Princess Margaret’s Secret Circle — The Lovers the Palace Paid to Silence

On June 14, 1953, twelve days after the coronation of Queen Elizabeth II, Group Captain Peter Townsend was summoned to the office of Sir Alan Lascelles, the Queen's private secretary.

Lascelles was 65 years old and had served the crown for 32 years across four monarchs. He had a reputation for cold precision. That morning, he looked at Townsend and said, "You must be either mad or bad."

Townsend recorded the meeting in his memoir, Time and Chance, published in 1978. He did not reproduce Lascelles's exact words at length. He did not describe the room or how he responded. What he described was the verdict.

A group captain with a decorated war record, twice shot down, awarded the Distinguished Service Order and the Distinguished Flying Cross and Bar, was being told by the man who ran the Queen's household that his feelings for Princess Margaret were not a private matter. They were a constitutional problem. And the response to a constitutional problem in this institution was not sympathy. It was removal.

He was posted to Brussels as air attaché. Fourteen days after that conversation, the posting was announced quietly. There was no press conference, no palace statement. Townsend simply left, the palace moved on, and coronation coverage dominated the newspapers.

For a moment, it seemed the management had worked. It had not worked.

What Lascelles's intervention in June 1953 actually started was a pattern. Not a tidy pattern, not one the palace had designed in advance. A pattern that accumulated across five decades, across four different men, and that cost the crown something it never quite got back: the ability to control what people who loved Princess Margaret were willing to say about her.

The men who came into her life between 1952 and 1981 are, taken together, a catalog of the palace's methods. Each was handled differently. Each was given a different instrument of silence.

Townsend was sent abroad. Robin Douglas-Home had his letters sought after he died at 36 with a bottle of sleeping pills and a drawer full of debts. Antony Armstrong-Jones received a title, a career, and eventually a settlement. Roddy Llewellyn received no formal arrangement that anyone has identified, and he has maintained silence for more than 40 years, which is its own kind of statement.

The word "secret" is not sensationalism. Several of these relationships were secret for years before the press found them. One never became fully public until after both parties were dead.

The word "paid" is used with deliberate precision. Not every payment was financial. Some came as continued social inclusion, decorations maintained, careers not obstructed. But each of the men who moved through Margaret's inner circle between the early 1950s and the early 1980s was offered some form of the palace's usual currency: we can make your life pleasant or difficult, and what we ask in return is your discretion.

Most took the offer. Some struggled with it. One did not survive it.

To understand what these men were walking into, it is necessary to go back to a house in Scotland and a six-year-old girl who had just been told that everything was about to change.

Margaret Rose was born at Glamis Castle in Angus, Scotland, on August 21, 1930. Her parents were the Duke and Duchess of York. Her father was Albert Frederick Arthur George, the second son of King George V, a quiet man with a stammer he had spent years training himself to manage.

Her mother was Lady Elizabeth Bowes-Lyon, daughter of the 14th Earl of Strathmore and Kinghorne. Her older sister was Elizabeth Alexandra Mary, born in April 1926, four years ahead.

By 1930, Elizabeth was already the one who understood what it meant to be the heir. Margaret did not have that clarity. She had something else. She had the run of the house.

Craig Brown's Ma'am Darling, published in 2017, describes a child who was quick, musically gifted, and aware of herself in a way her parents found both delightful and sometimes difficult. She could sing before she could reliably read. She did impressions of people she had just met and caught something true about them.

She was also the second, the spare, the one around whom the future was not arranged. For six years, this was fine. The family lived in a comfortable, almost private way. The main residence was 145 Piccadilly, an agreeable Regency townhouse rather than a palace.

Then in December 1936, Edward VIII abdicated and everything the family had arranged itself around was gone in 48 hours. Her father became King George VI. Her mother became queen. Elizabeth became heir. And Margaret became the sister of the heir, a position with enormous privileges and no formal purpose.

She was six years old. Tim Heald, in his 2007 biography Princess Margaret: A Life Unravelled, describes the effect of the abdication in terms of what it removed rather than what it gave. It removed the ordinary domestic arrangements that had made the family feel stable. It removed the father who had been genuinely available and replaced him with a king who now had to be publicly managed as well as privately loved.

After December 1936, Elizabeth was being prepared for a future that had a shape. Margaret's future had no shape at all. What she did with that shapelessness was characteristic. She turned it into performance.

Marion Crawford, the governess who later wrote The Little Princesses, the 1950 memoir that made her a pariah at court, describes a child who could hold a room with an anecdote and who understood the power of an audience before she understood very much else. The awareness was not vanity. It was survival.

The war came when she was nine. They spent most of it at Windsor. By the time the war ended in 1945, Elizabeth was 19 and being fitted into the public apparatus of an heir. Margaret was 14.

What she did have by 1945 was a father she was genuinely close to and a social life that was beginning to take shape. The parties at Windsor and Balmoral had started to include young men. She was quick and funny and beautiful, and the young men paid attention.

Then in February 1952, her father died. George VI had been ill since 1949. A cancerous lung had been removed in September 1951. He was 56 years old. On the morning of February 6, 1952, his valet found him dead in his bedroom at Sandringham.

Margaret was 21. She had been expecting this, but expecting a death and receiving it are different things. What she received that February morning was the simultaneous arrival of grief and the knowledge that her own position had just changed again.

Elizabeth was now queen. Margaret was now the Queen's sister, a distinct position from being the heir's sister. She was no longer second in the line of succession. She was a private person, more or less, with a civil list allowance, apartments in Kensington Palace, and a role nobody had written a job description for.

What she had in February 1952, apart from grief, was Peter Townsend. He had been her father's equerry since 1944, which meant he had been part of the household for eight years. He was a tall, composed RAF officer who moved through the palace with the particular calm of someone trained to perform under pressure. He was 37. She was 21.

During the king's illness, he had been present for many of the private family moments that illness creates. Townsend writes in Time and Chance that he and Margaret did not fall in love in a single moment. He describes something slower, an accumulation of time and proximity and recognition across the months following the king's death.

Anne Glenconner, a close friend of Margaret from the 1950s onward, makes clear in Lady in Waiting, her 2019 memoir, that the feeling was visible to people around them before anyone said anything. She describes a Margaret in 1952 and 1953 who was for the first time in her adult life genuinely occupied, not performing, not managing a room, occupied.

Then the coronation happened, and then a photograph was taken, and then Lascelles made his call.

The photograph is famous now. Princess Margaret on the steps of Westminster Abbey on June 2, 1953, leaning toward Peter Townsend and brushing what appeared to be a piece of lint from the lapel of his RAF uniform. The gesture was small. The intimacy of it was not.

Courtiers noticed it. Press photographers captured it. Within the week, the foreign press had begun to run stories. The Daily Mirror and the People ran them on June 14, the same day Lascelles had his conversation with Townsend.

The timing was not coincidental. The decision to post Townsend to Brussels had been made before the press ran the stories, not because of them. The palace was not reacting to coverage. It was acting in anticipation of it.

Lascelles's memo to the cabinet outlined the constitutional position with precision. Under the Royal Marriages Act of 1772, no member of the royal family could marry without the sovereign's consent before the age of 25. After 25, the family member could proceed only if Parliament did not object.

Margaret was 22. She needed the Queen's consent, and the Queen, whatever her private feelings, was being advised that granting consent would create a constitutional crisis, a Church of England crisis, and a Commonwealth crisis simultaneously.

Townsend was a divorced man. The Church of England, whose head was the sovereign, did not recognize divorce. The Archbishop of Canterbury was Geoffrey Fisher. And the Commonwealth, whose leaders had just watched the coronation of their new head, was not, in Lascelles's reading, prepared for the spectacle of the new Queen's sister marrying her divorced equerry.

Townsend left for Brussels on June 29, 1953. He was posted as air attaché to the British embassy. The role was real, but the posting was management. He later said he knew it at the time.

What he received in exchange was the continued respect of his rank, his decorations, and his status as a gentleman. Nobody called him a predator in official communications. Nobody cut him from invitation lists. He was simply moved.

The machinery of social management in this milieu does not, as a rule, destroy people openly. It relocates them quietly enough that no single action looks like punishment.

The two years of Townsend's exile in Brussels are sometimes described as a waiting period, two people counting the days until October 1955, when Margaret turned 25 and the consent requirement technically expired. This framing underestimates what those two years actually were.

Townsend was in Brussels, Margaret in London. They wrote to each other, though the full correspondence has never been published and remains in private hands. What Glenconner records is not the content of the letters but the mood they created in Margaret. She was not patient. She was not serene. She was someone doing the work of a public life while waiting for a private one, counting.

In October 1955, Townsend was permitted to return to Britain. He came back on October 12. He and Margaret spent the following 16 days seeing each other at various private locations, among them Clarence House and the home of Major John Wills in Berkshire, whose house at Allen Bay Park provided a place away from press surveillance.

During those 16 days, they had a conversation the content of which neither ever fully disclosed, and from which Margaret emerged with a decision.

On October 31, 1955, she issued a statement through the Press Association. "I would like it to be known that I have decided not to marry Group Captain Peter Townsend. I have been aware that, subject to my renouncing my rights of succession, it might have been possible for me to contract a civil marriage. But mindful of the Church's teaching that Christian marriage is indissoluble, and conscious of my duty to the Commonwealth, I have resolved to put these considerations before any others."

The statement ended in a single paragraph the most publicly discussed romance in Britain since the abdication. What most accounts get wrong is the emphasis. They frame it as a sacrifice, as Margaret giving up the love of her life for duty. That framing is not inaccurate exactly, but it is incomplete.

What the statement also ended was the palace's need to manage Peter Townsend. Once she issued it, he was no longer a problem. He was a man who could be allowed to exist comfortably, to write his memoir, to give his interviews, to live well in France, because nothing he said would change anything already decided.

He went back to Brussels. In 1959, he married Marie-Luce Jamagne, a 19-year-old Belgian woman who bore, as several observers noted at the time, a resemblance to the princess he had not been permitted to keep.

He wrote Time and Chance in 1978, 23 years after the statement. It is a scrupulous and careful book, protective of everyone named in it. He described Margaret as the most wonderful woman I have ever known. He died on June 19, 1995, in his house in the Normandy village of Le Moulinde-la-Rivière. He was 80 years old.

Margaret was still alive. She outlived the relationship by 40 years and the man who had made it by seven. She had three more men to go through first.

To understand why the palace needed to manage Margaret's relationships at all, you have to understand what kind of woman she was by her early 20s. The men who came into her life did not fall into a void. They fell into something specific.

By 1952, Margaret was 21 years old and already understood to be a problem in the particular way that brilliant women with no official purpose were problems in postwar Britain. She was not going to be the queen. She was not going to be prime minister. She was not going to run a company, because women of her class and position did not run companies in 1952.

She was going to be a princess. And a princess's function in the 1950s was to appear at the right events, marry the right person, and decline to have opinions about anything that mattered.

Margaret had opinions about everything. She had opinions about music, particularly jazz and musical theater, and they were educated opinions. She played piano to a standard that professional musicians found genuinely good. She knew the American songbook with the kind of precision that comes from actually listening.

Noel Coward, her friend since she was a teenager, told her in a letter in 1955 that her voice was effortlessly accurate in a way that most people who study singing for 10 years never achieve.

She had opinions about painting, literature, and film. She read broadly and retained what she read. People brought to dinner at Windsor or Clarence House to provide intellectual stimulus found that the younger sister was the one who had already read the book, already seen the film, already had a position she was prepared to defend across the table.

And she had opinions about men. Marion Crawford describes a teenage Margaret who assessed the young men who came to Windsor with a clarity that was sometimes uncomfortable for the young men being assessed. She was not cruel about it. She was accurate, which was worse.

The official royal family as a system is not designed for women with accurate opinions. It is designed for women who can channel what they know into the appropriate performance at the appropriate moment and otherwise maintain a composed silence. Queen Elizabeth II mastered this requirement so completely that 60 years of biographers have struggled to find the person underneath it. Margaret never mastered it and never wanted to.

What she wanted, and what the biographies and memoirs of people who knew her well make consistently clear, was a life she had chosen. Not the Kensington Palace apartment the Queen's household had arranged, not the civil list allowance that came with it, not the schedule of public engagements that arrived with her name preprinted on them.

She wanted the same thing any intelligent 22-year-old with strong preferences and a lot of energy wants: a life organized around what she actually cared about, with a person she had actually chosen to be with. The structure of her position made this almost impossible, but it did not make Peter Townsend impossible.

There is something else about the period before Townsend that gets underweighted. Her relationship with her sister. Elizabeth and Margaret were close in the way sisters who share an unusual world are close, with a closeness that contains inevitably a good deal of unspoken comparison.

Elizabeth was the one the country had organized itself around. Margaret was the one who, when she appeared in public, got the second paragraph. This was not Elizabeth's fault. It was not a dynamic either of them had chosen, but it was real and persistent, and it shaped the choices Margaret made.

When Elizabeth became queen in February 1952, the new queen's primary concern was the functioning of her role. She was 25 years old and had just inherited overnight the full administrative and ceremonial weight of the British monarchy. She had a husband with strong opinions about how a household should be run, two young children, a mother who was now the Queen Mother, and a sister who was in love with the divorced equerry.

The Queen dealt with the Townsend situation the way she dealt with most complicated human situations in this period. She referred it to the men whose job was to manage such things. Lascelles was the man. The Church was the institution. The memo was the instrument. And the memo said no.

Margaret heard the verdict through intermediaries. That detail matters. She was not called into a room and told by her sister directly. She heard it through the apparatus, which means she understood from this early point that when the family had something difficult to communicate to her, it would arrive through channels rather than across a table.

That understanding shaped every subsequent interaction she had with the palace on the subject of her personal life. She did not go to the Queen. She did not appeal to the Queen. She made her own decision in October 1955, issued her own statement, and received the result.

The result was loneliness. The specific loneliness of a woman who had spent three years loving someone she was not permitted to keep, in an institution where the official response to that kind of grief was a staff member with a composed expression and an appropriate sentence about duty.

The men who came after Townsend were in some ways responses to that loneliness. Not in a reductive psychological sense, as though she was simply seeking replacement. She was not that simple. But each of the subsequent relationships has in it something of the shape of a person who has learned that the standard arrangements on offer are things she cannot make herself want, and who is therefore looking for something outside the standard arrangement.

The years between October 1955 and May 1960 are not empty in her record. They are full. She traveled. She attended an extraordinary number of parties in London and in the country and abroad. She was photographed at nightclubs, at the theater, at private houses. She was the most photographed woman in Britain for much of this period.

She came very close to marrying Billy Wallace in 1956. Wallace was a friend, an heir to substantial money, and someone who had been in her social circle for years. She may have accepted a proposal from him or something close to it in early 1956. Then Wallace went on holiday, slept with someone else, and told her about it. Margaret ended whatever arrangement they had reached with an abruptness that friends of both parties later described as total.

There were others she allowed close before keeping a distance: John Turner, Mark Bonham Carter fleetingly. The connections were never formalized and never sustained past the point where they might have required a decision. She was waiting for someone who did not fit the approved category.

She was 30 and running out of time in the way women in her position in 1960 understood they were running out of time, when a man she had barely registered in her social circle sent her a note inviting her to dinner and everything changed.

The man was Antony Armstrong-Jones. The dinner was in February 1958 at the house of Lady Elizabeth Cavendish in Cheyne Walk, Chelsea. He was 27. She was 27. They had met before briefly at a party the previous year. Neither appears to have made much impression on the other at first meeting. At the Cheyne Walk dinner, they talked until long after the other guests had gone.

There is a version of the Armstrong-Jones story that is tidy. Boy meets girl. They are both unconventional, both creative, both drawn to the same world of art and music and bohemian London. They marry. The marriage is brilliant on the surface and privately combustible. After 18 years, it ends in the first royal divorce since Henry VIII. The tidy version treats the marriage as a misadventure. The truth is not tidy.

Antony Charles Robert Armstrong-Jones was born on March 7, 1930, in Kensington. His father, Ronald Armstrong-Jones, was a barrister who eventually became a King's Counsel. His mother was Anne Messel, daughter of Leonard Messel, whose country house at Nymans in Sussex had one of the great garden collections in Britain.

His mother later married Michael Parsons, the sixth Earl of Rosse, which gave Tony a half-brother in the Parsons family and a connection through his stepfather to the Anglo-Irish aristocracy. His parents divorced when he was five.

He was sent to Sandroyd School in Wiltshire and then to Eton. At Eton, he had a bout of polio that left him with a weakened right leg and required him to wear a built-up shoe for part of his life. He was not, by his account or the accounts of people who knew him at school, particularly interested in conventional success. He was interested in making things.

He went to Cambridge, to Jesus College, where he had gone to read natural sciences but left without completing his degree. After Cambridge, he had a brief period of uncertainty before deciding, with a confidence that turned out to be justified, that he would be a photographer.

By the late 1950s, he was one of the best photographers working in London. His portraits for Vogue and for the society press had a quality that the field's practitioners recognized and that his subjects often found uncomfortable. He did not flatter. He found something true about a face and put it on the page.

What is also clear about this period of his life is that his social world contained a significant number of gay men, and that his relationship to that world was not simply proximity. Jeremy Fry, announced as his best man when the engagement became public in February 1960, was withdrawn before the wedding with the explanation that he had a medical condition. The actual reason, which emerged later, was that Fry had a 1952 conviction for a homosexual offense. Fry and Armstrong-Jones remained close friends.

Nicholas Haslam, the designer, described his own friendship with Armstrong-Jones in terms that acknowledged the social world they shared. Craig Brown's research for Ma'am Darling is careful about this subject but does not elide it. The picture that emerges is of a man whose emotional and erotic life did not fit comfortably into the available categories and who had found in the London art and theater world of the late 1950s a circle where the categories were loose enough to accommodate him.

What Princess Margaret offered him was not a retreat from that world. She was part of it. She was the most famous person in it, but she was genuinely of it, someone whose interest in theater and photography and design was real rather than patronizing. She was also someone who understood the performance requirements of public life at a level most of the people around him did not, and who shared his preference for a private self that did not match the public one.

The courtship was conducted in serious secrecy for two years. He had a studio at 59 Rotherhithe Street, a converted warehouse on the south bank of the Thames, which could be reached by water and which the press for most of this period did not know existed. Margaret visited it. The secrecy required a level of logistical effort that, had it become public at the time, would have been a story in itself.

The engagement was announced on February 26, 1960. The press had not anticipated it. The wedding was set for May 6 at Westminster Abbey. It was the first royal wedding to be televised, and an estimated 300 million people watched it.

The marriage worked in its fashion for about eight years. What that means, in the context of two highly intelligent, highly competitive people who both needed to be the dominant presence in a room, is that it worked at the level of shared life, shared interests, shared social world. They built something real together.

The way Kensington Palace 1A was decorated, the people they had to dinner, the holidays in Mustique, where Margaret's friend Colin Tennant had established a private retreat on a Caribbean island not yet discovered by mass tourism. Their children, David and Sarah, were born in 1961 and 1964.

The photographs from this period, taken by people other than Armstrong-Jones, because he rarely photographed his own family, show two people who were often in the same room, often watching each other, sometimes with pleasure and sometimes with something sharper. From about 1967 onward, it was sharper more often than not.

The cruelty in the marriage has been documented by multiple sources. Armstrong-Jones, by various accounts, composed written insults he stored in a drawer and retrieved when he wanted to wound his wife at a specific point. One account, repeated in several biographies, has him leaving notes around the apartment containing sentences designed to destroy her on particular days.

Glenconner describes witnessing exchanges between them that she records as wounding in a way different from ordinary marital argument: calculated rather than reactive, as though both were operating from a list of each other's vulnerabilities they had spent years compiling.

The affairs on both sides were known within the circle before they were known outside it. Armstrong-Jones had his own life alongside the marriage for much of the 1970s, the most significant long-term relationship being with Lucy Lindsay-Hogg, an Irish actress and former wife of the film director Michael Lindsay-Hogg, whom he eventually married in 1997. Margaret had hers: Douglas-Home in 1967, Roddy Llewellyn from 1973.

The formal separation came on March 19, 1976. The divorce was finalized on May 24, 1978. It was, as the headlines noted, the first divorce in the senior royal family since the 16th century.

The financial settlement has never been fully disclosed. What is publicly recorded is that Armstrong-Jones retained his title, the Earldom of Snowdon, which the Queen had created for him in 1961 and which was his to keep regardless of the marriage's fate. He retained his MBE, his GCVO, and the various other decorations he had accumulated. He retained his career, which had never been dependent on the marriage for its substance but which benefited from continued access to a social world in which royal connections were a professional asset.

What the palace arranged in practical terms was the management of a man who knew a great deal about a member of the royal family and who had the profile, the wit, and the professional standing to make that knowledge extremely public if he chose to. He did not choose to.

The memoir he did not write has been noted by several journalists over the years. When asked about it in interviews, he gave answers consistent with someone who had made a private calculation about what was worth saying and what was worth keeping. He died on January 13, 2017, aged 86.

In his last years, it became known that he had fathered a daughter with Melanie Cable-Alexander, born in 1998, a relationship and a child kept from the public for nearly 20 years. The revelation added a layer to the picture of a man whose private life had always been more complicated than the official version suggested.

The palace's management of Armstrong-Jones was different in character from its management of Townsend because the circumstances were different. Townsend was removed before the relationship formalized. Armstrong-Jones was already the husband when the management became necessary, which meant the instruments available were different.

He could not be exiled to Brussels. He could be given a title and a career and a separation agreement and the understanding that the alternative to a well-managed divorce was a badly managed one, and that a badly managed divorce involving the Queen's sister would be worse for him than for the institution. He understood the calculation. He had been making calculations of that kind his entire professional life.

During the marriage, while Tony had his long-term involvement with Lucy Lindsay-Hogg, and while Margaret's friendships with various men in her circle were in various stages of seriousness, there was a man named Robin Douglas-Home. He was not, by most measures, the most significant man in Margaret's life. He was instead the man who died for it.

Robin Archibald Douglas-Home was born on September 1, 1932. He was the son of Henry Montagu Douglas-Home and Margaret Ogilvy, and the nephew on his father's side of Alexander Frederick Douglas-Home, who would become prime minister in October 1963 and who is remembered as the 14th Earl of Home and as the last prime minister to serve in the House of Lords before taking a disclaimed peerage to lead the country from the Commons.

Robin Douglas-Home did not have his uncle's political aptitude or his uncle's comfort with institutional life. What he had was a musical gift, a social fluency that came naturally rather than strategically, and a capacity for spending money he did not possess on a scale that accumulated debts faster than his income could address.

By the time he came into Margaret's orbit in 1967, he owed approximately £36,000, which in 1967 was a serious sum. He had been married once to Sandra Paul, who later married Michael Howard, the Conservative politician, and who has maintained for decades a dignified silence about the Douglas-Home years. They divorced in 1962.

By 1967, he was living in London, playing piano at private events, writing occasional journalism, and circulating in the society Margaret moved through. He had been around the edges of her world for some years before 1967. The friendship became something else during 1967, when the Armstrong-Jones marriage had settled into its Cold War phase and Margaret was spending increasing time in the company of people who were not her husband.

The love letters from this period are, by the accounts of the few people who have seen them or been told their content, among the most unguarded documents Margaret ever produced. She was 37 years old. She was writing to a man she trusted with something that, in the context of her position, was an extraordinary act of private expression.

She wrote about what her life felt like from the inside. She wrote about the protocol and the pressure and the permanent sense of being managed. She wrote in at least one passage about feeling as though her actual self existed somewhere the institution could not reach, and that Robin was the person she could send it to.

These letters were hers in the sense that she wrote them. They were also, in another sense, the property of a woman whose correspondence was understood by everyone in her circle to carry constitutional weight. What she put in writing was never entirely private, because the institution she was part of did not make full provision for private.

In 1967 or early 1968, Robin Douglas-Home approached Neue Revue, a German illustrated magazine, with an offer. He had material, he said, that would be of interest. He was understood to be offering either the letters themselves or information derived from them.

The exact nature of the approach has been described differently by different sources. What is consistent across the accounts is that the British establishment learned of it, and that the learning of it had consequences.

Whether the palace's response to Douglas-Home's approach contributed to the final state he was in when he died on October 15, 1968, is something no inquest and no subsequent investigation has definitively established. What is established is the sequence: the approach to Neue Revue, the establishment response, the continued deterioration of his finances, the debts at £36,000, the flat in Berkshire, the bottle of sleeping pills, the open verdict at the inquest. He was 36 years old.

There was no formal relationship to manage afterward because the relationship had never been public. There was no wife to whom arrangements needed to be explained because he was unmarried when he died. There were the letters, and the letters required attention.

The management of those letters after Douglas-Home's death has never been fully documented. What is known is that some material was acquired by parties with an interest in its suppression, and that some material survived in private hands and has been referenced by researchers and journalists without being published in full.

The nature of the Royal Archive, administered at Windsor Castle, means that documents touching on matters of this kind can be effectively restricted from public view for periods long enough to outlast the people most directly concerned.

Craig Brown was given access to more material on Margaret's life than most writers have in preparation for Ma'am Darling, and his account is the most careful and complete available. He does not publish the letters. He does not claim to have read them. He describes their existence and their general character from secondary sources, which is itself a careful navigational choice of the kind that writers working on royal subjects learn to make.

What remains is not the letters but the shape they leave. A woman who was, by all accounts of people who knew her in the late 1960s, in genuine distress about the state of her marriage and her life, found someone she could write to without the performance. The performance was what the institution required. The letters were the alternative to it. And when the alternative was threatened, the institution moved, as it always moved, to contain the exposure.

There is one more dimension to this story that has been largely absent from the accounts: the piano playing. He was described by people who heard him as genuinely talented, not professional concert level, but better than good in the way that people who have spent years making music seriously are better than good. He played jazz and popular songs and classical pieces by Chopin and Satie with the touch of someone who had been playing since childhood.

It is a small detail, but it matters, because it is part of what Margaret recognized in him. She was a serious pianist herself. The musical conversations they could have were real conversations, not the performances that social conversation about art so often becomes. When she described the relationship to Anne Glenconner, she described it in terms of connection rather than passion. He understood what she was actually saying. He could hear what she was actually playing.

The cruelty of what happened to him is not the palace's cruelty exactly, because the palace did not tell him to accumulate £36,000 in debts, and the palace did not tell him to approach a German magazine. The cruelty is the structure. The structure of a world in which a person could love a princess genuinely and find themselves at 36 in a room alone with an outcome that felt like the only available exit.

In August 1973, Colin Tennant held a house party at Glen, his family estate in Peeblesshire in the Scottish Borders. Tennant was the third Baron Glenconner, one of Margaret's oldest and closest friends, and the owner of the island of Mustique in the Caribbean, where he had been building villas and inviting people he found interesting since the late 1950s.

He had given Margaret a plot of land on Mustique as a wedding present in 1960, and by 1973 she had a villa there, Les Jolies Eaux, where she went for several weeks each year and where the press, in that pre-satellite era, could not easily follow.

The Glen house party in August 1973 brought together a collection of Tennant's friends that included several people in their 20s and 30s whose connection to the aristocracy was more social than ancestral. One of them was Roddy Llewellyn, 25 years old, the son of Sir Harry Llewellyn, the Olympic showjumper who had won gold at the 1952 Helsinki Games on his horse Foxhunter and who was one of the most celebrated sporting figures in postwar Wales.

Roddy Llewellyn by 1973 had not yet found what he was for. He had a good family, a good education at Shrewsbury, and an interest in gardens that had not yet crystallized into a profession. He was pleasant and gentle and possessed, by the accounts of everyone who knew him at the time, of the kind of manner that makes people feel attended to.

He was pretty and he was kind and he was 25 years old, and he was almost nothing like Peter Townsend, which may be part of what Margaret found when she looked at him across a room in Peeblesshire that August. She was 43. He was 25.

Glenconner describes the beginning of the friendship with a care that protects both parties without being evasive about what it was. He had an ease about him that the men in Margaret's life since Townsend had largely not had. Townsend had ease of a military kind, the composed calm of someone trained to function under pressure. Armstrong-Jones had a different kind of ease, the controlled assurance of a man who understood how he appeared in a room and managed the appearance.

Roddy Llewellyn had something older than either: an unstudied goodness that is rarer than either composure or assurance, and that is almost impossible to fake. Margaret recognized it. She had been surrounded for the better part of 20 years by people who were performing in one way or another. Roddy Llewellyn, in a Scottish house in August 1973, was simply someone who was glad to be where he was and who found her funny. She found this startling. She found it, if the accounts of people around her during this period are to be credited, something close to restorative.

The relationship that developed between 1973 and 1976 was conducted on Mustique, in London private houses, and at various country locations, with the assistance of the same network of friends who had been helping Margaret arrange private time for the previous 20 years. Glenconner was the primary facilitator, because Mustique was his island and the villa was his arrangement and the people he invited there could be trusted in the way that people who have been inside an aristocratic network for decades are trusted to keep what they saw to themselves.

What made Mustique work as a location was not simply its remoteness. It was a private island owned by a private individual who controlled who came and went. The British press in 1973, 1974, and 1975 could not land a correspondent at Mustique without Tennant's agreement, and Tennant did not agree. The villa Les Jolies Eaux was set on a hillside above a beach called Gelliceaux, and the beach was not accessible from the village. It was, in practical terms, as close to invisible as a royal residence could be.

This worked until February 1976, when it stopped working.

The Sunday People is a British tabloid newspaper. On February 22, 1976, it published a photograph on its front page. The photograph showed Princess Margaret and Roddy Llewellyn on the beach at Mustique. They were in swimwear. They were not touching, but they were together, and the context of the photograph and the caption made clear that they were together in a way that was not coincidental.

The photograph had been taken by a reporter who had arrived at Mustique by boat without Tennant's knowledge or permission. The paper ran the story with a headline whose essential content was the identification of the relationship and the identification of Llewellyn, including his age, which at 28 was 15 years younger than Margaret at 45.

The response was immediate. Within 48 hours, the story had been picked up by every major newspaper in Britain and by international press across Europe and North America. The specific character of the story, a separated princess in an apparent romantic relationship with a man 17 years her junior, was exactly the material the British tabloid press of the mid-1970s was structured to amplify.

Armstrong-Jones was in Nepal when the photograph appeared, working on a photographic project. He made a brief statement. The palace made a brief statement. Neither statement was illuminating, which is exactly what brief palace statements are designed not to be.

Willie Hamilton, the Labour MP for Fife Central, who had been the most consistent parliamentary critic of the civil list for years, made a statement that was rather more illuminating. He called for Margaret's civil list allocation, which was £55,000 per year, to be reviewed in light of what the photograph suggested about how it was being spent. Hamilton's criticism found an audience beyond his usual one. The photograph had arrived at a politically sensitive moment, when public patience with the cost of the monarchy was genuinely lower than it had been, and Hamilton understood how to work that moment.

The formal separation from Armstrong-Jones was announced in mid-March 1976, three weeks after the photograph. The timing has attracted comment from biographers and journalists ever since, because the question of which came first, the decision to separate or the photograph that made a continued fiction of the marriage impossible, has never been definitively answered. What is clear is that the two events arrived in rapid succession, and that the photograph accelerated at minimum the timing of the announcement.

Roddy Llewellyn's response to the exposure was characteristic of everything that was said about his character by the people who knew him. He did not sell a story. He did not give a press conference. He did not attempt to leverage what he knew into anything.

He went through a period of considerable personal difficulty in the years immediately following 1976, spending time at a place called Serendell Farm in Wiltshire, a smallholding run by a group of people attempting, with varying success, the kind of communal rural life that attracted certain people in mid-1970s Britain. He was not happy at Serendell Farm. He was not suited to it. He left.

What he found in the late 1970s was gardens. He trained at the Chelsea Physic Garden, the four-acre botanical garden on the Embankment that has been teaching horticulture in one form or another since 1673. He became a landscape garden designer, and he became a good one, building a practice that in the 1980s and 1990s was genuinely successful in the commercial world of private garden design.

He married Tatiana Soskin on July 9, 1981, a few months after his relationship with Margaret had come to its natural end. They had two daughters. He was, by the accounts of people who knew him in his professional and personal life after 1981, happy in a way that was accessible and calm and not contingent on anything the palace or the press could give or take away.

He attended Margaret's funeral at Windsor Castle on February 15, 2002. He was photographed arriving. He made no statement to the press. In the years between 1976 and 2002, he gave a handful of interviews about the relationship, always measured, always warm, never exploitative. He described her as someone he loved and whose company was genuinely one of the great pleasures of his life. He said she had the best laugh of anyone he had ever met.

He has not written a memoir. He has not sold his archive. He has not spoken at length about the relationship for many years. Whether this sustained silence reflects a private calculation, a continuing affection for Margaret's memory, a natural preference for a life outside the spotlight, or some combination of all three, is not something anyone outside his household can say with certainty.

What it is not is the palace's doing, at least not in any documented way. The silence of Roddy Llewellyn is the one in this catalog that does not appear to have a financial or institutional arrangement behind it. He simply chose to be quiet. In a story full of people who were paid or pressured or managed into discretion, he appears to be the one who arrived at it himself.

The four men in this account were handled by four different instruments, and the differences are revealing. Exile for Townsend. Pressure and poverty and death in a sequence that cannot be fully documented but cannot be fully dismissed for Douglas-Home. The gilded cage for Armstrong-Jones. And for Llewellyn, nothing formal, or nothing documented, and the voluntary silence of a man who appears to have concluded on his own that discretion was who he was.

Four instruments, four outcomes, all four effective.

To understand why the palace had four different instruments rather than one, you have to understand what the institution is actually managing when it manages these situations. It is not managing scandal, not primarily. Scandal is the word the press uses. What the institution is managing is institutional legitimacy, which is more specific than scandal and more fragile.

The British monarchy in the 20th century operated on an unusual compact. The country granted the institution an extraordinary level of privilege, social, financial, constitutional, and in return the institution maintained the appearance of a certain kind of life. The appearance was not of perfection. Nobody by the 1950s expected the royal family to be perfect. The appearance was of seriousness, of a family that understood the weight of what it was holding and carried it with the appropriate gravity.

Margaret was the fracture point in this compact, not because she was immoral or even particularly reckless, but because she was honest. Her life was not a life she could make herself perform without the performance becoming visible. And when the performance became visible, the institution's legitimacy was exposed, not by what she did, but by what her presence did to the logic of what the institution claimed to be.

The Royal Marriages Act of 1772, the formal instrument that governed Margaret's situation with Townsend, was itself a measure of how seriously the institution took the management of these situations. George III had passed it because several of his brothers had contracted marriages he found unsuitable, and the act gave the sovereign the power to prevent any member of the royal family from marrying anyone the sovereign did not approve of, at least until they turned 25. It remained in force, with modifications, until 2013.

But the act was only the formal instrument. The informal instruments were older and more efficient. The informal instrument was what biographer Theo Aronson, in his 1997 study of the royal family, called the machinery of discouragement: the combination of architectural, social, financial, and reputational pressure that the institution could apply to anyone who came too close to one of its members in a way it considered inconvenient.

Townsend received the architectural and geographical variant: Brussels, and then Johannesburg. The machinery moved him to where the problem couldn't develop.

Douglas-Home received something that looks from the outside less deliberate, but may not have been. He was a man with serious debts and a specific thing the institution wanted him not to do. When he approached the German press with an offer, the establishment response came quickly. What the response consisted of exactly has never been fully documented. What is documented is that he did not proceed with the sale, that his financial situation did not improve, and that he died with debts and sleeping pills a year after the approach.

Armstrong-Jones received the financial and social variant: a settlement that was adequate, a title that was irreversible, and continued access to the world that had made his career. The machine gave him a reason to stay quiet by making quiet the rational choice.

Llewellyn received apparently nothing formal, and yet he has been quiet for 40 years.

The mechanism underneath all four cases is the same, and it is older than the 20th century. It is the mechanism of social credit. Every person who lives inside a certain world, whether that world is the British aristocracy or a professional community or a small town, accumulates credit in that world through their behavior, their associations, and their adherence to the world's internal codes. The credit can be spent or maintained, and the choice of whether to spend it is constrained by the knowledge of what happens to people who spend it in ways the world does not endorse.

The specific social credit available to people who move through the world of the British royal family is not small. The invitations, the access, the professional associations, the sense of belonging to a set that considers itself the most significant set in the country: all of it is available to those who maintain the codes. All of it disappears for those who violate them.

Townsend understood this. He was a military man, which meant he had spent his career in an institution whose codes were explicit and whose sanctions for violating them were formal. The social machinery of the palace was by comparison gentler in its execution and entirely without formal rules. But the logic was identical. You maintained the code or you lost the world. He maintained the code. He lost, in one sense, the woman. He kept the world.

Armstrong-Jones understood it with even greater sophistication, because he had come from outside the world and had watched himself enter it with the analytical eye of a photographer who knew that the image of a thing and the thing itself were different categories of reality. He understood its rules not as nature but as convention. And the understanding of convention rather than nature is precisely what allows a person to work within convention strategically.

Llewellyn may have understood something simpler: that the alternative to silence was the destruction of the relationship's meaning. To speak about it commercially, to sell the interviews and the archive and the memoir, was to turn it into something it was not. The relationship had been private. The privacy had been the point. Selling the privacy was selling the thing that made it worth having.

This is not a naive position. It is a sophisticated one, and it may be the most complete explanation for his silence, because it does not require the assumption of any arrangement with the palace at all. It requires only that he was the person everyone who knew him described: a man of genuine feeling and genuine restraint, who understood that the value of a thing and the value of its public account are not the same.

The palace did not create the conditions that made these men silent by itself. It used conditions that already existed in the British class and social system. The deference to the institution, the understanding that access to the most significant social world in the country was conditional on not embarrassing it, the willingness of individuals to accept the institution's management of their private lives in exchange for the benefits of proximity: these conditions were not invented by the palace. They were inherited by it from a social structure that had been building them for centuries.

The four men in Margaret's circle were not victims of a conspiracy exactly. They were people who understood, with varying degrees of clarity, what world they were in, and they made the calculations that world made available to them. Townsend chose dignity over exposure. Armstrong-Jones chose professional continuation over a bitter divorce. Llewellyn chose privacy over money. Douglas-Home tried in desperation to convert what he knew into the money he needed, and the world made clear that this was the one conversion it would not permit.

The result, across 50 years of Margaret's adult life, was an almost complete absence of the male perspective on what being loved by her actually involved. The memoirs that exist are careful. The interviews that were given are measured. The letters are mostly inaccessible. The man who might have been least careful, who might have said the most, is the man who did not survive to say anything.

What we have instead are the accounts of the women who were closest to her: Anne Glenconner, whose Lady in Waiting is the most valuable primary source available; the women who served as ladies-in-waiting and who wrote or spoke after her death; the biographers who assembled the record from what was left.

What we do not have in any complete or unmanaged form is what these men actually experienced. The palace managed the men. The men managed themselves. And what remains is an outline of a life as it was understood by the institution that contained it, not the life as she actually lived it.

By the late 1970s, Princess Margaret was 48 years old, divorced, and in a relationship with a man who was 30 years old, and she understood that the circle she was in was closing. Not because of Roddy Llewellyn, who remained genuinely in her life until the early 1980s, but because of the accumulating weight of the institution's management of everything she had tried to be.

There is a tendency in accounts of her later life to frame the 1970s and 1980s as a period of decline. The parties that ran too late, the drinking that was real, the cigarettes, 100 a day at her worst, which her doctors told her would kill her. The holidays in Mustique, which the press covered as the holidays of someone avoiding the country rather than as the holidays of someone doing the only thing that felt like escape from the machinery.

The decline narrative is not wrong about the symptoms. It is wrong about the diagnosis. What was actually happening to Margaret in the 1970s and 1980s was not a personal disintegration. It was the result of 50 years of being managed by an institution that had never given her a function, only constraints.

She had been told at 22 that she could not marry the man she loved. She had been given at 29 a marriage that the institution approved because Armstrong-Jones was, whatever else he was, not a divorced commoner. She had been watched through 18 years of that marriage, had her affairs noted and her movements monitored and her spending tracked, and had been at the end of it the subject of the first royal divorce in 400 years, which the same institution that had prevented her first choice now found itself managing as a public relations problem.

She had then found in Roddy Llewellyn something that felt like the private life she had been trying to locate since 1952, and the press had published a photograph of it.

The consequence of all this was not a party that ran too late or a drink that was one more than was wise. The consequence was a woman who had been fighting the institution from the inside for 40 years, and who had in the late 1970s run out of positions from which to fight.

Glenconner, who was with her for much of this period, writes about it with a compassion that does not romanticize it. She describes Margaret in the late 1970s as someone who could still be the most brilliant company in any room she entered, and who was outside the rooms struggling with a life she had never chosen and that the institution was not going to change because she was struggling with it.

The smoking killed her slowly. The first stroke came in January 1998 at Les Jolies Eaux in Mustique. She was 67 years old. The stroke affected her vision in one eye but left her otherwise largely functional, and she returned to London and continued her schedule of engagements with the stoicism the family by then required as a minimum. The word for it in the official communiqués was "minor." It was not minor. It was the first in a series of strokes that would, across the following four years, reduce her from a woman whose physical presence could still command a room to a woman who could barely navigate it.

In 1999, she scalded her feet in a bath at Les Jolies Eaux. The scalding occurred because the stroke had affected her sensitivity to temperature and she did not register the water as too hot until the damage was done. The burns required treatment and left her with permanent skin damage. Her feet, which had been a private vanity of hers, something she had taken care of throughout her adult life, were burned. The detail is small and not small.

Further strokes followed in 2001 and early 2002. By the last year of her life, she had lost much of the use of her left hand. Her vision had deteriorated substantially. She could no longer play the piano, which was a loss she did not discuss in public, but which the people who knew her understood to be among the most significant things the strokes had taken.

She died at the King Edward VII Hospital in London on the morning of February 9, 2002. She was 71 years old. The Queen was at Sandringham. The family was notified. The announcement was made by Buckingham Palace at noon with the standard language of royal deaths.

Her funeral was held at St. George's Chapel, Windsor, on February 15, 2002. It was a private funeral, not the ceremonial state funeral that would follow her mother's death seven weeks later. But it was attended by the family and by the people who had been her actual circle across 50 years, which included Colin Tennant and Anne Glenconner and the household who had served her.

Roddy Llewellyn attended. He was photographed arriving. Peter Townsend was dead by then. He had died in 1995 in Normandy. Antony Armstrong-Jones, now the Earl of Snowdon, did not attend the funeral. This has been noted but not fully explained. The official reason given was that his attendance might attract disproportionate media attention on a day intended to focus on the family.

The accounting of what her managed life cost her is not in the end a financial one. The financial costs to the palace were real: the civil list allocation, the Kensington Palace apartments maintained across decades, the staff, the security, the management of various situations that required various instruments. These are documented in aggregate, if not in detail.

The accounting that matters is harder to quantify. It is the accounting of a life that contained in its first 50 years one man she could not have, and two marriages to men the institution approved and found workable for its own reasons, and the accounting of what that pattern produced in the person who lived it.

Craig Brown, in Ma'am Darling, refuses to reduce her to either a victim or a villain, and his refusal is the most accurate position available. She was not a victim of the institution in any simple sense, because she was also deeply embedded in the institution's privileges and genuinely attached to many of its elements. She loved being a princess. She loved the access and the attention and the rooms it opened. She did not want to give those rooms up. She wanted to have both the rooms and the life inside them.

And the institution's specific cruelty was not that it denied her the rooms, but that it made the life inside them conditional on a managed performance that ate across 50 years the private person it was managing.

What the private person was underneath the performance is what the letters that nobody has fully published would show us. The letters to Townsend, the letters to Douglas-Home, the notes and diary entries and private correspondence that sit in the Royal Archive and in private hands and in the estates of people who were in her circle and who have not yet decided what to do with what they hold.

We do not have those documents. What we have is what survives the management: the public record, the managed memoir, the measured interview. We have the shape of what was silenced, even if we do not have the sound of it.

The sound was a woman who could sing before she could read, who did impressions that made her father laugh, who played piano in a way that professional musicians found genuinely good, who understood which gap to find between what a person said and what they meant, who looked at every room she entered with eyes that were taking the full inventory of everything it contained, and who spent seven decades inside an institution that found all of those qualities easier to manage than to accommodate.

She died on a Saturday morning in a hospital in Marylebone. She was the last person to remember the abdication from the inside of the family. She had been watching since she was six years old.

The Royal Archive at Windsor Castle is one of the largest and most carefully maintained private archives in Britain. It contains documents from the reign of George III onward: official correspondence, private letters, diaries, photographs, files, memoranda, the accumulated paper of a family that has been producing paper of institutional and historical significance for more than 200 years.

Access to the archive is controlled by the Keeper of the Royal Archives, who is appointed by the sovereign and answers to the sovereign's household. The archive is not a public institution. It is not subject to the Freedom of Information Act. Documents held within it are not automatically released after a set number of years the way cabinet papers and other government documents are released under the 30-year or 20-year rules. The archive's own rules determine what is accessible to researchers, and those rules are the rules the archive makes for itself.

This matters for the story of Princess Margaret's private life, because the documents that bear most directly on her relationships, particularly the correspondence with Townsend and the letters written in the Douglas-Home period, are documents that, if held in a public archive, would have been available to researchers for decades.

The Royal Marriages Act documents held in the Cabinet Office have been released. The Lascelles memo is in the National Archives. The internal correspondence of the Archbishop of Canterbury's office from 1955 has been partially accessible to researchers through Lambeth Palace Library. But the private correspondence, the letters that passed between Margaret and the men she loved, is a different category.

The letters Townsend sent her are presumably part of her personal papers, which passed to her estate at her death in 2002. The management of her estate has not been publicly documented in a way that identifies the disposition of correspondence. The letters she sent to Townsend were in his possession until his death in 1995, and their disposition after his death has been handled by his family, who have maintained appropriate privacy.

The Douglas-Home letters present a more complicated situation. The letters Margaret sent to Douglas-Home were in his possession at the time of his death in 1968. His estate was handled at that point by whoever had authority over it, and the evidence of what happened to the letters is inconsistent across sources. Some accounts suggest that letters or copies of letters were acquired by persons acting on behalf of the palace in the weeks following his death. Other accounts suggest that the letters remained in private hands for a substantial period and that some have found their way into the hands of collectors or institutions through the normal market in historical documents.

What is established is that no major journalistic publication of the letters themselves has occurred. The references to their content that appear in biographies and journalistic accounts are based on descriptions provided by people who have seen them, or who have been told of their content by people who have seen them, rather than on the text of the letters themselves.

The absence of the published letters is itself a form of the palace's management. Not a management the palace necessarily orchestrated in this specific case, but a management that results from the combination of the normal instruments of privacy law, estate control, and the social dynamics of a world in which the people who hold historically significant documents often find that discretion is the wiser commercial and personal choice.

The question of what the archive holds that has not been made public is not, in this case, primarily a question about documents the palace has suppressed. The palace's specific suppression has been documented where it occurred: the Lascelles intervention in 1953, the management of the cabinet's involvement in the 1955 crisis, the arrangements surrounding the formal separation and divorce from Armstrong-Jones. These are all in some form in the record.

What is not in the record is the interior life, the specific texture of what Margaret felt across the decades of a managed life. The letters are the closest proxy for that texture, and the letters are not available.

What we have instead is the testimony of women who were around her. Glenconner's Lady in Waiting is the most substantial of these accounts, but it is itself a careful document, written by a woman who loved Margaret and who understood that what she published would be read against the full weight of Margaret's public reputation.

Glenconner describes a private person who was funnier, kinder, more genuinely curious, and more genuinely lonely than the public version suggested. She describes a woman who, when the household was not watching and the performance was not required, was someone worth knowing for the quality of what she brought to a room, not for the fact of who she was.

She describes a woman who understood her position in the family with a clarity that was sometimes painful to witness, who knew what she was not going to be, and who had the intelligence to identify what the knowing cost her, and the pride to pay the cost without complaining about it in company.

The private Margaret and the public Margaret were not two different people. They were two registers of the same person, one managed by the institution and one by her own preferences. The men who came into her life in the private register, who saw something of the unmanaged version, were the men the institution subsequently needed to manage. The circularity is its own kind of cruelty.

The television drama The Crown, which ran on Netflix from 2016 to 2023 and dedicated substantial material across several series to Princess Margaret's life and relationships, has, in the absence of the actual documents, become the primary source through which most people in the world now understand Margaret's story.

The Crown is a dramatic work, not a historical document. Its creators and its network have said so consistently and clearly. The conversations it depicts, the emotional interior lives it attributes to its characters, the specific dramatic moments it constructs from the general shape of historical events: these are inventions. They are informed inventions, produced from historical research and with historical consultants. But they are inventions.

What is interesting about The Crown's version of Margaret is not whether it is accurate, but what it tells us about the shape of what is available to dramatize. Because The Crown, working from the historical record that exists, arrived at something the record actually supports: a woman of considerable gifts, who was managed by an institution that did not know what to do with her, and who navigated that management across 70 years with varying degrees of success and dignity.

The drama also reflects, in its structure and its choices, the same absence this account has been tracking: the absence of the men's perspective. Townsend in The Crown is sympathetic but fundamentally exterior to the story, someone who happens to Margaret rather than someone who has his own equally full account of what passed between them. The drama can't show us what it doesn't have. And what it doesn't have is those letters.

The letters would change The Crown's version of Margaret if they were available. They would change it not because they would make her more sympathetic or less sympathetic, but because they would make her more specific. Specific is what the management has prevented. Specific is what the archive controls.

This is what the palace bought: not the preservation of a perfect reputation. Margaret's reputation has not been preserved as perfect. The biographies and the drama and the press coverage across 70 years have ensured that the cruelties and the difficulties and the complications are all on the record. What the palace preserved was the general outline at the expense of the specific. And the specific in a person's life is where the actual person lives.

The actual person, in Margaret's case, lived in the conversations after midnight, in the correspondence with men who saw her without the performance, in the hours on a beach in Mustique where nobody required her to be Princess Margaret. The management ensured that those hours, for the most part, are not in the archive. The archive holds the managed version. The managed version is what we have. We are working always from the outline. The actual person recedes.

At some point in the mid-1980s, when Margaret was in her mid-50s and the Llewellyn relationship had settled into friendship and the Armstrong-Jones divorce was seven years behind her and the Douglas-Home death was 17 years behind her and the Townsend statement was 30 years behind her, she told Anne Glenconner something that Glenconner reproduces in Lady in Waiting with the careful framing of someone who knows the sentence will travel further than the book.

She said, "I'm not as nice as people think I am."

Glenconner presents this as an act of self-awareness, which it was. It is also an act of self-definition, the kind of sentence a person says when they want to establish themselves against the hagiographic version that a warm public is sometimes constructing around them.

The accurate version: she was not as nice as people thought she was. She was also not as impossible as the staff memoirs suggested, or as reckless as the press coverage implied, or as defeated as the sympathetic biographies framed her. She was a specific person with specific qualities, some admirable and some not. And the managed life she had lived had produced a version of her that was legible to the public, but not quite the real one.

The sentence is also, and this is less often noticed, an act of protection. If she said it herself, no one else could use it to wound her. This is a habit of people who have spent decades managing what is said about them. They say the difficult truth first, and by saying it first, they control the framing of it in a way they cannot control the framing when someone else says it first.

She was extraordinarily good at this. The self-deprecation was real, but it was also strategic, which is another way of saying it was characteristic. She had been watching the gap between what people said and what they meant since she was a child, doing impressions in the nursery at Glamis, and she had spent 70 years refining the understanding of that gap into a social instrument of considerable precision.

The question of what she wanted at the most fundamental level, beneath the performance and the management and the specific relationships with the specific men, is a question the record doesn't quite answer. It gestures toward an answer through the pattern of her choices: the men she chose who were not what the institution preferred, the moments when she pushed against the management rather than accepting it, the late nights and the piano and the running-too-long evenings that were not accidents or failures of self-control but preferences, genuine preferences for a way of being that the formal structure of her public life could not contain.

What she wanted, taken at this level of abstraction, was what the institution specifically could not give her: a life that was hers in the sense of being organized around what she was rather than what she was supposed to be. The men she loved were people who had some access to the version of her that wasn't supposed to be. And the palace managed those people because it was managing, through them, that other version of her.

The management worked in the sense that the unmanaged version never fully escaped into the public record. The letters are not published. The diaries, if they exist, are not available. The conversations after midnight are in the memories of people who are themselves now mostly dead or who made their own choices about what to carry into print.

The management failed in the sense that the outline is available, and the outline is enough to understand what was happening. The specific person is gone. The shape of the specific person remains.

There is a moment in the archive that is available and that has been noted by several biographers. In the early 1970s, before the Llewellyn relationship began and after the Armstrong-Jones marriage had settled into its Cold War phase, Margaret gave a speech at a dinner hosted by a charitable organization she patronized.

The speech was not a state occasion. It was an after-dinner address to a room of people who supported a cause she had been formally associated with for years. The speech, in the accounts of people who were in the room, was extraordinary. She spoke without notes. She spoke about the cause with the specific knowledge of someone who had been paying attention for years, not performing interest but actually interested, and she made arguments that the professionals in the room said afterward they would not have been ashamed to make themselves.

The anecdotes she used were particular and well-chosen. The timing was the timing of someone who knows how to read an audience. The people in the room went home and told people about it. A few of them described it to journalists, and brief references appeared in print. The brief references were the standard thing: the princess was magnificent at the dinner. There was no transcript. There was no recording. There was no longer account.

She gave speeches like this throughout her life, in rooms that were not large enough or important enough to be recorded, to audiences who noticed what they were seeing but did not have the platform to say so in a way that would change the historical record. The speeches went into the people in the room, and the people carried them home, and in time the people died and the speeches went with them.

What remains is the outline: the form of a woman who was capable of real things, and an institution that arranged so that the real things were always done in rooms too small for the record to reach.

She was 71 when she died. The morning of February 9, 2002, was a Saturday. Outside the King Edward VII Hospital, London was doing what London does on a February Saturday morning, which is nothing connected to the specific life ending inside the building.

The Queen was at Sandringham and was reached by telephone. The announcement came at noon. It was in the standard language. It said that Her Royal Highness the Princess Margaret, Countess of Snowdon, had died peacefully in her sleep. "Peacefully" is a standard word. It appears in most royal death announcements. Whether it is accurate, or whether it is in this case another sentence the palace uses because the sentence was agreed upon long before the person it describes was born, is not something the record can tell us.

What is in the record is what she had been: the most photographed woman in Britain for long stretches of three decades, the subject of thousands of newspaper columns, the focus of parliamentary debate, the inspiration for two characters in a drama that a hundred million people watched.

She had been simultaneously one of the most publicly discussed people in the country and one of the least understood, because the discussion had been conducted with the managed outline and not the actual person.

The men who might have given the other account, the account from inside the private hours, were gone before her, or had chosen their silence, or had it chosen for them. Townsend died in 1995. Douglas-Home died in 1968. Armstrong-Jones outlived her by 15 years and said, when he said anything, what a man who has made a calculation about what is worth saying tends to say. Llewellyn attended her funeral and went home.

The account from inside the private hours is still waiting, in the archive or in private hands or lost. What we have is the managed version, which is extensive, and the outline of the unmanaged version, which is enough to know it was there.

It was there. The men who were silenced knew it. The women who were around her knew it. And the institution that spent 50 years managing them all knew it better than anyone, because it was what they were managing.

There is no tidy ending to this story. The circle that Margaret's secret relationships described is not one that closed on October 31, 1955, with a statement issued through the Press Association, or on May 24, 1978, with the finalization of a divorce, or on October 15, 1968, with an open verdict in a Berkshire inquest, or on February 15, 2002, when a quiet man arrived at Windsor for a funeral and then went home.

The circle is still open. The letters are somewhere. The diaries, if they survived, are somewhere. The full accounting of what was paid and to whom, and what the payments purchased, is in the files of institutions that are not yet required to share them.

What this account has tried to do is trace the outline. The outline is what the management made public by its actions: the exile, the titles, the settlements, the pressures, the managed silence across 50 years. Each instrument tells you something about what it was managing. The way a lock tells you something about what's behind the door.

What was behind the door was a woman who deserved a fuller account than the institution allowed, and who, in the absence of that account, has been reduced in the public imagination to a series of tabloid frames: the tragic princess, the wasted talent, the beautiful ruin. None of those frames is entirely wrong. None of them is remotely sufficient.

Peter Townsend, who was the closest to her for the shortest period of the actual peak of her youth, said in 1978 that she was the most wonderful woman he had ever known. He said it 23 years after the statement. He said it in an interview promoting a memoir that was careful not to say everything he knew.

The sentence is the sentence of a man who has maintained his discretion, but who is not willing to maintain the pretense, at the price of the memoir's honesty, that what they had was not what it was. It was what it was. The management did not change that. The silence did not change that. The exile and the title and the settlement and the £36,000 debt and the sleeping pills and the 40 years of private garden design did not change that.

What the management changed was what we know about it. That is the account as completely as it can be assembled from what the management has permitted to exist.

The island of Mustique still exists. Colin Tennant died in August 2010 in St. Lucia, at the age of 83, in the manner he would have chosen: dramatically, in appropriate circumstances, with an audience. Anne Glenconner was with him. The island he built from a private obsession into a private retreat was sold piece by piece over the decades following his development of it, to people who could pay for the specific combination of privacy and tropical beauty he had found there in 1958.

Les Jolies Eaux, the villa built to Oliver Messel's design, was sold by Margaret's estate after her death. It is now a rental property. Guests who book it can stay in the rooms where she spent her winters, on the hillside above Gelliceaux Bay, with a view of the water she watched for 30 years, and on a beach where a photograph was taken that changed the management's calculations and the public's understanding of what was happening, and probably the remainder of her life.

The water looks the same. The beach looks the same. The story that happened there is in the archive or in private hands or in the memories of people who are still keeping it.

Peter Townsend's house in Normandy passed to his wife and children. His medals are in various collections. The manuscript of Time and Chance is presumably somewhere in the holdings of its publishers or its estate. The correspondence with Margaret is somewhere he or his family put it.

Antony Armstrong-Jones, the Earl of Snowdon, died on January 13, 2017, in London at the age of 86. He did not publish a memoir. He left photographs.

Robin Douglas-Home's piano, if it was kept, is somewhere in an estate that has not been publicly cataloged. His letters from Margaret, or what remains of them, are somewhere the record cannot currently confirm.

Roddy Llewellyn is in his late 70s in England, tending something.

Princess Margaret is in the memorial chapel of St. George's Chapel, Windsor, in the vault that was opened for the interment of the King and Queen Mother before her. She was interred there in April 2002 in a ceremony that was private. There was no public footage.

This, in the end, is the palace's most complete management: the actual body of the person in a vault in a castle, at a ceremony nobody filmed. The woman who was the most photographed person in Britain for 30 years was buried without photographs.

The archive holds what remains. The rest is still waiting.