In WW2, Berlin-based company Heimsoeth & Rinke not only produced the Enigma, but also a cipher cylinder, about which as good as nothing is known. Can my readers help to research the history of this device?

In the Second World War, the Germans used or designed some 20 encryption machines, including the Enigma, the Lorenz machine and the T-52 (Geheimschreiber). If you want to know more, check the blog article about German WW2 encryption machines I published last year.


The HR cylinder

Among all these designs, one German encryption device from WW2 is especially puzzling: a cipher cylinder, about which as good as nothing is known. It bears a label of Berlin-based company Heimsoeth & Rinke, which is known to have manufactured the Enigma. I call this device “HR cylinder”. Only one copy of the HR cylinder is known to exist. It is owned by the National Cryptologic Museum in Fort Meade near Washington. Here is a picture I took there:

Heimsoeth-Zylinder (3)

Five years ago, I published my first blog post about the HR cylinder (in German). This article received many page hits and my readers published a few interesting comments. Anyway, the background of this device remained a mystery. It is not known when exactly it was built and who was supposed to use it. I am not aware of a written source that mentions the HR cylinder (except for a Cryptologia article by Louis Kruh and my book Codeknacker gegen Codemacher, but both contain no additional information).


Cipher cylinders

I don’t know how the HR cylinder was used. It probably worked like any other cipher cylinder (here’s a picture that explains how a text is encrypted). The history of cipher cylinders goes back to the 15th century, but this type of encryption gained popularity only after the First World War. The most successful cipher cylinder variant is the M-94, which was used by the US military in the 1920s:


The HR cylinder is bigger and includes more rings than the M-94, which was designed as a pocket device. Here is a closer shot:

Heimsoeth-Zylinder (4)

On the following picture, the name of the producer can be seen :

Heimsoeth-Zylinder (2)

Apart from the Enigma and the HR cylinder, I don’t know of any other cipher device Heimsoeth & Rinke produced.


A prototype?

The HR cylinder certainly was not designed as a high-security cryptograph and was not meant to encrypt long messages. Its cryptographic strength is probably lower than the one of the Enigma. As the HR cylinder doesn’t have a keyboard, encrypting a message with it is less convenient than with other encryption machines of the time. On the other hand, the HR cylinder was cheaper to produce and more robust than the Enigma and similar devices.

The HR cylinder of the National Cryptologic Museum comes in the following case:

Heimsoeth-Zylinder (5)

The fact that the HR cylinder owned by the National Cryptologic Museum is the only one known suggests that this copy is a prototype. If so, it must be an advanced prototype, as it bears a serial number, comes in a case and bears a producer label. It looks pretty mature to me.

Heimsoeth-Zylinder (1)

In the Cryptologia article mentioned above, Louis Kruh stated that the HR cylinder was probably not used by the military or the diplomacy. Then he wrote: “There remains (unlikely) RLM Forschungsamt, RSHA, Gestapo (Secret Police) and SD. Cryptographically mediocre institutions like Parteikanzlei (Nazi Party organizations), Reichskanzlei (Civil Administration), Railroad, Post Office which were denied unrestricted use of the Enigma and might have looked for other devices.”

Richard SantaColoma wrote: “This one looks to me… just a guess of course… that it was intended to be attached to some standard typewriter, and used as a ‘cipher guide’. The clamps at the ends point to this, as does the high number of wheels… maybe, one for each typewriter column? If this is correct, then the operation would be very slow, and manual. But also, it might make sense because the user would not have to carry a whole, big and heavy, typewriter to a remote location (or an enigma). They would only need to find a typewriter to use once they got there… and convert it to a cipher machine, with this. The accessory wheels could be adapters for different models?”


A few new questions

There are two interesting aspects I didn’t cover in my first blog post about the HR cylinder:

  • According to the German Wikipedia entry about Heimsoeth & Rinke, the devices produced by this company didn’t bear a producer label any more after 1940 (because knowing who produced such a device was an interesting information for a spy). Does this mean that the HR cylinder was built before 1940?

  • The HR cylinder owned by the National Cryptologic Museum bears two inscriptions: RDM 291 and 118. What does this mean?

Can a reader say more about this device?

Further reading: The Schlüsselrad: New information about a little known German WW2 cipher device


Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Kommentare (8)

  1. #1 Richard SantaColoma
    29. April 2019

    This one continues to drive me nuts. I actually open the high res pictures I took of it, once in awhile, and try finding some new clue in them.

    The confounding thing about it is that it is a contrasting combination of seemingly careful and complex engineering, but then with only the simplest of outward information. I mean, to show what it shows, it could be built much simpler than it is.

    I think this points to their being another layer of complexity, either within the unit, or when it is attached to a parent device. The problem with that, though, is that there does not seem to be any means of linking with another device… no gears or cogs, for instance, that might interface with another unit.

    And there doesn’t seem to be any electrical or mechanical linkage between the rings, which might cause them to interact with each other. At least, this cannot be seen from the outside.

    I do wonder if, if it were allowed, that a bit of dis assembly might offer an important clue. If there is certainly no connection between the rings, that leaves it as a display only. But if there are cogs or pins which cause the rings to interact in some way, with each other, then it would be valuable to know that, too.

    Maybe the museum would allow that, or perhaps someone on staff would be interested in trying to take a ring or two off, themselves.

  2. #2 Thomas
    29. April 2019

    This ist a pre-WWII device: It was produced between March 1936 and December 1938, since Heimsoeth und Rinke had the adress “Ludendorffstraße 6” only in this period,

  3. #3 Dan Durham
    29. April 2019

    It appears that the order of the characters on each wheel is different. Are the wheels numbered or marked in some other way to facilitate changing the order of the wheels?

  4. #4 David A Wilson
    30. April 2019

    How easy is it to change the order of the wheels?

  5. #5 Richard SantaColoma
    30. April 2019

    Dan wrote:
    “It appears that the order of the characters on each wheel is different.”

    Yes they are. And it seems that on the main wheels, the first half of the alphabet, to “m”, is all in red, and the second half, in white. Here is a close up I had taken of the device, in which you can see this, clearly:

    But also note that, on the left end, there are a set of four disks, two with different sets of numbers, but then next to two of the (seemingly) same main letter disks. I’d guess they serve some purpose different than the main row of disks, but?

    “Are the wheels numbered or marked in some other way to facilitate changing the order of the wheels?”

    I haven’t noted any such markings, but then such numbering marks could be on the sides of the disks, where they are hidden.

    One thing to note: In another picture I have, it shows the other side of that large knurled and plain aluminum ring shown above… and there is an embossed, red arrow pointing to one row of the letters on the disks.

  6. #6 Richard SantaColoma
    30. April 2019

    David wrote:

    “How easy is it to change the order of the wheels?”

    I don’t know if Klaus, or a staff at the museum, ever tried to take it apart, but of course I did not. However, the “kit” contains two unmarked disks, which the NSA labled, “2 spare spacer disks”. They are held under little canvas straps, with snaps, ready to use. This seems to imply that it can be disassembled by the user. Just a guess (again), but perhaps one or both of the aluminum ends are knurled so that one could get a grip on the, for unscrewing.

    One other thing to note: In my picture linked in the comment above, you can see one of the disks tilted slightly, and what looks like an aluminum pin in the side of it. The question would be the purpose of such a pin: Is it to fix the disks into a relationship once that relationship and order is chosen, so that they cannot move? Or is it a cog to actuate the neighboring disk, when they are rotated? That is, is this an active arrangement of disks, constantly advancing while being used; or is it a static display of the key of the moment?

  7. #7 Richard SantaColoma
    6. Juli 2019

    A (semi) Random Key Generator and Transmitter?

    I don’t believe the cylinder is used as a deciphering tool as other “cipher wheel” type machines are used. First of all, the operation would have to be manual, because there does not seem to be any sort of electrical contacts on the wheels; and also, there does not seem to be any kind of geared interface to allow the disks to be rotated by a machine (cipher, telex, typewriter). It is also large, and does not seem to be made for operation in the hands, nor do the wheels look easily turned even if mounted in another machine. For these reasons, it would just be too tedious and clumsy to use as a cipher wheel, turning the wheels for each letter of the message.

    I think this implies that the order of the disks would be set infrequently, and left in any one position for a time. But still, since the wheels are “knurled”, they are probably meant to be turned.

    Also, the letters on each wheel each seem to have a somewhat… seemingly, at least… random order. If this is so, then once the order and position of the wheels were set, without one knowing the order and position, even if an enemy had an identical device, they would find it very difficult, if not impossible, to determine the proper order and position on their (stolen) device: Even if they knew one string of characters, read along one row on the wheels, it would be hard to determine which wheel was in which position. This, because, say, W was 3 characters from B on one wheel, it may also be 3 characters from B on one or more other wheels, too. So an enemy would not know which wheel was being used in what position.

    This random order of individual alphabets on each wheel are, to me, another clue this is not a cipher cylinder in the traditional sense.

    Another feature of the device which I considered was the fact that it can be covered. Obviously, this is so “prying eyes” cannot easily see the order of the wheels, either in transport, or at its destination. But this would not matter so much if this were a cipher wheel, used to encrypt and decrypt, because the “secret part” of such a use is the key, not the device. Everyone can have, and see, a copy of a cipher wheel machine, but without the key, they are lost.

    So for these reasons, I wonder if this was a means of generating, then safely communicating, (semi) random short term keys, of up to 60 characters. The purpose may have been to solve the problem of communicating such keys over the radio or telegraph, because of the danger of interception. It could work like this:

    First, a central location prepares several cylinders identically: They order the wheels on the cylinder, close them up, and send them to the recipient’s locations. The serial numbers of each device are known, and kept track of, so that if any cylinder falls into the enemy’s hands, that wheel order will not be used again.

    Now say you want to transmit the daily… or hourly and so on… key to that day’s encrypted messages: You give a string of characters, which are NOT the key, they are just the order of ONE ROW of wheels. This can now be transmitted over the radio or telegraph, because an enemy… even if they have one of the devices, they do not know the proper order of the wheels.

    The intended recipients, however, have now set one row of their wheels all to that daily or hourly “string”. Once this is done, the daily key can be read off a designated numbered row… and this would be the purpose of those numbered and lettered wheels at the beginning of the cylinder. For this example, I’ll use only the first 10 lettered wheels, and the four letter/number wheels it does have. The recipients may receive a setting string of:

    29zy (row wheels), then “ofnyxdlzjz…”

    After all possessors of the device have set a row to that order, a new key can simply be chosen by designating an other row for it… say “20vv”. Then, the new daily/hourly key can be read off THAT row, and begin, perhaps, “gmsolactqd…”. And that would be the key used to encipher and decipher one or more messages. Of course it could be used for only one, and be very hard to decrypt by an enemy.

    Also, a new row could be designated with the same wheel settings, at any time, instantly changing the key when one did so. And of course, the whole key would not have to be transmitted… it never needs to be… only the row designator. And then, at any time, a new row setting can be transmitted again, and each recipient can set their device to the new row settings, and then it would be used for a multple number of new, 60 character keys.

    The fact that the device has one nice bar, with character positions numbered, would also fit with this daily/hourly key idea, I think, because a user can turn the whole assembly to that “aid”, more easily referring to the key without making a mistake.

    I chose my string examples above using this picture of the end of the device…

    … you can see the strings I chose on the #20 and #22 rows. So I suppose it is possible that it is these numbers, and not the four character string settings (such as “29zy) that designate the agreed-on key row. There is one row that is designated by an arrow… maybe this is the row everyone is to set to the daily string, to start. And also, perhaps, those smaller four wheels serve another purpose than key row designation, such as another level of encryption… such as a setting on the encryption machine used, or perhaps a radio frequency, or some other purpose.

    The dangers of using this system would be limited to an enemy capturing one device, then knowning the setting of the order of the wheels. They would then be able to set THEIR machine to the daily/hourly string. But if the device was known to be captured, the order of the wheels could be changed on all other devices, and the random nature of the letter strings would not be repeatable, or easily repeatable. But they could be, over time, by noting the multiple distances to the other characters in a string… but then, the only way to know them would be to decrypt the message first, and so determine those distances. I am no cipher expert, but from what I understand, this would be very difficult or impossible… as this would be, in effect, a resettable one time pad key.

    If the above purpose is correct, this may have been useful for shorter messages, and short term ones, such as battlefield information with short term value. It could have been used instead of more complicated and highly secret cipher machines, perhaps distributed to locations in which the loss of such a machine would be of higher risk…. such as for a front line telegraph or radio operator, or maybe a naval or land gun battery.

    A further possibility, but one I doubt (even assuming any of this is correct), is that a central location could transmit a NEW ORDER of the wheels… but this would be very dangerous, because it is that order that would then open up all the keys, on all the rows, to any enemy who possessed one of the devices, and heard the “daily/hourly string” over the radio. But then this means that all cylinders would have to be recalled, and reset somehow… or the information on wheel order hand delivered somehow, or by some other secure means. I don’t know.

    I am no cipher expert, and perhaps a reader will readily see a problem with my suggestion. Or maybe it is part of the answer… that this is still a key generator of some kind, but used in a different manner. But maybe the ideas above would help with search terms that may help find the actual patent for this device, or other information, even if partially or wholly incorrect. Perhaps someone can think of another cipher, or telex-type device which would rely on the attachment or proximity of a 60 character key, for easy reference by a user?

  8. #8 Klaus Schmeh
    6. Juli 2019

    @Rich: Thank you very much for this explanation. Sounds very interesting. If your guess is correct, this device works in a similar way as the Reihenschieber. The Reihenschieber, a pseudo-random generator, is sometimes erroneously regarded as a cipher slide. Maybe, a similar mistake happens here: this cylindrical device is not a cipher cylinder but a pseudo-random generator.