Enigma Settings Sheets

Here are some answers to Craig Heath's Enigma questions posted last week.

Phil Marks


-----Original Message-----
From: Bletchley Park discussion [mailto:BLETCHLEYPARK-0lvw86wZMd/[email protected]] On
Behalf Of Craig Heath
Sent: Friday, February 10, 2012 1:27 PM
To: BLETCHLEYPARK-0lvw86wZMd/[email protected]
Subject: [BPARK] Enigma Settings Sheets

<snip>

- Was there any particular logic governing when UKWD settings changed?  The
authentic settings sheets seem to have used variously 7, 8, 9 or 10 day
periods.


Not that I am aware of, other than the general observation I found in some
of the source documents that the Luftwaffe (GAF = German Air Force in GCCS
terms of the time) changed "about every 10 days".  The sample key sheet
shown in my Cryptologia article (originally provided to me by the late Tony
Sale) divides a 31-day month into UKWD periods of 9, 12 and 10 days.  Some
other sheets posted by Frode Weierud at:

http://cryptocellar.web.cern.ch/cryptocellar/Enigma/index.html

show shorter periods as Craig has noted.

At the time I wrote the article, I did not know much about the KD machine.
This was introduced by the Abwehr in late 1944 as a stopgap replacement for
the G-series Enigma (the "lobster machine" broken by Dilly Knox and team in
1941) until the much tougher SG41 machine could be brought into widespread
service.  Some details of the KD machine can be seen at:

http://www.cryptomuseum.com/crypto/enigma/k/index.htm

It also used UKWD, in conjunction with three multi-notch wheels selected
from a choice of 6.  It appears that with this machine, the UKWD wirings
remained in force for about 15 days each.



- Was there any physical limitation on which pairs of letters could be
plugged together?  The authentic settings sheets avoid alphabetically
adjacent pairs (like 'AB') but not all physically adjacent pairs ('IK' is
actually physically adjacent, because there is no 'J', and that's used on
the Luftwaffe settings sheet No. 651).


I do not believe so, other than the fixed B/O pairing (B/O in GCCS notation
- the relationship between this notation and the German markings is
explained in my article).  Replugging UKWD is a fiddly job, and there isn't
much room under the cover for the wires - but so far as I can see any other
pair of letters could be plugged together.

The Luftwaffe cipher office appears to have had some odd ideas about the
nature of randomness, and in fact managed to create patterns in their keys
(identifiable and actually used by GCCS) in what appears to have been some
sort of misguided attempt to improve security.  For example, not plugging
together consecutive letters on the Steckerboard.  This led to the addition
to the BTM bombe of the "CSKO" (consecutive stecker knock-out) feature,
which took advantage of the limitation by increasing the effective power of
the diagonal board.

In my article I noted that late in the war someone in the Luftwaffe cipher
became lazy and started constructing UKWD pairings by reusing stecker
pairings from other keys.  I will speculate that the phenomenon you are
seeing in the UKWD pairings in the sheets you reference (no consecutive
letters plugged together) is a by-product of reusing stecker pairings that
also contain no consecutive letters.



- Why did they avoid alphabetically adjacent pairs on the plug board in the
first place? (I assume this was a key generation practice that was simply
carried on when the UKWD settings were added.)

See previous answer.



- Were there any rules about which ring settings were permissible (such as
not using the same one two days running)?


I don't recall seeing anything of this nature.  A man called Reg Parker made
a minute study of German keys at GCCS (see Welchman, The Hut 6 Story); he
was the person who spotted the key repeats mentioned above - so if you have
access to any archival material written by him, that would be the place to
look.



- Are there any authentic settings sheets available for earlier indicator
procedures (with specified Grundstellung and/or fewer than 10 plug board
pairs)?


I don't recall seeing any, but I'm not the best person to answer this
question.




I have a couple of other random Enigma questions as well, so I may as well
ask while I'm here :-)
- What's the difference between an Enigma M2 and M3?


Sorry, can't help here.



- Deavours and Kruh 1985 (and hence Wikipedia) states that Enigma models
prior to I used a different rotor construction, where the turn-over notches
are fixed to the rotor core rather than to the settable ring.  Is this true?
(I haven't seen any other reference) and if so, which models?


D&K repeat this information in various articles.  I believe it to be true -
looking at Cryptologia 26:1, they seem to be saying it applied to the early
commercial models B, C, and D.  Can someone else on the list with more
knowledge than me pitch in here?

<ends>