The headless ghost associated with Mary Gallagher is one of Griffintown’s enduring stories. Its historical basis is a woman’s violent death in Montreal in 1879; its supernatural narrative concerns later appearances attributed to her. An early magazine account places a cluster of sightings in October 1928 and promises a seven-year cycle. Reading that text as a published ghost story, rather than as a police investigation, makes its claims and its gaps much clearer.
The historical event and the haunting narrative
The early printed account reviewed here is the unsigned piece “The Headless Woman of Griffintown” in the July 1929 issue of Ghost Stories, on printed page 51. It links Mary Gallagher’s death to Susan Kennedy’s conviction and commuted death sentence. The article invokes court records as its authority for the crime. It does not reproduce the proceedings or provide an archival reference that independently authenticates the later apparition stories.
The distinction matters because evidence of a murder does not establish the identity of a figure seen decades afterwards. Nor does a real conviction give automatic authority to every detail of a popular narrative about motive, relationships or a ghost’s intentions. This review does not reconstruct the murder trial from the magazine’s compressed description, or present the publication as a complete court-file audit.
The alleged October 1928 encounter
The magazine names Francis Johnson as a man walking towards home at dusk on 26 October 1928. As he turned into William Street, it says, he noticed a white blur taking the form of a tall woman. The figure was described as transparent, with street light visible through it, and then as headless. Johnson supposedly fled when it screamed.
The passage is written as a dramatic scene: the man’s relaxed whistling stops, recognition arrives in stages and terror follows the discovery of the missing head. These are features of the published narrative. The piece supplies no signed statement from Johnson, interview date, recorded questions or independent observation of the same event. A named character makes a claim more traceable than an entirely anonymous legend, but is not equivalent to having the underlying testimony.
The following night, the story says, an unnamed woman reached the Young Street police station with a similar account. Patrols were supposedly increased, while further people reported the figure over subsequent nights. Some encounters included screams and others were silent. The magazine asserts a considerable public response, but the reviewed page gives no incident numbers, officers’ reports or separately reproduced witness statements.
Police, reporters and St Anne’s
The narrative expands from individual sightings to police reserves, newspaper correspondents and residents recalling earlier appearances. It says the Montreal Star ran stories and the rector of St Anne’s Church wrote about the apparition’s supposed harmlessness. Those references identify possible avenues for checking the story. They are not the same as inspecting the cited newspaper issues or obtaining the rector’s original letter.
One episode is assigned to midnight on 30 October. A caller reportedly told the police that the figure was on the church steps. Officers and reporters rushed there and found nothing, while two unnamed people were said to see it elsewhere at that moment. The story offers no independent evidence fixing both locations and times. It is therefore inappropriate to turn the account into a demonstrated case of an apparition evading a police cordon.
The magazine also says officers considered a practical joker. It does not report someone being caught, a costume being recovered or a confession. That proposed explanation deserves the same distinction between possibility and demonstration as the supernatural interpretation. The page does not solve the appearances, but its failure to do so cannot establish a spirit.
The seven-year claim
The piece repeatedly presents the ghost as returning every seven years, with older residents supposedly recalling previous cycles. It traces this pattern back about fifty years to Gallagher’s death. The account provides neither a dated list of all the earlier sightings nor a method for deciding which reports belong to the pattern. It is evidence that the recurring-cycle motif was in circulation by 1929, rather than a verified schedule of appearances.
The October setting is also worth preserving. A later familiar version associates the return with the anniversary of the murder, whereas this early text specifies an October cluster. These details should not be silently harmonised into one precise chronology. A legend can retain its principal figure and recurring interval while changing the circumstances of individual tellings.
The incident-year filter records 1928 for the sightings explicitly claimed in this source. It does not use 1879 as though the murder itself were an apparition, or 1929 as though the magazine’s publication created a new incident. The underlying claims remain qualified. The existence of a dated printed account establishes when and how a story was told more securely than it establishes the events within it.
Later work on the story’s transmission
A 2011 Concordia University account of Alison Loader’s project Ghosts in the Machine: The Inquest of Mary Gallagher offers a different kind of primary context. Loader described comparing newspaper material and questioning how Kennedy had been represented. Her installation used actors, historical accounts and multiple perspectives. That work documents the story’s interpretation and the reuse of common newspaper sources; a staged performance is not another apparition or an independent witness to the 1879 crime. It also cautions against treating every later repetition as a new source of evidence.
The neighbourhood itself has changed greatly. The McCord Stewart Museum’s Griffintown exhibition juxtaposed historical photographs with modern urban transformation. Its collection identifies the 1896 Notman views used here for context. A surviving place-name or an old photograph can help a reader understand the historical setting, but neither confirms the presence of a ghost.
Assessment and location
The outcome is insufficient evidence. An identifiable early publication preserves specific apparition claims, names and dates, but does not supply the underlying police records or independent witness documentation needed to test them. The later cultural and creative record establishes the legend’s continuing importance rather than supernatural causation. This review demonstrates neither an apparition nor a hoax.
The story is also attached to a real victim and a convicted woman whose representation has been questioned. Repeating sensational claims about their lives as settled fact would add little to the assessment. The useful historical question is how a violent event became a recurring ghost narrative, and which parts of that narrative can be traced to particular publications.
The map is an approximate Griffintown neighbourhood reference, based on GeoNames 12626110 at 45.49172, −73.56211 and rounded to 0.01°. It is not a private address, a verified apparition position or a promise of access to any building. No current ghost-watch event or visitor venue is being listed. The historical image predates the alleged 1928 sightings and contains no photographic evidence of them.
The principal source read in full is the 1929 magazine piece, printed page 51 (PDF page 53), compared with the Concordia discussion and museum context. The original newspaper articles, rector’s correspondence and police records invoked by the magazine have not been independently verified here. For another Canadian account whose versions need to be distinguished, see the Wynyard apparition at Sydney.