An exhibition on “Alexander Graham Bell and the Origins of Recorded Sound” opened this week at the National Museum of American History, including audio recovered from select artifacts. Here I’d like to showcase one recording from the same group which we haven’t heard yet, since I want to stress the value of continuing efforts to bring the museum’s rich phonographic holdings to life in the future.
In 2011, with the support of a Lemelson Center Fellowship, I examined all of the museum’s experimental sound recordings and attempted to identify them, put them in context, and link them to relevant written documentation. The largest collections of this kind were the ones associated with Alexander Graham Bell and his Volta Laboratory colleagues Charles Sumner Tainter and Chichester A. Bell, which at that point had never been fully inventoried or studied. I’ve since completed an illustrated “discography” of them, which I’ve just posted online in conjunction with the opening of the exhibition. Here’s what a typical entry looks like:
Most of the individual media objects that fell into the scope of my project—discs, cylinders, and so forth—were stored in a cabinet up in Room 5400, where I spent many an hour examining them with the help and guidance of curator Carlene Stephens. However, some machines that belonged to the same collections were stored in cabinets down on the third floor, and a few of those machines happened to have recordings on them, so one day I was escorted down with a flashlight to survey those as well. Here’s one example of a machine with a recording mounted on it, NMAH 287655:
Discs that look like the one on this machine are pretty common among the Volta artifacts at NMAH, even if they’re scarcer than hen’s teeth out in the wider world. Their backing is made of rigid binder’s board—the kind of cardboard used to make book covers—about 25 centimeters in diameter, with a spindle hole about 2.5 centimeters across. One face is typically coated with a layer of wax—paler or darker yellow, dark brown, or something murkier with hints of green—except for the rim and center. The groove is vertically modulated, and its pitch varies from experiment to experiment, sometimes even on the same disc, but seems to remain consistent within any given recording. The few dated specimens bear the dates April 13 and 15, 1885, including NMAH 287881, which ends with the now-famous passage: “This record has been made by Alexander Graham Bell in the presence of Dr. Chichester A. Bell on the 15th of April 1885 at the Volta Laboratory, 1221 Connecticut Avenue, Washington D. C. In witness whereof, hear my voice: Alexander Graham Bell.”
The most useful complementary documentation about these discs is found in the notes of Chichester A. Bell (let’s call him “CAB” for short). The whereabouts of CAB’s original notebooks is unknown, but a carbon-paper typescript of at least some parts of them survives in Box 27 of the Alexander Graham Bell Family Papers at the Library of Congress. In them, CAB discusses recording on wax discs formed on glass and brass backings starting in November 1883 (6:4ff); two related artifacts are NMAH 287920 and 287688. However, the introduction of wax-coated cardboard discs seems to correspond to an entry of March 27, 1885, reporting developments since sometime in February (7:9):
Some wax-coated discs at NMAH with dates in February and March 1885 have the same dimensions and structure as the later binder’s board discs, but they’re on a thinner cardboard; CAB referenced the relevant change in backing material in another note dated April 13, 1885 (7:12):
Two more wax-coated discs with dates in September 1885 (NMAH 287909, 287910) are still on binder’s board, but they’re smaller in diameter—around 12.5 centimeters, with six-millimeter spindle holes, clearly made to fit a new and different machine. On those grounds, I’ve tentatively dated all binder’s-board discs with a diameter of about 25 centimeters and 2.5 centimeter spindle holes to mid-1885 (no earlier than April, no later than September). This is the broad category into which the recording found on NMAH 287655 falls.
The third-floor storage area was pretty dark—remember that reference to a flashlight?—so it was hard for me to see details clearly there. However, I took flash photographs of everything for future reference, and these often proved easier to study afterwards than the artifacts had been “in person.” One of my photographs of NMAH 287655 revealed the grooving on the binder’s board disc in sharp contrast, showing two discrete bands:
Most discs of the mid-1885 variety weren’t originally labeled in any way, so I hadn’t expected to spot any written inscription on this one. Only hours later, when I was studying the photographs I’d taken, did I see that there was a written inscription which my camera flash had fortuitously picked up for me:
In case you’re having trouble reading it, that’s “Killarney” and “Hot Shot March,” separated by a line. The former is presumably the popular song “Killarney” with music by Michael William Balfe and lyrics by Edmund Falconer, while the latter must be Thomas H. Rollinson’s “Hot Shot March,” with arrangements published for band (Boston: W. H. Cundy, 1881) and for piano (Boston: Oliver Ditson & Co., 1884). This is extremely unusual. No other Volta Laboratory recording at NMAH is labeled as containing music of any kind. The group’s experimental notes mention rare snippets of singing among other vocal tests; Charles Sumner Tainter describes an effort to record the music of a music box between October 27 and November 11, 1882 in his Home Notes (8:51-69); and CAB mentions some attempts to record notes and chords played on a piano in December 1884 (7:7-8); but otherwise music is mostly absent from such documentation as well. And yet here we have a Volta disc containing two musical selections, judging from the words scratched into the wax. What should we make of it?
Well, here’s a little more evidence to consider. A couple old photographs show the same machine with the same disc mounted on it and were presumably taken at the Smithsonian sometime after the materials were donated in 1914 (these prints are from negatives 27,656 and 27,657):
However, an even older photograph taken at the Volta Laboratory itself shows the same disc—the grooving and patterns of minor damage are unmistakable—mounted on a different machine that was later cataloged as NMAH 287656. The glass negative of the photograph survives among the Volta artifacts in a group alongside other negatives bearing dates in 1884 and 1885:
So it seems that this disc was selected back in 1885 when a machine was being posed for photographing, and that it was found again years later mounted on another machine. Given the dozens of similar-looking discs that survive today among the Volta artifacts, I doubt it’s a coincidence that the same disc turns up displayed twice on different machines, as opposed to two different discs. Rather, I suspect we may be dealing here with a specially prepared demonstration disc that was called into service whenever the Volta group wanted to show off this kind of disc graphophone to best advantage.
The inscription “Killarney / Hot Shot March” might conceivably have been a mere mock-up—a label added to a random experimental disc to show what a musical record might one day look like—but I’m inclined to take it at face value and suppose that there really is some music recorded here, and that it was prepared for exhibition, maybe to potential investors or collaborators. If so, it would be the oldest known surviving musical sound recording made in advance for exhibition during future events, rather than for immediate playback during the same event, as with the recordings made during the tinfoil phonograph exhibitions of 1878 (including this one, which opens with a little cornet music). If we’re looking for the earliest extant prototype of the musical sound recording as a commodity, this might well be it.
I believe it may also shed some light on the long-term goals which the Volta Laboratory Associates entertained for their sound-recording experiments during the mid-1880s. On June 14, 1885, Alexander Graham Bell prepared a message for his colleagues urging that they focus henceforward on designing a portable and convenient dictation machine, and the Bell-Tainter graphophone that was finally put on the market in 1887 was clearly conceived in that spirit. But early publicity also stressed the graphophone’s potential as a source of entertainment, as we see in this passage from Franck Z. Maguire, “The Graphophone,” Harper’s Weekly 30 (July 17, 1886), pp. 458-9:
The Graphophone is now prepared to represent all moods: it will tell you a funny story, and laugh with you in natural tones; it will repeat a tragedy that is blood-curdling in its nature; it will tell you a love story with all the ardor of a wooer; it will sing you an Irish song, or whistle a selection from the Mikado.
It is expected soon to be able to correctly reproduce the songs of great singers, and the recitations, dialogues, etc., of distinguished actors, and by a process already successful to copy the records of the songs or recitations and dispose of them at a trifle, thus enabling a person to enjoy at home such delightful singing as PATTI would render, or such elocution as we would listen to from EDWIN BOOTH.
This wasn’t just empty talk. The Volta Laboratory Associates had thrown a lot of time and resources into trying to work out methods of mass-duplicating sound recordings, none of which would have made sense if they hadn’t taken the idea of a commercial recording industry pretty seriously. I’ll cite just a couple instances here to prove my point, but I should emphasize that these represent a consistent and well-documented trajectory in the association’s work. Tainter’s Home Notes describe one set of experiments carried out in October 1881 using a copper electrotype negative made from a wax original record to form duplicates in modeling compound, beeswax, tinfoil, sealing-wax, and paper (3:53-55), and several of the duplicates survive at NMAH (see P* in my discography). Later, for a brief period in or around January 1886, the Volta group used wax-coated discs on thin cardboard about eight centimeters in diameter, and there’s clear evidence of a metal mold made from a record on one of these discs being used in turn to press duplicates in tinfoil-covered wax on cardboard.
It’s likely—though less certain—that similar experiments were being carried out during 1885 with discs identical in format to “Killarney / Hot Shot March.” There are several negative metal molds among the Volta artifacts at NMAH that appear to have been made by electroplating vertically-cut wax discs with the dimensions typical for early-to-mid-1885, although they could conceivably be older than that. Also, some of the discs found on thinner cardboard from February and March 1885 contain recordings embossed onto wax or other materials covered with tinfoil. The results are so poor (or poorly preserved) that it’s hard to figure out exactly what was going on with them, but these may sometimes have been attempts at making pressings from molds.
Overall, it’s plain that the Volta Laboratory Associates had been lavishing a lot of energy on trying to solve the problem of how to duplicate sound recordings in large quantities, presumably as the basis for a future recording industry. What’s less obvious is what they thought the content of those duplicate recordings would be. Most of the Volta recordings documented in laboratory notebooks, or played back thus far, don’t seem to model commercially viable content; instead, we get more or less elaborate equivalents of “testing, testing, one two three.” NMAH 287920—a recitation of Hamlet’s Soliloquy, probably recorded by CAB in 1884—may be an exception, but it too could have been intended merely as a test of intelligibility (listen to it here). By contrast, “Killarney / Hot Shot March” stands out for its conspicuously entertainment-oriented subject matter. As far as I can tell, it represents the only surviving effort of the Volta Laboratory Associates to record music, and they appear to have singled out the results for special treatment: this was the disc they displayed on one machine when photographing it, and the disc they left mounted on another machine for posterity when they moved on to other things. They were apparently using it as a showpiece of sorts, perhaps because they were especially proud of it or because they considered it a striking proof of concept. And it’s unquestionably playable: audio has already been recovered from several other Volta discs in the same format, and some of those were in worse condition than this one is. It looks like it should be easy enough to remove the disc from the machine for scanning.
Plenty of worthy candidates for future playback remain among the Volta recordings at NMAH. For instance, there are several more items that probably feature Alexander Graham Bell’s own voice: other early test recitations credited to him, such as 287907, 287909, and 287910, and a few later specimens of actual dictated correspondence, such as 287791. Even so, I think NMAH 287655, “Killarney / Hot Shot March,” would be my personal top choice for a playback queue if appropriate funding could be secured. It would be nice to verify whether it really contains what its label says it does, and if it does, hearing music from 1885 would be hard to beat in terms of gratification and popular appeal. Moreover, we’d then be able to answer some questions we can’t resolve in any other way. “Hot Shot March” is presumably instrumental, but what was the instrument, or was there more than one? Was “Killarney” sung or whistled, or is it another instrumental? Were there any spoken announcements, or did the performers launch right into the music? And how good are the recordings, technically speaking? Was this a momentary triumph from which the Volta Laboratory Associates were then lured away by other priorities, or was it so ghastly that they gave up on music and turned their attention to business dictation in despair and disgust? Inquiring ears want to hear—but for now, we can still look and dream.
- “A Discography of Volta Laboratory Recordings at the National Museum of American History” (January 2015), online here.
- “Phonographic Treasures of the Smithsonian,” The Antique Phonograph 30:1 (March 2012): 23-27 [part one]; 30:2 (June 2012): 21-25 [part two]; 30:4 (December 2012): 20-24 [part three], online here.
- “Trilled Rs and the Dawn of Recorded Sound in America,” Prototype (December 2011), also published on the National Museum of American History “Oh Say Can You See” Blog and in the online version of The Atlantic.
- “Experimental Recordings from the Volta Laboratory: Six Playbacks from a Pilot Project – Released December 2011,” at FirstSounds.org.