Nelly Butler at Sullivan: the cellar voices, the 1800 procession and Cummings’s disputed record

In Sullivan, on the Maine coast, people reported hearing a voice and seeing a white figure associated with a deceased woman known as Nelly Butler. Much of the surviving testimony concerns visits to Abner Blaisdel’s house in 1800, followed by a nighttime procession to a neighbouring household. The minister Abraham Cummings collected statements from both adherents and opponents and argued that the events demonstrated survival after death.

The record is unusually detailed about disagreement. Some people said that they saw a figure clearly; others in the same gathering heard a sound without understanding words, or saw nothing. Several witnesses who opposed the affair nevertheless described an appearance they considered disturbing. Their opposition did not always mean that they believed nothing had happened. The central historical questions concern observation, religious interpretation and the reliability of Cummings’s published presentation.

The book and the people behind its account

This account uses the 1859 Portland reprint of Cummings’s Immortality Proved by the Testimony of Sense, published by J. L. Lovell, together with an original scanned copy. Its title announces a conclusion, and its opening sections defend the possibility of apparitions through theological argument. It is an advocate’s case assembled from testimony, rather than an official finding that the claimed spirit existed.

The figure identified herself in the accounts by initials corresponding to Nelly’s maiden or married name. David Hooper regarded her as his daughter, and George Butler understood the voice as that of his deceased first wife. Lydia Blaisdel subsequently married George and is also called Mrs Butler in the text. Keeping those two women distinct is essential: references to the living Mrs Butler in the cellar or procession generally concern Lydia, not the woman whose return was alleged.

Cummings says that he wrote down statements from people he knew, mostly interviewing them separately, then read the wording back and obtained approval. He also says that he selected extracts and usually only a few witnesses for each important occurrence. That procedure supplies a valuable source trail, while leaving the reader dependent on his selection, transcription and commentary. The printed statements are not independently recorded audio, and they should not all be described as sworn affidavits.

Recognition in the cellar

David Hooper’s statement dates a visit to 2 January 1800. He says that messengers brought him at the supposed spirit’s request, and that conversation convinced him that it was his daughter. His later account mentions family property. Such references helped him interpret the voice as personal and familiar; they remain reported recognitions by an interested family member, rather than a separately authenticated identification.

George Butler described being reminded of a private conversation with his first wife about the possibility of remarriage. He said that no other living person had heard it. He also reported an appearance resembling a woman in a winding sheet, with a small child, and claimed that his hand passed through it without resistance. The text preserves his assertion of private knowledge and his interpretation of the visual experience, without supplying an independent record of the original conversation.

Paul Simson described a different experience on 3 January. He heard a voice but could not understand its first speech; companions supplied the words. At another visit, he again said others understood speech which was unintelligible to him despite similar proximity. This makes it unsafe to count everyone in a room as an independent witness to every reported sentence. Hearing a sound, understanding speech and identifying its source are separate steps.

Opponents who also reported appearances

Cummings placed Hannah Gatcomb among the opponents. She initially suspected a contrivance and said that she still thought nothing good had appeared. Her testimony nonetheless describes hearing a voice while Lydia went upstairs and, later, seeing a small white appearance which seemed to grow as it approached. Fear led her to leave the cellar before others said that they saw the figure more fully.

Paul Simson Junior similarly returned after an unsatisfactory gathering because he wanted to discover a deception. He described asking the family to sit together, using a candle and searching the cellar. After further sounds and conversation he reported a white appearance which enlarged and diminished. His statement supports a particular claimed experience under the circumstances he describes. It is not a documented discovery of hidden apparatus, nor does his initial suspicion make his later account a controlled experiment.

Thomas Uran’s deposition begins with a visit dated 11 August 1800. He says that he examined the cellar with a candle and later saw an appearance resembling a person wrapped in a white sheet near sunrise. He also said Lydia was absent during the reported speaking, and agreed to clear her of producing that voice. Cummings nevertheless introduced him as someone who considered the whole scene deceptive. A concession about one suspected speaker did not necessarily change the witness’s overall judgement.

The candle and the procession

Sally Martin’s account of 13 August gives concrete reasons for unease. She describes sounds moving rapidly around the house, then the company entering the cellar. Blaisdel arranged their positions, told them not to crowd the supposed speaker and ordered the candle extinguished. Martin heard unfamiliar speech and later reported seeing a white personal form beside Lydia during the walk to another house.

Cummings inserted a lengthy defence of the arrangements after her statement. He argued that order was necessary and that candlelight could itself assist deception or confuse a white dress with the alleged apparition. That is his response to an evidential objection, not a demonstration that darkness improved identification. The original dispute concerned who controlled the conditions and what the observers could reliably distinguish within them.

The procession on the night of 13–14 August included singing and movement in pairs. Gatcomb described seeing a white figure with Lydia outside Captain James Millar’s house. Millar reported an appearance there, but said that he saw nothing on the return because he walked near the front. Paul Simson likewise distinguished the outward journey, when others said they saw the figure behind him, from his later position near the rear, when he reported seeing it himself.

These differences give the crowd evidence more texture than a statement that nearly fifty people all saw one ghost. Some observations were firsthand; other parts of a witness’s account repeated what people elsewhere in the procession said. The number present cannot establish identical sightlines, continuous observation or agreement about the source of the appearance.

Marriage, predictions and the problem of dates

The narrative links the supposed spirit with directions that Lydia and George should marry. Cummings also says that, soon after the marriage, the voice foretold that Lydia would have one child and die, and that this happened ten months later. Those are retrospective claims about prediction and fulfilment in the published account. This source alone does not provide independently preserved, securely dated versions of every prediction before its alleged fulfilment.

The reprint has internal chronological problems. Its certification of four depositions is printed as 6 August 1800, although the preceding statements include Sarah Simson’s visit on 7 August and Uran’s visit on 11 August. Some explanatory letters also use August 1799 where comparable testimony uses August 1800. These are problems in the surviving printed record, not merely errors introduced by a modern summary. They prevent treating every date as a secure contemporaneous timestamp.

A further distinction concerns Cummings himself. In Letter VI he dates his own claimed field sighting to July 1806. He describes first taking a pale object for a rock, then seeing a globe rise and transform into a female figure which changed size. He says he did not speak to it, and that it had vanished when he returned with others. That personal report is separate from the main gatherings dated 1800; it cannot be silently inserted into them as another simultaneous observation.

What the surviving evidence supports

Cummings’s compilation supports a historical episode of reported voices, appearances, religious controversy and community involvement at Sullivan. It also preserves attempts to investigate, specific reasons for suspicion and substantial variation in what witnesses claimed to perceive. His theological confidence does not resolve those variations or the edition’s chronology. The record supplies neither independently established contact with Nelly Butler nor a demonstrated reconstruction explaining every report.

Sources and related reading

The principal text is the 1859 reprint, especially Section II, Letters I and VI, and Section III’s depositions and statements by Gatcomb, Martin, Millar and the Simsons. The original scanned edition was checked for its title page, Cummings’s 1806 date and the problematic oath certification. The Maine State Library catalogue confirms the edition, publisher and broad Sullivan setting. The complete book is longer than the sections used here.

Compare the Fox sisters’ rappings for a later dispute about voices, signalling and the interpretation of reported investigation. This record identifies the historical Sullivan locality in Maine. It does not supply a private address, speculative coordinates or permission to enter a cellar site.

Map location — approximate area: The pin uses Sullivan town area. It is a rounded locality reference, not the cellar, procession route or private property. Coordinate reference: U.S. Census Bureau 2025 Gazetteer representative area point. It gives no directions to private property.