The Littledean Roman Temple

Severn Panorama Pleasant Stile
“The most splendid location for a Roman monument in Lowland Britain.” Bryn Walters 1985 (Founding Director The Association for Roman Archaeology 1995 retired 2019)



Bryn Walters, Director of the  Association for Roman Archaeology and a leading authority for rural sanctuaries associated with water shrines has been involved with Littledean since its discovery in 1984. Following a searching re assessment of the evidence Bryn is confident the structure was an exceptionally large water shrine or temple built around a central formal pool with a shrine located at the west end within a large apse. This editorial content has now been modified and corrected to reflect this current view which  has been discussed with Professor Tony King, the leading authority for Roman ritual buildings in Britain, Mr Anthony Beeson author and leading specialist on Roman architecture and art and Luigi Thompson an artist renowned for his meticulous painting of Roman mosaics and skills of archaeological draughtmanship.

The Roman site was discovered in 1984. The remains were identified by the late Professor Barri Jones as belonging to a multi period Roman water shrine with Celtic origins and a possible continuity into the Early Christian period. The late Dr Anne Ross who was a renowned authority on the Celts considered the site had probably been a cult shrine venerating Sabrina the goddess of the river Severn. The site overlooking the great horseshoe bend of the Severn was considered by both Jones and Ross as outstanding and in 1984 Bryn Walters described the site as “the most splendid location for a Roman monument in Lowland Britain.”

Professor Barri Jones and Dr. Anne Ross discussing the Roman temple remains in 1985

The horseshoe bend of the river Severn is the largest topographical feature of its type in Europe set in a valley, flanked by the 800 ft high hills of the Cotswolds on the east and the 800 high hills of the Forest of Dean on the west. The  site is located near the 600 foot contour at the top of Dean Hill on the eastern flank of the Forest, directly overlooking the horseshoe bend. Immediately to the front of the eastern entrance to the site the remains of an artificial platform can still be seen, which Dr Ross identified as the probable Public place for ceremonies of propitiation to the river goddess.

Dr Ross and Professor Jones also considered the Severn bore and the filling of the horseshoe bend with the flood tide as central to the Cult of Sabrina. When full the horseshoe bend appears in certain lights as a silver ‘torc’ a sacred symbol of the Celts.

The visible remains probably date from the late 1st or earlier 2nd century. The archaeological evidence establishes that an earlier water shrine was replaced by a peristyle courtyard surrounding a large rectangular central pool. A large entrance court was situated at the east. The pool is the element which defines the building as a nymphaeum.   The outer walls surround an internal area including the entrance court of 39 metres x 19.5 metres with a large  centrally placed apse extending from the western end. Anthony Beeson considers the important surviving feature of the apse is the large foundation base for an ‘aedicula’. This has a direct parallel with Springhead Temple 1 in Kent. The aedicula was a free standing open fronted shrine centrally placed within the apse, probably with a pedimented roof and flanking pilaster columns. Set into the centre of the aedicula  and facing towards the pool would most likely have been a statue of the river nymph or goddess. The  central rectangular pool sited over the original springhead was augmented by a conduit from outlying springs, probably via a large header cistern, elements of which were excavated to the north east in 1986. Evidence produced in 1986 excavations suggests this conduit probably had a pipe encased within it in order to produce pressure for a fountain head or water spout entering the pool. The water left the pool  by way of a stone outfall drain probably via an overflow  basin within the corner of the pool.  Anthony Beeson considers a peristyle roof was carried either on columns around the pool or arcaded arches. Two surviving monolithic building stones, typical for use in a stylobate, along with two column drums, suggest the pool was flanked by columns. This would have formed a typical peristyle with walkways covered by roofs sloping into the centre. from which rainwater would have cascaded into the pool for extra effect during rainfall. Anthony Beeson also considers the space to the west of the pool was the  cella which may have been viewed through a raised triple arch supported on two larger columns, as one can see at Dougga in Northern Tunisia. Either side of the cella  transepts had niches in their west walls for statuary.

The courtyard and formal pool lay directly over the spring fed pool of the earlier shrine. The spring fed pool was notable for the collection of flint material and votive pebbles, dating from the Mesolithic, Neolithic and  Bronze Age periods, as well as fragments of iron arrow heads and  ‘ritually’ bent knives, dating from the late Iron Age.

A great deal more evidence has now materialised in 2018 and 2019 dealing with the prehistoric phases of the site and attention has also been focussed on possible solar and astronomical alignments.  Professor Josh Pollard is confident the evidence indicates a significant Neolithic/Bronze Age presence and Professor Miranda Aldhouse – Green considers the large collection of stone heads which display a particular emphasis on eyes, indicates the site was a water sanctuary with a focus on healing eye complaints during the Iron Age to early Roman period. She also draws attention to the horseshoe bend of the Severn and its symbolic shape of a torc. Bryn Walters believes it is very likely the site remained a healing sanctuary throughout the Roman period and is one of a complex of healing temples in the lower Severn region,. However at Littledean significant conversion in the late 4th century was probably the work of christians. The final period requires more study of both the artefactual evidence and the late Roman modifications. A final late timber built phase, which also awaits further study, may have been a church built in the latter first half of the 5th century AD. It probably had a relatively short life span and after its decay or demolishment the site may have been abandoned.


2020 -2022

Between 2020 and 2022 (March) targetted excavations have produced extensive new evidence, which reached a climax by the end of 2021 when it had been planned to finish excavations on the the actual site of the temple. However opinions varied over the interpretation of a ring of posts with a 15.25m diameter; was it a Bronze Age timber circle or a roundhouse? Given the wide spacings of the posts a circle was considered unlikely. A roundhouse also seemed unlikely because it enclosed the pools and springs and there was no hearth. A roofed late Iron Age temple was a possibility, in which case we would have to prove there was a wattle and daub wall connecting the posts. Good weather in February allowed the circumference of the south west quadrant of the circle to be investigated. Evidence was quickly found for interconnecting ‘panels’ of posts, stake holes and wattle holes, all consistent with a strong framed wattle fenced enclosure. This is considered to be a large circular Iron Age open air temple. It is a profound discovery as we can now also relate the evidence for a rectangular ditched enclosure, with outer bank, which may have dimensions c.24 x 24m. The circular ‘temple’ within the ditched enclosure surrounded three springhead pools in the eastern half and a Bronze Age cemetery, which has been radiocarbon dated within the range 1408 – 1286 BC, in the western half along with a collection of extraordinary pits, most of which had Neolithic/Bronze Age/Iron Age artefacts in them. The reason this evidence is profound is that there has been no explanation for how a small circular open aired wooden shrine around an equally small pool could jump to being replaced by the largest temple structure found in rural Britain. However now it seems likely that the temple mirrors nearly precisely an Iron Age ditched enclosure (temenos) and it is possible that the small water shrine initially stood within the large circular temple of the Iron Age (more research needed). Thus the parameters of scale can now be seen as a balanced continuum of ritual sacred development spanning the Iron Age to the end of the Roman Period. It is too early to be precise and the interpretation may be disproved. However if further discrete excavations can be implemented and we can establish the profile of the bank and ditch, we may also discover further dating material. It would be astonishing to find the internal ditch is bounded by a bank which may have a palisade trench, a feature identified only 20m to the west of the temple in 2016. Further important water features have been identified which now also appear to sit firmly in the Iron Age period. We hope to publish online papers relating to these findings fairly soon.