<<

. 15
( 118 .)



>>

Maryland, via the moon."
A few weeks later, the Oxford left Norfolk on its first operational
cruise, an eavesdropping sweep off eastern South America. After a brief
visit to Colon, Panama, it crossed the equator and sailed to Recife, Brazil;
Montevideo, Uruguay; Buenos Aires, Argentina; and Rio de Janeiro,
Brazil. Along the way, the ship successfully used its moon-bounce
antenna to send information back to Washington”another first.
In addition to speed, the moon-bounce antenna also provided the ship
with stealth. Unlike the standard high-frequency communications, which
were vulnerable to foreign direction-finding antennas, the moon-bounce
signal was virtually undetectable because it used hard-to-intercept
directional microwave signals. The moon-bounce system was also
immune to jamming. Ground stations for the system were located at
Cheltenham, Maryland, near NSA; Wahiawa, Hawaii; Sobe, Okinawa; and
Oakhanger, in the United Kingdom.
On June 20, 1962, Commander Thomas Avery Cosgrove took over as
captain of the Oxford. Cosgrove was a "mustang," an officer who had
previously served as an enlisted man; he was "as rough as sandpaper,"
said Aubrey Brown, one of the intercept operators on the ship. "He had
tattoos all the way on his arms down to his wrists. He had a tattoo
around his neck. And he had the language of a boatswain's mate."
About a month later, on July 16, the ship set out for another four-
month surveillance mission down the South American coast. But three
days later it received an emergency message to set sail immediately for
Cuba "in response to highest priority intelligence requirements."


By the summer of 1962, the shipping lanes between Russia and Cuba
were beginning to resemble a freeway during rush hour. On July 24, NSA
reported "at least four, and possibly five . . . Soviet passenger ships en
route Cuba with a possible 3,335 passengers on board." The passengers
may well have been Soviet military personnel brought to operate Soviet
radar and weapons systems. Over little more than a month, fifty-seven
Soviet merchant ships visited Havana. "In addition to the shipping


81
increase," recalled Admiral Robert Lee Dennison, who was in charge of
the Atlantic fleet at the time, "there were large numbers of Soviet-bloc
military personnel prior to August and then there was a buildup during
August and September when nine passenger ships arrived in Cuba with
a total capacity of 20,000 passengers. But at the time we didn't have any
way of really confirming how many people were on board these ships
because they would disembark at night."
At the same time, NSA began noticing increased use of deception.
Ships leaving ports in Russia listed destinations in the Far East and in
Africa. But NSA, with its network of giant elephant cages intercepting the
vessels' daily broadcasts and triangulating their positions, was able to
track them as they crossed the Atlantic en route to Cuba. NSA was also
able to detect ships loading far less cargo than their manifests called for,
thus leaving a great deal of room for weapons and military supplies.
Thus when the new Soviet cargo ship Beloretsk arrived at Archangel in
late May it was supposed to load about 7,800 tons of lumber, but only
5,240 tons were actually put aboard. That cargo filled only a third of the
Norwegian-built ship's 14,150-ton capacity. "It is therefore believed,"
concluded an NSA report, "that the Beloretsk may be carrying a partial
load of military cargo."
As the summer wore on, the signals became more ominous. About
forty miles off the westernmost tip of Cuba, an antenna-packed ferret
plane picked up the first telltale sounds of Soviet airborne intercept
radar. This meant that Cuban air defense bases could now accurately
target and shoot down U.S. aircraft flying over or near their territory,
thus increasing exponentially the risks of the eavesdropping missions.
That same day, intercept operators began hearing Russian voices over
Cuban internal communications links. "Comint sources reveal Russian
and non-Cuban voice activity on Cuban Revolutionary Air Force tactical
frequencies," said one report. Another troubling sign.
In May 1962, as the Soviet buildup in Cuba continued to look more
menacing, Vice Admiral Frost began touring listening posts in the Far
East, including the large Navy monitoring station at Kamiseya, Japan.
The next month he was suddenly booted from the agency and transferred
to the Potomac River Naval Command, a halfway house for admirals on
the brink of retirement. Director for less than two years, Frost bore the
brunt of the various inquests into the double defection of Martin and
Mitchell. Because of that, and disputes with the Pentagon, his cryptologic
career was terminated prematurely. Many also felt Frost had a problem
dealing with NSA personnel. "I thought Frost was one of the least
effective [NSA heads]," said former NSA research chief Howard
Campaigne. "I think his problems were communication problems."
Another former NSA official said Frost had trouble controlling his anger.
"I saw him chew out Frank Raven, Bill Ray [senior NSA officials], and


82
some Air Force brigadier general in a briefing," said Robert D. Farley, a
former NSA historian. "Just the finger-on-the-chest bit."
Replacing Frost was fifty-one-year-old Gordon Aylesworth Blake, an
Air Force lieutenant general. Blake knew what he was getting into; he
had earlier run NSA's air arm, the Air Force Security Service. As a
sixteen-year-old, he slipped into West Point under the age limit. "I hadn't
been north of Minneapolis, east of Chicago, south of Des Moines, or west
of Sioux City," he recalled. "I was pretty green."
Eventually awarded his pilot's wings, Blake went on to
communications school and was assigned to Pearl Harbor, Hawaii, in
1939. On the morning of December 7, 1941, he was on duty as the
airfield operations officer, waiting to make sure a returning flight of B-17
bombers was properly parked. They were due to arrive at 8:00 A.M. from
California. "So all of a sudden we hear this big 'karroppp,''" said Blake. "I
raced outside and here was a dive-bomber that had bombed a big depot
hangar at the south end of the hangar line. It pulled up and we could see
this red circle under the wing. Well, no guessing as to what the hell had
happened." Blake ran up to the control tower to warn the B-17s that
were due in and he eventually managed to land them safely. For his
actions during the Japanese attack on Pearl Harbor he was awarded the
Silver Star for gallantry.
Blake knew Frost was in trouble and was somewhat uneasy at moving
in as his predecessor was moving out. "Jack Frost was under some
nebulous status because of the Martin-Mitchell case," he said. "I very
much felt badly about coming in over his prostrate form, and he
understood that."
Blake kept Dr. Louis Tordella as deputy director and largely left to
him the agency's most secret operations. "It would be better for NSA and
for those activities if I left that to Tordella," Blake said. "And that was our
working relationship. So while I usually had a general knowledge of this
compartment and that compartment, I made no attempt to be really
knowledgeable about it and, therefore, just less involved security-wise.
Maybe that's an odd view but directors come and go and for them to
become a repository of every last little secret never struck me as being
really very useful." Tordella was on his way to an extraordinary reign as
NSA's chief keeper of the secrets.
Sensing the tremors of approaching war in the summer of 1941,
Tordella, then a thirty-year-old assistant professor of mathematics at
Loyola University in Chicago, one day walked into the nearby U.S. Fifth
Army Headquarters and volunteered his services. But after the professor
explained that he held a doctorate in mathematics, practiced crypt-
analysis as a hobby, held an amateur radio license, and wanted to serve,
the Army major in charge could not be bothered. Possibly thinking that



83
the new recruit would be far more comfortable with a box of chalk
instead of bullets, he brushed him off with a sneer: "When we want you,
we'll draft you."
Cold-shouldered by the Army, Tordella would soon be embraced by
the Navy. Spotting his background on a questionnaire Tordella had filled
out for the American Academy of Science, Laurance Safford, a naval
officer and father of the Navy's codebreaking effort, rolled out the red
carpet. By April 1942, Tordella was a lieutenant (junior grade) assigned
to OP-20-G, the Navy's cryptologic organization, in Washington. Working
out of a temporary building on Constitution Avenue, the lanky Hoosier
stood his first watch”supervising direction-finding operations”after one
eight-hour indoctrination.
But before long, Tordella was using mathematics like a burglar using
lock picks, looking for the array of numbers that would pry open the
hellish German cipher machine known as Enigma. In July 1942,
Lieutenant Tordella was transferred to Bainbridge Island in the state of
Washington, a key intercept station for eavesdropping on Japanese
communications. But after several years at the remote listening post in
the Northwest, and with the war beginning to wind down, Tordella ached
to test his skills closer to the front. The opportunity came in 1944, when
he received orders to China.
During a stopover in Washington, D.C., for meetings, however, he
learned that he had been bumped from the assignment. Instead of
heading to the war, he boarded a train to New York City and a special
twelve-week course at Bell Laboratories on new equipment that was
designed to decipher Japanese voice codes. Initially, Tordella was to
travel to the South Pacific to test various equipment and techniques.
However, before the system could be deployed the military situation in
the Pacific changed. Tordella was again reassigned, this time as the
officer-in-charge of a newly established Navy experimental intercept site
at Skaggs Island, an isolated, mosquito-infested wetland near San
Francisco. Here, amid the frogs and snakes and antennas, Tordella sat
out the remainder of the war.
Mustered out in October 1946, Tordella had not lost his taste for
codebreaking. Rather than return to the classroom in Chicago, he signed
on as a civilian mathematician with the Navy's cryptologic organization,
then called the Communications Supplementary Activity and later the
Naval Security Group. With the formation of NSA in 1952, Tordella
transferred over and became chief of NSA-70, which was responsible for
high-level cryptanalysis. A rising star, he was named deputy director in
August 1958.
Because Tordella had developed a close working relationship with the
CIA's chief of operations, Richard Helms, who would later go on to



84
become director, Blake also let the mathematician handle the problems
that occasionally developed with that agency. One difficult situation
came up when the CIA tried to muscle in on the NSA's territory by
putting out its own signals intelligence reports. "I left that one to Lou for
some reason or another to sort it out," said Blake. "He and Dick Helms
were thick as thieves."


With the enormous focus on Cuba, Blake barely had enough time to
find his office before the alarm bells began to sound. On July 19, Robert
McNamara pushed NSA into high gear. "NSA has been directed by Sec
Def [Secretary of Defense] to establish a Sigint collection capability in the
vicinity of Havana, Cuba," Blake immediately notified the Chief of Naval
Operations, "as a matter of the highest intelligence priority." He then
pulled the ferret ship USS Oxford from its South American patrol and
sent it steaming toward Havana.
The Oxford was ideally suited for the mission. Where once cases of
lima beans, truck axles, plumbing pipes, and other cargo had been
stored, the earphone-clad intercept operators now sat in front of racks of
receivers and reel-to-reel tape recorders. Up forward, near the bow, voice
and Morse collection specialists twisted dials and searched for signals.
Fortunately for NSA, the Cubans never tried to scramble voice
communications. In the background was the constant rapping of
teleprinters printing out intercepted Soviet and Cuban telexes and other
communications. One deck above was a steel forest of antennas. In the
stern, below another forest of spindly metal tree trunks and stiff wire
branches, the Elint specialists listened for the twitters and warbles of
Russian radar on Cuban air bases.
"From the ship we could look up and down the length of the island,"
said Harold L. Parish, a Soviet analyst. As if it were on a cruise to
nowhere, the Oxford would sail in circles and figure eights for weeks at a
time six miles off Havana's Morro Castle. The ship's slow and lazy pace
was especially good for loitering near key microwave beams, narrow
signals that were difficult for airborne ferrets to pick up. "The quality of
the intercept was good," Parish said. "Even with the C-130 you were
flying kind of fast and you flew through the [microwave] beams" so that
not enough signal was captured to decipher.
As the weeks went by, the intercepts became increasingly ominous.
On August 17, an Elint operator on the Oxford heard an unusual sound,
like the song of a rare bird out of its normal habitat. It was the electronic
call of a Soviet radar codenamed Whiff, a troubling sound that meant
Russian anti-aircraft weapons had now been set up.
At NSA, a number of Soviet Sigint experts in A Group were suddenly
told to report to the office of Major General John Davis, the operations


85
chief. "We were called down and told there was evidence of offensive
missiles," said Hal Parish. They were then sent to help out the Spanish
Sigint experts on the Cuba desk in B Group. "We all descended down
there and we formed what was the watch for the Cuban missile crisis. . .
. All the people who were previously associated with the Cuban target”
the management and so forth”kind of disappeared and went off to the
side. We came down and set up the round-the-clock activities and sort of
went from there." Parish said some friction developed between NSA's
civilian and military staffs. "There was some," he said, "there always is."
In Washington, within hours of receiving the CRITIC message
containing the intelligence, senior officials began scurrying to meetings.
The CIA director, John McCone, told one high-level group that he
believed that the evidence pointed to the construction of offensive
ballistic missiles in Cuba”missiles that could hit as far north as the
southern part of the United States. What else could the anti-aircraft
weapons be protecting? he asked. But both Secretary of State Dean Rusk
and McNamara disagreed, maintaining that the buildup was purely
defensive.
To try to coordinate much of the data collection, Blake set up NSA's
first around-the-clock Sigint Command Center, which later became the
present-day National Security Operations Center (NSOC). "It was for
most of us our initial contact by telephone with a customer on the other
end," said one of those assigned to the command center. "It was the first
time I had ever talked to colonels from DIA [the Defense Intelligence
Agency]. Same with CIA. . . . We turned on the heavy reporting, both spot
and daily report summaries at that time and twice-daily
summaries.....We worked between eight and twenty hours a day."
Blake spent much of his time in meetings with the U.S. Intelligence
Board. "We would recess for a few hours so the staff could type
something," he said, "and then we would come back again, and the basic
question we were addressing [was], If we belly up to the Russians, what
will they do? Well, I am sure you realize how hard that question is
because you talk about intent, you see, and you don't read any messages
that give you intent. And I recall our final paper on the subject to the
president, pretty much bottom line was 'We think the Russians will
blink.' "
Among the key problems NSA faced was a shortage of Spanish
linguists and, at least in the early stages of the crisis, a lack of intercept
coverage. "One collection facility . . . against x-hundred emitters that
were on the air at the time from the Cuban area," said NSA cryptologist
Hal Parish, "we were just a little short. So that was a problem." Still
another concern was the lack of secure communications between NSA
and the listening posts. "Communications were definitely a problem,"
Parish said. "Secure communications. I'd say we were doing advisory

<<

. 15
( 118 .)



>>