The Name Was Only Graffiti: Who Put Bella in the Wych Elm, 1943

On Sunday 18 April 1943, four boys were in Hagley Wood, on the estate near the Clent Hills in Worcestershire, when one of them reached into a hollow wych elm and found a human skull. Police recovered a nearly complete skeleton from inside the trunk, along with fragments of clothing, a shoe, and a cheap ring. The remains were those of an adult woman. The forensic examination concluded that she had been concealed in the tree within hours of her death, around eighteen months before the discovery, and that she had been gagged and most likely asphyxiated. She was murdered, and she was never identified.

The name by which the case is known did not come from any identification. From 1944, anonymous chalk graffiti began appearing in the area asking who had put “Bella” in the tree, and the name stuck. It is graffiti, not a record. In the decades since, two theories have carried the case’s fame far beyond the evidence that supports it: that she was a German agent dropped into a wartime spy ring, and that she was the victim of an occult ritual killing. Both outran the record. This is an account of what that record holds, what the evidence can and cannot establish, and why the most basic fact, who she was, was never settled. We keep three things separate, as always: what is documented, what the evidence shows, and what is still only a hypothesis.

The documented account

On 18 April 1943, four local boys were in Hagley Wood, variously described as poaching or bird-nesting. One of them, reaching into a hollow wych elm, found a human skull. The boys were trespassing and at first said nothing, but one of them told his father, and the police were informed. The four are named in the sources as Bob Farmer, Robert Hart, Thomas Willetts, and Fred Payne; the exact spellings rest on later retelling rather than on a primary record inspected here.

Police recovered a nearly complete human skeleton from inside the trunk, together with fragments of clothing, a crepe-soled shoe, and a cheap imitation-gold ring. A detached hand was reported found a short distance from the tree, a detail we return to below, because it is contested. The recorded clothing, per the Worcestershire Archive, was a mustard skirt, a blue and yellow striped cardigan, a light blue belt, a peach-coloured taffeta underskirt, and blue crepe-soled shoes.

The remains were examined by Professor James Webster, the Home Office forensic pathologist associated with the West Midlands laboratory and the University of Birmingham. His findings, as reported, gave a profile of a woman of roughly 35, about five feet tall, with light brown hair and irregular teeth, who had borne at least one child, and who had been dead at least eighteen months, placing her death in or before October 1941. A piece of taffeta found in the mouth indicated she had been gagged and had died of asphyxiation. Webster also judged that the body had been forced into the narrow trunk while still warm, before rigor mortis set in, which places the concealment within hours of death and argues against accident or suicide. We state the manner of death and the skeletal profile here, once and plainly, as the forensic record requires, and then move on; the mystery is not in the condition of the remains.

Police searched missing-persons records, circulated the dental pattern, and reportedly produced a facial reconstruction to aid identification. No match was ever made. From 1944, the chalked graffiti began to appear, the name “Bella” took hold, and the case settled into an open file that has never closed.

The physical record itself has not survived intact, and the way it was lost matters. The remains were reportedly held in a Birmingham police training collection at the Tally Ho! centre until roughly the late 1960s or early 1970s, after which their whereabouts became unknown. Webster’s original pathology report is likewise reported lost. That much supports the familiar claim that the evidence is gone. But the claim should not be overstated. The Worcestershire Archive records that the police case files were reviewed by West Mercia in 2005 and subsequently transferred to Worcestershire County Archives, held at The Hive in Worcester. So while the physical remains and the original report are the items reported missing, some documentary file material reportedly survives in an archive. The two are not the same thing, and the difference becomes important when the question turns to what can still be tested.

The evidence

What the record establishes is that an unidentified woman was murdered and concealed in the tree, that she was never identified, and that a phrase chalked on walls a year later gave her the only name she has. What it does not establish is who she was, who put her there, or even, on certain points, exactly what was found. Each item below is worth weighing for those gaps.

Webster’s findings and their limits. The broad profile, an adult woman of roughly middle height who had borne a child and had been dead well over a year, is consistently reported and is the kind of conclusion a competent examination of skeletal remains can support. The decisive limit is that the precise figures, the age of about 35, the height of about five feet, the eighteen-month estimate, all circulate through secondary retelling, because Webster’s original report is among the items reported lost. They cannot now be checked against the primary document. One later review of the case has noted that Webster’s work was not always reliable, which is a reason for caution about the exact numbers, though not a reason to discard the broad picture. We treat the profile as the reported forensic reading, not as independently re-verifiable fact.

The gag and the asphyxiation. The taffeta found in the mouth, read as a gag and as the cause of death by asphyxiation, is the single most important indicator that this was a homicide rather than a death by misadventure. It rests on the same lost report and should be held the same way: a strongly reported forensic finding, central to the case, and not something that can now be re-verified at source.

The detached hand. A hand was reported found near the tree, and that detail later became the foundation of the witchcraft theory. It is contested. Investigators are also reported to have attributed the scattered hand bones to animal disturbance rather than to deliberate severance. We do not assert that a hand was deliberately cut off and buried, because the record does not settle it. It is a genuinely disputed detail, and the theory that leans hardest on it leans on a premise that may not hold.

The graffiti. The first reported instances date to 1944, with variants recorded in the area, and the name “Bella” derives from them. Police are said to have taken the graffiti seriously in part because it was written too high to be a child’s work, raising the possibility that the writer knew something. Variants recurred for decades, later associated with the Wychbury Obelisk on Wychbury Hill, where the folklore spelling “witch elm” appears in some later instances. The tree was a wych elm, a species of elm; “witch elm” is a later folklore spelling, not a different tree. The exact wording, locations, and dates of the early graffiti rest on retelling and have not been confirmed here against contemporaneous press, so we give them as reported. What the graffiti establishes is the origin of the name and a durable social phenomenon. It is not evidence of identity. There is no proof the writer had genuine knowledge, and “Bella” was never matched to a missing woman.

The lost remains, and what they foreclose. With the physical remains and the original report missing, the most powerful modern tool, DNA analysis, has nothing to work from. In May 2023 a BBC podcast appealed to museums to help locate the remains so that DNA work might be attempted. The appeal underlines the limit rather than resolving it: as far as the reported record goes, the remains have not been found. The surviving file material at Worcestershire County Archives may yet hold detail, but documents are not tissue, and they cannot supply a genetic profile.

The 1953 letter and the unconfirmed intelligence interest. In November 1953, the journalist Wilfred Byford-Jones, writing as “Quaestor” in the Wolverhampton Express & Star, ran a series on the case. It drew a letter signed “Anna,” of Claverley, claiming inside knowledge, and the writer was later identified as Una Hainsworth, also recorded under her married name Mossop. Byford-Jones said some of her details were later verified, and both MI5 and the police are said to have looked into her claims. MI5 has never confirmed any involvement in the Hagley Wood case. This is the documented origin of the espionage theory, but it is a second-hand wartime claim made a decade after the fact, not corroborated physical evidence.

The honest summary of the evidence is short. The strongest things in the file are a homicide indicator that rests on a lost report and a name that came from anonymous graffiti. Even the most basic fact about the case, who the woman was, was never established.

The theories

Everything in this section is a hypothesis. None has been established, and the case’s central question, the woman’s identity and who killed her, is left open here on purpose, because that is where the honest record leaves it.

Wartime espionage. This is the most famous reading and the most thinly grounded. Its documented origin is the 1953 “Anna” letter, whose author, Una Hainsworth, gave an account of a West Midlands spy ring in 1941 that fed the Luftwaffe the locations of munitions factories, with the victim said to be a foreign woman who became entangled and was killed. A related family claim has Jack Mossop, to whom Una was married, confessing that he and a Dutch contact had put a woman in the tree, with Mossop reportedly dying in a mental hospital before the body was found. We present this as exactly what it is: a contested, second-hand family claim, attributed and not asserted. Both Mossops are long dead, and we make no assertion of proven guilt against anyone named in it.

The strand of the espionage theory that travelled furthest attached a name to the victim, and it is the part that most needs care. Some accounts identified Bella with Clara Bauerle, a German cabaret singer, by way of the executed German spy Josef Jakobs, who was said to have spoken of a performer being trained as an agent and possibly sent to England. That identification is disproven. The records cited by researchers establish that Clara Bauerle died in Berlin in December 1942, which is after the Hagley Wood victim was already dead, and she was reportedly far taller than the woman in the tree, around five feet ten inches against an estimated five feet. Clara Bauerle was a real person who was wrongly proposed as Bella and has since been cleared. She was not Bella, and nothing in this case is laid at her door. We name her only to retire the claim.

Witchcraft and ritual. The anthropologist Margaret Murray, a proponent of a now-largely-discredited “witch-cult” hypothesis, is reported to have suggested around 1945 that the detached hand echoed a “Hand of Glory” and that the killing might have been a ritual one. The press connected it to the 1945 killing of Charles Walton at nearby Lower Quinton, a violent death sometimes read as ritualistic. We give this as attributed speculation, and historians and folklorists have largely discounted it. The link to the Walton case is thematic, not evidential. Murray’s broader witch-cult framework is academically discredited. And the theory’s central premise, a deliberately severed hand, is itself the contested detail noted above, which investigators also attributed to animal disturbance.

An ordinary local murder. The least sensational reading is the most ordinary: a local or itinerant woman, killed and hidden, with the hollow tree serving simply as an effective place of concealment. A 2023 book reportedly notes that the family holding the timber licence to Hagley Wood had a connection to the travelling community, a thread some have pursued toward a local explanation. This category carries no famous name and no dramatic backdrop, which is part of why it is the least retold and arguably the most plausible. Its limit is the same one that defeats every theory here: no named local victim has ever been confirmed.

What remains unknown

She was never identified. That is the bottom line, and more than eighty years of retelling have not changed it. The murder was never solved. The detached hand that anchors the ritual theory may have been the work of animals rather than a human hand, and the question has never been settled. The graffiti that gave her a name proves only that someone wrote on a wall; whether the writer knew anything real is unestablished. And because the physical remains and the original forensic report are lost, the one tool that has cracked comparable cold cases, DNA, has nothing to work with, even though some of the police file reportedly survives in an archive.

So we will not tell you she was a German agent, because the spy theory rests on a second-hand letter written a decade later and on a confession reported within a family, and its one named identification, Clara Bauerle, is disproven by the record. We will not tell you she was killed in a ritual, because that reading depends on a severed hand that may never have been severed, and on a school of folklore scholarship that historians have set aside. We will not tell you she was a local woman who simply vanished into the trees, because no missing person was ever matched to her, however plausible the idea. And we will not tell you who she was, because the people who investigated her death could not, and the evidence that might have answered the question is gone.

What we can tell you is the documented shape of it. An unidentified woman was gagged, killed, and hidden inside a hollow tree in a Worcestershire wood around 1941, found by chance two years later, and given a name by strangers writing on walls. She is still unidentified, the file is still open, and the name on the case was never hers.

The Name Was Only Graffiti: Who Put Bella in the Wych Elm, 1943

The documented account

The evidence

The theories

What remains unknown

Sources

Primary / near-primary

Secondary / contextual