4

The Trident Conference’s Illusion of Agreement

One of the constant sources of danger to us in this war is the temptation to regard as our first enemy the partner that must work with us in defeating the real enemy.

—Gen. Dwight D. Eisenhower, letter to Brig. Gen. Thomas Handy, early 1943

On Tuesday, May 11, 1943, the American and British Chiefs of Staff, Churchill, and FDR were convening again, this time in Washington. A British delegation arrived at 6:30 p.m. and met their counterparts directly at a hotel cocktail party.1 Trident would be the follow-on conference for which the need had been so evident at the close of Symbol in Casablanca.

The Americans entered this conference with a determination not to be dominated, as they had been at Casablanca, by their British guests’ better staffing and preparation.2 The Americans were hopeful, though not confident, that this time Roosevelt would stand firm with them in his meetings with Churchill, particularly on strategy for Europe.

Certainly the Americans had prepared themselves as best they could, much better than they had for Casablanca. In advance of Trident, they produced and approved rigorous staff papers in six topical categories expected to be discussed. Better prepared, the American joint staffs were primed and standing by to make an agile response to any negotiating contingency as indeed they were called to do. By the time Trident concluded, the staff had produced thirty-seven formal studies in twenty days.3

Also stirring in the background in the spring of 1943 was a transformation in how American decision makers, political and military, viewed the security of the United States and its role in the world. Among shapers of U.S. policy could be heard rising calls for a newly internationalist and global outlook, particularly in the form of “Atlanticism.”

In April the influential newspaper columnist Walter Lippmann had published in Boston his new book, U.S. Foreign Policy: Shield of the Republic. On the basis of his sweeping summation of the sources and the evolution of U.S. security and foreign policy interests since the American Revolution, Lippmann forecasted global interests and prescriptions for securing them in the context of the twentieth-century industrialized world. He sought to shake off as forever obsolete the isolationist “hemispheric” perspective of the United States living shielded behind the barrier of two great oceans.

It its place Lippmann found “the shield of the republic” to be engagement in the world. Looking to win the war and embrace interests shared with postwar Europe, Lippmann argued that the United States could not allow any potential foe to control Western Europe. The freedom and security of nations on both sides of the Atlantic, he wrote, rest on an enduring interdependency. Therefore, the Atlantic Ocean should be viewed neither as a frontier for the United States nor as a geographic division between the Old and New Worlds. It should be “the inland sea of a community of nations allied with one another by geography, history and vital necessity.”4 A corollary was Lippmann’s projection of the essential U.S. security relationship to East Asia that pointed to a Pacific community.

As a manifestation of American nongovernmental, intellectual response to the global changes being thrust on the United States by war, Lippmann’s book also provided a welcome framework for operational needs. In March, expanding on a pre-Casablanca request from Roosevelt, the JCS had set planners to studying the parameters of a global postwar system of air bases. That was to be defined in the context of “the broader subject of worldwide military problems which will confront the Joint Chiefs of Staff as soon as the Axis power surrenders—and which, therefore, must be studied now in order that we may be fully prepared.”5

The military leadership of the United States found Lippmann’s book, published after their embarrassment at Casablanca and one month before Trident, to be exceptionally timely. Historian Mark Stoler concludes that JCS planners “certainly did” read Lippmann’s book, which would become a best seller, be popularized in national print media, and even be made available to U.S. troops in a 25-cent paperback edition.6 Lippmann’s book was read by Secretary of War Henry Stimson and Adm. Ernest King. Stimson recommended the book to Gen. John Hull, then a key U.S. Army planner, with his strong endorsement of the need for a new U.S. foreign policy.7 King, who sent Lippmann a note of thanks, told an off-the-record press conference, “We are grown up now and we have got to evolve a foreign policy and stick to it.”8

The British Chiefs of Staff arrived for Trident with their understanding of their country’s role in the world already long confirmed by centuries of imperial rule. They had two concerns, one basic and one immediate.

Most basic to the COS and Churchill was that Great Britain’s fighting capacity was fully engaged. In its fourth year of total war, having stood alone for a year and a half early in the struggle, Britain had nearly emptied its treasury. Britain was at the limit of the home islands’ manpower pool. The number of operational British divisions had peaked at eighteen. The day was approaching when some divisions would have to be disbanded in order to maintain the rest at full strength and reorganized as mobile expeditionary formations.9

Beyond these immediate and quite tangible military force issues, centuries of experience encouraged the British to a long view. They knew from current Allied planning that well over one million American troops would be inserted into Europe for a direct assault strategy. Obvious to them was the postwar political influence that would accrue to the United States from an American force of such size in Europe. The opportunistic strategy of weakening Germany by attrition and only then pouncing was the only military option remaining through which Britain could continue in a continental war as an equal partner in the Alliance. From the same long experience, British leaders, particularly Churchill, were acutely aware that the western Allies should conclude the war in Europe in a strong political-military position on the ground in Europe relative to that of Russia.

Flowing from their unvarnished understanding of the demands of the task still ahead and their finite means was the COS’s immediate concern, the military capability of their ally on whose military and materiel capacity they were critically dependent. Gen. Dwight D. Eisenhower had made serious mistakes while leading the first major Anglo-American ground operation in North Africa. These mistakes were military in nature against the Germans and political in dealing with collaborationist Vichy and the Free French. Their effect on the British General Staff was to fuel a conclusion that Eisenhower would have to be, and could be, managed by the British as a figurehead Allied military leader with British commanders of superior experience and ability actually running operations.10 Some went so far as to advocate putting well-equipped U.S. troops under British officers. In the ranks, seasoned British troops who had experienced the uncertain American efforts in North Africa could be heard in pubs singing a biting ditty that drew the title of a then-popular film into parody: “How Green Is My Ally.”11

Yet within forty-eight hours of the Combined Chiefs commencing their new round of talks at Trident, and despite Allied operational and political mistakes in North Africa, Axis resistance ended there in Allied victory. Allied forces, which now included Free French ground combat units, bagged 275,000 German and Italian prisoners, which exceeded the number of Axis troops killed or taken prisoner at Stalingrad in 1942.12

A U.S. Army planner present at both conferences concluded that, before victory in North Africa, the collective attitude at Casablanca had produced decisions limited to the year 1943 and characterized by “wishful thinking relative to our [Allied] Atlantic strategy and a nebulous approach to Pacific strategy.” When the CCS gathered for Trident, after the Tunisian victory, the thinking was “stronger, the decisions clearer, and the strategy more global and more realistic.”13 Logistics, critical to victory, got its due.

High on the agenda for the Trident Conference were the questions of which geographic axis the Allied armies should pursue for their reentry into Nazi-occupied Europe and when this reentry should take place. As at Casablanca, however, the strategy debate at Trident tapped deeper, divergent roots of national capabilities, priorities, and history on both sides. It was painfully evident in the tenor of discussion during its first week that the Trident Conference would be colored by personal suspicions of motives and bias deeply held on both sides.14 These would be tough negotiations.

At the first formal meeting of Trident, a plenary in the White House Oval Office at 2:30 p.m. on May 12, the CCS received guidance from the president and the prime minister. Roosevelt, joined by Churchill, began by remarking on how far the Allies had come in one year since the prime minister and General Brooke were last in that very room. In 1942, they had received the shocking news of the fall of Tobruk followed by Roosevelt’s immediate offer to divert Sherman tanks and armored artillery from then-scarce American stocks to the British in Egypt. There followed in this opening meeting of Trident Churchill’s and Roosevelt’s exposition of their respective views of the tasks before the CCS. Though put gracefully by the two leaders, clear to all was the conflict inherent in resumption of the strategy discussion.

The prime minister promptly set out the Mediterranean as “the first objective.” Among the five objectives that followed, Churchill raised a cross-Channel attack and the buildup of forces for that strategy fourth. Defeat of Japan was last.15

Roosevelt’s response to the heart of Churchill’s position began indirectly. He started by saying that he always had been a firm believer in attrition. He expressed surprise and optimism about the cordial reception that Churchill got on his post-Casablanca “fishing trip” to Turkey, always a potential key component in Churchill’s Mediterranean strategy. FDR stated that perhaps Turkey could be induced to adopt a favorable attitude toward the United Nations by diplomacy alone, implying a limit to his support for military operations as leverage. Then, referring to the scheduled Allied invasion of Sicily, the president asked rhetorically, “Where do we go from HUSKY?” By way of answer, FDR continued that he had “always shrunk from the thought of putting large armies into Italy.” Acknowledging the need to employ the surplus of Allied forces that would be in the Mediterranean following the conquest of Sicily, Roosevelt also warned of the humanitarian burden that would fall to the Allies if Italy were to be occupied. Roosevelt then made a firm statement in support of either a Sledgehammer or Roundup cross-Channel operation for 1944.

The meeting minutes summarize his remarks as being that “if one or the other were to be mounted in the spring of 1944, preparations should begin now. ROUNDUP and SLEDGEHAMMER have been talked about for two years, but as yet none of these operations had been accepted as a concrete plan to be carried out at a certain time. Therefore he wished to emphasize that SLEDGEHAMMER or ROUNDUP should be decided upon definitely as an operation for the spring of 1944.”16

Roosevelt’s comments then shifted to the Pacific theater and support for keeping China in the war. Taken together, FDR’s remarks summed up his Joint Chiefs of Staff’s priorities in competition for resources with what they saw as Churchill’s war-lengthening interest in the Mediterranean. Thus ended Trident’s initial White House meeting with each nation’s military chiefs having gotten their leader’s backing for their opening position, positions that were in direct conflict.

The first short week of negotiation by the Combined Chiefs of Staff began on Wednesday morning, May 13, in the Board of Governors Room of the Federal Reserve Building on Constitution Avenue. Quickly, along a path that led through the reduction of the German Air Force and its implications, the CCS came to their fundamental difference on strategy for Europe.

Admiral King stated that if reduction of the Luftwaffe was an overwhelming success, the Allies then must be in position and ready to exploit that with a cross-Channel operation on the scale of Roundup. JCS chairman Admiral Leahy, supporting King, added that it “might be unwise to divert or maintain in the Mediterranean forces which could be used in a cross-Channel assault.” General Brooke replied for the British: unless operations in the Mediterranean continued, “no possibility of an attack into France would arise.”17

There it was, as General Marshall then stated, “the heart of the problem . . . at the crossroads.” The British were insistent that an attack into France would be possible only if the German Army in France had been significantly weakened by redeployment of divisions to fight the Allies in the Mediterranean. The Americans were convinced that further operations in the Mediterranean would continue to vacuum up Allied troops and resources and thereby deny building up the concentration of forces in the UK for a cross-Channel assault ever. While the JCS at Trident spoke of the Roundup concept in terms of twenty to twenty-five divisions, not its earlier forty-eight, both sides continued to argue their position in absolutes: the Mediterranean or cross-Channel attack, formulated as either Roundup or Sledgehammer. Brooke stated that halting Mediterranean operations after Husky to cross the Channel would lengthen the war. Marshall said that further Mediterranean operations would commit the Allies through most of 1944 with serious implications for the Pacific. The war in Europe, he argued, would be prolonged, causing “a delay in the ultimate defeat of Japan, which the people of the U.S. would not tolerate.”18

General Brooke agreed that the war in Europe must be ended as quickly as possible. However, he said that “seizure of the Brest Peninsula, which was all we could now achieve, would merely lock up 20 divisions.”19 Arguing from the context of the size of the armies fighting in northwestern Europe in World War I, Brooke stated that no major operations would be possible until 1945 or 1946.

General Brooke’s elaboration went to a significant difference between the United States and British forces. The United States benefited from the opportunity to conceptualize and build its forces from the start for the expeditionary offensive, whereas Britain now faced a transition from a defensively structured force to an offensive force structure. Brooke stated that the RAF was a largely static force then transitioning to mobility to support an expeditionary army. British manpower was at its limit. To create the logistical services necessary to fight a continental war on the move, two of the twelve British divisions then in the UK probably would have to be cannibalized.20

Marshall’s response to Brooke sought to go to the essence of the COS’s thinking. To Marshall “it appeared that ROUNDUP was still regarded [by the British] as a vague conception.” Did the British chiefs, he asked, “regard Mediterranean operations as the key to a successful termination of the European war?”21 Apparently, yes.

The CCS took up other topics at length including China, Southeast Asia, and the Azores. On European strategy, however, they continued intransigently to argue variations of these opposing positions through the end of Trident’s first week. Their afternoon progress meeting with Churchill and Roosevelt at the White House on Friday, May 14, was consumed by consideration of China, Burma, and India: no progress on Europe and no easing of raw feelings. The following day, the CCS did agree to direct the British and American planners to produce two new studies in consultation with each other: defeating Germany “by concentrating on the biggest possible invasion force in the UK as soon as possible” and, alternatively, a plan to defeat Germany that “accepts the elimination of Italy as a necessary preliminary.”22 The CCS ordered that both studies be ready for presentation to them on Monday, May 17.

On Saturday, after a short meeting, the Combined Chiefs flew south into Tidewater Virginia for a weekend at the living history museum, Colonial Williamsburg, which was then in restoration. This was a well-planned diversion that by mutual agreement included no discussion of the war. Briefly, the chiefs found collegial respite from their terrible responsibilities. However, on return to business in Washington, May 17, nothing had changed. General Brooke wrote in his diary, “Another very disappointing day. . . . Again discussing ‘Global Strategy,’ which led us nowhere.” On Tuesday, the U.S. Joint Chiefs had to admit that their planners’ paper on the European theater was not yet ready and that they still were considering the British paper. Brooke concluded that “the Americans are now taking up the attitude that we led them down the garden path,” first to North African, then to Sicily, and now would do so again.23

Washington on Wednesday, May 19, was enduring the oppressive heat that can afflict the city as early as spring. In the Federal Reserve Board of Governors Room for a second week, the CCS discussion was hotter still. This decisive day for a very contentious conference began with sharp attacks by each country’s military chiefs on the other’s position. General Marshall, who held the conference’s rotating chairmanship that day, cleared the room of all but the British and U.S. chiefs. With staff gone, the dialogue continued to be frank and tough. Now “off the record,” however, the chiefs on both sides rediscovered their capability to be flexible.24 By 6:00 p.m., the Combined Chiefs had reached an agreement.

Agreed was that the Allies would assemble and prepare twenty-nine divisions for the invasion of northwestern Europe through France. The initial assault would begin with the landing of nine divisions on the target date of May 1, 1944. In the Mediterranean theater, General Eisenhower would be directed to capitalize on the conquest of Sicily with continued Allied assaults, at times and places that Eisenhower would determine, as pressure intended to eliminate Italy from the war. The now definitive Anglo-American commitment to an assault into northwestern Europe on the specified date and the direction to Eisenhower to follow up on Sicily with further action were put to paper.25 Not written down was an oral quid pro quo agreement that an Italian campaign would continue in exchange for this, the first of ultimately three British commitments to a cross-Channel assault in 1944.26

General Brooke considered the unwritten acceptance of continued operations against Italy to be a “triumph as Americans wanted to close down all operation in the Med after the capture of Sicily.” Brooke found the compromise to be “not altogether a satisfactory one, but far better than a break up of the conference!”27 From his perspective, General Marshall foresaw being dragged along by the British in a campaign up the Italian peninsula, at least as far as Rome, as a price he would have to pay for British concessions to the American preference for a cross-Channel assault strategy.

The JCS had succeeded at the Trident Conference in committing General Brooke, in writing, to more than just invading northwestern Europe via a cross-Channel assault on May 1, 1944, now designated Overlord. They had won British commitment to the transfer of seven combat-experienced Allied divisions, starting November 1, 1943, from the Mediterranean to the UK for Overlord, a commitment that would be tested again and again.28 As this day of hard-won if fragile military agreement ended, Brooke noted in his diary, a thunderstorm broke the heat.29

Over the following three days the Combined Chiefs turned their attention to the Pacific and to global shipping. Only to a degree were the issues less contentious than those of the European theater. Thanks to accelerated shipbuilding and recent success against the U-boats, the forecast for shipping was adequate to the CCS’s global plans.

On May 24, they turned to drafting the culminating Final Agreed Summary of Conclusions.30 All the national and service-favored arguments then rose back to the surface. Fortunately, if with difficulty, the Chiefs held to the agreements of the preceding days, notably those on European strategy. These already had been briefed in draft to the president and prime minister and accepted by both leaders on May 19. At 4:45 p.m., the Combined Chiefs traveled the short distance to the White House for photographs and a final meeting with Roosevelt and Churchill.

There, the torturously negotiated agreement crashed against the suddenly contrary position of Winston Churchill. Brooke recorded in his diary for May 24 that “the PM entirely repudiated the paper we had passed, agreed to, and had been congratulated on at our last meeting!! . . . Now we are threatened by a redraft by him and more difficulties tomorrow!” Reflecting on this White House meeting in 1957, Brooke recalled: “Some of the alterations [Churchill] wished to make were on points we had been forced to concede to the Americans in order to secure more important ones. From the attitude he took up the American Chiefs might well have believed we had gone behind their backs in an attempt to obtain those points through Winston.”31

*

Indeed, fairly or not, the U.S. Joint Chiefs arrived at the conclusion of the Trident Conference prepared to put little faith in the commitments of their British counterparts. However, viewing Trident through an aperture widened to include concurrent discussions at the White House, just six blocks from the chiefs’ negotiation, suggests a reason for Churchill’s sudden change of mind on May 24. As at Casablanca, there were two overarching issues under independent but parallel discussion during Trident. Less obvious than Allied military strategy, but vital to Churchill, was the issue of resuming Anglo-American atomic cooperation. This had become more urgent.

Churchill had left Casablanca in January believing that his raising of the atomic cooperation issue with Harry Hopkins had gained a promise of redress from FDR. He thought he had oral assurance, indirectly through Hopkins, that the matter would be handled “entirely in accordance with our wishes.”32 Instead, since January, the gap between the U.S. and British atomic programs had widened and acquired new substance.

Churchill went to Trident in May with the U.S. interruption of Anglo-American interchange of atomic information very much on his mind. The months since the conclusion of the Symbol Conference and the start of Trident had seen frequent high-level but fruitless communication between Washington and London on the issue of atomic information exchange. That impasse now combined with a sharpened British understanding of their own limited options to increase the level and urgency of Churchill’s concern.

Churchill’s confidence in the viability of an atomic bomb had been reinforced on April 7 by a Tube Alloys progress report from Lord Cherwell. Of the easier, more certain path to a bomb, using “light uranium” [U235], Cherwell described a bomb with 10 to 40 pounds of nuclear material “equivalent to 10,000 to 40,000 tons of T.N.T.” Although he himself would not go quite so far, Cherwell wrote, “The experts are prepared to bet 100 to 1 on its success.”33

At Churchill’s request, on April 15, the Tube Alloys leaders had assessed anew their prospects for developing on their own a British atomic bomb.34 Their conclusion was that the design and construction of a British bomb was unlikely to be completed in less than four years. The scientists believed that would be after the current war would end. They concluded that a postwar bomb would be possible with immediate construction of facilities in Britain and Canada. However, this could be done only at a staggeringly prohibitive cost in resources already critically needed by Britain for other uses to win the current war, a war that in May 1943 had only just tipped toward eventual Allied victory.

Beyond the financial expenditure, separate development of a British atomic bomb would require 20,000 laborers, a quarter possessing already scarce skills; 500,000 tons of steel already in heavy wartime demand; and an additional half million kilowatts of electricity. Parallel but separate British and American efforts to develop the bomb were foreseen by the British only to exacerbate the challenge of postwar atomic arms control already assessed to be difficult. The British scientists’ recommendation to Churchill, which he took with him to Washington as a task, was to redouble efforts to restore full Anglo-American interchange of atomic research information.35

When on April 29 Sir John Anderson forwarded to Churchill the assessment made in response to the prime minister’s request, he introduced a concern new to the dialogue, one certain to arouse Churchill. Anderson wrote: “We must always remember that the Russians, who are peculiarly well equipped scientifically for this kind of development, may well be working on the Tube Alloys project somewhere beyond the Urals and making great progress. It is incumbent upon us to make every possible effort to bring about an effective co-operation between the United States and ourselves.” At the top of the first page of Anderson’s cover note, Churchill wrote and then circled Trident.36

While at the Trident Conference, and well aware that his military chiefs and the Americans were in hot debate on strategy, Churchill received a cable on May 15 from Anderson. The encrypted cable made the task of quickly restoring information exchange and cooperation imperative. The United States had secured contracts that locked up the entire Canadian production of uranium and deuterium (heavy water). Anderson informed Churchill that, should the British opt for an independent program to develop an atomic bomb despite its already prohibitive requirements, they would have to do so without adequate sources for these essential ingredients.37

There was more to the information, received from Canada a day earlier, that had stimulated Anderson’s cable to Churchill, but he held that back. The prime minister would have found it an enlightening background to his attempts so far to move the Americans off their position. Dean C. J. Mackenzie in Canada had noticed a sea change in the confidence level of the American scientific leaders of S-1. He observed that Vannevar Bush and James Conant, who up until January 1943 had seemed doubtful of an atomic bomb’s practicability, had since March appeared “quite confident that Bombs will become available in reasonable quantities in time to be used in this war, and this has made all those concerned with the Policy direction of the American project absolutely firm on the question of Security.”38

Trident had been preceded by a month and a half of intensified communication between London and Washington on the suspension of atomic information interchange. Following Churchill’s lead, the British had gone through Harry Hopkins, Roosevelt’s closest adviser. On March 20, Hopkins took to Roosevelt another cable from Churchill that reminded him that the prime minister was still awaiting word from Hopkins of a resolution. At FDR’s direction, Hopkins sent the Churchill correspondence to Bush and requested a reply. Bush, who had become more confident of success in the quest for the bomb, according to C. J. Mackenzie, responded with a long defense of the U.S. “need to know” policy. Probably for the first time in his communication with Hopkins, Bush inserted a concern that he shared with his deputy, Dr. James Conant, and Manhattan Project director Brig. Gen. Leslie Groves. They suspected that the British were more interested in a postwar commercial advantage in nuclear power plants than they were in a bomb to help win the current war.39

As a result, no significant action on Churchill’s request advanced in Washington. That prompted Churchill to renew his request through Hopkins with another cable on April 1 that alluded to higher stakes. Churchill wrote that should the United States and Britain take separate paths to develop an atomic bomb, that would be a “somber decision.”40

Receiving no reply from Washington, there followed on April 13 a message to Hopkins from Foreign Secretary Anthony Eden. Eden’s query repeated Churchill’s hint at a possible parting of the ways: “Have you any news for me about the very secret matter we discussed? You will realize that we have various decisions to make if there has to be a separate development.”41

Hopkins did not receive Eden’s message until after he had met on April 13 with the British ambassador, Lord Halifax, at the embassy in Washington.42 There, according to Halifax’s account of the meeting in a letter to Eden, dated April 14, Hopkins “opened up on the secret scientific matter that [Eden] had spoken to him about.” Hopkins expressed to Halifax “very little doubt” that both the U.S. and British sides had “a tendency to hold back information from one another, since this scientific research was necessarily largely in the hands of persons who had been and would be again in the employ of big business, and therefore had their eye on postwar interests on either side.”

Hopkins’s understanding of the substance of the issue would evolve in the month to come. However, his approach to its resolution endured. As he expressed that to Halifax April 13, the ambassador wrote, Hopkins thought the only path to resolution was for “the President and Prime Minister to agree that the pooling of knowledge system should not be confined to the war, but to carry over into post-war times.” Halifax told Eden that Hopkins considered the issues involved too big to be decided at a level below Churchill and Roosevelt and that Hopkins “foresaw considerable possibility of trouble” if not resolved. Halifax recounted that Hopkins said he “had been having much talk with Vannevar Bush” and that “Winston was pressing him on the matter but that he had nothing definite to say at the present time.”43

The following day, Hopkins replied to Eden, via a message sent for him by Halifax in British channels: “Your message regarding secret matter received. I am going to send you on Monday [April 19] a full telegram about the matter. On further enquiry I find it has many ramifications and I therefore am anxious to send you my views fully.” Eden never received the later telegram from Hopkins, if ever one was sent.44

Harry Hopkins probably was not fully conversant in April 1943 on the specifics of the position of the British and that of the American scientists and General Groves. He knew enough, though. Hopkins’s response to Eden, through Halifax, directly communicated the position that atomic information interchange could be restored only through an agreement at the Roosevelt-Churchill level, and in doing so Hopkins signaled to the British the possibility that the agreement could take into account postwar considerations of the parties.

That this discussion did not progress further through the channel to Anthony Eden, or at all before the British arrived in Washington for Trident, suggests that Hopkins was acting with instructions from Roosevelt. As important as resumption of access to atomic information was to Churchill, there remained an open issue at least as important to the U.S. side, Churchill’s assent to a cross-Channel attack as the basis of the principal strategy for the victory campaign from the west in Europe.

As the Combined Chiefs worked at the Federal Reserve Building on Monday, May 24, to consolidate their many and varied conclusions into an agreed statement of Allied global military strategy, the open issue of atomic information sharing at last was under direct, high-level U.S.-British discussion at the White House. Churchill and his science adviser, Lord Cherwell, met for two and a half hours over lunch at the White House with Roosevelt, Vice President Henry Wallace (a member of the Military Policy Committee), and Harry Hopkins.45

Churchill was aware of the direction in which the CCS military strategy negotiations were proceeding, both from the Combined Chiefs’ briefing to him and FDR on May 19, and from daily direct reports from his own COS Committee. The prime minister knew that the military conclusions from those difficult talks were to be presented that afternoon to the two leaders for their endorsement.

Churchill put to the Americans, over lunch, the still-open issue of U.S. constraints on atomic research cooperation with British scientists. According to Hopkins’s summons to Vannevar Bush the next day, Churchill “formally raised the question of interchange on S-1.”46 Answers to Churchill’s satisfaction remained pending.

Two hours after the May 24 lunch, the CCS arrived at the White House and presented the final results of their collaboration. At this point, Churchill stunned the chiefs, British and American, by repudiating in some detail the agreements on Allied strategy that he had accepted in draft on May 19, only five days earlier.

Most interpretations of Churchill’s reversal of support for much of the chiefs’ compromise outcome, then and since, have been that he was reacting to the absence in the document of an explicit, written commitment to invade Italy and advance British objectives in the Mediterranean.47 That certainly was a central concern for the prime minister but not the only one in the moment.

A further explanation suggests itself for consideration, one not then apparent to the military chiefs, but given the separate White House discussion of joint atomic research at the lunch just concluded, likely very evident to Roosevelt and Hopkins. Was Churchill signaling resumption of full exchange of atomic information as the quid pro quo for his support of a cross-Channel attack in 1944?

The evidence for this is circumstantial, based on the ensuing meetings on May 24–26, 1943, and what the official histories have concluded from what is known about the context of these meetings. The meetings’ sequence is enlightening, however, and merits attention.

Following his outright dissent from the agreed strategy findings that afternoon, Churchill worked through the night of May 24, attempting to reword the Trident agreement to include a definite written commitment to invade Italy, while the British Chiefs of Staff continued attempts to persuade him to drop his main objections.48 During the evening, Harry Hopkins approached Churchill and said, “If you wish to carry your point you will have to stay here another week . . . even then there is no certainty.”49 General Brooke later credited Hopkins’s intervention with getting Churchill to scale back his changes, which in the end did not alter the principles of the written strategy agreement.50

Surely there was more to the Churchill-Hopkins conversation. Building on a month and a half of long-distance entreaties, culminating in their lunch meeting with Roosevelt that day, did Churchill raise again with Hopkins the open issue of resumption of atomic information interchange? Perhaps in the context of his assent to a cross-Channel attack?

What is known is that the following day, on short notice to Vannevar Bush, Hopkins convened a meeting between Lord Cherwell, Bush, and himself to address the issues in atomic information interchange. At some time between their lunch on May 24 and this May 25 meeting, Hopkins would have told Churchill that he would do this.

On May 25, with the last-minute issues of the Trident strategy document temporarily settled, but with no one confident of their durability, the military principals of the Trident Conference prepared for their departure from Washington. That day at the White House, as Hopkins had promised to Churchill, Bush and Hopkins met with Cherwell to seek an agreement to end the impasse on Anglo-American atomic information interchange.51

The meeting was frank, at times blunt. Bush reviewed for Cherwell and Hopkins the history of the U.S. position and defended the “need to know” principle for access to classified information as being applied to U.S. government and U.S. private sector participants in the Manhattan Project as well as to British participants. That proved to be too narrow an explanation to accommodate the larger issue. All of the participants knew that the operative justification being applied by the U.S. side was “need to know” in order to win this war. Creation of a British atomic capability for after the current war was not an access criterion acceptable to U.S. leaders for the project, unless they were so directed by the president. Asked by Bush whether he disagreed with this principle or its application, Lord Cherwell replied that he “disagreed with the principle itself.”52 After extended discussion of interest in commercial applications of atomic energy, of which the United States suspected the British, Cherwell stated to Bush and Hopkins that the United Kingdom wanted the data from the current research not for commercial purposes but to be able to build its own atomic bomb after the war.

Recounting the meeting, Bush wrote that in stating his understanding of what Lord Cherwell had said, “Mr. Hopkins reiterated it and emphasized it, that the reason the British wish [for] the information was so that in the period immediately after this war they would be able to develop the weapon for themselves very promptly and not after a considerable interval.”53 Bush and Hopkins were in agreement that taking the purpose of Anglo-American cooperation beyond the needs of winning the current war to address a British political-military interest in the postwar world was a matter that would have to be settled at the highest level by Roosevelt and Churchill.

Ending the meeting, Hopkins cautioned Lord Cherwell on the constitutional limits of the U.S. president’s war powers. Roosevelt could authorize sharing with the British through an executive agreement that, unless continued by whoever succeeded him as president, would be binding only on his own administration, Hopkins told Cherwell. Under the Constitution, Hopkins continued, FDR could not make international commitments binding on future presidents except by means of a treaty that would have to be ratified by the Senate.54 The compelling need to preserve the secrecy of the Manhattan Project put a treaty as the vehicle for a solution out of the question.

The meeting clarified and sharpened Hopkins’ understanding of the issue with the British on atomic research sharing. Asked by Bush if Hopkins wanted him to discuss the meeting with the other members of the Military Policy Committee that advised FDR on S-1, Hopkins told Bush, “No, nothing more to be done at this time.” Expecting Hopkins would brief FDR, Bush told Hopkins that he would “sit tight and do nothing unless and until I heard from him further on the matter.”55

Bush developed a belief, in which he probably was wrong, that Hopkins did not brief FDR on the meeting. Later communications to Hopkins from Cherwell and Churchill suggest that Hopkins briefed FDR within a day.56 With the two countries’ positions on atomic information sharing clarified but unreconciled, military strategy for Europe remained at issue at the highest level into the night.

Churchill’s commitment to his view of the path to victory ran deep, a path along which FDR had been unwilling to follow him this time. Pacing in his White House bedroom on May 25, before proceeding to a night meeting with FDR, Churchill said in exasperation to his physician, Lord Moran, “The President is not willing to put pressure on Marshall. He is not in favor of landing in Italy. It is most discouraging. I only crossed the Atlantic for this purpose. I cannot let the matter rest where it is.”57 And he did not.

Alone in the White House in their meeting that ran into the early hours of May 26, Churchill and Roosevelt struggled to draft language for a cable to Stalin to tell him the results of the conference. As they did so, they discussed the main outcomes of Trident. Churchill received Roosevelt’s oral agreement that Anglo-American full interchange of information on atomic bomb research should resume.58 Although the absence of a written record of this specific meeting leaves no evidence either way, FDR’s assurance possibly was understood to be in exchange for Churchill accepting the CCS compromise position at Trident, which included the commitment to Overlord for May 1, 1944.

In his May 26 message to Sir John Anderson, Churchill informed the Tube Alloys director of FDR’s agreement to resume sharing atomic information without stating that FDR’s assent was oral. Without referring to the limits of FDR’s war powers, and with a hint of tentativeness, Churchill told Anderson that he “understood that his ruling would be based upon the fact that this weapon may well be developed in time for the present war and that it thus falls within the general agreement covering the inter-change of research and invention secrets.”59

Churchill’s interpretation of the basis for FDR’s oral agreement in his cable to Anderson does not square with the assessment Hopkins gained from Lord Cherwell and reviewed with Cherwell at the time for clarity. This was that with restored access to information the British foresaw achieving their own atomic bomb “very promptly” after the war. That assessment is likely what Hopkins told FDR at some time in the interval between the conclusion of Hopkins’s meeting with Cherwell and the president’s late night session with Churchill, May 25–26, when FDR orally gave Churchill some assurance on atomic information interchange.60 There was advantage for both leaders in a loose assumption about when a joint Anglo-American endeavor might produce an atomic bomb. In any case, Churchill’s cable and Cherwell’s earlier thank-you to Hopkins, written May 30 after Cherwell returned to London, credit Hopkins with the result of FDR’s oral agreement.

*

Trident, the contentious third U.S.-British conference in Washington, officially had closed. But few if any on either side considered debate of its issues to have ended. Reluctantly, the British Chiefs of Staff had agreed at Trident to an Allied strategy that gave priority to an assault across the English Channel. Set for May 1, 1944, and now called Overlord, the goal of the strategy was to defeat Hitler by direct attack through northwestern Europe and into Germany. Overlord was to take priority over operations in the Mediterranean championed by Churchill, who, to the consternation of the British as well as U.S. Joint Chiefs of Staff, almost had upset the agreement in the last days of the conference. Few American participants in Trident believed that Churchill’s new commitment to cross-Channel attack would hold. To the Americans, proof of that came quickly in a de facto addendum to Trident engineered by Churchill.

Washington’s morning calm, May 26, 1943, shattered as the pilots of Winston Churchill’s camouflage-painted Boeing 314 flying boat, Bristol, revved one after another of the plane’s reciprocating engines. Then, all four engines at deep bass full throttle, the pilots set Bristol racing across the surface of the Potomac River. Bristol lifted through a steady rain and banked into the mist northeastward en route to Algiers via Newfoundland and Gibraltar.61

Along on the flight with the prime minister and Gen. Sir Alan Brooke were two last-minute, reluctant American passengers: Army Chief of Staff Gen. George C. Marshall, at Churchill’s insistent request to President Roosevelt, and Marshall’s aide, Lt. Col. Frank McCarthy. Marshall had been scheduled to take three days’ leave after Trident and then depart on a long overdue trip to the Pacific with Navy commander in chief Adm. Ernest King. Only six hours before the morning flight, Marshall learned that instead he would accompany Churchill in the opposite direction. At the last minute, President Roosevelt had relented to Churchill’s request to take Marshall along to continue Allied strategy discussions at General Eisenhower’s headquarters in North Africa.62

Marshall, the principal proponent for direct attack, was certain as to Churchill’s purpose in taking him along to North Africa. The prime minister intended to continue discussion to attempt realignment of Allied strategy for the European theater, supposedly already set at Trident, to restore precedence to Churchill’s preferred Mediterranean strategy. Indeed, seeing Marshall as his strongest opponent in the Allied strategy debate, Churchill intended to do precisely that.

General Brooke, Chief of the Imperial General Staff, had a more nuanced interpretation of Marshall’s presence. He was concerned that Churchill’s last-minute attempt to reverse the CCS decisions at the Trident Conference would rekindle American suspicions. Brooke worried that Churchill’s flight with Marshall would be interpreted “as an attempt to swing Eisenhower in our direction at the expense of decisions arrived at in Washington.”63 Marshall’s presence on the trip, Brooke felt, would be offsetting.

The generals’ misgivings were not the only ones likely to be found aboard Bristol. Flying away from Washington, as Marshall privately doubted the durability of Churchill’s commitment at Trident to a cross-Channel invasion and as Brooke fretted, Churchill probably pondered whether FDR’s spoken word would be confirmed by U.S. action to restore atomic information access.

Most immediate, however, was the task of framing the words to tell Joseph Stalin that the cross-Channel invasion, key to the second front promised to the Soviet premier for 1942 and then delayed until 1943, now had been reset at Trident for 1944. The night before, FDR and Churchill, each a master of language, had struggled together to find a way to convey this difficult message to Stalin. En route to Newfoundland, Churchill tried again to draw a cohesive message from the folder of cable drafts that he and Roosevelt had produced. Soon he gave up and turned the task over to Marshall.

Taking up a small, unlined pad and a pencil, the general started fresh. Two hours into the flight, Marshall returned to Churchill with his own draft, which survives and is remarkable for its few strike-outs and erasures.64 So impressed was the prime minister that without changing a word he sent it to Roosevelt from Newfoundland with the recommendation that Marshall’s version be their joint message to Stalin, and it was.65

Grasping at any possible diversion to continue delaying an inflight discussion of strategy for the European theater, Marshall drew Churchill into recounting the effects on Parliament of the late-eighteenth-century impeachment trial of the governor-general of India, Warren Hastings; then the 1941 landing in Scotland of Hitler’s deputy führer, Rudolf Hess; and then Churchill’s role in the crisis precipitated by the 1936 abdication of King Edward VIII. Each time, apparently unsuspecting of Marshall’s tactic, Churchill responded with extended fascinating remarks. By the time the steward called them to dinner, Marshall had succeeded in deflecting for that day any discussion of Allied strategy.66

Arriving with Churchill at Eisenhower’s headquarters in Algiers, Marshall sat down with the American Supreme Allied Commander for North Africa and his mostly British deputies, flush with the Allies’ fresh victory in Tunisia and its huge haul of POWs, for three days of conference with Churchill. In this reinforced company, Churchill pressed Marshall with his case for invading the Italian mainland.

Churchill’s private physician, Lord Moran, three months later heard directly from Marshall and wrote down the general’s account of how that went. To Moran’s recollection, Marshall told him: “I did not think that the moment had come for a decision. It would be better, I said to the Prime Minister, to decide what to do when the attack on Sicily was well under way. I wanted to know whether Germany meant to put up a stiff resistance on southern Italy or whether she would decide to retire to the Po [River] as Winston suggested. I wanted more facts.” To Churchill, Marshall seemed focused on details that the prime minister dismissed in favor of his own broader vision. Marshall told Moran, “I tried to argue that we must exercise great discretion in choosing what to do after the conquest of Sicily. I said to the Prime Minister that I would be content if Sardinia were taken before the invasion of France. He replied that the difference between taking southern Italy and Sardinia was the difference between a glorious campaign and a mere convenience.”67

According to Lord Moran, Churchill returned to London happy in his belief that he had won his point.68 That was an illusion. Marshall had decided not to continue the argument. On returning to Washington on June 7, Marshall reported that he believed he had protected the Trident decisions without damaging his good relationship with Churchill.69

In closer alignment among U.S. and British strategists was their contemplation of the Soviet Union. During their Trident meetings, their concern for keeping the Soviet Union in the war against Germany still was strong. Yet emerging was a parallel concern that would grow as the year progressed: the postwar political-military position of the Soviets on the ground in Europe relative to that of the western Allies.70 As the Combined Chiefs of Staff met, unbeknownst to them or Churchill, Roosevelt had initiated his own mission to engage Stalin.