< Main Page

Chapter 17

Rural Prosperity

There is no vast leap forward in agricultural technology during the Romano-British period. Research into the quality of Iron Age farming at Little Butser in Hampshire by the late Peter Reynolds led him to assert that it was very efficient and so we can assume that no great improvements could have come about without the aid of mechanisation which was not available except in the one case of a reaping machine invented on the plains of Picardy in Gaul but, as far as we are aware, not introduced into Britain.

Scale was the significant change in British agriculture. We cannot be sure how commercial agriculture was in the Iron Age but we do know from the evidence of the villas in the Romano-British period that large-scale production became common. This was brought about by the demand first of all from the military for food and raw materials and, later, from the towns and, also, we think, from export opportunities.

It is probable that large agricultural estates owned by the tribal nobles existed during the Iron Age but we have no proof of this apart from the existence of the establishments in the succeeding Roman period that are classified as ‘villas’.   However, it is likely that such estates (in reality, probably pre-existing estates belonging to the trbal aristocrats),  were in existence early enough in the Romano-British period for the first agricultural villas to appear in south-eastern Britain during the first century, the area that had been the most advanced during the late Iron Age. Villas, in the sense of buildings that were the centres of agricultural estates, existed on the Continent but the British villas were not identical to them nor is it likely that the estates were as big as the vast ‘spreads’ in Picardy nor were they run by managers as the continental estates, at least those in the south of France, seem to have been. Clearly, their establishment may have been inspired by the examples provided by these Gallic initiatives but they were British in character and they developed in an individual way. The standard work on villas is by A.L.F. Rivet.

‘Villa’ is really a portmanteau term that covers at least four categories of establishments.  That, most generally understood, is as an agricultural centre of varying size and quality, secondly were houses of industrial entrepreneurs like the several establishments in the iron-working area of Verlucio in north Wiltshire, none of which is far enough away from its neighbours to provide it with an agricultural hinterland. Thirdly are the really grand houses like Portchester, some of which would have had ‘home farms’ like the largest Victorian country houses  and finally, religious sites like Chedworth and Lydney Park which have been excavated and revealed in their true identities as healing and pilgrim sites. Perhaps Bath would have been classed as a villa on excavation if it had not been engineered by the Roman military into the Fosseway frontier and developed as a spa by the army.

As far as we can judge from the numerous excavations of the agricultural villa residences they were inhabited by the owner of the estate for the standard of comfort, except in the very smallest villas which may have been the farmhouses of small-time British farmers (erstwhile peasants), is of a higher standard than would be provided for a manager. They contain many of the conveniences like mosaic floors, underfloor heating, wall decoration and bath-houses that you would expect to find in the better-class town houses and one way of thinking of them is as town houses transferred to a rural setting. However,  so far, we have few examples of town houses in Britain but in Dover the ‘painted house’ dating from the second century AD is one with over 30 sq metres of painted walls that still stand in some places up to three metres high and gives some idea of how splendid such establishments could have been.

But, investigating the living standards of the British aristocracy during the Roman period does not take us very far in understanding the rural economy of the Romano-Britons. Unfortunately, the number of excavations specifically targeted at this aspect of the villa estate are very few. It means closing one’s ears to the siren calls of mosaics and other goodies in the villa residence and excavating amongst the ancillary buildings that were the working parts of the complexes.

They would include granaries, grain-dryers, barns, cart-sheds, stables, pig-sties and byres. No doubt some of these may be difficult to recognise but one would hope that when enough of these complexes have been examined a customary arrangement could be identified. Very little is known of the style of Roman agricultural buildings but in Britain it is likely that, like the villa residences themselves, they would be built of local materials and would vary in appearance with the district.

But to begin with the residence: The most characteristic form of the Roman villa residence in north-western Europe, the corridor house, soon appeared in Britain. In this form it consisted of an oblong block divided by cross walls into three or four rooms, sometimes with evidence of a verandah outside running alongside one long wall. The best early example is at Lockleys, Welwyn, in Hertfordshire, dating from AD65, built with dwarf flint walls on which the timber-framed building was erected. Similar buildings were erected at Park Street, St Albans, Hertfordshire (AD65) and Frocester Court in Gloucestershire (late-third century). The Park Street villa was of the same construction and materials as Lockleys but had five rooms and a cellar while Frocester Court, which demonstrates the longevity of the type, contained just four rooms in its earliest phase.

These simple buildings were often added to or rebuilt to bring them up to the latest and most fashionable form and this invariably meant that they were made bigger. At first all villas lacked the refinements of mosaics and bath-houses of later periods. It was not until the middle of the second century that these were generally considered to be essential by those who could afford them.  There were few exceptions to this generalisation. Eccles near Dartford in Kent, was an early luxurious mansion with bath-house, mosaics and ornamental garden and there was also Angmering on the Sussex coastal

Returning to less rarified altitudes, we find that during the second century the winged villa was becoming more common, either newly-built or developed by additions to the cottage house with the wings simply added on at each end of the oblong block. An early example is at Ditchley, Oxfordshire, built in the middle of the century, with six rooms in the oblong block and a single room built forward of the block at each end as wings. Park Street and Lockleys had wings added to the earlier houses. Sometimes the wings were added to the rear as at Walton-on-the- Hill in Surrey or at both front and rear, producing an ‘H’ plan as at Ickleton in Cambridgeshire. By this time, hypocausts and mosaics were becoming common.

Adding more rooms is a characteristic of this period and could be related to a desire for a better life-style - more room, more possessions, more slaves/servants or even an increase in family members. It might be the result of a change in the birth-rate or perhaps members of the extended family were being accommodated in the family residence. We might also suggest that the Roman use of buildings as status symbols was being transmitted to the Britons whom, as far as we know, seemed to have cared nothing for grand buildings during the previous Iron Age.

This construction activity, of course, provided employment for local builders working in timber, the traditional material. Other craftsmen could have been specialists like masonry builders, mosaicists, glaziers, hydraulic engineers and painters and these people would perhaps have been based in the civitas capitals and the same goes for those who provided the furniture and the fittings for the new rooms. No doubt by this time Britons had learnt these skills and were part of a new class of artisans/craftsmen in British society which included also carters, bargees, small merchants, shopkeepers, clerks and many others who had not existed before, at least, not in the same numbers.

However, the third century was a period of recession that certainly affected villa construction. The landowners in the south may also have suffered from competition from new entrepreneurs in the north. There is a considerable amount of evidence to show that increasing amounts of arable land were becoming available around the northern military forts and in Wharfdale in Yorkshire. The farmers there had the advantages of newer and, therefore, more fertile land, and also being nearer to the military who were important customers for their products.

But there were better times ahead, for recovery came in the late-third century. One of the phenonena of this period is the appearance of a hundred or so villas that can be ranked with the best continental houses in terms of size and quality. This development is so far unexplained. Some students have suggested a Gallic migration or a flight of Gallic villa builders/mosaicists etc from the troubles in Gaul of this period or even that some of these large later structures were religious centres rather than villas.

Interior decoration in villas is the most obvious demonstration of how ready some Britons were to adopt Roman ways and to cast off their own traditions. One would expect that in the decoration of their houses they might have preserved something of their own celtic art that was one of the most outstanding of their achievements. But there is no trace of it. All the designs are classical and would not look out of place in Rome itself.  This abandonment of such a heritage is one of the puzzles of the Romano-British period.

A good example of decoration in a British villa is at Lullingstone in Kent (Meates) for its situation beside the River Darenth ensured the preservation of its walls up to a considerable height under a layer of silt washed down the valley. In the second-century baths the exterior of the walls was cream-washed and stippled in red, yellow and blue and the fragments from the interior include realistic-looking fish, probably trout, swimming in blue water.

 In the basement was a nymphaeum, a shrine of the water nymphs. The white walls were decorated with panels lightly outlined in yellow, green and red with broader yellow stripes above a dado of rectangles and lozenges with ‘fried egg’ marbling in the same colours. At intervals, yellow trees, probably meant for date-palms, hung with scarlet fruit, separated the panels. In the south wall was a niche that had been filled in. When the blocking was removed it revealed a painting of three nymphs. One nymph wore a yellow cloak and necklace and bracelets and on her head an aureole of green leaves and a diadem. The others had red or blue cloaks. Later, this room was sealed and a room built above which was used as a place of Christian worship.

The south corridor of the villa contained a wall-scheme of white panels divided by deep red and purple pillars and covered with purple curvilinear motifs probably intended to represent veined marble. In the central reception rooms of the fourth-century house one  large, square apartment had a floor mostly covered with geometrical mosaics but with the central scene of the slaying of the two-headed Chimaera by the hero Bellephoron. He flies to the attack on his winged horse Pegasus and stabs his enemy with a thin red-coloured spear. The Chimaera is a small, whiskered animal with a second goat-like head on its back. Around are a number of dolphins. In three of the corners of the pavement representations of the Four Seasons survive shown as women. Winter is muffled up, Spring has a swallow on her shoulder and Summer has flowers in her hair.

Through this room and up a step in an apse is a semi-circular mosaic set in a plain border. This is the floor of the triclinium where diners reclined on couches around the mosaic that shows Europa being carried off by Jupiter disguised as a white bull with a wide smile. He is beckoned on by a winged cupid while a second cupid pulls his tail. The figures are shown in outline on a white ground with a deep blue sea. Europa’s transparent robe floats out behind her and the cupids have red stripes on their dark, blue wings. Above it is a Latin couplet which reads ‘If jealous Juno had seen the swimming of the bull, she would with greater justice on her side have repaired to the halls of Aeolus’. Figured mosaics like this tend to occur in the south, south-west and north-east of the area of villa distribution. A good many include representations of the ‘Orpheus and the Animals’ scene.

Gloucestershire is rich in villas and also in figured mosaics. Three featuring Orpheus are at Woodchester, Withington and Barton Farm outside Cirencester. In Somerset he probably appeared at Newton St Loe and Whatley. In the Isle of Wight he is playing his lute at Brading to a fox, some birds and a monkey in a red cap. At Littlecote in Wiltshire he is alone in a central medallion with the Four Seasons around him in separate panels, the Seasons are riding a bird, a panther, a bull and an antelope respectively. Further north in Lincolnshire two villas, Horkstow and Winterton, contain Orpheus pavements while in Yorkshire, Orpheus no longer appears but the mosaics at Brantingham, Aldborough, Malton and Rudston share certain characteristics with the Lincolnshire mosaics that suggest that they are the work of a mosaic firm based in York.

Another villa possessing mosaics which show a variety of scenes is Brading in the Isle of Wight. Others were at Bramdean, Frampton, Pitney and Keynsham.  Some of the most beautiful mosaic floors are at Bignor and Chedworth but very many have been lost in injudicious excavations in the past, having been uncovered and simply left to disintegrate in the weather, This happened after early excavations at Box in Wiltshire..

A considerable number of the largest and most sumptous villas are to be found in Gloucestershire and Somerset: Woodchester, Chedworth, Witcombe and Low Ham are some of the best. There are more in the Drenth valley of west Kent and include Lullingstone while others are in Hampshire, Wiltshire, Oxfordshire and in Sussex of which Bignor is probably the most important. More lie in Dorset and a further group extends through Northamptonshire into Lincolnshire and east Yorkshire.

In these later villa residences the wings were often extended to form, with the house, three sides of an enclosure around a courtyard with the fourth side closed by a wall as at Spoonley Wood in Gloucestershire, or a covered portico as at North Leigh in Oxfordshire or by a further range of rooms as at Bignor and Woodchester which are also provided with second courtyards.  (Rivet)

The view that some of these sumptuous establishments were religious centres rather than up-market residences has been suggested by Graham Webster and Martin Henig for the following villas:

Littlecote near Hungerford in Berkshire where bronze busts of Bacchus, a pottery face of the same deity, two pieces of pottery with bacchic inscriptions and the Bacchic mosaic and what is described as a sacred enclosure suggest at least a preoccupation with the god.

Gadebridge villa in Hertfordshire where the main structure was an elaborate 20.6m-long pool, perhaps for ritual bathing close to which a cache of votive offerings was discovered.

Chedworth villa in Gloucestershire was perhaps a pagan religious centre with votive altars, bathing facilities and probable accommodation for pilgrims where archaeologists have identified religious sculpture portraying seven pagan gods, including Bacchus and the Romano-Celtic god Lenus-Mars, as well as a fragment of a bronze hand of the Anatolian god Sabazius, who in Roman times became amalgamated with Bacchus. Bacchic scenes also appear in one of the mosaics. The site, perhaps a healing centre, was probably connected with one of Roman Britain’s biggest temples, a 21m-long building on a hillside 900m to the east.

Great Witcombe villa in Gloucestershire where the main structure was an elaborate bath-house. The centre was built on the side of a hill on top of a spring and included a substantial spring, perhaps a shrine.

Lufton villa in Somerset was another site with a bath-house attached to a shrine.

Box villa in Wiltshire, probably a very large cult centre, situated on a limestone bluff riddled with springs where religious sculpture of Neptune and other gods and a silver eye of a life-size cult statue have been found.

Frampton villa in Dorset with three Bacchic mosaics and no domestic structures such as bedrooms or kitchens.

Brading villa on the Isle of Wight. Another apparently Bacchic site which contains mosaics of the Bacchic prophet Orpheus and the paganised version of the Jewish god Yahweh, the Roman deity Iao who is sometimes identified as Bacchus.

Bacchic evidence is particularly important because in late-Roman times when paganism was facing competition from Christianity, the Bacchic cult developed into a sort of monotheistic religion in which a whole array of deities were merged into and absorbed by a god in the form of Bacchus. As in Christianity, death and resurrection are central to the Bacchic story and there is also a strong hint of the oneness of Father and Son (Bacchus was the son of the supreme god Zeus). The Greek author Plutarch described Bacchus as ‘the god who is destroyed, who disappears, who relinquishes life and then is born again.’

Resemblances between the plans of some of these villas and the pilgrim cult centre like Lydney in Gloucestershire parallel the similarities between the plans of the great villas of Picardy and the Gallo-Roman shrines of that region.

Not all the later villas were grand. Modest villas or farmhouses were built throughout the Roman period. The term ‘aisled farmhouse’ is normally used to describe these fourth-century examples since they differ from the early modest houses. Each usually consists of an oblong building divided longitudinally by two rows of posts into a central aisle and two side aisles like a church nave. They could represent barns but very often the posts are used to support partitions that divide the aisles up into separate rooms in which hypocausts, mosaics and frescoes can be found. Small bath-houses also appear yet the remaining space in the building looks as though it was devoted to farmworkers, storage and even animals. A good excavated example of such a building is at Clanville in Hampshire.

Romano-British villas must have employed a large number of workers, whether slave or free is not clear nor is their accommodation. Did they live in settlements on the estate away from the volla or were they accommodated in the villa complexes? The outer buildings are not usually excavated but often seem to contain a good many structures as for example, in the large complex of buildings that lie to the south-west of the enormous villa at Great Leigh in Oxon.  Many agricultural villas have one or more native settlements within 3-4 km of them. Could they have provided labour for their neighbouring villas? What may be a problem in this regard is that, in the south-west at least, many of the native settlements were first occupied early on in the Romano-British period and so precede the erection of the villas. (Branigan)

Most villas are found south and east of a line drawn from the Severn estuary to the north Yorkshire moors. Around some towns, notably Cirencester, Dorchester and Silchester, there is a distinct lack of villas and it may be that the surrounding land was farmed from the towns. A number of towns had farms or farmhouses within their boundaries. Silchester, for example, has near its north gate a house that is associated with barn- and byre-like buildings. Close to the Verulamium gate at Cirencester is a house that is not only laid out like a winged villa-house but has to the rear anciliary buildings that are probably workshops. One of them looks like an aisled farmhouse. Another winged villa-house plan is known in the town of Great Chesterford. In Wroxeter, aerial photographs show a building of similar plan. A corn-drying kiln was built into a small courtyard house during the early-fifth century that was situated in the centre of the town of Verulamium. However, the evidence for urban villas is not strong and, indeed, the phrase is a contradiction in terms.

So far, it has not been possible to define the perimeter of a villa estate  accurately but one can amuse oneself with an OS map using the natural features in the vicinity of a villa and try to suggest how they would fit as boundaries for, undoubtedly, natural features acted as boundary markers together with prehistoric man-made features like tumuli and were probably combined with artificial boundaries which no longer exist unless, as some would have it, they are preserved as later parish boundaries.

What were the products of villa estates? The army needed food, hides for leather and fats for lubricating their artillery and carts, both made from wood. So one would suggest that grain and cattle would figure largely in the list of products together with what would command a sale in the local town, the other main customer of the villa. Fieldwork at Stoke Ash in Suffolk has located a water-front settlement on a tributary of the River Dove which may have served as an outlet for produce from a Romano-British estate situated between Stoke Ash and Scole.  It is likely that a good deal of agricultural production was moved by water on rivers and even on streams that nowadays would be thought too small for practicable navigation as transport by road would have been expensive.

The following diagram (Branigan) is a list of products from excavations on some villas in the West Country.  (Branigan) Unfortunately, in the past, environmental evidence was not always collected so our knowledge from a good many major older excavations is very incomplete.

 A number of buildings were excavated at North Wraxall and were located by the excavator round a farmyard to the south of the main courtyard and the villa residence and of these one was provisionally recognised as a byre and another as a pigsty. At Barnsley Park there was sufficient evidence to identify corn production as a major part of the villa economy. Over100ha. of long, rectangular fields averaging about 60m by 27m have been recorded. Elsewhere on the Cotswolds there is evidence for cattle, pigs, sheep and wool production but no one product seems to have predominated.

Between the Cotswolds and the Severn there is some evidence of wheat production, spelt and either bread or club wheat, sheep and wool production, particularly at Frocester Court that also raised horses and grew oats like Hucclecote. In north Somerset, almost all villas were situated on the keuper marl which gives rise to well-drained brown earth soils. At Lye Hole the field system around the villa survives and is made up of long, rather narrow fields some 30m broad.

Reliable evidence of corn production comes from several villas in the region. At Star and Chew Park barley was grown as well as einkorn, emmer and spelt wheat. Peas and beans also come from Chew Park and also from Wookey. Chew Park produced evidence of cattle as did Brislington and Gatcombe. Star seems to have placed its main emphasis on wool sheep. In general on the keuper marl, farms seem to have been devoted to the production of mixed crops and beef cattle.

In south Somerset on the lias, grain seems to have been an important product with granaries or corn driers located at Bawdrip, Low Ham, Ilchester Mead and Pitney. Pigsties have been located also at Pitney. On the chalk south of Salisbury corn production is confirmed from Rockbourne Down and Downton with finds of corn driers. Spelt was grown at Rockbourne and wheat and barley at Downton where vetch and cleaver were grown for animal fodder. Animals at the two villas included cattle and sheep.

As Romano-British villas were self supporting in food one would expect to find a mixed economy. But they were geared to profit-making as well so one would expect to recognise a specialisation. Overall, wheat appears to have been the commonest specialisation with the role of cattle varying in importance from one region to another. However, we have some evidence for specialisation in sheep for wool and some pigs.

So, it is possible to say something of the products of villa estates but we know very little about the ancillary farm buildings used in this production. The best exidence is coming from excavations regularly carried out each year at Frocester Court in Gloucestershire. (Branigan)

In some places, around the small town of Verlucio in Wiltshire, for example, villa buildings are so close together that there is no room for each to have been surrounded by an agricultural estate. The area around Verlucio is littered with the evidence of Romano-British ironworking and it has been proposed that the villas belonged to the ironmasters with the workers living in the town. This has been suggested for other villas like Gayton Thorpe in Norfolk, Great Weldon in Northamptonshire, and Brislington, Gatcombe and Lansdown in Somerset.  All produce evidence for iron slag, furnace refuse and ironstone.

Cromhall in Somerset may have been involved in coalmining for the workings of the period are in the field next to the villa. At Chew Park there is evidence for lead-working and for pewter working at Gatcombe and Langridge, both in Somerset. Shale-working was carried on at Rockbourne and East Creech (Dorset) and perhaps the villa at Hazelbury near Box in Wiltshire may have been connected with the quarrying of Bath stone that was worked there in Roman times. The villa at Combe Down, just south of Bath, produced evidence of the manufacture of pewter where stone moulds have been found. From what has been said above, it is clear that some villas, like Rockbourne, were involved both in agriculture and in manufacturing.

We now know that villa residences were built for various purposes. Generally, they were centres of agricultural production but they could also be concerned with other functionss. Some, perhaps, were built as grand country houses like Fishbourne that belonged to some very prestigious politician or they could have been residences of high officials like the governor of the Isle of Wight who is said to have inhabited the villa at Brading, if it wasn’t, as some believe, a cult centre.

One does not know how long the tribal chieftains like Cogidubnus continued to be important figures in the administration of the civitates. If they maintained a position analogous to what their predecessors had during the Iron Age, they might well have built large and luxurious villas but this so far has not been confirmed by the archaeological evidence.

In the north and west, outside ‘villa country’ where peasants had not been organised to work on villa estates, settlements of the late-Iron Age continued to flourish during the Roman period. These include the rounds of the south-west like Carn Brae and Trevisker and courtyard houses like the group at Chycauster. There are also isolated huts like Port Godfrevy in Cornwall that produced coins, a samian dish, an amphora and iron nails which suggests that these traditional buildings should not be taken to indicate poverty or a lower standard of living than was enjoyed elsewhere.

A similar place in Yorkshire is at North Cave on the Humber west of Hull where there were two nucleated clusters of farmhouses linked by a trackway. Coins demonstrate that the inhabitants were active in the local economy, possibly in supplying grain via the River Humber to the market or direct to the Roman military base at Brough since grain-driers were being used. Farmers were also making pottery and perhaps iron tools.  

In Wales there are villages of rectangular wooden houses on the coast at Gateholme and Sheep Islan and large circular houses like those of the Iron Age. One at Dinorben was the property of a wealthy farmer who would have built himself a villa if he had lived in the south-east. Finds included Roman pottery, metalwork and ornaments, an up-to-date plough and a sickle of the latest Romano-British pattern.

On Holyhead Mountain there is a large settlement of round stone huts and at Din Lligwy on the mainland one of the polygonal farmsteads common in the region. A circular stone hut and some rectanguilar buildings were situated inside the five-sided stone enclosure with evidence of metalworking. It was the establishment of a well-off farmer comparable in status to the middling villa owner of the south-east.

Similar establishments are found in Northumberland at Elsdon Burn and at Ells Knowe. Nearer Hadrian’s Wall the round or oval enclosures are replaced by rectilinear ones as at Riding Wood where four houses were placed in a roughly trapezoidal enclosure with paved yards and paved entrance passages. Also in the north are loose collections of settlements with up to a hundred huts. No doubt these farmers were also engaged in producing goods for the Roman army that had a strong presence in this area. Carlisle and Corbridge would have been military markets and the presence of the rotary querns on neighbouring settlements shows that their technology was not behind the times and that their activity was not restricted to the raising of stock.

Rural settlements were not restricted to the north and west. In the south they existed. Typical examples are Chisenbury Warren on the north-eastern edge of Salisbury Plain in Wiltshire and Meriden Down in Dorset. The former had over eighty rectangular house-platforms as well as a system of lanes. Both these settlements are surrounded by tracts of celtic fields. It is possible that these agglomerations of small farmers were their response to the market economy of the period. Singly, they could not produce a surplus that would make much of a mark in the local market but collectively they could pool their several surpluses and put together substantial quantities of agricultural products that could command good prices in the market place.

A roadside settlement beside the Fosse Way in Somerset covering over 2 hectares has been investigated.  This is Fosse Lane, on the outskirts of Shepton Mallet, close to a Romano-British pottery-manufacturing site. Excavations revealed evidence of a small town or large village functioning as a local agricultural and minor industrial centre developing from the end of the 1st century and at its prime during the 4th century.  Streets, plot layouts, plans of both timber-framed and stone buildings, several small cemeteries existed and, in the latter phase of settlement, some continuity of use into the 6th century and the possibility of one Christian burial group.

A similar street was uncovered at Higham Ferrers in Northamptonshire. Along a stretch of 70m were the foundations of at least eighteen buildings, either houses/shops or workshops. On the other side of the road were two shrines where many brooches and pins were discovered, some ritually broken as an offering to the gods together with several  inscribed lead tablets that worshippers addressed to them with requests for favours. Finds on the other side of the street included tweezers, needles, chisels and pruniny hooks and other everyday articles while close by were two small cemeteries containing bath cremations and inhumations. The site was probably inhabited from the second century to the early fifth century AD.  

Salisbury Plain and Cranborne Chase are areas where few villas so far have been identified. This has led archaeologists to suggest that the region could have been an example of an imperial estate, the personal property of the emperor. Such tracts of land could contain reserves of important minerals but they were more usually areas of agricultural activity on the part of small-time farmers.

On Cranborne Chase the site at Iwerne was first occupied in the late Iron Age. Various rebuildings of the farm there resulted soon after AD300 in the construction of a narrow flint-built structure divided into rooms by transverse partitions. From one other long sides projected a tower-like structure with a raised floor which probably served as a granary. The largest room seems to have been an animal house for there was a stone-lined drain down the middle. Next to it was a smaller room with a roughly paved floor and beyond it a room which served as a living-room for it had plastered walls with painted decoration. This is an example of the simplest form of villa that developed on a traditional British farm site. There must have been many like it.

Other examples of this transition can be found in Sussex. In the area between the River Adur and Ouse over 22.5kms of lynchets are associated with at least 32 little farms. At Park Brow near Findon an earlier Iron Age settlement was succeeded by rectangular houses, one measuring 9m by 6m. The wattle-and-daub walls were plastered and colour-washed, the roof was tiled and some of the windows were glazed. A large door lock protected the front door.

Imperial estates were managed by the provincial procurator, the Roman ‘chancellor of the exchequer’ for the province who administered them through estate managers based on each of the estates. The home of one of these estate managers has been identified at Combe Down south of Bath where the villa produced an early-third century inscription which mentions the restoration by one of these estate managers of a building called a principia or head-quarters of an imperial estate probably at Combe Down itself. The estate could have comprised the lead and silver mines on Mendip to the south.

Another imperial estate is suspected in the Fen District. During the reign of Hadrian large-scale drainage took place there but there were no villas and the drained land was given over to small farms. At Godmanchester on the southern margin of the Fens there is a principia-like building constructed in the early-third century which could have functioned as the administrative head-quarters. In the Fens there are many farms and villages associated with extensive areas of field-systems. One of the nucleated sites is at Cottenham in Cambridgeshire that was laid out on both banks of the Car Dyke. Investigation has shown that while grain production was important, stock-raising was probably the most common activity.

Farming practice probably did not differ much over the British provinces. We can recognise elements of this traditional pattern in many places. Threshing floors have been found at several places, at Ditchley, for example, while corn-driers are known at a small farm at Wyboston in Bedfordshire and as a good example at Atworth in Wiltshire that shows that the flues were covered by two floors six inches apart made of stone slabs. The heat and smoke from the stokehole could be regulated by adjusting the flues. A wooden superstructure probably shielded it from any soot or smoke from the chimney. There is doubt of the exact purpose of some of these drying structures for experiments at Little Butser have shown that they are sometimes more efficient at malting barley than drying corn. Sites in Dorset have produced examples of such driers and they have also been found in towns.

Each town was surrounded by its belt of farmland, known as its territorium and this probably explains the lack of villas in the vicinities of some towns. Land around coloniae would have been allocated to the citizens by the government and should have been divided up by centuriation into a regular scheme of rectangular plots Traces of this scheme (centuriation or, more properly, cadastres) have so far been found in five areas: in South Norfolk, Gloucestershire and South Wales, Essex, Lincolnshire and Romney Marsh in Kent.

As time went on, land outside the military forts in Wales and the north of England was cultivated by the soldiers or by the inhabitants of the vici outside the gates. A late first-century granary has been identified at Caerleon and hillside slopes outside the fort at Housesteads were used for terraced production of cereals or vegetables.

Romano-British agricultural tools are known. Rakes and hoes from Newstead, a spade from Verulamium, a spade shoe from Suffolk, a ploughshare from Silchester, outsize scythes from Great Chesterford, smaller scythes from Newstead and Silchester and numerous bill-hooks and pruning knives have all demonstrated how little the designs of agricultural tools have changed over the millenia.

In fact, the agricultural technology established before the Roman period in Britain survived for many centuries afterwards. The villa residences were abandoned by the early part of the fifth century but the equipment, the use of slaves and the crops and animals that were produced on the farms changed little. The balance between crops and animals, essential before the days of chemical fertilisers and pesticides, and probably universal during the Romano-British period except on the most specialised villa estates, was maintained. However, there is still a good deal we do not know about Romano-British agriculture especially about specialisation. Horses, for example, were an integral part of military transport and of civilian transport too but traces of them on farms are sparse. Large numbers must have been reared. But where? They must have required a very large supply of food. Where were these oats grown?  For questions like this, archaeology has yet to come up with answers.
 

Building the New World
main page
Spiritual and Aristic Life