Oppenheimer's Elternton testimony and identification of Chevalier

Oppenheimer's Elternton testimony and identification of Chevalier describes how a brief, earlier conversation about sharing scientific information with the Soviet Union became a security case that Oppenheimer could neither clarify nor escape. In 1943, he disclosed George C. Eltenton's suspected effort to obtain project information, yet withheld the name of the friend who had conveyed the proposal. His insistence that he could distinguish the dangerous figure from innocent associates collided with the Army's demand for names. The resulting interviews were recorded, preserved in his security file, and later made his shifting account a lasting source of suspicion.

From the kitchen conversation to a security report

The episode began months earlier, in the winter of 1942–43, during a farewell dinner at the Oppenheimers' Berkeley home. In the kitchen, Haakon Chevalier told Oppenheimer that Eltenton had proposed a way to pass scientific information to a diplomat connected with the Soviet consulate in San Francisco. The details of their conversation were never fixed in contemporaneous notes. Chevalier later insisted that he had simply told his friend about Eltenton's proposal, not acted as its channel. Oppenheimer at the time treated it as a dead end and did not report it, in part because doing so would implicate a close friend whom he considered, at worst, an overenthusiastic idealist.

His understanding changed after John Lansdale spoke with him about the dismissal of Giovanni Rossi Lomanitz from the Radiation Laboratory. Oppenheimer suspected that union organizing, rather than the unspecified indiscretions behind the dismissal, might be at issue. Remembering that Eltenton had also been active in the union, he now reconsidered the kitchen conversation. What he had dismissed as a ridiculous proposal seemed serious enough to report.

General Leslie Groves later told the FBI that Oppenheimer first gave him Eltenton's name in early or mid-August. On August 25, 1943, while visiting Berkeley on project business, Oppenheimer went to see Lieutenant Lyall Johnson, the Army security officer for the Radiation Laboratory. He said that Eltenton, an employee of Shell Development Corporation who was active in the union, ought to be watched and might be trying to obtain information about the laboratory's work. Oppenheimer left without giving much more detail. Johnson called Colonel Boris Pash, who ordered that Oppenheimer return the next day. During the night, a microphone was installed in the base of Johnson's telephone, with a connection to a recorder in the next room.

Pash's recorded interrogation

Oppenheimer entered Johnson's office the following day expecting to discuss Lomanitz, and was startled to find Pash there. Pash conducted the interview. When asked about possible interest in the Radiation Laboratory, Oppenheimer said that a person connected to the Soviet consul had indirectly indicated, through people associated with the project, an ability to transmit information without attracting a leak or scandal. He said he had no firsthand knowledge, but described an effort to obtain information that he regarded as improper.

Oppenheimer's account distinguished between helping an ally through an authorized government decision and secretly passing information. He said he could see arguments for the commander in chief informing the Russians that the United States was working on the problem; he did not approve of information being moved “out the back door.” He described the people involved as sympathetic to helping the Soviet ally but bewildered by an unofficial channel, rather than cooperative with espionage. He identified Eltenton as someone who should be watched, but did not identify the intermediary who had brought the proposal to him.

Pash pressed him for the names of the people allegedly approached. Oppenheimer said that two or three cases had come to his attention, including two men at Los Alamos and a Radiation Laboratory employee who had left or was due to transfer to the Manhattan Project's Oak Ridge facility. He said that a third party, a Berkeley faculty member, had conveyed the approaches, but refused to name that person. He considered the intermediary innocent and believed it would be wrong to implicate people he thought had done nothing wrong. He explained the proposed route as he understood it: Eltenton would arrange contact with someone connected to the Soviet embassy, described as experienced in microfilm work. Oppenheimer acknowledged that the transfer would be treasonable, even as he placed the proposal in the wartime context of concern that official communication with the Soviet ally was inadequate.

The interview exposed the clash at the heart of his position. Oppenheimer told Pash that Communist association was incompatible with confidential war work, while also maintaining that many intelligent and thoughtful people had found value in the Communist movement. He urged attention to possible risks in the union and said that Los Alamos was “100 percent in order.” He even said he would be willing to be shot if he had done anything wrong. Yet he treated his own assessment of who was dangerous and who was innocent as a reason not to provide the names Pash wanted. The recorded exchange preserved that mixture of candor, confidence, and refusal.

Lansdale's follow-up and Groves's order

The recorded account alarmed the security officials. On September 12, Lansdale met Oppenheimer again, having read the transcript of the interrogation, and secretly recorded their conversation as well. Lansdale said that the government had known since February that people were transmitting information about the project to the Soviet government. Oppenheimer appeared surprised and said he had not known that. When Lansdale returned to the earlier approach, Oppenheimer continued to withhold the intermediary's identity, explaining that the other people he knew of seemed innocent and that the approach appeared erratic and unsystematic. He named several people when asked about current or former Communist Party members, including Lomanitz and Joseph Weinberg at Berkeley, and Charlotte Serber at Los Alamos, but maintained his refusal to name the intermediary or the other scientists he said had been approached.

Two days later, Groves and Lansdale discussed Eltenton with Oppenheimer on a train traveling between Cheyenne and Chicago. Oppenheimer said he would identify the intermediary only if ordered. At first, Groves accepted his reluctance as a principled reluctance to tell on a friend. Lansdale and Groves believed that Oppenheimer was telling the truth, but the investigation still lacked the name it required. In early December, Groves ordered him to provide it. Oppenheimer then reluctantly named Chevalier, while continuing to insist that his friend was harmless and innocent of espionage.

Conflicting accounts and the record's afterlife

Oppenheimer's naming of Chevalier did not resolve what he had said in August. Lansdale's December letter to the FBI reported that three members of the project had said an unnamed University of California professor had approached them about espionage, and that Oppenheimer identified Chevalier when ordered to name that professor. The letter did not name the other scientists. The FBI later pressed Groves for their identities, but its files contain no reply. The recorded testimony, the later letter, and Oppenheimer's refusals therefore left a consequential gap rather than a complete account.

A separate FBI memorandum, written in March 1944, offered a different version: Chevalier had allegedly approached Frank Oppenheimer in 1941, and Frank had immediately told his brother, who confronted Chevalier. The book identifies this account as incorrect. It conflicts with the timing of the bomb project and with the accounts of both Chevalier and Eltenton, who placed the kitchen conversation between Robert Oppenheimer and Chevalier in the winter of 1942–43. The FBI could not locate the original source for the claim about Frank. Yet its inclusion in Oppenheimer's dossier gave the version a life of its own.

In 1946, Oppenheimer told FBI agents that he had “concocted a completely fabricated story” about three unidentified associates in order to protect Chevalier. That admission reopened the central uncertainty: whether he had embellished his account to divert attention from himself, learned later that Chevalier had not approached several scientists, or changed his story to protect Chevalier and others. The competing explanations remained unresolved, and the contradictions would return during the 1954 security hearing. What began as Oppenheimer's effort to report a suspected approach while shielding people he judged innocent had become evidence that could be read either as loyalty and discretion or as deception. The concealed recording made his words durable; it did not make their meaning simple.