William Hope’s spirit photographs presented an apparently simple claim: a portrait could contain an additional face or form which had not been visibly present before the camera. The Crewe Circle’s reputation depended on interpreting those extras as something more than ordinary photographic manipulation. The investigations turned instead on plates, dark slides, lighting and who controlled the apparatus.
A particularly detailed primary account comes from a sitting at Crewe on 16 December 1921. Harry Houdini printed the report of his friend DeVega in A Magician Among the Spirits in 1924. DeVega described suspicious handling and inferred substitution, while explicitly acknowledging that he did not see the suspected exchange. That distinction should survive any retelling.
How the appointment was arranged
Houdini said that he had tried to obtain a sitting with Hope in December 1921 but was told that appointments were booked for months. He then contacted DeVega in Glasgow and asked him to arrange a photograph. After correspondence, Hope agreed to a sitting at Crewe. Houdini was the organiser and later publisher of the report, not a witness standing beside DeVega during this sitting.
DeVega’s account describes arriving in the morning and being asked to return in the afternoon. He had brought his own camera. Hope initially preferred his own equipment, while offering to allow examination. Mrs Buxton participated in the session. The preliminary proceedings included hymns, prayer and hands held above the packages of plates.
These ceremonial details establish the setting in which the photographs were made. They should not be confused with technical controls over the plates. A procedure can be meaningful to its participants without preventing access to a photographic surface, a holder or a previously prepared substitute.
The camera and darkroom
DeVega described a small darkroom, roughly six feet high, three feet wide and five feet long. Shelves held cloths, chemical bottles and a candle-lit lamp. With two people inside, he said there was little room to move. The studio was described as a small glazed structure attached to the house, with a green curtain behind the sitter.
Hope’s camera was an old quarter-plate studio model. DeVega said it lacked a shutter and normally used a lens cap, which was missing at the time. Its dark slide was an old double wooden holder. He examined the holder closely and reported finding no preparation in it. His initial glance at the camera itself was less searching, because he told Hope that he did not doubt his word.
This is a useful admission about the scope of an inspection. An invitation to examine apparatus does not mean that the visitor performed every possible examination. DeVega’s own report gives the reader reason to distinguish the closer inspection of one holder from his less detailed look at the camera.
The suspected substitution
In the darkroom DeVega opened his package and loaded two plates into Hope’s holder. It was put on a lower shelf where he could see it only faintly. Hope also asked him to open his own holders and sign the plates. DeVega used the name J. B. Gilchrist. As he wrote, the lamp was moved to improve his light, leaving the quarter-plate holder in shadow.
Hope then handed him the quarter-plate holder to mark those plates as well. DeVega believed the holder he had loaded had been exchanged while his attention was occupied. Crucially, he did not report seeing the exchange. His certainty was an interpretation of the sequence and the poor view beneath the shelf, rather than direct visual identification of two holders changing places.
He considered spoiling a plate so that it would have to be replaced from his own packet, but decided to let the sitting proceed. He later questioned why he had not simply been allowed to mark his plates before loading them and keep the loaded holder in his pocket. Those are concrete criticisms of custody. They do not retrospectively create the control he chose not to apply.
Making the exposures
Back in the studio, DeVega sat for Hope’s camera. With the cap missing, the dark cloth served to cover and uncover the lens. Hope and Mrs Buxton stood on opposite sides and moved the cloth during the exposures. DeVega’s report places the photographic process within the continued singing, rather than describing a single unattended exposure.
His own camera was then used. DeVega regarded Hope’s hand being held in front of its lens as suspicious and wondered whether a luminous substance could account for a later bright patch. That was his proposed mechanism. The report does not record a chemical sample, an examination of Hope’s hand or a reproduced effect demonstrating that such a substance was used.
It is important not to combine this speculation with the separate holder-substitution suspicion as though both were proved operations. DeVega raised different concerns about different exposures. His account supports the existence of those concerns; establishing a particular method would require evidence beyond the fact that he proposed it.
Development and the resulting images
The plates were developed in the darkroom. DeVega said one quarter-plate darkened much faster than its companion. He regarded that behaviour as evidence of prior exposure, reasoning that the two had come from the same packet and had undergone comparable loading, portrait exposure and development.
He also described the comparison conditions: the same subject, nominally similar exposure lengths, the same aperture and plates developed alongside one another. These details show the basis of his argument. His claim that the behaviour was impossible without previous exposure remains his photographic judgement, rather than a separately demonstrated conclusion supplied by this article.
One quarter-plate produced an extra face with drapery above DeVega’s portrait. One plate from his own camera showed a bright patch over his face. Mrs Buxton interpreted the patch as a spirit light, while Hope thought he could discern faint facial features. The difference between their descriptions illustrates that even an agreed visible mark can receive different interpretations.
The later Price controversy
A separate 1922 investigation brought the dispute into wider view. Efram Sera-Shriar’s archival study for the Science Museum Group Journal dates Harry Price’s sitting to 24 February 1922 in London. Price used secretly marked photographic plates and interpreted the absence of their marks in the results as evidence of substitution. Hope’s defenders challenged the handling and integrity of the evidence, particularly its packaging and custody. The study judges fabrication very likely while explaining why Price’s particular demonstration was contested.
That later test is relevant to the same photographer and wider claim, but it should not be silently substituted for DeVega’s Crewe sitting. The participants, dates and procedures differ. The accompanying historical Price portrait illustrates the disputed kind of image; it is not a photograph of DeVega’s session.
Assessment, location and sources
The classification is Disputed. The published accounts include specific allegations of photographic manipulation and substantial challenges to the evidential handling. This does not validate spirit photography or treat all interpretations as equally supported. It preserves the distinction between strong suspicion, a claimed exposure and what each surviving record directly establishes.
The map uses an approximate Crewe town-centre reference near the Municipal Buildings. It does not identify Hope’s private house or studio, and the later Price sitting took place in London. No private residential address or visitor access is supplied. The incident years recorded here are 1921 and 1922; Houdini’s 1924 book and the museum’s 2022 study are publication dates.
The primary source actually reviewed is Houdini, A Magician Among the Spirits (1924), pp. 129–133, including DeVega’s dated report and the Price photograph reproduced between these passages. Houdini’s hostile judgement is distinguished from DeVega’s observations and inferences. The museum study provides further reading on the 1922 dispute; this article does not claim to have re-examined every original plate or correspondence file. Sources checked 13 November 2026.