A Sixty-Eight Year Old Code

I admit, yesterday’s cipher, courtesy of the FBI and the Life Magazine photo archive, is neat, but not really all that exciting. That’s okay, because what I’ve got today is immensely awesome: An example of a pre-war German espionage code.

No, really…

The information for this picture says it was taken in 1941, and shows “An agent of the FBI decoding a message that was encrypted using the anagram encryption method”. Except… I’m fairly sure that’s wrong, and cool as the caption would be, were that true, reality, I believe you’ll find, is even better.

The date is July 24th, 1940. The place, the New York City Field Office of the FBI. The memo, from Director Hoover, is short and cryptic:

CLEARANCE AUTHORIZED FOR FOLLOWING MESSAGE: “DUNN SAYS FOUR BATTLESHIPS AND TEN DESTROYERS OF TEXAS CLASS GOING TO CARIBBEAN TO SCOUT FOR ENGLISH WITH SIX HUNDRED RESERVE OFFICERS AND SIX HUNDRED ENLISTED MEN.”

HOOVER

Below that appears to be the same basic message… in German. Above, a number of five-character letter groups. What’s going on here? A lot more than meets the eye, that’s what.

Students of intelligence history may know that “Dunn” was the code-name given by the Germans to Fritz Duquesne, infamous linchpin of the Duquesne spy ring in the early part of WWII. Duquesne was “played” by an American double-agent named William Sebold, who was instrumental in crushing Duquesne’s – and Germany’s – espionage operations in this country.

What you see here is a photo that never should have been allowed to be taken, and one which provides an amazing, one-of-a-kind glimpse into the world of WWII espionage and counter-espionage. As far as I can tell, what is shown in this picture is an FBI agent in New York encrypting a message, passed from “DUNN” – Duquesne – through Sebold, prior to transmitting that message to Germany via shortwave radio. Unlike yesterday’s photo, which was clearly staged, this appears to be real cryptology at work. Even the message is most likely real: it refers to the Neutrality Patrol undertaken shortly before the outbreak of war.

What I think is really cool, though, is that the photo also shows the agent’s worksheet:

I know it’s a little difficult to read, and we’re only looking at a “message depth” of one, but I’m pretty confident that some clever reader out there can use this photo to re-create the code in question. Admittedly, doing so is pointless, except as an academic exercise, but consider: how many other opportunities are budding cryptologists going to get to examine and play with a pre-war German espionage code?

Published in: Geekiness, General, History, Security | on November 26th, 2008| 65 Comments »

You can leave a response, or trackback from your own site.

65 Comments Leave a comment.

  1. On 11/26/2008 at 4:00 pm rmi Said:

    decoded text doesn’t make sense. there was no “texas class” for destroyers (then state names were reserved for battleships). and the “four battleships” cannot refer to the actual texas class battleships as there were only two: texas and new york.

  2. On 11/26/2008 at 4:13 pm Nemo Said:

    RMI: Yes, but the “texas class” battleships, like most of the other “neutrality patrol” vessels, were WWI-vintage, and I suspect that that’s what was trying to be conveyed: this is another, routine movement of old, outdated warships, not a deployment by modern vessels of the active fleet.

    You also have to remember that Duquesne was apparently communicating with Germany indirectly, through Sebold and the FBI; the latter were in a perfect position to play man-in-the-middle attacks against both sides – in this case, by sending Germany a slightly confusing, slightly confused report.

  3. On 11/28/2008 at 9:02 am Thiago Jung Bauermann Said:

    Hi,

    This is very interesting indeed, but it annoys me when you say that this is “pre-war” material. World War II started in 1939 (as Wikipedia itself says). To consider that it began when the US entered the war is a too US-centered worldview.

    (I hope you don’t receive this in duplicate. I had problems submitting it the first time.)

  4. On 11/28/2008 at 10:23 am jim Said:

    The German, as best I can make it out, reads: DUNN SAGT VIER KRIEGSSCHIFFE UND ZEHN ZERSTOERER KLASSE TEXAS NACH KARAIBISCHEN MEER FUER ENGLAENDER ZU SPAENEN MIT SECHS NULL NULL RESERVEOFFIZIER UND MANNSCHAFT SECHS NULL NULL

    I’m not quite sure about “spänen”–my German isn’t very good, but I think it means something like “scour”, so it would make sense in context.

  5. On 11/28/2008 at 10:26 am Jeff Dege Said:

    Are you sure that’s not anagramming?

    That looks very much like the hat diagram developed during the cryptanalysis of an irregular columnar transposition cipher.

  6. On 11/28/2008 at 11:53 am jim Said:

    It looks to be just a transposition cypher. The “key” is the arrangement of columns from 1-18. You write the message down with 18 columns across, then read down each column in the order given by the key, grouping in 5s. I suspect the caption is actually correct … you wouldn’t need such a complicated worksheet and all those typed strips for encryption, but they would make it easier to get the columns lined up when decrypting, so I imagine the cryptographer is decrypting the message to make sure it was encrypted correctly before sending it.

  7. On 11/28/2008 at 12:55 pm Craig Said:

    There is a link to this page and some commentary and discussion of the cipher at Bruce Schneier’s blog: http://www.schneier.com/blog/archives/2008/11/1941_pencil-and.html

    Thanks for posting this. It’s very cool.

  8. On 11/28/2008 at 1:00 pm Nemo Said:

    Thiago: It’s not meant to be Americentric: Duquesne entered German service in February 1939, and I’m willing to bet that the cipher in question dates back earlier than that, based on what I know of wartime crypto history. As I recall, at the time (1939-1941) the high-level military and diplomatic traffic was encrypted with Enigma machines, and everything else used codes that dated back to the early 1930s, or before (and were fairly easily broken by, e.g., Bletchley Park.)

    Hence, “pre-war”, as opposed to a later, military, war-time code.

  9. On 11/28/2008 at 1:02 pm Juergen Said:

    The German message reads:

    DUNNSAGTVIERKRIEGS
    SCHIFFEUNDZEHNZERS
    TOERERKLASSETEXASN
    ACHKARAIBISCHEMMEE
    RFUERENGLAENDERZUS
    PAEHENMITSECHSNULL
    NULLRESERVEOFFIZIE
    RENUNDMANNSCHAFTSE
    CHSNULLNULL

    Hope this helps

    Jürgen

  10. On 11/28/2008 at 1:49 pm Jeff Dege Said:

    “but they would make it easier to get the columns lined up when decrypting, so I imagine the cryptographer is decrypting the message to make sure it was encrypted correctly before sending it.”

    Remember when the article was written. “Decrypting” was used for both the process of reversing the encryption process so as to recover the plaintext, using the key, and of cryptanalyzing the cipher so as to recover the plaintext without knowing the key.

    From the photo, it seems the cryptanalyst has identified the cipher as an irregular columnar transposition, and has identified the width. The result is a grid of rows and columns, but with the columns having different lengths. He doesn’t know which columns have which lengths, so he extends the top of the grid by copying letters from the bottom. From these he generates strips for each column and then anagrams.

    IIRC, the process was described in Helen Gaines “Elementary Cryptanalysis”, which was published in 1939. Which makes me suspect that the worksheet – though not perhaps the message – was constructed for the news report. My guess is they made up an example of the cryptanalysis of one of the more sophisticated ciphers for which solutions were already publicly available. It seems unlikely that anyone was actually using a single transposition cipher in 1941 for anything serious in 1941.

    Read section 12-3 of FM 34-40-2:

    http://www.umich.edu/~umich/fm-34-40-2/ch12.pdf

  11. On 11/28/2008 at 2:37 pm Thiago Jung Bauermann Said:

    Nemo: Thanks for the explanation. I stand corrected.

  12. On 11/29/2008 at 4:21 am Jibberish Said:

    If you look for related photos you will find one of Hoover reading the same message, which to me implies it is all staged.

  13. On 11/29/2008 at 8:06 am jim Said:

    @Jeff Dege: Thanks. I expect you are right. The photo does look staged; it is too much work for ordinary encryption and decryption, but the fact that he’s got the original in front of him argues against his cracking it for real. It does seem like a rather questionable choice of material for a publicity shoot…

  14. On 11/29/2008 at 5:49 pm sehnsucht Said:

    Who’s taking the photo? Why? If it’s for real it must have been approved by someone. If so, why not release it earlier than 68 years ago? Are we sure it’s not a contemporary (faked) photo?

  15. On 11/29/2008 at 5:54 pm Nemo Said:

    sehnsucht: It was taken by a Life Magazine photographer, presumably for a feature in that publication ca. 1940-1941. I don’t know that it was ever published, however; it was recently made available – along with the rest of the Life photo archive – thanks to Google.

  16. On 11/29/2008 at 5:58 pm Hemo_jr Said:

    1939 is a Euro-centric date. WWII began with the Japanese invasion of Manchuria in 1931. However, if you consider that beginning of WWII to be the merging of the East Asian and European conflicts, 1941 is correct.

  17. On 11/29/2008 at 6:00 pm Biff Said:

    I believe it is staged. Any time an uncleared visitor, like a Life magazine interviewer, is in the area there is no way they would be allowed near real code unless they snuck in with a tiny camera. If it was a real interview they would have had someone pose for the photo in some sort of action position or near some amazing technology for effect.

  18. On 11/29/2008 at 6:05 pm Bob Said:

    Has anyone considered that the message may simply mean nothing? Perhaps it is just a created message with no historical factuality for teaching trainees how to encrypt/decrypt. Perhaps they were trained to know how to encrypt/decrypt both in English and in German.

  19. On 11/29/2008 at 6:15 pm Nemo Said:

    It’s somewhat obvious, but it’s worth remembering that if/when this image, or images like this, appeared in print in Life, documents shown, as here, wouldn’t have been legible; we’re fairly fortuitous to be able to look at these as fairly high-resolution scans, the equivalent of seeing these photos printed full-page, or even spread across two pages, in the original, intended magazine format.

    In other words, why go to the effort of fudging up some fake documents for a photo opportunity, when nobody’s ever going to be able to read the originals in print?

    Also, consider: if the whole thing were faked, why would the addressee be covered with a slip of paper, on the message from Hoover?

    They were security conscious – exactly as much as they had to be, at the time, IMO.

  20. On 11/29/2008 at 6:32 pm See Said:

    Biff:

    Not really… the Sino-Japanese war started in 1931, it became part of WWII either (depending on your view) when Japan attacked a few French colonies in 1940 (that is usually considered just part of the Sino-Japanese war), or in 1941 when war was actually declared between Japan and the allies as part of the same conflict that already opposed them to Germany. A war becomes a *World* War only once it involves countries on several continents.

  21. On 11/29/2008 at 7:21 pm Mick Mankford Said:

    It appears that the message was deciphered using the reverse narration technique, which was later found to be extremely vulnerable to crosswise hash attempts. This, as one could imagine, would render it useless for safeguarding national secrets.

  22. On 11/29/2008 at 7:38 pm Bill P Said:

    My guess is that we need to see who comments on this article so we can “recruit” you perhaps? Remember when “overlord” was published in a crossword puzzle before the event occurred? There is more espionage going on now than at any time in the history of the world. We need you help.. fin.

  23. On 11/29/2008 at 7:41 pm a_random_person Said:

    It’s somewhat obvious, but it’s worth remembering that if/when this image, or images like this, appeared in print in Life, documents shown, as here, wouldn’t have been legible; we’re fairly fortuitous to be able to look at these as fairly high-resolution scans, the equivalent of seeing these photos printed full-page, or even spread across two pages, in the original, intended magazine format.

    In other words, why go to the effort of fudging up some fake documents for a photo opportunity, when nobody’s ever going to be able to read the originals in print?

    Also, consider: if the whole thing were faked, why would the addressee be covered with a slip of paper, on the message from Hoover?

    They were security conscious – exactly as much as they had to be, at the time, IMO.

    ———————-

    If it was a posed shot, why make it too easy to determine the legitimacy (or lack thereof — a fake)? Of course there would be a strip covering the address, to make it seem more legit.

  24. On 11/29/2008 at 8:07 pm nick Said:

    I just finished reading Richard O’Kanes book ‘Clear the Bridge!’, about the war patrols of the USS Tang. He describes his XO’s decrypting received messages using this technique, in 43. They had machines that would do it, but they weren’t allowed on submarines, in case the sub sunk in shallow water, and the machines retrieved by the Japanese.

    The message was send in 5 letter words, and decrypted using strips of paper. Words were added to the message to make the columns line up, and the people decoding would have to discard them. He says that during pre-war years, officers were tested and expected to decipher the messages quickly, almost in lieu of being correct.

  25. On 11/29/2008 at 9:46 pm Keith Said:

    @Jim re:”It looks to be just a transposition cipher.” — true, but the simplest transposition cipher wouldn’t need all that’s going on in the middle of the page. Not that there is much going on there, but note how each column to the right “picks up” or repeats the following numbers of letters from the “bottom” of the preceding column: 0,1,2,3,4,5,6, then 7,7,7,7,7,6,5,4,3,2,1. So there is a bit more “scrambling” of the message than with the simplest columnar transposition cipher.

  26. On 11/29/2008 at 10:31 pm Charles Said:

    There’s another possibility for why this picture exists. You say it’s “a photo that never should have been allowed to be taken,” which is true unless it was intended for training material. It’s possible that this photo was intended for training cryptanalysts, to show them the layout for the grid, the strips, etc..

  27. On 11/29/2008 at 10:47 pm Keith Said:

    Note the “bleed-through” text in German from the opposite side of the page — just above the strips at the bottom of the page — I didn’t get the last word, but the text is something like, “bei schluessel system meinen sie das system welches sie inihr(?)” which Google translates as “For keys to my system the system which they…”

  28. On 11/29/2008 at 11:50 pm Matti Airas Said:

    Regarding suggestions that the photo would have been staged: I don’t think that is possible, due to one very obvious detail–the message mentions Dunn. There is no way the FBI would have intentionally revealed that they know anything about the Duquesne spy ring. I don’t believe such key information would have appeared in any training messages, either. So, the blog post is correct in stating that the photo should never have been allowed to be taken. There was just too much at stake.

  29. On 11/30/2008 at 12:02 am Les Coffman Said:

    CLEARANCE AUTHORIZED FOR FOLLOWING MESSAGE: “DUNN SAYS HE SAVED A LOT OF MONEY ON CAR INSURANCE BY SWITCHING TO GEIKO.?

    HOOVER

  30. On 11/30/2008 at 12:28 am KingFlathead Said:

    LOL, before the US entered the war it was mainly the Battle of Germany and Japan Ruining Everyone Else Mostly Unopposed. You’re welcome, guys. That’s why we get to set the dates – winners write the history books.

  31. On 11/30/2008 at 12:39 am Jess Said:

    @random_person:
    Why would you cover up the address and not the coded war message?

    There just must be something else going on that we’re not aware of. Still, quite an interesting find.

  32. On 11/30/2008 at 1:06 am theGeekPirate Said:

    Finally, I’ve figured it out!

    The encrypted message, once decrypted, reads: “All your base are belong to us”.

  33. On 11/30/2008 at 2:39 am Blonco Bars Said:

    Google has a tough job translating German. “bei schluessel system meinen sie das system welches sie inihr(?)? rather means “with ‘key system’ do you refer to the system which you [...] in your …”. Because in German the important verb is mostly placed at the end of a sentence it is missing here and so [...] would be the proper English translation of this missing verb.

  34. On 11/30/2008 at 2:41 am aawood Said:

    @kingflathead:
    Before going on about who gets to write what in the history books, you may want to try reading one. And no, the “Medal Of Honor” games don’t count.

  35. On 11/30/2008 at 4:20 am Anon Said:

    1940 was not pre-war! It doesn’t only become a war when the Americans join in. One year into a six year war that killed over 70 million people you can be sure there were a hell of a lot of dead people already, tell them there was no war on.

  36. On 11/30/2008 at 5:54 am Bornheim Said:

    schlüssel system = method of encryption

    So, the translation could be:
    By key encryption they refer to their encryption method, which
    is used within…

    greetings from heidelberg

    Did i miss the link to a hires version, so i could look at it
    myself in PS? Thanks.

  37. On 11/30/2008 at 6:38 am Eion MacDonald Said:

    Folk, The worksheet is re-opened for display , see crease fold! This is a posed ‘display picture’ not an actual work session. Bleed through from back side also confirms this as it would be inside the fold. Pecil in on ‘non-work page!
    regards
    eionmac

  38. On 11/30/2008 at 6:51 am Wodin Said:

    Blonco Bars: OK, so that sounds like instructions for how to use a key for some purpose. e.g. to decrypt a message. i.e. it sounds a bit like training material for how to use the vertical strips of paper in the decryption of messages.

  39. On 11/30/2008 at 7:44 am Arnold Reinhold Said:

    Regarding the comment by nick on the book “Clear the Bridge”: the cipher O’Kanes refers to is likely a strip cipher, which looks something like what is depicted in the article but is different in operation. The strips in the Life photo were made by the cryptanalyst to aid in unscrambling the shown cipher’s transposition by anagramming. The cryptanalyst tried various arrangements of the strips until words began to appear on some row. That gave him information on the key used to encipher the entire message.

    You can see a U.S. strip cipher of the type use in WWII subs on display in the NSA museum. There is a picture of that exhibit, with an explanation of how the cipher works, in the Wikipedia article titled “M-94″.

  40. On 11/30/2008 at 7:51 am HvdH Said:

    “bei schluessel system meinen sie das system welches sie inihr(?)”
    means “with ‘key system’ do you refer to the system which you […] in your …”

    In the german text “bei” is wrong, “mit” would be correct. This is a typical mistake of native english speakers writing in german.

  41. On 11/30/2008 at 9:13 am K Said:

    “LOL, before the US entered the war it was mainly the Battle of Germany and Japan Ruining Everyone Else Mostly Unopposed.”

    Except by the British Empire (almost 1/4 of the world’s population across 1/4 of the world, from the Pacific across Asia (the Japanese invaded British Crown holdings in Asia, remember, in the early 30s), Africa, and Canada [http://en.wikipedia.org/wiki/British_empire]), the French Empire (including Asian and African nations), the Poles (gunned down in their thousands), and the Soviets (in their millions). Just one or two people lying down and letting the Nazis beat them up while they waited listlessly for America to save them. Or something.

    “You’re welcome, guys. That’s why we get to set the dates – winners write the history books.”

    The Yanks did fine work when they turned up. However, I’d put better money on it being the Soviets who really swung it for us when the chips were down; but as the Cold War began towards the end of WWII, strategy and the American insistence that Britain pay their entire war debt immediately in order to cripple the Empire (then beginning to crumble under the weight of being the largest empire ever known) and put us in a secondary position in the Anglo-American Cold War alliance, ensured that America got to squander the gratitude of their allies by banging on about how everyone needed them. Not pretty, and not really terribly wise, but sic transit gloria mundi and all that.

  42. On 11/30/2008 at 9:20 am Jeff Dege Said:

    “He describes his XO’s decrypting received messages using this technique, in 43.”

    I am certain that this is a staged cryptanalysis of an incomplete columnar transposition. I think it’s possible that the inverted grid that is leaking through from back of the paper is where the message was originally encrypted.

    You don’t use strips with this cipher when enciphering or deciphering, you only need the strips if you don’t know the key.

    If someone wrote of using strips in enciphering or deciphering in 1943, I’d expect they were using the M-138 strip cipher device, which was a polyalphabetic substitution cipher — fundamentally the same as Thomas Jefferson’s wheel cipher, except using paper strips instead of wooden disks.

    http://en.wikipedia.org/wiki/M-94

    http://en.wikipedia.org/wiki/Bazeries_cylinder

  43. On 11/30/2008 at 9:38 am Clarke Schneider Said:

    Note that six hundred reserve officers and six hundred enlisted men for 4 battleships and 10 destroyers must mean something else, since a total of 1200 could not operate the ships specified.

  44. On 11/30/2008 at 10:23 am Seo expert Said:

    This is pretty neat, to find such a document, you’d be amazed at what you could find via Google because of people not being aware that their documents can resurface even if they are not linked from anywhere.

  45. On 11/30/2008 at 10:23 am Jeff Dege Said:

    “the simplest transposition cipher wouldn’t need all that’s going on in the middle of the page. Not that there is much going on there, but note how each column to the right “picks up? or repeats the following numbers of letters from the “bottom? of the preceding column: 0,1,2,3,4,5,6, then 7,7,7,7,7,6,5,4,3,2,1. So there is a bit more “scrambling? of the message than with the simplest columnar transposition cipher.”

    That complexity, the folding over of characters from one column to the next, is necessary only because the cryptanalyst doesn’t know which columns are of which length.

    A complete columnar transposition requires that the message be padded out to form a complete rectangle. An incomplete columnar transposition allows the last row to be incomplete. To anagram, the cryptanalyst has to guess the number of columns. There are statistical methods that can inform this guess, but it’s still a guess. In this case, he guessed (correctly) 18 columns.

    With that assumption, he knows from the length of the message that the last row has ten filled characters and eight empty ones. But he doesn’t know which columns are empty in the last row, so he has to build the “hat” at the top by adding the letters that might be at the top of each column and to draw the line at the bottom to mark the letters that might not be at the bottom of each column.

    If anyone has a copy of Helen Gaines’ “Elementary Cryptanalysis”, (reprinted in 1956 as “Cryptanalysis”) the use of hat diagrams is discussed beginning on p. 43. Helen Fouche Gaines was PICCOLA in the American Cryptogram Association, and “Elementary Cryptanalysis” was the ACA’s first published textbook. This cipher — and this technique for cracking it — has been a regular in the “Cipher Exchange” section of the ACA’s bimonthly newsletter (The Cryptogram) since the late 30′s.

  46. On 11/30/2008 at 12:40 pm john la berge Said:

    interesting letter and comments. consider this though
    1)would hoover haver added his /her own name to a document encrypted or otherwisae, i think given his /her previous actions that a most unlikely occurence
    2 the system used to lay out the cypher for rncoding and decoding is remiscent of the charts used pre and post war for the development of great grand master keying systems as had beeen developed by several lockmaking companies on both sides of the conflict note the frequent references to keys and keying systems especially when the standard u.s. lock system had a sipin level of keying the european had and still do retain a five and seven pin system cross ref that to the number of mariners in total and rank
    3 put the phot under a clearer light you might find the original paper has a different chemical composition than that used by big yellow perhaps iilfobraome then determine by if they are available when that type of paper was used.
    4 read the unauthorized versions of the histories of i.t.t. and i.b.m.
    better still it might be worthwile doing a bit of nosing around to consider as issuspect there may have been pre and post war the relationship between dupont vanden neymours and the f.b.i. to consider when looking at the photos
    5 how did the writer of the document arrive at the class of ships he/she does is interesting, someone suggested ww 1 influence, mayhap more research will offer a different position to consider
    6 and last may i suggest readers look for a better grade of german translation tools than the ones used by google for this event.

    i am not suggesting or claiming the photo was staged just asking you to consider and reply in aggreemnet and why you do or dont consider my observations and opinions on the photo as valid.

  47. On 11/30/2008 at 1:07 pm BlindWanderer Said:

    I’m inclined to say that it was contrived for the photograph but that they didn’t tell Life. That way if the images were printed, the Germans would see them, would waist time deciding if they were authentic and if they should try acquiring them. Give them a red herring and if they go for it, they end up on a wild goose chase.

  48. On 11/30/2008 at 2:29 pm Bruno Said:

    Was it a harmless bit of B.S. to make a good picture for a very popular national magazine? No, it was either a clever ruse upon those bumbling Germans, or a picture that NEVER should have been taken! Plus, the F.B.I. – which back then had quite the time nabbing Dillinger – was handling the pre-war “espionage and counter-espionage”, not the military.

  49. On 11/30/2008 at 4:38 pm Josh Said:

    “…winners get to write the history books”

    Thats just typical arrogant american BS. You obviously have never read a history book yourself that wasnt written by a fellow insecure american trying to make his history-penis bigger. Not to knock the great help the americans provided in WWII, but i might point out that there was great conflict before you joined. If anything i find it pathetic that the country turned a blind eye to the combat taking place around them, and only did something when THEY got attacked. But again, thats typical…you dont care much for anyone other than yourselves until its in your best interest to….looks like not much has changed either. And just because “winners” write the “history” books, doesnt mean that those history books are correct…i would sooner believe in the validity of a history book bought in Germany than one bought in Georgia or wherever the hell your ignorant and pompus ego hails from.

    But hey, take pride in the fact that you’re a “winner”. As a great man once said, a nation must first be destroyed from within before it can be destroyed by others… doesnt look like america is heading down that path now…or does it? I wonder what the history books will read in another 68 years….and just who the “winners” are then.

  50. On 11/30/2008 at 6:57 pm Keith Said:

    re: “bei schluessel system meinen sie das system welches sie inihr(?)?

    I know Google does not give the best translation — I offered it so that anyone can take a look for themselves — but the German isn’t the best in the first place, as HvdH pointed out — the original German for both the “secret message” and the text from the backside have obviously been loosely (poorly?) translated from English.

    There isn’t a “higher res” version of the picture — just copy and paste the image above into a viewer like IrfanView or similar, do a “Horizontal Flip” then a “Vertical Flip” and zoom in a bit and you can read the text that bled through from the backside of the paper – no magic.

    I think we are all looking to hard to find some magic here — this text from the opposite side of the page seems to show that this was not anything of high value, more likely a practice sheet or a Crypto 201 class — we might get sidetracked by the name “Dunn” or otherwise read in too much of our own information into the text.

  51. On 11/30/2008 at 9:45 pm Mark Said:

    I am not a student of crypto. I am a techno-history buff. I found the picture and everyone’s clues to the code fascinating. Being a history buff I checked out the links about Dunn and the Ring and this is what I am inclined to believe about the picture and the code.

    The code is undoubtedly pre 1939 German technology as many posters have pointed out. According to wikipedia the code was given to Sebold on microfilm in ’39 after he “reported to Hamburg, Germany, where he was instructed in such areas as preparing coded messages and microphotographs.”

    The picture was taken by a Life photographer. That means it was staged, and I propose well lit too. The Life archives were most probably scanned from original negatives and are uncropped. Since the picture is nearly square I believe this picture was taken by a twin lense reflex camera, a very popular camera type made by multiple manufactures. In order to achieve the depth of field displayed in the image good lighting would have been required or at very least a tripod.

    This picture was meant to exist. Why is the question. Possible theories are: public relations, war propaganda, a calling card to German Intel, or maybe the very evidence that busted them. The picture shows a complete message in all of its forms including intermediary steps and approved by J. Edgar Hoover for transmission. There is nothing in the picture that would have been a secret after the trial was ended on December 13, 1941. This was the week after Pearl Harbor and the propaganda machine was just getting started in the US. However, if it was taken before then it would have been a secretly staged photo shoot directed by the FBI with a Life photographer. Life had a talent for telling stories with photographs to the masses. The same could be applied to 12 jury members.

    Is there anyway to get the exact date of the photograph?

    Mark

  52. On 12/1/2008 at 12:47 am Nemo Said:

    Mark: Well-lit? Not really; there’s one flashbulb to camera right, in a bare reflector, and that’s it. In other words, pretty standard photojournalism stuff, for the period. No “good” lighting, or even a tripod, required – just that old workhorse, the #5 flashbulb.

  53. On 12/1/2008 at 11:11 am Mattias Said:

    Found the same message in english on this photo as well: http://images.google.com/hosted/life/l?q=encryption+source:life&imgurl=5f69a45a43a73dcb

    Perhaps these photos (and more?) were parts of a series of photos to show how the FBI worked. Why Hoover is reading the english message, and then have it encrypted into German might be interesting, but I would say it’s staged.

  54. On 12/1/2008 at 2:13 pm Keith Said:

    Did you notice there are two shots of the same scene?
    Here is the other one:
    http://images.google.com/hosted/life/f?q=encryption+source:life&imgurl=50ebffa0598db4bc

  55. On 12/1/2008 at 2:42 pm steve Said:

    # On 11/30/2008 at 9:45 pm Mark Said:

    Is there anyway to get the exact date of the photograph?

    Mark

    photo have any EXIF data? :p

  56. On 12/1/2008 at 3:16 pm M Said:

    “Be sure to drink your Ovaltine”

  57. On 12/1/2008 at 4:56 pm Mark Said:

    Keith, photographers typically take more than picture of the same thing. The other picture that you found could have been the the B shot, or maybe the A shot. Apparently everything went into the archives.

    Nemo, you are indeed correct about the flash. The second picture that Keith found is the proof.

    Steve, wouldn’t that data be for the scan and not the negative?

    Mark

  58. On 12/2/2008 at 2:44 pm Pijotre Said:

    its not Spaenen its Spaehen (or rather spähen) it means in this context “to scout out” for english in the carribean sea

  59. On 12/2/2008 at 10:03 pm JohnR Said:

    My late mother was a cryptanalyst for the FBI during WWII. Recently, while cleaning out some of her stuff I found her notes taken during her code training. If I have time, I’ll look through it and see if this method was mentioned.

  60. On 12/8/2008 at 9:26 am Jeff Dege Said:

    I’m revisiting this thread to admit I was wrong.

    I said, in earlier comments, that I couldn’t believe that the Germans would have been using simple transposition, as late as WWII.

    I was reading David Kahn’s “The Reader of Gentlemen’s Mail”, over the weekend. It’s a biography of Herbert Yardley, of “The American Black Chamber”. In 1941, Yardley was working for the Canadians, helping them set up a crypto bureau. From page 206:

    “The messages Yardley was solving were to and from German spies in South America. They were encrypted in a transposition cipher. Because it had to be easy enough for spies to use, the cipher was so simple that code puzzlers had long amused themselves with it.”

    My assumption, that no German spy ring would use a cryptosystem that was as trivial to break as a simple transposition, is clearly incorrect.

  61. On 12/9/2008 at 2:28 pm Peter Winkler Said:

    The German also doesn’t sound like real German to me. ‘fuer Englaender zu spaehen’ is not really German, more like a literal translation from English to German. A native German speaker would use ‘beobachten’ or ‘observieren’. also ‘Dunn sagt’ is a literal translation of ‘Dunn says’, but you wouldn’t use that in German either, you would say ‘Dunn berichtet’. Since making messages believable (faking the source), I think this is just something thrown together for the news report.

  62. On 1/14/2009 at 3:39 am Dan Fruchey Said:

    I just came across this thread and thought I would share what I know as I’ve been researching the Duquesne Spy Ring for almost a decade. This is a real message sent to Germany by the double agent William Sebold and his FBI handlers, a portion of a longer message that read:

    “Dunn says four battleships and ten destroyers, Texas class, going to Caribbean to scout for English; 600 reserve officers and 600 enlisted men. The Todd Dry Dock have contract to put anti-bomb decks on U.S. destroyers, 20 new destroyers, 20 from World War; half with laminated decks; explosion takes place before bomb reaches lamination. Conservative papers call for union between England and U.S.A. Articles were written to be published when Roosevelt nominated. British Embassy, United States State Department, British Bureau of Information, drew up articles, and State Department O.K’d them to publishers. Greetings.”

    This message was introduced as evidence in court (exhibit 14-K if you’re keeping track) and the sent date in the FBI record agrees with that in the photo (July 24, 1940). This was message #34 in a string of hundreds that were sent to Germany.

    How accurate is the information? Duquesne produced some astonishingly accurate intelligence that had to be edited by the FBI to protect secrets before being sent overseas but he also wasn’t above inventing material to insure a payday. Just two days before this message was sent Duquesne had been paid $250 in cash by Sebold for his services. Sebold even talked Duquesne into providing a receipt for the funds which he signed “Jimmy Dunn” an alias referred to in this message.

    The messages were usually broken into smaller sections for transmission, generally in logical blocks on a subject. The code was based upon a 1939 edition of Rachel Field’s novel entitled “All This And Heaven Too” (you need the UK edition to decipher the code, I picked up a copy from an online bookseller years ago).

    Virtually every author who has ever tried to describe the code has cited a September 9, 1941 article from the New York Times which stated,

    “The code system to which the novels were the keys was an ingeniously mystifying one, as it was outlined by the witness [Sebold].

    The solution lay in the date a radio message was to be received by the agent. The day and the month were totaled up and to this sum was added twenty. This indicated the page of the book on which the message was contained. Then, starting with the first line on the given page, the agent worked up and down the page in a complicated series of squares until the message was decoded.?

    Yes, they still really used these types of ciphers in those days. They recruited Sebold and many others in 1939 as they were reactivating elements of the German Military Intelligence apparatus and often called up old spies who hadn’t sharpened their skills since World War I. They taught what they knew and as it became apparent that someone had caught on they introduced new tricks. When caught, three of the spies in this case surrendered books used as the basis for their code. The technique was almost identical in each case and the same batch of spymasters trained many of the agents sent to South America.

    I have copies of the original prosecution files which detail the encoding method used but the prosecuting attorney crossed out all of the pages, believing the cipher too confusing to present in court to a jury. I have a few samples from FBI and court files which I ran through the book and their explanation is accurate but the information was a real pain to encode and decode. Typically two FBI agents would work with Sebold to create the messages and even then mistakes were made. The day after this message was sent Sebold received an administrative message from Germany to try to help correct some of the mistakes that were occurring in preparing messages.

    If I had to guess I would say that someone pulled out old documents used in preparing the court case and used a genuine artifact for the article. All the spies in this case had already been arrested in the summer of 1941 and the FBI probably felt comfortable flaunting the real thing as they trumpeted their success over the German secret service during prosecution later that year.

    As some of you have noted the odd syntax and speculated that this was a staged message let me assure you it was real and add an explanation along those lines that might be helpful. Some of the FBI agents who helped encode and decode messages were Americans who had been taught German for business reasons or were second or third generation German-Americans who had a middling grasp of the language at best. Sometimes the spies who gave Sebold the messages didn’t even understand German and the info was copied verbatim from reports written in English without any attempt at harmonizing the content. Some of the spymasters who sent the messages from Germany were more American than German, speaking German with a notable American accent, after having spent a good portion of their adult years in the U.S. They returned to the Fatherland with promises of rewards or a better life and were sometimes more comfortable composing messages in English than German. In one instance an individual associated with the case, who was still living in Germany long after the war, wrote his memoirs in English and his manuscript had to be translated into German for publication.

    Thanks for sharing this interesting artifact from a time long ago!

  63. On 4/15/2009 at 9:45 pm DpwnShift Said:

    Wow, fascinating picture and story! Also, thank you Dan Fruchey for all the additional material- it’s great to know the backstory and subsequent events!

  64. On 1/7/2010 at 4:21 pm searranty Said:

    http://gamrr.com/images/evid_blue/misc/logo.png

    GamRR.com Dedicated To Gaming, Offering Everything From The Oldest Arcade Rom Downloads To The Latest PS3 Gaming News.

    NES Roms

  65. On 7/12/2013 at 5:04 pm Berlin Ear Man Said:

    Looks to be a double, incomplete columnar transposition with the ‘twist’ of the truncated cone on the second transposition block. You can see that the top block of text has numbers across the top row. These divide the text into columns. Clear text is written across the rows of the top block (beneath the number key) in the normal fashion, sans spaces until complete. The column labeled “1″ is taken top to bottom and written into the second block from the first column on the left side, again from top to bottom. Next ’2′ from top, vertically entered into bottom block next column to the right. The only irregularity I see is the truncated cone of text in the second transposition block, which I presume was there to make cryptanalysis much more difficult then working against two rectangles. I had heard of unusual transposition block shapes, but never encountered them personally.

    Double incomplete columnar transpositions WERE actually rather common in the days before the SOE & OSS finally began using one-time-pads for agent communications, so it should be no surprise that German agents were also sing similar cipher systems – form fits function, as they say. The SOE sent their ‘pianists’ off with FAR WORSE than this in the early days, often using ciphers systems which were little more than puzzles of a grade found in the Sunday paper! This raises the question of how many brave young people died horribly due to their trusting those who sent them behind enemy lines lugging a 45 pound radio transceiver in a suit case and low grade cipher systems, until later in the war when these were greatly improved.

    …_ ._

Leave a Comment