British Archaeology banner

Cover of British Archaeology

Issue 65

June 2002



Wealthy early Roman graves near St Albans

Dig in West Midlands reveals empty landscape

Medieval enclosed garden found at Welsh border castle

Porcelain finds show changes in 18th century taste

Medieval parchment from site of Canterbury’s friary

In Brief


Fortress Britain
Simon Denison on Britain’s Second World War defences

Great Sites
Paul Bidwell on the Roman legionary baths at Exeter

Engines of Change
David Gwyn on the landscape archaeology of railways

Lord of the Hrungs
David Hinton on the influences behind Tolkien’s epic


Armoured jacks, human sacrifice, and a ‘lost’ Roman town


Stop demolishing Victorian terraces, by George Lambrick

Peter Ellis

Regular column


Mosaics of the Greek and Roman World by Katherine Dunbabin

Migrants and Invaders by Malcolm Todd

The Molecule Hunt by Martin Jones

Britons and Romans edited by Simon James & Martin Millett

The Archaeology of Shamanism edited by Neil Price

CBA update

favourite finds

John Lewis on realising the truth about the Stanwell cursus.


ISSN 1357-4442

Editor Simon Denison


Great Sites: Exeter Roman Baths

Paul Birdwell recalls the discovery in Exeter of one of the largest and most elaborate bath-houses of the early Roman Empire

In 1972 people passing through the Cathedral Close at Exeter were baffled to find the Victorian church of St Mary Major gone and in its place a huge hole. Notices explained that the remains at the bottom of the hole belonged to one of the largest and best-preserved Roman baths in Britain, but many visitors were more interested in the jumbled masses of human remains.

Although these belonged to a cemetery that had been in use for more than a thousand years, rumours soon spread that plague pits had been found. For one visitor, the poet Ted Hughes, the cemetery was 'the House of the Dead', and the baths emerged:

'Out of the dried blood of the redland marl -
Splayed, bleeding in rain, like an accident,
Gaped-at, photographed, commented-on and coddled
With waterproofs. Nobody knows what to think of it.'
(from 'Here is the Cathedral', Moortown, 1979)

At the time even the most expert opinion was puzzled by the discovery, though not quite so much at a loss as Ted Hughes suggested. The baths were very early Roman in date, of large size and richly decorated. A military connection seemed likely - but it was thought that the Roman army had just a token presence in the South-West at the time. Why had it embarked on such an enormous building project?

Devon and Cornwall, the territory of the Dumnonii, were less Romanized than the rest of southern England. West of Exeter there are only two certain villa sites, and the Roman occupation had little effect on the pre-existing patterns of settlement. Until the beginning of the 1960s it was even thought that the military campaigns had stopped short of Exeter, the implication being that the Dumnonii had acceded peacably to Roman rule.

We now know, however, that Devon has perhaps the largest concentration of forts in southern England - the transition to Roman rule had been achieved by military conquest. This better understanding of the Roman South-West began with the excavations carried out in the 1950s and 60s by Lady Aileen Fox on military sites in Cornwall and North Devon; while more recently, new forts have been discovered by aerial photography.

In 1964, on the outskirts of Roman Exeter, a ditch of military type containing pottery of the mid-1st century AD was found. In 1971, only a few years after the last of the areas devastated by war-time bombing had been rebuilt, another large part of the city centre was cleared to make way for a shopping centre. Within a few months excavations found the post-trenches of Roman timber barracks. The site was far enough away from the ditch found in 1964 to show that there had been two separate military sites at Exeter.

At the same time, archaeologists from the Exeter Archaeological Field Unit (directed by Mike Griffiths) began work in the Cathedral Close on a site intended for an underground car park. One of the first trenches located a late Anglo-Saxon burial, which was cut through a Roman floor edged on one side with a line of large blocks. Before long, the line of blocks emerged as the top tread of a massive flight of steps that must have been part of a large public building, later revealed as the town basilica. There was astonishment when the floor at the base of the steps - originally set within a colonnade round the town's forum - was removed to reveal a large brick hypocaust (or underfloor heating system) which had been part of an earlier building.

In the following year a much larger area was opened and the hypocaust proved to belong to the caldarium or hot room of baths which had been built about AD 60-65. The main part of the room measured 22.3m by 9.75m, excluding two large apses holding water basins which flanked a rectangular recess. Part of the adjacent tepidarium was also revealed, together with a furnace-house, service rooms and part of the palaestra or exercise-yard with the remains of a cockfighting pit. Demolition debris from the baths included fragments of polychrome mosaics (probably the earliest known from Britain), black and white stone floor tiles, panelling, mouldings and basin fragments in Purbeck Marble and painted wall-plaster. The roof-eaves had been decorated with moulded tile plaques with either the face of Medusa or two dolphins.

The construction of the baths would have been a formidable undertaking at any time - but especially so as it was amongst the first monumental masonry buildings to be built in Britain (and almost certainly the first outside south-east England). Almost everything had to start from scratch. Five separate quarries had to be located and opened for the different types of building stone. The most distant was at Purbeck in south-east Dorset, which supplied the marble-like stone to be sawn on the site of the baths into sheets for wall-panelling, or worked up into mouldings and basins. Production of lead and iron was stepped up and kilns built to produce the thousands of bricks and tiles required. An architect was needed, as well as mosaicists, painters and masons. These skills were probably beyond the competence of the army, experienced though it was in basic construction techniques.


Work in succeeding years showed that the baths had been part of the fortress of the Second Augustan Legion, the base from approximately AD 55 to 75 for some 5,000 legionaries and perhaps also a cavalry regiment 500 strong. They were accommodated in timber barracks tightly packed within defensive ditches and a turf rampart that enclosed an area of 41 acres (19.6ha). The baths were large enough to be used by several hundred bathers at any one time. They offered the legionaries some of the luxuries of the Mediterranean areas from which many had been recruited.

In fact the Exeter baths were far more advanced in design than most of those in their homelands. In the mid-1st century AD, baths of the Pompeian type were still prevalent. These were of no great size and had simple plans typified by a loose grouping of rooms, although their decoration might be very elaborate. More ambitious designs were slow to spread. For example, at Pompeii and Herculaneum at the time of the eruption of Vesuvius in AD 79, none of the baths were as large or as regularly planned as those at Exeter.

On the other hand, from the 2nd century AD, baths were built in the wealthier cities of the empire which far exceeded the Exeter baths in size and complexity. Largest and most elaborate of all were baths of the Imperial type, built for the populace of Rome and copied in other leading cities. As emperor succeeded emperor, the scale of these baths grew ever greater, ending with the immense late-3rd century Baths of Diocletian in Rome, a building that covered 9 acres (3.7ha) - or 33 acres if taken together with its exercise yards, libraries and shops.


Exeter is one of the small number of baths on the north-western fringes of the Empire which suggest that the Imperial type was already emerging in the mid-1st century AD. The others are the fortress baths at Vindonissa (Brugg) in Switzerland, which replaced timber baths about AD 45; at Caerleon in South Wales, built about AD 80; and the baths at the colonia (or settlement of military veterans) at Avenches in Switzerland, built in the AD 70s.

These buildings have many features in common, the most important being the regularity of their plans. Instead of the loose grouping seen in earlier baths, this new type displayed a completely symmetrical arrangement of its rooms along the central axis of the building. The most recently excavated baths in this group are at Caerleon, where excavator David Zienkiewicz, then of the National Museums of Wales, argued convincingly that the main rooms were probably roofed with intersecting concrete vaults. Concrete was the only material which could cover the huge spans of the rooms.

The heating system of the Exeter baths reflects its advanced architecture. Much larger spaces had to be heated in this new type of baths; and at the same time the contemporary taste was for higher temperatures than formerly. One technical innovation designed to channel hot air between the wall and its tile surface was the introduction of hollow rectangular 'box' tiles to line the walls and possibly also the vaults, an improvement on the earlier flat tiles with lugs or nipples used as spacers to separate the tile from the wall. Exeter is one of the earliest baths where box tiles were employed.

At the point where heat from the furnaces entered the hypocaust, iron frameworks were used to support the raised floor - presumably there were fears that heat from the furnaces would quickly disintegrate the bricks used elsewhere in the hypocaust. Iron bars were also used, laid diagonally across the tops of the hypocaust supports, to carry the floors of the rooms above. This use of iron to strengthen the construction of the hypocaust has not been seen elsewhere and hints at the builders' lack of confidence in traditional techniques of construction for a new type of building.

The theory that baths such as those at Exeter represented an early stage in the development of the Imperial type of baths, first proposed when the Vindonissa baths were dug in the 1930s, is still controversial. The suggestion was that baths on the frontiers were a reflection of new architectural styles at the heart of the Empire - although by the accidents of preservation and discovery, no early baths of this type have yet been found in Italy.


Some scholars have objected that military architecture on the frontiers would not be influenced so directly by metropolitan architecture. But evidence suggests quite strongly that it was. Work in recent years on Augustan fortresses in Germany, the Netherlands and the Czech Republic, dating to the end of the 1st century BC and the beginning of the 1st century AD, suggest how this might have happened. Some of the larger buildings, for example the houses of the tribunes (senior officers), are of great interest as they are interpretations in timber of the most up-to-date and luxurious stone houses of the Mediterranean aristocracy.

Most remarkable of all is the inclusion in these fortresses of baths almost as large as those at Exeter, but constructed almost entirely of timber and clay with baths and swimming-pools lined with sheets of lead. Some have hypocausts of timber; others, without hypocausts, must have been heated with braziers within the rooms. This peculiar experiment in building techniques never seems to have been repeated elsewhere.

These early fortresses show the lengths to which military architects went in order to reproduce the masonry buildings of a Mediterranean city in timber, and the plans of the tribunes' houses indicate that the architects involved were working in the contemporary Mediterranean style. In the same way, architects who were highly experienced in the most advanced architecture of their own time appear to have been brought in to oversee the complex building projects at Vindonissa and Exeter.

It is now 30 years since the Exeter baths were laid open to public view. Their site is marked only by a small plaque, but many people remember them. The juxtaposition of the great medieval cathedral, its stone weathered and its outline intricate with detail, and the crisp, geometrical plan of the baths with its walls of purple volcanic stone and hypocausts of red brick, was an unforgettable sight. In some respects the two buildings could not have been more dissimilar; and yet both were trailblazers for the architecture of their own day.

Paul Bidwell supervised and published the excavations at Exeter and is now Head of Archaeology at Tyne and Wear Museums

CBA web:

British Archaeology

Jan/Feb 2005
Mar/Apr 2005
May/Jun 2005
Jul/Aug 2005
Sep/Oct 2005
Nov/Dec 2005
Jan/Feb 2006
Mar/Apr 2006
May/Jun 2006
Jul/Aug 2006
Sep/Oct 2006
Nov/Dec 2006
Jan/Feb 2007
Mar/Apr 2007
May/Jun 2007
Jul/Aug 2007
Sep/Oct 2007
Nov/Dec 2007
Jan/Feb 2008
Mar/Apr 2008
May/Jun 2008
Jul/Aug 2008
Sep/Oct 2008
Nov/Dec 2008
Jan/Feb 2009
Mar/Apr 2009
May/Jun 2009
Jul/Aug 2009
Sep/Oct 2009
Nov/Dec 2009
Jan/Feb 2010
Mar/Apr 2010
May/Jun 2010
Jul/Aug 2010
Sep/Oct 2010
Nov/Dec 2010
Jan/Feb 2011
Mar/Apr 2011
May/Jun 2011
Jul/Aug 2011
Sep/Oct 2011
Nov/Dec 2011
Jan/Feb 2012
Mar/Apr 2012

CBA Briefing

Courses & lectures
CBA Network
Grants & awards

CBA homepage