Robert Amory, Jr., a solid Republican, the type who would run for office in Massachusetts, a Democratic state, did not shrink from expressing doubts to this new tribe of Democrats. Amory had already seen the depths of failure. Hungary, several years into his watch at the DI, had affected Amory deeply. In atonement, he and his wife Mary subsequently sponsored a Hungarian refugee into the United States. Amory had no desire for the same thing to happen to the anti-Castro Cubans. He also had a lot more experience with invasions than anyone involved in the Cuba project. In World War II Amory had been a landing-craft operator, finishing the war as a colonel in charge of a whole regiment of the vessels, a veteran of twenty-six assault landings in the South Pacific, many of about the same dimension as the planned CIA operation. Jack Hawkins, the actual invasion boss, had participated in exactly two, the massive Iwo Jima endeavor, where the United States had held all the cards and put ashore many thousands of troops, and the large Inchon operation in Korea. Amory thought Hawkins “just didn’t know beans about what a small self-contained beachhead would be like.” The Cuba plan failed Amory’s smell test. But no one consulted him. Perhaps they were scared off by his Castro costume.

  Amory, far from alone in his view, represented the element at CIA that was skeptical about the Cuba project. Some doubted the internal resistance could cripple Castro, others questioned that outcome even if the CIA sent in its exile brigade—in other words, they found the project dubious even with CIA at full stretch. Another DI unit was the Office of National Estimates (ONE), which produced the community’s official flagship reports, the National Intelligence Estimates. Without bringing it into the circle of knowledge on the plan, Dulles and Bissell asked ONE to assess Castro’s grip on power. ONE’s memorandum in late January could not have been reassuring to the secret warriors. The analysts judged that time was not on Washington’s side, that Castro had successfully consolidated his control.

  Barely a couple of weeks earlier Director Dulles had participated in the final reunion of President Eisenhower and his Hull Board watchdog group. The board had long complained that the Directorate for Operations did not incorporate intelligence analyses into its covert action plans. Dulles had defended his agency—and did so again at this January 1961 meeting—responding that the DO used intelligence estimates in all phases of planning. But in fact the ONE report had no discernible impact on the Cuba plan. Equally important, there is no evidence Allen Dulles did anything to ensure that it did.

  Two days after the inauguration, Project Ate received its first airing before the gaggle of officials who were Kennedy’s men. National security adviser McGeorge Bundy, Secretary of Defense Robert S. McNamara, Secretary of State Dean Rusk, and a host of others endured a long briefing where Brig. Gen. David W. Gray went over the Cuba project in eye-opening detail. For most the day brought their first knowledge that the United States had already created a secret army of Cuban exiles. Gray had poster-size charts and noted that U.S. participation was expected to be necessary.

  President Kennedy attended a full-dress presentation a week after the inaugural. While his Palm Beach conversation had been exploratory, the White House meeting on January 28 was specific. He listened as Allen Dulles mentioned what would soon be called the “disposal problem,” that the Cuban brigade had to get out of Guatemala soon, and what then? Dulles also told of Castro’s growing military power and, somewhat more fancifully, “a great increase . . . in popular opposition.” Discussion focused not so much on the invasion plan as on a comparison of that with six alternatives, including economic warfare or direct U.S. intervention. The official record notes the conclusion: “No course of action currently authorized by the United States Government will be effective in reaching the agreed national goal of overthrowing the Castro regime.”

  “It was very ethereal,” Bobby Kennedy recalled a few months later, but what stuck in his mind is that he remembers being told “it would be impossible to successfully overthrow Castro because of his control over his armed forces and the country in general, unless you had the invading force backed up by intervention by U.S. forces.” JFK received the ONE estimate judging time to be on Castro’s side. Kennedy ordered intensified political action, sabotage, overflights by CIA U-2 aircraft; and State Department preparation of an anti-Castro propaganda plan to be implemented throughout Latin America. He discovered that the U.S. military had not considered the feasibility of Project Ate and directed the Joint Chiefs of Staff to do such a review.

  Events began to move swiftly. A few days after Kennedy’s orders, the JCS were officially briefed on the CIA’s plan for the conventional invasion. This provided for a landing on the south coast, near the town of Trinidad and the Escambray Mountains. The Joint Chiefs’ official opinion, after a few days of study by a committee led by Brigadier General Gray, appeared in a paper titled “Military Evaluation of the CIA Paramilitary Plan—Cuba.”

  The seventeen conclusions in the JCS paper indicated continuing differences. On one hand, the military judged that if the airborne drop were successful, it would take several days for Castro to react to the landing, and thus, despite its shortcomings, the CIA had a fair chance of success. On the other hand, the Chiefs concluded that the Cuban army could reduce the beachhead. What constituted a fair chance? General Gray put it at 30-70. No one he heard went any higher than 40-60. Others estimated the chances against achieving surprise at a whopping 85-15.

  The military’s warning implied the need for rapid breakout from the landing site. But the CIA’s own view, articulated by Jack Hawkins in a January 4, 1961, report for Esterline, “Policy Decisions Required for Conduct of Strike Operations Against Government of Cuba,” held quite the opposite: Brigade 2506 should try to survive on the beachhead and not break out until the time became opportune or the United States intervened. Indeed, CIA planned to fly in Cuban politicians to form a provisional government while agency planners arranged for supply landings for a month on the beachhead. The conflicting views of military and CIA were not reconciled, and President Kennedy now lacked the supervisory staffs to tell him that.

  Allen Dulles meanwhile curried favor among the Kennedy people. At several morning staff meetings Dulles referred to the need to gain their confidence. One confidence-building measure came early in February when Dulles, a member of the exclusive Alibi Club, used that venue to host a dinner for key White House staff and top CIA spooks. Just over a block from the NSC staff offices, the Alibi Club, founded in 1884 by disaffected members of another Washington institution (the Metropolitan Club), has an unlisted telephone on a stand with a list of excuses next to it and the prices for each. The Alibi prided itself on decorating with oddments and things donated by members—just fifty at this time. Besides Dulles, future CIA directors Richard Helms, William J. Casey, and William Webster would all be members. Helms’s comment that “Very little plotting goes on at the Alibi Club” clearly does not apply to this evening in February 1961. Bissell brought down the house at this basic get-acquainted session, introducing himself by declaring, “I’m your man-eating shark!” But, even from this dinner, the secret warriors did not come away unscathed—someone asked them why the United States had gotten into so much trouble in the Indonesia operation, Project Haik.

  American intervention remained a sensitive matter. CIA understood the need to disable Cuban air forces that could disrupt the exile landing. Although a program of exile air strikes had been laid on, the secret warriors knew that Castro possessed some jet fighters. The exiles had no comparable aircraft. Support by jets, the most obvious form that U.S. intervention might take, had been mentioned both by CIA officers and Whiting Willauer. Kennedy rejected it, and the CIA knew of his reluctance.

  On February 9 Adm. Robert L. Dennison, commander of the navy’s Atlantic Fleet, sought clarification. At a discussion with Kennedy the admiral asked, “Am I likely to be involved in a bail-out operation?”

  “No,” replied the president. If there were any problems, the exiles would fade into the hinterland; American forces need not become overtly involved.

  The next day Dennison received a directive from Joint Chiefs chairman Lemnitzer defining the scope and restrictions on navy support, clearly to be minimal.

  By mid-February Esterline knew that Ate could not make the planned invasion date of March 5; Washington delayed it a month. The Guatemalans were asked to accept the delay. The military used this time to send three colonels to inspect Camp Trax and Retalhuleu. They assessed the brigade to be in good shape, but the odds were against surprise. The air evaluation stated that one Castro plane with .50-caliber machine guns could sink most or all the fleet, so if there were no surprise the operation would fail. Without air superiority, the invasion would also fail.

  On February 17 the brass had it out. By now the CIA had a paper arguing that the Cuban exiles could not infiltrate Cuba as guerrillas without huge losses, forfeiting effectiveness and being unable to rendezvous to conduct operations. In effect the disposal problem could best be solved by sending the brigade to invade Cuba. Bissell later recalled talk of loading the Cubans onto ships which, if not able to sail for Cuba after a reasonable time, could be escorted to a U.S. naval base (Guantanamo or Roosevelt Roads in Puerto Rico), and the exiles disarmed and set loose. No one wanted to think of the consequences of that. (Jake Esterline years later recorded that no one, from JFK on down, ever asked the WH/4 task force to develop a stand-down plan.) At the February 17 meeting Bissell stressed the need to decide when and how to surface the Frente political leaders. Secretary Rusk became the main adversary, arguing that the schedule left him little time to stoke up OAS support for regime change. Kennedy, having seen reports of Castro’s jet aircraft, asked if these were a result of the CIA effort. The reply was both encouraging and not: Castro’s jets had belonged to Batista, but if the United States waited much longer the Cubans would be receiving Russian jets, as they were already getting tanks and guns.

  In contrast to later decades, when Congress and the CIA tussled endlessly over current notification of covert operations, Allen Dulles told legislators about the Bay of Pigs in advance. On March 10 he appeared before the secret subcommittee of the House Armed Services Committee and informed the overseers not only of the timing of the original decision (“just about a year ago now”) but of the CIA’s efforts with the exile Frente. General Cabell followed to describe the exile army and air force. Answering questions, Dulles added that resistance to Castro would have to be sparked from the outside. Cabell claimed that increasing the size of the exile force would mean a lowering of military standards. The CIA deputy director also revealed the plan to seize and hold a position in Cuba rather than fan out to join the guerrillas. Legislators questioned size and feasibility. No one told the CIA it should forget about its invasion.

  How could the president interpret the conflicting reports that reached him? In the Oval Office one day Kennedy turned to Allen Dulles, asking him about the odds. The DCI alluded to his conversation with Eisenhower during Project Success. “I stood right here at Ike’s desk,” said Dulles, “and told him I was certain our Guatemalan operation would succeed, and, Mr. President, the prospects for this plan are even better than they were for that one.”

  Despite his doubts, Kennedy, McGeorge Bundy thought, kept looking for ways to make Ate work. Allen Dulles encouraged the president, treating him to another recitation of the disposal problem: “Don’t forget . . . we will have to transfer [the Cuban exiles] to the United States, and we can’t have them wandering around the country telling everyone what they have been doing.” Arthur Schlesinger believes that Washington’s desire to keep Project Ate going owed much to the embarrassment that would attend its cancellation. At Bissell’s full-dress presentation of the plan for an invasion at Trinidad in Cuba on March 11,* Dean Rusk again voiced objections. Rusk wanted an airfield there big enough to handle B-26 bombers so that strikes against Castro’s bases could be said to be coming from Trinidad. Told the field was not that long, Rusk wondered if the CIA could airdrop a bulldozer to lengthen it.

  “If I ever made a suggestion like that to Mr. Dulles,” Esterline interjected, “I should be summarily fired.” The Cuba task force chief wore out his White House welcome that day.

  JFK saw a daylight landing at a Cuban city as a “spectacular” invasion and asked for an alternative, convincing some that he sold out the Cubans that day. But Kennedy nevertheless issued a directive stating that he expected to approve the invasion. People outside the White House have a very different view of Kennedy’s decision: reducing the visibility of the invasion, an action tantamount to diminishing its chance of success, had roots in the president’s ambivalence. Kennedy’s decision on the “visibility” of the invasion is the second of the choices that have marked him with authorship of this disaster.

  Commentators on the Cuban operation often write of the Trinidad plan as if it represented a solution, that is, Project Ate would have been successful if only the CIA’s exiles had gone into Trinidad instead of their eventual target. But there were no panaceas. In fact the secret warriors overstated the possibilities at Trinidad while underestimating Castro’s defensive potential. An agent reported several thousand Cubans around the city ready to take up arms against Castro. With no means to cross-check, DO operatives took the dubious figure at face value. Meanwhile Castro, who had no spies in Guatemala but plenty in Miami, and whose security services carefully followed media accounts of exile activity, expected the CIA invasion at Trinidad. Where the Pentagon and CIA held that the nearest Castro militia or FAR troops were a hundred miles away, and evaluated their response capacity as a single battalion on the first day, Castro himself revealed at a 2001 conference that the FAR had concentrated two full brigades (six times the force) backed by thirty heavy cannon right at Trinidad, with observation posts overlooking the bay and preregistered artillery targets. Without counting Castro reinforcements, Brigade 2506 would have been landing against superior forces in prepared positions.

  But Trinidad, a big town, violated President Kennedy’s edict to reduce “visibility.” On March 12 Esterline got orders to redraft. In a frantic all-night work session, Jack Hawkins surveyed the Cuban coast for localities that met three criteria: not easily accessible to Castro forces; having an airfield capable of accommodating B-26 and C-54 aircraft, one that could be captured on the first day; and close to suitable beaches. Only a few places matched some of these criteria, and just one met them all: the Bay of Pigs, about eighty miles west of Trinidad. Two days later the paramilitary planners presented the Bay of Pigs option to the Joint Chiefs.

  On March 15 both options were outlined at the White House. JFK, to Schlesinger’s mind, listened somberly, again rejecting Trinidad as a “World War II assault operation.” He ordered the Bay of Pigs plan to be reoriented for a night as opposed to a dawn landing. No one told Kennedy that the United States had never carried out a major nighttime invasion. Other profile-lowering measures included halting rebel resupply flights at the end of March and cancellation of leaflet drops until after the landing. The CIA, whose chance of victory depended on mobilizing the Cuban population, foolishly accepted the flight stand-down, stopping the supply flow to the resistance. JFK also wanted to be able to call off the invasion up to the day before it happened.

  Finally it all came down to the people President Kennedy assembled in the Cabinet Room on April 5, his last real opportunity to cancel the invasion. Kennedy went around the table to poll participants. Fulbright adamantly opposed Ate, but the center of gravity was Adolf A. Berle, a former senior diplomat whom JFK had brought back to review Latin America policy. Esterline saw Berle as a strong supporter. Earlier, in a furious battle of memos, Dick Bissell had crossed swords with Tom Mann, the assistant secretary of state for Latin America. Now, like Mann, Berle delivered a lengthy disquisition, with a lot of “con” but a good deal of “pro” too. Asked where he finally stood, Berle declared, “I say, ‘let ’er rip!’ ”

  At this point task force chief Esterline returned to Guatemala for one final visit to the camps. As a morale booster he brought along rebel leader Jose Miro Cardona. The men flew black, out of Miami, crewed by some of the CIA’s Polish pilots. The secret flight became embarrassing when Cardona suffered a seizure as the plane neared Guatemala City. Oxygen had to be administered. Esterline sweated bricks that they would not need a hospital, which would have meant explaining their presence in Guatemala.

  President Kennedy continued wrestling with the go/no-go decision for a few more days, consulting Arthur Schlesinger again, but finally gave way. Kennedy even intervened with the New York Times when he learned the paper had a major exposé of the Cuba plan by reporter Tad Szulc ready to go to press on April 10. The president persuaded publisher Arthur Hays Sulzberger to downplay the story, though he failed to get it spiked.

  The image of the United States in world public opinion had to be considered. Self-determination had been bedrock American foreign policy since Woodrow Wilson. Plausible deniability remained necessary precisely so that leaders might deny U.S. involvement. At an important forum for world opinion, the United Nations General Assembly, then in session in New York City, a Cuba debate already figured on the agenda. Kennedy had appointed Adlai Stevenson his ambassador to the UN and believed Stevenson’s integrity to be a vital asset. Kennedy ordered Arthur Schlesinger to brief the ambassador together with CIA and State Department people.

  Schlesinger arrived late for the meeting in Stevenson’s suite at the Waldorf Astoria Hotel on the morning of Saturday, April 8. As a consequence, Tracy Barnes, CIA’s man, did most of the talking. Steeped in the agency’s “need to know” tradition, and despite close friendships with Stevenson’s senior assistants and his own acquaintance with the ambassador, Barnes imparted minimal information: this would be an all-Cuban affair merely helped by the United States, not a word about air strikes or other details. True, Bissell had told Barnes to say as little as possible, but Barnes made it impossible for Stevenson to play his role effectively. Barnes’s reticence was probably encouraged by the ambassador’s reaction—Stevenson doubted JFK had thought through the issues. Within days the effect of Tracy Barnes’s inadequate briefing became painfully apparent.