The name "Durotriges" appears on a map by Ptolemy and on building stones at Hadrian's Wall, but in no ancient narrative of a war. The image of a tribe slaughtered at Maiden Castle comes from an excavation of the 1930s, not from a text. This article separates what written sources, finds and later interpretation actually provide, and says openly where the picture stays unclear.
The Durotriges (Latin Durotriges) lived in the late Iron Age and early Roman period in what is now Dorset and in neighbouring parts of Somerset and Wiltshire, with access to the Channel coast. That is a reconstruction, not a boundary description: tribal borders are inferred from finds and from later Roman administrative districts, not handed down. No ancient author is known who travelled through Durotrigan country and described it. The tribe is known from three kinds of trace: a few lines of Greek and Latin geography and inscriptions, coins, and what archaeologists have found in the soil of Dorset.
That shapes everything that can be said about them. The Durotriges are a people whose name reached us from outside. What they called themselves, whether they saw themselves as one political unit, or whether Mediterranean authors condensed several groups into one label, cannot be established. Sentences about "the Durotriges" that credit them with attitudes, royal families or a system of alliances are, as a rule, interpretation. On this page each larger section therefore says what it rests on.
The neighbours: to the west lie the Dumnonii, to the north the Dobunni, to the east and north-east the territories of the Belgae and the Atrebates. An overview of the Britons as a whole is on The Britons, and the wider Celtic world is covered under The Celts.
The oldest traceable mention is in Ptolemy, the Greek-writing geographer of the 2nd century AD. In his description of the island of Albion he lists the Δουρότριγες in the south and gives them a town called Dunium (Δούνιον) with coordinates. He places the Dobunni, the Belgae and the Dumnonii in the same neighbourhood. In the usual editions the passage stands in the second book of the Geography, in the chapter on Albion; Stevenson's English translation, hosted on Bill Thayer's site at the University of Chicago, counts it as II.2, and Thayer himself notes there that the translation should not be used for serious purposes. The coordinates are rough approximations anyway.
Where Dunium lay is undecided. Hengistbury Head has been proposed, among other places in Dorset; scholars do not agree, and the entry should not be read as naming "the capital". Ptolemy names only one town, and nothing says it was the most important.
The second group of evidence is inscriptions, and they come from a very different time. Two building stones, catalogued as RIB 1672 and RIB 1673 in the Roman Inscriptions of Britain, name a civitas Durotrigum Lendiniensis, the "tribal district of the Durotriges of Lendinis". RIB 1672 was found at Cawfields on Hadrian's Wall; the catalogue dates it to AD 369 on contextual grounds. The stones show that a district of this name carried out building work on the Wall in the 4th century, and no more. The spelling also shows that the name was not written uniformly: Ptolemy has "-triges", the inscription "-trig-" in a genitive form.
What the name means is interpretation. A common derivation takes a Celtic duro- ("fort", or, on another reading, a word for "water") plus a second element that may mean "inhabitants"; "fort dwellers" would then be the reading. A second line of thought sees the word for "kings" in the second element. Both proposals are disputed in the literature and cannot be secured. Whether the name was already in use before the settlement of Dorset also remains open.
The most visible archaeological legacy of the area is its hilltop enclosures, called hillforts. These are sites on hilltops surrounded by bank and ditch; some are villages, some refuges, some meeting places, and scholars argue about which function predominated. From Durotrigan country Maiden Castle, Hod Hill, Hambledon Hill, Badbury Rings, Eggardon Hill, Poundbury, Ham Hill and Cadbury Castle are among the known examples. Most were built long before the Romans. For Hod Hill the current state of research gives a radiocarbon date for the main rampart of around 500 BC. Maiden Castle grew in stages: the first enclosure around 600 BC covered about 6.4 hectares, around 450 BC it was extended to roughly 19 hectares, making it the largest hillfort in Britain. Beneath it lies a Neolithic enclosure built around 4000 BC and abandoned around 3400 BC.
A common assumption holds that the Durotriges lived in hillforts and made war from them. The excavations paint a different picture: a large share of the sites were abandoned or only partly used during the last century BC; one overview gives "about 100 BC" as the point when most went out of use. There is also the surrounding landscape: excavations on Cranborne Chase and near Winterborne Kingston have uncovered farmsteads and a large late Iron Age settlement that earned the nickname "Duropolis" from Bournemouth University, which has been digging there since 2009. Life did not take place behind high ramparts alone. More useful than asking who lived in which fortress when is therefore the overall picture of an agricultural society with several settlement types. More on how such sites were built and used is on the page Castle and fortification studies.
On the coast, where the rivers Avon and Stour reach the English Channel, lies Hengistbury Head. According to the finds it was a trading place that was very active in the late Iron Age. The excavator Barry Cunliffe described it in his 1978 monograph. Finds show contacts with Armorica, that is Brittany, including Armorican coins and pottery; in addition there is a large quantity of wine amphorae from northern Italy, according to the summary of the finds more than at all other southern English sites combined, while such amphorae have so far been missing in Brittany itself. From this scholars infer that the wine did not come by way of Brittany but on a more direct route from Italy. Most of the amphorae date from before 50 BC; later forms are absent. Trade there thus declines around the time Caesar is fighting in Gaul.
Two ancient authors speak about trade with Britain, and both describe the island as a whole, not Dorset. Strabo writes in the fourth book of his Geography (4.5.2) that the island produces grain, cattle, gold, silver and iron, and that these goods are exported; he adds hides, slaves and hunting dogs. This cannot be tied to Hengistbury; the list applies to Britain overall. Caesar mentions in the third book of the Gallic War (3.8) the Veneti, a tribe on the Atlantic coast of Brittany, who have numerous ships in which they are accustomed to sail to Britain and who surpass the rest in seamanship. When they fight Caesar in 56 BC, he reports, they send for auxiliaries from Britain (3.9). That this help came from Dorset is not stated anywhere. Anyone who calls Hengistbury a "Veneti port" translates a body of finds into a tribal link that the texts do not explicitly make. More on the Veneti is on the page The Veneti.
The Durotriges are one of the tribes that issued their own coins before the conquest. The coins carry no inscription and no legible ruler's name. No king can therefore be derived from the coinage, and where the literature speaks of "Durotrigan kings", these are not attested persons. The development of the coin designs, by contrast, is clearly visible. The starting point was gold coins on Macedonian models: a laureate head on one side, a horse on the other, as on the Philip staters that were copied in Gaul and Britain over generations. In the chain of copies the images dissolve: laurel leaves become dots, the horse becomes a zigzag, a box with a circle and finally pellets. This is called abstraction, and the so-called Cranborne Chase and Starfish types are examples.
There are figures for the metal, though from an uneven body of sources. According to one overview gold was followed first by silver and billon (an alloy with a high share of base metal); silver staters are common from about 58 BC, silver striking ends about 30 BC, after which bronze is struck and later cast. This fits an observation discussed in research: the coins deteriorate steadily, which has been read as economic retrenchment and cultural isolation. Cunliffe assigned the tribe to a "periphery" around a core group in the south-east. At Hengistbury Head thousands of coins have been found, mostly Durotrigan; smelting debris and hearths point to striking on the spot. That is an indication, not proof: the mint is not secured, and the dating of individual types depends on a few dated pieces. In addition, many pieces come from metal-detector finds without secure context, which also limits dating.
Comparable coin histories, with other tribes and other images, are found for the Catuvellauni and the Iceni. The difference: there names appear on the coins, here they do not.
Durotrigan burials follow a recognisable pattern in the late Iron Age. The dead were, as far as is known, mostly laid in a crouched position in shallow oval pits, not cremated, with grave goods such as vessels and jewellery. A single woman's grave at Langton Herring in Dorset, found with a decorated bronze mirror, is radiocarbon-dated to AD 25–53. According to the state of research, the most striking grave goods occur more often in women's graves than in men's.
In January 2025 the journal Nature published a study led by Lara Cassidy (Trinity College Dublin), with excavators from Bournemouth University, analysing more than 50 genomes from burial grounds in Dorset, including some from the settlement near Winterborne Kingston. The finding: most individuals in one cemetery descend in the maternal line from a single woman who lived centuries earlier, while paternal kinship is barely detectable. The authors infer a matrilocal society, meaning that men married into their wives' communities. Land was possibly passed on through the female line. The authors themselves word this cautiously ("potentially", "possibly"): it is a single community, the comparative studies rest on smaller samples, and that women also held political power is an inference from the kinship pattern, not an observation. Miles Russell points out that Greek and Roman authors may have exaggerated the freedoms of British women.
The conquest of Britain began in AD 43 under Emperor Claudius with the general Aulus Plautius. Cassius Dio describes the course in book 60 of his Roman History: Plautius defeats Caratacus and Togodumnus, the sons of Cynobellinus (60.20.1), wins part of the "Bodunni" by capitulation (60.20.2), and then sends Flavius Vespasian and his brother Sabinus across a river (60.20.3). Dio does not mention the Durotriges. The name "Bodunni" is usually read as pointing to the Dobunni (see The Dobunni), and the struggle over the Thames, the death of Togodumnus and the arrival of Claudius fill the rest of the account.
The passage that keeps appearing in Dorset stories is in Suetonius, in the Life of Vespasian (4.1):
"Under the emperor Claudius, through the favour of Narcissus, he was sent to Germany as commander of a legion and from there transferred to Britain; there he fought thirty battles with the enemy. Two very powerful tribes, more than twenty towns and the island of Vectis, which lies nearest to Britain, he brought under Roman rule, partly under the leadership of the governor Aulus Plautius, partly under that of Claudius himself."Suetonius, Divus Vespasianus 4.1 (public domain; own translation from the Latin)
That is the whole ancient text. It names no tribe. It names no town. It also names no legion; it speaks of Germany, and of a legion without a number. The Latin oppida simply means "towns" here (Thayer's translation: "towns"), not "hillforts" or "fortified centres". The "Second Legion" does not appear in Suetonius. That Vespasian led Legio II Augusta goes back to Tacitus (Histories 3.44), where he is said to have been placed by Claudius at the head of the Second Legion in Britain and to have become famous in war. Vectis is the Isle of Wight. The "two powerful tribes" are often identified in research as the Durotriges and a second group (the Belgae or the Dumnonii are named), but that is a back-projection from the map, not a text: Suetonius gives no names.
Archaeology supports the impression of an advance to the south-west through the sites of military installations, and it dates the occupation of the region to about AD 43 to 50. A campaign of this kind does not end with a single battle but with bases: Alchester as the legion's first base, then probably Lake Farm and Bradford Abbas, small forts at Hod Hill (until about 50) and Waddon Hill (until about 60). From about AD 55 the legion was at Exeter, later at Caerleon in Wales, which is where the link to the Silures lies.
A regional link to Germany can be shown here, though only from overview literature: after the disaster in the Teutoburg Forest the Legio II Augusta was deployed in Germania and, according to its garrison list, lay from about AD 17 to 43 at Argentoratum, modern Strasbourg on the Rhine. The campaign in Dorset was thus led by soldiers who had previously been stationed on the Upper Rhine. That is a statement about a unit, not about individual men; for the stationing in detail no primary source was read for this article.
Hod Hill in the Stour valley is a place where Roman military settled directly inside an Iron Age fort. The Romans built their camp, about 200 metres across, in the north-western corner of the older enclosure. The camp was excavated in the 1950s by Ian Richmond on behalf of the British Museum; his final report appeared in 1969. Richmond read the garrison as a detachment of a legion of about 600 men and a cavalry unit of 250. Its use is usually dated to about AD 44 to 51; around 50 the troops were withdrawn for campaigns against Caratacus, to Waddon Hill.
Hod Hill became famous for iron missile heads. They are bolts of a torsion weapon known in English as a ballista: a kind of artillery that shot arrows or bolts. They were found in clusters around the so-called chieftain's hut, a large roundhouse inside the older enclosure. For a long time this was read as follows: the Romans shelled the house of the Durotrigan leader and took the fortress. The current state of research is more reserved. The bolts may also come from target practice at the later Roman camp; the capture of Hod Hill in 43 by Vespasian with the Second Legion is considered unlikely. How many bolts there were exactly is not stated in the sources read for this article, so no number is given here. On a hilltop slope it is hard to tell a siege from an exercise when the finds are not dated.
Maiden Castle, near Dorchester, was excavated between 1934 and 1937 by Mortimer and Tessa Wheeler, the first large-scale excavation of a hillfort interior in Britain. The work was largely paid for by public donations and followed by the press, and the Wheelers made the site known to a wide public. The name "Maiden Castle" first appears in 1607 as "Mayden Castell". The popular reading as "impregnable fortress" is a folk etymology; a Celtic mai-dun ("great hill") is also possible, but Richard Coates regards the name as medieval and applied to many similar places.
At the east gate Wheeler discovered a cemetery: 52 burials, among them about 14 people who had died violently. He linked one body, with a missile fragment in the vertebra, to a Roman catapult point, concluded that the Second Legion had stormed the site, and called the area a "war cemetery". The wounds, he wrote, were "the marks of battle". This picture, the last stand of the Britons against Rome, the victors with artillery, the defeated buried in haste, shaped the presentation of the conquest of Britain for decades and found its way into school books.
Wheeler's finds are real. Skeletons with sword and spear injuries lay in the ground, and nobody doubts the injuries. What is disputed is who inflicted them and when. That the picture shifted in the decades after was due to several steps: Niall Sharples's excavations and repairs of 1985/86, which included radiocarbon dating; Dave Stewart's geophysical survey of 2015, which found no Roman military camp on the site; and the renewed examination of the skeletons.
In May 2025 Martin Smith, Miles Russell and Paul Cheetham published a re-examination in the Oxford Journal of Archaeology titled "Fraught with High Tragedy". It uses new radiocarbon dates and reaches a result that contradicts Wheeler's scenario: the dead did not die in a single event but over decades and several generations, from the late first century BC into the early first century AD. The authors speak of repeated violence, possibly from local unrest, executions or dynastic quarrels before the Roman conquest, and Russell sums it up as "Britons killing Britons". A Roman massacre is not demonstrated; other atrocities of the Roman army are unaffected by this, the authors stress.
The missile fragment that Wheeler read as a ballista bolt was lodged in a person's spine. The re-examination classes it as a spearhead of local make. That is a single find, but it hit the core of the old account: the projectile that was meant to prove Roman artillery does not belong to it, by today's reading. How many skeletons carry wounds is given as "dozens" in the reports read for this article, without an exact number; Wheeler's own figures (52 burials, 14 with violent injury) come from overview accounts, not from his excavation report, which was not consulted.
Some things remain open. Wheeler excavated only part of the site, so further burials may lie in the ground. Martin Smith notes that Wheeler's view may have been shaped by the threat of invasion in his own time; that is an assessment by today's authors, not evidence of Wheeler's motives. Paul Cheetham points out that the mix of burial practices may indicate different cultures or complex social rules. The reassessment disputes a single Roman massacre; it does not dispute that violence occurred, nor that Rome acted brutally elsewhere.
After the conquest the centre of settlement shifted from the hillfort to the valley floor. Durnovaria, modern Dorchester, appears in the Antonine Itinerary, the Roman road list, and was a road junction and market place. The name is British; durn means "fist" and is interpreted as "place with fist-sized pebbles", with a possible echo of the tribal name. Durnovaria is considered the probable original centre of the civitas (the Roman administrative district of the tribe); a second district lay at Lindinis (Ilchester), as the inscription of AD 369 confirms. Roman Dorchester has town walls, an aqueduct, an earthen ring (Maumbury Rings) converted into an amphitheatre, and the only visible Roman town house in Britain, dating from the early 4th century. The aqueduct ran in a winding course of about nine kilometres to the town and was abandoned around AD 160 after the dam at Frampton failed.
Vindocladia, another place from the Itinerary, is usually equated with Badbury Rings in Dorset, where a Roman settlement lies beside the Iron Age hillfort. The Dorset heritage record writes cautiously that the site is "commonly held" to be the Vindocladia of the Itinerary. That is a widespread but unproven identification. And Poundbury, north of the town, has a hillfort with a V-shaped ditch from the 4th century BC and a large Roman-period cemetery whose burials mostly belong to the 4th century AD and were possibly Christian. When exactly the inhabitants became "Roman" cannot be pinned to a date; by about AD 70 the tribe was regarded as integrated into the province, and the Romans promoted local quarries and a pottery industry.
Behind the finds stand questions that are not conclusively answered. Why did the coinage depart so early from that of the south-east, and why did it get worse? Was the conquest of Dorset violent or negotiated, and how much of that is reflected in a cemetery? What role did Hengistbury Head play after 50 BC, when the wine trade apparently declined? How far does a DNA result from one community carry for a whole tribe? Anyone describing the Durotriges has to concede that written evidence is scarce and that the archaeology which takes its place is itself subject to interpretation. The reassessment of Maiden Castle shows that even an account regarded as secure for a century was assembled from finds and a hypothesis.
The neighbours and the story of the conquest from other directions are in the articles on the Dumnonii, Atrebates, Dobunni, Catuvellauni, Brigantes, Iceni and Silures. How Iron Age fighters were equipped is shown in Celtic warriors.
The Dumnonii · The Atrebates · The Catuvellauni · The Iceni · The Britons.
All History & Finds articles →
Three short mentions in foreign texts, a few dozen hillforts and a stronghold on the cathedral island of Głogów: that …
Read more →The Eburones were a tribe between the Rhine and the Meuse, in the area of today's Eifel, the Ardennes and the Aachen …
Read more →Most runestones tell of death. One in Uppland tells of a career. Ulf of Borresta, it says, took payment in England …
Read more →