A scorched summer and a chance observation at Badbury Rings helped reveal a lost Romano-British temple that had lain hidden in plain sight for centuries, says Martin Papworth

The dry summers of 1975 and particularly 1976 were good times for archaeological discovery. The grass died over buried walls revealing the outlines of long-forgotten buildings.

Just outside the outer rampart of Badbury Rings lies a roughly circular area of hollows, humps and bumps. Nobody was quite sure what they could be. One suggestion was that these earthworks marked the site of a ‘fancy’ type of Bronze Age burial mound known as a disc barrow. A spread platform in the middle surrounded by a low bank measuring about 60m across.

That was until 1975, when Bryan Pybus visited Badbury and noticed that the lack of rain had parched the outline of a clear rectangular building onto the central mound and revealed the surrounding bank to consist of a series of straight lines forming a rough octagon with an entrance facing the hillfort. Bryan thought that this site might be a Romano-British temple as he recovered small fragments of building stone and Roman pottery during his visit.

He sent his sketch map and notes to the Ordnance Survey archaeological records department and there they stayed until 1986 when I found them during the archaeological survey of the National Trust’s Kingston Lacy Estate.

I went to Badbury to see what I could find to support Bryan’s theory, and as I walked across the grass I found objects in mole hills which seemed to confirm that remains of a Roman building might lie buried here.

There were fragments of Purbeck limestone, brick, animal bones, oysters and Roman pottery. This could be a very exciting discovery, so I carried out further research and uncovered, in the Dorset Museum archive, records of small trenches dug near this spot in 1900 and again in 1952. From these investigations many Iron Age and Roman objects were recovered, including coins covering a time span stretching from the pre-Roman Durotriges tribe to the end of the Roman period.

Roman coin Uncovering a newly discovered Roman coin (Image: Martin Papworth)

This led to fieldwork across the site, mapping and geophysical survey, which confirmed the location of buried walls. Soon we had a good plan of the enclosure and buildings.

However, the many Roman objects turned up in the mole hills attracted metal detectorists to the scheduled monument. As we surveyed the site we saw the signs that items were being taken from the monument. There were small spade holes, the turf cut and placed to one side. There was no way of knowing what these finds might have been and separating an artefact from its location without record is a loss of valuable potential information.

The solution was to move the moles from the site and to cover the monument with chain-link fencing. The fencing would act as armour plating to prevent digging and also affect metal detector signal. The fencing would be pegged down and in a few years would grow over and its presence would not be visible.

Before this conservation covering was to be installed, Historic England gave the National Trust permission to carry out an archaeological excavation to discover whether this place really was the site of a temple.

In September 2000, using the evidence from our geophysical survey, we dug a long trench from the central platform to the northern edge of the enclosing bank. We decided to camp on site and my dad lent me our old blue family tent which formed the hub for a small village of tents. Long before drones, the Kingston Lacy building department built us a scaffolded photographic tower to enable overhead shots of the site. The fortnight of excavation began in torrential rain and ended with a storm that almost felled the tower, but I have good memories of walking Badbury in twilight, the white ghost of a barn owl gliding towards us one evening.

We cut the deep turf, and below the casual losses of recent visitors, the ring-pulls and silver paper, we found large lead bullets. These were explained by the 1880s Ordnance Survey map which showed a rifle range located in the fields west of our excavation.

Badbury temple trenchView of the Badbury temple trench from the scaffold tower. (Image: Martin Papworth)

Then we started finding chunks of rubble mixed with Roman pottery and, near the centre of the site, fragments of painted plaster and clay tile. It became clear that the buildings had been completely demolished and even the foundation trenches had been cleared with just a patch of flint and mortar left in place. Beside each robber trench was a little heap of debris where the best stone had been selected and the bits and pieces chucked to one site.

Once the building rubble was removed we could see that there had been a square building here measuring about 15m wide and this had been surrounded by a corridor. Surrounding this was a gravel yard and against the northern stone enclosure wall was another small building. This evidence could be compared with known Romano-Celtic temples which typically consist of a square tower-like shrine or cella surrounded by a roofed walkway known as an ambulatory. The building on the north side of the temenos or temple enclosure could be for storage or accommodation for a priest.

There were a few cubes of stone and tile found but no mosaics. The floor surfaces were of packed gravel or rammed chalk and had been left in place by the post-Roman temple robbers. Only the central area had been dug into and perhaps there had been something worth taking there, maybe a stone altar or possibly a lead water tank.

Whatever deity had been worshipped here had involved the donation of money and votive offerings. Our finds included many tiny blue and green glass beads and fragments of little jars. There were also interesting objects including a bronze ring which included a key, a pewter horse harness terret and an iron bridle bit.

An excavated Roman bridle bit An excavated Roman bridle bit (Image: Martin Papworth)

Coins were the most numerous finds, small change offerings concentrated within the central building. The upper chalk floor contained later 4th- and one or two 5th-century coins, deposited above it. When this was removed, the earlier gravel floor beneath revealed larger 2nd-century coins. Our assemblage spans emperors from Claudius to Honorius, along with a single pre-Roman Iron Age stater.

We had confirmed that this was a sacred site which lay beside the hillfort near the Roman road to Dorchester. Rituals were performed here for almost 500 years, though we found no statues or other votive objects which would give us an idea of which deities were venerated at this temple. Typically, the original native god was linked to a deity in the Roman pantheon and both were venerated. For example, we know that the gods worshipped at the temple complex built beside the hot springs at Bath included the Celtic god Sulis and the Roman god Minerva.

We invited Mr Pybus back to Badbury to see the excavation while we were there and he was pleased that his interpretation of the site, back in that dry summer, had now been confirmed..


Further Reading

Papworth M. 2014. ‘The Romano-Celtic Temple at Badbury Rings, Dorset’, Proceedings of the Dorset Natural History and Archaeological Society 135, 242-271.

Copies of the Dorset Proceedings are now being uploaded to the Dorset Museum website. Search on Dorset Proceedings dorsetmuseum.org/proceedings