Blind Man's Bluff: The Untold Story Of American Submarine Espionage

Three — Turn To The Deep

Flying on the wild success of his Polaris program, Admiral Red Raborn began looking ahead, thinking about new, imaginative ways of furthering nuclear deterrence. He quickly turned to the dreamer within his ranks, a young civilian whom the admiral had plucked from obscurity a few years earlier and anointed the chief scientist for Polaris.

John P. Craven was only in his midthirties when Raborn found him, but it was his job to look over the shoulder of everyone involved in the development of the missile subs, to find the problems, to come up with the answers. He was, as he put it, "chief kibitzer."

The moniker fit. Talking a torrent, his ideas usually overflowing, Craven was the kind of man who could dissect a blueprint and still have time to spout a few lines of poetry, biblical verse, or one of his endless series of self-scripted maxims of the sea. Sometimes he'd mix verse with maxim and sing the result aloud. He preached fantasy amid military discipline; he carried romance to the mechanics of nuclear war.

It was a role Craven had been bred to. He was the product of a family that reached back to Moorish pirates on his mother's side and was divided on his father's between Presbyterian ministers and Navy officers yawning in the family pew.

The Navy brass was the part of the lineage that most of the Cravens liked to boast about, the part that went back to Tunis Agustas MacDonough Craven, who skippered the Civil War Union ship Tecumseh when it was rammed by a Confederate mine during the Battle of Mobile Bay and inspired Admiral David Farragut's memorable cry to the remaining fleet: "Damn the torpedoes. Full speed ahead."

In noble tradition, Tunis drowned at the helm. Most of the Craven clan would stop the story there. John Craven, however, delighted in presenting a footnote: that Tunis died while fighting to get off the sinking ship ahead of the harbor pilot. And only John Craven boasted of what the rest of his family dared not even whisper: the pirate blood he inherited from his mother's side.

That John Craven was going to he different was evident from the moment he made his first appearance on the planet, landing in the Williamsburg section of Brooklyn. It was a Halloween night, a fact that his paternal relatives chose to ignore as they instantly christened him Navy, fully intending that he would live a life of rigid military discipline. That their plan was doomed to fail became clear some fifteen years later when Craven was rejected by the Naval Academy. It wasn't for lack of intelligence. He'd skipped through to high school by the time he was eleven years old. But once there, he took the rogue's route to popularity. He convinced his much older classmates that he was merely small for his age and then proceeded to win their respect by becoming the class wise guy, the kid who was too tough to do homework.

Ultimately, he fulfilled at least part of his family's expectations. He never earned a Naval Academy degree, but he did get his commission in the reserves and he became an ocean engineer. From then on, he took to sermonizing about the deep, about underwater maneuvers that most of the Navy passed off as impossible, or at least hugely improbable. He expected no easy converts. But like any minister preaching the coming of a miracle, Craven was drenched in the faith that he would ultimately be proven right.

Now Rahorn was handing Craven a nearly blank check to do what he did best-come up with ideas, as many as he could. By 1963 Craven was working hard on Rahorn's vision of an Advanced Seabased Deterrent Program. As his first step, he set aside $1 million a year, thinking that would he just enough to create a small political science program to dissect the strategy of deterrence. In the process, he discovered he had hired just about every political scientist specializing in strategic defense.

With the rest of his budget and his new platform, he began to peer into an untouched realm of the deep, working with his group to scribble out ideas: missiles that could he placed miles below the surface on the ocean floor; submarines that could reach down and see through the murky depths, carry cameras into untraveled and alien waters.

Most of the Navy greeted Craven's visions with hardly a yawn. What little study of the deep there had been before had long ago been shoved into a corner, the purview of a small group of oceanographers. Admirals saw operating in deep water as more difficult than the manned outer-space launches that, at that moment, held the nation's attention hostage. The Navy's best submarines could reach down just 1,000 to 1,500 feet or so. Go deeper, and there was certain death by implosion from punishing sea pressures great enough to quickly crush even the mighty Polaris subs.

The miles below the Navy's operational slice garnered about as much respect as the average landfill. The Navy's main design branch, the Bureau of Ships, listed deep submergence as tenth on its list of top ten priorities-giving the deep number ten only because the list wasn't any longer. Even Admiral Rickover, wrapped as he was in the public mantle of Navy innovator, was uninterested in plumbing the depths.

Craven's deep-submergence group was on the fringe, but eager to work. A team of his scientists was asked to help test the USS Thresher (SSN-593), the first of a powerful new class of nuclear attack submarines designed to go somewhat deeper than the other subs of the day. On April 10, 1963, Thresher failed during a test dive to 1,300 feet. As best as anyone could tell, a piping failure and a subsequent loss of propulsion set off a series of events that caused the submarine to sink, killing all 129 men aboard, including four men from Craven's team. Craven got the news as he was sitting with Harry Jackson, an engineering officer who had helped test the sub shortly before her last dive, and who had been present for every other deep dive.

Jackson sat, repeating over and over, "I should have been there." But Craven was relieved that Jackson had missed this, the nation's first loss of a nuclear submarine, along with three of Craven's own men who had been scratched from the test for lack of space.

It was only later that Craven realized that the disaster was about to mark him among the most important players in a new and dramatic chapter in this saga of undersea spying. Craven's opportunity would spring from the almost impossible promises the Navy made in its efforts at damage control.

In the wake of Thresher, the Navy promised a massive effort to learn about the unforgiving ocean depths. There would be a "Sub Safe" program. There would be "Deep Submergence Rescue Vehicles."

This was the Navy's chance to calm the public, a chance to erase tragedy with visions of ocean wonder, a chance to obscure submarine dangers beneath visions of safety innovations. Almost everyone involved recognized that some of the proposals were more science fiction than science, especially the prospect of deep-submergence rescue vehicles (DSRVs) for sunken subs. Anyone who was to be rescued would have to have the good fortune to go down over a continental shelf or atop an undersea mountain, in waters far more shallow than the two, three, or four miles of depth that made up much of the world's oceans. Most submariners knew that a severe casualty at sea almost always meant that they would disappear-no survivors, no rescue, nothing more to say.

Still, Congress okayed these popular proposals and offered up funding that caught the attention of the Office of Naval Intelligence. The Navy might have been promising an era that mirrored Jules Verne, but a few submarine espionage specialists now saw the means to launch a new age of spying that would he much closer to James Bond.

These intelligence officers were already crafting their plans when Craven began directing a massive post-Thresher study. He had also taken charge of the Deep Submergence Systems Project, a program created to design the Navy's promised deep-submergence rescue vehicles and to build an underwater laboratory, a habitat known as "SeaLab," where the Navy could study the physiological effects of deep-sea pressures on divers.

Craven saw opportunity, especially in the DSRV program. Like nearly everyone else with knowledge of the oceans, he knew that the DSRVs were largely fantasy. But he reasoned that maybe the push to build them might give him an edge in pursuing another of his dreams-a fleet of mini-submarines made of glass. Chemically, glass is a liquid, so Craven reasoned that glass submarines would be at their strongest under the most powerful deep-ocean pressures.

He wasn't the only one trying to sell the Navy on the idea of some kind of mini-submarine. Reynolds Aluminum Company was building its own boat, hoping to gain a lucrative contract. The Woods Hole Oceanographic Institution, with the Office of Naval Research, was designing the Alvin, a three-man submersible that could go down 6,000 feet. At this point, the only deep-submersible the Navy had inhouse was the Trieste If, a mini-dirigible that had to be carried or towed to dive sites. It had only limited maneuverability, but it could bring a crew of three down to 20,000 feet. The first Trieste had been lowered nearly 7 miles in 1960 to the deepest spot in the world-the Challenger Deep in the Marianas Trench, about 200 miles from Guam. Both Trieste I and II explored Thresher's wreckage.

It was just as Craven began to work out the mechanics of self-propelled, independent, deep-sea mini-subs that he was approached by a Naval Intelligence officer, one of the men who helped coordinate the submarine surveillance operations off the Soviet coasts. By now, those operations had been expanded to provide a year-round presence. Operating under the code name "Binnacle"-later "Ilolystone"-the Navy's growing fleet of nuclear subs and diesels were keeping constant watch on the Soviets as they aimed test launches of missiles from land silos and ships into the oceans. U.S. subs were also tracking the rapidly expanding fleet of Soviet nuclear subs as they finally began to venture out into the Atlantic and the Pacific. The Soviet Navy was beginning to enact its long-threatened plan to become a blue-water force.

With all this going on, the U.S. Navy nearly always had at least one surveillance sub in the Barents and two off the Soviets' Pacific ports, where they still had to dodge occasional Soviet depth charges. Even some of the early nuclear subs, like the USS Scamp (SSN-588), got chased with small depth charges, and more diesel subs, such as the USS Ronquil (SS-396) and the USS Trumpetfish (SS-425), got held down Gudgeon-style in the early 1960s. In addition to these operations off the Soviet coast, some diesel subs carried Russian emigres back to the Soviet Union to spy for the United States, and other diesel subs were landing commandos in places like Borneo, Indonesia, and the Middle East to track the expanding Soviet influence.[2]

Submarine spying had become so important that the chief of Naval Operations in Washington had taken charge of coordinating all operations, and a special undersea warfare office had been set up within the Office of Naval Intelligence to plan them.

Intelligence officials were so anxious to learn the latest about new Soviet subs and missiles that submarine spooks were under orders to flash off messages with mission highlights on the transit home. The Russian-language experts among them began transcribing tapes of stolen communications as soon as they left Soviet waters. Couriers met returning submarines at the dock, ready to whisk the intelligence directly to NSA headquarters in Fort Meade, Maryland. The spooks themselves were so valuable that the Navy ordered them to travel to and from ports by train rather than on commercial plane. The Navy wasn't willing to risk even a slim chance that they might be hijacked to Cuba.

Now the Naval Intelligence officer had come to Craven, asking him to help with a grander effort than any that had been tried before. The officer handed Craven a top-secret document, actually a very long wish list that Naval Intelligence had been amassing for several years, a document that had been touched by barely a dozen people before him.

Stamped across the front page were the words "Operation Sand Dollar." From there the list went on for pages. These were the splashdown points for Soviet ballistic missiles painstakingly monitored and noted by Navy surface ships and Air Force radar and underwater hydrophones, as well as the locations of planes and other Soviet military hardware glimpsed or heard plunging through the waves. Only a few miles away, three at most, lay the Soviets' most sensitive defense secrets: the best in Soviet missile guidance systems, metallurgy, and electronics-all of it tantalizing trash and all of it out of reach. No wonder the Soviet Union didn't even try to guard the cache. Nobody could have imagined an undersea raid through stars of luminescent plankton to the utter blackness of the deep.

But why not, intelligence officers reasoned, use the comforting notion of deep-submergence rescue vehicles to mask an effort to reach the items cataloged in Sand Dollar? Why not use the budgets of rescue gadgets that would hardly ever be used to create some tools that might just give the United States the definitive edge?

The Thresher tragedy would be the excuse, the new safety programs the stuff of a complicated cover story. And all of it was dependent on Craven's answer to one question. Could he manage a deep-water treasure hunt?

It was a matter of top national security, Craven was told. Left unsaid was that it was also a matter of pride, political standing, and turf. The intelligence arm of the Navy was in a desperate game of catch-up with that of the Air Force, which had just launched a new generation of spy satellites. With their growing coverage of the Soviet Union, these new "eyes in the sky" were sending back images of sites where the Soviets were digging silos for powerful landbased missiles and dry docks where the Soviets were preparing to create their own generation of Polaris-like submarines. The Polaris program had managed to prevent Air Force bombers and rockets from monopolizing the business of nuclear deterrence. Now maybe Navy spies could compete with the satellites, diving not for mere pictures but for actual Soviet arms and craft.

This was the opportunity Craven had been looking for, a chance to tap into his most fantastic plans. There was only one thing stopping him. He had no idea how to accomplish what the intelligence officer was asking for. Even Trieste II couldn't manage much of a secret undersea raid-it was too small, and the surface ship needed to carry the submersible out to mid-ocean would be a dead giveaway.

"Basically, we are developing the technology, but not the assets," Craven said, calling upon his best Navy-speak. Silence. Two beats, maybe three. No matter how officiously he said it, he was still admitting he had no way to do what he was being asked.

Then, Craven had a flash of inspiration. "Hey, look, we don't have anything that could do your operation because that requires things be clandestine." One more quick inhale and he came out with the kicker. "So it's really not worth doing Sand Dollar unless you do it from a submarine."

There it was, blurted out in desperation, the idea for what would become the Navy's most daring venture yet. A full-sized submarine, big enough to navigate the high seas, would be outfitted to hover in place in the upper reaches of the ocean and dangle cameras miles down, deep enough to scout the ocean bottom for Soviet treasures. It was inspired. Make the effort from below the surface, find a way to be nearly undetectable. Never let the Soviets know the Americans were anywhere near.

Actually all Craven was doing was rehashing his long-held belief that operating from the ocean surface was its own kind of hell. He had already included the concept in his self-scripted "Ten Commandments of Deep-Ocean Engineering." The way he said it was: "Remember that the free surface is neither ocean nor air and that man cannot walk upon it nor will equipments remain stable in its presence. So design your equipments that they tarry not long and that they need neither servicing nor repair at this unseemly interface."

Now, suddenly, he had not only the means to put that commandment to the test, he also had his chance to fulfill his favorite part of his lineage and plunder buried treasure. His pulpit secure, his corsair's blood aboil, all Craven needed was a submarine.

There were twenty nuclear attack subs in the fleet now, and more being built. But Navy admirals weren't about to give up a first-rate boat so that it could sit out in mid-ocean trolling with cameras. If Craven wanted a sub, he would have to take one of the Navy's two nuclear clunkers, the two failed experiments whose designs were never replicated. There was the USS Seawolf, a confused boat with the v-shaped bow of a destroyer and the top of a sub built to house a touchy reactor run with liquid sodium-a reactor that had been replaced early on. Then there was the USS Halibut, a boat with a grander, but short-lived, past. Halibut (SSGN-587) had been the only nuclear sub to carry the Regulus guided missiles, making seven missions off the Soviet coast. But that program had ended in mid-1964 when the Navy began basing Polaris subs in the Pacific. With the Regulus era over, no one knew quite what to do with Halibut.

She was a marine oddball, one of the least hydrodynamic of the nuclear fleet and one of the most ridiculous-looking creations ever born in a dry dock. Unlike the flat fish she was named for, Halibut wore a huge hump that might have been appropriate on a gargantuan desert creature except for the fact that it opened up into a large shark's-mouth hatch, part of the original missile hangar. Perhaps in another time, Halibut would have been quietly scrapped. After all, this boat was not only odd, she suffered from what was a near-fatal malady for a submarine: hydromechanical cacophony. Halibut was loud. Submariners heard the din, saw only potential flooding when they gazed upon that hatch, and shuddered when they examined her cumbersome ballast tanks, gaping caverns originally designed to allow her to surface fast, shoot a missile, and submerge even faster.

Craven took one look at the submarine it seemed nobody could love and was transfixed. All he saw were the possibilities, the strange and wonderful things that could be done with all of that excess space. And when he caught a glimpse of that gorgeous gaping mouth, it was enough to send him, like any self-respecting mad scientist, reeling with joy. No other submarine in the fleet boasted a hatch larger than 26 inches. Halibut's hatch was 22 feet.

It was settled: Halibut would be Craven's submarine, his laboratory, his ministry, his pirate ship. He would have $70 million to outfit her with electronic, sonic, photographic, and video gadgets. The Navy put out the word, and in February 1965 Halibut went into Pearl Harbor to be refitted as an oceanographic research vessel.

Less a lie than a huge omission, that was only one of several cover stories Craven would employ. The DSRV program and his other deepsea projects would add more layers, all hiding what Craven had proclaimed to be his "Skunk Works"-a term he borrowed for its drama from Lockheed Aircraft Corporation, the spy-plane manufacturer Craven would soon have working on the design of the DSRVs and also on a Deep Submergence Search Vehicle. The plan was that the DSSV would be able to sit on the ocean floor 20,000 feet deep and pick up objects with a mechanical arm. It was to travel to any recovery area mounted on the top of a submarine.

It would take two years to rebuild and test Halibut, but Craven would have little time to be impatient. From almost the moment the refit began, Craven's mass of cover stories began earning him notice outside the insular realm of Naval Intelligence. Suddenly he was being pulled into other high-profile projects.

Rickover, who had once done everything he could to limit Craven's interest in deep-diving mini-subs, now came to Craven asking him to help build the first nuclear-powered one, though one of steel, not glass. (The admiral would forever remain scathing about glass, to the point of insult.) But Craven was now working with Rickover, and the liaison would prove to be a crucial step in the scientist's education: crucial because it was through Rickover that Craven would learn how to mine the Navy's budget, deal with Congress, and handle the cadre of admirals who ran the submarine program.

It was a Faustian pact. Rickover may have been sixty-four years old, an age at which even less controversial officers have long been retired, but Craven, like just about everyone else in the Navy, could never quite learn to handle him. Rickover liked to begin their conversations in a way that showed just who was in charge: "Craven, my people are more competent than your people, but your shop is bigger, so I'm going to have to work with you." Rickover liked to try to throw men off balance just to see how they would handle themselves.

Rickover personally christened the mini-sub "NR-1." It might as well have been called the USS Rickover, for "NR" was the designation for the Naval Reactors Branch-Rickover's realm. If the president could have Air Force One, Rickover would have his NR-l.

Unlike Woods Hole's Alvin, which was completed in 1965 and was only 22 feet long, NR-1 was to be 137 feet, nearly half the size of an attack sub. NR-1 would be able to go down to 3,000 feet. Equipped with underwater lights, cameras, and a grappling arm to retrieve small objects, it also would have the potential to do some spying. One of the chief design problems was finding some way to shield the nuclear reactor in the NR-1. Standard sub reactors were shielded with a foot of lead on either end. But that would have made the NR-1 too heavy. Instead, Rickover, Craven, and the other designers decided it would have the standard lead shielding only in front where it faced crew compartments. The entire thirteen-foot area of the sub behind the reactor would be closed off permanently and flooded. The idea was to allow the wall of water to absorb any escaping radiation and work as a substitute for lead shielding. Craven had no doubt that environmentalists would cringe at the plan, but both he and Rickover believed it was entirely workable. Thirteen feet of water has the same molecular weight as one foot of lead. But when NR-1 was submerged, the water would add no weight at all-when water displaces an equal amount of water, the effective weight is zero.

But before NR-1 could be built, it had to be paid for, and right now there was little room in the budget for a mini-submarine among the plans for DSRVs and SeaLahs. The problem didn't faze Rickover, and he solved it at a meeting with Craven; Rear Admiral Levering Smith, Raborn's top deputy on Polaris; and Robert Morse, the assistant secretary of the Navy for research and development.

"You have any money we can get started with right now?" Rickover asked. Craven answered that his deep-submergence group could spare $10 million of its research and development money. Smith noted that the Polaris program had about $10 million of unused ship construction funds.

"How much is this submarine going to cost?" Morse asked.

Without hesitation, Rickover answered: $20 million. Morse went on to outline the tortuous process by which ships are normally built: contract definition, bidding, congressional approvals. Rickover cut him off before he could finish. "Just leave all that to me." Then Rickover turned to Craven and directed, "You call up Electric Boat tomorrow and tell them to get started."

Craven, Smith, and Morse exchanged looks of disbelief. Nobody believed this could be done for $20 million-the budget soon grew to $30 million. They also saw no way that Congress was going to stand for this. Less than a week later, Rickover called Craven and told him that the president was going to announce that afternoon that NR-1 was going to be built.

Upon hearing the news, Morse moved quickly from a state of shock into a state of panic. Up until that moment, NR-1 had been little more than an admiral's fantasy; indeed, Rickover had given only sketchy accounts of his plan to Paul H. Nitze, the secretary of the Navy, and Robert S. McNamara, the secretary of Defense. Though both had approved it, Morse knew Congress was not going to like hearing about a major project this way. As soon as the president announced the NR-1, the House Committee on Appropriations hastily called a hearing.

Craven, on Rickover's orders, had just a few days to come up with an official mission statement, a full-bore cost-benefit analysis, and a detailed study as to why the Navy needed the mini-sub.

"Well, you know, Admiral, that study really doesn't exist," Craven answered.

"It will exist by the time the hearing takes place," Rickover barked back.

Now the existence of NR-1, and perhaps his own career, rested on Craven's ability to spin visions from a black hole. He needed to prove that NR-1 was a crucial investment, one worth $30 million.

The appropriations committee wasn't fooled, but in the end it had no choice but to give in. NR-1 was now a presidential directive. No other submarine or ship had ever been authorized faster, or ever would he again. Later, the General Accounting Office, Congress's investigative arm, scrutinized the project and concluded that it was one of the worst managed programs its investigators had ever seen.

Rickover answered in typical form, firing off a letter to his critics that so amazed Craven that he committed it to memory. "I read the GAO report, and it reminds me of a review I read of Lady Chatterley's Lover in the magazine Field and Stream. The reviewer of that book knew as much about the real purpose of Lady Chatterley's Lover as the GAO knows about the design and development of submarines."

Rickover was no gentler on Craven. The admiral was infuriated that he had to share Craven with the Halibut refit, the DSRV program, and the other deep-ocean projects. As far as Rickover was concerned, none of those was more important than his NR-1.

Making the admiral even angrier was the fact that he was not cleared for the details about Halibut's new mission. Little went on within the submarine force that he didn't know something about, but the intelligence programs were one of the few areas in which he had no official "in," no real say. He took out his frustrations on Craven, who began to imagine that the admiral was waiting up nights before calling, waiting until Craven fell into a deep slumber or thought about romancing his wife. He was almost convinced that his time submerged conducting tests on Halibut was being monitored by Rickover, who seemed to time his calls for moments when it was impossible for Craven to answer. Craven always paid dearly for being unavailable.

One Friday, Rickover was giving a speech in New York City and he sent word to Halibut, which was out near Hawaii, demanding that Craven meet him in New York first thing Monday morning.

Craven hopped a plane, suffered a moment of panic when the flight got socked in by fog during a layover in Los Angeles, and finally landed in New York and raced breathless to Rickover's hotel suite, where the admiral was waiting. "You were out there playing golf with the beach boys," he said, mocking the cover story Craven had crafted for his trip to Hawaii.

He then turned to the house phone. "Bring this man the biggest lunch in the hotel." Craven waited for the punch line: he knew the admiral wasn't worried that he might he hungry after the long trip.

Sure enough: "For the next hour, you are going to sit and eat lunch," Rickover announced. "And I am going to bawl you out."

Appearances to the contrary, Rickover liked Craven almost as much as he liked making him miserable. Rickover was impressed that Craven had moxie enough to withstand his worst tantrums. The admiral also loved that Craven had never attended the Naval Academy. Rickover had been a loner as a midshipman, and now he made great sport of adding a little torture to the mix when he interviewed Naval Academy graduates for his nuclear program. Those entrance interviews had become more like initiation rites, in which the admiral took young men to the psychological brink in his quest for perfection. Trying to rattle his applicants, Rickover would spout obscenities, seat them in chairs with one leg cut short, or send them off to "Siberia," a storage closet where they would be left for hours.

Perhaps the all-time Rickover classic occurred when he squared off against a candidate and said, "Piss me off, if you can." The young man answered without hesitation and without a word. He lifted his arm and with one motion swept Rickover's desk clean of books, papers, pens, everything. The candidate was accepted.

For Rickover, torturing Craven was a mere sideline.

Craven, meanwhile, was increasingly on call as the Navy's resident deep-ocean expert. But there was one call that stood out from all the rest. It came on a Saturday morning in January 1966.

"This is Jack Howard," said an assistant secretary of Defense in charge of nuclear matters. "I've lost an H-bomb."

"Why are you calling me?" Craven asked.

"This one I've lost in the water, and I want you to find it." Craven was being assigned to work with a team hastily assembled by an admiral in the Pentagon. Another group was going to the site.

A B-52 bomber had collided with an air tanker during a refueling operation 30,000 feet in the air off the coast of Palomares, Spain, losing its atomic payload. Three of four bombs were recovered almost immediately. But a fourth was lost and had presumably fallen to the bottom of the Mediterranean. President Lyndon Johnson knew the Soviets were looking for the bomb, and he refused to believe the Navy's assurances that there was a good probability that it would never be recovered by either side. Indeed, that was the belief of most of the people assigned to find the bomb-but not Craven.

Craven called in a group of mathematicians and set them to work constructing a map of the sea bottom outside Palomares. That sounded reasonable enough, but Craven intended to use that map for an analysis that seemed more reminiscent of racetrack betting than of anything ever put down in a Navy search and salvage manual.

Once the map was completed, Craven asked a group of submarine and salvage experts to place Las Vegas-style bets on the probability of each of the different scenarios that might describe the bomb's loss being considered by the search team in Spain. Each scenario left the weapon in a different location.

Then, each possible location was run through a formula that was based on the odds created by the betting round. The locations were then replotted, yards or miles away from where logic and acoustic science alone would place them.

To the uninitiated, all this sounded like the old joke about a man who loses his wallet in a dark alley. Instead of searching the alley, the man chooses to search for his wallet yards away under a street lamp because the light there is better. But as far as Craven was concerned, there was good science behind his apparent madness.

He was relying on Bayes' theorem of subjective probability, an algebraic formula crafted by Thomas Bayes, a mathematician born in 1760. Essentially, the theorem was supposed to quantify the value of the hunch, factor in the knowledge that exists in people beyond their conscious minds.

Craven applied that doctrine to the search. The bomb had been hitched to two parachutes. He took bets on whether both had opened, or one, or none. He went through the same exercise over each possible detail of the crash. His team of mathematicians wrote out possible endings to the crash story and took bets on which ending they believed most. After the betting rounds were over, they used the odds they created to assign probability quotients to several possible locations. Then they mapped those probabilities and came up with the most probable site and several other possible ones.

Without ever having gone to sea, the team now believed they knew where the bomb was. According to their calculations, the most probable site lay far from where the other three bombs had been recovered and far from where most of the plane's debris had lilt the water. Worse, if Craven's calculations were correct, the bomb lay in a deep ravine and was all but unreachable.

The Navy had come across a Spaniard who was reputed to be the very best fisherman in Palomares, Francisco Simo-Orts. Simo Orts claimed to have seen the bomb fall into the water, and he pinpointed its location right over the same ravine. With no other leads, the team in the Med had no choice but to arrange a serious search of the ravine and began contacting the companies that had tried to interest the Navy in their deep-diving submersibles.

The Bureau of Ships agreed to pay to fly two submersibles to Palo- mares, Reynolds's Alurninaut and Woods Hole's Alvin. After several weeks and no success, President Johnson was furious. He demanded to know where the bomb was, and he demanded to know just when it would be recovered.

In answer, a copy of Craven's latest probability hill-altered to take the weeks of failures into account-was sent to the president.

Johnson blew up at the sight of Craven's curves and graphs. If the search teams couldn't give him instant answers, the president would find scientists who could. He insisted that another group of scientists be hired from Cornell and the Massachusetts Institute of Technology. They met in an all-day session. In the end they agreed that Craven's plan was the best anyone had.

Johnson didn't have much time to react. For that same day, the crew of the Alvin, on its tenth dive, sighted a parachute enshrouding a cylindrical object. It was 2,550 feet underwater wedged into a 70degree slope. The Alvin had found the missing H-bomb right where Craven's latest calculations put it. It would take several more weeks to recover the bomb. First the Alvin tried to hook it, but the bomb fell back into the water and was lost for another three weeks. Then the Navy dangled a robot, the cable-controlled underwater recovery vehicle (CURV), from a surface ship. The recovery team almost lost both the CURV and the bomb on April 7, 1966 when the robot failed to hook the bomb and instead became entangled in the parachute attached to the weapon. In desperation, the Navy decided to hoist both the CURV and the bomb up together, hoping the tangle was enough to bring both up to the surface. It was a less than elegant recovery, but it worked. More important to Craven, he had proven his theories. He was certain now that he could work miracles once he had Halibut.

He didn't have long to wait. Halibut was deemed finished just three weeks after the H-bomb recovery.

From the outside, she didn't look much changed. Her already towering sail had been raised to make room for extra masts that held periscopes and antennas to intercept communications to and from Soviet ships that might give chase. Atop her bow there was a small lump anyone could mistake for a misplaced dome of the type used to hold sonar arrays. In truth, that lump was something Craven called a thrust/vector control. It was a gadget he had originally doodled on the back of an envelope, and it allowed water to flow into Halibut's front and out her sides, causing the boat to hover nearly still in the water. Halibut could not only scan the ocean bottom, she could hang over objects, giving the Navy time to study them, perhaps one day giving Halibut divers the opportunity to slip outside of the sub and retrieve.

Inside, Halibut had been sliced, gutted, and given innards unlike any carried by other submarines. That camel-like hump with its gaping hatch had been transformed into a technological cavern now christened the "Bat Cave."

With gray, brown, and sky-blue laminates highlighting the stainless steel of its walls, the cave opened up 28 feet wide, stretched 50 feet long, climbed 30 feet high, and was divided into three levels.

There was a dark room, a data analysis room, and a computer room stuffed with one massive computer: the Univac 1 1 24. It was a huge machine with big tape reels and blinking lights, and it gave the cavern the feel of the science fiction-adventure realm for which it was named. (Still, Univac had only a tiny fraction of the power of the average modern laptop.) Crammed everywhere else were hunks, enough for a team of sixteen submariners and spooks.

Craven's crowning achievement was Halibut's "fish," which he hoped would swim through the deepest deep. Weighing two tons each, and spanning 12 feet long, these aluminum creatures had cameras with battery-powered strobe lights for eyes, whiskers of towed sonar arrays, and rudders and bow planes for fins. Designed to be towed from the bottom of the Bat Cave on several miles of cable, they had been spawned by Westinghouse Electric Corporation for $5 million each.

As Craven and company prepared the final round of tests on Halibut, he was meeting with his intelligence contacts almost daily in special soundproof rooms. He made a game of juggling his myriad other projects, all the while keeping the clearance-deprived within the Navy completely in the dark. There were cover stories within cover stories as he was called upon to solve various problems of the deep. There were the continuing demands from Rickover, as well as concern from Congress about why his deep-submergence projects were spending tens of millions of dollars more than anticipated. The overruns were, of course, being sunk into Halibut. But the project was one of the most highly classified in the Navy, and Craven could no more disclose those costs than he could his own whereabouts when he was on the sub.

Other programs became his casualties as he spread Halibut's costs in bogus budget items throughout the Navy. One poor captain was ordered to stash Halibut expenses in the accountings of a missile-warhead program, then faced weekly meetings at which he had to find some way to explain why his team was so badly overspending its budget. Another of Craven's favorite hiding places was the DSRV program. There was a certain poetry to this, since Craven was working on a fake DSRV that would one day be welded down to the back of Halibut to serve as a decompression chamber for divers. By the time Craven was done, the DSRV program had gone 2,000 percent over budget.

The sum so appalled Senator William Proxmire that the Wisconsin Democrat gave the project his "Golden Fleece Award." The DSRV, he declared, had one of the worst budget records in U.S. history. The Navy was horrified at the public dressing-down. Craven was elated. How many pirates get handed a cover story written by a senator?

Of course, Rickover eventually found out what Halibut was doing. He pushed until he knew most of the details. When he was refused by intelligence directors, he went straight to the admirals in charge of submarine operations. He would not accept that there were operations using his submarines that would take place without him. The admirals didn't dare turn him away. But intelligence officers bristled at his interference. For one thing, Rickover refused to sign the standard secrecy oaths, believing that his loyalty should be taken for granted.

Halibut's officers did little to make the job of appeasing the admiral any easier. When one of Rickover's inspectors sought to keep the submarine docked over concerns about the way the crew was handling the boat's reactor, Halibut's skipper, Commander Harold S. "Hank" Clay, refused to bow to Rickover's authority. Halibut operated under the highest priority code in the military, and the way the story was told on board Halibut, Clay barked at Rickover's man, "You want to fail me, fail me. You tell the president I can't get under way. This boat has Brick-Bat 01 authority."

Clay had enough problems without Rickover's interference. Halibut's test runs weren't going well. None of the spy equipment had been built to any of the normal military specifications. The military had, in fact, never devised a set of specs for anything that would operate 20,000 feet down. And so, by trial and error, mostly error, Halibut's crew tried their best to make all of her space-age equipment function. In these early days, the crew was becoming convinced that gremlins had moved into the Bat Cave. There were never-ending computer problems. The computer's "Interleaf" operating system needed more than the computer's 32 kilobytes of memory to operate. When computer components in the fish failed, new ones were secreted into Pearl Harbor in the luggage of American Airlines stewardesses.

Then there was the rest of Halibut's deep-sea equipment. Her crew was discovering that systems that functioned fine at a few hundred feet underwater just didn't work the same way 15,000 feet deep, where pressures were enough to crush any slight flaw or weakness into a full-scale failure. The tiny, gold-plated rubber connectors used in the fish's wiring failed at 10,000 feet when the gold and the wire began to compress at different rates, sending the gold flaking off and breaking the circuits.

The strobe lights, so carefully designed to ride the fish and light the sea floor, worked too well. They were so bright that they blinded the cameras. Ultimately, dimmer lights were built. Unfortunately, the video signal failed to survive the climb through the coaxial cable that toted the fish, one at a time. So on Halibut's early missions, the crew would have to make do with grainy sonar images of shadows, bright spots, and shapes. The crew would be able to grab hold of clearer photographs only once every six days, when a massive fish was hoisted back aboard, carrying its film to the surface.

"If something is worth doing, it's worth doing badly," Craven kept repeating, trying to ease the pain of failure. Meanwhile, he met weekly with the fish designers at the Westinghouse plant in Maryland, hoping to trade his tales of disasters for solutions.

"Okay, fellas, we are going to have a wire brushing, but I want you all to smile," he started each meeting, commanding grins at times, grimaces at others.

One day the engineers decided to answer his greeting in kind. They handed Craven a clear plastic box. Inside was a wire brush. His name was stenciled on the back of the brush which lay next to a small plate engraved with one word: "SMILE."

On one of the last runs to test the fish, a surface ship was supposed to drop an object into the ocean. The idea was for the fish to he employed in a scavenger hunt. Halibut's crew would have to identify the object, which would be hidden from periscope view by a huge box. The box would open from the bottom and drop the object unseen into the depths.

The day came, the weather was good. Halibut and the surface ship set out to sea. A crane on the ship lifted the box and lowered it until it barely dangled above water. Then the box opened from the bottom. Moments later, the bad news came over the ship-to-sub radio: the object that the Navy had taken such pains to hide was floating.

The crew on the surface ship hauled the object back aboard and began wrapping it in canvas and heavy anchor chains, lots of them. They threw it back overboard. Soon after, the Naval Investigative Service sprang into action, sending officers on board to force promises of confidentiality from all the men on the surface ship, who now knew exactly what their secret cargo was. Judging from the size of the box and the investigators' reactions, the secret object was probably designed to resemble a missile's nose cone.

For the next few days, Halibut searched. Somewhere along the line, a control rod got stuck at the bottom of Halibut's reactor chamber, shutting it down and forcing the boat to resort to diesel engines. Then one of the camera-toting fish was lost, joining all of the high-tech, sensitive trash it had been designed to find. Craven had expected some sort of fish disaster. He had ordered six of the contraptions, although Halibut was designed to carry only two at a time. As far as he was concerned, they had just dropped a spare-a very expensive spare.

Finally, the other fish was lowered and captured the images the men had been seeking. Later, and with some glee, the special projects crew proudly paraded a photograph of the object of their search around the boat.

Craven had just logged a major success, the first indication that Halibut might actually be able to accomplish all she had been rebuilt for. But Halibut's men couldn't see that. Most of the photograph had been blacked out for security reasons. As far as the men of the sub were concerned, they had just pulled off a massive search for nothing more than a lump of tangled anchor chain.