History of Birdoswald Roman Fort

Birdoswald’s history began when a wooded spur was cleared for the building of Hadrian’s Wall in AD 122. The fort, added to the Wall shortly afterwards, was garrisoned by an infantry cohort of over 800 men and remained in occupation throughout the Roman period. Its defences are the best preserved of any along the Wall. The fort was reoccupied in the Middle Ages and was the target of raids by border reivers in the 16th century.

Aerial view of Birdoswald Roman Fort, looking north from across the river Irthing
Aerial view of Birdoswald Roman Fort, looking north from across the river Irthing

Building the Wall and Fort

To the west of the river Irthing Hadrian’s Wall was at first built of turf. In this sector the regular fortlets known as milecastles were built of turf and timber, while the turrets between them were of stone.

At Birdoswald the builders of the turf Wall had to clear woodland and drain a small bog before construction could start. A ditched and palisaded camp for the builders is the earliest evidence for occupation on the site, and pieces of leather tent were found in its ditches.[1]

The fort was built astride the Wall, which, along with one of the stone turrets (no. 49a), was partially demolished to accommodate it. The fort may have begun as a turf and timber construction, but this phase was not completed and a stone fort was begun instead. At 2.14 hectares it was one of the larger forts on Hadrian’s Wall.

Find out more about Hadrian's Wall
Plan of Birdoswald in about AD 138
Plan of Birdoswald in about AD 138. The Turf Wall was replaced in stone, following a different course so that Birdoswald was completely contained behind the Wall

Completion and Initial Occupation

Work on the fort apparently stopped for a period of years at one point, but construction was completed before the end of the reign of Hadrian (AD 138) and most of the internal buildings date from this time. After the fort walls were finished, the first 5 miles of turf Wall were rebuilt in stone.[2]

We do not know what unit garrisoned Birdoswald under Hadrian, but it would have been an auxiliary unit raised from one of the peoples conquered by Rome.

The fort had been completed and a settlement had begun to develop outside the walls when Hadrian’s successor, Antoninus Pius (reigned AD 138–61), moved the army north to build another wall (the Antonine Wall) from the Forth to the Clyde. Birdoswald was certainly not given up during the brief occupation of this wall (about AD 142–60), but may have been held by a reduced ‘caretaker’ garrison.

Download a plan of Birdoswald Roman Fort
Inscription, flanked by a palm frond and a Dacian curved sword, recording the presence of the first cohort of Dacians at Birdoswald
Inscription, flanked by a palm frond and a Dacian curved sword, recording the presence of the first cohort of Dacians at Birdoswald
© Great North Museum: Hancock

Birdoswald’s Dacian Garrison

After the Antonine interlude Birdoswald was fully occupied, probably by the military unit that remained there throughout the 3rd and 4th centuries. This was cohors Prima Aelia Dacorum (the first cohort of Dacians, Hadrian’s own, 1,000 strong unit), recruited from what is now Romania.[3]

Two surviving inscriptions recording the presence of the Dacians at Birdoswald include reliefs of the typical curved sword of the Dacians (falx), together with a palm frond. One of these inscriptions commemorates the building of a granary in AD 205–8 in collaboration with a Thracian cohort and under the command of the tribune Aurelius Julianus.[4] During his time at Birdoswald Julianus lost his infant son, Aurelius Concordius, who was commemorated on a tombstone at the fort cemetery.[5] The second inscription, illustrated here, records building work at the main east gate in AD 219.[6]

The cohort is recorded in a series of 23 altars to Jupiter Optimus Maximus, five other religious inscriptions, one tombstone and one building stone.[7] The examples that can be dated span the entire 3rd century, after which the habit of making inscriptions on stone declined across the whole frontier.

A reconstruction of Birdoswald from the north-west, showing the fort and civilian settlement as they may have looked in the early 3rd century
A reconstruction of Birdoswald from the north-west, showing the fort and civilian settlement as they may have looked in the early 3rd century
© Historic England (illustration by Philip Corke)

THE 3RD AND 4TH CENTURIES

During the 3rd century repairs and rebuilding took place within the fort. Outside the walled fort the civilian settlement (vicus), which was made up of buildings with stone foundations set along streets, grew to the east and west and even to the north.[8]

South of the fort, timber buildings and a type of pottery associated with Frisia, in north-western Europe, show the presence of an irregular military unit which adhered to its traditional lifestyle.[9] Whether this was the unit named the Venatores Bannienses, or ‘Hunters of Banna’, as recorded on an altar found at the fort,[10] is not known.

The fort may have been briefly abandoned in the late 3rd century, to judge from the last Birdoswald inscription,[11] which records the rebuilding of the commander’s house that had ‘been covered in earth and fallen into ruin’. The Dacian unit, however, may have remained until the end of the 4th century, if information in the Notitia Dignitatum, a list of military and civil service posts across the empire, can be trusted – though there is doubt about its reliablity for the western empire.[12]

Read a description of Birdoswald Roman Fort
A reconstruction showing the timber building built on the site of the north granary in the 5th century
A reconstruction showing the timber building built on the site of the north granary in the 5th century
© Historic England (illustration by Philip Corke)

Post-Roman Birdoswald

One of the most important recent discoveries on Hadrian’s Wall has been that some forts continued to be occupied beyond the ‘end of Roman Britain’ in the early 5th century.

Some of the best evidence for this comes from Birdoswald where, on the site of the Roman granaries, a series of hall-type buildings was found. At first these reused the southern granary but later, when this became unviable, two phases of free-standing wooden buildings were erected on the site of the north granary.[13]

It is clear that this occupation continued without a break from the late Roman period, marking a radical change in the life of the fort after the collapse of the Roman administration and economy. The Roman military unit, already subject to late Roman local recruitment and hereditary service, perhaps became more akin to a war band, possibly even a small local chiefdom, whose members would have continued to regard themselves as ‘Roman’.[14]

This development is similar to that seen at the Roman town of Wroxeter, Shropshire.

Find out why Birdoswald matters
An Anglo-Saxon disc-headed pin from the 8th century, found to the east of the fort
An Anglo-Saxon disc-headed pin from the 8th century, found to the east of the fort

Birdoswald in the Middle Ages

Continuous occupation probably ended in the early 6th century. A stray Anglo-Saxon find of the 8th century may hint at some continued presence.[15]

Documentary references show that in the years between 1194 and 1232 Walter Beivin and his nephew ‘Radulpho de Bordeswald’ were witnessing charters and granting lands around the site and the Wall to the monks of Lanercost and Wetheral priories.[16] Birdoswald was part of the manor of Triermain, held from the barony of Gilsland. In 1295 John Gillett held land there.

John de Vaux is called ‘of Burdoswalde’ in a document of 1425[17] and there are other references to the Vaux family during the 15th century.[18] Archaeological evidence shows that a small tower house was built next to the main Roman west gate, which appears to have been in use at this time.

The farmhouse, with the excavated site in the foreground
The farmhouse, with the excavated site in the foreground

Later History

By the 16th century the west gate had collapsed, and a typical border bastle house was built within the walls of the Roman fort. Bastle houses had space for animals on the ground floor and accommodation above, acting as fortified farmhouses during the intense phase of raiding on the borders by the so-called ‘reivers’.[19]

This house was occupied by the Tweddle family, who complained to the Warden of the West March on three known occasions, twice in 1588 and again in 1590, following raids from the reiving families of Elliot, Nixon and Armstrong, when large numbers of cattle were stolen.[20] It was at this dangerous time that Birdoswald received its first antiquarian visitor, Reginald Bainbrigg, in 1599.[21]

From this time the names of owners of the site and their visitors are well recorded, and by 1849 the site was owned by Henry Norman. He created much of the Birdoswald we see today by excavating the fort walls and gates, and by building the tower on the existing farmhouse.[22]

    
About the Author

Tony Wilmott, a Senior Archaeologist at Historic England, has excavated at Birdoswald Roman Fort and is a leading authority on Hadrian’s Wall. He is the author of the English Heritage Red Guide to Birdoswald.

Buy the Guidebook to Birdoswald Roman Fort
Hadrian’s Wall running east from Birdoswald towards Harrow’s Scar
Hadrian’s Wall running east from Birdoswald towards Harrow’s Scar

Footnotes

    T Wilmott, H Cool and J Evans, ‘Excavations at the Hadrian’s Wall fort of Birdoswald (Banna), Cumbria: 1996–2000’, in Hadrian's Wall: Archaeological Research by English Heritage 1976–2000, ed T Wilmott (Swindon, 2009), 203–387 (accessed 8 Aug 2016); J McIntyre and IA Richmond, ‘Tents of the Roman army and leather from Birdoswald’, Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society, new series, 34 (1934), 62–90.
    T Wilmott, Birdoswald: Excavations of a Roman Fort on Hadrian’s Wall and its Successor Settlements: 1987–92, English Heritage Archaeological Report 14 (Swindon, 1997), 49–54.
    T Wilmott, ‘Cohors I Aelia Dacorum: a Dacian unit on Hadrian’s Wall’, Acta Musei Napocensis, 38/1 (2001), 103–22; Wilmott, op cit (1997), 195–6.
    RG Collingwood and RP Wright, The Roman Inscriptions of Britain, vol 1: Inscriptions on Stone (Oxford, 1965), no. 1909 (accessed 8 Aug 2016); now in Tullie House Museum, Carlisle.
    Ibid, no. 1919; now in the Great North Museum: Hancock, Newcastle upon Tyne.
    Ibid, no. 1914; now in the Great North Museum: Hancock, Newcastle upon Tyne.
    Ibid, nos 1872, 1874–96, 1898, 1904, 1912, 1918; RSO Tomlin, RP Wright and MWC Hassall, The Roman Inscriptions of Britain, vol 3: Inscriptions on Stone 1955–2006 (Oxford, 2009), nos 3478, 3479. Six of these were recorded in the past and are now lost. Those that survive are variously in the Tullie House Museum, Carlisle; Great North Museum: Hancock, Newcastle upon Tyne; Rokeby Park, Durham; Castlesteads, Cumbria; and built into later medieval and post-medieval buildings.
    JA Biggins and DJA Taylor, ‘A survey of the Roman fort and settlement at Birdoswald, Cumbria’, Britannia, 30 (1999), 91–110 (subscription required; accessed 8 Aug 2016); JA Biggins and DJA Taylor, ‘Geophysical survey of the vicus at Birdoswald Roman Fort, Cumbria’, Britannia, 35 (2004), 159–77 (subscription required; accessed 8 Aug 2016).
    Wilmott et al, op cit, 272–5.
    Collingwood and Wright, op cit, no. 1905; now in the Birdoswald visitor centre.
    Ibid, no. 1912.
    Bodleian Library, University of Oxford, MS Canon misc 378, fol 153v; M Kulikowski, ‘The Notitia Dignitatum as a historical source’, Historia, 49:3 (2000), 358–77.
    Wilmott, op cit (1997), 203–24.
    T Wilmott, ‘The late Roman transition at Birdoswald and on Hadrian’s Wall’, in The Late Roman Transition in the North, ed T Wilmott and P Wilson, British Archaeological Reports British Series 299 (Oxford, 2000), 13–24.
    R Cramp, ‘Anglo-Saxon pin from Birdoswald’, Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society, new series, 64 (1964), 90–93.
    JM Todd (ed), The Lanercost Cartulary (Cumbria County Record Office MS DZ/1), Surtees Society, 203 (Durham, 1997), nos i.10 and ii.168; JE Prescott, The Registry of the Priory of Wetherhal (Kendal, 1897), nos 125, 127, 128.
    THB Graham, ‘The Lanercost foundation charter, part II’, Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society, new series, 22 (1922), 24–43.
    Feodar of Gillesland and Feodar of Thomas Lord Dacre of Gillesland, Cumbria Archive Service.
    GM Fraser, The Steel Bonnets (London, 1971).
    J Bain, Calendar of Border Papers (London, 1894), vol 1, no. 356; vol 2, nos. 265, 466.
    F Haverfield, ‘Cotton Iulius F.VI. Notes on Reginald Bainbrigg of Appleby, on William Camden and on some Roman inscriptions’, Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society, new series, 11 (1911), 343–78.
    T Wilmott, Birdoswald Roman Fort: 1800 Years on Hadrian’s Wall (Stroud, 2001), 156–60.