Millbank: the grim fortress beneath Tate Britain

Place card

Place: the former Millbank Penitentiary site on the north bank of the Thames. Tate Britain occupies the river-facing centre of the historic footprint, while the Millbank Estate and buildings now used by Chelsea College of Arts cover other parts. The prison's plan still influences surrounding streets.

Address: Tate Britain, Millbank, London SW1P 4RG; the transportation memorial bollard is on the riverside near the gallery.

Nearest station: Pimlico; Westminster and Vauxhall are farther walks. River buses serve Millbank Millennium Pier, subject to current schedules.

Period: site purchased for Jeremy Bentham's proposed Panopticon in 1799; a different national penitentiary designed and built from 1812–13; first prisoners received in 1816; completed in 1821; epidemic and evacuation in 1822–23; reorganised as a transportation depot in 1843; transportation departure role ended in 1867; used as a military prison from 1870; closed and demolished around 1890; National Gallery of British Art, now Tate Britain, opened in 1897.

Themes: the Panopticon; penitentiary reform; separate and silent systems; architecture of surveillance; damp ground; disease; women prisoners; convict labour; transportation; Australia; prison records; burial; demolition; Tate Britain; memorial landscape.

Access: Tate Britain is a public gallery with changing opening arrangements, security rules and ticketed exhibitions; check its official visitor page before recording. The streets and riverside are public, but the Millbank Estate is residential and Chelsea College is an active campus. The surviving mooring stone or bollard is a memorial object, not a licence to climb on the river wall. Thames foreshore access is tidal and regulated.

The setting

Walk out of Tate Britain and turn toward the river.

Behind you is a national collection of British art: white portico, domed entrance hall, quiet galleries and carefully controlled light. In front of you the Thames moves past Westminster. Between gallery and water stands a heavy stone object linked to the landing place from which convicted people began the journey toward ships and, for many, Australia.

Now remove the gallery.

Replace it with a prison so large that Londoners compared its plan to a wheel, a flower or a fortress. Six pentagonal ranges spread around a central hexagon. Corridors repeated. Courtyards divided people into classes. A moat enclosed the outer wall. On damp ground at the edge of Westminster, more than a thousand cells formed an experiment in punishment by isolation, labour, religion and inspection.

It is tempting to call this Jeremy Bentham's Panopticon. That is only half true, and half-truth is the trap at Millbank. Bentham bought the land for his circular Inspection House, but government abandoned his scheme. The prison eventually built here was a different design—a sprawling modification influenced by inspection principles, not Bentham's exact machine.

It is tempting, too, to say that Tate Britain sits directly over every cell. The prison footprint was much larger. Housing, art school buildings, roads and the gallery share the site. Some foundations, a moat and structural remains have appeared in archaeology; most of the penitentiary is not an accessible tunnel system under the museum.

The reliable story is darker than the legend. Millbank promised humane reform in place of execution or exile. It cost an extraordinary sum, became difficult to navigate, sickened its inmates, evacuated women onto prison hulks during an epidemic and later turned into the national depot for transportation. The same river that gave the site a useful route carried people out of London under sentence.

The art gallery did not erase the prison. It changed what the site asks us to look at. Once the architecture watched prisoners. Today visitors look at pictures. The question is who remains outside the frame.

The story

Before the prison: marsh, mill and penal crisis

Late eighteenth-century Millbank was not central London as we know it. Tothill Fields and the riverside south of Westminster included marshy, poorly drained land, market gardens, scattered industry and unhealthy ground. The district took its name from a mill beside the Thames.

Britain was also confronting a punishment crisis. The criminal law threatened death for a wide range of property offences, though many sentences were commuted. Transportation had long sent convicted people to North America. The American Revolution closed that destination, and old ships converted into prison hulks became an allegedly temporary answer to overcrowding. Men laboured around dockyards while waiting for pardon, release or transportation elsewhere.

Reformers proposed penitentiaries: institutions where controlled solitude, work, cleanliness, classification and religious instruction would produce remorse and changed conduct. The word carried the promise of penitence. It did not yet mean simply any large prison.

The Penitentiary Act of 1779 had imagined national houses of hard labour, but implementation stalled. Jeremy Bentham offered an unusually complete answer—not just a building but a management system.

Bentham's eye that was never built

The Panopticon's central inspection principle originated with Jeremy's brother Samuel Bentham, who developed it while supervising workers in Russia. Jeremy adapted it into an institutional programme. Cells lined the circumference of a circular building. An inspector occupied the centre. Lighting and screens allowed the officer to see inmates while preventing them from knowing exactly when they were observed.

The uncertainty was the mechanism. A prisoner who might always be watched would internalise the guard. Bentham described an unprecedented power of mind over mind. He also promised economic efficiency, health, productive labour, detailed records and public accountability. The governor would manage under contract, with incentives intended to align order, economy and reform.

For years Bentham lobbied ministers, commissioned drawings, built models and negotiated for a site. In 1799 he purchased the boggy Millbank land for the proposed prison. Opposition to earlier sites, political delay and changing priorities defeated the project. UCL's Bentham Project emphasises a crucial fact: no prison was ever built in exact accordance with Bentham's design and management principles.

The government eventually compensated Bentham. Millbank remained available for a national penitentiary, but the building raised there was not his circular Panopticon. Calling it “Bentham's prison” erases his failure and disguises the gap between an ideal diagram and an administered institution.

Six pentagons and a building that defeated its builders

Parliament revived the national penitentiary plan in the 1810s. Architect William Williams developed the initial design; after construction problems and disputes, Thomas Hardwick took over and brought the vast complex to completion. Work began around 1812–13, prisoners arrived in 1816, and the institution was finished in 1821.

The plan was extraordinary. A central hexagonal space was surrounded by six pentagonal cell blocks, each with its own yards and internal divisions. Seen from above, the geometry resembles a flower with six thick petals or a wheel assembled from separate prisons. A perimeter wall and moat reinforced the fortress character.

This was inspection architecture, but not a single all-seeing eye. Warders in central points could observe radiating corridors and exercise yards, yet corners, repeated passages and multiple blocks created confusion. Mayhew and Binny later described the governor's house as the wheel's nave and the pentagons as forms converging around it. Visitors and staff reportedly struggled to orient themselves.

The site itself resisted. Marshy soil demanded expensive foundations. Defects and design changes drove cost to roughly half a million pounds, many times early expectations. The prison originally intended for about 1,200 people became a warning that rational geometry on paper can become wasteful, damp and disorienting in brick.

The often repeated phrase “largest prison in Europe” should be attributed rather than treated as a measured international record. What can be said securely is that Millbank was enormous: Westminster's conservation assessment gives a sixteen-acre institutional landscape, with seven acres occupied by prison buildings and yards in a later account.

The original discipline: separation, work and repentance

Millbank received men and women whose death or transportation sentences had been reduced or exchanged for imprisonment. The early regime divided a sentence into stages. During the separate period, a prisoner worked and slept alone in a cell. Religious instruction, regulated labour and controlled exercise were supposed to redirect thought. Later, selected prisoners could move into associated work.

Separation must be distinguished from the stricter “separate system” later associated with Pentonville. Millbank's design and administration changed repeatedly. Prisoners sometimes encountered one another at exercise, work machinery, worship or movement through the building. Officials attempted silence even when physical separation was incomplete.

Work included tailoring, shoemaking, weaving, laundering, cleaning and maintenance; assignments varied by sex, health and period. The point was simultaneously economic, disciplinary and moral. A clean cell and regular stitch could be presented as evidence of an ordered mind.

Religious performance became part of evaluation. Critics argued that prisoners learned to display repentance because good conduct affected classification and prospects. This does not prove every profession of faith false. It shows the danger of measuring inner reform through behaviour produced under surveillance.

Punishment for breaches included reduced diet, dark or refractory cells and corporal measures under changing rules. London Museum's Mayhew-and-Binny illustration shows a refractory cell with an outer wooden door closing over an iron grating, windows blocked to admit little light and bread-and-water rationing. The image and description refer to the mid-nineteenth-century prison, not necessarily every decade.

The epidemic of 1822–23

Millbank's grandest crisis came almost immediately after completion. In 1822 and 1823 large numbers of prisoners became ill. Contemporary investigators disagreed about diagnosis and cause. Accounts mention diarrhoeal disease or dysentery, scurvy-like symptoms, fever, poor nutrition, depression, cold, damp and inadequate ventilation. It is safer to call the event a complex epidemic than to assign one modern diagnosis to every case.

Architecture and diet could not be separated. The prison stood on wet ground. Long confinement reduced exposure to fresh air and sunlight. Monotonous or deficient food weakened bodies. Mental distress could worsen appetite and recovery. Once sickness spread, the system designed to isolate moral influence proved poorly equipped to protect health.

Parliamentary papers record medical correspondence and individual deaths. The response eventually required mass removal. In 1823 the penitentiary was evacuated and cleaned. Male prisoners went to other confinement, including hulks. A National Archives account records 167 women held on the Narcissus and Heroine hulks at Woolwich. Disease followed; most were pardoned and released by 1824.

This episode disrupts a common claim that women were never kept on prison hulks. They were not the normal hulk population, but Millbank's emergency placed women there. It also exposes the geography of the penal system: a failure in a model land prison displaced vulnerable people onto ships already associated with overcrowding and disease.

The prison reopened after changes to diet, health management and regime. Reopening did not prove that the original experiment had succeeded. It proved that government still needed the capacity.

Women at Millbank

Millbank was unusual among the institutions most often remembered in male convict narratives because it held a substantial female population during important parts of its life. Women lived in separate divisions under female officers and a regime shaped by assumptions about femininity, sexuality, domestic skill and motherhood.

Their work commonly included sewing, laundering and making prison clothing. Such labour was treated as useful training and moral discipline, but it also reproduced a narrow idea of acceptable women's work. Silence and separation could be especially severe for prisoners with children or family outside.

Records list offence, sentence, age, birthplace, physical description, health and conduct with varying completeness. They rarely preserve the full story of poverty, abusive relationships, survival economies or why a jury recommended mercy. A register can trace a woman from trial to Millbank and perhaps to a transport ship, but it cannot automatically explain her choices.

The 1823 hulk evacuation should remain a central women's story. One hundred and sixty-seven is not an atmospheric number. It represents individuals transferred from a supposedly reforming prison to vessels not designed as a normal women's institution, at a moment of illness and uncertainty.

Later Millbank continued to receive women, though its population and function changed. Mayhew and Binny's 1862 account includes a female convict division and period illustrations. Use these as observations of one administrative moment, not timeless scenes from 1816 to 1890.

From penitentiary experiment to convict depot

By the 1840s Millbank's original reformatory identity was losing ground to newer penal models. Pentonville opened in 1842 as a more systematic separate prison. In 1843 legislation renamed and reorganised the General Penitentiary as Millbank Prison, a central depot for people sentenced to transportation.

A depot is a sorting and transit institution. Convicts arrived from local prisons, were registered, examined, classified and assigned to their next destination. Some went to Australia. Others moved to hulks, public-works prisons, Pentonville or Portland. Some obtained conditional pardon, release or a changed sentence. A transportation sentence did not guarantee that a person physically sailed overseas.

The National Archives warns researchers against imagining one master file. A life may be distributed through criminal registers, court calendars, petitions, Millbank registers in HO 24, quarterly returns, hulk logs and transportation lists. Spellings vary, dates conflict and administrative transfer does not equal ocean departure.

From the river stairs prisoners could board a vessel for transfer down the Thames to a transport ship. The surviving riverside stone or bollard marks this relationship. Westminster's own heritage material describes one of two large bollards associated with the gateway through which prisoners passed before deportation to Australia.

Do not narrate every prisoner as walking directly from a Millbank cell onto a ship bound for Sydney. The route often involved intermediate boats, docks, hulks and different destinations. The bollard condenses a network into one object; the podcast should reopen that network.

Silence in the 1860s

When Henry Mayhew and John Binny visited and published in 1862, Millbank was a convict depot holding roughly a thousand men, with women in their own division. Their detailed survey describes reception, clothing, chains, cells, schoolrooms, exercise, workshops, infirmary, refractory punishment, associated rooms and the burial ground.

London Museum summarises the “silent system” they observed. A new prisoner might spend six months in a separate cell; many then entered shared rooms where they slept, ate or worked with others but were forbidden to speak. Talking could lead to punishment, including whipping under contemporary rules.

Silence was not solitude. Thirty people might occupy a room and still be denied ordinary conversation. This produced a peculiar social environment: bodies together, communication criminalised. Warders watched lips and gestures; prisoners invented covert means of exchange.

Mayhew and Binny should be read critically. They were investigators and reform-minded journalists, not recording devices. Their language sometimes reproduces Victorian hierarchies and criminal types. Yet they saw the institution in operation, questioned staff and prisoners and documented material details unavailable from architectural plans.

Their book also reveals classification's ambition. Clothing, badges, diet, work, health and behaviour created visible ranks. The prison attempted to convert a human biography into a legible administrative sequence.

Death and the prison burial ground

Millbank had a burial ground for prisoners who died in custody. Mayhew and Binny described a prison garden and churchyard; London Museum preserves an illustration of the burial ground from their book. The image shows an enclosed, managed landscape, not a Gothic pit.

The presence of graves raises modern questions about redevelopment. Demolition and later construction changed the site, and discoveries of human remains are sometimes sensationalised as a secret mass grave beneath Tate. Evidence should be tied to specific archaeological reports. A known prison cemetery is not automatically evidence of clandestine killing.

Archaeology around the former site has found portions of the moat, foundations and the concrete raft used to stabilise the prison. These traces lie across a large redevelopment area. They do not create a publicly accessible labyrinth beneath the gallery.

For a podcast, name the dead only when reliable individual records allow it. Otherwise resist turning anonymous burials into ghost inventory. The ethical subject is custody: the state controlled the person's movement, work, food, treatment and finally the ground in which the body was placed.

The end of transportation and the prison's last uses

British transportation to Australia declined in the 1850s and ended to the last receiving colony in 1868; Millbank's role as a departure point ended in 1867. Dates require care because sentencing, departures and colonial acceptance did not stop in one simultaneous act.

Without the old flow, Millbank continued in reduced and changing roles. It became a military prison from 1870 and also held civil prisoners in later phases. Its damp fabric, inefficient layout and cost made it increasingly obsolete beside newer prisons.

The institution closed in 1890 and demolition followed. Parliamentary debate in the 1890s divided the large state site among a new national art gallery, London County Council housing and institutional or military development.

This was not simply a prison replaced by Tate. The Millbank Estate rose as housing for working people, partly using materials associated with the demolished penitentiary. The Royal Army Medical College and other buildings occupied the southern area. Tate's gallery took the central river frontage and opened in 1897.

From prisoners to pictures

Sir Henry Tate funded the National Gallery of British Art, designed by Sidney R. J. Smith. Historic England describes it as the first purpose-built national gallery devoted to British art. It later became the Tate Gallery and then Tate Britain.

The transformation invites a neat line: darkness became culture. Reality is less comforting. Both prison and gallery are institutions of classification. Millbank sorted bodies by sentence, sex, conduct and health. A museum sorts objects by artist, date, school and national importance. The moral difference is enormous—artworks are not prisoners—but the spatial act of selection remains a useful, carefully limited comparison.

The old prison geometry still influences the district. Westminster's conservation area was first designated around the octagonal penitentiary site. Streets and later buildings respond to its former boundary. From above, one can read the memory more easily than from the gallery steps.

The riverside bollard, the blue plaque, archival plans, the housing estate and archaeological fragments form a distributed memorial. No single object tells the whole story.

The “Prisoner of Millbank” myth

A popular claim says “Pom” or “POME,” a term for a British person in Australia, came from initials meaning “Prisoner of Millbank” or “Prisoner of Mother England” stamped on convicts. Even Westminster material presents the Millbank version only as an allegation.

There is no sound evidence that such an acronym generated the word. Acronym origin stories are often invented long after a term appears. Treat this as folklore about transportation, not a fact to print on a caption.

The real linguistic legacy does not need the myth. Millbank entered Australian family histories through actual names in registers, sentences, ship lists and petitions. Those records are more moving than a clever but unsupported expansion of four letters.

People to remember

Samuel and Jeremy Bentham: Samuel developed the inspection principle; Jeremy turned it into the Panopticon programme and fought unsuccessfully to build it at Millbank. Their distinction prevents the story from becoming one philosopher's solitary invention.

William Williams and Thomas Hardwick: architects associated with the prison actually built. Williams's scheme and troubled construction were followed by Hardwick's completion, producing a complex that was influenced by Bentham without being his Panopticon.

The 167 women removed in 1823: transferred during the epidemic to the Narcissus and Heroine hulks. Their emergency confinement corrects the claim that prison hulks held only men.

Unnamed transported convicts: people whose journeys might run through a local gaol, Millbank, a river craft, hulk, transport ship and penal colony. “Transported” was a legal sentence and a chain of uncertain events, not one departure scene.

Prison medical staff and patients: participants in a crisis where diagnostic uncertainty, diet, damp, ventilation and mental health overlapped. Individual suffering should not be reduced to a picturesque “plague.”

Henry Mayhew and John Binny: investigators whose 1862 survey offers detailed observation of the late prison while also carrying Victorian assumptions that require critical reading.

The people buried at Millbank: prisoners who died under state custody and whose cemetery became part of a redeveloped urban site. Their graves are historical evidence, not raw material for ghost content.

Fact, interpretation and legend

Fact: Bentham acquired the Millbank site for a Panopticon, but his exact scheme was abandoned.

Interpretation: the prison eventually built borrowed inspection ideas while using six pentagonal ranges.

Legend: Tate Britain stands on a completed Jeremy Bentham Panopticon.

Fact: the prison opened in 1816 and was completed in 1821 on damp ground.

Interpretation: geometry intended to rationalise control produced navigation, construction and health problems.

Legend: one central guard could see every prisoner in every cell at all times.

Fact: a complex epidemic caused evacuation in 1823, including 167 women sent to hulks.

Interpretation: diet, damp, ventilation, confinement and multiple illnesses probably interacted.

Legend: one mysterious plague instantly killed most of Millbank.

Fact: Millbank became a depot for convicts sentenced to transportation.

Interpretation: the riverside bollard symbolises a multi-stage penal network.

Legend: every transportee walked from a Millbank cell straight onto an Australia-bound ship.

Fact: the site includes known prison burials and archaeological remains.

Interpretation: custody continued to shape prisoners after death.

Legend: public tunnels beneath Tate lead to an undisclosed mass grave.

Fact: “Prisoner of Millbank” is offered as a popular explanation of “Pom.”

Interpretation: the story expresses cultural memory of transportation.

Legend: the acronym origin is documented historical etymology.

The place today

Begin at the Millbank entrance of Tate Britain with a wide shot. Explain that the gallery occupies only part of the sixteen-acre prison landscape. Aerial plans work better than a ground-level claim that one room equals one cell.

Move to the riverside memorial bollard. Frame it with the Thames, then trace the likely sequence from prison reception to river transfer, hulk or transport ship. Read its inscription if visible, but do not climb on it or step into the carriageway. Keep the distinction between memorial and original working fabric clear.

Walk the public perimeter around Atterbury Street, John Islip Street and the Millbank Estate. Point out how the later streets and blocks respond to the former site without presenting residents' courtyards as a tourist attraction. Do not film identifiable residents through windows.

Inside Tate, follow current photography rules. The prison story does not grant special access to basements, service corridors or archives. Tate Archive material generally requires an advance Reading Rooms appointment; check rights before reproducing plans.

For the 1823 epidemic, use parliamentary correspondence or a map rather than sick-body reenactment. For the separate and silent systems, use Mayhew and Binny's illustrations with date and attribution. For graves, use the London Museum object record and avoid horror sound effects.

The Thames foreshore is tidal and can be dangerous. Do not descend unless access is lawful, conditions are understood and any required permit is held. There is no need: the strongest closing shot is the river passing the fixed stone.

Research bibliography

S0697. Westminster City Council, Millbank Conservation Area Audit — prison plan, opening, transportation role, site footprint, mooring stone and redevelopment.

S0698. English Heritage, “Millbank Prison” blue plaque — official commemoration of the 1816–1890 institution.

S0699. UCL Bentham Project, “The Panopticon” — Samuel's inspection principle, Jeremy's design, surveillance theory and the fact that no exact Panopticon was built.

S0700. UCL Transcribe Bentham, “The panopticon's failure and Bentham's anger” — Millbank purchase, boggy site, political obstruction and abandonment.

S0701. London Museum, “The punishing past of London's prisons” — Millbank chronology and Mayhew and Binny's account of separation, association and enforced silence.

S0702. The National Archives, “Prisoners and convicts” — Millbank registers in HO 24 and limits of individual prisoner research.

S0703. The National Archives, “Prison hulks” — 1823 evacuation, 167 women on the Narcissus and Heroine, disease and release.

S0704. The National Archives, “Criminal transportation” — post-American-Revolution crisis, hulks, Australia and record pathways.

S0705. Henry Mayhew and John Binny, The Criminal Prisons of London and Scenes of Prison Life (1862) — plan, regime, staffing, work, cells, women and burial ground, read as a contemporary investigation.

S0706. London Museum, “Millbank Prison” — 1816 coloured aquatint and visual evidence of the early institution.

S0707. London Museum, “Prisoner in refractory cell” — documented mid-century punishment cell, blocked light and bread-and-water ration.

S0708. London Museum, “Burial-ground at Millbank Prison” — museum record of the cemetery illustration in Mayhew and Binny.

S0709. Historic England, “Tate Gallery,” list entry 1222913 — gallery construction on the former penitentiary site and national architectural significance.

S0710. Archaeology Data Service, “Chelsea College of Art and Design: Archaeological Watching Brief” — excavated prison moat and concrete foundation raft.

S0711. Westminster City Council, “Riverside Walk Gardens” — Millbank name, transportation history, surviving bollard and present access.

S0712. Westminster City Council, Millbank Estate project history — transportation-depot context and attribution of the “Prisoner of Millbank” story as an allegation.

S0713. Tate Archive, “Memorandum of Arrangements: Millbank Prison Site and St George's Barracks” — archival evidence for allocation of the former prison land to the gallery project.