Listen to an Electric Fish from the 1870s!

It’s only a couple seconds long, but this is still an exciting snippet of historical audio, both audibly and conceptually.  You’ll find it at the end of this blog post, and if you’re feeling impatient, you can always scroll down and listen to it right now.  But as for what exactly it is—well, that’s going to take a little explanation.

As you may know, electric fish can produce high-frequency electric discharges, and they also have specialized organs for sensing electric fields much as human beings sense light and sound.  We can’t sense these electric phenomena directly in the nuanced way these fish can, but we can use speakers to educe them as sound waves in order to listen to them—a prime example of sonification.  Here’s a 1954 educational film from the Moody Institute of Science, called The Electric Eel, in which a loudspeaker is used to make electric discharges audible (starting at the 1:41 mark):

That film was made about sixty-two years ago (as of this writing).  But the practice of using audio technology to listen to the discharges of electric fish dates back much further, to around 138 years ago, when an article by Étienne-Jules Marey—entitled “Nouvelles recherches sur le poissons électriques: Caractères de la décharge du gymnote; effets d’une décharge de torpille lancée dans un telephone”—appeared in the Journal de physique théorique et appliquée series 1, vol. 8 (1879), 162-164, available in facsimile online here and here.  In English, the title translates to “New research on electric fish: Characters of the discharge of the gymnotus; effects of a torpedo discharge presented in a telephone.”  Marey was interested mainly in exploring an analogy he’d posited between muscular movements and the actions of electric organs.  As he wrote:

After having shown that muscular acts are complex, that is to say, that a muscle in tetanus or in contraction executes a series of successive small movements that I call tremors (secousses), which are added and combined to produce muscular shortening, I explored the discharge of the torpedo in order therein also to seek this complexity.  Causing this discharge to pass through an electromagnetic inscribing apparatus, I obtained the result I expected: I saw that this discharge is complex, consisting of multiple electrical fluxes, of which the frequency is about 150 per second.

Marey had secured trace A from the discharge of an Amazonian gymnotus and trace B from the discharge of a torpedo.  In both cases, the fish had been situated in a tank between two metal plates linked into a circuit with his electromagnetic recorder.  But he went on:

The difficulties of bringing exotic fish into France, and even the impossibility I found myself in this summer of procuring myself, on the coast of Normandy, a living batoid, made me look for another means of analyzing the discharge of electric fish. The telephone seemed to me to lend itself well to this analysis, since it makes a sound when it is traversed by successive currents of sufficient frequency.

Mr. G. Pouchet was then working at the aquarium of Concarneau; I sent him a telephone with the necessary instructions, and almost immediately I received the news that the discharge of the torpedo gives rise to a sound perceptible at a distance, but the pitch thereof is difficult to determine.

Just recently I had the opportunity to experiment myself on a torpedo and found that slight excitations of the animal provoke a relatively short croak, each of the small discharges provoked consisting of no more than about ten fluxes and lasting no more than 1/15 of a second. But, if one provokes a prolonged discharge by pricking the electric lobe of the brain, the sound which is produced lasts three to four seconds and consists of a sort of groan whose tone is close to mi₁ (165 vibrations), which substantially agrees with the results of the graphic experiments. The sound increases slightly in intensity and seems to rise a bit in tone when, by wiggling the needle, the electric lobe of the brain is excited.

The official publication date for Marey’s article was June 1879, but he had presented the same findings to the Academy of Sciences on February 17, 1879, which his telephone experiments must also predate.  That’s impressively early for a research project involving this kind of sonification.

True, efforts to use the telephone to sonify bioelectricity can be traced back even earlier—to at least February 3, 1878, in fact, when Ludimar Hermann reported listening to electric currents in the muscles of frogs and human beings.  Hermann’s experiments, along with a few others in a similar vein, have already attracted scholarly attention as very early instances of audification, which Florian Dombois and Gerhard Eckel define as “a technique of making sense of data by interpreting any kind of one-dimensional signal (or of a two-dimensional signal-like data set) as amplitude over time and playing it back on a loudspeaker for the purpose of listening” (see here, at page 301; and also Dombois’ separate piece here, at page 43).

Now, Dombois and Eckel claim that audification also encompasses the playback of sound recordings in general, writing that “every CD-player has an audification module,” even though they see “little special from the viewpoint of sonification” in such cases (here, at page 302).  I assume they’d make a similar claim about audification being inherent in all telephone signals.  But I see things differently.  I’ve argued elsewhere that there are two separate phenomena we should be careful to distinguish:

• Tympanic eduction, the technique of actualizing data as sound through the controlled rapid movement of a membrane.
• Sonification, the strategy of using a sonic parameter to represent something other than itself.

For me, audification—as a subcategory of sonification—is distinctive for using the sonic parameter of amplitude to represent some variable other than itself.  From this perspective, using sound pressure amplitudes to represent bioelectric voltages would count as audification, but the mere “reproduction” of sounds by telephones and phonographs wouldn’t because in that case amplitudes would simply represent amplitudes.  This is a bit like the difference between using a pebble to represent a pebble and using it to represent a bushel of grain.

In any case, Marey probably wasn’t the first person to experiment with audification when he set out to listen to the discharges of electric fish by telephone (unless he did his work over a year before he first presented on it).  But he also recorded phenomena of the same sort as the ones he was audifying.  And today we can audify those records in turn ourselves, listening to some of Marey’s data using a sonification technique he himself applied to the same sort of material.  Granted, the specific discharges Marey recorded wouldn’t have been the same ones he listened to by telephone; but even so, this is as close as we’re likely to get to hearing any of the pioneering audification efforts of the 1870s, rather than just reading about them.

As we’ve already seen, Marey published a pair of graphic traces of electric fish discharges in his 1879 article, and these could easily be transduced into sound using the techniques I described here.  I haven’t done so, though, because there’s an even longer and better trace available from the same course of experiments.

This had appeared in the first edition of Marey’s La Méthode graphique dans les sciences expérimentales (Paris: G. Masson), which is undated but thought to have been published in 1878.  Standard-resolution digital facsimiles are available online here and here, and I made a higher-resolution scan of the plate from the copy in the library of the Max Planck Institute for the History of Science in Berlin during a recent visit.  (For what it’s worth, the same material appeared unchanged in the second edition of 1885; see here or here.)  Here’s my own English translation of the accompanying text, found on pages 331 through 333:

On the intensity of current registered by the electrical rheograph.

This instrument is a Deprez electromagnetic signal to which I’ve made the following modification: between the armature and the soft iron is a compressible piece of variable elasticity which, being flattened due to the intensity of magnetic attraction, allows the writing stylus to make excursions of a greater or lesser extent.  So that the intensity of the current through the apparatus is reflected by the extent of the movements traced by the stylus.

Figure 176 shows the arrangement I used to determine the variations in the intensity of the electric currents of the torpedo.A thread of caoutchouc, reflected over two trestles, extends horizontally between the soft irons and the armature of an electromagnetic signal. The soft irons were filed down so as to present on top* a trough in which are engaged two half-cylinders of metal, welded to the bottom of the armature.

In this way, the drawing together of the parts subject to magnetic attraction will experience impediments increasingly large as the surfaces approach closer to one another.  Indeed, follow the armature through the different degrees of its lowering.  It will first meet the elastic thread by the convexity of the two half-cylinders which it bears on its lower side. At this moment, the extensibility of the thread will be very great; but, as the thread is lowered more, it will rest on points less and less removed from one another, thus it will become less and less extensible.  Further down, the thread of caoutchouc, stretched over the notch made in the soft iron, will be even less extensible; later, finally, when the thread, always pushed back by the armature, has taken the curvature of the parts that embrace it, it will oppose to a new descent of the armature the resistance which a taut thread of caoutchouc presents to flattening, a resistance which itself grows due to the deformation already obtained.

Figure 177 was traced by the instrument thus modified; a prolonged discharge of a torpedo has been provoked by the pricking of the electric lobe of the brain. From one end to the other of this discharge, the decrease in amplitude is considerable: from about 1 to 10.

In the first line of figure 177 are seen alternations of increase and decrease in the amplitude of electrical flux; they recur in a regular manner about every nine vibrations of the stylus. I have repeatedly observed a periodicity of this kind without knowing to what cause to attribute it.

The electric rheograph, with its current arrangement, cannot provide a very faithful trace of the phases of each flux, of its duration and of its shape; nonetheless it permits finding, in some cases, striking similarities between the form of an electric flux and that of a muscular twitch.

---

*Original text has somme at the end of a line with extra space at the end of it, suggesting a missing letter.  I’m reading it as sommet.

Marey’s account of his clever modification of a Deprez signal may not make for very gripping reading.  But here’s his Figure 177 converted into a playable sound file, with the speed set according to the scale at the bottom that shows how much space along the x axis corresponds to a tenth of a second:

---

That’s roughly what Marey would have heard through his telephone 138 years ago.  Meanwhile, the electric current in the wire leading to your speaker or headphones ought to be an approximate “reproduction”—at much reduced voltage—of the torpedo’s original electric discharge.