I have a theory about how the 340 was encoded. Without getting into to much detail, it is safe to say the 340 is not a single substitution cipher. There is software out there that can decode any single substitution cipher in a matter of seconds and the Z340 has been hammered by that software for years now.
The 340 must be some kind of double transposition cipher. And since it is highly unlikely Zodiac used any computer technology back in 1969 to encode it, he must have used a pen and paper type method.
Dan Olson of the FBI, has pointed out that:
the message may have been written, then split into two equal size parts and placed top over bottom.
Meaning Z may have encoded the first half (lines 1-10), then encoded the second half (lines 11-20) using new keywords.
Seems logical.
After one year of it never being cracked, Z sent a Halloween card to Paul Avery in 1970 that many consider to be a clue to the 340. (more on that here:)
http://www.kirps.com/web/main/resources … zodiac340/
In conclusion, I would guess two things about this cipher.
1) That Z used a double transposition method.
http://www.pbs.org/wgbh/nova/decoding/doubtrans.html
2) The hint Z sent in the Halloween card is the (2) keywords used for both the top and bottom section of the code.
These keyword could be: byfire & bygun or fire & gun for the top and byknife & byrope or knife & rope for the bottom section.
Hey Paul_Averly,
If you like you could encrypt a 340 text part of the 408 this way then we could see how much it would line up with the 340.
Here is the first 170 characters (same as half the 340) of the 408:
ilikekillingpeopl
ebecauseitissomuc
hfunitismorefunth
ankillingwildgame
intheforrestbecau
semanisthemoatdan
gertueanamalofall
tokillsomethinggi
vesmethemoatthril
lingexperenceitis
Here is that section encoded using Double Transposition with keywords: byfire, bygun.
iowiorgshnirchbae
kcdhlllfruviilsfo
pmegneskaleiultil
pertiesnaltlneame
eiuomkfamsnororgc
iimltnitaheettnms
bnsasisgnthliitex
aaemtehgelieofegn
iumoogeueeettleaa
itndtpuncahlmiski
A couple of interesting observations from this experiment.
Notice all the double letters. Could be that Z would use the double X in place of any random double letter.
Also, double trans is very difficult to decode because if only one letter is off it screws the entire decoding up.
If the Halloween card is in fact a clue, what was the clue? I would think it wouldn’t be to vague if he really wanted to assist in the decoding.
The idea of double transposition is interesting. Apply double transposition to the plaintext and then homophonic substitution.
I think that there is a tendency in the english language (or maybe languages in general) for there to be less double and triples, etc, than would occur in a random text with the same frequencies.
http://rumkin.com/tools/cipher/coltrans-double.php
Wow great call on the Halloween card hint…except the keywords where slave and paradice. Very intuitive though!
There was a further hint on how to read the 340 on the Halloween card envelope as well.
Does this mean the key to the 13 cipher is connected to writing "editor" upisde down on that envelope?
