In Warsaw in August 1923, the Polish engineer Stefan Ossowiecki reportedly described a drawing concealed inside three envelopes. The researcher who had prepared it stayed away from the sitting to avoid transmitting its contents through supposed thought-reading. Afterwards he inspected the packet and accepted the result as paranormal. The surviving accounts make this a substantial historical experiment to examine, while their partial results and overlapping testimony require more care than the familiar story of a man who could simply read sealed letters.
.jpg){kind=link}
A test brought to the Warsaw congress
The occasion was the Second International Congress of Psychical Research. Gustave Geley, director of the Paris Institut Métapsychique International, published a detailed account in the September–October 1923 Revue Métapsychique. Eric J. Dingwall subsequently described his preparations and assessment in the British Society for Psychical Research’s journal. These are contemporary research reports by interested investigators, rather than a later folktale. They are also connected accounts of the same event: Dingwall explicitly relied partly on Geley for what happened during the reading. Geley, pp.317–320; Dingwall, pp.259–263.
The concealed target combined several elements. A small sheet carried a French sentence about wine from the Rhine, Moselle and Burgundy, a deliberately rough bottle-like drawing within a three-sided border, and the preparation date, 22 August 1923. Folded paper went inside an opaque red envelope, then a close-fitting dark envelope, and finally an outer envelope. The two inner envelopes were not individually sealed. The outer one was pasted and wax-sealed, and four tiny holes pierced the packet at its corners. Their alignment was intended to reveal removal and replacement of the nested envelopes.
Geley reproduced Dingwall’s statement that nobody had watched the preparation or been told the contents. Until the handover in Warsaw, the packet was kept in a locked suitcase or carried with his passport. These details mattered to the investigators: they were trying to prevent an ordinary view through the paper or an unnoticed opening, and to separate direct clairvoyance from their alternative hypothesis of telepathy. This was their proposed reasoning, rather than proof that every possible information route had been eliminated.
The reading on 30 August 1923
Geley dates the visit to Ossowiecki’s home to 9pm on 30 August. Albert von Schrenck-Notzing brought Dingwall’s packet, along with two other letters prepared that evening by congress participants. Geley and René Sudre accompanied him. The two additional envelopes were white; the concealed drawing’s outer wrapper was described as grey by Geley and brown by Dingwall. Ossowiecki handled the three packets and selected the latter. The published record describes him concentrating, walking about and speaking in brief phrases separated by pauses, while remaining in view.
The account is more complicated than an immediate, exact recital. Ossowiecki initially spoke about the Hôtel de l’Europe and people connected with the other letters. He then referred to multiple envelopes, greenish material, red colour and a small bottle. He described Schrenck-Notzing’s study before returning to the target. Geley recorded the statements as they were made. The resulting transcript mixes descriptions of the concealed paper with impressions concerning people and settings; its different elements should not be treated as equally testable observations.
The most striking claimed correspondence was the sketch. Ossowiecki drew a bottle-like form within a border and specified an imperfectly drawn, slightly leaning bottle without a stopper. He also described the nested envelopes and a white sheet folded around the drawing. These were the details on which the investigators placed particular weight. The articles reproduce the target and response, allowing readers to compare the published figures rather than rely entirely on a verbal declaration of success.
There were clear limits. The date was incomplete, and Ossowiecki said that something before the year might be a date or a town. He recognised that additional writing was present but did not reproduce the French sentence. Geley acknowledged the missing date detail even while describing the experiment enthusiastically. The distinction matters: this was an alleged successful description of selected features of a concealed document, not a demonstrated verbatim reading of all its contents.
Opening the packet and judging the result
The participants decided to return the packet unopened to Dingwall. Geley says Schrenck-Notzing did so that evening. The envelope was subsequently opened before the congress, after the drawing and account had been presented. Geley described the inspection of the corner holes, the successive extraction of the envelopes and the comparison on a blackboard. His narrative ends with applause and Schrenck-Notzing thanking Ossowiecki in the name of science. That reaction records the reception among those present; applause supplies no additional experimental control.
The chronology contains a small but real discrepancy. Geley places the public opening two days after the sitting; Dingwall says the following day. Both identify the reading on 30 August, so the disagreement should be retained rather than silently resolved. The envelope-colour descriptions also differ slightly. These points do not by themselves explain the drawing, but show why later compressed versions should be checked against the originals.
Dingwall reported finding no sign of opening or tampering and concluded that the result was supernormal. His strongest evidence concerned the packet he had prepared and inspected. He had deliberately not witnessed the reading itself. His favourable conclusion therefore needs to be read alongside the observers’ account, not counted as a second independent observation of every stage. Neither report provides a statistical success rate from a prearranged series of comparable trials.
A further test in 1933
Theodore Besterman’s later original report supplies a useful comparison. A first packet was reportedly opened by the censorship authorities, so he sent a replacement. Preliminary sittings took place on 8 and 9 August; the final session occurred on 29 September 1933, with Lord Charles Hope observing. The target was an ink-bottle drawing with words beside it. The published response included SWAN and the incomplete IN from INK. Besterman recorded both correspondences and errors in descriptions of his surroundings and preparation. Hope also noted that Gravier suggested the colour orange. Besterman nevertheless judged the overall performance highly successful and said his private marks showed no opening. His report documents several sittings and some prompting, rather than one isolated perfect trial. Besterman’s report and custody statements, pp.345–351.
What the evidence supports
Zofia Weaver’s 2002 historical review makes the positive case for taking Ossowiecki seriously. She lists precautions against visible preparation, translucent envelopes, opening, unsupervised handling, stereotyped targets and information from the person administering the test. She regards the 1923 experiment as meeting these requirements, while acknowledging limited knowledge of failures outside recorded experiments. Her review helps explain the continuing interest, but is a later assessment of historical records, rather than a new replication. Weaver, relevant discussion pp.64–69.
Outcome: Insufficient evidence. The original publications establish that named investigators described impressive results and accepted them as evidence of unusual cognition. They do not establish a reproducible ability to obtain concealed information. Partial readings, the dependence between accounts and limits on reconstructing custody and observation prevent a stronger conclusion here. This review has not demonstrated a particular trick or ordinary mechanism for the whole case, so it should not be presented as a solved fraud either.
Location and source trail
The map uses a broad Warsaw city reference, rounded from the city’s recorded coordinates. It does not identify Ossowiecki’s private flat, the congress room or a visitor attraction. The modern photograph provides city context only. No access to a historic private residence is implied.
The core reading is Geley’s complete 1923 account, Dingwall’s May 1924 article and Besterman’s 1933 report, including the appended custody statements. The relevant portion of Weaver’s original historical review was also checked. Dates in this article concern the reported experiments: 1923 and 1933. Publication dates and modern archive uploads are kept separate. Related reading: UKPX Cases and Spiritual & Psychic. Sources checked 13 November 2026.