Devon in the Roman Period
The Roman army which invaded Britain in AD 43 had first to subdue the warlike tribes of the south-east, after which it diverged to the west, north-west, and north to complete the conquest of lowland Britain. The westward route fell to Legion II Augusta, under the later emperor Vespasian, with accompanying auxiliary troops. Sussex and much of Hampshire, under King Cogidubnus, allied with Rome and gave no trouble; but the Isle of Wight had to be conquered, and then the belligerent Durotriges of Dorset and southern Somerset who bravely but unavailingly defended their hill-forts against the long-range fire of Roman ballista artillery which swept the ramparts before a storming party went in.
Roman troops had to be left to garrison the conquered area, and it was probably only part of the legion which eventually entered Devon. About AD 47 a temporary frontier line was organised on the Fosse Way running south-westwards from Lincoln to the mouth of the Axe, its southern end marking the probable boundary between the conquered Durotriges and the Dumnonii. Penetration into Devon followed, with securing a crossing of the Exe as a first objective. The building of a legionary fort, with barracks, granary, workshops, and a massive bathhouse, covering the centre of the site of the later city, followed. It was smaller than the normal size for a full legion, and possibly part of the Second remained at Dorchester to overawe the Durotriges.
No clear evidence of armed resistance by the Du.mnonii has yet come to light, and they may, unlike the Durotriges, have been too thinly spread and weakly organised to resist effectively. But Roman troops spread well beyond Exeter, and eventually into Cornwall, where a fort was established at Nanstallon near Bodmin. Excavation, sometimes directed by aerial photography, is gradually revealing more of Roman military activity. Marching camps put up as temporary shelter for troops on the move have been traced at North Tawton (where a fort was shortly erected), at. Alverdiscott, and north of Tiverton; and the supply-base at Topsham, serving as port for Exeter, was also probably defended, though no enclosure has yet been traced. Dorchester had a similar port at Radipole near Weymouth, which may support the suggestion that the legion was divided.
A fortlet on Stoke Hill above Exeter covered a signal-station communicating with the estuary; and a similar one was established high above the north coast at Old Burrow near the Somerset border, to watch for and report any water-born activity by the Silures of South Wales. This was probably set up about AD 50, but the site must have been very inhospitable for its poorly-protected garrison. A few years later it was replaced by a better-built one at Martiphoe, which had timber barracks for a ‘century’ of eighty men. Timber barracks have lately been traced inside the ramparts of Hembury hill-fort, showing that a garrison was for a time placed there (as in Hod Hill fort in Dorset); and roads were engineered to simplify troop movements.
In AD 60 Boudicca raised her at first triumphant revolt, and orders were sent to Legion II to march for the defence of London. The commanding legate was then absent, and the senior officer in charge refused to move. Why is-not recorded, but one may guess the dispersed state of the troops and the threat of local revolt. On hearing of Boudicca’s defeat, he killed himself.
The considerable garrison at Exeter, with pay to spend, must have as elsewhere attracted civilians to meet its needs, who would have settled outside the fort ramparts, and these became the nucleus of the later city population. When the legion actually left on being moved to prepare for and mount the invasion of South Wales is still in some doubt, but there are traces of military occupation at Exeter to as late as about AD 75. The Legion’s eventual going left a vacuum which had to be filled with some other form of authority and administration, and this was done as elsewhere by organising the Dumnonii as a Civitas (canton) with Exeter as their capital.
Somewhere about AD 80 most of the military timber buildings in the fort were demolished, and civilian ones substituted; but the road running around the inside of the ramparts, and the ramparts themselves, were retained, the former serving as the basis for the grid layout of streets. This was part of a general policy of ‘romanising’ at least the upper class of Britons, carried out with the help of money and skilled surveyors and craftsmen provided by the government. A large central block was laid out and gravelled as the forum, surrounded by the usual shops and offices, with the basilica or city hail at the northern end. The legionary bath-house (whose hypocaust pillars remain under the turf just southwest of the cathedral) was reduced in size, but retained for the use of the townsmen.
Little is yet known of private buildings in Roman Exeter, but on evidence from elsewhere we may assume the development of long and narrow houses with shop-fronts on the main streets for traders and craftsmen, and the more elaborate town houses of tribal notables occupying blocks away from the centre. The place certainly grew beyond the area of the legionary fort, and late in the second century its ramparts were levelled and new ones built to enclose well over twice the original area. There is some evidence to suggest that the gates were built in masonry at the same time, but the front of the rampart was later cut back vertically to allow the building of a battlemented stone wall.
But beyond Exeter the romanisation of the civil population was only superficial, if not entirely lacking, and Iron Age farming settlements continued almost unaffected. The North Tawton fort seems to have survived in a new form, since its name of Nemetostatio appears in later road-books. Statio meant a posting station or tax-collecting site (or both), which from its position on the route into Cornwall seems likely. The fort recently traced at Okehampton, not far away, may have been a new site for the North Tawton garrison. The only villas in Devon were at Holcombe (Uplyme) and Seaton, in the extreme south-east of the county, and they were probably Durotrigan rather than Dumnonian. There are concentrations of Roman-type material at Plymouth and (to a lesser extent) at Tomes which may suggest partly romanised settlements; but elsewhere coins, pottery, and similar objects are thinly scattered and may not mean much romanisation. Their distribution in Devon is very sparse when compared to Dorset and other lowland counties.
With increasingly oppressive taxation after the Civil Wars of the mid third century, most tribal capitals in Britain show signs of strain. As the wealthy moved out to escape the burdens of public office (including responsibility for collecting taxes), by the later fourth century Exeter shows definite signs of decay. Grass and weeds grew in the forum, and pits were dug in the courtyard of the baths. Coins, fairly common up to about 350, then become scarce and have not been found dated after 388.
Imported amphorae (wine-jars) from the Mediterranean show trading contacts which may have introduced the series of plagues which hastened the fall of the Roman Empire and urban decline. Before the end of the Roman period the city was already in decay. The chi-rho symbol, found on a fragment of cooking-pot, shows there were some Christians by the middle of the fourth century. No temples or religious sites, apart from cemeteries have been discovered, though a number of religious objects have been found.
Isca Dumnoniorum was an outpost of romanised urban life amongst a people otherwise little affected, and geographically hardly a good centre for Dumnonia as a whole (which included Cornwall). With Exeter’s decline, the centre of tribal authority of the Dumnonii moved westwards, and later rulers appear to have been based across the Tamar. Some Britons however continued to live within the city walls – if only in patched-up ruins or native huts and continued there until expelled by Atheistan about 928; but urban life in any Roman sense had ceased by the early fifth century and long before the arrival of the Saxons.