Help us identify the function of an early WWI-era storage media device?
February 2, 2015 6:00 AM Subscribe
In the summer of 1915, a wireless telegraph station in Sayville, Long Island owned by the German company Telefunken was caught sending covert commands to U-Boats patrolling the Atlantic Ocean. The transmission device, Mystery Object 40.9.11, was recovered, but it still isn't clear what it is or how the messages were sent. Can anyone help us figure out what it is and how it worked?
This mystery box is in the Henry Ford Museum in Detroit, and though they haven't been able to identify the thing, it seems like it should be possible based on the specifics of the machine itself. The details that are available, plus pictures, are all here, and our site (including details about the project) is here.
We're hoping a specific identification of the object's mechanism can be determined based on the following clues (from the description pages, above): it is a "light-tight wooden box containing a neon yellow, paper tape reel. When the Museum’s conservation department initially opened the box, the paper tape began fading to a near-white pale yellow, leading them to believe that the tape was treated with cyanide. Kristen Gallerneaux, Curator of Communication and IT at the Henry Ford Museum, first thought of the cyanotype process. But her search for evidence of cyanotype paper-tape devices in radio and wireless history came up with nothing."
This mystery box is in the Henry Ford Museum in Detroit, and though they haven't been able to identify the thing, it seems like it should be possible based on the specifics of the machine itself. The details that are available, plus pictures, are all here, and our site (including details about the project) is here.
We're hoping a specific identification of the object's mechanism can be determined based on the following clues (from the description pages, above): it is a "light-tight wooden box containing a neon yellow, paper tape reel. When the Museum’s conservation department initially opened the box, the paper tape began fading to a near-white pale yellow, leading them to believe that the tape was treated with cyanide. Kristen Gallerneaux, Curator of Communication and IT at the Henry Ford Museum, first thought of the cyanotype process. But her search for evidence of cyanotype paper-tape devices in radio and wireless history came up with nothing."
The reel reminded me of half a cassette tape. There is a hole at the top where the tape can get through.
I checked some German sources and found this picture from a website of a Swiss organisation for historical war-equipment. The pictured object, Schnelltelegraphie-Anlage ("Quicktelegraphy-System") is from 1944+ but at least one part looks very familiar.
It was used to store the tape on which the machine ("Streifenlocher" / Tape Puncher) would make holes and patterns ("Ondulator- bzw. Radiogramme"), which needed to be translated by hand.
posted by KMB at 7:55 AM on February 2, 2015
I checked some German sources and found this picture from a website of a Swiss organisation for historical war-equipment. The pictured object, Schnelltelegraphie-Anlage ("Quicktelegraphy-System") is from 1944+ but at least one part looks very familiar.
It was used to store the tape on which the machine ("Streifenlocher" / Tape Puncher) would make holes and patterns ("Ondulator- bzw. Radiogramme"), which needed to be translated by hand.
posted by KMB at 7:55 AM on February 2, 2015
The box looks light-tight in the manner of contemporaneous photographic equipment: it's painted black on the inside to reduce reflection to a minimum in the case of a light leak, and even has a red-tinted viewport, probably to allow you to check whether a reel is in place without opening it. Other than that, there isn't much going on--it looks like there's a small jockey wheel to allow the tape to be off or on the reel while it's in the box. Probably off, now that I think about it, as doesn't look like there's any way to turn the reel while it was in the box (a light-tight coupling that passed the reel's spindle through the backside, for instance).
An uneducated guess: whatever method was used to treat the paper tape was photosensitive and intended to destroy the message after it was exposed to the light necessary to read or transcribe it. If this was the case, the paper reel (and the box itself) may have just been a carrier for the message, and whatever chemistry was involved was extrinsic to the message or cryptographic code. The code itself could have been anything that could be printed or exposed onto a chemical emulsion applied to the paper--letters and numbers, dots, handwriting, anything. It may be lost forever. Think of film removed from its canister and exposed to light before processing--as far as I know there is no way to forensically reconstruct the image it held.
posted by pullayup at 8:02 AM on February 2, 2015
An uneducated guess: whatever method was used to treat the paper tape was photosensitive and intended to destroy the message after it was exposed to the light necessary to read or transcribe it. If this was the case, the paper reel (and the box itself) may have just been a carrier for the message, and whatever chemistry was involved was extrinsic to the message or cryptographic code. The code itself could have been anything that could be printed or exposed onto a chemical emulsion applied to the paper--letters and numbers, dots, handwriting, anything. It may be lost forever. Think of film removed from its canister and exposed to light before processing--as far as I know there is no way to forensically reconstruct the image it held.
posted by pullayup at 8:02 AM on February 2, 2015
Looking at the photos, I think that the artifact is incomplete; it's not the entire mechanism. Looking at the hole in the side of the box, and then also at some of the discoloration to the edges of the box, it looks as though there was another component.
The artifact is basically one reel of a reel-to-reel type playback device, using paper tape. That's reasonable given its time—what we think of as "tape" wasn't invented until later. Paper tape was the dominant linear storage media for data and was heavily used for telegraphy and other purposes.
My guess is that the tape was fed out of the box and into some other device which 'read' the tape.
Based on the assumption that the paper was coated in something to destroy it if the box was opened (not an unreasonable assumption, given the machine's purpose, and also supported by the 'cassette' box being light-tight, which would otherwise not be necessary and adds bulk), then I would suggest that the reader was probably optical and didn't use punched holes.
If the data was encoded using punched holes, then opening the box wouldn't destroy it. But if the data was encoded as marks on the paper using what amounts to an unfixed photographic process, then opening the box would destroy it. (Similar intentionally-inadequate fixing schemes were used for photographic proofs, basically an early chemical "DRM" to prevent copying. So it was a well-known technique.) So the message goes onto the paper, the paper reel goes in the light-tight box, and the box gets shipped to the US probably via diplomatic pouch. If it's opened for inspection—poof. Nothing but blank paper inside.
If there are holes in the tape—and I can't tell from the photos—they are probably for timing. Think of an old 8mm movie projector; the sprocket holes are used to drive the film past the gate at a reliable, fixed rate. (They aren't actually used to pull the film on or off of the reels.) This is because if the data was encoded using the holes, then it's hard to destroy... and also then you could have just used a standard paper-tape system, like those common for telegraphy; you wouldn't have to go to the work of making a crazy unique system just for the purpose.
So here's my WAG: Attached to the cassette was some sort of optical playback device which used a focused beam of light and some sort of electro-optical photodetector to 'read' the tape. In reading the tape, it destroyed it... so it was a read-once medium. Very clever. Alternately, and I don't know enough about early paper cyanotype processes to know which is more likely, maybe the people on the receiving end had to run the paper through some sort of chemical bath to 'develop' it into something with enough contrast to read; in other words, they were basically being sent exposed but undeveloped film.
What I am not sure of, though, is what the reader would have been like. An "electric eye" based around a photomultiplier tube would have been very high tech for 1915. Not impossible—the fundamental knowledge was all there—but it would suggest that the Germans were at least 5-10 years ahead of the rest of the world at the time, which is not something I have ever heard before. That would be very interesting if it was the case. If that was true, it could have been fed directly to the radio transmitter electrically by simulating a Morse code key.
What I think is more likely, is that the tape was just colored in dark and light patches or stripes, and that it was decoded manually. As the tape went through the 'reader' device, the light reflected from the tape probably just illuminated a window, and a human had to decode it just like they would have decoded Morse code. (The Germans had their own dot-dash code similar to Morse, but not exactly the same.) It was probably itself encoded and meaningless to the person doing the transcription. They would have simply transmitted it onwards, and the final decoding to meaningful instructions would have happened in the actual U-boats.
This wouldn't have required much special training; being able to transmit and receive Morse-type codes using flashing lights was a common naval skill (up through WWII actually).
posted by Kadin2048 at 8:35 AM on February 2, 2015 [2 favorites]
The artifact is basically one reel of a reel-to-reel type playback device, using paper tape. That's reasonable given its time—what we think of as "tape" wasn't invented until later. Paper tape was the dominant linear storage media for data and was heavily used for telegraphy and other purposes.
My guess is that the tape was fed out of the box and into some other device which 'read' the tape.
Based on the assumption that the paper was coated in something to destroy it if the box was opened (not an unreasonable assumption, given the machine's purpose, and also supported by the 'cassette' box being light-tight, which would otherwise not be necessary and adds bulk), then I would suggest that the reader was probably optical and didn't use punched holes.
If the data was encoded using punched holes, then opening the box wouldn't destroy it. But if the data was encoded as marks on the paper using what amounts to an unfixed photographic process, then opening the box would destroy it. (Similar intentionally-inadequate fixing schemes were used for photographic proofs, basically an early chemical "DRM" to prevent copying. So it was a well-known technique.) So the message goes onto the paper, the paper reel goes in the light-tight box, and the box gets shipped to the US probably via diplomatic pouch. If it's opened for inspection—poof. Nothing but blank paper inside.
If there are holes in the tape—and I can't tell from the photos—they are probably for timing. Think of an old 8mm movie projector; the sprocket holes are used to drive the film past the gate at a reliable, fixed rate. (They aren't actually used to pull the film on or off of the reels.) This is because if the data was encoded using the holes, then it's hard to destroy... and also then you could have just used a standard paper-tape system, like those common for telegraphy; you wouldn't have to go to the work of making a crazy unique system just for the purpose.
So here's my WAG: Attached to the cassette was some sort of optical playback device which used a focused beam of light and some sort of electro-optical photodetector to 'read' the tape. In reading the tape, it destroyed it... so it was a read-once medium. Very clever. Alternately, and I don't know enough about early paper cyanotype processes to know which is more likely, maybe the people on the receiving end had to run the paper through some sort of chemical bath to 'develop' it into something with enough contrast to read; in other words, they were basically being sent exposed but undeveloped film.
What I am not sure of, though, is what the reader would have been like. An "electric eye" based around a photomultiplier tube would have been very high tech for 1915. Not impossible—the fundamental knowledge was all there—but it would suggest that the Germans were at least 5-10 years ahead of the rest of the world at the time, which is not something I have ever heard before. That would be very interesting if it was the case. If that was true, it could have been fed directly to the radio transmitter electrically by simulating a Morse code key.
What I think is more likely, is that the tape was just colored in dark and light patches or stripes, and that it was decoded manually. As the tape went through the 'reader' device, the light reflected from the tape probably just illuminated a window, and a human had to decode it just like they would have decoded Morse code. (The Germans had their own dot-dash code similar to Morse, but not exactly the same.) It was probably itself encoded and meaningless to the person doing the transcription. They would have simply transmitted it onwards, and the final decoding to meaningful instructions would have happened in the actual U-boats.
This wouldn't have required much special training; being able to transmit and receive Morse-type codes using flashing lights was a common naval skill (up through WWII actually).
posted by Kadin2048 at 8:35 AM on February 2, 2015 [2 favorites]
Also, in 1915 the actual transmission to the submarines would have almost certainly been by continuous-wave Morse-type code.
According to this page, the Sayville Telefunken station used a rare type of transmitter, called a Joly-Arco. You can read about it in Popular Science Monthly (Vol. 91, July-Dec 1917, p.453), which notes that it transmitted on 32 kHz. (It seems to me that it would have been fairly easy to make it transmit on any sub-multiple though, if needed. E.g. 8 kHz, 4 kHz, etc.)
The US Navy favored the Alexanderson alternator system, by contrast.
At any rate, while it might have been technically possible to record actual sound on paper tape, if the tape was moved fast enough and if we make some really generous assumptions about German photomultiplier tube technology in the field in 1915, I don't know why you'd want to do that. The quality would have sucked (if it was possible to get good quality recordings on paper, they wouldn't have gone to the trouble of building expensive and dangerous metal-tape recorders) and it just wasn't necessary. A binary code would have conveyed the information and been much less complicated to decode.
posted by Kadin2048 at 8:49 AM on February 2, 2015
According to this page, the Sayville Telefunken station used a rare type of transmitter, called a Joly-Arco. You can read about it in Popular Science Monthly (Vol. 91, July-Dec 1917, p.453), which notes that it transmitted on 32 kHz. (It seems to me that it would have been fairly easy to make it transmit on any sub-multiple though, if needed. E.g. 8 kHz, 4 kHz, etc.)
The US Navy favored the Alexanderson alternator system, by contrast.
At any rate, while it might have been technically possible to record actual sound on paper tape, if the tape was moved fast enough and if we make some really generous assumptions about German photomultiplier tube technology in the field in 1915, I don't know why you'd want to do that. The quality would have sucked (if it was possible to get good quality recordings on paper, they wouldn't have gone to the trouble of building expensive and dangerous metal-tape recorders) and it just wasn't necessary. A binary code would have conveyed the information and been much less complicated to decode.
posted by Kadin2048 at 8:49 AM on February 2, 2015
Seems to me like you should be contacting German sources -- e.g. the history of cryptography people at the Deutsches Museum in Munich. or perhaps the NSA's National Cryptologic Museum.
posted by crazy with stars at 10:14 AM on February 2, 2015
posted by crazy with stars at 10:14 AM on February 2, 2015
I think this has nothing to do with cryptography (or nothing cryptography-specific). If you look at patent US1818585 you see that Telefunken is interested in sending and receiving morse code messages faster, so that they are less susceptible to time-varying ionospheric conditions and thus they wouldn't have to repeat as many garbled messages. Doing this requires having a way to record the message so it can be decoded (no cryptological meaning implied) "offline". The patent discusses exposing a light-sensitive paper, which would then need to be taken up into a light-tight film canister and then developed. I think this is that canister. Tape gets fed in through the port on the side from the radio receiving equipment, there's a little colored-glass window to see inside the canister, and then it can be taken over to a developing station and either unreeled or opened in the dark, developed, then decoded.
I think there's no reason to think that the possibility of opening the box and destroying the film was a feature rather than a drawback.
posted by kiltedtaco at 10:46 AM on February 2, 2015 [4 favorites]
I think there's no reason to think that the possibility of opening the box and destroying the film was a feature rather than a drawback.
posted by kiltedtaco at 10:46 AM on February 2, 2015 [4 favorites]
It may be very worth your time to reach out to the folks at reddit.com/r/AskHistorians, there's a lot of folks over there with very niche German WWII backgrounds.
posted by TomMelee at 10:56 AM on February 2, 2015 [1 favorite]
posted by TomMelee at 10:56 AM on February 2, 2015 [1 favorite]
hades: A reel of paper tape which could possibly hold information that vanished when exposed to light? Sounds like a great place to store your one-time pad, except that 1915's a little early for one-time pads to have been in use.One-time pads date back at least to Julius Caesar's day.
posted by IAmBroom at 11:15 AM on February 2, 2015 [1 favorite]
I think that kiltedtaco is onto something, and agree with Kadin2048 that the artefact is incomplete. Good straight-on photos of all the sides of the box, along with a ruler, shared with other acrchives that might have the other pieces could be very productive. The position of the screw holes, alignment holes, vias for wires and paper tape, and the marks near the screw holes showing where the box was flush with brackets or other pieces are keys.
posted by Good Brain at 10:43 PM on February 2, 2015
posted by Good Brain at 10:43 PM on February 2, 2015
hades: Wikipedia has steered me wrong, then:Confusing wording: it wasn't described or named as such by Iivlivs Caesar; just used (by his men).First described by Frank Miller in 1882,[5][6] the one-time pad was re-invented in 1917 and patented a couple of years later.
posted by IAmBroom at 7:02 AM on February 3, 2015
Ivlivs Caesar.
Iivlivs was his less-well-known half-brother.
posted by IAmBroom at 7:09 AM on February 3, 2015
Iivlivs was his less-well-known half-brother.
posted by IAmBroom at 7:09 AM on February 3, 2015
Well, codebooks were common before 1919, and used carefully, a pre-distributed codebook can functionally be as secure as a one-time-pad. It's just that a one-time-pad, by definition, isn't ever reused: there is as much keying material as there is message content. Traditional codebooks are not decryptable mathematically on first usage, but via traffic analysis you can figure out what they mean if the codes are reused.
It's the decision to reuse the codes that makes them distinct from the modern concept of OTPs, although in practice many naval codebooks were big and complex enough that they were effectively unbreakable over short timescales—hence there was no small amount of attention paid to attempting to recover them from captured ships, and trying to make sure they were destroyed before surrendering (leaded bags thrown overboard, printing them on water-soluble paper, etc.).
The British captured at least one copy of the German Navy codebook in the early months of the war, and they weren't well-designed at all.
posted by Kadin2048 at 6:54 PM on February 4, 2015
It's the decision to reuse the codes that makes them distinct from the modern concept of OTPs, although in practice many naval codebooks were big and complex enough that they were effectively unbreakable over short timescales—hence there was no small amount of attention paid to attempting to recover them from captured ships, and trying to make sure they were destroyed before surrendering (leaded bags thrown overboard, printing them on water-soluble paper, etc.).
The British captured at least one copy of the German Navy codebook in the early months of the war, and they weren't well-designed at all.
posted by Kadin2048 at 6:54 PM on February 4, 2015
This thread is closed to new comments.
I think that understanding the precise nature and manner of the messages will help you speculate about the use of the reel. From the documents you've linked to, it's unclear precisely how the messages were transmitted, much less what messages there were. It seems that the "extra dots" graphic is just speculation. (The spacing theory seems more plausible to me, and it would be captured by a phonograph. If I were sending a hidden wireless message, though, I'd try to encode the message by subtly tinkering with the modulation of the signal, but that's just me.)
Complete speculation: If you've got enough bandwidth and transmission time you could certainly imagine transmitting images -- crude photographs. (This was done commercially only a few years later, e.g. http://en.wikipedia.org/wiki/Bartlane_cable_picture_transmission_system ). Maybe this device was involved in a similar system?
Another possible angle for you to develop is to use FOIA for records having to do with the seizure of the station. Secret Service (US Treasury), Navy, and Department of Commerce appear to have been involved; maybe there are other agencies you could contact to get their files?
posted by cgs06 at 6:34 AM on February 2, 2015 [1 favorite]