In response to:

Were the Rosenbergs Framed? from the July 21, 1983 issue

To the Editors:

The updated, expanded edition of our book, Invitation to an Inquest, reviewed by Ronald Radosh and Joyce Milton [NYR, July 21], contains new chapters based on some 200,000 pages of FBI and other government documents relating to the Rosenberg-Sobell case. These documents were obtained under the Freedom of Information Act by means of an eight-year-long legal battle waged—against the determined opposition of the FBI—by attorney Marshall Perlin on behalf of the Rosenbergs’ sons. The bureau still is withholding tens of thousands of files and a substantial proportion of those released are heavily censored. Clearly, in working with such rich yet problematic source material, questions of accuracy and authenticity are bound to arise and some honest differences of interpretation between researchers would hardly be surprising. What separates us from Radosh and Milton, unfortunately, are not such honest differences. Their attack on us—both in their review and book—frustrates rational discourse because their arguments rely on distortions, omissions, and, all too often, outright falsifications.

It is not true, as Radosh and Milton claim, that we have failed to undertake “a full-scale review” of the FOIA files concerning Elizabeth Bentley, the prison informer Jerome Tartakow, or the alleged spy ring. What is true is that some material on those subjects that they have included in their own book, material crucial to their conclusion that the Rosenbergs and Sobell were guilty, does not appear in our new chapters. The reason, to put it bluntly, is that this material also does not appear in any of the FOIA files: It has been invented by Radosh and Milton. Detecting these derelictions is not easy because many of their statements are not supported by any source citation, some citations are erroneous, and their identification of FBI files is so inadequate (providing neither file nor serial numbers) that it can take hours or even days to locate a single item—even when one is as familiar with the archive as we are.

A further problem in evaluating their accuracy arises from their use of interviews (Radosh has indicated that he will not make transcripts available to other historians). Our decision to conduct no interviews for the new chapters in Invitation to an Inquest, because of our reluctance to rely on unsupported recollections of events that happened over thirty years ago, is faulted by Radosh and Milton in their review. They quote derisively from the statement in our book that we chose “to listen instead to the voices that speak from the files in words unaltered by the passage of time.” Their book, on the other hand, is replete with interviews, some of which—we have discovered—were altered considerably, even without the passage of time. For example, on pages 56-57 of their book are three seemingly decisive interviews with ex-Communist Party functionaries Max Gordon, Junius Scales, and John Gates. The latter two are quoted to the effect that they knew of the Rosenbergs’ involvement in espionage long before their arrest. According to Radosh and Milton, these interviews “provide circumstantial corroboration of the Green-glasses’ version of events.” But when we contacted Gordon, Scales, and Gates, each in turn unequivocally denied the statements attributed to him.

An added impediment to follow-up scrutiny of the Radosh-Milton sources is their frequent resort to quoting unidentified persons. In their book, they present material from a “well-known left-wing lawyer,” a “well-known and influential left-wing journalist, whose confidentiality we are respecting,” an “anonymous source who is close to [Roy] Cohn,” a “lawyer who has requested anonymity,” a “venerable and well-established Communist party lawyer,” and “a source close to the Rosenberg defense effort over the years.” They cite the latter source twice: to substantiate a close friendship between prison informer Tartakow and Communist Party General Secretary Eugene Dennis (which they say explains why Rosenberg confessed his crimes to Tartakow), and to back their astonishing claim that Dennis gave Tartakow “a letter of reference” to Rosenberg defense attorney Emanuel Bloch (pp. 293 and 314). In this instance we know that the unidentified source is an individual who was previously named by Radosh in the New Republic but who firmly repudiated in a letter to that publication the information attributed to her.1 In any event, this individual could not possibly have had any direct knowledge of the relationship or incident for which she is cited as an anonymous source, since she was less than ten years old at the time.

Exemplifying Radosh and Milton’s brand of historiography is their treatment of Elizabeth Bentley, a major prosecution witness at the Rosenberg-Sobell trial in 1951. The files show that in 1945 she gave a statement to the FBI about a group of New York City engineers, one of whom she knew as “Julius.” (“However, I do not believe this was his true name.”) In her FBI statement, Bentley said that in 1943 “Julius and the others in the group” moved to Norfolk, Virginia, “where they secured employment of some kind” and, apparently, obtained naval data for the Russians.2 Julius Rosenberg could not have been the “Julius” described in Bentley’s story. He never lived or worked in Norfolk, nor was he ever employed on any naval projects. Nor did any of the other men the FBI named in the files as suspected members of a Rosenberg spy ring ever work in Norfolk.


Radosh and Milton refuse to face these simple facts. In a chapter titled “Elizabeth Bentley: The Norfolk Connection,” they admit that “Julius Rosenberg had never worked in Norfolk.” But they insist that Bentley’s story about people who moved to Norfolk in 1943 did apply to Rosenberg and offer the following “evidence”: A Rosenberg classmate, William Perl, resided during this period in Hampton, Virginia, while employed by the National Advisory Committee for Aeronautics; in 1945, a Rosenberg acquaintance, Alfred Sarant, wrote a letter to the Navy Bureau of Ships; at some undesignated time and place, an unidentified CCNY classmate of Rosenberg’s, “never involved in espionage,” did work for the Navy on sonar; and Morton Sobell once visited a ship at Norfolk without proper authorization (pp. 230-231). Radosh and Milton do not dignify this mélange with a single source citation.

Their pièce de résistance, of course, is the last item—the only one that claims the actual presence of someone in Norfolk. They write: “…the Bureau knew that on one occasion…Sobell had gone to Norfolk and boarded a ship that housed fire-control devices and other material of a classified nature and received a briefing for which he technically did not have clearance.” Even if Morton Sobell had once boarded a Navy ship at Norfolk it would hardly establish that in 1943 Julius Rosenberg had a “Norfolk connection.” But Sobell never did. The item is false. No document describing a visit by Sobell to Norfolk exists in the FOIA files.3

In their review, Radosh and Milton find our utilization of FOIA files on Tartakow and the spy ring “highly selective,” presumably as compared with their book. But on these subjects also their book contains invented data and omits relevant information. Regarding the prison informer to whom Rosenberg was supposedly confessing the innermost secrets of his spy ring, Radosh and Milton write: “One of the things the Bureau wanted most from Jerry Tartakow was a lead that might help them to build an espionage case against William Perl.” Eventually, “Tartakow came through by telling his interviewers of a specific incident.” He claimed that on a July 4th weekend Perl “removed some secret files” from a Columbia University physics laboratory and took them to Rosenberg’s apartment to photograph. The bureau found, “to their amazement,” that Tartakow’s story “fit the facts.” Agents discovered that at Columbia’s Pupin Lab during the appropriate period “Perl had checked out and signed for a huge amount of classified material” and “had also checked out substantial numbers of reports in May and June of that same year (1948)” (pp. 297-300).

This is not proof of espionage, but it is as close as Radosh and Milton get anywhere in their book. However, their statement that the FBI discovered that Perl “checked out and signed for” classified material is false. No FOIA document that we have seen makes any such claim, and the source cited in the Radosh-Milton book (p. 537) could not be located despite an extensive search. In any event, a Justice Department report on the subject notes that “there was no system for checking out material,” and observes: “Investigation at Columbia University failed to disclose evidence that Perl removed any classified material at Columbia University.”4

The only verifiable parts of the Columbia story were Perl’s presence at the school and the existence of classified material there. Radosh and Milton believe that Tartakow could only have learned these facts from Julius Rosenberg. However, a more likely source was the FBI, which already had the requisite information. (A year before Tartakow told his story, FBI agent John P. Buscher had made a detailed review of all material forwarded to Columbia’s Pupin Lab during the period Perl worked there.)5 Radosh and Milton object that “it makes no sense” to suggest that the FBI fed data to Tartakow (p. 534). But one value of Tartakow’s statements to the FBI became manifest in 1953 when Hoover, to forestall the possibility of presidential clemency, dispatched a top-secret memorandum to the Attorney General citing the informer’s charges against Perl and others as extra-judicial proof of the Rosenbergs’ guilt.6

Radosh and Milton’s contention that Perl and others were part of a larger spy ring was first voiced by the FBI over thirty years ago. The problem today remains the same as it was then: suspicion without evidence. In addition to the defendants, the FBI regarded more than a dozen men and women as spy ring suspects. Nearly all stayed in the United States, where they underwent years of investigation, job loss, and other hardships—but never were charged with espionage. Two became expatriates whose whereabouts were unknown. Joel Barr, an engineer fired for his left-wing politics, had been living in Western Europe since early 1948. Alfred Sarant, also a leftist, was an engineer working as an Ithaca, New York, painting contractor. In 1950, after Rosenberg’s arrest, the FBI searched Sarant’s home for a week (with his permission) and he was interrogated continually, surveilled, and accused of espionage, which he denied. Three weeks later he and his lover, Carol Dayton, eluded the bureau and went to Mexico. Each abandoned a spouse and young children. A few years ago the former Carol Dayton contacted family members in the United States for the first time and she has since corresponded and met with some of them. According to family members, she and Sarant lived in Mexico for six months during which time she decided not to return. They married, moved to Prague and later the Soviet Union, and had four children. Sarant had a successful career as an engineer-scientist and died of a heart attack in 1979. Less is known about the wanderings of Barr, who has also been out of touch with relatives, but he is reportedly living and working in the Soviet Union. How they chose to live their lives is a story with many human mysteries but, as far as we have any basis for knowing, it is not a spy story. If the FBI has any evidence to the contrary, it should be released. But this seems unlikely. In March 1951 Assistant Attorney General James McInerney inquired of US Attorney Saypol “what action should be taken by the FBI in the event Alfred Sarant re-enters the United States” and Saypol responded: “…there is insufficient evidence at the present time to warrant filing a complaint against Sarant in this district on any possible federal charge.”7 That statement is not mentioned in the Radosh-Milton book.


Harry Gold’s alleged June 3, 1945, Albuquerque espionage meeting with the Greenglasses, including the Rosenbergs’ part in the arrangements, formed the core of the prosecution’s atom-spy case. Radosh and Milton maintain that Gold told the truth about this episode. We do not envy them their need to defend Gold’s credibility. For the files show that Gold lied repeatedly to his FBI interrogators, as when he gave them vivid descriptions of three espionage confederates and then said all three were “nonexistent.”8 As for voluminous accounts FBI agents found in Gold’s home—wonderfully detailed stories about espionage recruits—he admitted they were “absolutely fictitious.”9

On May 22, 1950, Gold confessed that he was the man who had met with Klaus Fuchs in Santa Fe and elsewhere. He said he had visited Albuquerque briefly in June and September 1945, but mentioned nothing about an Albuquerque espionage encounter. Not until ten days later, on June 1, 1950, did Gold begin telling about a meeting with an unknown GI in Albuquerque; he had “completely forgotten about” it, he explained. Thus, the files confirm precisely the timing we outlined in the original edition of Invitation to an Inquest (p. 368). (We never wrote that Gold first told the Albuquerque story on June 14, as the Radosh-Milton review infers, and then refutes.)

In our new edition we made use of a unique primary source, one that has the virtues of immediacy and directness: handwritten logs kept by two FBI agents while interrogating Gold.10 Radosh and Milton inexplicably make no reference to the logs in their book, though they refer extensively to them in their review. We discovered that the logs do not contain a single one of those details of Gold’s trial testimony which, according to the prosecutor, “forged the necessary link in the chain that points indisputably to the guilt of the Rosenbergs”: no cut Jell-O box; no password “Julius”; no brother-in-law “Julius”; no “woman who was supposed to go.” Radosh and Milton label these a “few lapses of memory,” and state: “The Schneirs apparently believe that a witness’s inability to remember certain details of an incident during his first interrogation renders his entire testimony invalid.” But the logs also cover the questioning of Gold over the entire two-week period prior to the first interrogation of Greenglass during which time, according to a June 9 FBI memo, Gold was “milked dry” about the Albuquerque episode.11 In fact, some of Gold’s “details” that sent two people to their deaths and a third to Alcatraz on a thirty-year sentence were not recalled by him until shortly before the trial nearly nine months later.

Radosh and Milton believe in Gold’s veracity because “he gave the FBI sufficient information to lead them eventually to David Greenglass.” Actually, less than twenty-four hours elapsed between Gold’s mention of an Albuquerque meeting with an unknown GI and the bureau’s decision that David Greenglass was the man.12 Radosh and Milton explain the bureau’s amazing speed by claiming that Greenglass’s name was “on a list” of people previously questioned about thefts of souvenir samples of uranium. But they cite no source and we doubt that there was such a list. The reason the FBI did not need Gold to “lead them” to Greenglass is revealed in our new edition: David Greenglass was a suspect in a Los Alamos espionage investigation at least five weeks before Gold “recollected” meeting with him.

Radosh and Milton say in their book that the “agents who interviewed Harry Gold in Philadelphia on June 2, 1950, had never heard of David Greenglass” (p. 46), but the logs prove that they are wrong. A June 2 log discloses that on that date, to assist Gold in “remembering” the name David Greenglass, the agents presented him with a playful list of twenty names that included, along with Greenglass, a famous baseball player, the name of a restaurant, monikers such as Uriah and Isiah, and ten names arranged in crypto-alphabetical order (Arnold Blume, Charles Drusher, Ezra Finkelstein, etc.). The Radosh-Milton review complains that we do not mention that Gold’s FBI interrogators “thought the unknown GI was named either Hall or Sax.” However, neither the name Hall, nor the name Sax, nor the name of any FBI spy suspect other than Greenglass was on that list.

Harry Gold’s entire testimony about a hotel stay in Albuquerque was that “on Sunday morning” he had registered at the Hilton in his own name and that same afternoon departed for New York by train. After Gold left the stand the prosecutor, with defense permission, introduced as a government exhibit a photostatic copy of Gold’s registration card without calling any hotel employee as a witness. In our new edition we report our discovery that “other than in his trial testimony, Harry Gold never stated that he had registered at the Hilton Hotel in Albuquerque on June 3, 1945” (p. 450). We remark particularly that no such statement is recorded in the agents’ handwritten logs. Radosh and Milton tax us with having used a partial quotation followed by four dots from a log entry dated eleven days after the June card came into the FBI’s possession. They comment coyly:

Wondering what followed the four dots at the end of the [Schneirs’] quote, we checked our copy of the same log and found that agent Miller actually wrote the following: “Has no memory of regis’tring at Hilton, on June trip [this is where we broke the quotation with four dots], but sd he easily could have done so—possible that reg’d.”

Checking our copy of the same log, we find that Radosh and Milton also cut the sentence—but without using four dots. The final portion reads: “…and gave up possibly about Mid.—didn’t pay anything.” The totality of the sentence does not change the basic idea conveyed: Gold did not remember registering at the Hilton in June. Moreover, we quote on the same page of our book a later, more coherent statement by Gold regarding a possible unsuccessful midnight “registration” (p. 449).

Reproductions of the front and back of the June 3, 1945, hotel card photostat and of another Gold Albuquerque Hilton card dated September 19, 1945, appear in our book. We examined both cards and, based on everything we learned, concluded that the June card is a forgery. Radosh and Milton devote a chapter of their book (with no photos of the cards) and the longest section of their review to a scattershot attack on our conclusion, but their argument is marred by serious misstatements of fact. For example, in 1961 we asked document examiner Elizabeth McCarthy to study the hotel cards. She reported that a Hilton Hotel clerk, Anna Kindernecht, probably had written on and initialed the lower half of the September card, but probably had not written nor initialed the June card. Now, Radosh and Milton flourish the report of another handwriting analyst, Hanna Sulner, done in 1966 for Morton Sobell’s attorneys and never used in court. They accuse us of “hiding this contradictory statement.” But when we obtained a copy of Sulner’s report from the Wisconsin State Historical Society we learned that her opinion—that “the lower portion of the two cards were not written by the same person”—is in accord with McCarthy’s key finding. The Radosh-Milton review further claims that Sulner stated that “handwriting differences on the two cards had to be disregarded because of possible stains on portions of the paper that would not be visible on photostats.” This is a blatant distortion. What Sulner actually wrote was that “a defect in the printing”—not the handwriting—on the September card—not the two cards—might have been caused by stains not visible on photostats. Radosh and Milton parade as gospel Sulner’s comment that she could not compare the 1945 cards with samples we had obtained of Anna Kindernecht’s writing in 1961 because too much time had intervened. They failed to notice that we later made available to McCarthy a 1952 sample. Moreover, they failed to do any research in journals or texts in the field of document examination or they would have found that, while some handwritings do change with age, others remain amazingly similar, even after fifty years or more.13

Sulner and McCarthy both complained of limitations in working with photostats. In 1966 attorneys for Sobell requested the originals of both cards and were informed by the Department of Justice that they had been destroyed. The FBI said that it had returned the June 1945 card to the hotel (where it would have been promptly destroyed) only four months after the trial, before argument of appeals. The original of the September card, which was not a trial exhibit, was retained by the bureau for ten years and then destroyed.14 Radosh and Milton conceal from their readers the bureau’s hasty disposal of the original June card by reporting falsely that “some years after the trial” the originals of both cards were returned to the Hilton and “routinely destroyed” (pp. 461-462). Neither do they discuss many other questions, including why there is no customary FBI identification on the June card (the September card is initialed by two FBI agents and dated by them).

They do comment at length about one puzzling circumstance we consider in our book: the machine date-time stamp error on the reverse side of the June 3 card, which says June 4, 1945, 12:36 P.M., and thus does not conform with the handwritten date on the face of the card or with Harry Gold’s trial testimony. They claim Albuquerque FBI agents immediately noticed the date-stamp error; engaged in “feverish activity” to solve the mystery; interviewed twice the Hilton manager (who “explained” that a mechanical error was at fault); and confirmed this explanation with the assistant manager. Their version of events simply is not supported by the files. The hotel manager never “explained” the cause of the dating error, he offered an “opinion” about it; a report by agent McConnell cited in their review is incorrectly dated and badly misinterpreted (McConnell never “reinterviewed” the hotel manager, he only summarized an earlier interview); the assistant manager was not questioned on this point—but why go on?15

Rather than argue this minutiae endlessly, we prefer to raise a broader question. We are convinced, after careful study, that at least a few FBI documents in the Rosenberg-Sobell files were prepared solely to create a paper history and in some instances were produced or amended by the bureau at a time subsequent to their stated date. Radosh and Milton, on the other hand, ridicule us as “authors who unhesitatingly brand FBI documents as forgeries whenever it suits their purpose.” Many FBI reports assert that Harry Gold first mentioned a June 1945 stay at the Hilton on June 1, 1950. They offer as their source a teletype sent from Philadelphia to Washington, DC on that date, stating that “Gold said that…in June, 1945, he stopped at Hilton Hotel, Albuquerque.”16 The problem here is that there exists overwhelming evidence in the FBI logs and elsewhere that on June 1, 1950, Gold said no such thing. This teletype, unmentioned by Radosh and Milton in their review or book, poses a fundamental challenge to their apparent assumption that every single document in the FBI files is authentic. In our book, we offer the explanation that this fake June 1 teletype is part of the paper history of the forged Hilton Hotel card. How do Radosh and Milton explain this document?

We will have more to say on these subjects, and on others we have not had space to touch on here, on October 20 when a debate between ourselves and Radosh and Milton will be jointly sponsored by the Nation and New Republic magazines at Town Hall.

Pleasantville, New York

Ronald Radosh and Joyce Milton reply:

Our review of Invitation to an Inquest criticized the book’s methods, its use of evidence, and its analysis of issues. Rather than respond to our major points, Walter and Miriam Schneir have sought to use their rebuttal space to launch an attack on our recently published book, The Rosenberg File. Not content to take up our serious differences over interpretation of the evidence, they rush to charge us with falsification of evidence and literary dishonesty. Ironically this only confirms our original contention that their writing is a brief for the Rosenberg defense—not objective analysis—and that they fling about baseless charges of forgery whenever the facts do not happen to fit their preconceived theories. This will be made clear by our comments on each of their charges.

The Schneirs write that we have “invented” material that does not appear in the FBI files. This is scurrilous and totally unfounded. Our book is extensively footnoted and draws on many thousands of pages of FBI files as well as on numerous other sources. The fact is the Schneirs have already argued with us on some of these issues and lost—see the exchange on the credibility of the informer Jerome Tartakow in The New Republic, August 4-11, 1979. Now, rather than take up the substance of our discussion once again, they simply assert that we invented the material. Any FBI document we cite can be retrieved by referring to the particular file from which it comes: e.g., Rosenberg HQ File, D.M. Ladd to Hoover, June 2, 1950, etc. And in fact, we used serial numbers when we felt that they were necessary to avoid confusion.

As it happens, FBI serial numbers are useful to pinpoint the location of documents in a huge file—but they are not essential. They are most useful in this case if one has access to the repository of Rosenberg FBI files held in New York City by the FOIA, Inc., the group that won the files from the government and whose board of directors includes the Meeropols and their counsel, Marshall Perlin. The Schneirs undoubtedly have this resource available to them. FOIA, Inc., has catalogued and cross-indexed the material, making it easy for researchers to use. But we did not have access to this collection. Although FOIA, Inc., purportedly holds these files for the benefit of all working scholars and journalists, Ronald Radosh and Sol Stern were summarily refused permission to continue working with the collection back in 1978, when it became apparent that they were not prepared to hew to the line of the Rosenbergs’ total innocence. This required us to research the rest of the book at the FBI building in Washington, DC, and to obtain our own Xerox copies of the material, at considerable expense. If the Schneirs and others who use FOIA, Inc., now find it inconvenient that we do not use serial numbers throughout, they should not blame us.

The Schneirs go on to charge us with fabricating interviews. This is an old technique—getting those who have contributed interviews, but feel embarrassed by what they said now that it is in print, to repudiate their statements. Indeed, James Weinstein has told us that the Schneirs called him as well, trying to get him to say that we had distorted his interview. They did not receive satisfaction in his case, but apparently got disclaiming statements from a few they spoke to. Let us be clear. We interviewed Max Gordon, John Gates, and Junius Scales on the dates indicated in our book. We spoke to John Gates on three occasions. Our quotes are accurate and in context, and we are fully prepared to defend their accuracy if need be.

However, the Schneirs’ summary of these quotations is distorted. The Schneirs write that Scales and Gates “are quoted to the effect that they knew of the Rosenbergs’ involvement in espionage long before their arrest.” If this is what Scales and Gates were asked to confirm, no wonder they balked. Both men stressed to us that they had no personal knowledge of this, and we never claimed they did. What Gates said was that he assumed, when the Rosenbergs became inactive in the Communist Party, that “there was only one explanation”—espionage. After the Rosenbergs’ arrest, Gates recalled being told by the Party cultural official V.J. Jerome that the Rosenbergs were “heroes” for going to their deaths without saying a word (see page 329 of our book).

Scales told us that he knew that the Rosenbergs’ Daily Worker subscription was canceled by order from the Party leadership and that hence he was not surprised when they were arrested. Max Gordon gave us background on how communists in general who had been recruited for spying were handled by the American CP. Again, Gordon stressed that he had no specific knowledge of the Rosenbergs’ activities. Indeed, the point we make in our book is that Party leaders were generally aware of the existence of these illegal activities, and sometimes condoned them at a distance, but looked the other way when it came to the specifics.

The Schneirs try to make it appear that our book is based on anonymous sources. Any reader of the book can easily see otherwise. They cite a few instances in which such unnamed sources are used—two of them from Ronald Radosh’s foreword, who are cited for the express purpose of illustrating the cast of mind and the double standards employed by some of the Rosenbergs’ more prominent defenders. These quotations were not used as the basis for the analysis and conclusions in the body of our book.

In a third instance—referring to one of the very few anonymous attributions in the body of The Rosenberg File—the Schneirs are clearly playing a game with us. We report that an unnamed source told us that Eugene Dennis had given the FBI informer Jerome Tartakow a recommendation which helped Tartakow to obtain employment with Emmanuel Bloch. The Schneirs excoriate us for not naming this source and then go on to claim that he or she could not possibly have had direct knowledge of any recommendation by Dennis. The Schneirs know very well that this same source previously sued us for libel—in a baseless but costly harassment suit which was eventually thrown out of court on First Amendment grounds. In fact, this source not only talked to Ronald Radosh and Sol Stern but gave them copies of two letters from the source’s law office—not FOIA—files. One of these letters, from Tartakow to Emmanuel Bloch, is reprinted in its entirety on pages 541-543 of our book. Moreover, the authenticity of this letter, which sheds considerable light on the motivations of Tartakow and his relations with others connected with the case, was confirmed for us by Tartakow himself.

On Elizabeth Bentley: the evolution of the Schneirs’ treatment of Bentley illustrates their partisan approach. In the original edition of their book they tend to dismiss Bentley’s testimony as phony and to cast doubt on whether she ever mentioned a “Julius” in her original 1945 statements to the FBI. The answer, they wrote then, “still remains hidden in the FBI files.” If it were up to the Schneirs, the answer would be hidden still. They did not choose to reprint Bentley’s dramatic 1945 statement in the latest edition of their book even though it has now been available for years.

Only after our book, which uses the statement, became available to them did the Schneirs rush to blunt the impact of our revelations by giving their own analysis of the statement in the June 25, 1983, issue of The Nation. Since it is no longer possible to claim that Bentley never told the FBI about a “Julius,” the Schneirs now resort to picking out inaccuracies in her description of him. But these inaccuracies are not disputed by us. We make clear in our book that Bentley’s information was secondhand, that it was not definitive or hard legal evidence, and that Julius Rosenberg himself never worked in Norfolk, Virginia.

But there was a Norfolk connection. If the Schneirs have studied the Rosenberg case documents they must know well the name of the CCNY classmate of Rosenberg’s who worked at Norfolk on classified research having to do with the development of sonar. We chose not to name this person. Furthermore, Morton Sobell, Max Elitcher, and William Danziger, all associates of Rosenberg’s, worked for the Navy Bureau of Ordnance—which had its offices in Washington, DC, but which carried out projects in Norfolk. The Schneirs are correct on one point—it was Max Elitcher, not Sobell, who boarded a Navy ship in Norfolk. However, Sobell worked for the firm in New York which was carrying out the classified research for the same project. The point we make in our book is that, to the disappointment of the prosecutors in the Rosenberg case, Bentley insisted on sticking to her 1945 story, errors and all. We write: “But much of what Bentley had said—that ‘Julius’ was the contact for a group of engineers and that he was connected with the theft of Navy secrets—was secondhand information and hence inadmissible.” The record of Bentley’s refusal to flesh out the story or to claim that she had this information firsthand convinced us that the testimony she gave during the trial was truthful.

On William Perl’s checking out classified documents from the Pupin Lab at Columbia University: the Schneirs assert that “no FOIA document that we have seen makes any such claim.” In fact, the report by Agent O’Donoghue of July 2, 1951, cited in our book, not only says that Perl signed a receipt on July 3, 1948, for checked-out materials but actually lists these materials by serial number and gives a description of their contents. (We quote from three such descriptions in our book.) The report also confirms that Perl had the combination to Von Karman’s safe and the key to his file cabinets, where classified reports were stored.

While accusing us of “suspicion without evidence,” the Schneirs seem curiously lacking in suspicion and indifferent to the evidence. Here and in their book they never bother to advance an explanation why Vivian Glassman approached the same William Perl after Rosenberg’s arrest with $2,000 and instructions for fleeing to Mexico. Yet this story came from Perl himself. Rather than try to explain why this episode happened, if there was no espionage going on, the Schneirs keep insisting that Perl and others were never indicted for espionage—as if the failure of the government to indict them proves that they were wholly innocent. On the contrary, FBI files make clear that the reason why arrests and indictments were not forthcoming was that the government believed it needed the testimony of Julius Rosenberg to make its case—hence the harsh treatment of Rosenberg in an effort to pressure him into confessing.

Astonishingly, the Schneirs have now chosen this moment to inform us, in passing, that Joel Barr and Alfred Sarant—both of whom were sought by the government in connection with the Rosenberg espionage conspiracy—actually defected to the Soviet Union. This is what the Rosenberg prosecutors charged all along, but the Schneirs said in the original edition of their book only that Barr and Sarant were “presumably living abroad”; and they mentioned a 1957 Justice Department report that both men were “behind the Iron Curtain” only to make derisory and skeptical comments about it. Even now, they do not bother to inform the readers of The New York Review that the two men lived in the Soviet Union under false identities—Sarant as Philip Staris and Barr as Joseph Berg—and that Sarant’s “successful career as an engineer-scientist” happened to be in the field of microelectronics research for the Soviet defense establishment. That the Schneirs find this story innocent of any implication about possible involvement in espionage, and worth pondering only for its “many human mysteries,” speaks for itself.

We also wonder why, if they knew all this, they did not tell the story in the addendum to their latest edition. The new edition says of Sarant and his lover Carol Dayton only that “whether they made new lives in Mexico or elsewhere, or were victims of accident or foul play, would be interesting to know.” Of Barr they write vaguely (but presciently, it seems) that perhaps he “moved to Eastern Europe or the Soviet Union, where his technical skills would have made him a welcome immigrant.” They then suggest that Barr may also have settled in Sweden. Could it be that the Schneirs have decided to tell the facts now only because they heard that we recently had obtained the same information from other sources and released it to the press?

In any case, when an American who has worked in military research makes available his technical skills to the Soviet Union—and does so while living under a false identity—doesn’t this sound as if he may have some connection with espionage? And why do the Schneirs continue to obfuscate the fact that both men’s disappearances coincided with the arrests of David Greenglass and Julius Rosenberg? (Barr disappeared on June 15, 1950.) They now write, correctly, that Sarant crossed into Mexico approximately three weeks after Rosenberg’s arrest. In the new material of their book, however, they assert that Sarant was “accessible to the Bureau for many months.” Which months?

The Schneirs continue to argue now that the FBI files “confirm precisely the timing” outlined in their original theory of how the FBI set up Harry Gold to frame Greenglass. We can only invite readers to examine this outline, which is repeated on pages 397-399 of the latest edition of their book. The Schneirs write: “Pursuing a mass of newly obtained data lent us by Ballard [Gold’s attorney] we found that Gold had not written anything about his mission to Albuquerque prior to the time that Greenglass signed his initial FBI statement at 1 A.M. June 16.” (It is strongly implied that Gold had not said anything either, prior to his June 14 meeting with counsel.) The entire point of the ensuing argument is that Gold did not tell his attorneys anything he could not have learned during his two weeks in FBI custody. Yet, as the Schneirs reluctantly acknowledge now, Gold gave a statement descriptive of Greenglass as early as June 1, 1950. The two-week interval has disappeared. What Gold may have omitted from his account of Greenglass on this date is irrelevant to this particular argument—the fact is he gave sufficient information that could have applied to Greenglass and Greenglass alone.

On the matter of the June 17, 1950, log, it is the Schneirs and not us who are being coy. The phrase they omitted from their quotation clearly supports our argument that Gold—while he may not have remembered staying at the Albuquerque Hilton in June 1945—certainly never denied he had done so. The following phrases—“and gave up possibly about midnight—didn’t pay anything”—are ambiguous but apparently refer to Gold’s story that he tried to register at the Hilton on the previous day, Saturday, but was turned away for lack of room—an incident entirely consistent with the version of events Gold gave in court and confirmed to the end of his life.

Since FBI memos, signed statements, and summary reports do not support their conspiracy thesis, the Schneirs have now retreated to the position that the agents’ handwritten logs are the only truly credible evidence. We agree that the logs project “immediacy and directness.” They are also sketchy and incomplete. There is no reason whatever to infer that because a particular phrase does not appear verbatim in the logs its appearance in a memo or a signed statement of the same date is a forgery. As for the “details” that the Schneirs cite as missing—the specific identification of the Jell-O box as the recognition signal and the use of the name “Julius” as a password—these are not in dispute. We note in our book that Gold originally described the Jell-O box as simply two matching pieces of cut or torn cardboard. We also treat at length the belated evolution of Gold’s memory of using the name “Julius.” This particular aspect of Gold’s testimony is indeed suspect. But even if Gold lied when he claimed to recall using “Julius” as a password, this would not justify the conclusion that he never met David and Ruth Greenglass, that he was never in Albuquerque in June 1945, and that he was never a Soviet agent.

As to how David Greenglass’s name first came up as a candidate to be Gold’s unnamed GI, the Schneirs themselves acknowledge that Greenglass had been investigated the previous January on suspicion of having stolen uranium souvenirs from Los Alamos. Why do they tax us now for saying the same? The Schneirs further assert in the new material of their latest edition that Greenglass was also “named in connection with the FBI’s giant atomic espionage probe by the early spring of 1950.” If they have information to support this, why don’t they share it with us and their readers? It would be interesting to know what basis the FBI might have had for connecting Greenglass to atomic espionage at that date. Until the Schneirs convince us otherwise, however, we will continue to suspect that documents naming Greenglass and “hundreds of other former Los Alamos employees” really refer to the uranium investigation mentioned above. And whatever Greenglass was suspected of, why did the FBI quickly single him out for connection to Gold—particularly if there were “hundreds” of other suspects, as the Schneirs say? One would think that they would be more comfortable with our version: that there were only two prime suspects, namely Hall and Sachs (not Sax). We did not make these men up. Why should we? Whether there were two other men or hundreds makes no difference. David Greenglass was the one man who fit Gold’s original description.

Let us now turn to Hannah Sulner’s analysis of the photostats of Gold’s June and September Hilton Hotel registration cards. The Schneirs report Mrs. Sulner’s opinion that “the handwriting on the lower portion of the two cards were not written by the same person.” This observation, however, is not evidence that a forgery took place. As we noted in our own book, Judge Edward Weinfeld took note of these same differences and disposed of that point in his decision on the Sobell appeal. Weinfeld wrote:

The circumstances that at a public and busy hotel the same initials appear on the two cards…[while] the data as to rate, stay, departure and room number are in different handwriting, does not, in one fell swoop, permit the inference that it was “forged, that the government knew it was forged…[and] that Gold did not register at the Hilton on June 3.”

The Schneirs correctly identify one error in our passage on the Sulner letter. In the process of retyping various drafts of our review a reference to a “defect” in printing—correctly described on page 463 of our book—was transformed into “handwriting differences.” In any event, the Schneirs have nothing to say about the substance of Sulner’s conclusion that the “defect” or difference “does not indicate to me forgery” but may actually be the result of a smudge or a grease mark. Further, whether the Schneirs eventually showed Elizabeth McCarthy a 1952 sample of Anna Kindernecht’s handwriting has no bearing on Mrs. Sulner’s comment that she (Sulner) found no basis for comparing the 1945 and 1961 samples that she saw.

Nor do we “parade as gospel” Mrs. Sulner’s conclusions. On the contrary, we cited her letter to demonstrate that two highly respected experts were able to examine essentially the same material and come to opposite conclusions. While handwriting analysis may have its forensic uses, it is by no means an exact science, and one expert’s opinion about a very few scrawled notations cannot support the sweeping conclusions the Schneirs draw. No amount of “research in journals and texts in the field of document examination” is going to convince us otherwise, and as we noted, even Mrs. McCarthy did not contend to us that her analysis amounted to definitive proof of forgery.

With regard to the circumstances of the original registration cards’ destruction, the point is not precisely when the June 1945 card was returned to the Albuquerque Hilton but why it would have been returned at all if it was a forgery. Wouldn’t the Hilton employees, who worked with such cards every day, be the first to spot a fake? Why not just destroy the card? If the June registration card was returned more promptly, as the Schneirs say, this may well have been because Emanuel Bloch had stipulated at the trial that the photostat of this card was authentic. He therefore did not have the option of changing his mind and making the card a subject of appeal. Short of powerful evidence of government chicanery, no appeals judge would have been likely to let Bloch raise questions about the card after the fact.

The Schneirs simply cannot accept the machine date-time stamp error as an explanation for the difference in dates. They criticize us for claiming that Albuquerque FBI agents “immediately noticed the date-stamp error,” as if the reports of the agents prove they did not. But they did. The Schneirs are wrong. The material cited in our review and in our book establishes this firmly. Having failed to repudiate our evidence, the Schneirs resort to embarrassing plays with semantics. Thus they write: “The hotel manager never ‘explained’ the cause of the dating error, he offered an ‘opinion’ about it.” Readers of our book will find on page 460 the exact words of the Bureau’s interview with Brumit:

[Brumit] explained that all registration cards received on June 3 were stamped “June 4.” He could only offer the opinion that through mechanical error the time and date stamp machine had erroneously set the date as June 4, rather than the correct date of June 3.

The Schneirs prefer to accept Brumit’s lack of memory when they interviewed him for their 1966 book—when he told them he “had no recollection” of having initialed the card—although the FBI files show that he had done so—and his apparent belief that the discrepancy was due to clerical error (p. 382). The files show that the Schneirs’ conclusions are faulty. They seek to ignore the decisive pieces of evidence and, by chicanery and guile, hope to rescue a now discredited and ludicrous conspiracy theory.

They quibble instead over the date of Agent McConnell’s report. Having looked again at the poor quality of the FBI Xerox, we defy them to show whether the date is June 10 or June 16. It is, of course, irrelevant—as is their unsubstantiated claim that the agent is simply summarizing an older interview with Brumit. What the report says, in no uncertain words, is that Brumit told them “that an examination of other registration cards for guests registering at that hotel on June 3, 1945, showed that the date stamp on these cards carried the date of June 4, 1945, as in the case of this particular card for GOLD.”

The Schneirs’ response to our review offers one final thought: the cause they are trying to sustain is disintegrating; and the desperate efforts of these writers are only having the unintended effect of making it disintegrate more quickly.

This Issue

September 29, 1983