bronze regalia and a silver mask with plaque with inscription and deity. Considerable quantities of decorated wall plaster were recovered. <15> <16> See also <17> <18> <19> for more on this excavation. 1983-1984. Archaeological excavation. Continued excavation in 1983 and 1984 uncovered parts of the precinct wall (the precinct was recorded by cropmarks). Features recorded include: a building (function unknown), a gateway, paths, kilns and pits. Painted plaster was found in the building. An aerial photograph indicates the precinct was surrounded by a ditch - not yet confirmed by excavation. <20> <21> <22> 1983: Investigations to the east of the temple precinct revealed the continuation of a 3m-wide path heading westwards to an unknown gate in the precinct wall, and measuring 5.5m wide with semi-circular buttresses. On the outside and sealed by the path was a ditch and kiln. The east and north walls were recorded from crop mark evidence and the south wall were confirmed by excavation. Partial excavation took place on an internal building. <20> 1984: Excavation on the internal building continued revealing it to be 9 x 2.5m with wings. Also more of the precinct wall was excavated. <20> Note: The results of excavations in the temple precinct in 1983-1988 are recorded in EHER 7311 and EHER 13962. The next section provides a general description of the site gathered from several sources. The temple is 1km to the east of the Roman town, on the southern slope of a slight hill next to the River Slade, with extensive views to the west and south. The location marks the junction of the gravels and the chalk ridge to the north. The site has been archaeologically investigated three times: Braybrooke uncovered the temple walls and recorded the mosaic floor in 1847, the upper levels of the walls and the mosaic subsequently being demolished by the landowner, after which the site reverted to arable agriculture; the temple was re-excavated in 1978 by the GCAG under the direction of A.E. Collins; and between 1983 and 1988 the GCAG, under the direction of T.E. Miller, excavated a series of trenches both within and outside the temple precinct. <6> Two aerial photographs in the EHER of temple under excavation. <21> Roman pottery and coins from about the site were presented to Saffron Walden Museum. <22> <23> (TL 51404361) ROMAN BUILDING (R) (Site of) (NAT). <24> See also <25> - <31>. See aerial photograph <32>. The precinct enclosure is mapped on the NMP cropmark plot. <33> A Google Earth aerial image from 2007 shows the enclosure ditch. <34> The archaeological evidence was re-examined by Maria Medlycott for the EAA Report No. 137 (<1>). The remaining text provides a broad summary, by period, of the temple precinct as a whole and provides references to other monument records which provide detail on features within the precinct. The phased summary below has been informed by the discussion section in Chapter 10 of the EAA report, which pulls together the excavated evidence from 1847, 1978 and the 1980s for the temple and its precinct. The 'phases' are determined by the phases of temple construction and alteration discussed in this record above, and initially set by the 1978 excavations. Phase 1 - Late Iron Age-early 1st century AD: A three-sided structure measuring 6 x 5m is thought to be the earliest phase of the temple or 'shrine'. Surrounding it was a wide, low bank, faced with flints on its inner edge and with a post-hole setting at its terminals and on each of its internal corners. To the north of the structure was a pit or post hole containing a votive deposit of a platter and bowl. There are no features within the temple precinct that can be dated to Phase 1. Phase 2 - Late 1st–early 2nd century AD: In the late 1st or early 2nd century the masonry temple was built in the classic Romano-British style of a square-within-a-square. It measured 17.6 square metres externally, with an 8.6 square metre cella. The doorway faced due east. Immediately outside the western wall of the temple, facing towards the town, was a substantial base for an altar or statue. A gravel path was laid in front of the temple. The precinct, or temenos, was laid out by a large ditch and in the form of a parallelogram (approximately 96 x 94m), with a gateway facing east. A series of pits were dug in the south-west corner, two of which were exceptionally deep. Most, but not all, of the pits contained large quantities of animal bones, mostly sheep. Outside the temple to the south-east a single large and deep pit was excavated and contained pottery. It appears that Nauheim derivatives and Colchester B/BB brooches may have been sold at the temple specifically for use as votives, and they may have been made locally, possibly in the temple’s own workshops. The Nauheim derivatives, in particular, may have been specifically made for deposition as votives, as some of the examples recovered were non-functional, with unfinished catch-plates. Phase 3 - Mid-Late 2nd century AD: Soon after the construction of the temple structural modifications were made, including a porch added to the temple itself. The paths were regravelled and extended to encircle the temple, and a path was laid linking the precinct entrance with the new porch. Immediately to the south of the temple were several small pits, one of which had been used as a hearth and contained a 'placed deposit' of a bowl, butt beaker and burnt and cremated bone. A palisade trench defining the precinct area was dug within and parallel to the precinct ditch, see EHER 7311. Further pits were dug in the south-west corner, including one very large example, which contained animal bone deposits. Outside the temple to the south-east, in Site B (EHER 13962), a number of structures were built. The use of symbolic votives became more overt during the 2nd century, with the deposition of sheet leaves and plaques. Phase 4 - Early-Mid 3rd century AD: Further repairs and renovations took place: the paths were regravelled and the plinth behind the temple was removed. A number of votive deposits, in the form of silver plaques, a silver pin, coiled silver wire and a miniature arrow-head, were placed in or on the paths. Two more pits were dug in the south-west corner of the precinct. Activity also continued outside the precinct on Site B (EHER 13962). Phase 5a and 5b - Late 3rd-early 4th century AD: The temple fell into a state of disrepair, with the external plaster and roof tiles falling off. A programme of repair and renovation was therefore undertaken (Phase 5a), as part of which the roof battens were removed and burnt and, presumably, a new roof was built. The existing walls were thickened and strengthened. The porch was dismantled and the resulting cleared flat surface used as a mosaicist’s workshop. It is presumed, therefore, that the mosaic excavated by Braybrooke dates to this period. A mosaic and tesserae paving was placed within the cella. The interior of the cella was plastered and painted with what Braybrooke recorded as patterns representing fruit and flowers; the ambulatory was also plastered and painted with a red border. On the southern side of the temple, a new structure or lean-to was built against the existing temple wall (Phase 5b). A new building was erected in the precinct, between the precinct gateway and the temple, resembling in plan half a small square-in-square temple; it is not clear whether it does represent half a temple, with the remainder having been destroyed, or whether it is an unusual plan-form, presumably for a shrine. The structure was plastered and painted internally with panels of pink faux marble. A hearth or kiln was located within the structure, although it is possible that it actually preceded the construction of the building. The palisade which delimited the precinct was replaced by a stone wall built of mortared flint. Outside the precinct, on Site B (EHER 13962), occupation continued. Phase 6 - 4th century AD: There is little evidence for any activity on the site in the later 4th century, apart from in Site B, see EHER 13962. Phase 7 (early Saxon period): The only evidence was the discovery of a spearhead leaning against the inner south-eastern corner of the precinct wall. It is not known whether this was a deliberately placed offering to the site or a simple loss. The mosaics were destroyed by someone smashing in the centre of the patterns, but at what date this was done is unknown; it is possible that this took place as part of a Christian-inspired iconoclasm. <1> Original HER entry merged with NRHE data transfer. NRHE data transfer: The buried remains of a Romano-British temple complex located within an arable field on the sloping hillside on the east side of the River Cam. The central building was discovered in 1847 and excavated under the direction of the Hon RC Neville, who incorrectly interpreted the remains as part of a Roman villa. He exposed the foundations of a square cella about 7 metres in width, surrounded by a ambulatory. The remains of two elaborate mosaic floors were found within this structure. The re-excavation of the temple in 1978 provided far greater details of the date and evolution of the structure. Evidence was found of a small Late Iron Age ditched enclosure and structure, similar in outline to the temple and probably a precursor to the Romanised building which was erected on the same spot in the period AD 60-90. The temple, with walls of mortared flint and tiled roof, remained in use throughout the later first and early second centuries. The building is thought to have been abandoned in the mid-second century and allowed to decay until major refurbishment took place between AD 280-370. The restored structure remained in use until the late fourth century. Pits and hollows of various sizes were found throughout the excavated areas, but a small excavation in the south western corner of the precinct revealed a distinct concentration of pits containing accumulations of ash, animal bone, oyster shells and pottery. These are thought to have been used for the disposal of waste from religious feasts which appear, from the evidence of animal remains, to have coincided with the culling of lambs in the spring and autumn. Other ritual activities are indicated by the large number of votive offerings including coins, brooches and other items of personal adornment. The most spectacular object related to the religious nature of the site is a silver mask with Celtic-type lentoid eyes and moustache, discovered during the 1978 excavations. Scheduled.