This short series focusing on the Rossett Roman Villa began yesterday with Part 1 – What is a Roman Villa? , which was an overview of Roman villas in general, looking at how they are defined, their key features, what is known about who lived in them, how they changed over time and how they are dated.
Today’s post, part 2, looks at the background to the decision to start excavating at the site, information assembled from press releases, the villa project’s Twitter releases (impressive!) and the information imparted by Stephen Grenter, Heritage Services Manager at the Wrexham Museum on the Open Day. Part 3 will describe the excellent Archaeological Open Day that took us through what was happening in the three big trenches opened in the last two weeks, as well as plans for the future. Again, I just want to say a huge thank-you for such a great tour of the site and the sheer amount of knowledge imparted in such a relaxed but professional way. I have included some of the information communicated on the day in this post, but any errors are my own.
The strategically important legionary fortress at Chester means that a Roman presence in the surrounding area was almost inevitable, and it has been known for a long time that there was a civilian settlement south of Chester at Heronbridge, Roman industrial activity in Holt (a brickworks and tileworks) to the east and at Halkyn Mountain to the west. Also in the west, Ffrith has produced Roman remains, but it lies under the village so not much is known about it. Up until now, however, nothing concrete was known about Roman activity in the Rossett-Burton area.
The Rossett site is located to the west of Burton and is the first villa known from northeast Wales, making it of particular importance. Prior to any major discoveries, the existence of a Roman presence of some description in the area had already been inferred by archaeologists who had found Roman objects in local ploughed fields. Ploughed fields are excellent for field walking, as the action of ploughing draws artefacts from lower down up to the surface of the field, and they are often clearly visible against the dark soil. When the field is recently ploughed, there are no distractions like crop stubble or weeds.
Confirmation that from the 1st Century onwards Romans had at least passed through the Rossett-Burton area came with a discovery made by a responsible metal detectorist who reported an important find: an inscribed lead pig that turned out to date to the 1st century AD, the century in which the Romans first arrived. All of a sudden, the Burton area was in the archaeological spotlight. A survey and excavation of the ingot site followed, funded partly with a grant from the Roman Research Trust and carried out by archaeologists from Wrexham Museum, the University of Chester and Archaeological Survey West. There was sufficient time and funding remaining after the ingot investigations had been completed for further geophysical survey work to be carried out in a nearby field and this revealed a beautifully delineated buried structure with the typical layout of a Roman villa, staggeringly clear on the survey image shown below. Additional structures were evident, but not so easy to interpret, and some of those too are now under excavation.
The Rossett Ingot
The first indication of a site near Rossett was the discovery of a lead ingot or pig. A pig is a roughly rectangular bar of mined metal that is shaped to be convenient for transportation to a location where it can be processed. Its discovery was reported on the Archaeology Chester (University of Chester) blog:
Our story begins in September 2019 when a lead pig (ingot) marked with the name of Trebellius Maximus, the Governor of Roman Britain from AD 63 to 69 was found near Rossett, Wrexham County Borough, Wales. A responsible, skilled, and knowledgeable local metal detectorist found an impressive metal signature while out detecting. He immediately contacted the local Finds Liaison Officer based at Wrexham Museum and the object was subsequently excavated with the help of staff at Wrexham Museum and the Clwyd-Powys Archaeological Trust.
The find generated a lot of interest in academic circles and was widely reported in the media, because this is the first inscription known in Britain that mentions Marcus Trebellius Maximus, who was the governor of the province of Britannia between AD 63 and 69 during the reign of the Emperor Nero, one of Britain’s first governors after the AD 43 invasion. An administrator rather than a soldier, he made no push to gain more territory, placing emphasis on consolidation and economic growth. He was unable to secure the respect of the military and in AD 69, after the death of Nero and a period of instability in Rome, a mutiny in Britain forced him to flee. He was replaced by Marcus Vettius Bolanus who had the twin assets of being both a Roman Senator and a soldier.
Although as a legionary fortress Chester was an important Roman centre, with roads radiating out of it towards other Roman sites in all directions, northeast Wales itself has not provided much data to support the idea of a significant Roman interest in the area, so the find raised two important questions that led to the decision to carry out further surveys and excavations in the area. The first question raised by the pig itself was about the extent to which the Romans were exploiting local mineral resources from early on during their occupation of Britain. The second concerned how a major new Roman find might shed light on the Roman occupation of northeast Wales, data for which is extremely thin on the ground to the northwest of Wrexham and southwest of Chester.
The initial fieldwork, phase 1 of the project, took part during partial lockdown with atrocious weather conditions during September and October 2020, with financial support from the Roman Research Trust, the University of Chester and Wrexham Museum. Although the surveys suggested some promising features, excavation by a small 6-person team, battling with rapidly flooding trenches only revealed remains from mainly much later periods. The absence of Roman period finds during the excavations was, however, informative:
The absence of Roman archaeology and confirmation of alluvial deposits highlighted the likely watery or marsh-like setting that existed during the Roman and later periods. In turn this tells us that the ingot is reflective (perhaps) of a stray loss since no evidence of deliberate deposition or lead processing could be found nearby. [Pudnesy 2021]
The conclusion is that the lead was mined elsewhere and was lost in transit on its way to its intended destination. The analysis of the lead at Liverpool University, which hopes to narrow down a source, is still ongoing but initial work suggests that it may have been mined from elsewhere in northeast Wales, perhaps at somewhere like Ffrith, where Roman remains have been found, including indications of lead mining, or Minera:
That the Romans mined lead at Minera has long been inferred; the mineral veins would have been easily discovered at outcrop, a Roman road passes close by, and residues of lead smelting have been recorded in a Roman context only three miles distant. Proof
remains elusive though ancient working is inferred by the discovery of a stone mortar. [Peter Appleton]
No further archaeological remains were discovered at the site during excavation.
Other pigs have been found in the Chester area. David Mason shows one in his book Roman Chester, excavated with the remains of a timber jetty at the Roodee (Chester racecourse on the side of the river Dee) in 1886 dating to AD74. The text is abbreviated but reads “[Cast] while the Emperor Vespasian Augustus was consul for the fifth time and Titus, acclaimed Imperator, consul for the third time. On the side is another inscription that reads “Deceangl” meaning that it was mined on Deceangli territory. The Decaengli territory of northeast Wales ran along the borders of the Cornovii territory that occupied what is now West Cheshire, and probably extended up the Wirral.
Britain’s mineral resources were amongs the properties of Britain that was extremely attractive to Rome, and the territory of the Deceangli had numerous stone and metal resources including lead. Lead was used in building projects, but some of it was also a source of silver when subjected to a process called cupellation. Lead mines at Prestatyn were established in c.75AD. Others have been found at Meliden, Pentre-Oakenholt, Halkyn and Ffrith. Clwyd and Powys Archaeological Trust describes an ore vein that runs east to west at Halkyn. A Roman ‘pig’ or ingot of lead was found in 1950 inscribed with the letters C NIPI ASCANI, the abbreviated name of a private lead producer, C. Nipius Ascanius, the lead thought to have been mined and smelted on Halkyn Mountain. Excavations in the Pentre Oakenholt area of Flint have provided evidence of lead smelting, presumably from ores from Halkyn Mountain. Roman domestic buildings at Pentre Farm, Flint may have been the home of a mine supervisor.
There was sufficient funding from the Roman Research Trust grant left over for additional geophysical survey. Stephen Grenter had visited the field in which the villa was found at an earlier date because pottery sherds and other small finds had been made there, and found additional objects that suggested that it would be worth carrying out additional fieldwalking and geophysical survey, so this was carried out.
The field walking recovered a total of 181 artefacts from the ploughsoil. A large proportion of artefacts were ceramic, including brick and tile (CBM). A total of 76 sherds of pottery, 23 fragments of worked stone, 4 metal objects, 5 fragments of glass and one fragment of animal bone were also retrieved. Together with fragments of painted plaster and opus signinum, the assemblage reflected the likely presence of a Roman building, but potentially of higher status than we’d initially suspected. [Pudney 2021]
Geophysical survey (magnetometry) followed. Geophysical survey results can be remarkably difficult to interpret, but the amazing scan of the villa’s foundations, was phenomenally clear, showing the perfect layout of a wing and corridor villa with rooms behind. Other features suggested by the geophysical survey were not nearly as clear, and some of those are now under excavation.
The Rossett Villa
To the west of Burton Green, the villa is described as Rossett Villa. Clear evidence of Roman occupation in the immediate area had been indicated by objects produced in the process of agricultural ploughing as well as metal detecting. These items included pieces of samian ware (terra sigillata, a Roman luxury ceramic), box tiles, fragments of mortaria (food preparation mortars) and quern stones. Other Roman objects found in the general vicinity had been registered with the Portable Antiquities Scheme including Roman brooches and coins. The presence of a villa had not been suspected as they are extremely rare in the northeast Wales/Cheshire areas.
The Rossett villa is unusual in that its location is outside the main distribution area of villa sites. The densest concentration of known villas is in the south of England. Rossett is only one of two villas known in the area that potentially fell under the influence of the Chester legionary fortress, the other being located at Eaton-by Tarporley. The discovery of signs of a hypocaust at Crewe-by-Farndon have led to suggestions that there may have been another a third one in that area (mentioned in the the Farndon Archaeological Assessment). The nearest villa in Wales is at a substantial distance from Rossett, in a remote part of Ceredigion, near to Trawsgoed Roman Fort.
Some of the sites in the Chester area are connected to one another by the Roman road network, but it is not yet fully understood how northeast Wales was reached from Chester and how it was connected to west Wales, including the sites at Ffrith and Halkyn.
The Rossett villa is located just off the proposed route of a potential Roman road from Chester. It has been suggested that the road may have run south from the fortress, across the bridge shown in Baum’s reconstruction of Chester above, before turning southwest and passing through Ffrith, where there is plenty of evidence both for Roman settlement remains and a stretch of Roman road, before proceeding via Bala to the fortlet at Brithdir to the south of Dolgellau. This presumably also connected with the Cefn Caer fortlet at Pennal (about which I posted on another blog here), which guarded a crossing over the river Dyfi, connecting north and south Wales. It is hoped that future LiDAR research will clarify the location of the road. It worked a treat with clarifying Roman road 6A (also known as Watling Street West) that runs south from Chester via Aldford and Malpas to Whitchurch and beyond to Wroxeter (about which I have posted here).
The nearest villa to Rossett, as the crow flies, was actually at Eaton by Tarporley in Cheshire, excavated 1980-81 and again in 1982. As far as I know, it remains Cheshire’s only known Roman villa, as reported by Morris in 1982 and 1983, and summarized on the Heritage Gateway website. The summary is copied here because it provides a useful illustration of the often multi-period character of villas:
During the laying of the Lake Vyrnwy-Liverpool water main in 1886, Roman tiles, mortar and a coin of Marcus Aurelius were found on the western fringe of Eaton-by-Tarporley (a). A field-walking programme in 1980 to investigate the context of these finds, led to the discovery at SJ 57176341 of a Roman winged-corridor villa, the first villa to be identified in Cheshire. Excavations were conducted on the site from 1980-82. These revealed 4 Roman phases.
Phase 1. Only two post-holes were found relating to the primary occupation of the site, perhaps beginning c. AD 150. The building was probably short-lived, quickly succeeded by the phase 2 construction on a different alignment.
Phase 2. A timber building was erected delimited to W and N by ditches perhaps serving to convey water to the site from a nearby spring rather than for drainage. Again the building seems to have been short-lived, this time destroyed by fire.
Phase 3. About the last quarter of the 2nd century, the first stone-built villa was constructed, of winged-corridor plan and of a single storey only. The S wing formed a baths suite. All rooms in the main range were decorated with painted wall-plaster and had floors of opus signium or mortared pebbles. One room here was heated, plus two in the N wing.
Phase 4. c.AD 350 the villa underwent thorough reconstruction. The colonnade was demolished and the living space extended out to this line. Thickening of the walls indicates a second storey was added at this time. No evidence survived for the destruction/abandonment of the villa due to Medieval stone-robbing and PM ploughing.
Medieval. Large numbers of pottery wasters were recovered from the villa, and excavation SW of it located a complex of 14th century pottery kilns. At some later date but still within this period, a building of unknown function was erected out of re-used Roman materials over the SW corner of the villa.
The villa included a bath suite including a calidarium and tepidarium sitting over hypocausts (raised floors on short pillars, the spaces created heated with fires), and a frigidarium for cooling off.
In Wales itself, the only other villa site known north of south Wales is in Ceredigion, less than a mile from Trawsgoed Roman Fort, and in a very remote area. Abermagwr villa was first identified from aerial photographs taken during the drought of 2006, was subjected to geophysical survey in 2009 and was partially excavated in 2010 by by Jeffrey Davies and Toby Driver. Described by its excavators as “a comparatively modest late third- to early fourth-century AD house,” was established around 230AD, which is interestingly around a century after the Trawsgoed fort was abandoned, and it is suggested that building material from the fort’s bathhouse was used to build the villa. It had a very fine slate roof, and finds included pieces of a remarkable glass bowl that was made in Germany’s Rhineland. The villa burned down in c.330AD, and was abandoned.
This is a very poor showing for villas in northeast, northwest and mid Wales, and for Cheshire as well. The scarcity of villas in this area seems to require an explanation, particularly as Chester was such an important fort, there was a civil Roman settlement at Heronbridge just to the south of Chester near Ecclestone, a tile and pottery manufacturing base was located immediately to the north of Holt and there was another pottery production centre at Plas Coch on the outskirts of Wrexham. This was an area of prime agricultural land that one would have thought would be ideal for the establishment of one or more potentially profitable estates. There are two primary reasons why sites do not occur on distribution maps. The first is because they were simply not built in certain areas, and the second is that they have not yet been found. There are more reasons too, such as sites that have been completely destroyed, or those that whose building materials were robbed for the building of other buildings, but a complete absence of evidence in an area tends to fall into one or other of the first two categories. In practical terms, this means that a gap on a distribution map is a question mark, not a sign that nothing was build there. This is perfectly demonstrated by the Eaton-by-Tarporley, Abermagwr and Rossett villas, all of which turned up in places that were empty patches on villa distribution maps. With more grants for future research, a lot more field work and a bit of luck thrown in, the Rossett and Burton areas may reveal more previously unrecognized archaeology, including that from both previous and later periods. Indeed, Toby Driver has recorded cropmarks at Rossett similar to those at Roman villas in other parts of Wales (noted on the Coflein website). However, with the discovery of the Rossett villa, it seems likely that others will now turn up.
Even before I went to the Open Day, the team had made it clear in their reports that the Rossett and Burton Green finds are exciting hints of a greater Roman presence to the southwest of Chester than had previously been suspected. Both the original discoveries and the work that has since taken place will hopefully form a platform for the launch of future survey and excavation work that will help to clarify how northeast Wales fits into the bigger Roman and Romano-British picture. The team is hoping to reconvene next year for a six week dig, assuming that funding is forthcoming.
The last words today go to Dr Caroline Pudney:
Both the lead pig and the villa whisper to us of great potential. The prospect that this villa complex does not exist in isolation is very real. There are not many Roman villas known across north Wales. North east Wales specifically, was until now, yet to reveal one buried beneath its soils. Who knows how many more lurk beneath the surface? There are also a surprisingly low number known further west and south into Cheshire and Shropshire. This is strange considering the presence of a whacking great Roman fortress (Deva Victrix) and the civitas capital at Viriconium (Wroxeter). One would surely expect a richer character of rural settlement in this area than is presented in the known archaeological record to date. [Pudney 2021]
Follow the Roman Villa excavations, their post-excavation findings and their news about future work on Twitter using the hashtag #rossettvilla.
The main source of information about the Rossett discoveries is Dr Caroline Pudney’s report on the Archaeology Chester (University of Chester) website, which has been quoted extensively above: The highs and lows of archaeology: In the footsteps of Trebellius Maximus. By Dr Caroline Pudney, 16th Apr 2021
Additional background information as well as some notes about the villas in Cheshire and Ceredigion have been sourced as follows:
Books and Papers:
Clark, J. 2003. Cheshire Historic Towns Survey. Farndon. Archaeological Assessment. Environmental Planning, Cheshire County Council
Clark, J. 2003. Cheshire Historic Towns Survey. Tarporley. Archaeological Assessment. Environmental Planning, Cheshire County Council
Greene, K. 1986. The Archaeology of the Roman Economy. Batsford
Johnston, D.E. 1994. Roman Villas. Shire Archaeology
Mason, D.J.P. 2007 (2nd edition). Roman Chester. City of the Eagles. Tempus
Morris, M.G. 1982. Eaton By Tarporley, SJ57176341. Cheshire Archaeological Bulletin 8, p.49-52
Morris, M.G. 1983. Eaton By Tarporley, Roman Villa. Cheshire Archaeological Bulletin 9, p.67-73
North, F.J. 1962. Mining for Metals in Wales. National Museum of Wales
Pudney, C. 2021. The highs and lows of archaeology: In the footsteps of Trebellius Maximus. Archaeology Chester, 16th Apr 2021
Rowe, J.E. 2015. Roman Villas of Wales. M.A. Thesis, University of Regina, Saskatchewan
Cefn Caer, the Roman auxiliary fort at Pennal. By Andie Byrnes. 3rd February 2019
Based in Churton
A touch of Rome just east of Churton #1 – Background to the Roman Road. By Andie Byrnes. 3rd April 2021
The Abermagwr Roman Villa, Cerdigion
Lane Farm Cropmarks, Rossett
CPAT Regional Sites & Monuments Record
PRN 100020 – Ffrith Roman site (multiple site). Scheduled Ancient Monument FL164(FLT)
PRN 86912 – Ffrith, Roman Road
Holywell Common and Halkyn Mountain
Historic England Research Records – Monument Number 71430 (Eaton by Tarporley villa)
Royal Commission on the Ancient and Historic Monuments of Wales
The Roman villa that made history: Abermagwr Villa, Ceredigion. By Toby Driver, 24th July 2020
U3A Ruthin and District
Mineralisation and Mining at Minera, North Wales. By Peter Appleton. Date unknown.
Rossett Roman villa dig underway in ‘history-changing project. 6th September 2021
Wrexham County Borough Museum and Archives
The Rossett Lead Pig