4. The foregoing represents a means of comparing the correct
setting with random, settings. However, many of the wrong settings
are not truly random because of the self match of the patterns which
is inevitable owing to the limtations on their composition. The
British call this self match "slide", A "slide" of 6 means that there
is a strong self match at a dispIacement of 6. This is a factor
that cannot be lost sight of in appraising the reliability of results
and the true significance of scores. Some theory has been developed
on this phase of the problem which I will try to go into at a subsequent
date.
5. Bream was the only circuit solved in March. Strenuous efforts
have been and are made to determine the Jellyfish X patterns
statistically. The circuit is Paris - Berlin and the traftic is be-
lieved to be very important. Messages with the same QEP numbers will
not read in depth so presumably Jellyfish has some auto-key element.
Very likely it is the P5 limitation but this is not certain. The circuit
started early in 1944 and the auto-key feature has always existed.
Since the traffic has never been read nothing is known about its plain
text characteristics. Virtually the entire attertion of Mr Newman' s
section has been devoted to operational work on Bream and research on
Jellyflsh. A slight amount or work has been done by Maj. Tester's
section on Stickleback (Koenigsberg - Poland) which uses the X2 limit-
ation but not the P5. No other circuit is being worked on at all.
6. I have asked Mr. Newman to collect some suitable messages
which his group is unable to work on, to be sent to A. H. with the
hope that our people can find the X wheels statstically. He Is very
anxious to have us try. I don't kriow whether the messages will be old
Jellyfish or some other circuit but will send them on as quickly as
possible.