Excavating the CA archive: Roman villas – part 3

6 mins read

Joe Flatman explores half a century of reports from the past.

A selection of articles mentioned by Joe Flatman in this month’s column below can be accessed for free for one month via Exact Editions, from 2 January. Use the links within the text to jump to the individual articles, or click on the covers below. Print subscribers can add digital access to their account for just £12 a year – this includes everything from the last 50 years, right back to Issue 1! Call our dedicated subscriptions team on 020 8819 5580, quoting DIGI359, to add digital access to your account, or click here for more information.

In the last two columns I explored some of the Roman villas of south and central England; in this third column on such sites I turn my sights north and westwards, into the frontier lands where traditionally villas have not been expected to be found. But, as CA has reported over the years, archaeological fieldwork increasingly tells a different story, identifying villa sites linked to native elites tapping into the new economic opportunities represented by the emerging military presence in such regions. I will also look at how archaeologists have examined the structure of Roman villas though experimental reconstructions, and how more recent finds of villa sites continue to crop up, thanks especially to the hard work and dedication of voluntary groups up and down the country.


CA 220 (July 2008) focused on Holme House in County Durham, where excavations during gravel extraction in 1971-1972 on the site of Roman Piercebridge revealed what was initially thought to be part of the military infrastructure south of Hadrian’s Wall, but which subsequently turned out to be something much more unusual. Issues 40 (September 1973), 55 (March 1976), and 113 (February 1989) all told parts of this story, but CA 220 returned to explore the full record of these landmark excavations.

An article in CA titled 'The rise and fall of Roman Piercebridge'
Excavations at Piercebridge presented a fascinating but dramatically evolving story, with remains initially interpreted as a Roman fort subsequently identified as a villa – as reported in CA 220.

What was revealed was not, as originally thought, a ‘missing’ Roman fort but rather a native aristocratic settlement that began as a roundhouse sitting in the centre of an enclosure, and which then evolved into a villa from the 1st century AD onwards. The villa was subsequently enlarged so that by the mid-2nd century it had become a villa with an apsidal wing and a bath-suite wing. The roundhouse, meanwhile, was rebuilt in stone. Finally, and only in the middle of the 3rd century AD, were the fort-like defences built, and Piercebridge started to look more like a typical military site – the one that the excavators originally thought that they had found.

Expanding on the example of Holme House, in CA 239 (February 2010) David Mason, Durham County Council’s lead archaeologist, argued for a wholesale reassessment of the north-west frontier of the Roman Empire, moving beyond the traditionally bleak vision of a landscape of forts and farmsteads. David argued that the villa at Holme is not an isolated exception. In recent years, other villas have turned up in the lower reaches of the Tees Valley: for example, two have been excavated on the south bank, one at Ingleby Barwick and the other at Dalton-on-Tees. Both appear to have functioned as estate centres, to have had winged-corridor houses for the principal residences, and to date from the mid-2nd century AD. And indirect evidence for high-status country living has come from elsewhere too. As David concluded: ‘the evidence of villa estates, industrial parks, and roadside service-stations are changing our view of Roman County Durham radically. Some sections of the native population seem to have been cashing in on the economic opportunities represented by a strong military presence’.

Turning westwards, CA 240 (March 2010) focused on a site at Abermagwr, near Aberystwyth in mid-Wales. Aerial photography commissioned by the Royal Commission on the Ancient and Historical Monuments of Wales in 2006 revealed the full extent of what was clearly a villa, an extremely rare survival in south-west Wales and before this time a site-type entirely unknown in the Ceredigion area, with mid- and west Wales functioning (so we all thought) as a ‘military zone’ for much of the Roman occupation, distinct from the main block of known Welsh villa sites in lowland Gwent and the Vale of Glamorgan, south Carmarthenshire, and Pembrokeshire.

News article on 'the most north-westerly Roman villa in Wales'
CA 247 returned to the site of a Roman villa located at Abermagwr, near Aberystwyth.

The villa had first been discovered when the Gwynedd Archaeological Trust was commissioned to carry out a magnetometer survey of a site revealed in August 1979, but alas it was then forgotten for over a quarter of a century. Only with the help of Royal Commission aerial photography during the drought of 2006 was the full extent of the large enclosure, with the footings of a stone building within, appreciated. CA 247 (October 2010) returned to the site where excavations had taken place earlier in the year. Trenching established that the building had a winged-corridor plan and was fronted by a cobbled yard. It had walls of local stone built on cobble foundations, glazed windows, and a roof of local slate, cut to form a highly decorative roof of pentagonal shapes, common amongst villas in south-west England and the Isle of Wight.


While CA has traditionally focused on field archaeology, the magazine has always paid wider attention to other topics, including that of experimental archaeology undertaken to inform both researchers and the public alike. Roman archaeology, and especially Roman villas, offer particularly rich inspiration in this light, and it is a theme that CA has turned to repeatedly over the years. Issues 143 (June 1995) and 188 (October 2003) both visited sites in the UK where ‘reconstructed’ buildings of many different periods, including Roman, have been built. CA 143 paid special attention to Upton Romano-British Farm near Poole in Dorset, representing the simplest kind of structure, a wattle-and-daub cottage once inhabited by a family working on the nearby marshes, either making pottery or producing salt. Sadly, the building, set within a wider country park managed by the local authority, was destroyed by arsonists in November 2014.

A happier tale then comes in CA 188, from the much-loved site of Butser Ancient Farm in Hampshire. This is an archaeological site (perhaps ‘experience’ is a better word) beloved of generations of visitors, and originally dating back to work led by the Council for British Archaeology in 1970. In the early 2000s, however, they undertook arguably their most ambitious works to date, reconstructing a wing of the villa found at Sparsholt in Hampshire. This was part of a TV series entitled Rebuilding the Past (narrated by ‘Python’ Terry Jones, no less), which was shown on the Discovery Channel during November 2003 – perfect winter viewing for all of us sofa-bound archaeology addicts. Many lessons were learned about both Roman construction techniques and archaeology-media interactions, as well as more prosaic problems common to us all – with how to get planning permission from a local authority paramount among them.


As any regular reader of Current Archaeology knows, new Roman villas continue to be discovered thanks to development, research, and voluntary archaeological work alike. In 2016-2017, for example, CA broke the news on multiple new villa sites. CA 311 (February 2016) reports from Otford in Kent, where the West Kent Archaeological Society discovered a substantial new late Roman villa of 3rd- or 4th-century date; while CA 323 (February 2017) visited Druce Farm in Dorset, where the East Dorset Antiquarian Society led tremendous work at a multi-phased villa originally identified through metal-detector finds. Meanwhile, CA 332 and 333 (November and December 2017) reported from Boxford in Berkshire, where a community-led project initiated by the Berkshire Archaeology Research Group uncovered the remains of a late-Roman villa containing what has been hailed as ‘the most exciting mosaic discovery made in Britain in the last 50 years’.

Article on Druce Farm villa
In CA 323 we visited a multi-phased villa at Druce Farm in Dorset.

More recently this latter project has also appeared in issue 357, and is nominated for a 2020 CA Award, but all three of these discoveries represent volunteer-led fieldwork of the highest standard, a reminder of what is possible when determined individuals gather together and decide to make a difference. Current Archaeology has had a history of championing grassroots archaeology of this type since the magazine’s inception in 1967, and long may this continue – there’s room for many different people and approaches in the continually evolving story of our nation as told by archaeology, and local community groups lie at the heart of that narrative.

Discover old issues
Read a selection of articles discussed by Joe for free online at www.archaeology.co.uk/archive359. They will be available for one month, from 2 January. Print subscribers can add digital access to the entire back catalogue of CA for just £12 a year – simply call us on 020 8819 5580 and quote ‘DIGI359’.

About the author
Joe Flatman completed a PhD in medieval archaeology at the University of Southampton in 2003, and since then has held positions in universities, and local and most recently – central government. Since March 2019, he has been a Consultancy Manager in the National Trust’s London and South-East Region, leading a team working on Trust sites across Hampshire and the Isle of Wight. You can follow him on Twitter @joeflatman.

Leave a Reply

Your email address will not be published.