What MAGIC and ULTRA information went to President Franklin
D. Roosevelt during World War II, and how did he use it?
It may as well be said at once that exhaustive research provides only
a sketchy answer to the ﬁrst question and merely unproved hypotheses
for the second.
Extant records specify only a few of the solved German and Japanese
messages that went to the president. Moreover, no documents or rec-
ollections have come to light suggesting any action that he took based
on, or even inﬂuenced by, these intercepts.
Despite these disappointing results, it may be worthwhile to set out
what has been learned about the production of MAGIC (Japanese
solutions) and ULTRA (German) and their selection for and presen-
tation to Roosevelt. Doing so will at least outline how one intelli-
gence organization operated in preparing information for the highest
authority in the land. And it will provide a framework in case any
papers are discovered that answer, or at least shed more light on, the
fundamental questions about Roosevelt, MAGIC, and ULTRA.
Foreign intercepts began going to the White House for the ﬁrst
time in 1937.
ey came from the Army’s Signal Intelligence Service,
the Signal Corps’ codemaking and codebreaking agency. Founded
under the great cryptanalyst William F. Friedman, SIS
directed its ﬁrst nontraining codebreaking at Japan, starting in 1932.
At ﬁrst it solved Japanese diplomatic codes.
But these turned out to
encrypt only insigniﬁcant messages. High-level, important traﬃc was
* From Cryptologia, XVI(4), October 1992, 289–313. Used by permission.
WO RLD WA R II' S G RE ATES T SP Y
apparently protected by a machine cipher.
In 1935, when SIS real-
ized this and had obtained more experience and greater knowledge
of Japanese communication habits from its code solutions, it attacked
By 1936, cryptanalysts Frank B. Rowlett and Solomon
perhaps helped by information about a simpler Japanese
Navy machine cipher that the U. S. Navy had solved,
the Japanese Foreign Oﬃce called its “A” machine and what the
Americans called RED. With this, the United States gained access to
more and better intelligence than ever before.
is included advance information starting in March 1937 about
Italy’s possible adherence to the German-Japanese anti-Comintern
pact of the previous November. e intercepts revealed the early indi-
cations that Italy might join, the start of negotiations, the discussion
over whether Italy should accede to the existing treaty or sign a sepa-
rate agreement with Japan, a statement that Hitler wanted Italy to
participate in the existing agreement, Italy’s apparent acceptance, the
emperor’s approval, and part of the text of the treaty, which was signed
6 November 1937.
U. S. diplomats did not start reporting until October on this rap-
prochement between three aggressive nations, to the east and to the
west of the United States—and even then their information was far less
speciﬁc than the intercepts.
So these seized the attention of oﬃcials in
Army and Navy intelligence, in the State Department, and in the White
House. Information from solved foreign cryptosystems, which had never
before been of suﬃcient value to reach the topmost level of government,
began going to the White House,
probably to the president.
On 20 February 1939, the Japanese introduced a new and more com-
plicated cipher machine to supersede RED, whose mechanisms were
wearing out. e Japanese called it their “B” machine; the Americans,
ere followed nineteen months of concentrated cryptana-
lytic attack by a team under the general supervision of Friedman and the
immediate direction of Rowlett, with some help from the Navy (which
resumed diplomatic cryptanalysis to assist with PURPLE after having
abandoned that ﬁeld around 1938 or 1939 to concentrate on Japanese
naval solutions). eir eﬀorts were crowned with success in September
us, even after Japan discontinued use of the RED machine on
21 August 1941,
the PURPLE solution assured American oﬃcials of
being able to read Japan’s most secret diplomatic messages.