Blondlot · Source catalogue

London: Longmans, Green, 1905, English translation of Blondlot's Comptes Rendus notes of 1903-1904. Body is PDF pp.14-116 (book pp.i-83); PDF pp.118-161 are a Longmans catalogue. Page convention: PDF p.N is the scan page; book page differs by +23 (book pp.3-16), +25 (pp.17-62) and +27 to +33 afterwards because of unnumbered plates, so the rule 'book = PDF - 25' holds only for the middle of the body. A Preliminary Notice (PDF pp.16-17) says the papers are reprinted as published, that the first paper's results on 'X' (Roentgen) rays really concern a different radiation named N rays (after Nancy), and that Blondlot's earlier velocity-of-propagation measurements for 'X' rays (C.R. t.125, pp.666, 721, 763) 'apply in reality not to X rays but to N rays', the velocity being 'the same as that of Hertzian waves, and consequently of light'; the book prints no velocity data for N rays. Sequence of the nine papers: (1) 2 Feb 1903, polarization of X rays: a very small spark gap near an X-ray tube is brighter when parallel to the cathode-ray direction and unaffected when perpendicular, varying like an analyser behind a rotating Nicol prism; quartz and lump sugar rotate the plane by 40 degrees, S rays by 18 degrees in the opposite sense [read: p.26-29]. (2) 23 Mar 1903, a new species of light: refraction by quartz prism and lens, regular reflection by glass, diffusion by ground glass, elliptic polarization with mica; no fluorescence or photographic action. (3) 11 May 1903, Auer burner: quartz-lens focus found with the spark; the text prints one pair (lens-to-slit l = 26.5 cm, image l' about 13.9 cm, quartz lens of 12 cm focal length for yellow light) and then says substituting 'other values' of l' gives a mean index 2.93, a second lens (f = 33 cm) giving 2.942, plus three further species at 2.62, 2.436, 2.29; the radiations pass metals, wood and glass but not rock salt 3 mm, platinum 4 mm or water [read: p.36-38]. The printed pair alone does not reproduce 2.93: 1/26.5 + 1/13.9 gives an effective focal length of about 9.1 cm for the N rays against 12 cm for yellow light, i.e. a quartz index near 1.72 on the usual thin-lens scaling [calc, from the printed numbers; quartz yellow index about 1.544 is [memory] high], and the other pairs behind the 2.93 are not printed. (4)-(6) May-July 1903: gas flames, heated metals, Nernst lamps and the sun also emit them; the calcium sulphide (CaS) glow is increased by the rays and diminished by a lead screen or the hand; a Rubens thermopile shows no heating effect (3 July 1904 addition inside a note dated 1903) and a platinum-wire resistance test puts any temperature rise below about 1/60 of a degree [read: p.55-57]. (7) 2 Nov 1903: dimly lit paper, polished steel needle and bronze mirror brighten when N rays act; Blondlot admits that aptitude for seeing small brightness changes 'is very different in different persons' (p.62). (8) 9 Nov 1903: storage of N rays by quartz, metals, CaS, stones, sea water, sodium hyposulphite. (9) 23 Nov 1903: sun-exposed brick brightens dim paper and makes a clock dial visible at 4 m; an ox eye is transparent and stores the rays. Later papers (deep read): strained bodies emit them (hardened steel, Rupert's drops, a Gallo-Roman knife after over twelve centuries; penetrate 1.5 cm aluminium or 3 cm oak); a second species N1 acts inversely (prism indices 1.004-1.041, wavelengths about 0.003-0.0083, same unit question as below); an aluminium prism of 27 deg 15 min disperses N rays into pencils of index 1.04, 1.19, 1.29, 1.36, 1.40, 1.48, 1.68, 1.85, wavelengths 0.00815 to 0.0176 from gratings of 200, 100 and 50 lines/mm and Newton's rings, relative error 'does not exceed 4 per cent' although he 'cannot with certainty' state the approximation (book p.58-60); wavelength rises with index, opposite to light. The unit of the wavelength table is not printed in the OCR; micrometres is an inference (about 8-18 nm). Photography (about forty experiments, one unsuccessful; 'several eminent physicists' unnamed witnessed them) records the spark brightened by N rays, not the rays; the printed figures carry the note that striae and most spots 'do not exist in the original photograph'. Named third parties: Rubens, Bichat, Charpentier, Gutton, Macé de Lépinay, Sagnac, Becquerel (see key_claims and relevance). The appendix gives a CaS/collodion screen recipe and observing instructions: dark room, silence, do not fix the eye, look 'like an impressionist painter', 'some people never succeed'; Blondlot writes that it would be illusory to show an audience, because effects depend on each person's position. The only detectors are the eye judging faint brightness changes, a small spark, and photographs of that spark; no instrument measurement independent of these appears. The book contains no account of Wood's 1904 test.

Key claims

• Blondlot reassigns his earlier velocity-of-propagation measurements (C.R. t.125) from X rays to N rays and states the velocity equals that of Hertzian waves, i.e. light (stated in the Preliminary Notice; no data given in this book).
• A small electric spark gap near an X-ray tube changes brightness with its orientation, showing that the emitted radiation is plane-polarized at emission (later re-attributed by Blondlot to N rays).
• Quartz and lump sugar rotate the plane of polarization of the radiation by about 40 degrees, and secondary 'S' rays by 18 degrees in the opposite sense.
• Radiation from Auer burners, Nernst lamps, flames, the sun and focus tubes includes N rays that pass aluminium, wood, black paper and metals (copper even at 65 cm, note 5) but are stopped by water, wet paper, platinum 4 mm and rock salt 3 mm.
• N rays are refracted by quartz lenses with quartz indices of about 2.93, 2.942, 2.62, 2.436 and 2.29 for different species, and by aluminium prisms with indices from 1.04 to 1.85.
• N rays have wavelengths of about 0.008 to 0.018 (unit not printed in the extracted text; presumably micrometres), measured by gratings and Newton's rings to within 4 per cent, and the wavelength increases with the refractive index.
• N rays increase the glow of a small spark, flame or dimly phosphorescent surface seen normally and decrease it when seen tangentially; they produce no heating effect on a thermopile and no photographic action of their own.
• Sun-exposed bricks, stones, sea water, quartz, metals and an ox eye store N rays and re-emit them; compressed or strained bodies (hardened steel, unannealed glass, Rupert's drops) emit them spontaneously and indefinitely, even after twelve centuries.
• A second species, N1 rays, acts inversely (dims rather than brightens), has indices 1.004-1.041 and wavelengths about 0.003-0.0083, and is emitted by strained ethyl ether and by copper, silver and hard-drawn platinum wire.
• The effect of the spark under N rays can be recorded photographically as a more intense impression (about forty experiments, one failure), N rays themselves having no photographic action.
• Colleagues at Nancy and elsewhere reported related effects: Bichat (N1 emission by stretched ethyl ether; lead's opacity due to oxide, C.R. t.138 p.548), Charpentier (physiological N-ray research, C.R. t.137 p.1049; wet cardboard transmits knife-blade rays), Gutton and Macé de Lépinay (magnetic field, electromotive force and sound brighten a feebly luminous surface with the same normal/tangential pattern), Rubens (a private letter, ambiguously reported: Blondlot writes that Rubens 'made the same observation', with a more sensitive apparatus, after describing both the null thermopile result and the brightened CaS tube; the book gives no quotation, date or data, so it is not independent confirmation of either); Sagnac proposed long N-ray wavelengths from quartz-lens multiple images, which Blondlot's grating results contradict; Becquerel appears only in a radioactivity comparison.

Relevance: Blondlot's own primary text and the whole evidentiary base of the project. It shows what was claimed, that each detector was the unaided eye judging faint brightness changes or a spark whose sensitivity is found by trial, and Blondlot's own admissions (observers differ, only some succeed, an audience could not be shown the effect). The Nancy colleagues named in the book show the 'over-credulous first' half of the story: confirmations were reported, but all inside Blondlot's circle and none with independent numbers. It is the textbook example of pathological science and observer bias.

Mainstream conflict: Strong. Mainstream physics holds that N rays do not exist: Robert W. Wood's 1904 test removed the aluminium prism (and swapped the supposed source for an inert piece of wood) without changing the observers' reports, and later work attributed the effect to subjective perception (see [web]). X rays from a tube are partially polarized (Barkla, 1906), but not by the spark-analyser mechanism described and not at the 40 degree quartz rotation claimed; a quartz index of about 2.9 belongs to the claimed N rays, not to X rays, and is far above any known quartz index except near strong absorption bands in the far infrared. Wavelengths of 8-18 nm would be extreme ultraviolet, which does not traverse metals, wood or 65 cm of copper. An index above 1 for aluminium at those wavelengths is possible only just below the Al L absorption edge near 72.7 eV (about 17 nm), not across the whole band, and 'wavelength rises with index' is the reverse of normal dispersion. The unit of the wavelength table is an inference from the text.

Non-substantive pages bundled in the same PDF: Google usage notice, blank leaves, dark cover plates, a frontispiece photograph of a phosphorescent screen made of about 25 calcium sulphide spots, and a Longmans science-book catalogue from PDF p.118 to the end. The book body ends at PDF p.116 ('THE END').

Relevance: None beyond showing the frontispiece is the recommended screen layout in the author's own instructions.

Mainstream conflict: none