Some info regarding letter and word frequencies:

http://norvig.com/mayzner.html

According to this tableau, the most common 5-gram that consists out of actually two 3-grams (combined to each other, comparable to the ‘IOFBc’ in line 13 of the 340) would be:

CTI + ION = CTION
ECT + TIO = ECTIO
INT + TER = INTER
ITI + ION = ITION
RAT + TIO = RATIO
STA + ATE = STATE
THE + ERE = THERE
THI + ING = THING
TIO + ONS = TIONS

while all other would not be amongst the 50 most frequent ones (all, the two 3-grams and the 5-gram combined of those two 3-grams).

IMO this one could be a great approach and I’d like to thank ‘Zambac’ for this valuable hint.

And we got the 3rd letter of such a 9-letter-structure as the last but one letter before the IOFBc structure, which again has the ‘+’ as the second but last symbol before it. So we got ‘+NIOFBc’ with ‘+’ being our frequent double letter and ‘N’ being the 3rd letter in the repeating D-N—-o structure. Also the fourth letter is a ‘p’, which is the second double letter in the cipher, appearing in line 4:

QT

Hello…

Zambac did support us with another hint – and I got to say that I am completely baffled..

So far I had thought that the 340 would be similar to the 408. However, when compared to the 408, massive similarities in the cipher structure can be found. With ‘massive’ I consider structures that show up with an e.g. 63^5 + 63^5 + 63^4 + 63^4 + 63^3 chance, therefore 1:2,016,629,055 or 1:2 billion.

And not only that, there are even more structures repeating inside the 340 itself. Those are numerous, too, so that we have to rethink completely about the 340 structure itself.

Thanks to Zambac for this excellent work, imo its a brilliant breakthrough in cracking the 340, if possible at all (Zambac, I modified your work slightly as two symbols or so differed and I could figure out another ‘B’ in one of the last rows..):

Identical structures can be found in both ciphers

Additionally, some structures are repeating in the 340

And this was not the only one…

QT

…instead we find even more repeating structures.

At this point it should be clear that those structures do not repeat accidentially. One or two, ok..but all those structures of how the identical symbols are placed, with identical distances to each other, should convince that we deal with a certain method of how those symbols had been placed in the 340.

Still unclear, however, how to derive a possible cleartext from those structures. I wonder if there is a possible punchcard cipher behind those structures? IBM had started with punchcards in 1928, so it might have been a time when punchcards had their early development. Such systems might have had a military use, e.g. for observatories, radio technology or military.

Please feel free to share your opinion on that, in any case I am convinced that Zambac has found a major clue here.

Btw, transforming the position of the relevant symbols over the cleartext of the 408 leads to the letters:

EPIETSUOOSMEHSNNGPLAY

This might be nonsense, however ‘play’ does appear at the end of this letter row. Anagramming could lead to SOMETHING EPESSOUN or SEEN SOME SOUP THING or something like that..not sure if this way is a good one, but the structures itself are really striking. Baffled.

QT

Hi Quicktrader, You’re welcome. I do as much as I can. Thank you.

Excuse me. Because my Einglich is not good, so I feed User Quicktrader with my approach, because he speaks the German language as I do.

It makes me happy when I find something that solution to the Z is 340.

Zambac

Incredible work !!!

I found now another Structure +

absolutely great…now I am really curious about the meaning behind those weird symbol structures..

QT

absolutely great…now I am really curious about the meaning behind those weird symbol structures..

QT

Have you tried to plot in the 408 symbol=letters solution from the 408 symbol structures, into the 340 cipher symbol structures ?

Is it your theory that Zodiac is using the same words or phrases repeatedly?

absolutely great…now I am really curious about the meaning behind those weird symbol structures..

QT

Have you tried to plot in the 408 symbol=letters solution from the 408 symbol structures, into the 340 cipher symbol structures ?

I have tried..it’s mentioned above – but not with the new structures yet – will do so immediately..

Update: EIPIETSSISNUOOSMEHSMNENRPOEILAY

Only nonsense coming out of it (transferring the structures to the 408)…also transferring the cleartext letters vice versa from the 408 to the 340 doesn’t work out as e.g. for the symbol ‘p’ there are at least three different alphabetical letters present in the 408.

No idea how to continue with that one..

Be aware that there is two different structure types found:

1.) Structures that can be found in both ciphers (8)
2.) Structures that repeat itself in the 340 only (8)

Latter one can be found above (structure2 & structure3), the first one (updated, structure6) can now be seen here:

So far there are 8 structures identified for each, therefore a total of 16 structures found. One or two being accidential, ok, would be a chance of 1:20,000 or so. But 16?

This is so weird..all this looks like a fingerprint, imo someone could secure nuclear weapons using such ‘sheets’, too.

QT

Ps.: U may want to use ‘Imagus’ as a Chrome browser extension to see pictures in full extent when moving the mouse over it, so you don’t need to download the pictures.

Ps.: U may want to use ‘Imagus’ as a Chrome browser extension to see pictures in full extent when moving the mouse over it, so you don’t need to download the pictures.

Cool.

Sorry Quicktrader, it is back to work for you.

I’m going crazy, it is I think even more. ı get headaches in the search.

PS : It might not be important, however, there are more connections than just three letters, but these are shifted only by a letter, either to the left or right and up or down.

So that’s the last thing I could find – enough for today.

Interesting findings. Have you tried doing the same exercice of finding repeating patterns inside the z408? I had a quick look and I could find this 9-symbol one:

For such intra-cipher repeating 2D patterns, I imagine it is similar to finding repeating n-grams that contain many many wildcards. If you approach this programmatically, you can probably find many of those. Using doranchak’s Cryptoscope gives an idea of how many you could find by looking at the "Find repeating ngrams with gaps" section; I believe it only displays patterns with a maximum of 3 consecutive wildcards so you could find many more than this.

Now, trying to find repeating 2D patterns between the z340 and the z408 is probably like finding repeating n-grams in the concatenation of both ciphers. As a very basic exercice, I used ZKD to find the repeating n-grams of such a z748(!).

z340: 9 repeating n-grams
z408: 42 repeating n-grams
z748: 57 repeating n-grams (6 more than the sum of the individual ciphers’ totals)

So, concatenating both ciphers augments the total amount of repeating n-grams by 12%. And these are n-grams with a maximum of 1 consecutive wildcard. Allowing more wildcards would augment the number of newly created repeating n-grams by combining both ciphers.

Based on all of the above, my opinion is that these intra and inter repeating structures look cool but are probably coincidental.

Interesting findings. Have you tried doing the same exercice of finding repeating patterns inside the z408? I had a quick look and I could find this 9-symbol one:

For such intra-cipher repeating 2D patterns, I imagine it is similar to finding repeating n-grams that contain many many wildcards. If you approach this programmatically, you can probably find many of those. Using doranchak’s Cryptoscope gives an idea of how many you could find by looking at the "Find repeating ngrams with gaps" section; I believe it only displays patterns with a maximum of 3 consecutive wildcards so you could find many more than this.

Now, trying to find repeating 2D patterns between the z340 and the z408 is probably like finding repeating n-grams in the concatenation of both ciphers. As a very basic exercice, I used ZKD to find the repeating n-grams of such a z748(!).

z340: 9 repeating n-grams
z408: 42 repeating n-grams
z748: 57 repeating n-grams (6 more than the sum of the individual ciphers’ totals)

So, concatenating both ciphers augments the total amount of repeating n-grams by 12%. And these are n-grams with a maximum of 1 consecutive wildcard. Allowing more wildcards would augment the number of newly created repeating n-grams by combining both ciphers.

Based on all of the above, my opinion is that these intra and inter repeating structures look cool but are probably coincidental.

Good post and good find…but the last sentence I’d like to discuss: In the repeating 9-symbol structure (biggest one so far!) we can see that some symbols MUST repeat in such a structure as various letters simply are represented by one homophone only. According to the frequency table Z had used for the 408, the letters B, C and U would (usually) be represented only by one homophone. The word "BECAUSE" would therefore indeed show up with such a structure, and so it does in the 408 with the ‘V+eGY’ coming up twice with ‘VZeGY’, both representing the word ‘BECAUSE’.

This, btw, is a typical approach to crack a homophone.

QT