< Main Page

Chapter 17

Roman Conquest, the Army and Villas

The south of France had long been of interest to the Romans. For one thing it provided the only land route to the Roman province of Spain and the prosperity of the area, which had blossomed after the foundation of the Greek colony of Massalia (Marseilles) in the early sixth century, made it seem to Roman eyes to be already half-civilized. This region between the River Rhône and the Alps was nominally in the control of the Greek city but the Romans had already intervened there on the colonists' behalf to discipline the native Ligurians. A permanent Roman settlement was established at Aix-en-Provence (Provence-Alpes-Côte d'Azur) round about 120BC. On the western side of Massalia the city of Narbonne (Languedoc-Rousillon) was similarly initiated a couple of years later. Further west at the foot of the Pyrenees, St Bertrand-de-Comminges (Midi-Pyrénées) was founded by Pompey in 77BC during his campaign in Spain. By Caesar's day the region was accounted as part of the Empire and it was not long before Massalia lost its independence (49BC) and became a Roman town like any other.

Caesar's conquest of Gaul was first restricted to the eastern and northern part of what is now France and adjoining parts further east together with the area we now know as Switzerland. In the eastern area, along the Rhône-Saône route northwards from the Mediterranean, had flowed the luxury goods and wine that the Gaulish chieftains loved and as a result the aristocracy there were to some extent predisposed to view the Romans as the harbingers of civilisation. The leaders of the tribes were flattered to be given command of auxiliary cavalry units in the Roman army and at the same time gain Roman citizenship. Others were encouraged to involve themselves in the creation and administration of Roman-style towns which could lead to either Roman or Latin citizenship.

Caesar's campaign in Brittany was an altogether bloodier affair, only completed in 56BC with the aid of the Roman fleet. Further south in Aquitaine Crassus led an army which effectively won the whole south-western area of Gaul for Rome in the same year. In 53BC Caesar savaged the Belgae of north-eastern France and the lower Rhine and in 52BC he consolidated the conquest of the Rhône-Saône-Seine corridor in a campaign which culminated with the siege and capture of Alesia in Burgundy, the oppidum of Vercingetorix, Caesar's greatest military opponent. In 55BC after eighteen days in Germany, Caesar marched to the Channel coast and prepared a fleet for an invasion of Britain, the first of two landings. Late in the season, in August, a force of two legions made the Channel crossing to Dover. After being battered to near disaster by tides, storms and hostile tribesmen, Caesar returned to Gaul just in time to suppress a revolt among the Belgae. A major force of five legions (half of the army in Gaul) was prepared for a second invasion of Britain in July of 54 BC. After a successful landing, the Romans marched as far as the Thames and secured tribute from several British tribes, among them the Trinovantes under Cassivellaunus. Before a permanent camp could be established in Britain however, Caesar had to return to quell disturbances in Gaul and never went back to Britain.

Romanization in Gaul proceeded region by region, the Romans winning the support of the local elite so assimilation started at the top of society and percolated downwards. An example of this policy at Saintes is provided on the Triumphal Arch in a tablet of dedication in 18 or 19BCE to the Emperor and his heirs by Caius Julius Rufus, a local magnate, whose great-grandfather was a Gaul, Epotsoviridos, who was granted Roman citizenship under Caesar. His son adopted a Roman name, Caius Julius Gedomo, which incorporated his Gaulish surname. His son was known as Caius Julius Otuaneunos, again using a Gaulish name, but his son who built the Arch was completely Romanized, the culmination of a process that took three generations and some seventy-five years.

The Roman Conquest of Britain of AD43 was probably inspired more by the Emperor Claudius' ambition than by any other factor although there certainly were a number of contributory reasons. At this time, Cunobelin, the leader of the major tribe in Britain, the Catuvellauni, was hostile to the Romans. His expansionist policy resulted in the overrunning of the northern half of the Atrebatic kingdom in Hampshire causing Verica, its king, to flee as a supplicant to beg aid from the Emperor.  As a result the diplomatic relations with Britain that the Romans had tried to foster for a hundred years broke down and the Romans were faced with a hostile power beyond the English Channel instead of an ally happy to maintain the peace on the Roman northern frontier. To add to this, Britain was the power base of the druids who were keeping alive nationalistic fervour on the Continent.

Under the command of Aulus Plautius came four legions, one fewer than the force commanded by Julius Caesar a hundred years before, but equipped and trained in much   the same way. The four legions were the Second Augusta (The Emperor's Own), the Fourteenth Gemina (Gemina indicates that this was a force formed from two amalgamated legions), the Twentieth Valeria (raised originally in Sicily), all withdrawn from the Rhine frontier where they had spent forty years and the Ninth Hispana (The Spaniards), in all about 40,000 men. The Roman army consisted of more than one type of soldier. Besides the fleet, known in Britain as the classis Britannica, there were units made up of Roman citizens, the legions itemised above; and other units raised from the provinces whose veterans, when discharged, would be granted Roman citizenship. These units were the auxiliary units.

The Fleet was based on the Channel coasts of both Britain and Gaul and operated as a transport and supply arm of the military. Its craftsmen helped to build Hadrian's Wall and worked the iron industry in the Weald of Kent and Sussex where the site of Beauport Park in East Sussex has been associated with it on the basis of a large number of tiles bearing the CL(assis) BR(itannica) stamp of the organisation, a military-style bath house and a slag heap covering a hectare. Occupation there lasted from the second quarter of the second century AD through to the middle of the third. Later on in the Roman period the functions of the fleet became more operational and it patrolled the seas around the coasts in camouflaged craft and uniforms during the third and fourth centuries to monitor the movements of Irish and Saxon pirates.

A legion had a fighting strength of about 5000 men subdivided into ten cohorts. Each  cohort other than the senior cohort consisted of six centuries. The senior cohort had five cohorts of double strength. A normal century contained about eighty men in the charge of a centurion. Commanding the legion was a senator of praetorian rank accompanied by a personal staff of six military tribunes. 

Auxiliary troops were organised into cohorts of infantry or alae of mounted troops and were usually specialist soldiers. There were units of archers, slingers, boatmen, swimmers and light infantry. A further military arm of the later Roman Empire included the part-time troops called federates settled as farmers along the borders of the Roman provinces on the understanding that they would help in their defence when called upon.

The main invasion force landed in Britain at Richborough in Kent, with probable subsidiary landings at other points, one of which may have been in the territory of the Atrebates in Hampshire. After the flight of their leader to Rome, the southern Atrebates were being ruled by Cogidubnus. His name appears in an inscription found at Chichester that presumably identifies him as a collaborator for he is described as being honoured by the bestowal of the Roman titles of King and Imperial Legate.  It is probable that a force, after landing in his friendly territory near Portchester and setting up a supply base, marched westward parallel to the coast to attack hillforts of the Durotriges in present-day Dorset including Hod Hill where Roman ballista bolts have been found and Maiden Castle where Mortimer Wheeler excavated a war-cemetery which contained some of the slain British warriors. The general in charge of this operation is likely to have been Vespasian, later to become Emperor.

As Caesar had done a hundred years before, the main force marched through northern Kent  and established itself just north of the Thames within striking distance of the British oppidum at Camulodunum and there awaited the arrival of the Emperor Claudius who turned up to lead the capture of the place. Thereafter a frontier zone was established along a line from Exeter to Lincoln backed by a military road linking the two places with a string of earth-and-timber forts between them. This military zone was supplied by army-built roads radiating out from the Thames crossing where the settlement of Londinium was speedily growing up. Roman roads were primarily military routes built by the army in order to move troops and supplies quickly from south-eastern Britain to the west and, later, to the north.

The early network was made more elaborate as time went on and towns were very often constructed at nodal points. In some cases there is evidence of later villas being linked to  Roman road networks as at Estrées-sur-Noye (Picardy) and at North Wraxall in Wiltshire which suggest roads were also to serve important economic roles.. As the roads were built for military purposes, they went directly to their destinations often bypassing Iron Age civilian centres, so that some important earlier settlements like Maiden Castle in Dorset simply withered away in favour of new towns built on the roads. In this particular case Dorchester was in a more economically viable location and so superseded the Iron Age centre. The same thing happened on the Continent where the Roman road network superseded the earlier one. On the Continent, a good example is the area of central France where several ancient strongholds went into terminal decline. Of course some Iron Age centres were fortunate enough to be on the Roman routes and had a new lease of life - Silchester in southern Britain is an example. A large part of the later  road network of the Empire including Britain is described in the Antonine Itinerary of the early third century which gives the names of Roman towns, forts and villages which lay on the routes, the distances between them and, usually, the total distances between terminals.

Roman roads were constructed by clearing a wide strip of land and draining it by digging ditches on each side of it. If marshy areas could not be avoided by the engineers, a substructure of brushwood and baulks of wood was first laid down. This technique was used on Watling Street at the crossing of the River Medway. Material excavated from the ditches was piled on the intervening strip to form the agger. Large stones were usually placed on top of this to form the road foundation and it was generally topped with a rammed layer of finer aggregate for the road surface. In all cases local materials were used and the construction varied to suit whatever was at hand. Whenever possible, bridges were built over rivers. In Britain these were usually of timber but masonry examples survive on the Continent as at St Charnas (Provence-Alpes-Côte d'Azur) in France. Where bridges were not used, a stream would be forded and its bed reinforced with a layer of stones. Milestones were an integral part of the main road system. Some were uninscribed but many carried an inscription mentioning the emperor's name, sometimes the name of the place from which the distance was measured but often only a mileage.

Rivers were used for the carriage of heavy goods. Britain has a good network of navigable rivers and flat-bottomed barges of the sort discovered in the Netherlands and at Magor in Gwent would probably have been used on them. In one area at least artificial rivers were built to link the rivers: around the Wash, in eastern Britain, the Car Dyke and the Midfendic were canals probably used both for navigation and drainage. Sea-going vessels for coastal work are known from Britain. A small carvel-built ship, found near County Hall in London, can be dated to the late-third century. Another found at Blackfriars apparently sank while unloading a cargo of building stone from Kent.

Wharves and harbours have been identified in many places, one of the most impressive  waterfronts being at Billingsgate on the north bank of the Thames close to the Roman London Bridge. At Dover, on Castle Cliff, there still stands one of a pair of lighthouses (pharos) marking the entrance to the harbour. Transport of heavy or fragile goods around the coast must have been common. A good example is the shipping of Black Burnished Ware from its place of manufacture in Dorset up to Hadrian's Wall.

The military zone with its road which we now know as the Fosse Way became the jump-off line for further advances into the west and north of Britain. It was at an early year in this phase, in AD60, that the Boudiccan revolt took place and the incipient towns of Camulodunum (Colchester), Londinium (London) and Verulamian (St Albans) were sacked. Evidence of the conflagrations caused by the rebels appears in the trenches of archaeologists excavating in these towns today.

Various governors of Roman Britain took over the conquest of successive areas of the west and the north. One of the best known is Agricola, the 'fort-builder', responsible for a line of forts along a road which nowadays would run from Carlisle to Newcastle-on-Tyne, a frontier for a brief period, and for the conquest of Scotland where he built numerous earth-and-timber forts including the legionary fortress at Inchtuthil and a further frontier line from Clyde to Forth.

The evidence of the presence of the Roman army at different times in various parts of Britain takes the forms of military roads, milestones and fortifications, ranging from simple earthwork camps that still lurk in the recesses of ancient woodlands to the great stone forts some of which today stand to their full height like that constructed at Portchester in the fourth century. An example of a fort of the standard type is Housesteads on Hadrian's Wall equipped with barrack blocks, headquarter's building, granaries, hospital, commandant's house, latrine, drainage system and ovens. Forts capable of housing a legion are referred to as fortresses, others sometimes as vexillation (detachment) forts.

During the later Roman period the authorities found it necessary to build forts on sites not of their own choosing: marshy coastal locations along the eastern and southern English coasts and the northern shores of France where it was likely Saxon pirates would land. At the same time the riverside wall and a watchtower downriver at Shadwell on the Thames were constructed as a  protection for the commercial city of Londinium against the same threat.

These Saxon Shore Forts were irregular in shape to accommodate the sites and their foundations were supported on timber frameworks. Examples of this device has been found at Burgh Castle in Suffolk and Portchester in Hampshire. Such coastal works and other defensive frontiers like Hadrian's Wall and the turf wall of the Antonine period in Scotland both built to replace  Agricolan forts survive to impress us today but we tend to forget the tileries, potteries, iron-works, timber industries and so on that the army was also responsible for.

The army must have played a part in the setting-up of the coloniae (sing. colonia), the retirement towns for soldiers who were given a plot of ground inside the walls and an allotment outside. Camulodunum (Colchester), Lindum (Lincoln), Glevum (Gloucester) and Eboracum (York) were all towns of this sort. Another aspect of the army's activities  lies in the temples and altars dedicated to their favourite deities. Further evidence of the roles of the army is in the military documents that have come to light in excavations at Vindolanda, one of the forts built by Agricola and superseded by the Hadrian Wall complex.

Soldiers must be credited with the introduction of the first 'market-friendly' coins of small enough denomination to be used by the whole population. They contrast with the 'prestige' issues of the late-Iron Age kings. Similar useful low denomination coins do not appear in Britain until the new-style coinage of Edward I in AD1279. An example of a place where the practice of using small change would have been introduced to local people quite unused to coinage is the vicus (civilian settlement) which grew up outside the western gate of the fort of Vindolanda to service the off-duty demands of the garrison. An equivalent excavated settlement has been investigated at Zugmantel (Hessen), a frontier fort on the Rhine limes similar in location to Vindolanda on the northern British frontier.

This list of military activities shows how far-reaching the influence of the Roman army was on Britain, particularly in the early years of the province. Even so it would have taken some time for the native British aristocracy and middle class to find their feet in the new Roman environment and they must have spent a good many years at first learning and copying from the military. At the same time they had to accustom themselves to the Roman style of administration. The original provinces of Roman Britain were ruled by a governor who controlled both the civilian and military bureaucracy as well as being the Commander-in-Chief. He was assisted in financial matters by a Procurator who reported direct to the Emperor. Decrees and tax demands emanating from their offices were passed to Coloniae, Municipia (sing. municipium) and Civitates (sing. civitas) to be carried out by the British leaders in each area.

Municipia were towns with official charters while the civitas were the tribal areas. Each of these was governed by the local British aristocracy with an Ordo, a council of one hundred elected decurians, which chose from its own members two Duoviri Iuridcundo and two Aediles. The former, senior magistrates, administered the law and tried minor cases, while the latter junior magistrates looked after everyday affairs, maintained roads and were in charge of public buildings. Within each civitas there was an administrative town which replaced any Iron Age tribal capital as well as minor settlements which were very often industrial centres like the iron-working site of Verlucio on the Bath road in Wiltshire and the lead-mining site of Charterhouse-on-Mendip in Somerset.

If the army was a considerable or perhaps even the major romanizing influence, it could not have been the only one. Without other help, towns and villas would not have been built nor would coins and imported pottery have circulated in the province. As it was, British local authorities had the example of the coloniae to follow and by the time the civitates towns were being built some time after the Conquest, architects, builders, surveyors and other experts would have arrived to assist in planning them and advise on building construction, layout and the running of the first British towns and agricultural villas. Entrepreneurs of other kinds may also have flocked to Britain, glimpsing an opportunity of enriching themselves on the 'new frontier'. One such, identified from documentary evidence, was Viducius Placidius who traded between Britain and the Rhineland.

Villas were appearing in the landscape in western Europe by the first century AD. Investigation of them on the Continent is not as extensive as in Britain but in the Somme department of northern France an  aerial survey has produced a map of the location of the villas which reveals a surprisingly dense distribution on the chalk plains, each villa commanding the surrounding rich wheatlands which formed its estate and provided its wealth.

Villa types are similar to those found in England. A large example is the Estrées-sur-Noye villa which can be compared to an equivalent British villa at Bignor in Sussex.  In the Somme region, villas did not survive as long as those in Britain. They disappeared in the upheavals which afflicted northern Gaul in the latter part of the third century.  According to French archaeologists they were replaced with agricultural settlements which lie under modern-day villages. Unfortunately, in this area the remains of the villas are being ploughed out of existence very quickly by modern deep-ploughing farming techniques.

In the Rhineland, a typical villa is that of Kôln-Mungersdorf which started as a simple farmhouse and grew over a period of years to house a prosperous landowner of the third century. The house, with bath suite, painted plaster and a limited hypocaust system, stood within a walled precinct which also enclosed gardens, orchards, two barns, a grain store, a pig-sty, a cattle byre, a stable, servants' quarters and a cemetery. In some ways this arrangement is similar to many of the villas in Picardy (like Warfusee) and Aquitaine that were contained in similar, but larger, rectangular courtyards up to two hectares in area around the walls of which were built the farm buildings with the owner's house inside a smaller, individual courtyard at one end. Those in Aquitaine contain more good-quality mosaics than any elsewhere in Gaul. A similar type of 'walled' villa appears in Britain at Gatcombe in Somerset and perhaps at Badbury in Wiltshire, both originating in the late-third century AD.

Large open halls form part of some villa residences in the German provinces. In Provence villas tend to be more like the Italian types, very often less luxurious than those further north, presumably because they many were run by managers producing wine and olive oil for the commercial market. In these there is evidence for wine and oil storage.

Estimates of villa population are difficult to make but archaeologists in France offer the suggestion that a large villa may have housed up to a hundred people including servants and estate workers. Certainly in some parts they could be very populous establishments, the equivalent of a large agricultural settlement elsewhere. A recent find of 700kg of metal, more than 1000 items, in the Rhine at Neupotz near Karlsruhe was probably lost while being loaded onto the boats of a party of Alamannic looters. It represents the complete metal contents of a large villa apart from the personal silverware and jewellery of the villa owners. The objects include kitchen implements, agricultural and carpentry tools, weapons, relics from a shrine and even heirlooms - Greek and Celtic antiquities! It is suggested that the looted villa stood in the region of Champagne during the third century AD.

Few villas were built in Britain before the middle of the second century when towns were being extensively developed to provide both local market-places and luxurious town houses for upper-class Britons who were encouraged to take on the roles of local government officials.

But gradually prosperous landowners in the countryside of lowland Britain were abandoning the simple, rectangular cottage-farmhouses which had replaced the late-Iron Age roundhouses and were equipping grander buildings with some of the town-house luxuries like mosaic floors, painted plaster, hypocaust heating and bath-houses. These comfortable establishments with agricultural outbuildings set in the middle of their estates are the characteristic villa complexes of the Romano-British period.

As you journey further north the villa phenomenon dies away. There, and in Wales and Cornwall, substantial farmers were more conservative, and lived in traditional farm complexes often built in stone and surrounded by an enclosing wall. A well-preserved site of this sort is at Din Lligwy on Anglesey where the complex includes two round buildings and a number of rectangular structures, two of which were used for smelting iron.

During the troubled third century agricultural producers seem to have had a difficult time, some of the villas like Lullingstone in Kent and Ditchley (Oxon) were temporarily abandoned. But the golden age of the villa was still to come in the fourth century when the great courtyard villas appeared in considerable numbers. Many of these were built in the West Country. In most cases the courtyard was probably a farmyard but sometimes, as at Chedworth in Gloucestershire, there was a second court that perhaps served that function leaving the main court perhaps for gardens and gracious outdoor living.

So large are some of them with so many rooms that it has been suggested that they were not residences but religious centres with accommodation for pilgrims. This idea has been put forward in the case of those sites where there is evidence for the veneration of Bacchus and other gods in the form of mosaics (Framton, Dorset), religious statuettes (Littlecote, Berks), temples (Chedworth, Glos) or baths or bathing pools (Gadebridge, Herts). Other large villas suggested for this category are Box, Wilts, Great Witcombe, Glos, Lufton, Soms and Brading in the Isle of Wight. By the time the sumptuous courtyard villas appeared in the fourth century perhaps a degree of amalgamation of estates had taken place but it cannot be proved. The fact the suggestion that some of these villas might be cult centres can be made shows how little we know about the functions and workings of our Romano-British villas.

Excavation on villa sites has often been aimed at uncovering mosaics and hypocausts in the residence. Not enough has been done to examine the ancillary farm buildings or to investigate the environmental evidence. In the West Country where some of this work is in progress the following table summarizes what has been discovered so far about the products in villas of all periods:

 (Adapted from Branigan)

Some of the above obviously reflects the degree of rigor of the excavation process but a picture does emerge of a mixed farming economy with the emphasis on the grain, leather and animal fats demanded by the army. The picture is biased in favour of the type of environmental evidence which is easiest to recognise and deal with on excavation, that is, the animal bones so we are left with a view in which the crop evidence is under-represented but we do get a blurred snapshot of the agricultural products that were on the market and we are provided with some idea of the activities that were being undertaken in the Romano-British countryside. . But, subsequent to the sale of the agricultural products, much processing must have been carried on in the villas. Flour-milling, malting, corn-drying, threshing, leather-working, shearing, fulling, spinning, weaving can be suggested. We have evidence for a few of these activities from Frocester Court villa in Gloucestershire.

During Roman times all such activities were very labour-intemsive so that each villa would have had a large work-force, either slave or free, or a mixture of both. Whether these workers were commonly accommodated in the villa complex or housed elsewhere in a village on the estate we do not yet know as the outbuildings of Romano-British villas are not commonly excavated.  Picardy villas are often found adjacent to modern villages. This gives rise to the suggestion that the village, if its origin goes far enough back, could have been contemporary with the villa. In Britain we do not have the same close siting association and one wonders whether the most fertile Romano-British countrysides would have had many settlements in them apart from the villas.

Other, more industrial activities, like iron-working, coal, lead, stone and shale mining and pewter manufacture are known from some of the same villa estates. In addition a great deal of timber must have been grown and processed both by the army and by civilian entrepreneurs.

Another unknown area is the extent of the villa estates. So far no villa estate boundaries has been fully identified although the boundaries of a villa in Hampshire come closest to it. Perhaps more intensive fieldwork would be the best way of solving this question - natural features such as streams and man-made features like roads would have been obvious boundaries and perhaps could be still recognized as such today. We do have some idea of the sizes of the fields: in north Somerset, fieldwork alongside two villas has identified long rectangular fields approximately 80 metres by 27/30 metres wide. These, or enclosures of similar area may have been characteristic of the time. It is difficult to visualize commercial farming being carried out in the old-style celtic fields.

Roman-period fields have been identified around the settlement at Chalton in Hampshire. They are rectangular, like those above, in contrast to the square shape of the celtic fields associated with settlements of the Iron Age, and may indicate that a new type of plough which turned over the soil with a mould-board was being used by peasant farmers as well as by villa owners. If the new plough had been adopted it would account for these fields being considerably larger than the cross-ploughed enclosures of the Iron Age. An area outside Britain, where survey has identified similar fields is near Mayen by Koblenz in the Rhineland. There the fields measured 180m by 20/25m, and were associated, as at Chalton, with a native site.

Settlements in the Romano-British countryside as in other archaeological periods are often best preserved on the chalk, particularly in the Ministry of Defence training area of the western half of Salisbury Plain where aerial photographs of sites like Knook West show the complete layout of a Romano-British peasant community.

An excavation has recently been carried out at Boreham in Essex of a Romano-British farmstead of the first to fourth centuries AD. It might be thought of as being well on the way to achieving villa status for in the last phase there was a bath house and the food debris included imported olive stones and pine kernels which so far are known only from the colonia at Colchester. This could mean that the original owner was a retired Roman soldier and the idea is strengthened by possible evidence of falconry, a sport foreign to Britain. At the bottom of the farm well, apart from the exotic foods mentioned above, there were cherry and plum stones, walnut shells, fish bones, bones of geese, duck and woodcock, hare bones and oyster shells. The farmhouse was timber-framed and wattle-and-daubed in the usual Romano-British manner with the bathhouse attached to one end which had hot and cold rooms and a furnace and water that was drained into a nearby pond. Running a bathhouse was an expensive business and suggests that the farmer was fairly well off. Evidence from a burnt-down granary shows that spelt wheat, barley and peas were grown. Around the farm some evidence has been found for fields and trackways.

On the Continent as well as in Britain some celtic fields survived into the Roman period and examples on Romano-Gallic villa estates have been identified at Allain (Centre) and Chatillion (Centre) in forested areas which have preserved the boundaries. Further south, centuriation, a system in which large areas of countryside were laid out in a regular grid system around coloniae, is much more evident and reflects the considerable number of Roman coloniae in Provence. An actual map of a system around Orange (Provence-Alpes-Côte d'Azur) still exists.

As can be expected, grain was the chief crop in the north of Gaul. New cereals in the Roman period include breadwheat, durum wheat, rye and oats. Emmer and spelt wheat became less common. Carrots, cabbages, apples and hazel nuts have been recognised in excavations. A Belgic innovation at this time was the vallus, a mechanical reaper, pushed by draught animals ahead of them into the standing grain crop to remove just the ears of corn. At Barbegal near Arles (Provence-Alpes-Côte d'Azur)) in the south is a bank of seven water mills which could produce nine tonnes of flour in a day. So far this example of large-scale manufactory is unique in the western Roman Empire.

Wine and olives in the south were probably introduced into the area by the Greek colonists in pre-Roman times. Production of wine gradually spread north up the Rhône and Saône valleys where the glory of Burgundy was established and even further north to where vineyards were introduced to the Moselle valley. As a result, in these areas and in the south common finds in villas are wine and olive presses, wine vats and amphorae potteries. Later, in the north, the tendency increasingly was to transport wine in barrels. Two transhipment sites where wine was emptied from amphorae into barrels have been identified at the riverside towns of Toulouse and Chalon-sur-Saône. In both places large quantities of smashed discarded amphorae have been found.

Agriculture was a vast industry in Gaul, Germany and in Belgic lands at all levels from the humble peasant farmstead to the grand villa. As a result, much forest was cleared so that the countryside would probably have carried no more woodland than it does today. Pigs were enclosed in the surviving forest while cattle and horses grazed on the pasture. Sheep and goats were less common in the north of Gaul. In the homesteads domestic cats were rare, dogs ubiquitous and domestic fowl becoming more common. Hunting was the perquisite of the wealthy.

It is interesting and perhaps surprising to realize that fish farming was practised around the coasts of Brittany and in the estuary of the Rhône. There was apparently a good demand for dried fish while of course, anybody who has ever excavated a provincial Roman site will know of the appetite for oysters and other shellfish.   

The main economic axis of Roman Gaul was the route from Arles up the Rhône to Lyon thence up the Saône to Chalon-sur-Saône (Burgogne)) where goods would be transferred to wagons or pack horses for the trek to the head of the Moselle and thence by water down to Metz, Trier, Cologne or Mainz. All of these places were redistribution centres radiating Mediterranean goods outwards and collecting the local produce in exchange. The latter two towns, Cologne and Mainz, were also the gateways to trade-routes beyond the Roman frontier into barbarian Germany.

Another important route ran from Narbonne (Languedoc-Rousillon) to Toulouse (Midi-Pyrénées) and so to the west coast of Gaul at Bordeaux. Along the waterways of the Garonne and its tributaries barges carried bulky goods as well as lighter breakable articles like pottery. Amongst the heavier goods were minerals and metals. From Limoges  came gold from its mines, from the Massif Central came iron, lead, and silver which could also be found in the Pyrenees, Brittany contributed tin and copper while both Spain and Cornwall were tin producers at different times during the Roman period. Cornwall could be reached by the sea-route around Spain as well as the route via Bordeaux.  Roman iron mines can still be seen at Le Goutil (Midi-Pyrénées) in the Pyrenees. Smelting was done at many sites. In Britain the area around Verlucio in Wiltshire is littered with tap slag from shaft furnaces and at Les Martyr (Languedoc-Rousillon)) in France it has been estimated that between 8 and 10 million tonnes of ore were processed in 230 years.

Stone was an important product. In Britain we know that the Bath stone mines near Box in Wiltshire produced the material for constructing buildings in the nearby spa and the surrounding villas. Roofing and floor tiles came from quarries in the pennant sandstone near Keynsham on the river Avon and was widely distributed along it in the same way as the stone from Seyssel (Rhône-Alpes) was along the Rhône in eastern France. In western France, the St Béat marble dug near St Bertrand-de-Comminges (Midi-Pyrenées) was used for building purposes and is found widely in the region. Evidence from some sites in Roman Gaul demonstrates that architectural features were being ready-made in the quarries. Stone was quarried in for building in Germany also but very often as well for making mill- and grindstones. The Niedermendig lava was well-known for this purpose but similar stone was mined at Kruft near Mayen with the Rhine serving as the artery of distribution for both materials.

Rivers were commonly used for the inland carriage of pottery which certainly was too fragile to travel very far in any other way. The best-known pottery in western Europe is samian, made at various times in three locations in Gaul. Its most prolific manufacturing site was in the area around La Graufesenque in the south. Two other areas were around Lezoux near Clermont-Ferrand in the centre and in the east around Lyon. This industry was probably established in Gaul by potters from the area of Arezzo in Italy whose Arretine ware was the predecessor of samian and was also widely popular. Samian manufacture increased sharply from the early second century AD but from the later part of that century it began to lose its popularity and at about AD200 disappears from archaeological sites in Britain although what replaced it in popular esteem as tableware is not clear.

The Iron Age in Britain
main page
Religion, Towns and Frontiers