In the autumn of 1878, a household in Amherst, Nova Scotia, became known for reports of violent knocks, moving objects and fires associated with Esther Cox. An actor, Walter Hubbell, later boarded with the family and turned the disturbances into a book. The resulting Great Amherst Mystery survives through several layers of evidence: his early narrative, an expanded edition, interviews conducted decades afterwards and a detailed criticism of the records. Following those layers reveals both the attraction of the story and the difficulty of establishing exactly what happened.
Map location. The pin is an approximate Amherst town-area reference, rounded from GeoNames’ Amherst record. It does not identify Esther Cox’s home or an exact reported incident site.
The household before Hubbell arrived
Hubbell’s 1879 book, The Haunted House, places Esther with her married sister Olive and Olive’s husband Daniel Teed, a shoemaker. The crowded cottage also accommodated children and other relatives. Its opening chapters describe ordinary domestic routines in exceptional detail, including conversations and private thoughts. Those passages are reconstructed storytelling. Hubbell was not present when the reported disturbances began in 1878; his own account places his arrival the following June.
The early book names Esther’s elder sister Jane; later accounts commonly call her Jennie. This small difference is a reminder to identify the edition being used. The early narrative gives Esther’s birth as March 1860, making her eighteen when the story begins. It connects the onset with a frightening buggy journey involving a young man named Bob McNeal and a reported threat with a revolver. That episode reaches the reader through Hubbell’s account of what Esther later disclosed, rather than a surviving statement independently taken on the night.
The box, the bed and the writing
In Hubbell’s version, the first bedroom disturbance seemed like a mouse moving in the mattress. A subsequent rustling came from a box of patchwork under the bed. The sisters moved it into the room and, according to the narrative, watched it jump. Daniel initially dismissed their alarm. The story then escalates into Esther’s distress, reports of bodily swelling, explosive sounds and bedclothes being pulled away. These are historical claims about an individual’s experiences and behaviour; they do not provide a basis for diagnosing her today.
A local physician, Dr Caritte, appears in the 1879 text as a sceptical visitor who becomes puzzled by the household events. Hubbell describes a pillow moving, noises following the doctor across the room and threatening writing appearing on the wall near the bed. The book’s dialogue turns the visit into a dramatic conversion scene. It is evidence of how Hubbell presented the doctor’s involvement, not a verbatim medical record or a controlled demonstration. The different spelling Carritte appears in later sources.
The family also began treating knocks as answers. One episode is particularly useful for comparing editions. In the 1879 text, five knocks answer a question about how many people are in the room, and the narrative names five people. The revised 1888 book, at printed page 58, instead gives six knocks and six people. Both versions present the answer as correct. The discrepancy does not resolve the haunting, but it makes the exact count unsuitable as a stable piece of corroboration.
Fires and movement between households
Hubbell’s early narrative reports pauses as well as escalation. Esther’s illness and time away form part of its chronology, while the disturbances are repeatedly described as dependent on her presence. The story follows her to the household of John White and his dining establishment, where further incidents are attributed to the same agency. Its repertoire includes disappearing objects, thrown utensils and fires. The apparent movement of a haunting with a person made Amherst a different kind of story from a permanently haunted building.
The named spirits belong to the participants’ interpretation of the disturbances. The texts describe messages or appearances attributed to Bob Nickle, Maggie Fisher and others. It would be misleading to turn those names into independently established identities. Equally, an assertion that a fire appeared without anyone lighting it depends on what was actually watched, from which position and for how long. The narrative often moves from an unexplained occurrence to a spirit explanation more quickly than the evidence warrants.
Hubbell’s own visit in June 1879
Chapter V gives a precise beginning to Hubbell’s personal participation: seven in the morning on 21 June 1879. He describes an umbrella flying over his head, a knife arriving nearby and a chair striking the one on which he sat. He also claims that answering knocks gave the number of his watch and dates on coins. Later he watched Esther while pretending to sleep and reported a glass paperweight narrowly missing his head. These episodes are stronger as evidence of his stated experience than the earlier reconstructed scenes, but they remain one observer’s published account.
Hubbell stresses daylight and his attention to Esther’s hands and feet. That states the control he thought he had; it does not document continuous observation by an independent recorder. His description also shows how quickly he accepted an intentional ghost as the agent, including an intention to injure him. The distinction between seeing an object arrive and seeing its complete movement from its starting point becomes central in the later investigations.
Carrington returns to Amherst
Hereward Carrington’s Personal Experiences in Spiritualism (1913), chapters VII–VIII, preserves a later attempt to question surviving witnesses. The hosting filename misleadingly begins “1884”; the book itself is the 1913 publication. Carrington says he spent most of 26–27 January 1907 interviewing people and visiting the cottage. By then many potentially useful witnesses had died or moved away, and alterations to the house had covered the reported marks.
Olive Teed retained her belief in the disturbances, yet told him that Hubbell had dramatised and embellished parts of the account. Carrington reports that, when pressed about flying objects, she could not remember seeing an object start its journey and generally noticed it after it had arrived. He nevertheless found her manner convincing and recorded other specific recollections, including a moving dishpan. His favourable impression of a witness and the limitations of her reported observation are both part of the evidence.
Carrington also sought Esther, then living in Massachusetts. He describes her reluctance to discuss the affair and an interview that ended without the full account he wanted. This was a retrospective enquiry, almost thirty years after the events, rather than an observation of the original disturbances. Her wish to avoid the subject cannot fairly be treated as either a confession or confirmation.
Prince compares the story with its records
Walter F. Prince took a markedly more critical approach in “A Critical Study of ‘The Great Amherst Mystery’”, Proceedings of the American Society for Psychical Research, volume XIII (1919), pages 89–129. He challenged the supposed independence of much of the testimony, tracing several newspaper accounts back to Hubbell. A crowd of visitors, he argued, was not equivalent to a collection of precise, independent observations.
His comparison of Hubbell’s journal and published book raises concrete problems. A sugar-bowl lid’s route and landing place change between versions. More substantially, Prince says the journal places Esther’s departure on 11 July, while the book describes an intensification during the latter part of July. These criticisms concern the reliability of the surviving narrative. Prince’s proposed ways an event might have been produced remain hypotheses, rather than evidence that he caught Esther performing that particular act.
Two endings and an unresolved record
The 1879 narrative ends with Hubbell visiting Esther at the Van Amburgh farm on 1 August, finding her quilting and describing her as settled. The 1888 edition adds a darker aftermath at printed page 148: it reports a barn fire, her conviction and imprisonment, while insisting that a ghost was responsible. It then reports marriage and the cessation of the disturbances. The court record was not consulted for this article; Hubbell’s explanation must remain separate from his report of the conviction.
The National Library of Medicine’s historical discussion provides context for that later edition. For the original publication and its provenance, Canadiana preserves the 1879 work from a National Library of Canada copy. Read together with Carrington and Prince, the books preserve an important haunting controversy, with substantial disagreement about the quality of its testimony. They do not establish a supernatural agent, nor justify treating every later retelling as independent support.
Related cases and location
For comparisons with other household disturbances and their records, see Enfield and the Cheltenham “Morton” haunting. Amherst is listed at town level. No private residential address or unverified map pin is provided. Sources and image provenance checked on 5 November 2026.