Eusapia Palladino’s New York sittings of 1909–1910 brought a celebrated physical medium into direct conflict with American investigators. Tables rose, objects moved behind a curtain and sitters reported touches while believing her hands and feet were controlled. What those observations established became the central dispute. Critics described substitutions and physical interference; her defender Hereward Carrington argued that familiar tricks did not explain the entire record.
This case concerns the New York investigation, including Hugo Münsterberg’s sittings at the Lincoln Square Arcade and the Columbia University tests reported in 1910. It is distinct from Palladino’s much earlier visit to Cambridge, England. Münsterberg’s association with Harvard must not move the American séances to Cambridge, Massachusetts: his own account explicitly identifies the New York setting.
The room and the proposed controls
In American Problems, published in 1910, Münsterberg describes accepting Carrington’s invitation despite his previous reluctance to investigate mediumship. The relevant chapter, My Friends, the Spiritualists, records both effects he found startling and his reasons for rejecting a paranormal explanation. It is a participant’s argumentative account, not a neutral transcript agreed by everyone present.
The room had a partition and a small cabinet, black curtains, a light table and musical instruments behind the medium. A stenographer recorded the proceedings. Sitters examined the space and the lighting arrangements; Münsterberg also describes an alarm on the windows intended to exclude outside assistance. He did not accuse every participant of collusion. He explicitly rejected the idea that Carrington knowingly helped to produce fraud.
The principal control depended on touch. A sitter on each side was supposed to monitor a hand and foot, with additional attention to the knees. That arrangement mattered because apparently occupied limbs were central to the argument that Palladino could not physically reach a moving object. But feeling a shoe or believing a hand remained in place was not equivalent to continuously observing the actual limb.
What Münsterberg said he experienced
His account begins with table movements in strong electric light. After the room was darkened, he describes a breeze, movement of the curtain, sounds from a guitar, touches on sitters and a small table moving from the cabinet. He acknowledged feeling pressure through the curtain which he could not immediately explain. His rejection of the spirit interpretation did not require denying that objects moved or that people experienced the reported touches.
He also describes finding that Palladino had exchanged the feet used in the tactile control without his noticing the change at first. In another informal test, he and a neighbour secretly broke the supposed chain of touching hands while making it look intact. The effects continued. Münsterberg interpreted this as evidence against the claimed necessity of an unbroken chain, rather than a demonstration of the exact means behind every effect.
Some of his proposed mechanisms remained conjectural. He suggested concealed bellows or tubes for breezes, and discussed movements of clothing and the body. He did not report recovering such an apparatus. His own warning that he might be on the wrong track is important: a possible ordinary explanation should not be rewritten as a piece of equipment actually found in the room.
The concealed observer and the foot
For a later sitting, described as a week before Christmas, Münsterberg arranged for an unnamed man to observe secretly from the floor and cabinet area. Münsterberg still believed he could feel the medium’s shoe pressing on his foot. Behind the curtain, according to his account, the hidden observer saw an unshod foot reaching backwards towards the guitar and small table, seized it and prompted a scream.
Münsterberg interpreted the earlier touches on his arm and sleeve as the same foot at work, and the incident as a decisive exposure of the cabinet performance. He suggested that the other shoe could have held down the empty one, but that detail was a proposed explanation, not a recovered hook documented in the account. The essential reported observation was the freed foot and its seizure, despite an apparently continuing shoe contact.
The distinction between that concrete claim and his wider conclusion matters. The account describes a method by which tactile supervision could be defeated. It does not provide a simultaneous photographic record of each earlier table movement, nor does every theory in the chapter acquire the same evidential standing as the reported seizure. His broad dismissal was stronger than the limited observation alone.
The Columbia report of May 1910
The report in Science, 20 May 1910, pages 776–780, concerns sittings in Columbia University’s Physical Laboratory during January. Its opening statement says the undersigned attended three sittings to seek evidence of previously unknown forces. They found no convincing evidence of such forces, but did find indications of trickery. The conclusion was expressly tied to the data supplied by those sittings.
The named signers included Charles L. Dana, William Hallock, Dickinson S. Miller, Frederick Peterson, Walter B. Pitkin, Augustus Trowbridge, Edmund B. Wilson and Robert Williams Wood. A further statement signed by six participants says that, during a fourth sitting, something approaching control which excluded trickery was exercised and the evidential effects did not occur while it lasted. The paper also contains individual assessments which differ in emphasis.
Miller regarded the substitution of hands and feet as a recurring weakness in earlier favourable reports. He credited experienced conjuring advisers and watchers for their assistance. His conclusion was forceful, but other contributors asked whether a brief negative statement without full methods and observations deserved the finality readers might attach to a scientific publication.
Disagreement among the critics
W. P. Montague said the sittings had not convinced him of an unknown force and that repeated trickery made ordinary causes highly probable. He nevertheless thought the experiment had failed to deliver the rigorously conclusive test he had wanted. He favoured mechanical controls strong enough to eliminate familiar substitutions. That reservation was about the adequacy of the experiment, not a declaration that paranormal powers had been demonstrated.
Pitkin signed the majority statement while criticising its lack of detailed methods and results. He distinguished a strong judgement based on experience and probability from the stricter scientific conclusion the investigators had hoped to produce. These objections belong inside the history of the negative report. Describing all the contributors as expressing one identical degree of certainty would erase a substantive disagreement.
Trowbridge supplied a particular example from 22 January. Objects moved behind the medium while her limbs were restrained in ways which initially appeared impressive. But several observers thought the back of her chair had knocked objects in the cabinet; when the chair moved openly soon afterwards, more objects fell. His point was that a control of limbs could still leave another physical route for interference.
Wood’s observations and the X-ray test that never happened
Wood described watching the cabinet from above, using a weak reflected light on its floor. A grille concealed the light from the medium’s position while allowing the observer above to see movement against the illuminated background. He reported a dark object entering behind the curtain and groping towards the small table, but did not regard its appearance as proof of a supernormal additional arm.
He also described table elevations in bright light which he found difficult to explain, including an occasion when he felt the table legs were clear of contact with the medium’s person. This was a reservation in a critical account, not an independently established violation of physics. His report records both repeated trickery and some observations he considered unresolved.
Wood had prepared an X-ray apparatus and fluorescent screen to investigate the cabinet activity, but explicitly said it was never used because the sittings ended. The equipment had been set up and tested; the proposed test of Palladino did not take place. There is consequently no X-ray image from this investigation identifying either a concealed ordinary limb or a paranormal third arm. A planned experiment cannot be counted as a completed result.
Carrington’s answer
In his introduction to the 1911 English-language edition of Flournoy’s Spiritism and Psychology, Carrington acknowledges that Palladino failed to convince American investigators. He nevertheless rejects the claim of a complete exposure. His argument is that observers detected already familiar tricks, while leaving other effects unexplained, and that American investigators drew conclusions too quickly compared with prolonged European work.
Carrington had brought the medium to America and had a direct stake in defending the investigation and his own judgement. His published reply is valuable evidence of the dispute, but not a fresh experiment which resolves it. Nor does an unexplained remainder establish his preferred interpretation. The New York record is strongest where it identifies a particular observation, a particular control and the limits of the conclusion drawn from them.
Sources, image and related reading
The sources used here are Münsterberg’s complete chapter, the complete five-page Science report and its individual statements, and the relevant pages of Carrington’s 1911 introduction. His introduction is distinguished from Flournoy’s main text. The photograph, published in that edition, is captioned as showing Palladino in the room used for most New York sittings; its photographer is uncredited. It documents a person and setting, not paranormal motion.
Read Palladino’s Cambridge experiments of 1895 for the separate English investigation, and Margery’s Boston tests of 1924 for another dispute over controls and observation. No modern private room or speculative séance-site coordinate is identified.
Map location — approximate area: The pin uses Manhattan borough area in New York City. It is a rounded locality reference, not a particular historical séance room or the Columbia laboratory. Coordinate reference: U.S. Census Bureau 2025 Gazetteer representative area point. It gives no directions to private property.