N-rays

N-rays

Prosper-René Blondlot, 1903

Figures 2 and 3 from Blondlot's own collection of papers: a spark photographed through a slip of cigarette paper, dry and then wetted. Dry paper passes N-rays and wet paper stops them, so the first ought to be the brighter. "The photo-engraving," he writes, "shows that in the first case the spark is notably brighter."^[11]

Type	Radiation
Were one	No
Announced	1903, at Nancy[1]
Named for	The university, not the discoverer
Emitted by	Nearly every substance, and people
Not emitted by	Green wood; some treated metals[2]
Detector	The observer
Papers published	About three hundred[3]
Authors	About a hundred and twenty
Ended by	A prism, removed in the dark[4]
Still in Webster's	1946

N-rays were a form of radiation announced in 1903 by Prosper-René Blondlot, professor of physics at Nancy, and named for his university rather than for himself.^[1] Over the following three years they were investigated, confirmed, extended and disputed in about three hundred papers by something like a hundred and twenty scientists, and they do not exist.

The case is not one of a crank ignored by his betters. Blondlot was a corresponding member of the Académie des Sciences, had taken two of its prizes, and had done work on the propagation of electromagnetic waves that J. J. Thomson and Poincaré both thought well of. What went wrong is narrower and more uncomfortable than incompetence: the detector was a person. Every other property N-rays were found to have follows from that one fact about the apparatus.^[5]

The season for rays

It is worth remembering how reasonable this looked. Vacuum ultraviolet had been found in 1893, X-rays in 1895, radioactivity in 1896 and the electron in 1897. Four unsuspected things in five years, all of them real, several of them invisible, and every one announced by a respectable physicist.

A working physicist in 1903 who suspected there was another ray still out there was not being credulous. He was extrapolating, and the recent record was entirely on his side. Blondlot's error was not looking for a fifth. It was finding one.

What they did

Blondlot was trying to polarise X-rays when he noticed that a small electric spark seemed to brighten when the X-ray beam fell on it. He concluded that some further radiation was at work, and set about characterising it.

The characterisation proceeded quickly, and its results were these. N-rays were emitted by nearly every substance tried, including the human body; they passed through aluminium and through black paper; they were refracted by an aluminium prism into a spectrum, which was mapped; they could be stored in a brick and released later. They were not emitted by green wood, nor by certain metals that had been treated beforehand.^[2]

That last sentence is where the whole business is visible, and nobody at the time saw it. A real radiation has no reason to distinguish seasoned timber from green. An expectation does.

The detector

The apparatus varied. The detector did not. In every version, the final reading was a human being in a darkened room looking at a faint phosphorescent surface, or a small spark, and forming a judgement as to whether it had got brighter.

It is only fair to record that Blondlot saw this objection coming and tried to answer it. From 1903 he worked on registering the effect photographically, exposing a plate to the spark for a few seconds with the beam passing and then blocked, precisely so that the result should not depend on anybody's impression at the time. The plates were published. They are reproduced above, and they resolve the difficulty by moving it: the reader must now decide which of two grey rectangles is the paler.^[11]

This is a task the eye is unfit for, and the unfitness was already documented. The dark-adapted eye's estimate of small differences in the brightness of a dim source is unreliable, drifts with time and attention, and is powerfully suggestible.^[6] What was being recorded, in three hundred papers, was the state of the observer.

The consequences are exactly what one would predict of such an instrument. Substances the experimenters expected to emit rays emitted them. The rays got stronger when the experimenters expected strength. And their intensities, when written down, were reproducible within each laboratory and would not travel between laboratories, which was attributed for three years to the difficulty of the technique.

Kelvin, Crookes, Lummer and Rubens all tried and all failed to see anything. Rubens's failure was the most public: the Kaiser had asked him to repeat the French results, and after a fortnight he had to go back and report that he could not.^[7]

The commission on priority

While this was going on, Gustave le Bon and P. Audollet each claimed to have discovered N-rays before Blondlot, as did the spiritualist Carl Huter.

The Académie des Sciences convened a commission to determine which of them had precedence.^[8]

Wood in the dark room

Robert Wood of Johns Hopkins had already tried and failed to see N-rays, an attempt he described as wasting a whole morning. Nature asked him to go to Nancy and look properly. He suggested Rubens should go instead, as the more aggrieved party; Rubens replied that it had better be Wood, since Blondlot had been courteous to him throughout and it would look better coming from someone else. Neither of them wished to be the man who went, and one of them had to be.

In the darkened laboratory, during a demonstration of the N-ray spectrum, Wood reached into the apparatus and removed the aluminium prism, which was the component that produced the spectrum, and put it in his pocket. The demonstrator continued to read off the positions of the lines, and they were the same positions as before.

Later, asked to observe rays from a large steel file, Wood substituted a piece of wood, and the rays were duly observed. He reported all of this to Nature in a letter dated 22 September 1904, in tones of some restraint.^[4]

After

Outside Nancy the subject was finished within a year. Inside Nancy it was not: Blondlot is reported to have been still satisfied of the rays' existence in 1926, more than two decades on.

It is easy, and cheap, to make him ridiculous. The more useful reading is Langmuir's, who filed N-rays as a specimen of pathological science: not fraud, and not stupidity, but a genuine effect at the threshold of detection, studied by honest people whose instrument was their own expectation.^[9] Accounts that his career's collapse killed him are told with more confidence than the evidence supports.

The word outlived the thing. N-ray went into the dictionaries on announcement, and Webster's was still carrying a definition in 1946: an emanation from certain hot bodies which increases the luminosity without increasing the temperature, as yet not fully determined.^[10]

Blondlot left his house and garden to the city of Nancy, which made them into a public park. The park is still open and still bears his name, and when a visitor asked around the university some decades later, he found that hardly anybody there could say what the name was for.

See also

• Dayton Miller – thirty years of a real signal, read as the wrong thing
• The 21 grams experiment – a result that also outlived its refutation, and by a wider margin
• Vulcan – a body reported by good observers that was not there either
• Curved visual space – on what the eye actually reports, measured properly
• The Rayleigh criterion – a real and hard limit on seeing
• Dunning–Kruger effect – on the relation between confidence and competence
• Michelson–Morley experiment – a null result that was accepted
• Grimble's razor – where the consequence set aside was the prism itself
• The double-slit experiment – where the observer is an interaction and not a person
• Neith – where the observations were sound and the object was not
• Bathybius haeckelii – where the jar did what the observer did here
• The Allison effect – the other case in Langmuir's lecture, which produced two elements that were not there
• The Millikan creep – the same self-deception spread across a discipline instead of a laboratory
• Vic Tandy – a laboratory apparition with a real physical cause, and a mechanism for it that did not survive

References

1. ^ Named for Nancy. Blondlot's restraint in not naming them after himself has been remarked on since, generally by people making a different point.
2. ^ The exclusions are the tell, and they are not a later gloss: they were published as findings. A radiation that declines to leave green wood is a radiation with a view about carpentry.
3. ^ Roughly three hundred papers by roughly a hundred and twenty authors, in the ordinary journals, over about three years. Whatever this was, it was not suppressed.
4. ^ R. W. Wood, "The n-Rays", Nature, September 1904. The letter is short, and its restraint is the reason it worked.
5. ^ Which is why this article files under what the eye reports rather than under fraud. Nobody is known to have lied. There is some reason to think Blondlot's assistant confirmed whatever was wanted, which is a different failure and a commoner one.
6. ^ This was known at the time, and was said at the time, and made no difference at all for three years.
7. ^ The detail is worth keeping because it shows the pressure the thing was under. A head of state had taken an interest, a first-rate experimentalist had been set on it, and it still could not be reproduced.
8. ^ The commission's difficulty was not that the claims were hard to adjudicate. It was that all of them were claims to have been first to observe nothing.
9. ^ Langmuir's category covers effects near the limit of detection, claimed at low signal, defended with ad hoc explanations for why others cannot reproduce them, and abandoned without ever being disproved to their proponents' satisfaction. It describes several subjects in this encyclopedia.
10. ^ Webster's, 1946. Forty-two years after the prism went into Wood's pocket, the dictionary was still describing the rays as a phenomenon awaiting full determination.
11. ^ Blondlot, "N" Rays, trans. Garcin (1905), the plate facing p. 16, and the passage on the cigarette paper at p. 16. The wording is his: the engraving shows that the first spark is notably brighter. It is the only claim in this article that the reader can test personally, at the top of the page.

External links

• R. W. Wood, "The n-Rays" – Nature (September 1904) – the letter that ended it
• I. M. Klotz, "The N-Ray Affair" – Scientific American (May 1980) – the standard modern account
• P.-R. Blondlot, "N" Rays: A Collection of Papers Communicated to the Academy of Sciences, trans. J. Garcin (1905) – the plates and the papers, from the author
• N-ray – general reference