Note on the First Microscopical Images of Captured-and-Reproducible Sound? (1878)

JF Ptak Science Books   Post 2743

Only months into the invention of the phonograph by Thomas Edison (at Menlo Park in December of 1877 and patented in January 1878) Alexander J. Ellis reviewed a version of it constructed in London by Mr. Stroh, and found it, well, wanting.  Ellis (1814-1890) was a product of Eton, Cambridge, Trinity, and was a gifted mathematician, philologist, and a groundbreaking ethnomusicologist, and did see some limited utility of the machine to restricted areas of his research. In the short note he wrote for Nature in their April 18, 1878 issue¹, he mostly found that the phonograph was generally flawed in its reproductive capacity, and did not venture far from this interpretation so far as the possible applications of the machine was concerned, which seems iconically short-sighted from my perch here in the future. Even though Ellis recognized that "the effects produced are sometimes startling (as in cries, coughs, laughter, music), the philosophy of the process (making a permanent impression of a very complex compound vibration, and using it as a mould to reproduce that vibration is exceedingly attractive, but at present the instrument--at least the one that I saw..." he concludes that the status of the invention "...has not risen beyond a lecture illustration or a philosophical toy".  

Five weeks after this article and about four months following the patenting a visually-arresting article appeared in Nature² which contains what I am assuming to be the first microscopic images of recorded and reproducible sound. I know that is a pretty deep qualification, but I think that's its accurate. The article appeared as "Examination of the Phonograph Record under the Microscope" by Persifor Frazer³, who investigated what the impression of the stylus looked like as known sounds are recorded on tin foil. And so this remarkable image, showing what vowel and some dipthongs "look like".  

There were a number of early attempts to study what sounds look like by Lissajous and Konig, though these instruments did not produce reproducible sound but the  experimenters were able to produce characteristic traces of sounds which were of high value to folks like von Helmholtz. The premier pioneer in this area must be Édouard-Léon Scott de Martinville  (1817 –1879), a Parisian printer and bookseller [and also described in other places as a typesetter, stenographer, and “tinkerer”], who was the inventor of the earliest known sound recording device (patented 25 March 1857) known as the “phonautograph”. He was a pre-Edison Edison and was a major influence on the man, though Edison was the man who produced playback from his recordings, and Scott did not. 

• "Scott called this process “phonauto­graphy”—the self-writing of sound. Of course he didn’t expect perfectly formed letters of the alphabet to emerge from the stylus. But he did believe the calligraphy inscribed in the soot—”the words that wrote themselves”—embodied a form of “natural stenography” that would some­day be read as easily as a sten­ographer deciphered his own jottings."--https://www.nps.gov/edis/learn/historyculture/origins-of-sound-recording-edouard-leon-scott-de-martinville.htm
• “Sound had been invisible and transient since the beginning of time. Scott’s phonautograph recorded it and made it both visible and perm­anent. It was a technological breakthrough, ahead of its time. He did not intend for his phon­autograms to be played back; that concept was another 20 years away. “-https://www.nps.gov/edis/learn/historyculture/origins-of-sound-recording-edouard-leon-scott-de-martinville.htm

This instrument was not as I said designed to play back the recording; rather it was intended that the sound's images and not the auditory experience would be studied. Scott achieved fame, recently, when archephonists were able to math the hell out of the recorded record and cranked an image to sound result—and thus, the earliest recording, Scott de Martinville  (or someone) singing a bit of “Clair de la lune”, made on 9 April 1860, which was presented for the first time in 2008.

I saw a reference and then promptly lost it stating that the first appearance in a journal of Scott's work occurs in 1857—except that in a later paper (1861) Scott references this event and describes giving the Academie a “paquet cachete”, a "sealed document", which I think means that it wasn't read to the assembly or published and was private. This turns out to be the case, the 1857 document being a manuscript, and presented at the excellent First Sounds website, here: http://firstsounds.org/publications/facsimiles/FirstSounds_Facsimile_01.pdf   It does seems though that the first and second published descriptions of the machine occur in 1858 and 1859, with Jules Lissajous’ "Report to the Society" (1858) and in Scott's own paper in Cosmos,  "Phonautographe et fixation graphique de la voix" (1859).

As luck would have it I own a copy of Scott's third paper on his instrument, which appeared in the Comptes Rendus in 1861 (15 July 1861, vol 53 #3, pp 77-128), which seems in its few pages to contain a good description of the device, which would render, as Scott said, “the imprudent idea of photographing the word.”

 Notes:

1)  "The Phonograph", being an article in the April 18, 1878 issue of Nature, volume 17, No 442, pp 485-6 in the weekly issue of pp 481-500. 

2)  Abstract of a paper presented at the Franklin Institute April 17, 1878, by Persifor Frazer, "Examination of the Phonograph record under the Microscope", in Nature, May 23, 1878, p. 102. In this same issue is an excellent paper by Francis Galton, "Composite Photographs..."

3)  Persifor Frazer was of the Philadelphia Frazers, of long and interesting lineage. A Gettysburg vet, he was in professional life mainly a geologist; he did become expert in handwriting and the detection of forgeries using photographic means. 

4) Scott, M. E.-L. "Inscription automatique des sons de l'air au moyen d'une oreille artificielle",  which is translated along the lines of  "Automatic Registration of Sound by an Artificial Ear". Evidently the idea came to Scott while working/reading a textbook and having an aha! moment while reading about the human ear.  Comptes Rendus, pp 108-110, volume 53, 1861.