Until the middle of the eighteenth century, an extraordinary Roman stone building stood near Stenhousemuir beside the eastern end of the Antonine Wall. Known for centuries as Arthur’s O’on, it had survived the end of Roman rule, the formation of medieval Scotland and generations of agricultural change. Antiquaries measured it, drew it and argued about what the Romans had built it for. Then, in 1743, Sir Michael Bruce of Stenhouse had it dismantled. Its stones were taken to build a dam on the River Carron. Five years later, a flood carried much of them away.
When Sir Robert Sibbald examined the Roman antiquities of Scotland at the beginning of the eighteenth century, there was one structure near the River Carron that required little excavation or imagination. It was still standing.
Arthur’s O’on was a small circular stone building near Stenhouse, a short distance from the eastern end of the Antonine Wall. Its walls rose vertically before curving inward into a corbelled stone dome. An opening remained at the top. The entrance was narrow and the masonry substantial. Although its original Roman purpose had already been forgotten, the building itself had endured sufficiently well for eighteenth-century observers to enter, measure and draw it.
Its survival was exceptional. Roman Scotland is now encountered largely through foundations, earthworks, ditches, roads, inscriptions and objects recovered by archaeology. Arthur’s O’on belonged to another category. Here was a Roman structure surviving above ground to something close to its original height more than a millennium after the Roman army had withdrawn from northern Britain.
It survived until 1743.
A Roman building beside the Antonine frontier
The building stood in the landscape around Stenhouse and Carron, close to the eastern sector of the Antonine Wall. The wall had been constructed during the reign of Antoninus Pius after Roman forces advanced north from Hadrian’s Wall in the second century AD. Running for roughly 60 kilometres between the Firth of Forth and the Clyde, it marked the northern frontier of the Roman Empire for part of the second century.
The country around the eastern end of the frontier contained forts, roads, temporary camps and other traces of Roman military occupation. Carriden, Mumrills, Rough Castle and Camelon formed part of the wider archaeological landscape. Arthur’s O’on stood south of the wall itself, near the Roman road system and the River Carron.
Its date and purpose have never been established with complete certainty. The physical evidence that might have answered some of those questions was largely destroyed with the building. Early antiquaries variously described it as a temple, shrine or monument. Later interpretations have included the possibility that it was a Roman victory monument associated with military campaigning in northern Britain.
What can be established more securely is its form. Accounts and drawings made before demolition describe a circular building with very thick masonry walls and a domed roof constructed by gradually corbelling the stone inward. Measurements vary slightly between observers, but the structure was approximately 28 to 30 feet across externally and rose to around 22 feet.
Its appearance was unusual enough to generate its name. The Scots word “O’on” or “Oven” reflected the resemblance of the domed building to an oven. The association with Arthur belonged to a much wider medieval and early-modern tendency to attach the legendary British king’s name to ancient structures whose true origins were no longer understood.
By 1707, Scotland’s antiquaries were trying to understand it
Sir Robert Sibbald included Arthur’s O’on in his Historical Inquiries, Concerning the Roman Monuments and Antiquities in the North-part of Britain Called Scotland, published in 1707. Sibbald was a physician, geographer, natural historian and one of the founders of the Royal College of Physicians of Edinburgh. His antiquarian work belonged to an early attempt to reconstruct Scotland’s Roman past by combining classical texts with surviving remains on the ground.
He visited and described monuments at a time when archaeology had not yet developed its modern methods. Excavation was unsystematic, stratigraphy was poorly understood and dating often depended upon inscriptions, coins and classical authors. Yet Sibbald and his contemporaries were increasingly treating physical remains as evidence capable of being measured and compared.
Arthur’s O’on presented an unusual opportunity because so much of the monument remained visible. Sibbald recorded marks and worn carvings within it that he interpreted as possibly including the head and wings of an eagle and a representation of Victory. Such observations have to be treated cautiously. The building was already ancient and weathered when he saw it, and later scholars could no longer examine the surfaces for themselves.
Reports also circulated of Roman objects associated with the structure or its immediate surroundings, including an altar and what was described as a brass finger from a statue. Their precise archaeological contexts were not recorded to modern standards, limiting what can now be concluded from them.
The antiquaries nevertheless understood that the building was Roman and that its preservation was unusual.
They measured it before it disappeared
The early eighteenth century produced a surprisingly substantial record of Arthur’s O’on. In 1719, the architect Andrews Jelfe examined the monument on behalf of the English antiquary William Stukeley and prepared measured drawings. Alexander Gordon subsequently included the structure in his Itinerarium Septentrionale, published in 1726, an ambitious survey of Roman remains in northern Britain.
Gordon regarded the building as a Roman sacellum, or small sacred structure. Other interpretations followed. The monument appeared on maps and in antiquarian correspondence and became known among scholars studying Roman Britain.
Those drawings eventually acquired a value their makers could not have intended. They became evidence for a building that later generations would never be able to examine.
Modern archaeological records still depend substantially upon observations made before 1743. Dimensions, construction, appearance and some of the reported internal features survive because eighteenth-century scholars recorded them before the structure was destroyed.
The case is an early example of a problem that would repeatedly confront archaeology: documentation becoming the surviving monument after the monument itself has gone.
Sir Michael Bruce wanted the stone
Arthur’s O’on stood on the estate of Stenhouse, which came into the possession of Sir Michael Bruce. In 1743, Bruce ordered the building to be dismantled.
The reason was practical. Stone was required for a dam or weir associated with a mill on the River Carron.
The Roman masonry was available nearby. The building was deliberately taken apart and its stone transported for reuse in the new structure. A Roman monument that had remained standing for perhaps fifteen centuries was converted into construction material.
Stone reuse was hardly unusual in Scotland. Medieval castles, abbeys, churches and Roman structures were repeatedly quarried after losing their original functions. Dressed stone represented labour already performed. Where a ruin no longer possessed an economic purpose, its masonry could be more valuable than the structure it formed.
Arthur’s O’on was unusual because antiquaries had already recognised what it was before the demolition occurred.
The destruction caused outrage at the time
The loss was not discovered generations later by archaeologists wondering what had happened. Bruce’s contemporaries protested.
Sir John Clerk of Penicuik, one of Scotland’s most knowledgeable eighteenth-century antiquaries, reacted with anger. Clerk had spent years studying Roman remains and understood the unusual character of the building. In correspondence with the antiquary Roger Gale, he condemned Bruce’s decision and complained that the stone obtained from the monument could have been acquired cheaply elsewhere.
Arthur’s O’on was not destroyed in an age when nobody understood that ancient monuments possessed historical value. Some of the most knowledgeable people in Scotland understood precisely what was being lost. The disagreement reflected competing ideas about the value of old structures. There was no system of statutory monument protection capable of resolving that conflict in favour of preservation. Britain would not develop comprehensive legal protection for ancient monuments until much later. The Ancient Monuments Protection Act of 1882, almost 140 years after Arthur’s O’on was destroyed, represented the first significant national statutory attempt to protect selected prehistoric monuments. The system that eventually produced scheduled-monument protection developed through later legislation.
In 1743, antiquaries could object. They could not prevent the owner from dismantling the building.
Then the stones disappeared for a second time
The masonry taken from Arthur’s O’on did not remain permanently incorporated into the mill dam.
In the summer of 1748, about five years after the demolition, flooding on the River Carron damaged or destroyed the structure into which much of the Roman stone had been placed. Accounts of the episode report that a considerable part of the reused masonry was swept away.
The chronology is sometimes compressed into a story in which the dam was destroyed immediately after the Roman monument was demolished. The surviving accounts place the flood several years later.
The result was nevertheless severe. The monument had first been dismantled into individual stones and those stones were then dispersed again by the river.
Material that might otherwise have survived in a later wall, bridge or estate building became substantially harder to identify.
By the middle of the eighteenth century, Arthur’s O’on had therefore disappeared not only as architecture but largely as recoverable masonry.
One antiquarian family decided to build it again
The reaction of the Clerk family at Penicuik produced one of the strangest episodes in the monument’s history.
Sir John Clerk’s son, Sir James Clerk, commissioned a reconstruction of Arthur’s O’on on the family estate at Penicuik. The replica was based upon the surviving measurements and drawings of the Roman building and was incorporated into the stable complex associated with Penicuik House.
It was adapted as a dovecot.
The reconstruction survives. Historic Environment Scotland includes the structure within the Category A-listed former stable block at Penicuik House and records its design as a replica of Arthur’s O’on.
The loss made later interpretation much harder
Had Arthur’s O’on survived into the modern archaeological period, the questions surrounding it could have been approached with methods unavailable to Sibbald, Gordon, Jelfe and Clerk.
Its masonry could have been examined geologically. Construction sequences might have been identified. Mortar could potentially have provided chronological evidence. The ground around and beneath the building could have been excavated systematically. Finds could have been recorded by archaeological context rather than collected as isolated curiosities.
The relationship between the building and the Antonine Wall landscape might also have been investigated with modern survey techniques.
Instead, scholars are required to reconstruct the monument from drawings, measurements, descriptions, scattered reports of finds and the archaeological evidence surviving around its former site.
This is one reason its purpose remains debated.
The traditional description of Arthur’s O’on as a temple is plausible but not certain. Gordon’s interpretation of it as a sacellum belonged to eighteenth-century Roman scholarship. Later researchers have considered whether its architecture and position might be better understood as a commemorative or victory monument.
One possibility associates it with the Roman advance under Quintus Lollius Urbicus, governor of Britain during the reign of Antoninus Pius. Urbicus led the campaign that preceded construction of the Antonine Wall in the AD 140s. A monumental structure near the frontier could conceivably have commemorated Roman military success or imperial authority.
The evidence does not permit the building to be assigned confidently to that interpretation. The demolition removed much of what might have allowed the competing theories to be tested.
Its name survived better than its purpose
Arthur’s O’on had already acquired layers of folklore before antiquaries identified it as Roman. Its association with King Arthur belongs to a broad family of British place names in which ancient monuments became connected with legendary figures after knowledge of their original builders had disappeared.
The “O’on” element was more descriptive. The domed form resembled an oven, providing an immediately understandable name for an otherwise inexplicable stone building.
That folk name survived the rediscovery of its Roman origins. Eighteenth-century scholars could discuss Roman temples and military monuments while continuing to call the structure Arthur’s O’on.
The combination is characteristic of early antiquarian Scotland. Classical scholarship, local memory, medieval legend and physical archaeology existed alongside one another rather than replacing each other cleanly.
Stenhouse itself belonged to a changing landscape
The destruction occurred during a century in which Scottish landowners were substantially remaking the countryside. Estates were reorganised, fields enclosed, roads redirected, woodland planted, mills developed and agricultural production altered. Ancient structures could become obstacles, quarries or convenient sources of material within that process.
The Stenhouse landscape changed as well. Later estate development altered roads and approaches around the property. Industrialisation transformed the wider Carron valley even more dramatically.
Only a few years after Arthur’s O’on disappeared, the Carron Company was founded in 1759. Its ironworks became one of the most important industrial enterprises in eighteenth- and nineteenth-century Scotland, producing cannon, machinery and the short naval gun that gave the carronade its name.
The district in which a Roman monument had survived through the medieval period was entering one of the most intensive periods of industrial change in Scottish history.
The Antonine Wall itself would suffer from similar pressures. Agriculture, roads, canals, housing, industry and later urban expansion altered or destroyed sections of the frontier. Other Roman remains were excavated without modern recording or simply removed.
Arthur’s O’on disappeared shortly before that transformation accelerated.
Scotland had no machinery to save it
The history of monument protection in Scotland developed long after the antiquarian movement had begun documenting what was being lost. By the nineteenth century, archaeological and historical societies were creating increasingly systematic records, but ownership rights remained powerful and statutory protection limited.
The first Ancient Monuments Protection Act of 1882 covered only a restricted selection of monuments. Later legislation expanded state powers and eventually produced the scheduling system through which nationally important archaeological sites can now receive legal protection.
The Antonine Wall is today part of the Frontiers of the Roman Empire World Heritage Site. Surviving forts, stretches of rampart and ditch, camps and associated remains are protected and studied through archaeological and planning systems that did not exist when Bruce ordered his workmen to dismantle Arthur’s O’on.
Had the building survived another century and a half, its fate would almost certainly have been different.
The original survives largely on paper
There is no standing Roman dome at Stenhouse today. What remains of Arthur’s O’on is dispersed across several kinds of evidence.
Sibbald’s 1707 account preserves the observations of a scholar who saw the building. Jelfe’s measurements preserve its dimensions. Gordon’s Itinerarium Septentrionale records it within the early study of Roman Britain. Antiquarian correspondence records the anger surrounding its destruction. Archaeological inventories preserve the former location and later interpretations. The Penicuik reconstruction preserves an eighteenth-century attempt to reproduce its architecture.
Together they make Arthur’s O’on unusually well documented for a monument destroyed nearly three centuries ago.
They cannot provide what the building itself would have provided.
The most consequential part of the loss is not that modern Scotland lacks an impressive Roman curiosity. It is the archaeological information contained within the original masonry, foundations and surrounding ground that disappeared before scientific archaeology could examine it.
By 1743, Scotland already had scholars capable of recognising that information as part of the country’s history. They could measure the monument, draw it, correspond about it and protest against its destruction.
They could not stop its owner taking it apart.
Sources
Historical Inquiries, Concerning the Roman Monuments and Antiquities in the North-part of Britain Called Scotland
Sir Robert Sibbald — 1707
https://books.google.com/books?id=1KFMaCfRLeMC
Itinerarium Septentrionale: or, A Journey Thro’ Most of the Counties of Scotland, and Those in the North of England
Alexander Gordon — 1726
Internet Archive search: Alexander Gordon, Itinerarium Septentrionale
Britannia Romana: or Roman Antiquities in Britain
John Horsley — 1732
Internet Archive search: John Horsley, Britannia Romana
The History and Antiquities of Scotland
William Maitland — 1757
Internet Archive search: William Maitland, History and Antiquities of Scotland
The Antiquarian Repertory: A Miscellaneous Assemblage of Topography, History, Biography, Customs, and Manners, Volume III
Francis Grose, editor — 1780
Internet Archive search: Antiquarian Repertory Volume III
Memoirs of the Life of Sir John Clerk of Penicuik, Baronet
Sir John Clerk; edited by John M. Gray — Scottish History Society, 1892
Internet Archive search: Memoirs of Sir John Clerk of Penicuik
Arthur’s O’on, Stenhouse
Historic Environment Scotland / Trove — accessed September 2026
https://www.trove.scot/place/46950
Arthur’s O’on Again: Newly-Discovered Drawings by John Adair and Their Context
Iain G. Brown and Peter G. Vasey — Proceedings of the Society of Antiquaries of Scotland, Volume 119, 1989
Proceedings of the Society of Antiquaries of Scotland, Volume 119
The Antiquarian Rediscovery of the Antonine Wall
Lawrence Keppie — Society of Antiquaries of Scotland — 2012
Society of Antiquaries of Scotland publications
‘Gothicism, Ignorance and a Bad Taste’: The Destruction of Arthur’s O’on
Kenneth A. Steer — Antiquity — 1975
Cambridge Core: The Destruction of Arthur’s O’on
More Light on Arthur’s O’on
K. A. Steer — Glasgow Archaeological Journal — 1976
More Light on Arthur’s O’on
New Penicuik House (Formerly Stables), LB14635
Historic Environment Scotland — designation record, accessed September 2026
New Penicuik House designation record
Penicuik House, Stables — Arthur’s O’on Replica
Historic Environment Scotland / Trove — accessed September 2026
Penicuik House stables archaeological record
View of the Replica of Arthur’s O’on
Historic Environment Scotland Archives / Trove — accessed September 2026
Arthur’s O’on replica archival record
The History of Stirlingshire
William Nimmo; revised by Robert Gillespie — 1880
The History of Stirlingshire — Internet Archive
Round About Falkirk: With an Account of the Historical and Antiquarian Landmarks of the Counties of Stirling and Linlithgow
Robert Gillespie — 1879
Round About Falkirk — Internet Archive
Lands and Lairds of Larbert and Dunipace Parishes
John Charles Gibson — 1908
Lands and Lairds of Larbert and Dunipace — Internet Archive
