TOP SECRET CX/MSS
SUIWECT: Fish Notes Report #F 46
TO : CO, SSA, War Dept. 12 June, 1944
1. SSA 161 left me in some doubt as to Just what was wanted and also
as to the urgency of the request. I assumed that a cabled reply was not
expected because of the quantity or data requested but suggest that in
the future requests of this nature indicate whether a reply is wanted by
cable or through a report sent via regular channels.
2. I also assume that the inquiry was directed only at patterns used
on links employing the P5 limitation and that you are not interested, for
example, in Stickleback patterns. The scope of your questions was puzzling
not only to me but also to Maj. Tester and Mr. Newman because the number
of crosses in X, Psi and DeltaX patterns is always the nearest integer (one
side or the other) to half the period of the wheel (necessarily an even integer
in the case of DeltaX). It is conceivable that there may have been an
isolated instance or two in which this rule has been violated but no-one
here remembers such a case and the patterns have not been checked from
this viewpoint. It is difficult to see how an occasional variation in
this respect could be of cryptanalytic significance.
3. Even as to the DeltaPsi patterns. the reason for the inquiry~is not
clear because the average answer is dependent on the number of dots in
Mu37 and I have been sending data on This regularly in my reports.
Presumably, however, you are interested in deviations from the average so
this information is set forth below. The people here are curious as to
whether you have in mind any techniques based on these deviations.
B R E A M . JELLYFISH
Jan. Feb. Mar. Apr. May June Mar. Apr. May 2nd May June
4. Obtaining the correct answers to the "Questions on Tunny Conven-
tions which Capt. Seaman sent me as Annex G was rather difficult. There
appears to be considerable misapprehension on the subject in the sections
themselves and this accounts for the fact that the illustration given in
Report #F 5 is not wholly correct. I believe, however, that the following
description is completely accutate. The Xs; Psis and motor all start at
the first position. The motor is not written above the column it is
about to affect, as suggested in Annex G, but rather, as stated in Report
#F 5, the motor elements determine whether the Psis in the column
immediately