by Gregg Herken
* * *
Within hours of the meeting between Nelson and Joe, King had couriered a transcript of the conversation to G-2 at the Presidio, where the news hit Pash like a thunderclap. Flying to Washington the following day to personally brief Groves and Lansdale on the incident, Pash believed that the FBI bug provided proof of Oppenheimer’s involvement in an espionage plot.93
Pash assigned Lyall Johnson the task of discovering the identity of Joe. The conversation with Nelson had yielded several valuable clues: Joe had moved to California in 1939, and his wife was from Wisconsin; he had two sisters living in New York City, one of them a teacher.
At the San Francisco FBI office, King asked Branigan and Cassidy to begin immediate round-the-clock physical surveillance of Nelson.94 Aware that the FBI director had been trying since the 1930s to outlaw the Communist Party in the United States as a haven for spies, Pieper wrote Hoover: “I believe a definite possibility exists of developing an espionage case against the CP, USA and the Bureau’s advice is requested.”95
The stakeout of Nelson’s house on Grove Street yielded results the very next morning, April 1, when Nelson walked to a corner drugstore and telephoned the Soviets’ San Francisco consulate from a pay phone. Nelson was overheard using the name “Hugo” and asking to speak with Ivanov; the two agreed to meet at “the usual place” a few days hence.
On the evening of April 6, Branigan and Cassidy followed Nelson to the grounds of St. Joseph Hospital in San Francisco, where Nelson met Ivanov in an open area surrounded by trees.96
Four days later, Cassidy captured another visitor to Nelson’s house on film—a burly figure in an ill-fitting suit whom the FBI later identified as Vassili Zubilin, the recently appointed third secretary of the Soviet embassy in Washington. Despite his lowly title, the bureau suspected Zubilin—who in the United States went by the name “Mr. Cooper”—of being the new NKVD rezident at the embassy.97
The bug in Nelson’s house picked up the Russian counting out bills or bundles of currency, and Branigan recorded this exchange:
NELSON: “Jesus, you count money like a banker.”
UNKNOWN MAN: “Well, after all, I told (deleted), I used to pay out at Russia.”98
For more than an hour, Nelson and Zubilin talked about the Soviet espionage apparatus in America. Nelson said that he had been recruited at the end of 1942 by “a man from Moscow” and that Earl Browder, the head of the Communist Party in the United States, was aware of and supported his mission. He was miffed, Nelson said, that Soviet officials had begun short-circuiting the established apparatus by recruiting rank-and-file members of the party to spy, and instructing them not to inform party superiors of their assignments. Nelson suggested that the Soviets instead choose a trustworthy comrade in every important city or state and allow that person to handle contacts with the local spies. He proposed himself as the contact in the Bay Area.99
Several current and prospective agents were also mentioned by their code names during the conversation. Nelson complained bitterly about the incompetence of two of the couriers who were responsible for liaison between the West Coast espionage apparatus and Browder. Bill Schneiderman, state secretary of the party, was said to be reluctant to involve Communists in such “special work.” But Nelson told Zubilin that he nonetheless had a new recruit, one that “you can add to your company—her name is Bernstein.”100
* * *
The conversations intercepted by the FBI put a prompt end to the battle between the army and the bureau. On April 5, 1943, G-2’s Strong acknowledged to Hoover that the military was engaged in a secret “experimental program.”101 The following morning, Hoover’s personal assistant, D. Milton “Mickey” Ladd, and the FBI’s foremost Soviet expert, Lish Whitson, met with Lansdale and Groves in the general’s office at the New War Department Building. Without disclosing its goal—he spoke only of producing an unidentified “material”—Groves outlined the scope, cost, and significance of the Manhattan Project, adding, in notes taken by Ladd: “General Groves advised that which ever country got the material first would win the war and could dictate the terms of the peace. He further said that if Switzerland could make only so much of the material as would fill a small room, Switzerland could rule the world.”102
Groves’s speech was the first that Hoover would hear of the Manhattan Project—and the fact that it was already the target of Soviet spies. Following the meeting, the FBI director sent a bulletin to the bureau’s field offices, instructing them to report any instances of Communists soliciting information “re scientific experiments.”103 On May 7, Hoover sent FDR aide Harry Hopkins a memo outlining the Nelson-Zubilin conversation. (“I thought the president and you would be interested…”)104
The FBI intercepts also galvanized the army into action. The day before his meeting with the bureau, Lansdale had submitted detailed plans for a major counterintelligence effort in the Bay Area; Strong now approved it immediately.105
With the cooperation of the Rad Lab’s personnel director, Lyall Johnson placed army undercover agents on the research staff. One, an engineer, joined the local chapter of FAECT. Another secret informant, a secretary at the lab, reported to the army on particular people and events up on the hill.
Across the Bay in San Francisco, Pash set up a dummy business office—the “Universal Subscription Company”—in a building just off Market as a staging area for his undercover agents. An army lieutenant, James Murray, headed the plainclothes operation under the nom de guerre of Paul Sheridan.106
Two enlisted men, former telephone repairmen, installed wiretaps and bugs for the army in cooperation with the local telephone company. Under the arrangement agreed to between Hoover and Strong, the army focused upon university employees under contract to the Manhattan Project, while the bureau concentrated upon known or suspected Communists with connections to the Rad Lab.107
As its surveillance effort grew, the army rented a two-story house on Forest Avenue, a few blocks south of the Berkeley campus, to serve as a listening post. An undercover agent and his family lived downstairs; upstairs, in a back room, officers assigned to the Military Intelligence Division’s Counter-Intelligence Corps (CIC) recorded the telephone calls of selected Rad Lab employees.108 (On an early inspection visit, Groves nearly compromised the operation by arriving at the house in full uniform. Belatedly realizing that the neighbors might wonder what an army general was doing there, he wrapped himself in the diminutive Pash’s raincoat and dashed back to his car.)109
Lansdale ordered Pash to launch his own investigations at Berkeley—not realizing that the eager counterspy already had one under way.
Pash did not have to be persuaded that Oppenheimer was a security threat. Under the pretext of protecting Oppenheimer against Axis assassins, he had earlier assigned the physicist a pair of bodyguards, who reported regularly on Oppie’s activities. (Oppenheimer, however, foiled the eavesdroppers by keeping the back windows of the car open and by speaking in whispers.)110 A few weeks before, Pash had ordered a young MID lieutenant assigned to the Presidio, Peer de Silva, to begin a personal investigation of Oppenheimer. De Silva, just two years out of West Point, quickly came to share Pash’s suspicion that Oppie was a Soviet agent.111
* * *
Despite the army’s efforts, it would be another two months before “Joe” was identified.112 The break in the case came about purely by chance. Early in June, a commercial photographer stationed near Sather Gate snapped a picture of Rossi Lomanitz arm-in-arm with David Bohm, Max Friedman, and Joe Weinberg. After the group left, an army undercover agent standing nearby purchased the negative.113
Several days later, CIC agents followed Lomanitz, Friedman, and Weinberg home to Berkeley from San Francisco, where the trio and a woman friend had attended a celebration in honor of Soviet writer Maxim Gorki. The address on Blake Street was passed along to Johnson, whose check of personnel records at the lab showed an exact match between Weinberg’s background and that of “Joe.”114 Army investigators also
located a handwritten note that Weinberg had written to Oppenheimer on April 12—two weeks after Joe’s meeting with Nelson, and only two days before Weinberg filled out the required security questionnaire at the Rad Lab—indicating that he was already involved in the bomb project and eager to do more.115
Oppie had put Weinberg to work on magnetic field calculations—making a science out of what, for Lawrence, had been the art of focusing the cyclotron beam with iron shims.116 Despite Oppenheimer’s earlier reservations about hiring Weinberg, the urgent need to get the Calutrons running had short-circuited caution and the lab’s rudimentary security procedures.117
Army agents began following not only Weinberg but the other three in the photograph as well. Since Joe had indicated in his talk with Nelson that future secrets might be passed through his family in New York, the FBI also wiretapped Weinberg’s parents and his two sisters.118
In Pash’s view, Oppenheimer’s role in hiring Weinberg helped seal the case against the man Groves had chosen to lead Los Alamos.119 But the identification of Joe was overshadowed by the news—received that same day, via a wiretap on Oppie’s telephone—that the Los Alamos director was planning a surprise visit to Berkeley. Pash excitedly cabled Lansdale that he was preparing a reception: “Oppenheimer will be covered on arrival. Details by airmail. This office setting out all leads.”120
* * *
Oppenheimer arrived in Berkeley on Saturday, June 12, ostensibly to recruit a personal assistant.121 But his real purpose was to visit Jean Tatlock, whom he had avoided seeing before he left Berkeley.122 Tatlock was being treated for depression at Mt. Zion Hospital in San Francisco and had recently sent word to Los Alamos via Oppie’s former landlady that she wanted to see him.
For his assistant, Oppie chose David Hawkins, a young philosophy professor from Berkeley whom his brother, Frank, had known at Stanford. Army CIC agents followed Oppenheimer to Hawkins’s apartment in the Sunset district of San Francisco and later put a tap on Hawkins’s telephone.123
On Sunday evening, Oppenheimer had dinner with Lawrence, who was, as usual, cheerily optimistic about the progress of his Calutrons. Army agents in an adjoining booth at the restaurant overheard Ernest report that the lab was within sight of another milestone: 1 gram of U-235 per day.
On Monday morning, Oppie took the Key System train across the bridge to Jean Tatlock’s apartment on Montgomery Street, with Pash’s agents in tow. The couple had dinner at the Xochinilco Café on Broadway before returning to her apartment, where Oppenheimer spent the night. Army agents waited in a car parked on the street below.
Tuesday morning, Oppenheimer took the train back to Berkeley. He and Tatlock met again that evening for dinner, and afterward she drove him to the airport. Since Oppie had originally planned to be back in Los Alamos by Monday night, his tardiness required him to break Groves’s rule that grounded laboratory directors.
Oppenheimer’s rendezvous with Tatlock and his hiring of Hawkins—another suspected Communist—added further fuel to G-2’s suspicions.124 Pash warned Washington that the Los Alamos director was deliberately luring left-wing associates to the lab. Pash even hinted to Lansdale that Oppie might have agreed to work on the bomb just so he could give it to the Russians.125 The army asked the FBI to investigate Hawkins and to plant a bug in Tatlock’s apartment.126 Pieper—protesting that Pash was asking the bureau to “run errands”—initially refused both requests but later relented.127
Aware that Groves seemed committed to Oppenheimer, Pash broached with Lansdale the possibility of gradually easing Oppie out of his job at Los Alamos and replacing him with another scientist.128 The FBI’s Whitson also raised the question of firing Oppenheimer. Groves, however, would have none of it, as Pash and the bureau learned from the project’s head of security: “Lansdale stated that General Groves claims flatly that Oppenheimer is irreplaceable and that if anything happened to Oppenheimer, the project would be set back at least six months.”129 Moreover, Groves had argued, such a delay could be catastrophic—since recent intelligence showed that the Germans were laying new high-tension wires, leading the army to conclude that the Nazis might be building their own Calutrons.
A few weeks later, on July 20, 1943, Groves put a quick and decisive end to the long agonizing over Oppenheimer by summarily ordering the district engineer to issue a security clearance for Oppie “without delay, irrespective of the information which you have concerning Mr. Oppenheimer. He is absolutely essential to the project.”130
6
A QUESTION OF DIVIDED LOYALTIES
LANSDALE DECIDED UPON a quick visit to the desert mesa later that summer, to take his own measure of Oppenheimer. Without alluding to the FBI’s wiretaps, Groves’s security man told Oppie that Lomanitz had remained active in FAECT and other radical causes. Oppenheimer expressed anger at the graduate student, saying that he wanted no Communists working on the project, for “one always had a question of divided loyalties.”1 Lansdale was just about to ask Oppenheimer for the names of party members at Los Alamos—he had told Whitson that he intended to fire them all—when the conversation was interrupted. The question remained unasked.
Lansdale came away from his visit convinced that Oppenheimer’s ego and Kitty’s vicarious ambition for her husband were the two best reasons for believing that neither would endanger the success of the project by allowing its secrets to be passed to the Russians. When the head of MED security shared this view with Groves, the general agreed.2
At Los Alamos, de Silva had already alienated the scientists by imposing new restrictions on travel to Santa Fe and by confiscating their cameras—measures for which Groves, as usual, got the blame.3
* * *
The bad news had come to Berkeley early in 1943, when Oppie had informed Lawrence that calculations indicated the fissionable material for the bomb had to be more than 90 percent pure U-235. Using uranium of any lower enrichment would mean a weapon too big and too heavy for a bomber to carry.4 Accordingly, the metal produced by the Alpha Calutrons would have to undergo a second, more careful separation. Lawrence and the boys had already designed a new machine for the purpose.5
Groves, looking ahead, also wanted to use the so-called Beta Calutrons to further enrich uranium that had been separated by other methods like gaseous diffusion. Work was begun on two Beta racetracks alongside the five Alpha Calutron racetracks already under construction on the 825-acre plot of land southeast of Oak Ridge. Lawrence’s cyclotroneers had meanwhile given Y-12 their own, more affectionate appellation: “Dogpatch.”6
By any scale, the operation there was mammoth. Plans called for installing a pair of 500-tank Calutron racetracks end-to-end in twin two-story buildings, each measuring more than four football fields long. The racetracks were on the second story; pumps and plumbing for the vacuum system occupied the first floor.7 A third building would hold the fifth and last Alpha Calutron. The army purchased 85,000 electronic vacuum tubes for the Calutrons—commandeering the nation’s entire annual production of one type.
Logistic and personnel requirements were in proportion. Every pair of vacuum tanks required an individual operator seated at a console, continually adjusting the current to focus the beam. An army of technicians was needed to monitor the orange uranium-oxide “feed” material for the beam and later scrape the errant green “gunk”—uranium salts dissolved in carbon tetrachloride—from the insides of each tank. An army of chemists would separate out the silvery powder containing U-235 that was left in the receivers following each weeklong run.
Initial guesses were that 2,500 people would be necessary to operate the production plant; by spring 1943 that estimate had already swelled to 13,500.8 (A year hence it would exceed 25,000.) But the shortage of technicians turned out not to be the problem that Lawrence had anticipated: unskilled workers—mostly young women, recruited from the Tennessee hills—proved at least as adept at running the Calutrons as degree-bearing physicists, and posed less of a security concern for the army.9
Meanwhile,
the output of Berkeley’s scale-model Calutron was still meager at best. By mid-April, following its first full week’s run, the machine had produced just under 4 grams of uranium metal, enriched to an average of 20 percent U-235. The “great optimism” that the boys had heard Lawrence express hardly seemed warranted. The beam had actually been too intense, eating its way through the receivers, and many vacuum tubes had burned out. “We’ve got to get the wrinkles out as soon as possible,” Ernest needlessly admonished the boys.
But the results two weeks later seemed more dismal than ever. One vacuum tank had worked for only three days, whereas shorts and leaks had prevented a second tank from operating at all. Nonetheless, Ernest assured Groves—as he had Oppie—that the goal of 1 gram of U-235 per day was now in sight. “The thing that interests me,” the general replied gruffly, “is where do we go from here.”10
Groves had already decided that the fifth racetrack, still under construction, would be modified to become the first Alpha II Calutron. After Oppenheimer reported from Los Alamos that the gun-type bomb might require three times the amount of uranium originally estimated, Lawrence launched a campaign to double the number of Calutrons. Consideration was given to tearing down all the Alpha I racetracks and converting them to the new, four-beam Alpha II design.
Groves also authorized the building of two additional Beta units to process the expected additional output from the Alpha IIs—even though, as Lawrence pointed out, the design of the Betas remained unproven; not even a prototype had yet been tested.11 The first full-scale Calutron was scheduled to go into operation that fall. On November 4, 1943, Lawrence reminded the Coordinating Committee that the “zero hour is approaching.”12
