If we are not prepared to accept the risks, face the difficulties, suffer the casualties, then let us concentrate at once exclusively on the production of heavy bombers and think in terms of 1950.
—Lord Beaverbrook to Harry Hopkins, June 1943
In London, on July 22, the day Stimson and Churchill conversed at No. 10 Downing Street on atomic bomb cooperation and war strategy, two blocks away in the bunkered Cabinet War Rooms in Great George Street, Lieutenant General Morgan, COSSAC, met with the British chiefs. Still awaiting the chiefs’ endorsement of his team’s plan, which had such critical relevance to the imminent conference in Quebec, COSSAC now got further instruction from Churchill’s military leaders.
Having received, on July 15, COSSAC’s final draft for the Overlord Outline Plan, the British Chiefs of Staff, four days later, directed Morgan to assign a small COSSAC team to accompany and support the British Quebec Conference delegation. Morgan responded later that day by asking permission to send a similar team to Washington to enlighten the U.S. Joint Chiefs of Staff as well.1 Now, attending the July 22 meeting of the war cabinet COS Committee, Morgan and his deputy, Barker, received their answer. Permission withheld.
Chief of the Imperial General Staff, Gen. Sir Alan Brooke told Morgan that the COS had not yet examined the Overlord report. In the presence of U.S. Army theater commander General Devers, General Brooke told Morgan that “no discussion should take place in Washington on plans for OPERATION OVERLORD and no [COSSAC] Staff Mission should proceed to America until the plan had been examined by the Committee.”2
During this meeting, in addition to the constraint on sharing its Overlord plan with Washington, COSSAC also received a telling query. The COS wanted to be told more about COSSAC’s planning for its other task, Operation Rankin, the contingency for responding to an internal collapse of the German Reich.3 Rankin was preferred by Churchill and his chiefs as an alternative to Overlord’s direct assault at the time and place determined by Allies. However, Rankin could be implemented only in reaction to developments that might be influenced but not controlled by the Allies.
From their representatives in Washington, but not recorded as being discussed in the London meeting, the British chiefs knew about the dissent percolating among the American planners. Should that dissent prevail in Washington, the chiefs surmised, priority in Allied strategy would shift to the British-preferred Mediterranean strategy and cross-Channel assault would be subordinated and recast into a Rankin-like operation.
Denial of permission to send a briefing team to Washington thrust the London community of planners for Overlord into turmoil. By COSSAC’s charter, the planning group reported to the Combined Chiefs of Staff in Washington through the COS Committee in London and received their orders through the British chiefs. If the Combined Chiefs “were in fact combined,” Morgan later reasoned, denying COSSAC permission to brief the U.S. Joint Chiefs of Staff on the Overlord Outline Plan in advance of the Quadrant Conference was within the authority of the British chiefs.4 Reaction against the COS order immediately began to build among the American officers in London and within COSSAC’s primarily Anglo-American staff.
*
Stimson was flying his long, dog-leg course to meet with General Eisenhower in Algiers when Allied sparring over strategy sharpened at transatlantic distance between London and Washington. On Saturday, July 24, two days after withholding for the moment permission for COSSAC to share its draft Overlord plan with Washington, the British Chiefs of Staff leapt to take advantage of Marshall’s expressed, though constrained, interest in an invasion of the Italian mainland. The COS responded to the Combined Chiefs of Staff with a message of strong support for invading Italy. Without waiting for Washington’s reaction, they acted to give their position substance. British senior commanders in the Mediterranean were sent an order to “stop own movement of forces.” From London the COS suggested that the United States do the same. This triggered a sharp reaction from the U.S. Joint Chiefs of Staff.5 Even if meant as support for a new move into Italy, the order also could be interpreted as an act in opposition to Overlord. So it was seen in Washington.
In acting unilaterally to stop redeployment of their forces from the Mediterranean to Britain for Overlord, the British Chiefs of Staff had laid bare again the deepest anxieties of the U.S. Joint Chiefs. The JCS put no confidence in Churchill’s acceptance, at the Trident Conference only two months earlier, of an Allied cross-Channel invasion of northwestern Europe in 1944. The JCS saw at best very fragile support among the COS for Overlord’s direct assault, versus waiting for the opportunistic conditions for Rankin. Nagging at the JCS was their counterparts’ demonstrated willingness to go along with strategy proposals from Churchill in which, they believed, the British chiefs themselves sometimes put no faith.
From experience, the JCS and, as Marshall knew, Stimson and Hopkins were deeply worried about Churchill’s ability to turn their president’s position away from what they saw as the best strategic interests of the United States. Now they knew for certain that the prime minister intended to come early to meet with Roosevelt before Quadrant. Perhaps his greatest concern, Marshall also knew that with Quadrant less than a month away, the U.S. strategy for Europe and for the conference was not firm; U.S. military planners, themselves, were sliding into a crisis of doubt about cross-Channel attack.
Events in the Mediterranean advanced independently from the Allies’ control. On Sunday, July 25, the Italian dictator Benito Mussolini was ousted. On orders from King Victor Emmanuel III, Mussolini was arrested and replaced by Field Marshal Pietro Badoglio.6
As news broke in Washington and London of Mussolini’s arrest and formation of the new Italian government, General Marshall traveled by car seventy-five miles to Shangri-La, the president’s retreat in Maryland’s Catoctin Mountains. Anticipating the imminent conference in Quebec, Marshall intended to impress on FDR the Joint Chiefs’ objections to the British strategy. In his notes for the meeting, Marshall’s objective was unequivocal and emphatic. Typed with underlining, that was: “To prevent, prior to the Quadrant conference, the making of commitments, verbal or written, affecting operations now under way or future operations.”7
Basing his case directly on the COS “stop own movement” order and their message to the JCS of the previous day, Marshall argued that the British were spring-loaded to exploit for their own objectives any statement by the president or U.S. military leaders. He told FDR of the JCS’s continuing “great difficulties” in determining “whether the proposals [the British chiefs] advance are those of the prime minister or are those of his military advisors. Are the ideas based on sound operational and logistics considerations from a military viewpoint or are they accepted by the COS (in some cases possibly against their better judgment) only because the Prime Minister wants them?” Marshall warned FDR, “Proposals contained in conversations or communications between the President and the Prime Minister prior to the conference will, if accepted by the latter, definitely tie the COS and thus interfere with what should be a calm and reasoned consideration of the issues involved.”8
Fixed on the goal that out of Quadrant “must come a really firm decision as to whether or not we will cross the Channel against determined resistance” (emphasis in the original), Marshall appealed to FDR. If Roosevelt found it necessary to reply to any proposal from Churchill before the conference, Marshall implored, the president first should absolutely obtain the advice of his Joint Chiefs of Staff.9
Marshall’s meeting with Roosevelt had been held at Shangri-La and away from White House bustle, but amid distractions from the news of Mussolini’s overthrow. Marshall left not confident that his message had registered with the president. Context compounded his disappointment. In addition to the British “stop own movement” gambit, Marshall knew that, among his own planners, support of Overlord was wavering. Within the Pentagon, eight days earlier while Stimson and Churchill toured the Dover Cliffs area, the initial foray had begun into what the official U.S. Army history later would title the U.S. staff “Revolt Against Overlord.”10
Allied troops deployed in the European theater, as of the summer of 1943, were split and neither the concentration of forces in the United Kingdom nor in the Mediterranean was sufficient in itself to implement a strategy for total defeat of the Axis in Europe.11 Some U.S. strategy planners had become hesitant about reaffirming cross-Channel attack as the principal Allied strategy and were becoming persuaded that the successful landing in Sicily pointed to the Mediterranean-focused strategy as the bird in hand. The advocates for the cross-Channel attack-based Overlord strategy opposed the British Mediterranean strategy as too lengthy, too susceptible to still more diversions to the east, and too reliant on the hope of inducing defeat of Germany by political-economic collapse.12
The dissenters from Overlord, however, had a respected champion. Brig. Gen. John E. Hull was the chief of the Army’s Operations Division (OPD) theater group and one of the early authors of the cross-Channel strategy. On July 17, Hull had stated his new view that “we should now reverse our decision and pour our resources into exploitation of our Mediterranean operations.”13 General Hull’s was an opinion that could not be ignored.
As the Army chief of staff made his apparently fruitless round-trip to Shangri-La that Sunday, July 25, the seed of a dissent from Marshall’s strategy that Hull had planted on July 17 was taking root. The senior Army and Army Air Force members of the Joint War Plans Committee (JWPC), Cols. William W. Bessell and Richard C. Lindsay, issued a paper attacking the Trident priority for the cross-Channel strategy. Bessell and Lindsay contrasted Bolero’s goals to its limited accomplishment in building a U.S. force in the United Kingdom and compared that to the size of the Allied force in being then in the Mediterranean. They recommended that the U.S. acknowledge that the Soviet Union had been, and likely would continue to be, the main source of ground pressure against Germany and, given that, continue with Britain to bomb Germany from the air.14
The full JWPC acted to formally question the cross-Channel strategy in a report to the Joint Planning Staff. Building on Bessell and Lindsay’s argument, the committee recommended that future Allied operations emphasize the Mediterranean and landings in southern France. They proposed that the cross-Channel strategy become the “final” action, not “the opening wedge for decisive defeat of the German armies.” The JWPC alternative recommended leaving in place the seven combat-experienced divisions slated for transfer from the Mediterranean to the UK. Instead, they recommended that the shipping thus released should be used to bring seven fresh divisions from the United States to Europe.15
Monday morning, July 26, Bessell and Lindsay gained support for their position from the senior Navy member of the Joint Planning Staff, Rear Adm. Charles M. “Savvy” Cooke Jr. Cooke believed that Overlord, burdened by many operational and logistical challenges, should be subordinated to more certain operations in the Mediterranean and particularly in the Pacific.16 In its sum, the Joint Planning Staff position advocated reversing the Trident decisions and accepting the British position, championed by Churchill and supported by his senior commanders, along with strengthened advocacy for U.S. Pacific operations.
As the U.S. Joint Planning Staff made its recommendations, challenging the position preferred by General Marshall and the other Joint Chiefs, Anglo-American skirmishing began for the control of the agenda for the Quadrant Conference. Typical of all international conferences is that this minuet of process-setting can steer the substantive outcome. While moving to check the staff crisis of confidence, the U.S. Joint Chiefs also began to reveal their hard-learned better footwork to those masters of the dance in London.
The British Chiefs of Staff submitted to the U.S. Joint Chiefs their “Proposed Agenda for Quadrant.” Paragraph One inferred that the purpose of Quadrant would be to examine the Trident decisions “in light of the situation existing in [July–August 1943].” Paragraph Two was “Strategy in the Mediterranean.” Paragraph Three was “Defeat of Japan.” Overlord and Bolero-Sickle only appeared as subheadings in Paragraph Four, “Operations from the UK”17
Within a day, the U.S. Joint Chiefs of Staff responded by submitting to the CCS the U.S. Proposed Agenda for Quadrant (CCS 288/1). The JCS intended to come to grips first with the Allies’ strategic goals that in turn would define objectives for operations. Their intent was to not allow the COS an opening to employ their tactic of flooding the discussion with comparative details so as to avoid commitment to larger strategy decisions. In order by its roman numerals, the U.S. proposal was: I Conference procedure; II Consideration of Overall Objective, Overall Strategic Concept, and Basic Undertakings in support of Overall Strategic Concept; III Consideration of specific operations for 1943–44 in the European-Mediterranean area with (1) Bolero-Overlord and (2) Sickle-Pointblank leading the list. Antisubmarine warfare, critical to success of transporting across the Atlantic of the massive force required for Overlord, was listed third.18
The British chiefs’ rejoinder to the JCS came from London to the CCS in Washington on July 29, as their Proposed Agenda for Quadrant (CCS 288/2). The COS expressed their hope “that it may be possible to confine the Agenda as far as possible to those specific issues on which decisions are required to govern operations in the comparatively near future.” They asked that the U.S. Chiefs of Staff “dispense with lengthy discussions on over-all strategic concepts or global strategy.” They requested that instead of exchanging more cables, the agenda be settled at the first meeting of the conference, and again they asked that the U.S. chiefs join the conference by August 10 or as soon thereafter as possible.19
Confident that their preferred agenda was firmly lodged, assigning priority to Overlord, the U.S. Chiefs had no intention of sitting down with the British in Quebec before the U.S. positions were embraced in their full development by a unified U.S. team that included the president. Marshall, with the full support of the JCS, was determined that this would be a team intent on winning.
Unfortunately, as the British chiefs knew, the U.S. team was not unified. The U.S. planners below the level of the Joint Chiefs were debating the relative merit of the Mediterranean strategy and cross-Channel attack, FDR’s position remained ambiguous and, however delayed and curtailed in duration, a pre-Quadrant meeting between Roosevelt and Churchill could not be prevented by the U.S. chiefs.
In Washington, the absence of a viable Overlord plan—COSSAC’s plan held in suspense in London—made both the British chiefs’ and now the internal American challenges to the case for Overlord difficult to refute. London had gained knowledge from side conversations among CCS representative that, at staff-level in Washington, the internal cracks in U.S. support for giving priority to cross-Channel assault appeared to be expanding.
*
At Norfolk House, disturbed though COSSAC was by the chiefs’ stricture against sharing with Washington their draft Overlord plan, COSSAC’s mandatory, immediate task was clear: move ahead urgently on planning for Rankin in order to answer the COS question. Their Rankin planning had fallen well behind Overlord for a combination of reasons. Overlord was tangible, a defined opening phase starting point just across the Channel. The COSSAC staff had become united and stimulated by: the very challenges of Overlord; anticipation of the enormity of the still-pent-up but inevitable surge of forces to concentrate for a defined date for action, May 1, 1944; and their own shared desire to take the fight back to the enemy on the Continent.
In contrast, Rankin was a fog of “what if’s.” Practical operational responses to Rankin’s three contingency cases were difficult to conceptualize. Each case was intended to react to a situation of “disintegration” determined not by the Allies but by the circumstances of unpredictable dissolution within enemy forces and leadership. Earlier, on the third of May, Lieutenant General Morgan had pointed out to the Joint Planning Staff in London that there was no official definition of “disintegration.” He appealed to the Chiefs of Staff Committee to “lay down objectives . . . to enable him to decide what would constitute disintegration.”20 Set against the context of RAF bombing of Germany since 1941 with the intent of causing political-economic collapse, the absence in 1943 of a definition for disintegration, at a minimum, suggested a strategy unmoored. Compounding this was awareness that the Allied resources to respond to any Rankin option were assumed to come not from a planned buildup against a specified date, but from what forces would be at hand in the United Kingdom for action who knew when?
Accepting COSSAC’S “Overlord Outline Plan” for their review in the July 22 meeting, the British chiefs had asked COSSAC for a Rankin plan that could be made operational on August 1, nine days hence,21 and that “if undertaken in 1943, should be planned on the basis that there would be no opposition to our landing.”22 The COS wanted a Rankin plan, operational and ready to be initiated, in their quiver when they arrived at Quadrant as an opportunistic alternative to Overlord.
Despite the British chiefs’ urgent deadline, Rankin planning advanced only haltingly. The COSSAC staff was fighting the ambiguity inherent to their Rankin task and an undercurrent of consternation that sharing of their jointly developed Overlord plan with Washington was being held up.
General Morgan saw the immediate impediment to planning for Operation Rankin to be in the wording of COSSAC’s implementing directive. In a July 27 memorandum, intended to accelerate action to craft a Rankin plan, by then due to the chiefs in five days, Morgan told his staff: “The operation that was described to me as ‘A return to the Continent in the event of German Disintegration’ [was] not in fact an operation of war.”23 He directed COSSAC’s principal staff officers to identify in the abstract military operations by type that would be required to varying degrees but in common by all the circumstances reasonable to a drawing inward by the Reich. Although abstract, identification of needed operations would provide a basis for assessing the resources necessary. Acting on the intelligence assessments he was given,24 Morgan impressed on his officers the need for urgency in developing solutions to the amorphous demands of Rankin. Acknowledging the differences in Germany’s situation in 1943 from those which brought on that country’s chaotic implosion in the previous war, Morgan concluded that “nevertheless the sum total of all the various factors operating cannot be far from that of the factors which caused the German collapse in 1918. Once again, therefore, emphasis must be on speed in the evolution of a plan.”25
Three abstract operational requirements were apparent immediately. To have a hope of averting chaos, the Allies needed to avoid being surprised by a German collapse or partial collapse, as they had been in 1918. That pointed to a need for sustained reconnaissance in strength. Allied aerial assets, signals intelligence, and the growing array of British Commandos and expected U.S. Army Rangers could meet that need. A collapse would trigger a need for rapid deployment to the Continent to fill the vacuum left by withdrawing German forces, particularly to capture strategic airfields from which a weakened Germany could be “overawed” into complete surrender. Allied airborne forces, to be transferred or returned to the UK for Overlord, could take that mission. Ports and transportation systems would need to be restored to operation rapidly to handle the surge of ground troops to complete the occupation. That would require prompt, accurate damage assessment and action by specialized engineering and salvage units. Across every circumstance would be the need for a robust Allied civil affairs capability attuned and responsive to the situation of the populous in each liberated country.26
In response, the planners developed three cases for Rankin. Case A considered a return to the Continent before the date for Overlord in the event of a substantial weakening of German capacity for resistance in France and the Low Countries. Case A was not considered feasible before the end of 1943, unless Germany was very near collapse and an adequate naval assault force was available. Case B assumed a return to the Continent if Germany withdrew from the extremities of the areas it occupied to make a defense at the Siegfried Line located east of the Rhine River. Case C responded to an unconditional surrender by the Germans mandating the need to occupy the Reich itself and other areas as rapidly as possible to enforce the surrender.27
Development of civil affairs planning and capabilities, which later would become a major effort within COSSAC, lagged across the three cases for Rankin. Civil affairs depended on the policies of the United Kingdom and the United States toward specific occupied countries. These policies, however, were still developing. COSSAC’s own history of Rankin planning, written in May 1944 in the midst of war and near the eve of the D-Day invasion, provides a glimpse into both the quandary of civil affairs and wartime emotion, not without cynicism:
That some delay should elapse before the main [Rankin] forces should follow those advance guards was inevitable but it might not be a disadvantage. There was bound to be a certain amount of “blood-letting” among the liberated peoples eager for revenge upon their quislings, and the delay in completing full occupation would both save the Anglo-American authorities from contamination and yet avoid the necessity of interfering and being compelled morally to save their enemies from very well-merited fates.28
In the event, civil affairs planning by COSSAC, SHAEF, and the Allied governments was overwhelmed at liberation by the enormous scope of Nazi barbarity, general destruction, loss, and the revenge, and sometimes-violent pursuit of pent-up political agenda that followed in Europe.29
*
As the month of July drew to its close, General Morgan recounted, “The British Chiefs of Staff had had ample opportunity, of which good use had been made, to familiarize themselves with our project in all its weak points as well as its strong particulars. In discussion the weak points naturally had bulked largest.”30 On July 27, COSSAC submitted to the COS the final draft of the Overlord plan and its appendices.
Knowledge of the nascent American staff hesitancy was spreading in London. When on July 27 Morgan asked Lt. Gen. Jacob Devers about the rumor of appointment of a U.S. Supreme Allied Commander, his friend demurred. Devers had been with Morgan in the July 22 meeting in which Gen. Sir Alan Brooke withheld permission to share the draft Overlord plan with Washington. Now Devers replied that he felt unable to answer Morgan’s question in part because he was not sure of the attitude in Washington toward COSSAC. Washington’s attention seemed to be shifting to the Mediterranean and the Pacific. Devers told Morgan he was very conscious of being at the end of a “3,000-mile-long limb.”31
Among U.S. officers in London knowledgeable of Overlord, both within and outside the planning group, vigorous protests mounted against General Brooke’s July 22 order constraining COSSAC, which they considered to be inequitable. Plainly evident to these officers was the COSSAC crash effort in response to the COS to develop a Rankin alternative to Overlord. Morgan’s American deputy, Maj. Gen. Ray Barker, knew that General Marshall and the U.S. Joint Chiefs did not want to arrive at the Quebec Conference uninformed about the substance of the Overlord Outline Plan already presented to the COS.32
Less than a fortnight earlier, Morgan and Barker had given Stimson a full, candid briefing on the essentials of the plan that the British chiefs were now holding close. The two generals had revealed their own concerns about assembling the concentration of forces required for Overlord’s success. They were acutely aware that withholding full details of the plan now would invite trouble with Stimson, who, while impressed by COSSAC’s ability and commitment, was known to hold suspect the British commitment to Overlord. Through General Devers’s remarks to Morgan on July 27, they also knew of the questioning of priority for the cross-Channel attack strategy then stirring within Washington, in part stimulated by advocates for the Pacific theater who saw their own opportunity in reacting to the British.
Certain that time was of the essence, Morgan and Barker also knew that direct appeal in London or Washington of the British chiefs’ decision only would inflame the situation. The two generals found in the details of COSSAC’s work grounds for taking action. They concluded that many key elements of the draft plan, now submitted to the COS, would have received, and should have had, closer examination and authorization from the U.S. planners in Washington had more time been available to do so. These elements affected critical U.S. long lead-time decisions on troops, logistics, and civil affairs that would have to be addressed and, they reasoned, “did not lend themselves to treatment over the trans-Atlantic telephone.”33
Morgan had been given a legitimate order that thrust upon the career soldier an exquisite dilemma. Under the Combined Chiefs of Staff chain of command, COSSAC was responsible to and reported through the COS. Before COSSAC, Morgan had commanded a corps and had expectations that his next British Army assignment would be to take up a combat command. After consultation with the other British and American advocates for Overlord in London, Morgan acted on this inequitable treatment of his country’s ally. Mindful of the bureaucratic reasons not to act, and potential consequences to his British Army career, on July 28, Morgan dispatched a team to Washington anyway, headed by Barker, with copies of COSSAC’s plan to brief to the Americans.34
Military flights from Henley, outside London, direct to Prestwick, Scotland, to catch a transatlantic flight westward were an option for the team, provided they applied for and received an air priority order.35 The Washington-bound COSSAC team, instead, chose a lower profile option for leaving London. They bought train tickets to Glasgow.
At 8:00 p.m., Wednesday, 28 July, the London Midland & Scottish Railway’s Night Mail eased out of London’s Euston Station on its express, limited-stop, overnight run north to Glasgow. Not listed on passenger timetables, the Night Mail was known informally as “the Ghost Train,” because limited passenger accommodations indeed were available. In addition to mail cars, the train included a single rider coach with three to four passenger compartments. One was used by the guard (conductor). The other compartments often were used by the king’s messengers who favored anonymous travel with speed and security.36
These attributes made “the Ghost Train” the perfect accommodation for Major General Ray Barker and three U.S officers from COSSAC who boarded that night to slip out of London on the first leg of their trip to Washington. Copies of the secret draft Overlord Outline Plan were locked in their government-issue briefcases of thick brown leather.37 Barker and his team arrived in Glasgow at 5:00 a.m., July 29. Intent on making their flight west across the North Atlantic to Washington, the team took a staff car to the air terminal at Prestwick where their progress halted abruptly. Westbound planes had stopped flying.
The weather that had barely favored Secretary of War Stimson’s homeward flight only hours earlier had closed in. A major tropical depression was advancing eastward over the North Atlantic inflicting chaos on flight operations. Flying out of southern Newfoundland already had been suspended and flying south of Iceland would be affected by July 31. Pilots in Prestwick were advised that they would have to wait two to three days for favorable flying weather westward.38
*
On another route, undeterred by weather or anything else, Anglo-American atomic negotiations also were transferring west across the Atlantic. As July moved toward its end, the next stage of talks on atomic research sharing was about to open, now in Washington. Friday, July 30, the White House received Churchill’s reply to FDR’s invitation to send “your top man” to Washington “to get full understanding from our people.”39 The cable informed FDR that Sir John Anderson would arrive in Washington by air on the coming Monday or Tuesday. Anderson would make contact with Wallace Akers, the senior British representative for Tube Alloys/S-1 in the United States, and then be ready to talk with anyone FDR wished.40
The Pentagon already had received a July 29 letter from Churchill to Henry Stimson to tell the secretary of war of FDR’s invitation and travel by Anderson in response. Churchill informed the secretary that Anderson would get in touch with Stimson in Washington on the second or third of August and enclosed for Stimson his “Draft Heads of Agreement” setting out the four points Churchill proposed in their July 22 meeting in London.41 Stimson still was en route home from North Africa and would not see Churchill’s letter until he went into the Pentagon on Sunday, August 1, the day before Anderson would arrive in Washington.
Anticipating Anderson’s arrival, opposition to an Anglo-American sharing agreement also went on paper that Friday, July 30. Dr. James Conant, Vannevar Bush’s deputy at the Office of Scientific Research and Development, sent Bush a memorandum as deputy chairman of the Military Policy Committee. Noting FDR’s directive to Bush to exchange information with the British on an “inclusive basis,” the memorandum put on record Conant’s “conviction . . . that a complete interchange with the British on the S-1 project is a mistake.” Conant authorized Bush to quote him to “higher authority if you see fit” meaning the president.42 The president, however, had plans.