My principal objective had been to visit troops. But when I reached London the PM virtually took possession of my movements for the first week. . . . These unexpected subjects were so important that I devoted the bulk of my time to their consideration and altered my trip accordingly.
—Secretary of War Henry L. Stimson, August 4, 1943
Henry Lewis Stimson, at seventy-five years of age in 1943, was serving his fourth president. For the second time Stimson was the U.S. secretary of war. Patrician in background and education, Stimson had been appointed U.S. attorney for the Southern District of New York by President Theodore Roosevelt. He had served as President Taft’s secretary of war. Having risen to the rank of colonel as an artillery officer in France in World War I, Stimson subsequently was appointed governor-general of the Philippines and then secretary of state by President Hoover. An internationalist and proponent of aid to Britain and France, Republican Stimson again was appointed secretary of war by Franklin D. Roosevelt, a Democrat, in June 1940. He took up his post as Britain was in the last days of a scramble to rescue its army from the beaches of Dunkirk and France was collapsing under the German Blitzkrieg.
In the new War Department Building and later at the completed Pentagon across the Potomac, Stimson and Army Chief of Staff Gen. George Marshall worked in close partnership from adjoining wood-paneled offices between which the door never was closed.1 Stimson became a strong advocate for Marshall’s strategy for taking the war in Europe to Germany through a cross-Channel attack into northwestern Europe. Both Stimson and Marshall served on the Military Policy Committee formed by FDR to advise him on the project to develop an atomic bomb.
Thursday morning, July 8, 1943, Secretary Stimson, his special assistant, Harvey Bundy, and his aides gathered at Washington National Airport in the Inter-Continental Hangar of Trans World Airlines. They were to begin the secretary’s long-delayed trip to Britain to meet with U.S. forces and commanders in the European Theater.2
Their plane was a four-engine Douglas C-54 crewed and operated by TWA on contract to the Army Air Force. The plane had been modified to provide “every imaginable convenience aboard,” and for Stimson’s trip it carried extensive survival gear for a variety of climates from tropical to subarctic.3 At 9:20, Stimson’s plane took off into an overcast sky on the first leg. Theirs would become a 12,000-mile trip after Winston Churchill took over Stimson’s schedule in Britain and, as a result, General Marshall urged Stimson to fly on to North Africa to get another opinion from Gen. Dwight D. Eisenhower.
Separately making their way by air from Washington to London, via New York, Lisbon, and Ireland, were FDR’s science adviser, Dr. Vannevar Bush, and two antisubmarine warfare technology experts. Bush had been invited by Sir Stafford Cripps to attend a meeting of the War Cabinet Anti-U-boat Committee on July 15. The three Americans expected to use their time in Britain to work with their British colleagues in the Allied Anti-Submarine Working Group.4 However, the coincidence of Bush and Stimson’s presence in London would lead to a meeting with Churchill on July 22 critical to both Anglo-American cooperation on atomic bomb research and Allied strategy for the European theater. Bush, who had lunched with FDR on June 24, believed that he was prepared for an atomic research sharing discussion, but only if need be.5
Delayed by weather, Stimson’s party finally took to the air again from Gander, Newfoundland, headed for Iceland on the evening of July 9. As their C-54 flew northeastward through the dusk of the subarctic summer night, 4,000 miles to the southeast, ten Allied divisions of Operation Husky began assaulting beaches and parachuting onto landing grounds in the invasion of Sicily.6 The Allies were stepping onto a doorstep into the Continent from the south while, far to the east on the steppes of Russia, the titanic Battle of Kursk raged between 2.7 million German and Soviet soldiers.
After twenty-four hours inspecting troops and meeting with U.S., Danish, and Icelandic officials, Stimson flew on to Prestwick, Scotland. Arriving in Windsor by overnight train from Prestwick early on July 12, they took up residence at Stanwell, a sixteenth-century shooting lodge built for Henry VIII.7
Upon completing a day of initial meetings with British and American officials, Stimson went to No. 10 Downing Street to dine with the Churchill’s, Ambassador and Mrs. John Gilbert Winant, and Foreign Secretary Sir Anthony Eden. Discussion quickly turned to Operation Overlord, which Stimson referred to in all his reports on this trip as “Round Hammer.”8 Churchill again expressed his apprehension about direct attack, and Eden “painted a rosy picture of the possibilities of stimulating trouble in the Balkans and Greece against the Nazis.”
Stimson responded in the context of the approaching presidential election. He described to Churchill and Eden “the political danger of a delay, pointing out the tenseness of the situation in the U.S. and the danger of the people not understanding or approving further penetration of the Eastern Mediterranean; the consequent possible loss of prestige to the President, with the immense consequent damage to the Common Cause.”9 Stimson explained that “only by an intellectual effort had [the American people] been convinced that Germany was their most dangerous enemy and should be disposed of before Japan.” Their anger at Japan’s aggression, he said, competed with their acceptance of the “Europe first” strategy.10 In his report on the trip, Stimson reflected that he did not think Churchill had had the situation in the United States explained to him that way before.11
After departing from the dinner, Stimson and Ambassador Winant told each other that they had been encouraged by Churchill’s support for Bolero and Overlord, “with conditions . . . less destructful than before.” Over conversations in the week that followed, but from this dinner particularly, Stimson formed an assessment of Churchill that he would express to General Marshall in a secure telephone call on July 19. The prime minister, Stimson observed, was prepared to keep his commitment to cross-Channel assault but that his impulsive attraction to tangential commitments could make preparations for the cross-Channel assault impossible. Churchill’s commitment to advancing in Italy was strong.12
The next day, Stimson met with Brig. Gen. Ira Eaker, the commander of the U.S. Eighth Air Force. Stimson found Eaker “intensely interested in a thorough softening-up of the opposition on the Continent, and gave every evidence of desiring to go on according to the directive,” meaning the Combined Bomber Offensive agreed upon at Casablanca in January.
Stimson sat next to Churchill again at a dinner in the secretary of war’s honor given by Winant at Claridge’s Hotel on July 14. The dinner was attended by most of the senior British and American officials whom the secretary would be seeing in the days ahead. Stimson observed that he and Churchill further discussed the matters they had gone over two nights earlier; the substance of their remarks to each other is not in the record.13
*
Now emerged the issue of sharing Anglo-American atomic information that soon would draw in Stimson. About to attend a July 15 meeting of the War Cabinet Anti-U-boat Committee, Dr. Vannevar Bush called on Churchill at No. 10 Downing Street. Escorted by Minister for Air Sir Stafford Cripps, Bush had expected the visit to take thirty seconds “to pay his respects.” Churchill, however, confronted the American on the breakdown in atomic sharing and then floated his approach to a solution.
The essence of their exchange can be reconstructed from fragmentary accounts in British and American records. Bush wrote in 1970 that Churchill upbraided him for “ten to fifteen minutes” on the atomic information interchange issue describing the U.S. position as unfair, unreasonable, and nonsensical.14 To support his position, Churchill produced a copy of the January 1943 Conant memorandum. According to the British official history, Bush replied that he was “shocked at the document and doubted whether it had even existed,” adding that he himself had never seen it.15
Bush, however, had communicated with James Conant about the drafting of his memorandum. He was involved in drafting the Military Policy Committee’s December 15 recommendation to FDR, which was identical in substance if more constructively worded. Bush knew the committee’s recommendation had been conveyed to the British and certainly knew of James Conant’s communication with the British.16
According to Bush’s own meeting notes, handwritten on No. 10 Downing Street stationery, the prime minister had more fundamental points. Churchill told Bush that he “had the Pres[ident’s] word of honor to share equally.” Churchill then alluded to “the threat from the East,” underlined by Bush. Concluding his notes, Bush recorded Churchill’s statement that the “Pres[ident] agreed that interchange should be resumed.” According to Bush’s notes, Churchill believed FDR’s oral commitment had been made in an undated telephone call between the two leaders in response to a July 9, 1943, cable from Churchill.17 According to U.S. records, FDR’s response to Churchill’s cable was to consult Harry Hopkins, who reminded the president of his oral commitment to Churchill, May 25–26, at the close of the Trident Conference.18 FDR did not issue instructions to Bush to “renew” the information exchange until July 20, 1943. The textual meaning of that message from the president to Bush apparently was intentionally altered before it left Washington.
Conversing with Bush, Churchill introduced most of the elements of an Anglo-American agreement to resolve the matter. Churchill’s points would evolve and expand. In this first airing, they included interchange of all information, commitment never to use an atomic weapon without the consent of the other party, and deferral to a decision by the U.S. president on British commercial use of atomic energy. Churchill avowed “no interest” in commercial aspects.19
When the prime minister had finished, Bush replied, “The American atomic energy development is now under the Army. The Secretary of War of the United States is in London, and I certainly do not propose to discuss this subject in his absence.” Having asked for a follow-up talk early in their exchange, Churchill replied that he would defer the matter to “a full-dress discussion” later.20 That afternoon, Bush apprised Harvey Bundy, Stimson’s aide, of the meeting and the gist of Churchill’s case. Bush used the next several days to prepare his own response.21
On the day he upbraided Bush, July 15, Churchill was generally combative and assertive. He approved Operation Gomorrah, a series of four night bombing raids by the RAF that would devastate the German city of Hamburg with fire storms now estimated to have caused between 37,000 and 40,000 deaths.22 On July 15, Churchill also wrote to South African field marshal Jan Smuts, a close adviser, with regard to Allied strategy: “I will in no circumstances allow the powerful British and British-controlled armies in the Mediterranean to stand idle. . . . Not only must we take Rome and much as far north as possible in Italy, but our right hand must give succor to the Balkan patriots. . . . I shall go to all lengths to procure the agreement of our Allies. If not, we have ample forces to act by ourselves.”23 Although Churchill’s conditions for Bolero and Overlord, stated to Stimson and Winant three nights earlier, may have been “less destructful than before,” in his letter to Smuts, the prime minister was resolute in pursuit of his Mediterranean goals.
*
A short distance from No. 10 Downing Street, Stimson was visiting COSSAC, the Overlord planning office at Norfolk House, for what Stimson later described as “the most important meeting I had had.” Stimson received a briefing by the U.S. deputy for COSSAC, Maj. Gen. Ray Barker, on the just-completed “Overlord Outline Plan” followed by a frank discussion with Lieutenant General Morgan and Barker. Both men impressed Stimson not only with their confidence in the plan but also with their deep concern about distractions that could dissipate the forces intended to be available when the time came for the cross-Channel attack. According to Stimson, Morgan candidly was “very fearful of delays caused by getting too deep into commitments in the Mediterranean.” In particular, he felt that it was imperative that the divisions being released from the Mediterranean for Overlord on the first of November actually should be free to move back to the UK starting on that date and not merely to plan to come back.”24 Though probably known to the British COS Committee informally, these concerns had not been in Morgan’s cover letter when he submitted the plan to them earlier that day.25
The fact that Morgan and Barker had presented the “Overlord Outline Plan” to the U.S. secretary of war soon would influence the two generals to send the plan to Washington before they had approval to do so from the British COS Committee. What Stimson heard from Morgan and Barker in London would also add critical weight to his report to Roosevelt.
Subsequent to meeting with Morgan and Barker, Stimson also talked with Lt. Gen. Jacob Devers, commander of ETOUSA, who was concerned about the delay of a large contingent from the United States that included one division. Stimson found Devers’s fears to be similar to Morgan’s.26 A week later, on July 22, Stimson had lunch with Lt. Gen. John C. H. Lee, commander of the U.S. Army Service Forces. A concerned Lee wanted no curtailment of the flow of the logistical buildup of weapons and materiel into the United Kingdom, a position that complemented exactly the concern for the troop buildup expressed by Devers and the COSSAC planners.
*
That day from Washington, President Roosevelt continued his quest for a private meeting with Joseph Stalin. In a cable to Stalin, apologizing for a friendly fire incident in the North Pacific, FDR wrote, “I hope to hear from you very soon about the other matter which I still feel to be of great importance to you and me.” Handwritten notation on the surviving copy of this cable reads, “In President’s letter to Stalin via Amb. Davies,” who delivered it in Moscow on May 5, 1943.27
The following day Roosevelt cabled Churchill again. FDR had “still heard nothing from U.J.” When he did hear from Stalin, he would let Churchill “know at once about ABRAHAM [the Quebec Conference].” He liked Churchill’s suggested date.28 In a breezy way, the president had reminded Churchill that the upcoming Anglo-American strategy conference, precise dates still to be determined, could be influenced by an exclusively bilateral meeting between the U.S. president and the leader of the Soviet Union.
*
Blocks away from the White House, the Combined Chiefs of Staff permanent representatives were meeting at the old Public Health Building. The American representatives had a new initiative. With Admiral King’s support, General Marshall proposed to the CCS that Eisenhower be sounded out about a landing in Italy south of Rome as a flanking attack. Marshall’s intent was to knock Italy out of the war. The Allies then could go on the defensive in Italy and concentrate on the buildup in the United Kingdom for the cross-Channel invasion. The British representatives at the CCS staff meeting enthusiastically endorsed Marshall’s proposal and suggested landing south of Rome in late August.29 The report from the July 16 meeting in Washington would not take long to reach Churchill in London.
Since Roosevelt’s June 28 proposal to Churchill for a full-dress conference at the Citadel in Quebec City,30 the U.S. military leadership and Harry Hopkins had been anxious to learn precisely when and where Churchill would land in North America for the gathering. Their concern was that Churchill would come early to meet privately with Roosevelt and—once again—influence him in favor of the British position on Allied strategy in advance of the Joint Chiefs’ battle to control the outcome of the formal conference. Should that happen, the timing could be bad for the U.S. side. Marshall and King knew that internal support for the U.S. strategy for the conference was not yet solid. Some staff organizations in the Pentagon were experiencing doubts of their own about their commitment to cross-Channel attack.
For both countries’ leadership, the regular CCS staff meetings in Washington were not just a forum for formal exchanges but also a fertile source of informal information. The Americans were alert for hints as to Churchill’s next move in the contest for FDR’s support for either of the strategy alternatives. At this CCS meeting, side conversations with the British about Churchill’s intentions may have stimulated concern among American officers for an early visit to FDR by Churchill and triggered action to forestall that.
What is known from Harry Hopkins’s telephone records is that late on Friday afternoon, after the close of the CCS meeting, the president’s closest adviser spoke with the head of FDR’s Secret Service detail, Michael Reilly, and the JCS chairman, Adm. William Leahy. Late the next morning, Hopkins spoke again by telephone with Reilly.31 Reilly and Dewey Long of the White House Travel Office planned and organized the president’s travels away from Washington.
Taken off the shelf at the White House was a proposal for a fishing trip to McGregor Bay in Ontario, Canada, 760 miles to the northwest, first put forward in April 1942 by Zenith Radio Corporation founder E. F. McDonald.32 Coincidental or not, presidential travel began to be arranged with dates that would narrow Churchill’s opportunities to meet with Roosevelt by hanging a “gone fishing” sign on the White House. Although FDR eventually would share advanced knowledge of the trip with Churchill, the out-of-country excursion deep into Ontario’s north country would be kept unusually secret.
*
Saturday morning in England, July 17, the prime minister took the U.S. secretary of war on a daytrip to inspect coastal defenses. Churchill’s special train steamed out of Victoria Station and, clear of London, picked up speed eastward through the countryside of Kent as its passengers sat down in wing chairs to breakfast. They were bound for Dover and the Channel coast.33 The train was equipped with a dining car, a generator, and communications that could be connected to London from anywhere on Britain’s railways. At the heart of the train were two splendid early twentieth-century coaches that had been intended for royalty. Elegantly paneled, carpeted, and upholstered, the cars offered Churchill and his guests open saloons, bedrooms, a conference room, and an office for the prime minister.34
Detraining at Dover, the party inspected coastal artillery batteries that sometimes dueled with German long-range guns across the Channel, shelters for Royal Navy motor torpedo boats that went out to stalk and skirmish with German E-boats, and subterranean command bunkers “newly enlarged for major trans-Channel operations.”35 Churchill’s agenda for Stimson, however, was anything but favorable to cross-Channel attack.
Back aboard the train, over lunch, Churchill pressed Stimson hard on alternatives to Overlord, producing a British cable that he had received based on the previous day’s CCS staff meeting in Washington. Stating that Marshall wanted a study of Avalanche, an amphibious assault onto the Italian mainland at Salerno near Naples, Churchill argued to Stimson that the cable meant Marshall had come over to support Churchill’s Mediterranean strategy. Stimson responded that Marshall only wanted to get the drive to Rome out of the way so as to focus on Overlord.36
Churchill then took Stimson into his onboard office for an earnest conversation alone and “at length” about the issue of atomic information sharing. Stimson recorded that Churchill was “most anxious that I should help him by intervening in that matter and he would hold himself open at any date to meet me, Dr. Bush, and Mr. Bundy during this coming week. He was to have present with him a few members of his staff, and the Lord President of the Council [Sir John Anderson, director of Britain’s atomic program].”37 Very likely, the prime minister used this conversation to sketch to Stimson the elements of the Anglo-American agreement forming in his mind which he had tried out on Vannevar Bush on July 15. Now expanded by Churchill beyond resuming information interchange, the agreement he proposed entered into aspects of a postwar atomic relationship between the two countries. In its scope, the proposed agreement could test the legal limits of FDR’s war powers.
Already apprised by Harvey Bundy of Churchill’s confrontation with Bush two days earlier, Stimson probably anticipated Churchill’s approach on the train. Now, direct from the source, Stimson had a fuller understanding. As a member of the Military Policy Committee advising FDR, Stimson was well informed on this highly secret and complex topic. Stimson also was attuned to the tension between Churchill’s enduring resistance to the U.S. strategy of direct attack and his intense desire for Britain to be restored to access to atomic weapons research.
*
In the afternoon on July 19, Stimson went to the U.S. Army’s headquarters for European theater operations where, after generally reviewing his meetings in London and the state of CCS decisions on strategy with Lt. Gen. Jacob Devers and Maj. Gen. Alexander Surles, Stimson conducted a secure radiotelephone call over the new Sigsaly system with General Marshall in Washington. Stimson told Marshall of his impression of Churchill, that “he was honestly ready to keep his pledge as to ‘Round Hammer,’ but was impulsively likely to branch out into commitments which would make it impossible, and further that he was very set on a march to Rome.” To Stimson’s description of his Saturday response to Churchill, that Marshall’s intentions for Italy were limited to facilitating renewed concentration on Overlord, Marshall replied, “That was exactly right, you were quite right, that was what I meant.” Marshall then urged Stimson to travel to North Africa to get Eisenhower’s view.38
*
The parallel, but separate, Anglo-American negotiation on resuming atomic cooperation, which came ever closer to the strategy negotiations, now advanced in London. That, however, began with a curious turn in Washington.
Prodded by Churchill again, and on the advice of Harry Hopkins,39 President Roosevelt took action on Tuesday, July 20, to move forward the issue of atomic information interchange. FDR dictated an instruction to Vannevar Bush that would cause consternation among the leadership of the Manhattan Project. Knowing Bush was in London, the note to his science adviser was sent by courier to the Office of Scientific Research and Development, located at the Carnegie Institute at 1530 P Street, N.W., in Washington, for forward transmission:
Dear Van:
While the Prime Minister was here we discussed the whole question of exchange of information regarding tube alloys, including the building project.
While I am mindful of the vital necessity for security in regard to this I feel that our understanding with the British encompasses the complete exchange of all information.
I wish, therefore, that you would renew, in an inclusive manner, the full exchange of information with the British Government regarding tube alloys.
Sincerely yours,
Franklin D. Roosevelt40
The note was received at OSRD by Bush’s executive assistant, Dr. Carroll L. Wilson. OSRD’s communication procedure required paraphrasing the substance of a classified message, as a common additional security measure, before submitting it for encryption and radio transmission, in this case to the OSRD Liaison Office’s London Mission.41 Wilson was responsible for paraphrasing FDR’s message, and he anticipated trouble.
As Bush’s executive assistant, Wilson was involved closely with discussions and deliberations between Bush and James Conant. Wilson had drafted the original uranium development agreement with the British. On receiving FDR’s note to Bush on July 20, Wilson immediately recognized the significance of Roosevelt’s instruction as a major step beyond the “need to know” policy toward the British. Wilson knew that Conant, Bush, and Groves were opposed to sharing further atomic information with the British.
Wilson’s personal position on the matter is not evident in archived OSRD papers, but his actions are. Wilson consulted at least with Conant before forwarding FDR’s instruction. He paraphrased the president’s text for transmission in a manner sympathetic to the position of its opponents. On July 27, OSRD had the Navy encrypt and transmit the president’s July 20 instruction to Bush in the following paraphrase:
To Bush.
Letter to you from your Chief requests you review in an inclusive manner full exchange with the U.K. on Essone. Conant taking no action and will not discuss with anyone awaiting your instructions on return.
From Carroll L. Wilson42
In paraphrasing Roosevelt’s text, Dr. Wilson replaced the critical verb, renew, with review, thus altering the instruction’s intent. Wilson added for Bush information about how Conant would handle the matter until Bush returned. In a note to the president sent July 28 confirming that he had paraphrased and sent FDR’s message, but not saying how it was paraphrased, Wilson wrote, “Tomorrow [July 29] I shall bring your letter to the attention of Dr. James B. Conant upon his return to Washington.”43 Apparent from the paraphrase sent to Bush, however, is that Wilson already had communicated with Conant about Roosevelt’s instruction to Bush. That Conant knew the content of FDR’s letter is evident in Conant’s cable to Bush on July 29, received in London after Dr. Bush had left for Washington. Conant asked Bush if he wanted him to convene a meeting on the issue with the Military Policy Committee and General Groves in Bush’s absence.44
Onward transmission of the president’s instruction to Bush to “renew” full exchange with the British was held in suspense by Wilson for seven days, a week in which Vannevar Bush and Secretary of War Stimson would have a critical meeting with Churchill unaware of the president’s new instruction. When the message did reach Bush, he still would be misinformed by its altered language. Not until Bush returned to Washington and saw FDR’s original note was his misperception clarified.
*
Unaware that two days earlier Roosevelt had stepped into the issue but that his instruction was sitting in a safe in Washington, Bush sat down in London with Churchill, Stimson, and Lord Cherwell for their meeting on atomic information sharing to which, not by chance, Stimson added a separate meeting with Churchill on strategy for Europe.
The morning of July 22, Stimson, Harvey Bundy, and Bush met at Claridge’s Hotel to prepare for their meeting that afternoon with Churchill and his key scientific advisers on the highly secret atomic sharing issue. On grounds of law and policy, particularly the limits of Roosevelt’s war powers, Bush and Bundy advised caution. Neither wanted to see resolution of postwar issues with a foreign power, such as an atomic military capability or atomic power commercial advantage for Britain, determined by presidential executive action that might challenge congressionally authorized war powers constrained to steps necessary to win this war.45 The essence of the information sharing policy in the 1942 Military Policy Committee recommendation to the president, then being implemented across the Manhattan Project, was access to information on the basis of “need to know” in order to win the current war. Without referring to other rumblings within the U.S. atomic leadership, if he knew about them, Bundy advised Stimson to frame his responses to Churchill within the limits of FDR’s war powers.
Bush was taken through his position on interchange of information rigorously by Stimson. The secretary of war used the adversarial format of a mock trial to test FDR’s scientific adviser’s argument. Later in the afternoon, as they walked to No. 10 Downing Street, a satisfied Stimson told Bush to take the lead in the discussion of the S-1 matter according to Bush’s approach to how atomic information sharing should be handled.46
Bush, who opposed sharing with the British, was surprised by Stimson’s delegation to him of the discussion on S-1. The secretary of war earlier had made clear to Bush that he wanted a postwar Anglo-American partnership and considered Churchill’s position on atomic sharing to be correct.47 Bush asked Stimson directly if his intent was for Bush to speak for the American side from his own oppositional perspective of the issue. This Stimson affirmed. From that Bush concluded that the purpose of the exercise at Claridge’s had been to confirm that “I had my arguments in order.”48
True enough. However, with much of the summation that he himself planned to offer to Churchill already formed in his mind, and tasked by Roosevelt to press Churchill on cross-Channel assault, Stimson probably saw Bush’s tactical role in the larger context of a strategy of quid pro quo.49 There is irony in the fact that the boundaries of the OSRD director’s extensive wartime responsibilities left Bush outside European strategy discussions. For those, Bush lacked the “need to know.”
Stimson, Bush, and Bundy were received at No. 10 Downing Street by Winston Churchill, Sir John Anderson, and Churchill’s science adviser, Lord Cherwell. Anderson and Lord Cherwell were the top leadership of Tube Alloys. Absent from accounts of the discussion that followed are allusions to Britain launching an independent atomic bomb development program. These certainly were in each participant’s mind from previous discussions, as recently as one week earlier.
Beginning the meeting, Churchill recounted FDR’s oral assurances to him on three occasions that atomic bomb research would be a joint enterprise. None of the president’s assurances had been committed to paper and, obvious to the British, none had been acted on by U.S. Manhattan Project leaders. Churchill said the British were deeply concerned when they received James Conant’s January 1943 memorandum “rigidly limiting the exchange of information” through a unilateral U.S. application of the concept of need to know.50 Churchill, Cherwell, and Anderson all feared for Britain’s position in a postwar world between the United States and Soviet Union, each atomic-armed, in the absence of a British atomic deterrent.51 According to Harvey Bundy:
The Prime Minister took the position that this particular matter was so important that it might affect seriously British-American relationships; that it would not be satisfactory for the United States to claim the right of sole knowledge in this matter. The Prime Minister further said that Britain was not interested in the commercial aspects but was vitally interested in the possession of all information because this will be necessary for Britain’s independence in the future as well as for success during the war; that it would never do to have Germany or Russia win the race for something which might be used for international blackmail; and that Russia might be in a position to accomplish this result unless we worked together.52
Taking the lead in responding for the U.S. side, per Stimson’s instruction, Bush emphasized that the U.S. position was based on cooperation to facilitate winning the current war, “that post war problems were separate, and that the difficulties of complete exchange lay in respect to post war matters, both political and commercial.”53 Churchill, supported by Cherwell and Anderson, replied that commercial interests were not a factor in the British position and that the United Kingdom would be willing to enter any agreement FDR thought fair, taking into account the relative contributions to the Manhattan Project of the two countries.54
Secretary of War Stimson then offered his own summation of the situation. With the advantage of five days to have considered what Churchill had said to him on the train trip to Dover, and having used Bush to present the U.S. case, Stimson now read from his prepared notes:
1. Two Governments in possession of an unfinished scientific hypothetical formula on which they are working.
2. Both Governments continue working on the development of that formula and are ready to interchange reports of their respective developments.
3. U.S. at large expenditure of public monies sets on foot construction out of which these formulae may be transformed into practical products; on the understanding that the U.K. may share these products for the joint object of winning the war.
4. U.K. now asks U.S. for running reports on its constructive designs and other manufacturing experience, in order that UK after the war is ended and its present strain of other construction is over, may be in a position to prepare itself to promptly produce against the danger of a new threat or a new war.
5. Should the U.S. grant this request unequivocally? Should it seek safeguards against any use of product except under political restrictions? Should it refuse the request as entirely uncalled for, under the original agreement between the President and the Prime Minister?55
Responding that Stimson’s “was a trenchant analysis of the situation,” Churchill offered to accept an agreement between himself and Roosevelt that the prime minister then outlined with five points:
1. A Free interchange to the end that the matter be a completely joint enterprise.
2. That each Government should agree not to use this invention against the other.
3. That each Government should agree not to give information to any other parties without the consent of both.
4. That they should agree not to use it against any other party without the consent of both.
5. That the commercial or industrial uses of Great Britain should be limited in such a manner as the President might consider fair and equitable in view of the large additional expense incurred by the U.S.56
The essence of a resolution of the atomic information sharing impasse had been put into play—in draft—but with political and military elements that could be agreed to only at the presidential–prime ministerial level. Stimson offered to convey Churchill’s proposed agreement to Roosevelt. Churchill agreed to put the proposal in writing.
*
With the discussion of atomic weapons cooperation concluded, Stimson sent the others out of the room so that he could meet alone with the prime minister on the issue of equal importance and urgency to the United States, which was clearly still open, despite the Trident Conference. That was reaffirmation of cross-Channel assault as the primary Allied liberation strategy.
The two men began with by then familiar lines of opposition and “had it hammer and tongs,” according to Stimson. Churchill, citing strong German resistance in Sicily, told him that “if he had 50,000 men ashore on the French Channel Coast, he would not have an easy moment, because he felt that the Germans could rush up in sufficient force to drive them back.” His anger rising, Stimson accused Churchill in strong terms of reneging on the commitment to Overlord decided at the Trident Conference two months earlier, telling Churchill that his statements were “like hitting us in the eye.” Questioned by Stimson, Churchill responded that he would not have Overlord if the decision were his, but would “loyally” implement the agreed strategy. However, Churchill remained eager to advance to Rome. Stimson replied that, while he did not question the sincerity of Churchill’s promise, he also worried that Churchill did not allow for the long-term planning and preparation essential to an operation like cross-Channel attack. He told Churchill of his meeting a week earlier with COSSAC at which Generals Morgan and Barker expressed confidence in their Overlord Outline Plan but also concern for diversion of resources. Stimson closed the meeting by telling Churchill he would be “taking every step to try to prevent any further encroachment into the plans made for Round Hammer [Overlord].”57
Churchill’s expression of intent to “loyally” follow the cross-Channel assault strategy well might have been a backhanded allusion to American disloyalty in not putting into effect FDR’s oral—not written—promises in October 1941 and May 1943 of full exchange of atomic information.58 As a former U.S. Attorney for the Southern District of New York, Stimson could neither miss nor ignore the lawyerly quality of Churchill’s response on Overlord.59 From his perspective, despite the light that seemed to be breaking early in his meeting with Stimson, Churchill almost certainly held a no less critical opinion of the “need to know” rationale for U.S. suspension of atomic information sharing with the British. From Stimson’s last substantive meeting with Churchill on this trip, the secretary of war left for North Africa.
The meeting had produced a path that could resolve the issue of atomic research sharing with further negotiation. However, Stimson drew two conclusions with respect to Overlord that altered the impression of Churchill that he had expressed to General Marshall three days earlier and that would shape Stimson’s recommendations to Roosevelt. Stimson was convinced that the British would continue every effort to find an alternative to Overlord, and knowing their prime minister’s reticence, he moved toward the conclusion that the cross-Channel invasion could not succeed with any British commander.60
*
From No. 10 Downing Street, Vannevar Bush returned to the U.S. embassy. There he prepared a cryptic memorandum, to be sent by diplomatic pouch on July 23, to alert his deputy, James Conant, that an Anglo-American agreement on atomic information interchange was in preparation and probably would be delivered to FDR by the secretary of war. Citing pressure he encountered from the British, but believing that “the outcome is not going to be bad at all,” Bush lamented, “I have had to go alone for there was no way whatever of stopping the progress nor would it indeed have been advisable to do so, and I merely hope that I have not made a lot of mistakes in advising the secretary.”61 Bush wrote to Conant while still ignorant of FDR’s new instruction and then turned to the antisubmarine and other research business that had caused his trip to Britain.
As Stimson and Bush negotiated in London with Churchill on a way forward on atomic information interchange, FDR’s new instruction to Bush of July 20 remained with Carroll Wilson at OSRD headquarters in Washington. Events, however, moved on.
Bush learned from Sir John Anderson on July 27 that another message to Churchill had arrived directly from FDR. Apparently assuming that his July 20 instruction had been received by Bush as he had written it and had been acted on, Roosevelt cabled Churchill that he had “arranged satisfactorily for the TUBE ALLOYS” and recommended that Churchill send over his “top man in this enterprise” to get “full understanding from our people.”62 Mystified as to the genesis of FDR’s cable to Churchill, and concerned that the agreement outlined in the July 22 meeting with Churchill not be undercut, Bush sent an anxious radio message to Harvey Bundy, who was with Stimson in North Africa. On July 28, six days after meeting Churchill, Bush received in London the transmission of FDR’s new instruction, but paraphrased in a way that altered the president’s intent. Reading the misleading cable after the meeting, Bush recalled in his memoir that he concluded at the time, “Well this simply tells me to do what I am doing now.”63
Scheduled to leave for Washington early the next morning, Bush still wanted instructions. He sent another radio message, now marked “urgent,” to Bundy in North Africa. “Cable now informs me that my chief by letter requests me to review in inclusive manner full exchange with [UK].”64 Secretary Stimson replied in an urgent message that evening, sent under Eisenhower’s signature. He recommended that Bush “advise Hopkins for the President that the Prime Minister opened the matter with the Secretary and subsequent discussions went well, but emphasize that of course no commitments were made.” The message closed with Stimson’s assumption that he would be carrying back to FDR immediately Churchill’s suggested settlement.65 FDR’s cabled invitation in hand, however, Churchill would not wait for that.
Concurrent with the meetings in London, preparations began for the fourth full-dress Anglo-American conference, Quadrant. On July 20, Churchill sent a cable to Roosevelt in Hyde Park that Canadian prime minister William Mackenzie King welcomed the proposal to hold the Quadrant meeting in Quebec City.66 The following day, Churchill cabled FDR, “Planning arrive ABRAHAM 11,”67 meaning the Plains of Abraham (Quebec City) August 11. Quadrant would commence in approximately three weeks, but of concern to the JCS, Henry Stimson, and Harry Hopkins, substantive talks at the top of the Alliance might occur earlier. Stimson’s recollection was that Churchill told him on July 22 or 23 of his intention to travel to North America early in August in advance of the Quadrant Conference, not information that Stimson kept to himself.68 Churchill’s early arrival could only mean that the prime minister intended to meet with Roosevelt before the conference.
Actions initiated the previous week at the White House already had accelerated, ostensibly to give “the Boss” some rest in a remote venue before Quadrant. Thomas Beck at Crowell-Collier Publishing in New York had been the intermediary for the idea of an FDR fishing trip to the Great Lakes. On July 20, at 10:20 a.m., Harry Hopkins telephoned Beck to get the number for the originator of the proposal for the trip, E. F. McDonald in Chicago. Five minutes later, Hopkins called McDonald.69 The next day, as Mike Reilly of the Secret Service was consulting with Hopkins about the trip, McDonald mailed a special delivery letter describing updated arrangements for fishing by the president in McGregor Bay, Ontario, “when and if the trip materializes.”70 The morning of July 22, Brig. Gen. John Deane of the JCS Secretariat met Hopkins at the White House without an appointment, and at 2:30 p.m. Hopkins met General Marshall at the Pentagon.71 Probably on Friday morning, July 23, when McDonald’s updating letter was received, the decision was made at the White House. Before the Quadrant Conference, FDR would be going fishing.
*
Secretary of War Stimson, on General Marshall’s strong recommendation, had reached Scotland to embark on the long flight to North Africa to consult with Eisenhower. Taking off from Prestwick for Marrakesh on the evening of July 24, the Stimson party’s C-54 flew west across Northern Ireland in the near-daylight of northern summer and west over the Atlantic before turning south. The navigator had planned to fly south along 14 degree longitude, but Stimson moved the southward leg further west to the 18th degree “for safety.” This change put their course well beyond German radar range from the French coast and that of Luftwaffe JU-88 fighter bombers hunting for patrolling Allied antisubmarine aircraft, but still within the outer edge of normal operating area of Luftwaffe four-engine FW-200 maritime reconnaissance bombers flying out of Bordeaux.
There were reasons for prudence. With some frequency, Allied antisubmarine patrol aircraft were intercepted and shot down by the JU-88s in the Bay of Biscay, as had been a KLM airliner, en route from Lisbon to London with actor Leslie Howard on board only three weeks earlier. Less frequently, farther west, Allied very long-range, four-engine Liberators on antisubmarine patrols and the Luftwaffe FW-200s, hunting targets for the same U-boats, sometimes encountered each other and fought aerial duels. Stimson also knew from U.S. actions that, given foreknowledge of their movements, senior leaders were wartime targets for assassination. Only three months earlier in the South Pacific on April 18, in an intelligence-based planned ambush, American fighter planes had shot down two Japanese bombers, killing Combined Fleet commander and Pearl Harbor attack planner Adm. Isoroku Yamamoto and most of his senior staff.72
Stimson’s aide, Colonel Wright, recounted, “The weather was most favorable for our purpose, since we passed the most dangerous area west of the Bay of Biscay with good cloud cover above and below,” cruising at six thousand feet.73 Drawing closer to the North African coast in the morning, Stimson’s plane came within range of a protective response from Allied fighters if needed.
Events indeed were moving forward in Washington, stimulated by Stimson’s report from London on Churchill’s anticipated travel. On Sunday, July 25, from Shangri-La, FDR responded to Churchill’s cable that Lady Churchill and daughter Mary would be coming with him to Quebec by inviting Churchill and family to Hyde Park for a visit to be concluded by August 15. On that day, Eleanor Roosevelt would leave Hyde Park on an inspection trip to the Pacific. Roosevelt also told Churchill, “I hope to go on a short trip to fish and sleep next Saturday [July 31].”74 Based on subsequent British communication of new dates for Churchill to be at Hyde Park, FDR’s message must have communicated something of the duration of the fishing trip.
A fishing trip? As Marshall in Washington viewed with increased foreboding the weeks ahead, across the Atlantic, the president’s almost nonchalant mention of a fishing trip must have stimulated Churchill to wonder if he was the recipient of a political euphemism.
For two months, while their military chiefs grappled over strategy, the two western leaders had engaged in earnest communication about Roosevelt’s desire to meet Stalin alone with almost no staff and without Churchill. FDR believed that such intimate informality would allow him to build trust with Stalin and, unburdened by an agenda, make progress on East-West issues in advance of the Big Three leaders meeting to address the critical war years ahead. After a strong protest in June, Churchill had accepted, grudgingly, the potential value in FDR meeting alone with Stalin in the context of a Churchill-FDR bilateral and Big Three meeting quickly following. Churchill never relaxed his concern for the effect on Britain’s standing as an ally or its postwar position, subsequent to any wartime meeting with the Soviet leader that did not include Britain at the table.
The expectation, outlined by the Soviet premier in May, was that he would give FDR two weeks’ notice of when he could be at a meeting place, probably in August. On that basis the president would organize the venue, suggested by Stalin to be Fairbanks, Alaska.75 Roosevelt had told Churchill that he did not expect to hear from Stalin before July 15.76 Then came word to Churchill from the president, setting back the date after which Churchill would be welcome at Hyde Park, that he would begin travel into a remote area of Canada a little more than two weeks beyond the date on which FDR had anticipated a response from Stalin. FDR had used a “fishing trip” in the vicinity of Cape Cod as a ruse to cover his meeting with Churchill at Argentia, Newfoundland, in 1941.77 At Trident in May, FDR had called Churchill’s post-Casablanca trip to entreat Turkey’s president to enter the war on the Allies’ side a “fishing trip.”78 FDR had demonstrated by his use of it a liking for this Americanism for an exploratory meeting. Given Churchill’s heightened appreciation of the importance of Quadrant, a pre-Quebec meeting between FDR and Stalin was a source of anxiety for him.
Whatever concern he might have had, Churchill remained resolute in his position on strategy. The next day, building on Mussolini’s demise, Churchill sent an effusive letter to FDR describing prospects for a Mediterranean strategy.79 The British Chiefs of Staff held to their July 24 “stand fast” order to British forces in Mediterranean, pending a reappraisal of strategy at Quadrant.
Uncertainty about just when Churchill and the British delegation for Quadrant would arrive in North America was resolved in the United States by a message to the Joint Staff Mission from British COS for delivery to General Marshall. Received in Washington the afternoon of July 26, the message stated that “‘Colonel Warden’ [Churchill] will arrive at Halifax evening of the 9th by the same method as last time [aboard the Queen Mary]. Colonel Warden and his lady will then go to Hyde Park until 15th, since P.Q. [Roosevelt] does not come to Abraham until 17th.” The British delegation would go directly from Halifax to Quebec, arriving there early on August 10 and “hoped all their opposite numbers from Washington will be able to arrive that day.”80 Sent Most Secret and Most Immediate, the message was hand-delivered to Col. Frank McCarthy, Marshall’s aide.
The U.S. Joint Chiefs of Staff had their own conception of when the U.S. delegation should arrive in Quebec City and would not allow themselves to be rushed by the British. A new complication for the JCS, emerging from subordinate staff, was dissent from their goal for the fast-approaching conference. The JCS also remained uncertain about their commander in chief’s commitment to that goal, reassertion of Overlord’s primacy, and FDR was about to go fishing.
Mike Reilly and two Secret Service agents departed the White House on July 26 for Chicago, on the first leg of a reconnaissance to arrange for FDR’s trip to Ontario. Uncertain about the dates for Churchill’s trip to North America—and the possibility of travel by Stalin—Reilly and Dewey Long had stood by in suspense on the dates for FDR’s trip until then.81 During this interval, by some means, Churchill was informed that FDR could not receive him at Hyde Park before August 12 because he would be “fishing” or in Washington and that he would depart Washington on July 30. For thirteen days before Quadrant, the president would not be available to the prime minister except via cables.
Back at the White House from Shangri-La, Monday, July 26, FDR cabled Churchill that he had “arranged satisfactorily for the TUBE ALLOYS” and recommended that Churchill send over his “top man in this enterprise” to get “full understanding from our people.”82
*
Working with Maj. Gen. Alexander Surles on Tuesday, July 27, in Algiers, Secretary Stimson compiled his notes on three days of meetings in North Africa with Generals Dwight Eisenhower, Carl Spaatz, and Jimmy Doolittle.83 Stimson agreed with the generals he consulted in Algiers that, following the conquest of Sicily, an operation on the Italian mainland “seems not only advisable, but essential to cover the interim before mounting a powerful [cross-Channel] invasion offensive for a knockout.” If the operation Eisenhower was developing for an amphibious landing at Salerno (Operation Avalanche) was feasible, this would be preferable to a long, costly slog up the boot of Italy.
However, Stimson concluded that the operation was being advocated for two other reasons that were in conflict. He interpreted the British concept to be “the use of the favorable conclusion of such an operation as the basis for a ground operation in the direction of the Balkans or of the Mediterranean avenues of invasion.” As to the goal of the operation for the American commanders in North Africa, Stimson interpreted that to be shaped instead by the needs and contribution of air power.84
Highly desired by Eisenhower and Spaatz (a desire shared in Washington by Marshall) were air bases on the Italian mainland as far north as Rome from which heavy bombers could attack southern Germany. Flying from Britain, its operations limited by weather, the Eighth Air Force’s losses to steadily improving German air defenses were “approaching the margin of safety” at 10 percent, whereas flying in the south incurred losses of less than 1 percent and benefited from better weather. The airmen did not believe German air defense in the south could be developed to the extent it was in northwestern Europe. If this second axis of air attack could be established concurrent with the mounting of Overlord, Stimson and Major General Surles believed the Allies in the west “would have the maximum advantage in effect upon Germany both psychologically and militarily.”
“However,” they concluded, “the Southern effort must be strictly confined to this objective of the air attack. There must be no further diversions of forces or material which will interfere with the coincident mounting of the Round Hammer [Overlord] Project.”85 In Stimson’s interpretation of what he had heard in Algiers, the American goal for assault on the Italian mainland left scant room for the British conception.
Stimson returned to Washington on Saturday, July 31, and went to his office in the Pentagon on Sunday.86 He had missed President Roosevelt, who was by then 760 miles to the north in the Canadian bush.