Mary Allen, Charles Stanhope, and the Child at Holsworthy

In my family, one whisper has been passed down for generations:

That William Allen, born in Devon in 1810, was the illegitimate son of Charles Stanhope, 3rd Earl Stanhope.

It is the kind of story that sounds dramatic enough to doubt, but specific enough to investigate.

• A young woman from Devon.
• A powerful aristocrat from Kent.
• A child born without a named father.
• A rumour that survived generations.

And one question that keeps pulling me back:

Was there ever a real connection between Mary Allen and Charles Stanhope — or has my family been carrying a myth?

That is exactly the kind of question The Settlement Project was built to explore. Not to force a rumour into a family tree, but to map the world around it: the houses, estates, roads, parishes, employers, servants, neighbours, landowners, records, and silences.

Because sometimes a family mystery is not solved by asking only:

Who was the father?

Sometimes you have to ask:

Who was nearby?

The rumour

The woman at the centre of this story is Mary Allen, sometimes recorded as Mary Allin.

My current family research identifies Mary as the daughter of William Allen and Elizabeth Dennis. She was born on 10 December 1786 in Stoke Damerel, Devon, England, and later lived with her parents in Holsworthy, Devon.

The family story says Mary worked as a servant for Charles Stanhope.

That one detail matters.

If Mary was simply a young unmarried woman living in Devon, the story could float forever as a family rumour with nowhere to go. But if she was connected to a Stanhope household, estate, or residence, then the rumour becomes something else.

It becomes a research case.

William Allen's later death certificate gives his birth date as 18 February 1810. The Holsworthy baptism return gives me the first hard parish record in the case: on that same date, a child named William was recorded as the base child of Mary Allin. A note beside the entry says he was privately baptised and later received into church on 11 March 1810.

That line does not name Charles Stanhope.

But it does prove something important: in Holsworthy parish records, William was recorded as Mary Allin's child, born outside marriage.

The father is the mystery.

The family story says the father was Charles Stanhope.

I do not yet have proof of that. I do not yet have a record naming Charles as William's father. But the story has been repeated through generations, and it has enough detail to deserve investigation.

So the question becomes:

Could Mary Allen of Holsworthy and Charles Stanhope of Chevening have crossed paths at the right time?

The geographic problem

Charles Stanhope was born in London on 3 August 1753 and died at Chevening, Kent, on 15 December 1816. He was styled Viscount Mahon before becoming the 3rd Earl Stanhope in 1786. Britannica describes him as a radical English politician, experimental scientist, inventor, and "brilliant eccentric."

His public world was Chevening, Parliament, science, invention, reform, and aristocratic family papers.

Mary's world was different.

Holsworthy. Devon. Parish records. Domestic service. A child born outside marriage.

At first glance, those worlds seem far apart.

That is why the first map matters.

View map on The Settlement Project

This screenshot does more than decorate the story.

It shows exactly what The Settlement Project does.

A family tree can tell me that Charles Stanhope lived at Chevening. It can tell me Mary Allen lived in Devon. But the map lets me place those lives in relation to each other. It makes the distance visible. It turns a vague rumour into a geographic problem.

And once the problem is visible, the next question becomes sharper:

Was there a reason Charles Stanhope would have been in Devon?

That is where the story changes.

The place that changes the question

Charles Stanhope was not only connected to Chevening.

He was also connected to Soldon Manor, near Holsworthy, Devon.

Soldon Manor is a real historic manor house in Holsworthy Hamlets. Historic England lists it as a Grade II listed building in Devon, describing it as a manor house with mid-sixteenth-century origins and later alterations.

The Stanhope connection to Holsworthy was also real. The Stanhope of Chevening Manuscripts include estate papers and accounts, and the collection is held by Kent History and Library Centre under reference U1590. The catalogue arrangement includes manorial records, title deeds, estate papers, accounts, correspondence, maps, and plans.

That changes the whole question.

The mystery is no longer:

Why would a Devon servant know a Kent earl?

The better question is:

Was the Kent earl present at, visiting, managing, planning around, or corresponding from his Devon estate at the right time?

That is a question you can investigate.

And that is exactly where The Settlement Project becomes useful.

This second map narrows the case.

The first map shows the problem: Chevening and Holsworthy are far apart.

The second map shows why the rumour cannot be dismissed too quickly: Soldon Manor and Mary's Holsworthy world are close enough to investigate together.

And that tiny label — employer — matters.

It is not proof. It is a research hypothesis.

On The Settlement Project, I can record the relationship exactly as it currently stands: not "father," not "confirmed," not "proved," but employer, based on the story passed down through the family.

That is the power of the platform.

It lets the uncertainty stay visible.

It lets a rumour become structured evidence.

It lets me map the relationship I am trying to prove, without pretending I have already proved it.

The Chevening problem

One later family-history account, Cynthia Glover's The Allen Family, says Mary Allin was employed as a maid at Chevening House.

I have not yet found the primary evidence behind that statement, and I have not found evidence that Mary left Holsworthy for Kent.

That does not mean Cynthia was wrong. She may have had access to a family source, letter, memory, or record I have not yet seen. But it is also possible that Chevening entered the story because it was Charles Stanhope's best-known residence.

For now, I am treating the Chevening detail as a research lead rather than a proven fact.

The Settlement Project lets me hold both possibilities in place.

• Mary may have worked at Chevening.
• Mary may have worked at Soldon Manor.
• Mary may have worked for a Stanhope-connected household, tenant, agent, or estate.
• Or the family story may have preserved a real relationship but attached it to the wrong house.

That uncertainty is not a weakness in the research.

It is the research.

The canal clue

This is the part of the case that made me stop and look again.

Charles Stanhope was not just a distant landowner whose name happened to be connected to Holsworthy. In the 1790s, he was actively thinking about a canal project involving Holsworthy.

Britannica notes that Stanhope projected a canal between the Bristol Channel and his estate at Holsworthy, Devon.

A 1914 biography, The Life of Charles, Third Earl Stanhope, gives more detail, stating that in 1793 Stanhope projected a canal beginning from Bude Haven and passing Holsworthy, "of which he was lord of the manor." It also says that when Robert Fulton wrote to him about canal engineering, Stanhope replied from Holsworthy.

That does not prove he was in Holsworthy in 1809.

But it proves something important:

Holsworthy was part of Charles Stanhope's active world.

• Not just a name in an estate list.

• Not just a distant property.

• Not just a vague family connection.

• A place where he had plans.

• A place tied to his inventions.

• A place from which he corresponded.

• A place connected to land, engineering, money, labour, and estate management.

By 1796, Stanhope was still working on drawings, sections, models, and estimates for the Bude Canal. The canal itself was not actually constructed until after both Stanhope and Fulton had died.

This is the first real hinge in the story.

It does not prove paternity.

But it makes the rumour geographically and socially plausible enough to investigate.

The missing year: 1809

William Allen's death certificate gives his birth date as 18 February 1810, and the Holsworthy baptism return places him in the parish record on that same date as Mary Allin's "base child."

That means the likely conception window was around late May or early June 1809.

So the key question is not simply:

Did Charles Stanhope have a connection to Holsworthy?

He did.

The key question is:

Can Charles Stanhope be placed in Devon, Holsworthy, Soldon Manor, or a Stanhope-connected household around 1808–1810?

That is the gap.

And it is the gap the whole investigation now has to focus on.

By February 1810, Charles was about 56 years old. Mary was about 23.

Those ages do not rule the story out.

The geography does not rule the story out.

The estate connection does not rule the story out.

But none of that is proof.

The evidence still has to place the two people in the same world at the same time.

Charles's family situation in 1810

Charles Stanhope married twice.

His first wife was Lady Hester Pitt, whom he married in 1774. After her death, he married Louisa Grenville in 1781. Biographical sources name his legitimate children from these marriages, including Lady Hester Stanhope, Griselda, Lucy Rachel, Philip Henry, Charles Banks, and James Hamilton Stanhope.

That legitimate family structure matters because, if the family rumour is true, William Allen was born outside an already established aristocratic household.

By 1810, Charles had adult or near-adult legitimate children. His eldest surviving son, Philip Henry Stanhope, would inherit the title. His second son, Charles Banks Stanhope, had been killed at Coruña in January 1809, only months before William Allen's likely conception window.

I am not suggesting that event caused anything in Mary's story.

But timelines matter.

When you are testing a rumour, every date becomes part of the evidence board.

What was Charles Stanhope like?

This part of the story needs to be handled carefully.

The public Charles Stanhope was a reformer, scientist, inventor, and politician. UK Parliament describes him as being involved in issues including abolition of the slave trade, parliamentary reform, and religious toleration.

But the private Charles Stanhope appears to have been more complicated.

By 1806, Charles and Louisa had separated. I have not found evidence that they reconciled. They appear to have remained legally married until Charles died in 1816, while living apart.

A contemporary diary report also suggests that the separation was surrounded by serious social and legal tension. Joseph Farington recorded what he had been told about Lady Stanhope obtaining separate maintenance and about threatened adultery proceedings. I am treating that carefully: it is not proof of adultery, but it does show that rumours and pressure surrounded the breakdown of the marriage.

That does not prove anything about Mary Allen.

But it gives context.

If William Allen was Charles's child, he would have been born into a complicated private world: a separated marriage, family disputes, inheritance concerns, estate management, servants, accounts, and silence.

In other words, if there was evidence, it might not look like a neat record saying:

Father: Charles Stanhope.

It might be hidden somewhere less obvious.

• A payment.
• A letter.
• A servant account.
• A solicitor's note.
• An estate ledger.
• A family instruction passed down quietly.

That is what I am looking for now.

The deaths in December 1816

Then there is the part of the story I cannot stop thinking about.

Charles Stanhope died at Chevening, Kent, on 15 December 1816.

Mary Allen died in Devon on 28 December 1816. I have a copy of her parish burial record from Holsworthy Parish Church, which confirms her death.

• The same month.
• The same year.
• Less than two weeks apart.

I am not saying those deaths were connected. At this stage, I have no evidence that they were.

But if the family tradition is correct, then William Allen lost both his mother and the man named by family memory as his father within the same month.

William was only six years old.

That is not proof.

But it is haunting.

And it is exactly the kind of detail that makes a family story survive.

The "paid off" rumour

The rumour does not end with William's birth.

Family tradition says that some members of the Stanhope family knew about William — possibly one of Charles's sons — and that William was later "paid off."

The story says this may explain how William eventually immigrated to New Zealand and had enough money to purchase farms in 1841.

Again, this is not proven.

But it gives the investigation a second trail.

If William received money later in life, especially around emigration, it may not have come directly from Charles. Charles died in 1816, when William was still a child. If the payment story is true, the trail may involve executors, trustees, estate agents, solicitors, or the next generation of the Stanhope family.

That means I need to look not only for evidence around William's birth in 1810, but also for evidence around his departure, migration, land purchase, or unexplained financial support decades later.

Because sometimes family stories preserve the meaning of an event long after the document itself has disappeared.

"He was paid off" may not be the exact legal truth.

But it may point toward something real.

Why The Settlement Project matters here

A normal family tree can show me this:

Mary Allen → William Allen → descendant line.

But that is not enough.

A family tree wants a father's name.

This mystery does not give one easily.

The Settlement Project lets me ask a different set of questions.

It lets me see Charles Stanhope's public world at Chevening. It lets me connect him to Soldon Manor. It lets me place Mary Allin in the approximate area where she lived in Holsworthy. It lets me record Charles as "employer," not as a conclusion, but as a relationship to test. It lets me measure the distance between the places. It lets me see that a rumour has geography.

The maps do not prove the story.

But they transform it.

They take a family whisper and give it structure.

They turn a rumour into a research plan.

And as a proof of concept, these two screenshots say almost everything I want The Settlement Project to say:

**- People are not just names.

• They lived in places.
• Those places created proximity.
• And proximity can reveal connections a family tree cannot.**

That is why The Settlement Project exists.

Because our ancestors did not live as isolated names on a chart. They lived in settlements, households, parishes, estates, churches, lanes, farms, manors, and communities. Sometimes the people around them are the only way back into the truth.

Where the missing evidence may be hiding

The next step is not to keep repeating the rumour.

The next step is to search for records that could prove or disprove it.

On the Stanhope side, the Stanhope of Chevening Manuscripts, held by Kent History and Library Centre, look especially important. The collection includes correspondence and papers of the Earls Stanhope, and its catalogue arrangement includes manorial records, title deeds, estate papers, accounts, correspondence, maps, and plans.

That is exactly the kind of archive where this story might leave a trace.

If Charles was at Holsworthy around the right time, there may be correspondence. If Soldon Manor was being managed through agents, there may be estate accounts. If Mary worked in a Stanhope-connected household, her name may appear in wages, servant lists, household expenses, or local accounts. If money later passed to William, the trail may sit not in a baptism record but in a solicitor's paper, executor's note, trustee account, or family ledger.

On the Devon side, the search begins with the parish world around Mary and William. William's baptism matters because the exact wording matters. The line "William base child of Mary Allin" gives me a fixed parish clue: Holsworthy, February 1810, Mary Allin, an illegitimate child, no named father.

Mary's life in Holsworthy matters because if she needed parish support, or if the parish tried to identify William's father, there may once have been poor-law, settlement, or bastardy-related material. Her parents matter too, because any payment or arrangement may not have been made directly to Mary. It may have gone through her father, William Allen, her mother, Elizabeth Dennis, another relative, a parish officer, or someone responsible for the child.

If local parish records are missing, the next legal route is the Quarter Sessions. Not every bastardy case went that far. Many were handled locally by parish officers if they could identify the putative father and arrange support privately. But some cases did reach the county sessions, especially where payment, enforcement, or legal orders were involved. In those records, a case might appear in an Order Book, or a payment might appear in an Estreat Book.

And William's later life matters because the "paid off" rumour points forward, not just backward. If he left for New Zealand with unusual means, then migration records, land-purchase records, early New Zealand property records, or family correspondence may be just as important as the Devon material.

This is no longer one record to find.

It is a web of possible evidence.

• A parish record in Devon.
• An estate account in Kent.
• A manor record.
• A canal file.
• A servant payment.
• A solicitor's note.
• A Quarter Sessions entry.
• A New Zealand land purchase.
• A family story that may have preserved the shape of something real.

That is the work now.

What I have done next

Because this is still unproven, I have started contacting the places most likely to hold the missing evidence.

I have contacted Holsworthy Museum to ask about local files, name files, family traditions, parish notes, Soldon Manor material, and any references to Allen/Allin or Stanhope.

I have contacted Devon Archives / South West Heritage Trust to ask about Holsworthy parish chest material: poor-law, bastardy, settlement, churchwardens', overseers', vestry, and maintenance records.

Their reply clarified something important. The main Holsworthy parish collection, 2901A, is where any surviving Overseers' records would normally be expected. Those would have been the most obvious records for bastardy, poor relief, settlement, and removal matters. Unfortunately, no relevant Overseers' records appear to survive for Holsworthy for the period I need.

That does not prove nothing happened.

It means one possible paper trail may no longer exist.

Devon Archives suggested one remaining legal route: the Devon Quarter Sessions Order Book and Estreat Book. Because of that, I have now requested a search for any bastardy, affiliation, maintenance, parish support, payment, or reputed-father entry connected to Mary Allin / Allen, William Allin / Allen, or Holsworthy around late 1809 to early 1811.

I have also contacted Kent Archives about the Stanhope of Chevening Manuscripts, especially the Holsworthy estate papers, accounts, correspondence, and estate records that might show whether Mary was employed at Soldon Manor, Chevening, or somewhere else connected to Charles Stanhope.

This is the part of genealogy people do not always see.

• The waiting.
• The archive enquiries.
• The missing records.
• The surviving fragments.
• The carefully worded questions.
• The hope that one ledger, one order book, one letter, or one parish note might still survive.

What I know, what I suspect, and what I still need to prove

Here is where the case stands.

I know that Mary Allen or Allin lived in the Devon world connected to Holsworthy.

I know that family tradition says she worked as a servant for Charles Stanhope.

I know that William Allen's death certificate gives his birth date as 18 February 1810.

I know that a Holsworthy baptism return records William, "base child of Mary Allin," on that same date.

I know that Charles Stanhope is strongly associated with Chevening House in Kent.

I know that the Stanhope family had a real connection to Holsworthy and Soldon Manor.

I know that Charles was actively involved in a canal scheme connected to Holsworthy in the 1790s.

I know that The Settlement Project map places Mary's approximate Holsworthy location only a few miles from Soldon Manor.

I know that Charles Stanhope and Mary Allen died in the same month in 1816.

I know that family tradition says William may later have been financially supported or "paid off."

I know that Devon Archives has confirmed no surviving Holsworthy Overseers' records appear to exist for the relevant period.

I know that the remaining Devon legal lead is a search of the Quarter Sessions Order Book and Estreat Book.

But I do not yet know whether Charles was at Soldon Manor or Holsworthy around 1809.

I do not yet have proof that Mary worked for him.

I do not yet have a record naming Charles as William's father.

I do not yet have evidence of the alleged payment.

I do not yet know whether Mary or William appear in the Devon Quarter Sessions records.

So for now, the story is not proven.

But it is no longer just a rumour floating in the family.

• It has names.
• It has dates.
• It has places.
• It has a parish record.
• It has a map.
• It has a relationship to test.
• It has a research trail.

And somewhere, perhaps in an estate account, parish record, solicitor's letter, household ledger, or family paper, the missing piece may still be waiting.

Why I am sharing this

I am sharing this story because this is exactly what The Settlement Project is for.

Not just collecting ancestors.

Not just building trees.

But reconstructing the worlds around unresolved family stories.

Mary Allen's story cannot be understood by looking at her name alone. William Allen's father may not be found by looking only at William's baptism. Charles Stanhope's connection to this family, if there was one, may not appear in a public biography.

The answer may be in the settlement.

• In Holsworthy.
• In Soldon Manor.
• In Chevening.
• In the estate papers.
• In the servant accounts.
• In the road between a town and a manor house.
• In the records that sit around the people, not just directly on top of them.

That is the work.

That is the mystery.

And that is why I keep searching.

Call for help

If you have information connected to Mary Allen, Mary Allin, William Allen, William Allin, Soldon Manor, Holsworthy, Charles Stanhope, the Stanhope estate, the Bude Canal, or Allen/Allin migration to New Zealand, I would love to hear from you.

I am especially interested in servant records, household accounts, Holsworthy parish material, Quarter Sessions references, poor-law or bastardy records, Stanhope estate correspondence, solicitor records, migration papers, and New Zealand land-purchase records.

Sometimes a family rumour is only a rumour.

But sometimes it is the memory of a record no one has found yet.

And this is the case I am trying to solve.

Appendix: Case Timeline

3 August 1753 — Charles Stanhope is born in London This establishes Charles's age and generation.

c.1764–1774 — Charles spends much of his youth in Geneva This helps explain his education, scientific interests, and wider European outlook.

19 December 1774 — Charles marries Lady Hester Pitt His first legitimate family begins.

1780 — Lady Hester Pitt dies Charles becomes a widower.

1781 — Charles marries Louisa Grenville His second legitimate family begins.

7 March 1786 — Charles becomes 3rd Earl Stanhope He inherits title, status, and estate responsibilities.

10 December 1786 — Mary Allen / Allin is born in Stoke Damerel, Devon, according to current family research This establishes Mary's age and Devon background.

1793 — Charles projects a canal from Bude Haven past Holsworthy This shows Holsworthy was part of his active estate and engineering world.

7 October 1793 — Charles reportedly replies to Robert Fulton from Holsworthy about canal engineering This places Charles at Holsworthy, though much earlier than William's birth.

1796 — Charles continues working on Bude Canal drawings, sections, models, and estimates This shows continuing interest in the Holsworthy/Bude canal project.

1806 — Charles and Louisa separate This is important context for his later private life.

Late May / early June 1809 — Likely conception window for William Allen This is based on William's 18 February 1810 birth date and is the crucial missing period.

18 February 1810 — William Allen is born, according to his death certificate The Holsworthy baptism return records William, "base child of Mary Allin," on the same date. This is the strongest fixed point in the case: Holsworthy, Mary Allin, illegitimacy, and no named father.

11 March 1810 — William is received into church This was not a second full baptism, but the public church reception after private baptism.

15 December 1816 — Charles Stanhope dies at Chevening, Kent If the family tradition is correct, William lost the man named as his father when he was six.

28 December 1816 — Mary Allen dies in Devon I have a copy of her parish burial record from Holsworthy Parish Church confirming her death. The timing is striking, though it is not proof of connection.

1841 — Family tradition says William later had money to emigrate to New Zealand and purchase farms This may be a second evidence trail for the "paid off" rumour.

2026 — Devon Archives confirms no relevant Holsworthy Overseers' records appear to survive They suggest Quarter Sessions as the remaining Devon legal lead. This explains why the search has moved from parish poor-law records to Quarter Sessions Order Books and Estreat Books.