Druce Farm Roman villa lies 6.5km to the north-east of Dorchester in the parish of Puddletown, Dorset. The site could be described as a classic villa location, enjoying maximum sunshine on south-facing chalk downland that slopes to the River Piddle. There is a chalk spring close by, providing water of such purity that it was channelled via a leet in the early 19th century to feed watercress beds in the valley below.
Mrs Ann Ridout, the owner of the modern farm on the site, gave permission to the Stour Valley Search and Recovery Club to metal-detect on the farm in the early 2000s, but the group did not need sophisticated kit to locate the villa: large fragments of Roman roof tile – both limestone and ceramic – were already visible on the soil surface. A geophysical survey revealed the unambiguous evidence of buildings set around a courtyard.
Mrs Ridout and her family were eager to find out more. She approached the East Dorset Antiquarian Society (EDAS), and they eagerly accepted the opportunity to excavate the site under the direction of Lilian Ladle, a self-taught archaeologist, who had previously directed and published multi-period sites at Bestwall Quarry, near Wareham, and Football Field, Worth Matravers, both in Dorset.
The site was excavated between 2012 and 2018 by a core team of 20 regular volunteers, who worked Monday to Friday, and a further 90 people who spent time on the project. Over 250 students from 17 schools also took part, including all the pupils from the local Piddle Valley Primary School and members of the Dorchester branch of the Young Archaeologists’ Club.
This was a volunteer-led project with few resources, so the donations given by the large number of visitors who attended the regular open days made a vital contribution towards the post-excavation costs. Many local people gave their time and loaned essential equipment: JCBs to strip and backfill the site, a microlight and cherry-picker for photography, laser-scanning kit and total station equipment for topographical survey, digital cameras and metal-detectors to check spoil heaps and surrounding fields. In sum, the Druce Farm excavation was that rare thing: a genuine community excavation that gave many local people the opportunity to experience hands-on archaeology and get to know more about the hidden history of their own patch of the world.
The recently published report on the project (see ‘Further reading’ on p.30) sets out the questions that Lilian Ladle and her team aimed to answer. They wanted to determine the chronology of the villa and compare that with other examples, local and further afield, to assess changes in the settlement’s status and function; to examine cultural associations, human remains, and non-domestic activity; and to explore the evidence for environmental change, agricultural practice, food-production, and craft and trade activity.
Enigmatic enclosures and uncertain origins
The land on which the villa would eventually be built was already under cultivation in the early Neolithic, c.3941-3648 BC, as represented by a timber structure and pits containing carinated bowl sherds and flint tools, ranging from scrapers to axe-heads. It was also part of the extensive Dole Hill Bronze Age field system that the English Royal Commission surveyed in 1970, when it was found to extend for 809 hectares to the north-east of the villa site, and which probably dates from the 16th and 15th centuries BC by analogy with similar dated field systems elsewhere.
Two large rectangular enclosures were then created on the south-facing slope in the mid-1st century AD, on the same alignment as the field system, enclosing 1.38 hectares. Forming an outer and an inner enclosure, the ditches had steep sloping sides and a narrow flat or gently rounded base, and this profile was true of every section excavated (75 sections amounting to c.12% of the total). This consistency, combined with the sharpness of the corner angles, speaks of a carefully planned single system laid out over a short period of time. The ditches did, though, vary in width, from 1.3m to 2.4m, and even greater in depth, from as little as 0.6m to 1.88m below present ground-level.
The ditches show up with remarkable clarity in geophysical surveys, and all were found to be filled with building debris, pottery, charcoal, slag, butchery waste, and animal bone, including sheep/goat, pig and cattle, red deer, dogs, and horse. One artefact-rich section of the ditch contained pottery ranging in date from c.AD 120 through to c.450-650. All this suggests that the ditches were not scoured out for long, and they were used for the disposal of domestic rubbish continually from the mid-1st century through to the early medieval period.
Lilian Ladle admits to being puzzled by these enclosures, which appear to be without parallel for Roman villas in Britain. Boundaries were a powerful landscape feature signifying possession, and a number of early villa sites (such as Ditches in Gloucestershire; see CA 217) are set within late Iron Age enclosures, but in association with a pre-existing settlement, whereas this was a greenfield site when the ditches were dug. Similar enclosures have been found to be military in character; here, the refuse is domestic.
They might have been laid out with an eye to the future use of the land – marking out a development plot, as it were. And associated with the enclosure ditches is a single-storey structure built with coursed flint walls, probably with a tiled or thatched roof: a rare masonry building at a time when military buildings were constructed of timber. Again, there are no parallels for a building of this type at this date, and it has not been possible to assign a function to the three internal rooms – domestic or agricultural – though a likely explanation is that it was used for storage.
The dating evidence suggests that the building and ditches date from the immediate post-Conquest period, AD 42-70. There is much material of this date from the ditch and from pit-fills, including imported tableware, amphorae (implying the consumption of wine, olive oil, and fish sauce), Continental mortaria for preparing food, and glass, as well as crucibles and metal-working slag and even coinage. All this paints a picture of military or official contact at an early date, and raises the possibility that there is more to discover in other unexcavated parts of the site. Indeed, the excavators call this early structure the ‘ancillary building’, implying that it served to support some primary activity located elsewhere.
Two further buildings were constructed in the next period, from AD 70-120, consisting of separate northern and eastern ranges of a ‘proto-villa’, with mortared flint walls. The quoins forming the building corners of the north range were made of limestone ashlar from Purbeck, an expensive, high-status material. This four-room range also had a sunken room cut into the chalk. Such a structure on a military base would indicate a strongroom, though it could have been used as a cold store for perishable foods. Again, masonry structures are unparalleled locally at this period: only timber structures have been recorded in Durnovaria (the future Dorchester), the developing civitas capital and administrative centre for the region, which lay some 10km to the south.
Finds from the ditches and pits include brooches, a gaming counter, a box-fitting, Samian and amphorae, shale platters and trays, and glass imports. Yet the buildings contain pits full of industrial waste (charcoal, ash, slag, and burnt limestone) and the eastern range has two sets of ovens. Charred seeds from the fills include emmer/spelt grains, and there was a millstone fragment and part of an imported lava quern. Dyer’s weed and flax seeds indicate cloth-working as well as crop-processing. Whoever occupied this site was able to gain access to new materials and skilled artisans, while the imported luxury goods imply Romanised dining habits. But the excavated buildings do not fit with this image, and some of the other finds from this period – including box-flue tiles that might have been derived from a bathhouse – further lend weight to the idea that somewhere there might be additional buildings that are more domestic in character.
Two centuries of extravagance and expansion
In the period AD 120-200, the north range was enlarged by the addition of a room with a large apse, decorated with painted wall plaster and mosaic flooring. The east range was enlarged, too, and a tank installed in the larger of its two rooms. That tank, and the floors of both rooms, were lined with opus signinum, a mixture of lime-based mortar and crushed ceramic tile. A new single-room western range was constructed to replace the ‘ancillary building’. Although this range continued to be used as a store and workshop, its construction matches that of the other two wings, suggesting that aesthetic considerations were important. These separate ranges together formed the three sides of a courtyard – the first Dorset villa to replicate, albeit on a smaller scale, the arrangement of large palatial residences such as Fishbourne.
The finds associated with this period in the development of the site include hairpins, a child’s finger-ring, and perinatal infant bones, and it is likely from the quantity of material – pottery, glass, shale, and jewellery – that at least one family was living in the northern range, with servants and labourers in the other two wings, living on top of their work. Among the faunal remains, a rack of red-deer antlers suggests hunting trophies, and there are wild-bird bones from both garden species and wildfowl. Chicken, duck, and pigeon were on the menu, as were their eggs, no doubt, along with cattle, sheep/goat, pig, deer, eel, and trout, seasoned with wild celery seed. Hazel, cherry, apple, pear, and plum hint at an orchard, while fig seeds indicate a supply of dried fruit. Also present are tools used for the processing of wool and cloth, but on a domestic rather than industrial scale.
The theme of the period AD 200- 250 is one of further aggrandisement to upgrade the accommodation to the status expected of a wealthy Romano-British family, probably with trade or official connections to the nearby civitas capital. The north wing was remodelled and enlarged to provide three reception rooms, all with painted walls and mosaic floors, fronted by a narrow corridor. Rather than hypocausts, the rooms have fireplaces – not at all a common feature in Romano-British buildings. Social stratification is indicated by the assemblages of expensive imported drinking vessels associated with the residential block, in contrast to the cheaper locally made examples associated with the east range, which was demolished at this time and replaced with an aisled hall where servants and slaves might have been accommodated.
The west range was rebuilt to form a substantial four-roomed building containing a series of ovens, perhaps for cooking, although two of them were used for iron-working. Consistent with the continued use of the site for industrial and agricultural activity, a T-shaped grain-dryer was built to the south-west of the villa complex – again something of an innovation, as grain-dryers generally date from the 3rd or 4th centuries and are more common in Britain than on the Continent, possibly for climate-related reasons. The scale of the construction, using limestone blocks, flint nodules, and ceramic tiles, indicates crop-drying on a larger scale than previously.
Cultural climax and collapse
The period that saw the most extensive building works took place in the period AD 300-370. The residential north range was extended and given new mosaics and wall paintings. A catastrophic fire necessitated the rebuilding of the aisled hall, which resulted in a longer building with two additional bays. All three ranges were reroofed in expensive Purbeck limestone tiles, and the aisled hall roof incorporated decorative slate from quarries in Somerset or Devon. A perimeter wall was constructed to create a greater sense of a courtyard complex, with a formal entrance constructed over the infilled ditches to the south. Considerable wealth must have been available to the owners to enable them to commission the skilled masons, roofers, mosaicists, carpenters, plasterers, and painters, along with large quantities of mature timber and building materials.
There was no change in what was essentially a life of luxury for the villa owners and of continued investment in parts of the complex, though the period also sees the abandonment of the grain-dryer. Perhaps as part of some kind of closing ritual, the stoke- pit was then used as a grave for an adult male in his late 20s or early 30s, whose body was covered by part of a cattle carcase. As well as suffering from arthritic conditions, he had healed bone fractures and had suffered at least two severe blows to the head. According to the pathology report, the degenerative changes combined with his injuries point to a strenuous working life, as well as long-term and repeated ‘interpersonal violence’. This strange burial could be interpreted in different ways, but it looks suspiciously like the interment of somebody ill-treated and regarded as an outcast, possibly an enslaved person and perhaps even the victim of sacrifice or capital punishment.
It is not the only sign from this period of ritualistic deposits. The body of a juvenile owl, minus head and feet, had been placed in the chalk infill of the floor of a room in the north range, while five juvenile sheep were deposited in the tank in the aisled hall when that was filled in. Deposits of horse skulls were discovered in two ditches, and infant human burials were found in and around the service rooms of the northern and western ranges.
In the final Romano-British phase, AD 370-430, some of these refurbished buildings began to deteriorate, and it is assumed that the lack of maintenance and the crude repairs to mosaics implies a shortage of materials and skilled artisans. Indeed, pits were dug through some of the mosaic floors in the north range, suggesting a radical change of use and of priorities – perhaps a switch from the use of the buildings for social purposes and display to storing and processing agricultural products.
Among the finds were fragments of a high-status military belt-set, which could be interpreted in a number of ways relating to the splintering of central administration and military authority at this time. Was the villa owner seeking to maintain the culture of Romanitas at the beginning of a period of far-reaching change? Or was this a sign of the manoeuvring for power that probably ensued as former Romanised officials and military leaders took over the levers of power locally?
Despite social, political, and economic turmoil, those who lived at the villa did not experience hunger. Cattle, sheep/goat, and piglets were consumed in abundance, and more millstones and querns were recorded for this period than in any other. Roman-style agriculture continued, albeit on a reduced scale, now producing only for local consumption, with landowners relieved of the necessity to produce a surplus to pay taxes previously levied by the now-absent Roman authorities.
The pattern of contraction and decline continued into the final period, AD 430-650, characterised by the gradual decay and collapse of all the buildings. Pits filled with demolition debris testify to occasional efforts to clear the site and keep the west range and aisled hall clean in order to provide some kind of storage and living space. Late Roman amphorae sherds show that Druce Farm was linked to the economy of the South-West and to the continued supply of wine containers from the eastern Mediterranean. Druce Farm is also one of only three sites in Dorset from which hand-made late Roman pottery has been recovered. Found in the upper layers of the pits, these ceramics represent the final activities at the site before it was abandoned at some time after AD 650.
In many ways the site superficially fulfils the traditional model of a villa: a set of buildings at the centre of an estate that combines agricultural production with display and consumption, a place inhabited by the elite and their retainers, as well as by farm labourers and slaves. Yet it has many unusual characteristics, too. Though small, it has the earliest masonry structures in Dorset and continued in use throughout the Roman period and beyond. It lacks some of the building types, such as bathing facilities, that one might expect of a classic villa. Luxury and squalor co-exist – what must it have been like to have lived surrounded by ditches full of rubbish? One possibility is that the residential wing of the villa was not occupied all the time. Did it serve as an elite hunting lodge with only periodic occupation? Was it the second home of a dynastic family based in the civitas capital? Was it even in one family’s ownership for the entirety of its very long existence? Questions such as this cannot all be answered by archaeology, but it is clear that we do not yet know everything there is to know about villas, and that this high-quality volunteer excavation has added much valuable new information to extend our knowledge of the Romano-British countryside.
All images: Lilian Ladle and East Dorset Antiquarian Society, unless otherwise stated
Further reading
Lilian Ladle (2022) The Rise and Decline of Druce Farm Roman Villa (60-650 CE): excavations 2012-2018 (BAR Publishing, ISBN 978-1407360010, £95).
