2015-02-26

The Colmar Incident

The Colmar incident concerns a true episode in World War II which took place in and around Colmar, France, in which a top-secret cryptographic machine called a SIGABA (or an ECM Mark II) went missing. This event is now declassified and this blog post follows very closely a document published on the NSA website (author unknown). There is no originality here, but the story has the makings of a Hollywood movie and I only hope that this post is entertaining to others as well. Maybe a Hollywood scriptwriter!



caption: The ECM Mark II or SIGABA machine. source: wikipedia

The SIGABA machine was used during WWII by the US Army in its most highly classified operations. Unfortunately, it went missing around the time of the Yalta Conference, and allies were very concerned that the Nazi's had gotten a hold of the machine (Colmar is located in eastern France about 10 miles from the border with Germany).


caption: The Yalta Conference was held February 4-11, 1945, in the Black Sea, about 1500 miles east of Colmar.

The main duty to track down the machine was given to Army COL David Erskine, who, with Navy LT Grant Heilman, eventually found it in the Gresson River. This is the story of its recovery. Below, we follow almost verbatim the NSA document mentioned above [CI].


The Colmar Incident


Introduction


As the Allied armies were preparing for the final assault into Germany during the waning months of World War II, an event occurred which had the makings of a major disaster. An advancing U.S. Army division lost a SIGABA, the cryptographic machine used for the highest level of U.S. communications. This type of machine was a vital part of a world-wide combined (UK/US) communications system, and should it fall into enemy hands, the results could be devastating. All of the Allied war plans for the spring assault into Germany, for example, had been enciphered by SIGABA, as well as arrangements for President Roosevelt's pending trip
to Yalta. And it was the only highly secure U.S. cipher system available to many U.S. units in Europe at that time.

After news of the incident reached Arlington Hall, the most important question was whether the equipment had fallen into enemy hands. And, under wartime conditions, it had to be assumed, until proven otherwise, that the SIGABA machine had, in fact met this fate. As a consequence. there was a massive effort to fill this cryptographic void while the matter was being investigated. And the investigation itself, in addition to diverting men and equipment from the war effort, deeply involved persons at the highest levels of the U.S. and Allied governments and military establishments.


Events leading to the disappearance


On 6 February 1945, Headquarters, Communications Zone, European Theatre of Operations (Paris) dispatched an Urgent~Secret message to all strategic commands in Europe, informing their signal officers to suspend use of certain cryptographic systems and materials due to a possible compromise. An information copy of the message, which read as follows, was also sent to the War Department and Arlington Hall:


Suspend use immediately of editions 18 and 19, of systems 2201, 2222, and 2242, editions 10 and 11 of CCBP 0125 and 0126 and edition 2 of 514.

Grave possibility of compromise of all mentioned exists, all Commanders to be informed of possibility of enciphered traffic via radio in above systems being read by enemy. Reserve systems now being distributed.

Use only secure landline. courier, or reserve systems where available until security restored. Nets 12, 13. 14, 1S and 17 not affected as well as CCBP 0101, 0102, and 0103 by above. Inform all subordinate units concerned.

But it was not until 10 February, when the officer in charge of the War Department Code Center called persons at Arlington Hall to alert them that an extremely important message concerning the compromise had just been received, that any specific details became known concerning the systems involved in the compromise:


Parked truck belonging to two-eight infantry divisions containing following documents stolen in Colmar France: SIGABA, .... Use of all systems listed has been suspended in this theater and every possible action being taken to restore security. CCBP0131 has been distributed to all holders. Secure landline and courier being used in most cases. Full investigation of loss being made by Sixth Army Group. Will keep you informed of development and action taken.

A trans-Atlantic enciphered phone conversation on the 11th, between Arlington Hall Station and the Signal Intelligence Division, ETO, Paris, afforded additional information. It noted, among other things, that a 2.5-ton truck had been stolen from a city street in Colmar, France, containing the SIGABA and associated, and other, equipment and documents. It further noted that the SIGABA and these other materials were in a locked safe in the ``code room'' carried by the truck.


caption: A GMC 2 and a half ton US Army truck from WWII.

The theft involved the 28th Infantry Division, which was being transferred from Kayersburg, France, to the city of Colmar. Colmar had been evacuated by German troops several days before, and on the day of the theft the enemy was only about four miles to the north of the city.

The event had its beginning on 4 February, when the chief of the message center of the 28th Division's Signal company departed from Kayersburg, with an advance station of his cryptographic team and equipment, to set up a message center in Colmar. When the men arrived, they found there were no billets ready or even allocated to the signal company. Having found an unoccupied house at No. 16 Barbarassastrasse, and having obtained official permission to occupy it, they unloaded their cryptographic equipment and established communications.

The next day, 5 February, the second half of the team, with spare sets of cryptographic equipment, including a SIGABA, departed for Colmar. Upon arrival, the men located the first group. (Unfortunately, though the signal company's motor pool section, which had the responsibility for establishing a guarded truck park, had not yet arrived.) After locating the division's message center, which had been established by the advance party the day before, the men were told that their billets were at No. 16, and were instructed to park as near to the house as possible. Subsequently, the truck (with SIGABA and associated equipment) was parked for the night in front of No. 20 Barbarassastrasse, but without guards, as normally required.

The next morning, when two officers went to the truck, to make some repairs, it was gone. After reporting to the message center chief that the vehicle was missing, the two men began searching for the truck.

Note added: David Kahn (The Codebreakers, [K], page 510) gives another account: Two SIGABA guards parked the truck outside a brothel in Colmar on the evening of February 3 to visit some friendly ladies. When they emerged the truck was gone. One version of the story says these guards were sent to Leavenworth for the remainder of the war, and the pardoned later.

About 11:15a.m. they found the trailer to the truck abandoned on a dead-end road. Tire tracks at the scene indicated that the trailer had been unfastened, and the truck had been turned around and driven out of town. A quick inspection revealed nothing was missing from the trailer, which had contained only unclassified material.

G-2 (Intelligence staff) of the Division Headquarters in Colmar, coordinating with the Military Police, ordered a complete search of the area, and the Inspector General, 28th Division, ordered a thorough investigation of the responsible personnel. He also sought anyone who might have any information concerning the lost truck. This request had some results, revealing, among other things, that a radio car from the same company had been parked across the street from the 2.5-ton truck and that men of the company had made frequent trips to the car throughout the night, for the purpose of occasionally running the motor. However, these men had neither seen nor heard anything and, in fact, had taken no notice of the truck in question.

It was further directed that all units and divisions conduct officer-supervised searches to include all sheds, barns, woods, mountain areas, etc.. SHAEF (Note added: SHAEF - Supreme Headquarters, Allied Expeditionary Force - European Theater of Operation, General Eisenhower's Headquarters.) was also asked to conduct a complete theater-wide search and the Sixth Army Group directed all headquarters troops, and subordinate units as well, to inventory the motor numbers and inspect the content of all 2.5-ton trucks. Also, the help of the local police and the Surete Militaire (CIC's French counterpart) was promptly enlisted. Descriptions of the truck and safes, but not the contents, were given to the French officials. (American and British headquarters elements in Europe received more specific information about the contents of the truck.) General Eisenhower personally became involved and ordered that a vigorous investigation and search be made to locate the missing vehicle. With the possibility of the disclosure of all of the plans for the spring offensive, the Supreme Allied Commander in Europe gave top priority to the recovery of the truck containing the SIGABA; all Allied high commands participated in the search.

Note added: According to [K], Eisenhower ordered General Jacob L. Devers, commander of the 6th Army Group to find the missing safe(s). Devers assigned this to his chief intelligence officer Colonel David G. Erskine.


Crypto-equipment found


Extensive air and ground reconnaissance was conducted while motor and foot patrols continued their more intensive searches throughout the Sixth Army Group area during February and the early days of March. During this time, several reports of incidents involving the French came to light, casting suspicions that the culprits could very well be Frenchmen. One involved a truck, resembling the stolen 2.5-ton GMC, which had run through a road block at Montbilliard, 35 miles north of Colmar. On another occasion, on an overnight visit to Colmar, the Seventh Army's CIC Detachment had three jeeps stolen by French units. Also, as soon as it became known that Colmar was being occupied by Allied troops, French soldiers from various units not with the French and American divisions occupying the town, suddenly appeared on the scene and "appropriated''
whatever trucks and vehicles they could.

Although such activities strongly suggested that the French were involved with the theft, there was another equally suspicious - and far more ominous - possibility.

Because members of the 28th Division's signal company had been captured before 5 February by the Germans, the enemy could have known that such highly sensitive cryptographic material was in the Colmar area, and made a special effort to acquire it.

Although many reports continued to come in during February, the results were all negative. However, the vehicle inspections did turn up some enlightening facts. Stealing trucks from front-line positions was obviously common practice, as evidenced for example by the recovery of six 2.5-ton trucks found in the possession of the wrong units during the serial-number checks.

Four days later, on the morning of 9 March 1945, the truck was finally located, abandoned in a wooded area several miles north of the city of Rambervillers, approximately 45 miles northeast of Colmar. The bumpers of the truck had been removed, and the numbers had been painted out; otherwise, there was no other apparent damage to the truck. However, the safes containing the SIGABA and the other equipment were gone.

Later that same afternoon, a search party from the Deuxieme Bureau, II Corps, First French Army, discovered the upper half of the safe (designated SIGRINO, Chest CH-76), as well as a small company field safe, submerged in the Gressen River, a small mountain stream which runs between the towns of Chaenois and Scherweiller. The safes were immediately placed under guard and turned over to the Sixth Army Group. The upper half of the SIGRINO contained the SIGABA and other classified equipment. The lower half of the safe, which had not yet been recovered, contained instructions, pamphlets and rotors for the SIGABA.

Meanwhile, the search for the missing safe and equipment intensified. The river was dredged near the bridge where the first discovery had been made, and plans were even made to divert the water flow. Working on the suspicion that the safes were probably thrown into the river in early February, at a time when the water level was much higher, investigators theorized that the swift current would have carried them a little farther downstream. Therefore, river dredging efforts were concentrated a few yards downstream from the bridge.

While this operation continued, Headquarters, Sixth Army Group, began a thorough inspection of the recovered. equipment. The vital question was still whether the safes, rotors, and associated material could have been compromised. Fingerprinting and other investigative analyses were virtually impossible, because of the deteriorated condition of the mud·and-water-soaked equipment. However, after a thorough inspection, G-2 and Signal Intelligence decided that, since the settings in the safes had not been re-adjusted, and since there was no evidence suggesting that attempts had been made to open the safes, no compromise existed.

Note added: Examination proved that the rotors were set up in arrangement for 5 February. (This was, however. in violation of strict security rules for moving the machine, as rotors were never to be left in place when transporting such equipment.)

Finally, eleven days later, on 20 March, the lower safe was recovered from the river bottom about 100 feet downstream from the bridge.

Note added: The version of the story in [K] is similar: Erskine was joined by 2-star General Fay B. Prickett and Navy Lieutenant Grant Heilman. Based on a tip from a French source, Erskine and Heilman found the machine and safe(s) in a river.


source: Johnson [J]

Efforts to open the safe were unsuccessful because the combination was clogged with silt. Subsequently, the safe was sent to Paris where it was "opened'' at a "ceremony'' attended by at least half of the intelligence staff in Paris. Attempts were first made to open the safe with a torch, but, fortunately, the torch ran out of oxygen, after cutting about two inches, for when the safe was finally cut and drilled open later, long after the VIP guests had departed, it was found to contain two 55-pound thermite bombs and 14-pound blocks of TNT, all wired to detonators. It was also fortunate that these explosive had not been designed as a booby trap; they were simply the devices normally carried with cipher machines to destroy them if they were in danger of being captured by the enemy.


source: Johnson [J]

Suspects


Although no written records can be found regarding who had actually stolen the truck, an interview with an officer who was active in G-2 efforts at Headquarters, ETO, Paris, during the Colmar incident did uncover some interesting information. He stated that during the Inspector General's investigations, two French peasants commented that the American Government could find its truck in the "woods,'' and, further, that the "boxes'' in the truck had been thrown into the river. Although the woods and river were not identified by name, their description of the surrounding area was sufficient to permit the truck and equipment to be found and these two men were subsequently apprehended on the suspicion they participated in the theft. During their interrogation, however, it was learned from these Frenchmen that it, was actually their friend, a farmer, who had taken the truck to move his household furniture.

Note added: Kahn [K] relays a story told to Erskine that it was "borrowed'' by a French military chauffeur who lost his own truck. When the investigation started, he was afraid of getting caught, so dumped the boxes in a river.

The farmer's name was never disclosed by his friends; consequently, he was not apprehended for questioning. But in view of these revelations, the consensus at Sixth Army Group Headquarters was that the vehicle had been stolen for the value of the truck itself, and that the thieves were unaware that the truck contained highly valuable and sensitive cryptographic equipment.

Note added: According to Johnson [J], Erskine was awarded the Bronze Star and the Legion of Merit, and was eventually promoted to Chief of Staff for Intelligence for all the Army ground forces. He retired in 1952.


source: Johnson [J]

Bibliography:

[CI] "Colmar Incident,'' Colmar_Incident.pdf

[K] David Kahn, The Codebreakers, revised and updated edition, Scribner, 1996.

[J] Thomas M. Johnson, "Search for the stolen SIGABA", Army, Feb. 1962, pp 50-56.

See also the references in:
[SC] M. Stamp and W. Chan, "SIGABA: Cryptanalysis of the Full Keyspace", available as a pdf.

Introduction to the script for His Girl Friday

This is an edited version of an article first published in the MWA's KIH (part 1 is here, part 2 is here).

First, what is a script? It is a specially formatted way of telling a story with a visual emphasis. For the "grammatical" rules of screenplay structure, see Christopher Riley's The Hollywood Standard, David Troltter's The Screenwriter's Bible (now in the 6th edition), or his Ask Dr. Format, or try to find a free source on the internet, such as Deemer's site. Here is a very simple example:

Example



INT. JONES ELEMENTARY SCHOOL, CLASSROOM - DAY


School is in session and dozens of STUDENTS, including JANE (7, female), are working at their desks. First grade teacher MRS SMITH (40) is looking over the shoulder of Jane who is drawing something.

MRS SMITH

That's lovely! What are you drawing?

JANE

I'm drawing God!

MRS SMITH

Oh but deary, no one knows what God looks like.

JANE

They will in a minute.


In the beginning...


Odd formatting aside, the child Jane is drawing a picture in Mrs Smith's classroom. Pretty clear isn't it? Yes, the formatting is funny at first, but simply put, the basic idea is to tell a compelling story in a visual medium, and the formatting is there to efficiently give the cast and crew necessary information on how the scenes should be set up and performed.

Suppose you are interested in learning about movie scripts. The way I look at it, reading and writing a movie script is, in the end, not much different than reading or writing a short story or a novel. You need (a) a compelling story and (b) some knowledge of the genre you are writing in. If you are really new to this, the format might look strange at first, but you'll get used to it. Writing the story down is the hardest part, and besides, there are lots of computer programs out there that will help format it for you as a script (for example, both Trelby and Celtx are free).

How to start? Scott Myers, a very experience teacher and screenwriter, whose blog Go Into The Story is one of the most popular screenwriting blogs on the planet, has a recommendation that he expresses succinctly as "1-2-7-14":

  • Read 1 screenplay per week.
  • Watch 2 movies per week.
  • Write 7 pages per week.
  • Work 14 hours per week prepping a story.

You all know how to write and prep, but how do you find scripts to read? You would be amazed how easy it is to find good scripts for free on the internet - just google "(movie name) script". For example, if you google "Casablanca script" you will find the script that the Writers Guild of America voted as the best script of all time. Sadly, not every script ever written is on the internet but there are plenty out there to learn from and enjoy. Check out dailyscript.com or imsdb.com and try to find your favorite movie or TV show there.

Homework: watch a movie!


Here is an assignment: pausing as necessary, watch a movie while reading the script! If you want to get started, here are two movie \& script suggestions:


  1. His Girl Friday, by Howard Hawks, starring Cary Grant and Rosiland Russell. The film in public domain in the US and can be watched online, or downloaded to your computer, and the script is available online as well . All free, no hassles with registration
    or anything like that.

    Logline from imdb.com:
    A newspaper editor uses every trick in the book to keep his ace reporter ex-wife from remarrying.


    caption: Cary Grant and Rosalind Russell in His Girl Friday (source: Wikipedia, http://en.wikipedia.org/wiki/File:HisgirlFriday.jpg)

  2. Stanley Donen's Charade, starring Cary Grant and Audrey Hepburn. Like HGF above, this film in public domain in the US and can be watched online, or downloaded to your computer, and the script is available online as well. All free. Logline from imdb.com: Romance and suspense in Paris, as a woman is pursued by several men who want a fortune her murdered husband had stolen. Who can she trust?

    caption: Audrey Hepburn in Charade (source: Wikipedia, http://en.wikipedia.org/wiki/File:Charadehepburn.jpg)
    Interesting story about Peter Stone's script: Apparently, he and Marc Behm had tried to sell a script to Hollywood executives, and failed. They gave up and Peter Stone rewrote the script as a novel and sold it to Redbook (a women's magazine which still is published today), where it was serialized. Someone in Hollywood who reads Redbook liked it. They bought the book rights, from which Peter Stone wrote the script!



Note the two "loglines'' above. A logline is essentially a very concise description of the story and plot. Lots more examples are given in the Script Lab's Logline Library.


His Girl Friday


His Girl Friday (1940), a film in the public domain whose script is also online. It is a screwball comedy directed by Howard Hawks, starring Cary Grant and Rosiland Russell. This is one of my favorite movies and the rest of this post will talk more about that film and its script.

His Girl Friday is based on the 1928 play The Front Page, by Ben Hecht and Charles MacArthur (also made into several other movies). MacArthur won one Academy Award for writing (with Ben Hecht) and was nominated for two others. His brother is John D. MacArthur, co-founder of the John D. and Catherine T. MacArthur Foundation, the benefactor of the "genius awards". Ben Hecht, another widely admired writer, won two Academy Awards and was nominated for four others (all for writing).



FADE IN:

INT. ANTEROOM CLOSE SHOT SWITCHBOARD

Two telephone operators sit at switchboard busy plugging in and out answering calls.

1ST OPERATOR
This is the Morning Post... The City Room? Just a
moment, I'll connect you.
(plugs in call)

2ND OPERATOR
Morning Post... Sports Department? Just a moment --
(plugs in call)
CAMERA PULLS BACK to disclose the rest of the anteroom. To Camera left are the elevators ­at back wall directly behind switchboard are chairs and a table for visitors. Next to switchboard are stairs leading downward to the next floor. A waist­high iron grill with a gate in it separates the switchboard from the anteroom, a similar grill separating it again from the city room which stretches on beyond switchboard. At a table in the switchboard enclosure sits an office boy, about fifteen, doing a crossword puzzle. The big clock on the back wall shows that it is nearly one o'clock.

CLOSE SHOT OFFICE BOY

as he bends over paper. We catch a glimpse of the squares of a crossword puzzle.


Although the movie did not follow this verbatim, it is pretty close to what the first 20 seconds of the film. A few things are worth pointing out.

  • First, notice the font. For reasons not entirely clear to me, two and only two fonts are "allowed" - Times Roman and Courier (12 pt). No bold, italics, etc. Richard Walter, who teaches screenwriting at UCLA, reads and comments on so many scripts that he has deveoped a system of notation for "script mistakes" to make his suggestions for improvement more efficient. When Walter reads a script which has a font error he writes nofx in the margin, meaning no word processor "effects". Only what the manual typewriter can do is considered standard.
  • Second, notice the camera directions ("close shot", "camera pulls back"). This indicates that this script is a "shooting script". A shooting script is written after the original script is finalized and the film is moving into "pre-production." Basically, it gives the crew directions on what scenes to light and how the camera operator should get his shots. Please ignore these. Although the original script is preferable, we are lucky that any version of the script is available for this great film!
  • Third, notice that the movie begins with motion and fast-paced dialog. Charles Lederer, the screenwriter who adapted the play into the film script, is known for his pioneering role in screw-ball comedies. When Walter reads a script which is not moving fast enough (in proportion to the time and money a potential audience member spends watching it), he writes $? in the margin ("is this part of your script worth the money?"). The nice thing about His Girl Friday is that it moves quickly starting at the beginning, all the way to the end.
  • Fourth, I'd like to point out something that you might not have noticed. For each scene, there is generally a "slug line" or scene heading, such as


    INT. ANTEROOM CLOSE SHOT SWITCHBOARD

    (INT. means that the location is inside, "interior," as opposed to outside, or EXT.). This is generally followed by the "action lines" - a short scene description, which could include a description of any characters, at least those with a speaking part or some other important role, who first appear in that scene. For example,


    Two telephone OPERATORS sit at switchboard busy plugging in and out answering calls.

    However, in the version script we are looking at, the word "operators" is not in all upper case letters, as is typical for characters who first appear in that scene. I'm not sure why that is. (Maybe it was not the standard back in the 1940s or possibly the caps were left out in the shooting script?) In any case, after this first appearance, usually the character name does not appear in all caps again. However, this is not universal either for older scripts and some use caps for all characters. (For example, the online script by Peter Stone for Charade (1963), does this.)

These are some basic comments on the part of the Lederer script covering the first 20 seconds of the movie His Girl Friday. I hope this encourages you to watch the film, which you can view for free online.

2015-02-11

Elizebeth Friedman and the Lew Kim Yuen case

This is an undated transcription of a few (incomplete) pages of notes by ESF on the Lew Kim Yuen case. The memo (and other information in this blog post) can be found in Box 6, file 26 of the ESF Collection at the George C. Marshall Foundation library.

We worked under some pressure all the time trying to find out when the S/S Taythybius was due at Seattle carrying the very large amount of ''white stuff''-- heroin -- mentioned in the message. Corrections of letters-to-digits-to-Chinese-to-English had to be made because of errors, or ``garbles'' as it is known in the trade, after we had received an actual photostat of the original message seized in Portland, Oregon. But all that was possible to do was finished and turned over to the Bureau of Customs in Washington on February 12th. We later learned that the S/S Taythybius actually docked at Seattle on February 13th. The peculiar nature of the Chinese language, with its inability -- rather its complete lack of ability -- to express many sounds or even concepts makes for greater length of text in the Chinese version of a message. For example, the literal text of this message was perhaps five times the length of the final English translation, but this single message furnished to the Customs Agency forces in the Northwest area an enormous fund of information which could be linked in with the growing bulk of knowledge concerning the operation of narcotics smuggling by utilizing crew members of legitimate vessels operating in the Pacific.


Image source: Chris R. Pownall

My office determined that the letters were monoalphabetic encipherments for the digits 0 to 9:

Q U O N G S L A W K
1 2 3 4 5 6 7 8 9 0
I had known and had in my possession a Chinese commercial dictionary -- code-book, if you prefer. It is actually a dictionary of 10,000 Chinese characters commonly used in business or commercial cable and telegraph, because obviously the Chinese characters themselves cannot be send in Morse or International signals. Each of these Chinese characters in this abbreviated dictionary has four-digit number-groups assigned to it and when the message is telegraphed it is the number-groups which were used. In this case, a simple cipher was used to convert the digits of the code-book into letters for sending. After my office had converted the four-letter groups into four-digit numbers each number-group was found in the small Chinese dictionary and sit down in columnar fashion the better to study the message for translation.
The memo continues on another page which is missing from the file. Here is a typed version of the encrypted message:


Here is one page from ESF's notes on the decryption of the message:

Here is the final decryption: