Milsington Leg Linked to Antoninus Pius Statue in Scotland

National Museums Scotland curators now interpret the gilded Milsington leg, found near Hawick in 1820, as the surviving fragment of a monumental equestrian statue of Emperor Antoninus Pius erected in southern Scotland in the AD 140s. The local-origin case is stronger than the old York-loot theory, but the emperor’s identity remains an attribution rather than an inscription-backed fact—and the scholarly shift began years before this week’s exhibition.

 
 

The Leg, the Horse, and the Lost Monument

The object is less ambiguous than its imperial name. It is the lower right leg and boot from a gilded, larger-than-life bronze figure. A fitting for a spur shows that the figure was mounted, while the scale and expensive surface treatment point to an elite public monument rather than a portable statuette. Its scale separates it from the smaller objects that dominate discussions of Roman artifacts: this was monumental public sculpture, not personal equipment.

The Trimontium Museum archive record gives the bronze a broad date of AD 100–180 and describes the rider as presumably an emperor or prominent military leader. The narrower date in the 140s, and the name Antoninus Pius, belong to the new curatorial interpretation. Keeping those two levels of certainty separate is essential.

The fragment was dug from boggy ground at Milsington in the Scottish Borders in 1820. The Roman Inscriptions of Britain record for a bronze base found with it says the objects remained at Dalkeith Palace until 1920. The leg is now in the National Museums Scotland collection.

It is the centerpiece of Roman Scotland: Life on the Edge of Empire, which opens at the National Museum of Scotland on November 14, 2026, and runs through April 18, 2027. The exhibition presents the fragment within a reconstruction of the lost mounted statue.

Infographic by Art in Context, based on STV News / PA Media, Proceedings of the Society of Antiquaries of Scotland, Newcastle University eTheses.

 
 

What the Evidence Supports—and What It Does Not

The curators’ case combines physical observation with a changed understanding of sculpture on Roman frontiers. No single piece of evidence identifies both the original location and the rider; the force comes from how the clues fit together.

Clue	What it supports	Limit
Milsington bog findspot	A Scottish origin is plausible	Objects can be transported
Spur fitting and monumental scale	A mounted elite or imperial figure	Not a named emperor
Top preserved as cast, without cut marks	Careful dismantling, not violent hacking	Does not identify who dismantled it
Imperial bronzes at frontier forts	Such a monument could stand in occupied Scotland	No exact Scottish site is known
AD 140s date plus Pius’s Scottish campaign	Antoninus Pius is a coherent candidate	The public archive still dates it AD 100–180
Bronze base found nearby	More than one sculptural object was deposited	Its connection to the leg is unproven

The strongest update concerns process. Current examination found no chopping at the top of the leg; the edge remains as cast. That observation favors deliberate disassembly over the dramatic picture of a limb hacked off as war booty. A wider comparison with bronze statues from frontier forts also removes the old assumption that monumental imperial sculpture was inherently out of place north of Hadrian’s Wall.

Those points make local erection credible. They do not, by themselves, prove that the statue stood at Milsington, that Romans conducted the dismantling, or that the bog deposit was merely practical disposal.

 
 

The York-Loot Theory Was Already Losing Ground

The older interpretation was not based on an inscription naming York. In 1932, archaeologist James Curle called the find “clearly enough the evidence of plunder” and argued that a gilded monument required a settled walled town rather than a volatile frontier fort. His published argument turned an expectation about where Roman art belonged into the object’s presumed biography.

The RIB 2215 database entry still says the leg and nearby bronze base “seem to have been looted from a military headquarters.” That wording records the established interpretation; it is not fresh material evidence for a journey from York. The new frontier comparisons directly challenge Curle’s premise that a large gilded statue could not have stood in Roman Scotland.

The change also predates the November 2026 exhibition. A 2022 Newcastle University doctoral thesis cited Fraser Hunter’s 2017 research for both careful dismantling and possible votive deposition in the wet ground. The thesis compared Milsington with other sculptural fragments recovered from rivers and watery sites and concluded that ritual deposition was a serious alternative.

That chronology changes what is genuinely new. The local-origin and careful-dismantling arguments have circulated in scholarship for years. The exhibition’s sharper contribution is the full reconstruction and the explicit identification of Antoninus Pius.

A surviving section of the Antonine Wall at Rough Castle near Falkirk. Photograph by Kim Traynor, Wikimedia Commons, under CC BY-SA 3.0.

 
 

Why Antoninus Pius Is Still an Attribution

Curators date the leg to the 140s, when Antoninus Pius’s armies occupied southern Scotland and pushed the frontier north. A PA Media report carried by STV says their identification rests on Pius’s particular attention to his Scottish victories, together with the date and the frontier-statue comparisons.

That is a persuasive historical fit, but it is not a named portrait. The surviving fragment has no face, and no inscription on the leg identifies the rider. The separate bronze base bears only the letter L; RIB notes that even its proposed legionary meaning is uncertain, while Trimontium’s record says the relationship between base and leg is unknown.

The surviving equestrian bronze of Marcus Aurelius in Rome gives a useful sense of the type and scale that curators are reconstructing, not a template that proves the Milsington rider’s identity.

As an artwork, the lost monument mattered through visibility. Gold over bronze would have caught light, while a naturalistically modeled mounted ruler projected a human body at superhuman scale. Within the broader language of Roman art, it was both portrait and state image. In an Iron Age visual culture where naturalistic human representation was rare, the sculpture’s foreignness would have been part of its force.

The second-century bronze equestrian statue of Marcus Aurelius in Rome, photographed by Merulana. Hosted by Wikimedia Commons under CC BY-SA 4.0.

 
 

What the Next Tests Could Resolve

Trimontium’s archive says X-ray imaging and metal analysis are planned. Imaging could clarify hidden casting joins, repairs, wall thickness, and the method used to separate the leg. Metal analysis could compare alloy composition and gilding with other Roman bronzes. Either result may strengthen or weaken parts of the proposed object history.

Neither technique can name the rider on its own. A secure identification would be much stronger if another portrait fragment, a connected inscription, or a documented erection site emerged. The location of the rest of the statue is unknown; it may have been recycled, dispersed, or remain undiscovered.

The bog context is equally unresolved. Careful removal is consistent with orderly Roman decommissioning, but a subsequent ritual deposit and a practical concealment are still possible. The physical condition of the joint answers how the leg was separated more readily than why it entered wet ground.

 

A marble bust of Emperor Antoninus Pius in military dress, photographed at the National Museum of Scotland by Osama Shukir Muhammed Amin FRCP(Glasg). Hosted by Wikimedia Commons under CC BY-SA 4.0. Cropped.

 
 

A Local Monument Is Likely; the Emperor Is Not Yet Certain

The surviving bronze securely establishes a larger-than-life gilded equestrian sculpture in the Milsington story; the spur fitting and scale do that work.

The clean cast edge and frontier parallels make a locally erected Scottish monument more convincing than George Macdonald’s 1926 York-loot scenario and Curle’s 1932 southern-plunder theory, although the precise original site remains unknown.

The public evidence supports Antoninus Pius as the curators’ best candidate, not as an inscription-proven identification. The broad AD 100–180 archive date, the uncertain bronze base, and the absence of a named fragment keep that distinction open.

The 2026 exhibition makes the attribution newly visible, but careful dismantling and local or ritual deposition had entered scholarship by 2017. X-rays, alloy analysis, and any future fragments could show how much further the reconstruction can move from plausible history to identified monument.