Danum shield

Coordinates: 53°31′31″N 1°08′02″W / 53.5254°N 1.1340°W / 53.5254; -1.1340

The Danum shield was a Roman shield found in the Danum Roman fort at Doncaster in 1971. It was discovered amid the remains of a bonfire and may have been intentionally disposed of during the partial abandonment of the fort. The shield was rectangular in shape and measured approximately 0.65 metres (2.1 ft) by 1.25 metres (4.1 ft). It is considered to have been part of the equipment of a Roman auxilium (non-citizen) soldier. An assessment in the 1970s considered it to have had an unusual vertical hand grip, suggesting a possible use by cavalry. However a recent work suggests the handgrip may have been horizontal and the shield used by an infantryman. The shield was covered in leather and was probably painted, though its original colour is not known. The outer face was decorated with bronze sheeting, possibly incised with a Celtic pattern.

The possible appearance of the Danum Shield based on Buckland's 1978 work. The original paint colour of the shield is not known

The iron boss and handgrip, together with other iron and bronze remains, are on display at the Doncaster Museum and Art Gallery. A reconstruction of the shield, made in 1978, is also on display. The exhibition is due to relocate to the new Danum Gallery, Library and Museum later in 2020.


Map showing the Minster and Danum fort outline

The Danum shield was discovered during excavations near St George's Minster, Doncaster in 1971. The excavations investigated the city's Roman fort, Danum, ahead of the construction of the Inner Ring Road.[1]: 247  Danum was constructed and occupied as early as the Vespasian era (69–79 AD) but was abandoned when the army campaigned in Caledonia (modern Scotland) on the orders of Gnaeus Julius Agricola (governor of Britannia 77–85 AD).[2]: 39 [1]: 247  The fort seems to have been demolished by the Romans in a controlled manner at the time of its abandonment.[1]: 247  After 87 AD the fort was reconstructed on the same site to house a larger garrison made up of soldiers who had by then been withdrawn from Caledonia.[2]: 39 [1]: 247  This Flavian-Hadrianic era timber fort was replaced by a stone fort in the Antonine era.[2]: 39 

St George's Minster and Inner Ring Road (right)

Excavations through one of the Antonine fort's ramparts found that much of the 2-metre (6.6 ft) rampart had been obliterated by a mediaeval cellar.[1]: 247  Evidence showed that this site lay near to one of the internal roads of the Flavian-Hadrianic fort.[2]: 20, 39  Beneath the cellar floor a layer, a few centimetres (approx one inch) thick, was found that proved to be the remains of a bonfire, possibly associated with the abandonment of the fort. The remains of the Danum shield were discovered close to the edge of the bonfire layer, lying on a north-south orientation.[1]: 247–248  The shield's iron shield boss and hand grip were first noticed, and further excavation revealed the remains of a full shield lying face down and partially turned to charcoal.[1]: 247  The remains had been partly disturbed by a trench associated with the Antonine fort and a wild animal run.[1]: 247, 253  The charcoal preserved details of three layers of wood used to form the board of the shield and also a black, vitrified, vesicular material that was thought to be the remains of a leather outer covering.[1]: 247 

The shield was photographed in situ with a gamma camera in an attempt to discern the arrangement of any metal fittings on the front face.[1]: 247  This was largely unsuccessful as the bronze fittings were too corroded to show up on the photograph and the iron fittings, which passed right through the shield, had already been mapped from the back (top) face. The shield was considered too delicate to recover by hand excavation so was lifted in a single block, encased in plaster and weighing 30 long hundredweight (1,500 kg).[1]: 248  This block was then examined at the Ancient Monuments Laboratory in London by Leo Biek.[2]: 41 


A depiction of a Roman auxilium in Britain, with rectangular shield (note this example has a curved profile and iron rim)

Fewer than ten Roman shields have been discovered by archaeologists so the Danum shield represents an important example of this type of artefact.[3] The Danum shield has been dated to the late 1st century/early 2nd century and attributed to a Roman auxiliary soldier (auxiliarius), though because of the variation in equipment among these forces this identification cannot be certain.[4][1]: 264, 269  Archaeologist Paul Buckland published his evaluation of the shield in 1978, based on evidence from the original excavation and the subsequent investigations carried out at the Ancient Monuments Laboratory.[1]: 247 

Buckland (1978)Edit

In Buckland's assessment the board of the shield, which had largely deteriorated, was broadly rectangular, with a curved top and bottom, and measured at 0.64 metres (2.1 ft) in width and a maximum of 1.25 metres (4.1 ft) in length.[1]: 248, 251  Based on the size of iron rivets used in its construction the board measured around 10 millimetres (0.39 in) in thickness.[1]: 249  The shield was flat in profile, not curved as the famous Roman scutum shield. The board was made from three layers of wood; a centre of oak with outer layers of alder joined by glue. The centre was made from six staves of wood, approximately 3.8 millimetres (0.15 in) thick, arranged in a vertical orientation, while the outer layers were of staves around 3.2 millimetres (0.13 in) thick laid in a horizontal orientation. The staves varied but measured 103–110 millimetres (4.1–4.3 in) in width.[1]: 251  The shield board was not banded in iron along its edge as some other Roman shields were. This would have made it less durable in battle against opponents with slashing weapons such as swords. It may be that the design was intentional and worked to trap an opponent's sword if it struck the shield's edge, or that the shield was intended for use only against spear-armed opponents.[1]: 258 

The domed shield boss was 200 millimetres (7.9 in) in diameter and made of 2-millimetre (0.079 in) thick iron.[1]: 249, 251  It was positioned 50 millimetres (2.0 in) vertically above the centre of the shield and affixed to it by four iron rivets driven through a 28-millimetre (1.1 in) flange.[1]: 249  The off-centre position of the boss would cause the bottom portion of the shield to tip towards the user, protecting their legs.[1]: 259  On the rear of the shield was a 0.8-metre (2.6 ft) vertical iron hand grip, secured by six iron rivets.[1]: 249  The rivets were hidden on the front of the shield by a spina, a wooden rib 28–33 millimetres (1.1–1.3 in) wide and 5 millimetres (0.20 in) thick running vertically down the centre of the face, this had no structural value.[1]: 253, 259  The handgrip showed evidence of being repaired, possibly due to breakage of the grip or because the rivets pulled out. The spina was removed, and the hand grip reattached with nails driven through the rivetholes and bent over, the spina was then reattached.[1]: 256, 263 

Microscopic fragments showed that the centre of the grip, behind the boss, was bound with leather where it would have been held by the soldier's hand.[1]: 249  During combustion the boss and the grip had sprung away from the shield board, taking some of the wood with them.[1]: 248  The short depth of the boss would have left the user's shield-holding hand vulnerable to impact from blows against the outside of the boss; it is possible that a padding of horse-hair was used here as with the Caerhun shield.[1]: 259  The boss showed evidence of an attempt being made to remove it prior to the disposal of the shield by burning, possibly to salvage it for reuse.[1]: 248  A leather thong may have been provided from the grip to the soldier's wrist to spread the weight of the shield and to prevent its loss in battle.[1]: 259  An iron ferrule was found on the handgrip and, with an eyelet affixed to the back of the shield board, may have been used to affix a leather shoulder strap for carrying the shield.[1]: 249 

There was evidence that a leather cover was affixed to the inner and outer faces of the shield, which could have been formed from a single cow hide.[1]: 251, 258  This was likely to have originally been painted, though no evidence of the original colour was found.[1]: 259  The leather covering was pierced by decorative lines of iron studs around 10 millimetres (0.39 in) in diameter and protruding from the face by 6 millimetres (0.24 in). One surviving run of studs radiated from around the boss and was probably mirrored in all four quadrants of the shield face.[1]: 251  Other studs were found in 60-millimetre (2.4 in)-long crescent positioned 70 millimetres (2.8 in) from the base of the shield. It was assumed that a similar group would have been positioned at the top of the shield.[1]: 253  These studs do not appear to have helped much with tensioning the leather cover, which would probably have had to have been glued to the board, but may have helped keep the separate wood layers bound together.[1]: 258 

Modern Roman cavalry re-enactors with rectangular shields

The rivets through the handgrip retained fragments of bronze sheeting that probably decorated the shield face.[1]: 249  Most of the bronzework seems to have been removed prior to the disposal of the shield.[1]: 256  The original pattern of this cannot be known for certain but elements can be discerned from green corrosion products left in the underlying sand.[1]: 256  The sheeting was secured by a number of 5-millimetre (0.20 in) and 8-millimetre (0.31 in) diameter bronze studs with a mixture of flat and domed heads.[1]: 253  At least four studs radiated from the boss as extensions of the lines of iron studs on the shield face.[1]: 256  The bronzework may have been inlaid with a chased pattern of curves and lines in a Celtic-style pattern, as the fragments had fractured along curved lines, and may have been used for tactical recognition on the battlefield.[1]: 259, 263 [4]

The vertical grip used on the shield is rare in Roman finds, most others having horizontal grips with the exception of a shield, virtually identical to the Danum example, recovered at Trimontium fort in the Scottish borders and one near Strasbourg, France.[1]: 263  The vertical grip would have made it awkward for used by an infantryman as it is more difficult to apply force through the shield than with a horizontal grip, as would be required in a melee. Buckland considered that the shield may have been used by a Roman auxiliary cavalryman, though noted that its weight was heavier than other known cavalry shields, which were made largely from leather.[1]: 260 

Buckland considered that the shield might not have been Roman in origin, potentially being a trophy taken from a Gallic tribe, as it bears some resemblance to examples known from European iron age tribes.[1]: 260  However, Buckland stated that the number and position of rivets on the shield probably gave it a Roman origin and it may have been brought over by an auxiliary soldier from Western continental Europe.[1]: 264  Buckland considered it likely that the shield had been disposed of in the early 2nd century AD, possibly around 105 AD when the Roman army was withdrawn from Caledonia.[1]: 269 

Travis & Travis reinterpretation (2014)Edit

The Dura shield

Hilary & John Travis discussed the Danum shield in their 2014 book Roman Shields and suggested that Buckland's findings could be open to reinterpretation. Because the handgrip had become detached during combustion of the shield it is possible that it was, in fact, originally orientated horizontally. This would have lent it better to use by infantrymen. To match the orientation of wood grain recovered from the boss and grip for this interpretation to be true the shield would have had to have been of two-ply construction, with the inner layer of wood orientated vertically and the outer layer horizontally. Travis & Travis suggest that overall thickness of the shield would still have been 10 millimetres (0.39 in) and that the two layers would have been approximately 5 millimetres (0.20 in) each, rather than the 3–3.8 millimetres (0.12–0.15 in) of Buckland's three layers.[2]: 41–42  A horizontally orientated handgrip would match the shields recovered at Faiyum Oasis and at Dura-Europos as well as six iron handgrips recovered elsewhere.[2]: 42 

Travis & Travis also suggested that the shield could have originally been of the more traditional curved rectangular shield shape and that it had flattened out after disposal in the fire. A horizontally-orientated handgrip would have been more likely to spring away from a curved shield board in a fire than a vertically orientated grip. That the shield had been curved had been the original working assumption of Buckland during the 1971 excavation.[2]: 42  Travis & Travis also suggest that if the shield had been disposed of intentionally the handgrip and boss would have been recovered and propose that the shield may have been accidentally consumed by the fire.[2]: 43 


The Doncaster Museum which held the remains of the Danum shield until 2020

The boss and handgrip from the shield were preserved, along with some iron and bronze fragments, and have been on display at the Doncaster Museum and Art Gallery.[1]: 248  Buckland made what he described as a "reasonably accurate overall reconstruction" of the shield in 1978 which has been displayed alongside it.[1]: 248, 256  The reconstruction draws inspiration from a shield depicted on the Triumphal Arch of Orange. The tabula ansata detail at the top of the bronzework decoration was based findings from the Roman camp at Vindonissa. The reconstruction weighs around 9 kilograms (20 lb) which is probably a little lighter than the original.[1]: 259 

The remains will be transferred to the Danum Gallery, Library and Museum being constructed on the site of the former Doncaster Girls' High School, whose 1910 frontage is being retained, later in 2020.[5][6]


  1. ^ a b c d e f g h i j k l m n o p q r s t u v w x y z aa ab ac ad ae af ag ah ai aj ak al am an ao ap aq ar as at au av Buckland, Paul (1978). "A First-Century Shield from Doncaster, Yorkshire". Britannia. 9: 247–269. doi:10.2307/525941. ISSN 0068-113X. JSTOR 525941.
  2. ^ a b c d e f g h i Travis, Hilary; Travis, John (2014). Roman Shields. Amberley Publishing Limited. ISBN 978-1-4456-3843-0.
  3. ^ Eriksen, Thomas Hylland (2018). An Overheated World: An Anthropological History of the Early Twenty-first Century. Routledge. p. 237. ISBN 978-1-351-72483-8.
  4. ^ a b "Roman power". Doncaster Council. Retrieved 27 June 2020.
  5. ^ "New library plans for old girls' school". BBC News. 27 June 2017. Retrieved 27 June 2020.
  6. ^ Kessen, David (17 June 2020). "Picture gallery shows stunning transformation of Doncaster town centre as major building projects progress". Doncaster Free Press. Retrieved 27 June 2020.