Talk:Jefferson disk

Cipher System

The link to "cipher system" now points to a Swedish melodic death metal band. I don't think that's what's desired.129.42.208.183 (talk) 18:16, 18 October 2013 (UTC)Dave

Merge discussion

I don't think the two articles need to be merged. There's a small amount of duplication but space is not short. The two articles are adequately linked so that anyone reading one and interested in the other can easily find it. --Wtshymanski 03:38, 17 Apr 2005 (UTC)

Well, both articles are describing the exact same topic — a merge seems necessary. — Matt Crypto 10:16, 17 Apr 2005 (UTC)

Someone who feels more bold than I can merge them. The Bazeries cylinder article is a little on the long side, but I suppose the short paragraph that uniquely describes the Jefferson invention wouldn't hurt it. If it's merged, I think that the photo must also be moved. --Wtshymanski 16:46, 17 Apr 2005 (UTC)

DON'T MERGE THEM!! Jefferson needs to be honored with the invention of this particular machine, not Bazeries, and vice versa.

But aren't they the same machine? We don't keep seperate articles simply to "honour" people. — Matt Crypto 18:52, 8 February 2006 (UTC)

I have gone ahead and done the merge as I felt they definitely should be the same article - the duplication was confusing. I hope I have done it in a way that highlights both people's contributions. I placed the page here as Jefferson was the original inventor, but kept the use of the Bazeries name through the discussion as it is his version that is used. Kcordina 10:10, 20 March 2006 (UTC)

Alice and Bob are commonly used names to denote the message encoder and decoder. Unfortunately, their usage is both unclear and American biased (maybe British biased? but definitely not worldly). I suggest we remove all their references and replace with message encoder or something similar. —Preceding unsigned comment added by 71.232.103.200 (talk) 21:26, 23 April 2008 (UTC)

I also completely agree and I have changed it :) Logan5671111^Talk_{Logan5671111#0001} 17:30, 27 January 2022 (UTC)

Commons link

I've removed the Wikipedia Commons link as there wasn't a page there and the only search result for "Jefferson Disk" on Commons was the image that's already on the page. Jobarts^-Talk 23:27, 31 October 2006 (UTC)

Nazi Reference

Is there any need to have 'heilhitler' be the message to be decoded? I tracked down the history of it, and it seems to be a remnant from the original initial Bazeries cylinder page (now a redirect), merged into this title and apparently imported from some unnamed 'public domain source.' I don't feel there's any reason to keep the code there. TJEFFERSON would be perfectly fine I think. —Preceding unsigned comment added by JJansen (talk • contribs) 01:45, 3 March 2009 (UTC)

I totally agree. I'm a teacher at a german school and right now we're talking about different cypher methods. I couldn't explain that Hitler-reference to the student who found this. That's plain unnecessary since Jefferson has a complete different background and time of life. Mana (talk) 12:32, 29 June 2011 (UTC)

Poor writing

This article is riddled with an awkward present tense voice, and cast in terms of "and then Holmes does this, and then Holmes does that".

It's awkward and distracting. We can do better. This is largely a stylistic point and so need not be repaired instantly, but is less than good work. WP should be better served.

I will add it to my list, but doubt if I will be able to get to it. ww (talk) 20:08, 29 October 2009 (UTC)

There are 36 disks

The Jefferson disks had 36 disks, not 26 (there are 26 characters on the periphery of each of the 36 disks). See The Codebreakers by David Kahn (page 194 of the 1st edition).

You can also count 35 disks on the image attached to the article. That image is from the National Cryptologic Museum (the 36th was taken out to show a disk in its entirety, it is in placed in front of the set of disks on display at the Museum).

I will fix the error Econrad (talk) 19:21, 5 November 2009 (UTC)

The montecello.org page says Jefferson used 26 disks. The model shown uses 36. I clarified this in the article. I also did a major copy edit of the intro, which was repetitious and misleading regarding security.--agr (talk) 21:50, 5 November 2009 (UTC)

The Codebreakers is a far more reliable source than the Monticello.org web site. The Codebreakers cites Jefferson using 36 disks. Just because someone built a modern set of 26 disks does not mean Jefferson used 26 disks.

See page 194 of this edition of the Codebreakers on Google Books: http://books.google.com/books?id=SEH_rHkgaogC&lpg=PA194&ots=_0iql8t0v4&dq=jefferson%20cypher%20codebreakers%20kahn%2036&pg=PA194#v=onepage&q=&f=false

Econrad (talk) 21:56, 5 November 2009 (UTC)

Precedence of European designs.

A device similar to, but more advanced than, that of TJ was designed in Sweden by Fredrik Gripenstierna in 1786.

Since the Jefferson cylinder was never constructed precedence might also be given to Gottfried Leibniz who designed a machine c1670 that had some similar features. AnnaComnemna (talk) 19:07, 8 February 2015 (UTC)

Further information on the Jefferson disk

Even Jefferson failed to recognize the brilliance of his invention; he praised the Playfair cipher to its inventor, disparaging his own effort.

While his description involved a wooden version, the WWI version was made of metal, considerably smaller, and easier to carry and hide.

See "Codebreakers" by David Kahn. — Preceding unsigned comment added by Lucky Vassall (talk • contribs) 19:00, 16 June 2019 (UTC)

==Wiki Education assignment: The Age of Revolution and Historical Memory== This article was the subject of a Wiki Education Foundation–supported course assignment, between 20 January 2022 and 4 May 2022. Further details are available on the course page. Student editor(s): Lexi Armstrong (article contribs). Peer reviewers: Mallorybrennan.

Jefferson disk in the NSA museum (picture)

The cipher cylinder on display there (picture in this article!) originally had 40 discs; discs 35, 36 and 38 to 40 were lost over the years. Each of the discs also has 42 positions, with provision for all the letters of the French alphabet. After all, the device originates from the southern states of the USA! (In French, in addition to the 26 letters of the standard alphabet, there are 16 further letters: à, â, æ, ç, è, é, ê, ë, î, ï, ô, œ, ù, û, ü, ÿ). These 42 characters can be arranged on a disc in 41! = 33,452,526,613,163,807,108,170,062,053,440,751,665,152,000,000,000 different, fully cyclic permutations. The internal key space required to produce the 40 discs is therefore (41!)^40 = 9.5 × 10^1,980 (approx. 6,581 bits). By contrast, the actual external key space of an existing machine of this type was ‘only’ 40! = 8.2 × 10^47, or approx. 159 bits.

Contrary to what might be assumed, there is no evidence that this device ever had anything to do with Thomas Jefferson, although je worked as US Ambassador in Paris for four years (1785 - 1789). Compared with the cipher machine described in Jefferson’s own notes (he suggested 36 - 46 disks), which were presumably found amongst his papers, this device had 40 discs instead of 36, with 40! permutations instead of 36!, and an additional 16 letters per disc. Furthermore, for every 40 plaintext letters in a line, there were 41 possible ciphertext lines available, from which one was selected at random, with the process repeating for each block of 40 letters. It is completely unknown who used this device for encrypted communication and with whom. However, it seems rather unlikely that this device was built without first having seen Jefferson’s cipher cylinder and having its function explained. Perhaps it was an unauthorised copy at the time, a plagiarism? In any case, it is not Bazeries’s; he reinvented the principle of the cipher cylinder 96 years after Thomas Jefferson (in 1891 rather than 1795), without being aware of Jefferson’s invention. Permissiveactionlink (talk) 13:29, 13 October 2026 (UTC)

Diagnosis - Treatment !

"One major weakness of the Bazeries cylinder is that the offset from the plaintext letter to the ciphertext letter for the cipher alphabet on each disc will be exactly the same. In the example shown above, this offset is six letters." This serious security flaw can be remedied very easily (on slotted cipher devices; less so on cylinder devices): with an ADJUSTMENT KEY. For a Jefferson cylinder with 36 discs, this key consists of exactly 36 numbers (not digits!), e.g.: 13, 05, 24, 19, 02, 00, .........., 17. There are 26^36 = 869,041,521,632,724,682,610,610,260,053,736,824,323,722,262,347,776 different keys for this. (additional keyspace : 169,2 bit)

The Transmitter of the encrypted message proceeds as follows:

1. Prepare the device with the correct disc permutation.

2. Set the plaintext line on the cylinder to a single line.

3. Using the adjustment key, turn all the discs in accordance with the key by the appropriate number of positions.

4. Select the ciphertext at random from one of the 26 (not 25) rows.

The receiver follows the procedure in reverse:

1. As before/above

2. Instead of the plaintext line, set the ciphertext line on the cylinder in a single line.

3. Using the adjustment key, turn all the discs BACK (!) by the corresponding number of positions !

4. Scan the 26 lines of the cylinder visually: the plaintext will now be found in one of the lines.

A significant difference now: de Viaris’s method no longer works, because a plaintext letter can easily be encrypted with itself (possibility or "fixed points" !) Permissiveactionlink (talk) 14:17, 13 October 2026 (UTC)

Jefferson disk encryption method

The Jefferson disk works cryptographically as follows: Each plaintext character is replaced, in each of the 36 cipher alphabets, by other letters which always occur n = 1 to 25 characters further along in the cyclic sequence of characters on the discs. Although n can be chosen at random, it remains the same for all 36 characters. The cipher operations therefore consist of 36 substitution cipher operations, each shifting by n positions in each of the 36 multiplex-alphabets. If n is chosen at random, the mapping between plaintext characters and ciphertext characters is defined, mathematically spoken : bijective. Contrary to what is sometimes stated in cryptology textbooks, this cipher is neither homophonic nor polyphonic ! Since one is completely free to choose any one of 25 substitution ciphers for each plaintext position from a heavily scrambled alphabet, provided that n – that is, the number of substitution shift positions – remains constant for 36 characters, this method could instead be called "RANDOM HOMOALPHABETIC". Please do not confuse this with monoalphabetic! Permissiveactionlink (talk) 14:41, 13 October 2026 (UTC)