This paper adds nothing to the technical and organisational his-
tory of the Fish problem but is concerned with the experience and opinions
of those Americans who worked on that problem. Although the American Con-
tingent was very few in number, the training and the efforts were of great
value in what may well be termed the future trend of cryptanalysis. The
use of rapid electronic counting, the development of new statistical ap-
plications to the solution of cipher traffic and the research in abstract
mathematics itself all reached a peak in the operational work on the Fish
problem. The records speak for themselves in their contributions to the
science of cryptanalysis.
Organisation and Task of Block F. To give any description of
Block F, a brief picture of the organisation and its problem is necessary.
The scope of Fish was greatly enlarged and intensified because of its high
value in intelligence; and the diversified nature of the various approaches
to solution have already filled the pages of books without any general
concept of how the section actually worked to obtain results.
Essentially the Fish people were divided into two camps, the
"Newmanry" and the "Testery", between whom rivalry reigned. The first
handled the statistical work, usually consisting of the solution of the
patterns of five wheels and also setting those wheels on each message.
The Testery was concerned with the final stage of solution by linguistic
efforts, that is, the solution and setting of the seven other wheels used
by the cipher machine, and also the deciphering of the plaintext which was
sent to Hut Three. The division of the work was never as clearly detailed
as above stated because in many cases there were several means for arriving
at a solution. All twelve wheels of the cipher machine could be set
statistically in the Newmanry, the keys obtained from cribs could be
solved by either group, and the solution of new daily patterns could be
obtained independently by either group from depths or entirely from statis-