OVERLORD in the Balance
When the Combined Chiefs of Staff parted at Washington in May 1943, they left with the sense that many of their decisions had been necessarily tentative and would have to be re-examined in the light of what happened in the next few months in Russia and in the Mediterranean.1 Near the end of July the Joint Strategic Survey Committee reviewed the principal developments for the guidance of the Joint Chiefs and found especially noteworthy the "substantial improvement in the submarine situation and prospects, satisfactory progress of the bombing offensive, unforeseen degree of success attained with HUSKY [the invasion of Sicily], indications of the impending collapse of Italy, decision to carry the war to the mainland of Italy . . . and apparent loss of the initiative by the Germans on the Eastern Front."2 This was actually a cautious statement of successes which in two and a half months had definitely tipped the balance of military power in favor of the Allies throughout the world.
The Quebec Conference
At the Casablanca Conference, it will be remembered, the Allies had decided that the first charge on their resources should be the defeat of the German submarine. By the summer of 1943 that defeat had all but been accomplished. Sinkings of Allied ships would continue, but on a diminishing scale, while the destruction of enemy U-boats accelerated. The turning point in the war on U-boats came dramatically in the spring of 1943.
The Allied counteroffensive began in a small way in the early months of 1943. Although sinkings continued heavy, reaching a peak in March of 141 ships lost, the Germans were also beginning to lose more heavily, and, though the Allies were then unaware of it, the German Admiralty was worried.3 Early in February 1943 Grossadmiral Karl Doenitz expressed alarm at the fact that the Allies now seemed to have exact knowledge of the disposition and in some cases of the number of German U-boats in operational areas. At the end of the month he reported to Hitler that nothing had been sunk that month because nothing had been sighted.4 He attributed the failure to bad weather and lack of reconnaissance. But it was not bad weather that destroyed
nineteen U-boats in February and fifteen in March.
For the Germans much worse was to come. In March on the suggestion of the Canadian Chief of Naval Staff, American, British, and Canadian naval commands held a conference in Washington to reorganize the Atlantic convoy system. Up to that time Atlantic convoys had been operated on a combined basis with ships of all three nations participating. Friction and inefficiency resulted. It was decided therefore to turn over naval responsibility for the North Atlantic entirely to the British and Canadians. The United States Atlantic Fleet would escort convoys between the eastern seaboard and North Africa, as well as certain special fuel convoys. The change, effective at the end of April, contributed to the conspicuous success of the Allied antisubmarine offensive beginning in May. A still greater contribution, however, was made by the long-range aircraft which the conference recommended be allocated to coastal areas for antisubmarine work. The Combined Chiefs of Staff assigned nearly all the aircraft requested (about half from British and half from American sources) and specified that all planes would be equipped with radar and operated by crews specially trained in antisubmarine warfare. Steps were being taken also to supplement land-based air patrols with carrier aircraft. During the winter four baby flattops were in process of conversion for Atlantic convoy use. The first was delivered in March. In June, three were operating in the South Atlantic.5
With a better-organized convoy system, new aircraft, and new detection methods, the Allies struck hard at the enemy's underwater fleet in May. That month German U-boat losses soared to 30 percent of all boats at sea. Admiral Doenitz told Hitler: "These losses are too high. We must conserve our strength, otherwise we will play into the hands of the enemy." But as to the cause of the disaster Doenitz could only speculate. He knew that British aircraft were using a new microwave radar and suspected that the same device was employed on Allied surface vessels. He knew that, whatever the new device was, it apparently enabled planes and destroyers to hunt out German submarines in the dark and in the fog and destroy them. He knew also that the submarines had no radar detection set that could warn them of the impending attack. In fact, he said, "We don't even know on what wave length the enemy locates us. Neither do we know whether high frequency or other location devices are being employed. Everything possible is being done to find out what it is." Hitler added a dark intuitive note. Perhaps, he suggested, the new device involved principles with which the Germans were not familiar. "The crisis must be overcome by all possible means."6
But it was not overcome. It deepened and spread. In June, German submarines chalked up only nineteen Allied sinkings; in August, only three; and in September almost all U-boats had been withdrawn from the Atlantic, with the result that the U.S. Navy could virtually abolish the convoy system in coastal waters.7
GERMAN SUBMARINE UNDER AERIAL ATTACK
The reckoning at the end of 1943 was that three million gross tons of Allied and neutral shipping had been sunk as compared with more than eight million the year before.8 But enemy success was much less than even these figures show, for during the year Allied shipyards had put fourteen million tons of new shipping on the seas as against only seven million in 1942. In other words the net loss in 1942 of about a million tons became in 1943 a net gain of almost eleven million. At the same time estimated sinkings of U-boats increased from 85 in 1942 to 237 in 1943.
The spring of 1943 marked a sudden intensification of the air war against Germany almost as dramatic as the antisubmarine campaign. After the reduction of U.S. Air Forces in the United Kingdom in the fall of 1942 to furnish air support for the North African campaign, American air power in Great Britain remained virtually static until the spring. From December 1942 through April 1943, the Eighth Air Force based in the United Kingdom had just six groups of heavy bombers operational each month. During May, the number of groups doubled; in June, thirteen groups were operational, in July, fifteen, and in August, sixteen and three-quarters.9
The build-up was reflected gradually in intensified operations. March and May each saw nine operations carried out, but in March they involved 956 sorties and the dropping of 1,662 tons of bombs. In May 1,640 sorties were flown and 2,851 tons of bombs dropped.10 In short, the average mission in March involved about 106 planes; in May it involved 180. This was still a long way from the June 1944 peak when the average operation was undertaken by 1,000 heavy bombers. Nevertheless the sudden spurt forward in the spring of 1943 was a step decisively in the direction of bringing crushing air power to bear on the enemy's war machine.
A similar increase in effort by the British Bomber Command at the same time was noted by the Chief of Air Staff, Sir Charles Portal. The bomb tonnage dropped on enemy territory during the second quarter of 1943, he reported, more than doubled the tonnage in the first quarter. Even after making allowances for weather, this achievement, together with the accelerated build-up of the Eighth Air Force, represented a notable net increase in the pressure of the combined air offensive. It was too early to see certain or decisive results, but the promise of things to come was clear. Air Marshal Portal remarked that already Germany was being forced into a defensive air strategy.11
What the Joint Strategic Survey Committee referred to as an "unforeseen degree of success attained with HUSKY" was the relatively quick victory in Sicily accompanied by signs of complete Italian demoralization. The forces of Field Marshal Harold R. Alexander (General Patton's U.S. Seventh Army and General Montgomery's British Eighth Army) invaded
Sicily on 10 July 1943; thirty-nine days later (on 17 August) organized resistance had ended. The overrunning of the island had the immediate effect of convincing the Italians that their cause was hopeless. On 5 July Mussolini was forced out. On 8 September the Italian government formed by Marshal Pietro Badoglio capitulated. The imminence of Italian collapse had been apparent to both the Allies and the Germans some weeks in advance, and both sides were hurriedly preparing for the next move.
The Germans, believing that there might still be some fight left in the Italians although their leadership had altogether broken down, in May had made plans for moving Army Group B, under Generalfeldmarschall Erwin Rommel, into Italy. The move was accomplished during July and August while a pretense of cooperation with the Badoglio government was maintained.12
The prospect of an early Italian collapse convinced the U.S. Chiefs of Staff that the time had come for bold action. At the Washington Conference in May and for some weeks subsequently the principal Mediterranean post-HUSKY operations envisaged were the seizure of Sardinia and limited-objective attacks on the toe or heel of the mainland. Both the Combined Chiefs of Staff and General Eisenhower favored the mainland operations as more likely to produce decisive results, but feasibility depended on the progress of HUSKY.13 Planning therefore proceeded concurrently for all contingencies, and decision was deferred. General Marshall thought the choice might be made toward the end of July. It was actually about the middle of the month that the Combined Chiefs proposed scrapping all their earlier conservative plans in favor of a direct assault on Naples with a view to full-scale exploitation up the peninsula.14
Before this decision was made, General Eisenhower had asked that he be given additional troops to follow up the victory of Sicily. Specifically he asked for the diversion to the Mediterranean of one convoy scheduled for the United Kingdom. The Combined Chiefs of Staff approved his request though it meant a net loss to the United Kingdom build-up for OVERLORD of 66,000 troops.15
British planners then proposed that General Eisenhower's striking power be still further increased.16 But as their proposal would have reduced the OVERLORD build-up by another 50,000 troops and might have caused postponement of either OVERLORD or operations in the Pacific, the U.S. Chiefs of Staff demurred. They pointed out that, no matter how important an action in the Mediterranean might be, it could not substitute for action required in the Pacific or northwest Europe. In advocating the assault on Naples, they said, they had intended a bold plan which involved risks, justified,
in their opinion, by the Italian collapse. They criticized the British proposal as overinsuring the operation at the cost of prejudicing agreed strategy elsewhere. They could not approve the Naples assault so conceived.17 On this point the Joint Chiefs of Staff stood fast and on 26 July the British yielded.18 The order went out to General Eisenhower to plan an attack on Naples with only those resources already allotted to his theater. These, of course, now included 66,000 more men that had been earmarked for the Mediterranean by the Washington Conference.
Pressure on Germany from the west and south by the spring of 1943 was at least sufficient to blunt Germany's offensive capabilities in these directions. But from Germany's standpoint much worse was happening in the east. In the two preceding summers the Germans had set their armies rolling across huge expanses of Russian territory, though the decisive victory on both occasions remained just beyond their grasp. In July 1943 they tried again. They massed armor south of the Russian salient at Kursk. The armor broke through only to be destroyed in great tank battles in the rear of the Russian main lines. The Russians waited only a week for this situation to develop before they launched their first major summer offensive with three armies attacking about fifty German divisions in the vicinity of Orel. On 5 August the city fell. More important, the Russians had finally and permanently seized the initiative and had begun the dogged battle of attrition in which the prize was not territory but destruction of the German armies. The seesaw campaign in which the Germans advanced in the summer and retreated in the winter was over. Henceforth the main German armies in the east would walk in only one direction-back toward the homeland.
The war was, of course, not won in August 1943, and it could not have seemed so. But at least the brightening of Allied prospects was unmistakable. The optimistic might see victory dead ahead; the most pessimistic could feel maneuver room between their backs and the wall. It is certain that when the Combined Chiefs of Staff met at Quebec for another of the great plenary conferences they faced the future with a new sure knowledge of their power. This knowledge in turn sharpened the urgency to crystallize strategy. It was more clearly than ever up to the Allies to determine when and where the ultimate decision of arms would be sought.
The Quebec Conference, which opened on 12 August, had, unlike the earlier meetings at Casablanca and Washington, no decisions to make on immediate operations in Europe. The next step in the Mediterranean, the invasion of Italy in the Naples areas, had already been decided. This fact cleared the decks for a full-dress debate on European strategy. Fundamental disagreement between the U.S. Chiefs of Staff and the British on how to defeat Germany had been adumbrated in September 1941, and thereafter had underlain all their discussions on strategy. But during the first half of 1943 the opposition of strategic principle had been blurred because the immediate and pressing issue had always been to decide
QUEBEC CONFERENCE. Seated, left to right: Prime Minister Mackenzie King of Canada, President Roosevelt, and Prime Minister Churchill. Standing: General Arnold, Air Chief Marshal Portal, General Brooke, Admiral King, Field Marshal Dill, General Marshall, Admiral Pound, and Admiral Leahy.
whether or not to carry out a specific operation in the Mediterranean. Once the Allies were committed in North Africa, to reject exploitation in the Mediterranean meant to choose idleness for a considerable Allied force and to surrender the clear military advantage of maintaining momentum against the enemy. But attacks in the Mediterranean had always been conceived as having limited, not decisive, objectives. The ultimate aim, the British had argued, was to knock Italy out of the war. With the invasion of the Italian mainland, the course of action begun by TORCH would therefore reach its goal. After the defeat of Italy, continued attacks into the Axis underbelly would require new tactical and strategic justification. With the close of the 1943 summer offensive, furthermore, the old argument in
favor of Mediterranean ventures as interim operations designed to maintain pressure on the enemy while preparations for OVERLORD were completed would no longer apply. On the contrary, for the first time since TORCH, Mediterranean operations and the cross-Channel attack became clear-cut alternatives competing for consideration as the Allied main effort.
From every standpoint the Quebec Conference seemed to present both the urgent need and an ideal opportunity to seek a firm resolution on future strategy in Europe. Aware of this, the Joint Chiefs of Staff made particularly careful preparations to present their views. They analyzed at length the procedure of previous conferences, the debating techniques of the British, and even the precise number of planners required to cope on equal terms with British staffs. Quite frankly they went to Quebec determined to make their ideas prevail by all the means at their disposal.19
This was to be, it must again be emphasized, a friendly debate between Allies seeking like same end. Nevertheless it was serious, for each Ally brought to it deep and honest convictions. It therefore became a matter of grave importance for each to find the strongest possible grounds on which to rest his case. The American ground in earlier conferences had been relatively weak. In the first place the British Chiefs of Staff had more readily available and more complete information on the European war because they were closer to the scene, because they had presided over planning for European operations almost two years before the Americans became involved, because they were served by a well-developed intelligence system for which the newly arrived Americans had no equivalent, and finally because their own planning staffs were generally more numerous and more experienced.20 These inequalities evened out only gradually as the U.S. Army grew and organized itself in the European and Mediterranean theaters.
In the second place, the British Chiefs of Staff were a much more tightly knit organization than the U.S. Joint Chiefs. This also was due to a longer familiarity with the problems of the European war, which had permitted them to iron out through long association and discussion both individual and service differences. It was more directly due, however, to the influence of the Prime Minister and the War Cabinet, who provided a large amount of daily guidance in strategic and political matters.21 Through Lt. Gen. Sir Hastings Ismay, Military Chief of Staff to the Minister of Defence, who attended most Chiefs of Staff meetings, the military and political leaders were kept in close contact with each other and welded into a relatively homogeneous directorate
of the total British war effort.22 The unity of British political and military policy was further insured by the British system of cabinet responsibility. Churchill, as Minister of Defence, together with his War Cabinet, was directly answerable to the House of Commons for military decisions. It behooved him then to keep in close touch with his military chiefs and to make certain that he fully understood and approved their decisions, even as to detail, in order that he could defend them later, if necessary. Churchill was personally inclined, furthermore, to occupy himself even more intimately with military affairs than his position required. As a long-time student of tactics he could contribute substantially at the planning level to the conduct of the war. He attended and took active part in many of the regular Chiefs of Staff Committee meetings. He intervened constantly in the planning of operations, not only to enunciate principles but to suggest tactical detail. As already noted, he played a leading role in the development of landing craft and the artificial ports. He concerned himself in the details of artillery and armored tactics.23
In all these matters he had a lively personal interest. He sent along to the Army a stream of memoranda containing all manner of suggestions, some practical, some not. But all had in common a bold enthusiastic approach designed either to shock the planners out of what Churchill always feared was a narrow professional view of the war or to cheer up his subordinates when the outlook was black.24 After Dunkerque, when dispatches of disaster were piling up in the War Office and men were returning from the Continent stripped of their weapons and often of their clothes, when there seemed to be no hope of giving them even rifles with which to defend their country, the Prime Minister wrote: "We are stronger than ever before. Look how many extra men we have here now. Form Leopard Brigades to tear and claw enemy."25
Churchill's close personal leadership never amounted to dictatorship. Although he generally called the British tune in discussions with the Americans, it would be a mistake to assume an identity between his views and those of the British Chiefs of Staff. Similarly it should be noted that what appeared as "British" views at the conferences were not necessarily endorsed by all British military men. "British" doubts concerning the feasibility of OVERLORD, for instance, were never shared by General Morgan and his staff.
Despite these differences of opinion, it remains true that there was a British point of view on strategy which Churchill and the British Chiefs of Staff consistently maintained. That view favored opportunistic, peripheral operations aimed initially at the reduction of German
power by indirect attack (by air over western Europe and by sea, air, and land in the Mediterranean), with the ultimate object of so weakening German defenses in France that a cross-Channel attack might be launched in the final phases of the war against an enemy whose will to resist was gravely shaken.26 It was never clear precisely what degree of deterioration this involved and no doubt opinions would have differed widely. What was central, however, to British strategy and consistently maintained by its advocates was that the date for the final assault across the Channel had to be contingent and indeterminate.
On the American side there was rather less unanimity in the direction of the war. In the first place, traditional American isolationism had made this country wary of political involvement in Europe. The Joint Chiefs of Staff tended therefore to develop a purely military perspective that considered political implications chiefly with an eye to avoiding them. This perspective accurately reflected the popular obsession with winning the war as quickly and as cheaply as possible-an obsession which allowed little room for consideration of America's postwar position.27 One effect was to make the Joint Chiefs of Staff relatively independent of the President and State Department in formulating strategy. The President, unlike Churchill, seldom intervened in detailed military planning. He tended to make only the large decisions as between fully developed alternative courses of action. He was occupied almost exclusively with large perspectives and the political problems involved in the liberation of occupied countries, the employment of United Nations28 forces, and dealings with foreign governments affected by proposed military operations. The President generally appeared at the Allied conferences as a defender of the strategy worked out by the Joint Chiefs of Staff.29 Though apparently reasonably convinced of its soundness, he was still in the position of a man who having accepted persuasion from one quarter was psychologically prepared to listen to advice from another. The President, in short, because of his position, his experience and personality, and American opinion, did not create a unified American front comparable to the British front molded and inspired by the Prime Minister.30
In part, the lack of complete unity in American views can be traced to the divergent
interests of the services. Before the establishment of the Joint Chiefs of Staff in 1942, the U.S. Army and Navy had had little experience in the sort of close cooperation now demanded of them. Agreement on joint strategy did not prove too difficult, but day-to-day implementation continually revealed disparities between the Army and Navy points of view. It was not only that the services had different habits of thought and organizational jealousies; they were split by the geography of the war. The Army's primary interest was in Europe where the bulk of its forces would be used. The Navy had a similar primary interest in the Pacific. The naval members of the Joint Chiefs of Staff, Admirals Leahy and King, were not opposed to OVERLORD, nor to the basic strategy of defeating Germany first. They simply lacked the U.S. Army's single-minded enthusiasm for European projects. On the one hand, while there remained any doubt as to whether OVERLORD would be mounted, they questioned the wisdom of hoarding resources in the United Kingdom and so delaying the day when the all-out offensive in the Pacific could be started.31 On the other hand, they were more open than the Army to persuasion that OVERLORD might be unnecessary.
In July 1943 naval representatives on the Joint War Plans Committee and the Joint Planners actually proposed shifting U.S. strategy away from the OVERLORD idea.32 They argued that world conditions which had made the cross-Channel project the best means of seeking a decision with Germany had so changed as to "cast grave doubt over the ROUNDUP concept." The decisive defeat of the German armies, they believed, would, if necessary, be accomplished by Russia. They accepted the original British idea of a cross-Channel operation in the final stages of the war, and concluded that, "without prejudice to Pacific and Burma operations, the continuing success and great momentum of our Mediterranean operations must carry on, at least until after Italy is knocked out of the war." The decision by the Combined Chiefs of Staff in May, to withdraw seven divisions from the Mediterranean for use in OVERLORD appeared to them unsound, since it would remove pressure from Germany, preventing a maximum effort in Italy and a possible opportunistic exploitation of Italian collapse through invasion of southern France. "If Germany chooses to fight in Italy," they wrote, "that is the place to fight her. There is, indeed, no doubt but that the combination of this constant wastage of Axis air power [through the combined bomber offensive] plus the elimination of Italy represents the maximum possible 1943 Anglo-American contribution towards the defeat of the European Axis."
The proposals, in short, completely reversed the American strategic principles consistently defended up to that time by the Joint Chiefs of Staff. The paper, representing "a new trend in U.S. thinking," went informally to Generally Marshall.33 It was considered so seriously that
when General Barker, deputy COSSAC, arrived in Washington with the OVERLORD plan he observed a "weakening" of the American attitude toward OVERLORD. He reported to General Morgan that the weakening was induced, "as previously," by U.S. naval authorities. He thought he had come just in time to strengthen the War Department's hand.34 Actually the War Department planners had not weakened. They had already begun a thoroughgoing attack on the naval position and had presented a separate analysis of strategy for consideration by the Joint Chiefs of Staff. This returned to the traditional U.S. view on OVERLORD and reaffirmed the May decisions of the Combined Chiefs of Staff, especially in regard to the withdrawal of the seven battle-trained divisions from the Mediterranean. Admiral King did not defend the naval planners' report in the Joint Chiefs' next meeting on 7 August, and the traditional War Department view was accepted again as U.S. strategy.35
The flurry of planners' memoranda at the end of July 1943 was the Navy's only open challenge of General Marshall's European strategy. Admiral King himself was apparently content that, in discussions with the British, U.S. views on European operations should be the War Department's view for which General Marshall was the spokesman.36 As European matters were the principal and by far the most controversial concern of the Combined Chiefs of Staff, General Marshall assumed the chief responsibility for defending U.S. strategy in debate with the British. To this responsibility Marshall brought both conviction and diplomatic skill. He was firm in his belief that the strategy evolved by the War Department was the correct strategy, yet he was always willing to attend the possible merits of rival suggestions. He fought hard for ROUNDUP in July 1942 and lost primarily to the President's insistence on immediate action. Thereafter Marshall rode two horses, struggling to keep tight rein on Mediterranean commitments in order that BOLERO could proceed apace. He managed this feat by a series of compromises. But, at Quebec, as has been suggested, the roads diverged. It seemed essential to choose between them. For
the first time since July 1942 Marshall accepted the need for a showdown with the British and undertook to lead the fight. To his colleagues he said: "We must go into this argument in the spirit of winning. If, after fighting it out on that basis, the President and Prime Minister [decide] that the Mediterranean strategy should be adopted . . . the decision [should] be made firm in order that definite plans could be made with reasonable expectation of their being carried out.37
Deeply as General Marshall opposed the Mediterranean strategy, he still preferred to settle on that rather than face a series of emergency operations and a generally hand-to-mouth strategy. This opposition between a settled plan and a dependence on opportunity was probably the most serious difference between the American and British points of view-a rift wider and more difficult to reconcile than any differences over where and when the campaigns in Europe were to be fought.
The British said in effect, "How can we tell what we should do six months or a year hence until we know how we come out of next month's action?" The Americans retorted, "How do we know whether next month's action is wise unless we know where we want to be a year from now?" The positions were difficult to reconcile because ultimately both stemmed from the respective resources and experiences of the two nations. For a year and a half Great Britain stood alone on the edge of a hostile Europe fearful a good part of the time of being herself invaded by the Nazi conquerors. She stood counting up her resources and matching them against her responsibilities for the defense of her own shores and her life lines throughout the world. The deficit in means must have seemed appalling.
From 1940 on it was obvious to British strategists that, if the war was to be won, eventually British armies would have to return to the Continent. But all their plans for such an eventuality had to recognize that a maximum effort might hurl about twenty divisions against the Atlantic Wall. Planners in 1941 and 1942 were bound to face strategy as primarily a problem of what could be done with such a relatively small army. Their invariable conclusion was that nothing could be done except under conditions of German collapse or a very great weakening in German morale. It should not be forgotten that the ground war in Europe at this time was being fought by massed German and Russian armies of hundreds of divisions. How could the scales of this conflict be tipped by the addition of twenty divisions? On the other hand if those twenty divisions (the core of an expeditionary force of about one million men) were committed in an unsuccessful enterprise, it was probable that they could never be reconstituted. England's striking power on land would thus be fruitlessly dissipated. It is not strange then that the strategists looked to the fringes of the Continent, outside the area where the bulk of the German Army was engaged. Their best hope for eventual success seemed to be to nibble the Nazi monster down to size. When this might happen no one could foretell. It depended partly on Russian success, partly on their own success. If they could get away with the first nibble they might try
a larger bite. Militarily they did not want rigid plans and long-range target dates. They did not want what General Brooke later called "a lawyer's agreement" to tie their hands.38 They wanted maximum freedom to exploit whatever situation arose and promised the best chance of furthering their initial object of wearing down the German military power in preparation for the final blow. The Mediterranean, because of the weakness of Italy and because of the unusual number and variety of legitimate military objectives, was an ideal area in which to work out their policy. Moreover, since the eighteenth century the Mediterranean had been an area of British political interest. Military aims and traditional political concern thus both focused on the European underbelly.39
America came into the war with historic consciousness of her great, unbeaten power. Although she was dangerously unprepared, her psychology from the beginning was offensive. Americans were confident they could build huge armies, navies, and air forces. For example the War Department "Victory Program," prepared in September 1941 in answer to the President's request for an estimate of the Army's needs for waging war, contemplated an army of 215 divisions.40 The thought was then that the USSR might succumb and that America had to be ready to take on the German Army alone, except for the relatively small contributions of Great Britain. After Pearl Harbor the War Department published a Troop Basis calling for an army of more than ten million men.41 As the war progressed Soviet success made a huge U.S. field army unnecessary. But the very fact that Americans at the beginning contemplated taking on the German Army by themselves and felt it possible is a clue to their psychology. As U.S. war power accumulated, Americans became impatient to bring on the decisive trial of strength. It was evident that such a trial could never be made in the Mediterranean. Moreover, the attempt to nibble away at the German power, the Joint Chiefs of Staff believed, would only result in attrition of their own strength. However great the reservoir of American military power, its application was limited by the logistical difficulties of long sea communications and the chronic shortage of shipping. Every division sent into the Mediterranean was a division lost for the main battle. It could not be moved to another theater without expending the same amount of shipping as
would be required to send a fresh division overseas from the United States.42 Its very presence in the Mediterranean sucked in shipping to sustain it and more divisions to reinforce it. The Mediterranean, General Marshall thought, was a vacuum into which American's great military might could be drawn off until there was nothing left with which to deal the decisive blow on the Continent.43
And what if, after committing the country's resources to the Mediterranean, they should be boxed up there, and destroyed? The Americans could not forget that the Mediterranean was practically a closed lake to which Spain held the western key. What if the Allied armies, after fighting hard for minor victories, should be cut off from the sources of supply? The possibility did not seem unlikely. Franco was a Fascist who had certain reasons to be grateful to Hitler and might be persuaded at least to turn the other way while German armies marched through to Gibraltar. This fear of being locked up in the Mediterranean was never shared by the British, who had long been accustomed to faith in Gibraltar and diplomatic bulwarks in the Iberian Peninsula. But it remained a strong factor in American hostility to Mediterranean commitments until well into 1943, when it was finally concluded that the Germans no longer had enough spare divisions to seize the Iberian Peninsula.44
The differences of strategic concept were fully recognized by both the U.S. and British Chiefs of Staff. At Quebec they were squarely faced in a debate brought on by the Joint Chiefs' submission of a paper outlining strategy for the defeat of the Axis in Europe.45 Their paper called for no action. It was a simple statement of principle, giving "overriding priority" to OVERLORD over all other European operations. At first reading General Brooke found it for the most part unexceptionable. He agreed that OVERLORD should constitute the major offensive for 1944 and that Italian operations should be planned with this conception as a background."46 That agreement, however, was not so positive as it sounded. The British view was that the Italian operations were essential steps on the road that led to OVERLORD. Being essential rather than diversionary, they acquired in British eyes at least an equal importance with the operation for which they prepared. The British objected therefore to assigning "over-riding priority" to the cross-Channel attack. The Joint Chiefs believed that, without such priority, Mediterranean operations would compete successfully for Allied resources and so perhaps make OVERLORD impossible when the time came to mount it. For three days the debate precipitated by that phrase "over-riding priority" raged, often behind closed doors with no secretaries present.
It was no mere verbal dispute. It tackled head on the issue of whether or
not the Allies should base their strategy on a single-minded concentration on a cross-Channel attack in 1944. That there should be no doubt of the seriousness of the issue, General Marshall stated the American view in strong terms. If the Allies relied on opportunism, he said, they would be "opening a new concept which . . . weakened . . . [the] chances of an early victory and rendered necessary a reexamination of . . . [the] basic strategy with a possible readjustment towards the Pacific."47 The threat had been used before, and was probably not taken very seriously. But there was no mistaking the new firmness in the American stand. "The United States Chiefs of Staff believe that the acceptance of this decision [to grant over-riding priority to OVERLORD] must be without conditions and without mental reservation. They accept the fact that a grave emergency will always call for appropriate action to meet it. However, long range decision for the conduct of the war must not be dominated by possible eventualities."48 The debate continued.
On 17 August 1943, the Combined Chiefs at last agreed to a statement of policy which on the face of it was acceptance of the American principle, but the wording was pregnant with mental reservations. Instead of giving OVERLORD "over-riding priority," they proclaimed: "As between operation OVERLORD and operations in the Mediterranean, where there is a shortage of resources, available resources will be distributed with the main object of insuring the success of OVERLORD. Operations in the Mediterranean Theater will be carried out with the forces allotted at TRIDENT except as these may be varied by decision of the Combined Chiefs of Staff."49 Possible eventualities still were riding high.
American attempts to crystallize strategy were, of course, at bottom merely efforts to force the British Chiefs of Staff into a hard and fast commitment on OVERLORD. How far they succeeded at Quebec is difficult to appraise. General Brooke, when directly challenged as to his faith in OVERLORD replied that he thought OVERLORD would succeed provided that the conditions of restricted enemy opposition as laid down in the plans were met. The Prime Minister in the course of both plenary sessions underlined these conditions. If they were still unfulfilled when the target date became imminent, then he asked that the Combined Chiefs review the operation. So that they would have a second string to their bow, he suggested that planning be directed now for an alternative operation against Norway. In conversation with Marshall, on the other hand, Churchill said that he "had changed his mind regarding OVERLORD and that we should use every opportunity to further that operation."50 He also told General Ismay after the conference that "the thing looked good."51 He would not, however, commit himself on the issue of the relative priorities of OVERLORD and the Mediterranean. In general it may be said that the British Chiefs and the Prime Minister at Quebec endorsed the principle of a cross-Channel invasion in 1944 but still
entertained considerable reservations as to its feasibility at that date. It would take another stormy meeting in the winter to settle these for good.
In the meantime, the reservations were less important than the decisions. In effect the Combined Chiefs at Quebec abided by their agreements of the Washington Conference. They resisted suggestions that the seven battle-trained divisions ought not to be moved out of the Mediterranean. They approved General Morgan's OVERLORD plan and ordered him to continue plans and preparations. They further suggested that the plan should be enlarged. The Prime Minister asked for an increase of the assaulting forces by at least 25 percent. He also desired a landing on the east coast of the Cotentin. General Marshall agreed. Although these recommendations were not embodied in a directive, they constituted an authorization to Morgan to overstep the limitations imposed on his planning by the Washington agreements.52
The OVERLORD planners had also asked for a diversion against southern France to prevent the Germans from moving troops from the south to meet the main Allied assault in Normandy. The Quebec conferees accepted with very little discussion the desirability of such a diversion and ordered General Eisenhower, then commanding general of the North African theater, to draw up plans for an actual operation. It was to be timed to coincide with OVERLORD and have as its objective the establishment of a lodgment in the Toulon-Marseilles area with subsequent exploitation northward. The British expressed some tentative doubts about the plan. General Brooke believed that the invasion should be attempted only if the Germans had been "forced to withdraw a number of their divisions from that area,"53 and the Prime Minister wondered whether air nourishment of the French guerrilla forces in the region might not constitute an acceptable alternative. In the absence of a specific plan and before the Italian venture had been undertaken, the whole project was evidently highly speculative and the most that could be decided was to explore the possibilities.
Even the possibilities were admittedly very limited. General Eisenhower was directed to plan on the basis of resources already allotted to his theater, and it was estimated that this would allow him an amphibious lift for only 27,000 troops and 1,500 vehicles-a lift, in other words, for about one division. It was further noted that "the ships and craft shown do not provide a balanced assault lift...." However inadequate these resources might be, the Quebec planners ruled that nothing more would be available. "Augmentation is not considered practicable without drawing from OVERLORD."54
Despite this ruling, planning would reveal in the months ahead that augmentation was essential, and in consequence the southern France invasion, though conceived by the Americans as an integral part of OVERLORD, would actually become another competitor for scarce resources.
On the other hand, by providing a means to employ forces already located in the Mediterranean in direct support of the attack against northwest Europe, it would prove a useful talking point against British advocacy of Balkan adventures or an increased effort in Italy.
For the fate of OVERLORD, the casual suggestion to increase its size was probably the most significant product of the Quebec Conference. It set in motion plans and preparations that acquired their own momentum, which with the mere passage of time became increasingly difficult to reverse. This fact underlines an important characteristic of all the debates among the Combined Chiefs of Staff. The debates always prefaced action. Action was required by circumstances regardless of whether minds met or not. There was never any question of allowing differences to split the Anglo-American alliance. However great was the disparity in strategic thinking, the community of basic interests ran deeper. The need for common action was more vital than any principle under debate. Thus, although neither Americans nor British perceptibly yielded their separate points of view, practical decisions were made as if such yielding had taken place. The deadlock of ideas was broken by the practical politics of compromise, enforced by the necessity of continuously acting against the enemy.
Compromise, furthermore, was made possible, as well as necessary, by the equality of the Allies. The realization that neither America nor Great Britain could act alone was implicit in every discussion and every decision. England's base and America's resources were alike indispensable for a decisive trial of strength with Germany. Each nation separately would have been doomed to a long and possibly futile struggle.
Quebec was as inconclusive as the previous meetings had been in attempting an unalterable definition of European strategy. But operation OVERLORD came through the debates unscathed by any serious criticism. OVERLORD was the weapon. Once forged, it would argue for its own use more eloquently than all the words in council.
General Morgan did not wait for the decisions at Quebec to begin a re-examination of his completed plan with the view to strengthening it. One of his first steps after the plan had been delivered to the British Chiefs of Staff was to ask his naval staff to estimate, first, the number of additional landing craft that would be required to load the vehicles of the two follow-up divisions tactically so that they could be used on D plus 1, and, second, the number of craft needed to lift another assault division. The figures supplied him on 7 July 1943 almost immediately became irrelevant.55 Instead of being able to contemplate a larger assault, Morgan saw his already inadequate supply of landing craft dwindling away.
Landing Craft Again
Of the 653 LCT's which the Washington Conference allotted to OVERLORD and which General Morgan from the outset complained constituted a bare and dangerous minimum for the task, 44 had
been taken by the British Navy for net protection duties at Scapa Flow. It was possible, though not certain, that some might be released in time to take part in OVERLORD. Still more serious, a large number of LCT's had to be converted into close-support craft for the assault because the Combined Chiefs had made almost no provision for such craft. Although the burden of neutralizing the enemy coastal and beach defenses would be undertaken by the air forces and naval vessels, the heavy preparatory fires would have to be lifted some minutes before the assault craft actually touched down. It was never imagined that preparatory fire could destroy any significant proportion of the enemy defenses. The most it could do would be to prevent the enemy from manning his guns effectively while the bombardment was in progress. It was therefore essential for the assaulting force to keep up neutralizing fire until the last possible moment. That meant small craft, armored and equipped with guns and rockets, which could maneuver close to the shore and fire as the assault waves rolled in. To perform this vital task, the Washington Conference allotted OVERLORD just thirteen LCG's (Landing Craft, Gun). Early in August 1943, COSSAC was informed that only seven of these craft could be delivered in time; the rest would be a month late. General Morgan took the opportunity to raise the whole question of the inadequacy of the planned close support. As a result, the British Chiefs of Staff invited him on 12 August to prepare a detailed report on the shortage of landing craft "and other naval equipment" for OVERLORD.56
General Morgan rolled up his sleeves and accepted the invitation with zest. He told 1 Army Group, which was particularly concerned over the shortage of support craft, that he would take full advantage of the opportunity offered him to expound the subject of landing craft shortages. "I am aiming," he said, "to give the Chiefs a proper earful."57
Facts for the earful were supplied by his naval staff, which calculated shortages based solely on current allocations and the three-division assault plan. Altogether about one-quarter of the 648 LCT's planned for the assault lift were, for various reasons, no longer available. In addition to the 44 already noted which were being used at Scapa Flow, 36 were to be converted to LCT(R)'s (rocket-carrying craft), 48 would be armored to carry direct-support high-explosive weapons, and 36 would probably be needed for close support of the U.S. assault division. The latter estimate, in reality, was very low. General Devers, ETOUSA commanding general, subsequently calculated that the American assault division would require 56 support craft of the types that used LCT hulls or the equivalent. The net deficit Navy planners set at 164 LCT's as well as 7 LCI(L)'s that had been converted into headquarters ships.58
How was this deficit to be made up? The U.S. Chiefs of Staff held out little hope that the craft might come from
American sources. The United States, they decided, could not assign additional LCT's to the United Kingdom to replace those diverted from OVERLORD to Scapa Flow defense duties.59 The Joint Chiefs declared that the United States had never built any gun-support craft and did not intend to.60 On 10 August Admiral King wrote to Admiral Stark: "Production in the U.S. and requirements of other theaters will permit no advancement of present schedule of BOLERO landing craft to come from the United States mainland. The above matters are being studied at the present time and you will be informed when decisions are reached."61 The decision was reached a few days later to explore the possibility of increasing production, and Admiral King then said he felt sure some increase could be achieved. But there was considerable doubt whether any increases, even if effected, could be felt in time for OVERLORD. The program of U.S. landing craft production remained for several weeks clouded with uncertainty.
British production offered no better hope. The shipbuilding industry in the United Kingdom was near the saturation point, especially as regards the employment of available skilled labor. The manufacture of landing craft had already been expanded to the point where it was consuming a quarter of all the steel worked into new hulls. To increase it further would have involved a host of technical difficulties which seemed uneconomical to British authorities.62 Not until later in the fall did the need for more craft appear so urgent that the Prime Minister intervened to cancel certain ship construction in order to make way for an increase of about sixty landing craft for OVERLORD. At the moment, General Morgan's request for a revision of the whole shipbuilding program was turned down by the British Chiefs of Staff.63
The problem of how to make good the deficit in landing craft for a three-division assault had received only a preliminary examination when word came from Quebec that the Prime Minister wanted the assault increased to four divisions.64 The British Chiefs of Staff asked General Morgan to report on that possibility. By the end of September, COSSAC completed a thorough re-examination. So much debate and hypothesis had by that time clouded the issue that General Morgan undertook to start afresh, recalculate his requirements, and restate his tactical thinking.
"My original Directive" he began, "placed at my disposal a quantity of landing craft which bore little or no relation, as to numbers and types, to the actual requirements of the proposed operation." Now, he continued, it was proposed to strengthen the operation, which he fully conceded needed strengthening. But did this mean adding a fourth division to the assault? General Morgan thought not. He pointed out the dangerous weaknesses of the plan as it stood, weaknesses which were thrust on it by the
inadequacy of the landing craft allotted. "Detailed analysis of the present plan shows that while the three assault divisions are only barely adequately mounted in craft of suitable types, the immediate follow-up formations are most inadequately mounted, and there is a dangerous gap on D-plus-1 day." A large proportion of the follow-up forces being mounted in ordinary shipping could not be tactically loaded and thus would not be operationally available on the far shore until twelve hours after landing. In short most of them were "follow-up" in name only; they would not be in position immediately to reinforce the assault troops. General Morgan therefore recommended that, before any consideration be given to increasing the number of assaulting divisions, all additional landing craft that could be raked up should be put in to strengthen the follow-up and provide a floating reserve. "We already have far too high a proportion of our goods in the shop window," he said. "To consider any increase in this proportion without adequate stocking of the back premises would in my opinion be basically unsound."65
Morgan's new calculations of craft needed to permit the landing of two full divisions in the follow-up for use on D plus 1 showed a deficit of 251 LCT's for a three-division assault and 389 for a four-division assault. In addition, for a four-division assault there would be a shortage of more than 150 support craft using LCT or equivalent hulls.
Quite apart from this very large landing craft requirement, the four-division assault struck General Morgan as unwise because it would necessitate broadening the assault front. Extension to the east he believed would bring the assaulting troops within range of the Le Havre coastal guns, which were among the most formidable in the Atlantic Wall. Extension on the right flank of the assault would involve landing on the beaches northwest of the Carentan estuary. Reversing his previous stand that this would be desirable, General Morgan now noted that the Germans had already begun flooding the hinterland of the Cotentin and that therefore the contemplated assault in that area was unsound.
The British Chiefs of Staff were not impressed with this argument and continued to advocate a four-division assault, the fourth division to be American and to be employed against the east Cotentin. No suggestions were made as to how the evident difficulties might be overcome.
However the strengthening of OVERLORD was to be accomplished, there was general agreement that strengthening was needed, and that this would require large new increments of landing craft. In the urgent need to find these craft, the question of their tactical employment for the moment took a back seat.
In September Donald Nelson, chairman of the U.S. War Production Board, went to London and talked to General Morgan and his staff about landing craft requirements. As a result of his conversations he cabled Charles E. Wilson his conviction that LST's and LCT's were the "most important single instrument of war from the point of view of the European
Theater," and that the requirements for them had been "grossly understated."67
Planning adequate supplies of landing craft, however, was still complicated by the competition of other parts of the war production program for critical war materials, particularly steel plate and marine engines. Equally important was the competition between landing craft and other shipping for priority in the nation's crowded shipyards. The Navy's 1944 shipbuilding program called for a 50 percent increase of tonnage over the previous year, while the Maritime Commission's schedules for merchant shipbuilding remained about the same as before. To superimpose on this large shipbuilding commitment an increase in landing craft production schedules required a careful scrutiny of strategic needs for various types of shipping. Furthermore, if changes were to be made in apportioning materials and facilities, they had to be made long in advance of anticipated needs since readjustments in the program required time--time to cancel contracts for one type of vessel and let contracts for another, time to complete construction already under way on one type and initiate construction of another. A large part of the Navy's 1944 shipbuilding program consisted of small vessels and destroyer escorts which used mostly the same tools, materials, and yards as landing craft.68
Time was already short when the question of increasing landing craft production was officially raised at Quebec and Admiral King told the Combined Chiefs of Staff that he was examining the possibility of building more landing craft by halting the construction of 110-foot subchasers. This, he thought, might result in an increase of 25 percent, but the estimate was not yet firm.69 On 20 August 1943, planners, after tentative studies, reported that any acceleration in landing craft production would probably not be felt before 1 April 1944.70
On 13 September the Chief of Naval Operations submitted a new schedule calling for a 35 percent increase in landing craft production, saying that it "was designed to meet the requirements of the operations agreed upon in QUADRANT." The Joint Staff Planners commenting thereon agreed, with qualifications.71 "Landing craft destined for OVERLORD," they pointed out, "will arrive in time for the operation [that is, craft already allotted] though the complete allotment of some types from the United States will not arrive as early as desired by COSSAC. The most critical situation will exist in LCT (5 and 6)'s. The only solution (if the late arrival cannot be accepted) seems to be additional withdrawals from the MEDITERRANEAN. There is no increased U.S. production in this type of craft.... The QUADRANT decision relative to OVERLORD will therefore not be affected
by the increased production of landing craft."
At the end of October 1943 logistical planners had prepared another study which indicated that landing craft production could be further increased, but no decision was made and the 1 November production schedules virtually repeated the 1 October schedule. In the meantime General Morgan, visiting Washington, was led to believe that "if sufficiently powerful pressure was applied at the right spot, U.S. landing craft production . . . [could] in fact be increased...." Although he observed that it was "extremely difficult to find out just exactly whence to have this pressure applied. . . ," he still hoped to be able to "bring matters to a head."72 In that hope he was evidently disappointed. As time passed and action was postponed the prospects of producing more craft in time for OVERLORD faded.
Thenceforward procurement of landing craft for OVERLORD proceeded on the basis of making the most of existing supply. The principal methods of stretching the supply were reallocation from the Mediterranean theater, a comb-out of training facilities in the United States and United Kingdom, increase in serviceability rates, and increased loading. All these expedients were fully explored in the months following General Eisenhower's assumption of the Supreme Command in January 1944, but up to the last minute the situation remained tight and threatened to compromise the tactical dispositions for the assault.
For OVERLORD planners the months following Quebec were filled with frustrations. The order to re-examine the strength of the assault brought them up against the apparently unanswerable question of landing craft availability. Still more baffling was the seemingly simple directive to carry on with plans and preparations. What did it mean? General Morgan's original job-the only job for which he had a clear-cut directive-was finished. He had been told to study the possibilities of mounting a cross-Channel invasion of a certain size at a certain date and say by means of an outline plan whether such an operation was feasible. The verdict had been returned. To continue planning could mean only to plan tactically, to fill in the detail of how the operation as outlined could be carried out. But this was no job for a small planning group designed as the nucleus for the Supreme Headquarters. It could be done only by the tactical headquarters which would have direct control of the operation: army group, army, and corps. The Quebec order, therefore, was actually an order to COSSAC to farm out and coordinate tactical planning. Such a task obviously required executive authority. General Morgan was legally only the chief of a planning group, without command functions. He was, as he pointed out, a relatively junior officer.73 It was out of the question to give him the substance of command, but he was at length given the shadow. The
Questions of Command
Combined Chiefs of Staff in September made him temporarily responsible for "taking the necessary executive action to implement" the COSSAC plans, pending the appointment of a supreme commander.74 That this was a stopgap expedient scarcely adequate for the situation was apparent to everyone and to no one more painfully than to General Morgan himself. He wrote to General Devers: "While I hate the sight of this whole business I am completely at a loss to suggest anything better, short, of course, of appointing the great man himself which appears to be utterly impossible."75
The quasi command thus conferred on General Morgan had grave limitations. Although, with the backing of the Combined Chiefs, he could now take "executive action" he could not, on his own, make command decisions. One command decision, in particular, was immediately required if preparations for OVERLORD were to go forward: what was to be the organization of command of the ground forces in the assault and in the successive phases thereafter?
The basic principles of combined command had been outlined in 1942 and agreed to by the Combined Chiefs of Staff in the fall of that year.76 They stemmed from the principle approved at the ARCADIA Conference that one Allied commander should have supreme command in each theater of operations.77 But agreement on principle still left a wide area for debate on the detailed organization of the command for specific operations. The COSSAC plan for OVERLORD contemplated the carrying out of a single tactical mission by a task force of army size and mixed nationality. There was no question of the need for a single unified command over all ground forces in the assault phase. The closest coordination was essential, both because the immediate objective was the establishment of a single bridgehead and because the enemy could be expected to probe out the boundary between U.S. and British forces and, reasoning that it represented a point of special weakness, attempt to drive a wedge into it. There was, on the other hand, no question of the necessity for having both American and British troops take part in the assault.78 Military and political considerations both required it. The administrative and logistical organizations of the U.S. and British armies were so different that to attempt to pass one army through the beachhead established and organized by the other would have involved critical difficulties. Even if these difficulties could have been accepted, it was still unthinkable that the first blow of the supreme offensive of the
Western Allies should be struck by the armies of only one of them.
The situation envisaged by OVERLORD was actually not very different from that existing in the attack on Salerno where a British corps was subordinated to an American army controlling the assault. Apparently, however, it seemed different, for the Salerno experience with mixed tactical command was conspicuously absent from the discussion of the OVERLORD problem.79
In the OVERLORD plan, COSSAC recommended that one U.S. division be employed on the right of the assault, two British divisions on the left, and that all three be initially under a British army commander.80 As soon as an American army was established on the Continent, Allied field command would pass to a British army group, which would continue to exercise operational control until the capture of the Brittany peninsula or the establishment of a U.S. army group in France, whichever occurred first. Morgan's reason for recommending initially a British chain of command was his feeling that it would be easier for British commanders to organize and coordinate an assault from a British base. It is clear, further, that General Morgan wrote the OVERLORD plan under the impression that the supreme commander would also be British.81
Planning recommendations are one thing; decision another. It was true that the OVERLORD plan had been approved without change by the Combined Chiefs; but, when General Morgan pressed for a ruling on whether this over-all approval could be taken to apply to the recommendations on command, he was told that the Combined Chiefs had not been aware of any "special implications" of approving the command system, and that in any case details of the plan were subject to change by the supreme commander when appointed.82 General Morgan replied that a firm decision was required at once if the operation was to take place on schedule, and he asked the Combined Chiefs to make that decision.83
The British Chiefs of Staff then decided to accept General Morgan's proposals in the OVERLORD plan as representing their own point of view, and asked for American comment.84 The Americans at once began formulating counterproposals. They had not agreed on any formal reply to the British, however, when the whole question was deflected by the introduction of a new and more pressing problem: the need for immediate establishment of an over-all tactical air command.
A tentative organization of the British tactical air forces for support of the invasion had already been set up under Air Marshal Sir Trafford Leigh-Mallory, who had been authorized to make planning decisions on the use of the combined U.S. and British tactical air forces without prejudice to the later appointment of
another air commander-in-chief.85 At Quebec in August 1943, the Combined Chiefs agreed to name Leigh-Mallory Commander-in-Chief, AEAF (Allied Expeditionary Air Force), but delayed writing a directive.86 This action, in effect, formalized his authority to make combined planning decisions on air matters for OVERLORD, but he remained without a combined command. About the middle of September Maj. Gen. Lewis H. Brereton arrived in England from the Middle East with orders to organize a U.S. tactical air force in the United Kingdom. This became the Ninth Air Force, formed initially with the headquarters of the Ninth Air Force from the Middle East and planes from the Eighth Air Support Command taken over from General Eaker's command.
In October the British Chiefs of Staff pressed for the immediate integration of U.S. and British tactical air forces under Leigh-Mallory's command, and submitted to the U.S. Chiefs on the 12th a draft directive outlining AEAF powers and responsibilities.87 Their earlier request, meanwhile, for U.S. views on Morgan's command and control recommendations was still unanswered. On 19 October the Joint Chiefs commented on both British papers in the same vein. To the proposed directive to the Commander-in-Chief, AEAF, they replied, "It is the view of the United States Chiefs of Staff that the issuance by the Chiefs of Staff of directives to subordinates of the Supreme Allied Commander is unsound." This principle applied equally, the Joint Chiefs believed, to the proposed successive command of ground forces in the assault. To specify the organization of command under the supreme commander would be to encroach on his prerogative.88
These two negative replies funneled the discussions on OVERLORD command into a single debate over the authority of the supreme commander, as the Combined Chiefs began the delicate job of writing a directive for him. When they undertook the task, the identity of the supreme commander was still unsettled, but it was known at least that he would be an American and there was a strong presumption that the choice would be General Marshall.89 It was therefore a good bet that decisions left to the discretion of the supreme commander would in fact reflect American views on command.
In contrast to the British, the U.S. Army always insisted on permitting a field commander the maximum freedom and discretionary power in the exercise of his command. The Americans believed that it was sufficient for the Combined Chiefs to assign the supreme commander a mission and leave to his discretion all the details of how that mission should be carried out. They viewed with alarm the British tendency to extend the control from the highest level down through the echelons of command, narrowly specifying the functions of subordinate commanders. When Eisenhower
heard, after his appointment to the supreme command of OVERLORD, that the British Chiefs of Staff proposed to dictate the detailed composition of the tactical air forces under him, he complained not only against the specific proposal but against what he called the British tendency to freeze organization so that commanders could not use trusted subordinates in their proper spheres.90 The British position seems to have been precisely that they were hesitant to trust their subordinates. The Prime Minister wrote: "In practice it is found not sufficient for a Government to give a General a directive to beat the enemy and wait to see what happens.... The General may well be below the level of his task.... A definite measure of guidance and control is required from the Staffs and from high Government authorities. It would not be in accordance with the British view that any such element should be ruled out."91
It was against this background of differing conceptions of the basic principles of command that the specific debates on OVERLORD command took place between the Americans and the British. The principal objections which the Americans advanced to COSSAC's proposals on successive command in the invasion were, first, that they would have subordinated American units of division size and smaller to direct British command in violation of the Casablanca declaration, and, second, that they seemed to exclude the supreme commander from operational control of the assault and early build-up phases.
General Devers, in the first formal commentary on the COSSAC proposals, argued that both objections could be met by having the assault divisions controlled by U.S. and British corps commanders whose efforts in turn would be coordinated under direct command of the supreme commander through an advance headquarters. COSSAC replied that the Devers plan was probably unworkable. Supreme Headquarters would not be set up for tactical command. It would not be organized to control the detail that normally flowed through a field army. Furthermore it would be impossible for the supreme commander to be physically located so that he could exercise field command of the assaulting forces and at the same time carry out his basic mission of coordinating the whole operation through relations with ETOUSA, the air and naval commands, and U.S. and British army group headquarters and service ministries, all either located in London or having established liaison and communication with the capital.92
Despite this criticism, the War Department continued to advocate some measure of direct control by the supreme commander in the assault. The Devers plan, however, was modified to restore an army commander to take direct command. The War Department believed that the assault army should be American, not British, and this view was apparently accepted as a basis for discussion although it was not at once embodied in a written agreement. War Department thinking, reflected in a series of draft
memoranda and reports during September and October,93 envisaged a command setup somewhat as follows: the assault would be initially under the Commanding General, First U.S. Army, whose forces would consist of airborne troops, a British corps, U.S. corps, and Canadian corps, each to contribute one division to the assault. "During the period from a date prior to the assault to be designated by SAC94 until CG, First U.S. Army relinquishes the overall command of forces on the continent to British 21st Army Group, control of operations will be exercised directly by SAC to First Army."95 After the first three corps had built up on the Continent, a British army headquarters would be established. At the same time the Continental field command would pass to 21 Army Group. The supreme commander's control of the operation would then be exercised through army group to U.S. and British armies. On the arrival of the second U.S. army, First U.S. Army Group would assume direct command of all American forces under the supreme Allied commander.
There was no argument about the arrangements thus outlined for the build-up phase. General Morgan, however, was still unable to accept the concept of the supreme commander's direct intervention in the assault. He was, in fact, unable to understand just what was meant. After considerable discussion of the matter in Washington during his visit in October, he wrote:
"There persists here the thought that S.A.C. should have some direct connection with the army of assault. Now, what this direct connection is to consist of I cannot for the moment say. I am completely bogged down for the moment in the differences between our two languages. It seems that the word "command" has two different meanings in our two services. All I can get out of General Marshall is that he has a sensation that he should in some way control the assaulting army, although I am quite sure that his conception falls far short of what we understand by the term "command". As far as I can make out, it also falls far short of the American conception of command. But how to define it I must confess completely puzzles me. . . "96
General Barker, acting head of COSSAC in Morgan's absence, took up the problem with his staff in London and replied at length. He did not feel that General Marshall need be apprehensive about the degree of control he could exercise over the assaulting army. But the control should be that appropriate to a supreme commander, not the control normally exercised by a tactical headquarters. General Barker wrote:
"I think it goes without saying that the assault must be directly commanded by an Army Commander. Furthermore, the fact that an Army Group Commander must perforce take over direct command of the operation at a fairly early stage, say D plus 5 or 6 or thereabouts, makes it essential that said Army Group Commander should be closely associated with the planning and execution of the assault, otherwise there is likely to be a break in continuity of command, or at least some friction in the change-over as well as in the planning for the buildup. After all, the SAC can and would intervene at any time the situation seems to warrant it. The Salerno operation is a good example of the direct
intervention of a Commander-in-Chief when the operation gets sticky. He would, of course, have a senior staff officer from his staff on liaison with the Army Commander.
You will recall that Alexander and Eisenhower both intervened in Clark's battle at Salerno, and as the result, Eisenhower ordered the whole weight of the Air and Naval forces to concentrate on the battlefield. In the assault stage of OVERLORD, it is these two weapons, the Air and the Navy, that SAC would employ to influence the course of the battle. Our headquarters will have direct telephonic, telegraphic and radio contact with Army, Army Group and the Air and Naval Headquarters, as well as with our liaison officers at these headquarters. In consequence, the SAC will be in the closest touch with the battle and can intervene quickly should the necessity arise."97
General Morgan agreed and wrote that he would "make arrangements for [General Marshall as SAC] to be able to participate directly in Bradley's battle when it takes place. What though the planning must have been through the other system, that is, through the 21 Army Group, I am myself satisfied that this is the practical solution having in view the fact that the Commander himself will not be present throughout the whole proceedings."98
American concern with expanding the authority of SAC took another direction. In the same memorandum in which the Joint Chiefs criticized the COSSAC proposed chain of command, they recommended defining the supreme commander's sphere of command to include the whole of Germany for purposes of conducting air operations.99 The British at once demurred, saying that the proposal amounted to giving the supreme commander all responsibility for strategic bombing. This responsibility, they felt, should remain with the Combined Chiefs of Staff. General Marshall observed that operation POINTBLANK [the strategic air offensive] is intended as a preparation for OVERLORD and may now be extended to operations from Italy. Thus it is a part of OVERLORD.100
Although there was no argument over SAC's right to control the strategic bomber forces once the invasion had begun, the British strenuously objected to turning over command before that date. The furthest they would go at the moment was to specify in the draft directive to the Commander-in-Chief, AEAF, that the strategic air forces would be "detailed from time to time by the Combined Chiefs of Staff to operate with all or part of their effort to meet the requirements of the Supreme Commander."101 They believed that SAC control of the Eighth Air Force and RAF Bomber Command was unsound. Air operations were specialized; air channels had been organized through four years of war; air operations affected all enemy fronts; and, finally, the air offensive already in progress was designed to establish conditions for OVERLORD. This system could not be improved on by putting SAC in command.102 The Joint Chiefs could not accept the argument. Whatever
the technical objections might be, they held it was unthinkable that the supreme commander of any operation should not have absolute command of all forces needed by him to carry out that operation. "A committee," said General Marshall, "cannot fight a battle."103
The upshot of the debate was a temporary deadlock. It was decided then to put the question aside until it became critical. At the moment it was more important to issue a directive to the Commander-in-Chief, Allied Expeditionary Air Force, authorizing him to organize the U.S. and British tactical air forces into a single command under SAC. The essential outlines of this command were agreed to by the Combined Chiefs of Staff, and, despite U.S. objections to issuing directives to subordinate commanders, they were embodied in a directive to Air Marshal Leigh-Mallory early in November.104 The AEAF headquarters was established as of 15 November, to take immediate control of the RAF Tactical Air Force and the Air Defence of Great Britain. The date at which the Ninth Air Force would pass to AEAF command was tentatively set for 15 December.105
Designed first as an appendix to the directive to the supreme commander, the AEAF directive was issued separately as it became apparent that full agreement on the definition of SAC's authority could not immediately be reached. Consideration of the latter problem, in fact, was deferred to January.106 In the meantime at the Cairo Conference in December the supreme commander was at last appointed. General Eisenhower, then commander in chief in the Mediterranean, was notified of his appointment about the middle of December and arrived in England to take command the middle of January. It was another month before the formal directive from the Combined Chiefs was issued to him. But the debate over the directive after his appointment was mainly verbal. The few substantial issues at stake were settled by being dropped. In effect the whole directive was trimmed down to what both sides could agree to and the settlement of such questions as the control of strategic air was left for later discussion in which the supreme commander's determination of his own requirements would be the decisive factor.107
The appointment of General Eisenhower ended almost a year of uncertainty which kept many of the preparations for invasion in long suspense. When the question of selecting a commander for cross-Channel operations was broached at the Casablanca Conference, it was shelved on the grounds that it was unnecessary to appoint a commander so early in the proceeding. Informally it was agreed that the commander, when appointed, would be British. Casablanca's perspective, it will be remembered, was confined principally to 1943.
If a cross-Channel invasion had been attempted that year, the British would inevitably have made by far the greater initial contributions of resources. It was on that basis, and in consideration of the fact that General Eisenhower was commanding in the Mediterranean, that the agreement was made. COSSAC was given to understand immediately after Casablanca that the matter was settled. The Prime Minister informed General Sir Alan Brooke that he would be appointed supreme commander as soon as it seemed advisable to make a formal appointment.108
These arrangements were much less settled than they seemed. Churchill, while agreeing to the choice of a British commander, formally enunciated the principle that the nation contributing the majority of the forces to any combined enterprise should command it. When it became apparent that no cross-Channel operations would be launched during 1943 and that for the 1944 operation the United States would furnish the bulk of the troops and materiel, the Casablanca decision on the high command was changed. As early as July 1942 the Prime Minister had suggested to President Roosevelt the appointment of General Marshall to command cross-Channel operations.109 Churchill apparently repeated the recommendations at Quebec a year later.110 Although no agreement was reached, rumors of Marshall's appointment leaked to the press, and received increasing credence even from those officially concerned with OVERLORD.111 When President Roosevelt was asked by General Morgan in October for confirmation that General Marshall would be given the command, the President replied only that he was still not sure.112
Behind the President's reluctance to appoint General Marshall to the OVERLORD command was the feeling that no one could satisfactorily replace him as Chief of Staff of the Army and member of the Combined Chiefs of Staff.113 In particular it was felt that General Marshall had established a working relationship with Congress which would be jeopardized by bringing in a new man whom Congress did not know. There was also a very general sense in the American camp that a field command in one theater was not a big enough job for the man who for four years had occupied the top position in the U.S. Army and had taken a leading part in formulating strategy throughout the world. Partly in an effort to create a job suitable for their Chief of Staff, the Americans at the Cairo Conference suggested combining the European and Mediterranean theaters under a single commander who would have the direction of all operations in Europe.114 The suggestion
stemmed also from American conviction that a single command over related operations was more efficient than cooperation between coordinate commanders. It was turned down by the British on the grounds that a single field command over such disparate operations was unfeasible and that the proposal therefore would only mean the creation of an unnecessary intermediary between the Combined Chiefs of Staff and the theater commanders.115
Shortly after the British had refused to entertain the idea of amalgamating the Mediterranean and European commands, President Roosevelt decided that General Marshall should remain at his post as Chief of Staff, and that General Eisenhower, the next logical choice, should take command of OVERLORD.
While principles were being discussed on the highest levels, the organization in the theater of various headquarters that would assume control of the operation proceeded. In the early summer the British had constructed the skeleton of their tactical command for OVERLORD, complete with the creation of headquarters of the Second British Army, the First Canadian Army, and 21 Army Group. At that time the highest U.S. ground force command in the United Kingdom was still V Corps. In May, General Devers strongly urged the War Department to establish a U.S. army headquarters.116 It was required to parallel British organization, to "initiate actual planning for the 1944 operation," and finally as part of the scheme to make the Germans believe that an attack across the Channel would be made in 1943. Two months later General Devers further recommended that a skeleton headquarters for a U.S. army group also be sent to England.117
Despite this urging, which was reinforced by General Morgan, the War Department did not become convinced of the necessity for an immediate appointment of an army commander until the end of August 1943, and insisted then on still further delay in naming an army group commander.118 The choice of an army commander was unanimous: he was to be Lt. Gen. Omar N. Bradley, American commander in the battle of Tunisia.119 His release from the command of II Corps then in Sicily was arranged during the last week in August, and effected on 7 September. After conferences in Washington, Bradley arrived in the United Kingdom in October. On the 20th of that month the headquarters of the First U. S. Army was at last opened. The headquarters had been activated in the United States during September with cadre from the Eastern Defense Command.120
Army group headquarters was to be allowed to develop gradually out of the theater as the need for it grew. At the time that General Marshall arranged for the appointment of General Bradley as army
commander he was still not ready to name a commander for army group. General Devers then proposed that a single superior U.S. headquarters, a U.S. GHQ, be established to direct both operations and administration, following the precedent of World War I.121 Devers' proposed GHQ was to consist of a field headquarters and a rear echelon to handle theater functions. It was to be formed gradually by doubling the staff sections of ETOUSA to constitute forward and rear components. Once the field headquarters moved to the Continent, the theater functions might be redefined to permit, presumably, a greater independence of operational and administrative commands.
The Devers proposal did not suit General Marshall, who believed that a maximum separation of operational and administrative functions was desirable.122 On 18 September the War Department agreed in principle to the establishment of a U. S. army group headquarters, and planning for it was delegated to General Bradley in addition to his duties as army commander.123 So strongly did General Marshall feel that the new headquarters should not be burdened with theater responsibilities that he recommended to General Devers a physical separation of army group from ETOUSA. "I desire that the organization of the Army Group Headquarters be initially controlled directly by Bradley under your supervision and that it not be merely an offshoot, or appurtenance to ETO Headquarters."124
First U. S. Army Group (FUSAG) was activated on 16 October. Its first assigned task was operational planning under the direction of ETOUSA. The operational missions of both FUSAG and First Army were to be assigned later by COSSAC. By this time, however, it had already been decided that First U. S. Army would command at least all American troops in the assault and that 21 Army Group, chiefly because it was early on the scene and had participated in COSSAC planning, would have over-all ground command in the assault and early build-up phases. In effect, those decisions meant that the role of FUSAG would be to take over command of U. S. troops when two American armies had become operational on the Continent, that is to say, after the establishment of the initial lodgment area.
It does not appear that the decision giving initial ground command to 21 Army Group was ever formally confirmed by the Combined Chiefs of Staff. When General Morgan returned to London in November after more than a month's stay in Washington, he carried back the instructions and full confidence of General Marshall. "He knows exactly what I want," General Marshall told General Devers in asking the latter for full cooperation with COSSAC.125 It was presumably in the light of General Marshall's instructions that General Morgan, shortly after his return, began to draft a directive to the Commander-in-Chief, 21 Army Group, which finally fixed the chain of command for OVERLORD.
That directive, issued on 29 November, significantly under the letterhead of "Supreme Allied Headquarters," told the 21 Army Group commander that he would be "jointly responsible with the Allied Naval Commander-in-Chief and the Air Commander-in-Chief, Allied Expeditionary Air Force, for the planning of the operation, and, when so ordered, for its execution until such time as the Supreme Allied Commander allocates an area of responsibility to the Commanding General, First Army Group." He was also told that operations in the assault phase would be carried out by U. S. and British corps under "the unified command of the Commanding General, First (US) Army," and that "the Commanding General, First (US) Army . . . [would] remain in immediate control of land operations until such time as the forces landed warrant, in your opinion, the introduction of a second army headquarters . . . to take over a portion of the front."126
It will be noted that this directive made the commander of 21 Army Group an over-all ground commander in the initial phases coordinate in authority to the air and naval commanders in chief, but it specifically limited his tenure to the first part of the operation. In the earliest attempts at devising a formula for unified combined command, logic and-it might be added-the natural rivalry among the services for coequal dignity had seemed to demand separate commanders in chief under the supreme commander for all three services, Air Force, Navy, and Army. The Casablanca declaration on command had provided for three "assistant" supreme commanders for each of the services. Mediterranean operations had in effect practiced the doctrine of three coequal commanders, although the operations themselves were small enough so that the "ground commander" was never more than the equivalent of an army group commander. The issue of a ground commander for OVERLORD was raised in intramural discussions in the Operations Division of the War Department in September and came before the Joint Chiefs, but was shelved there. The concept of a ground commander seemed objectionable on practical grounds. Since the supreme commander would be American, it was considered in September that the ground commander, if there was one, would also have to be American. But, as one officer in OPD pointed out, the "ruling factor" determining in practice the nationality of the ground commander would be the availability of a suitable individual to fill the position. He observed further that no U. S. commander had the battle experience and reputation to challenge the qualifications of the British generals, Montgomery and Alexander, for the job. The conclusion was obvious: it would be impolitic of the Americans to suggest the creation of the job.127
The command organization for OVERLORD, which was at last agreed on in the closing months of the year, was retained in principle after the assault was broadened from a three- to a five-division front. The successive command already established was then phased forward. Two armies (U. S. and British) were then to make the assault under the direct command of 21 Army Group. General Montgomery, commander of the army group,
GENERAL MONTGOMERY, Ground commander of the Allied assault on the Normandy beaches.
was given a de facto ground command for the assault phase-the same kind of unified command that under the COSSAC plan would have been exercised by First Army.128
In late October 1943, the Combined Chiefs of Staff began looking forward to their next staff conference and the issues which they expected to be raised there. Most of the indications were that the Quebec compromises would be reopened and the debate would go on as before.
The Cairo-Tehran Conferences
In proposing an agenda for the conference, the British Chiefs of Staff emphasized their opinion that the most important point for discussion would be a "general review of our future strategy in the light of the events which have taken place since QUADRANT."129 That this meant a further British attempt to shift additional Allied weight to the Mediterranean had already become clear through an exchange of views on the disappointingly slow progress in Italy.
On 24 October, General Eisenhower sent a long cable to the Combined Chiefs of Staff outlining the situation in Italy.130 It was, he said, much changed since the Salerno landings, and the change was seriously and dangerously to the disadvantage of the Allies. At the moment the Allies had eleven divisions in the peninsula. They were opposed by nine German divisions, which, it was estimated, could be reinforced by fifteen more that were located to the north. At the current rate of build-up the Allies would have only sixteen to seventeen divisions by the end of January. There was thus a real possibility that the Allies might lose the critical build-up race and be not only stopped but defeated. What General Eisenhower feared most of all, however, was that, although he might be able to batter his way up the peninsula to Rome, his armies would arrive on their objective too weak to hold it. He could not stand south of Rome because it was important to take the capital as a symbol of victory in Italy, to capture the airfields north of it, and above all to retain the initiative at least until after OVERLORD. Otherwise, the Germans would be free to withdraw divisions from Italy to oppose the Normandy landings and the main purpose of the Mediterranean campaign would be compromised. General Eisenhower concluded that he must keep pushing north; to do so without fatally exhausting his armies he needed landing craft to take advantage "of the enemy's inherent weakness," the exposure of his flanks to amphibious envelopment.
After receiving Eisenhower's report, the British Chiefs of Staff urgently cabled Washington asking that General Eisenhower be backed to the full. Specifically they wanted delay in the agreed schedules for transferring landing craft from the Mediterranean to OVERLORD. They wanted this "even if the 'OVERLORD' programme is delayed."131 For the moment the Joint
Chiefs simply replied that the British exaggerated the seriousness of the Italian situation and reminded them of their agreement to consider that OVERLORD and not the Italian battle was the "primary ground and air effort against Germany."132 But it was clear that strategy was on the block again. Again a relatively minor practical decision, this time involving the convoy schedules for moving landing craft, raised the large issue of how the war in Europe was to be fought.
The landing craft problem was temporarily settled by an agreement that General Eisenhower might retain, until 15 December, sixty LST's scheduled for transfer to the United Kingdom a month earlier.133 But no one doubted that the stopgap settlement was as unreliable as a cork in a volcano. In fact, the cable to General Eisenhower conveying the Combined Chiefs' decision134 had no sooner been drafted than the British representatives began to argue that General Eisenhower had evidently not asked for enough resources, that he should be provided with at least a two-divisional amphibious lift even though he had requested only one.135
There were other straws in the wind that pointed to the resumption of the debate on European strategy. Immediately after the surrender of Italy in September 1943, the British had attempted to pluck off the Dodecanese Islands at the entrance to the Aegean Sea. They thought that the islands would fall readily into the vacuum of the general Italian collapse. German reaction, however, was unexpectedly swift and determined, and Italian assistance to the British, on the other hand, was of even less value than anticipated. As a result, within a few weeks the British had been ejected from their bridgeheads with a loss later estimated by the Prime Minister at 5,000 first-class troops.136 Before this happened, in late September, the commanding generals of the Middle East and Mediterranean theaters had fully examined the possibilities of reinforcing the Aegean attack and saving the forces already committed there. They concluded that the only operation which could reestablish the Allied position was the capture of Rhodes. In view of heavy German air activity and the enemy disposition to resist strongly in the area, this operation was envisaged as requiring a very considerable contribution of resources from the Mediterranean which could be taken only from the Italian campaign. General Eisenhower, further, saw in an Aegean offensive a constant and probably mounting drain on Mediterranean resources which would seriously affect his capacity to prosecute the war in Italy. The other commanders concerned agreed and it was decided to postpone the proposed invasion of Rhodes. At the moment the Combined Chiefs did no more than take note of these decisions.137 Since the operation was "postponed," however, and not canceled, it could be expected to come up again, particularly since the Prime Minister
had always taken a special interest in the eastern Mediterranean.
One of the arguments for pushing operations in the eastern Mediterranean was that success in that area might persuade Turkey to enter the war on the side of the Allies. The project of using diplomatic pressure on Turkey to bring her into the war had been approved in principle by the Combined Chiefs at Casablanca.138 Since that time various diplomatic overtures had been made through the British. Although Turkey remained coy in the face of both implied threats and open blandishments-threats that without active participation in the struggle against the Axis Turkey's position might be weak at the peace table, and direct offers of military assistance if she did come in-it still seemed possible at the end of 1943 that an alliance with Turkey might be achieved. The project was given impetus by reports that the Soviet Union was especially interested in getting direct and immediate support from the Mediterranean area so as to draw off more German strength from the main Eastern Front.139
There were thus three post-Quebec developments that would tend to supply fresh fuel for the British argument for heavier Mediterranean commitments. To these a fourth consideration, indirect and intangible, but nevertheless important, may be added. Red Army successes in the 1943 summer offensive had been impressive. The German armies were backed up to the line of the Dnepr. The Russians did not pause. Bridgeheads were thrown across the river. The Germans fought to avoid annihilation on various parts of the front and in the process fell back as much as a hundred miles in the south. By November the main German armies were split by the Pripet Marshes; Kiev had fallen; the forces in the Ukraine were threatened with encirclement; the Crimea was cut off. Although the German armies gave no sign of collapse, they were fighting to escape destruction and steadily retreating. It seemed impossible that they could any longer hope for victory. In the meantime the tempo of the Allied air offensive from Great Britain had stepped up. There were increasing reports of fading morale among enemy prisoners, partly induced, it was thought, by letters from home describing the devastation rained by Allied planes on German cities. In London hopes were raised that enemy disintegration was near, that the closing months of 1918 were about to be reenacted. There was some feeling that plan RANKIN (the plan for return to the Continent in case of enemy collapse) was more likely to be executed than OVERLORD.140 If collapse were, in truth, so near, then the strategy of throwing everything into the immediate battle in the Mediterranean and so avoiding, perhaps, the necessity of ever having to strike the big costly cross-Channel blow took on a new plausibility.
In the continuing policy discussions during the weeks before the Cairo Conference the Joint Chiefs of Staff found only one new consideration important enough to warrant a possible re-examination of their positive commitment to the OVERLORD concept: that was the suggestion
of a Soviet preference for immediate support via the Mediterranean over the stronger but delayed attack on northwest Europe.141 If the Russians were actually to urge accelerated operations in the Mediterranean, it would be a complete reversal of their previous stand that they wanted a second front, that they wanted it in northwest Europe, and that they did not consider Allied attacks in the Mediterranean as a substitute. Yet in November it seemed quite possible that the Soviet Union might make just this reversal. The circumstances are worth examining in some detail.
In October, U.S. and British military and diplomatic representatives met with the Russians in Moscow to discuss various problems of military collaboration.142 Marshal K. E. Voroshilov put before the conferees three Soviet proposals for hastening the successful conclusion of the war: (1) preparations by the United States and Great Britain during 1943 to insure an invasion of northern France; (2) inducements to Turkey to enter the war on the side of the United Nations; (3) inducements to Sweden to permit the use of air bases for the war against Germany. In answer to the first proposal Maj. Gen. John Russell Deane and General Ismay outlined the conclusions of the Quebec Conference and reassured the Soviets that the decision to mount OVERLORD was firm. General Deane found it worth remarking that these assurances were accepted by the Russians and that the Russians did not press to have the target date on OVERLORD put forward. Still more remarkable to General Deane was their calm acceptance of statements that OVERLORD might have to be delayed. Reporting on this to the Joint Chiefs of Staff, General Deane said that it opened up the possibility that the USSR's attitude on a second front in the west had been altered by recent successes of the Red Army, that the Russians were now laying greater stress on immediate assistance. General Deane's impression was strengthened when Molotov complained that Allied pressure in Italy had been insufficient to prevent the Germans from moving divisions to the Eastern Front. Soviet leaders could not understand why two nations of the combined resources of the United States and Great Britain could not tie up more than a handful of enemy forces.143
General Deane was convinced by all these signs that the Russians were working around to a demand that American and British armies intensify the campaign in Italy or perhaps launch an invasion of the Balkans. He cabled the Joint Chiefs of Staff on 9 November to be prepared for such a demand at the Tehran Conference. He told them that it was quite likely that Russian enthusiasm for a cross-Channel attack had cooled. He believed the Red Army was now confident
of its ability to move into Berlin without benefit of the squeeze from the west.
The prospects of mounting OVERLORD as planned could not have seemed very bright to the Joint Chiefs of Staff as they traveled to Cairo for the first conversations with the British before the meeting with the Russians.144 Shortly after their arrival the Joint Chiefs asked Ambassador John G. Winant for his impressions of British thinking so that they might be prepared for the inevitable debate.145 Mr. Winant began by saying that "he thought the British had no idea of abandoning . . . [OVERLORD] but that they did oppose a fixed date for it." The British, he said, were anxious to mount the operation at the right "psychological moment" and they didn't think that moment could be predicted so far ahead of time. British defeat in the Dodecanese had much upset the Prime Minister and had confirmed him in the opinion that the Germans were still superior to both Great Britain and the United States on land. British military men, Ambassador Winant added, felt that the Prime Minister's views on the Dodecanese defeat were "considerably out of perspective," but they agreed in fearing the German capacity to build up rapidly on the ground to oppose the OVERLORD landings, despite Allied superiority in the air. In short, he believed, the British still doubted that the Allies could attack successfully at a fixed date which was dependent only on the state of Allied preparations; they still wanted to wait for a moment of German weakness.146
Winant's interpretation of British views was confirmed in the course of the conference. Early in the proceedings the Prime Minister emphasized the British commitment to OVERLORD but begged for some "elasticity" in order to expedite Mediterranean operations.147 The necessary elasticity he thought could be achieved by delaying OVERLORD about five or six weeks. Landing craft destined for transfer from the Mediterranean to the United Kingdom could be held in the Mediterranean long enough to carry out amphibious assaults behind the German lines in Italy and undertake the invasion of Rhodes.
The Prime Minister's views were further elaborated and formally presented by the British Chiefs of Staff the following day.148 They proposed the following actions: push the offensive in Italy until the Pisa-Rimini line was reached, nourish guerrilla movements in Yugoslavia, Greece, and Albania, induce Turkey to enter the war, open the Dardanelles supply route to the Soviet Union, and promote chaos in the Balkans. They added that, if these actions meant delaying OVERLORD, then that delay should be accepted.
These proposals, counter to U.S. strategy built around the absolute priority of OVERLORD, should have called forth a debate at least as vigorous as that at Quebec. Actually, the Joint Chiefs of Staff promptly accepted the British program as
a basis for discussion at the Tehran Conference and afterward. The Joint Chiefs' acquiescence did not mean that they had relaxed their conviction of the transcending importance of OVERLORD. In accepting British proposals as a basis for discussion with the Russians, they were underlining the fact that final decision had to take into account Soviet views. They did not thereby agree to argue for the British stand; they intended only to present it in order to elicit Soviet reaction.149
There the matter rested when the preliminary Cairo conferences broke up and the Combined Chiefs of Staff traveled to Tehran for the central and critical meeting with the Russians. The keynote of the Tehran meetings was set by the first plenary session on 28 November 1943 between President Roosevelt, Prime Minister Churchill, and Marshal Stalin.150 Because the main object of the Tehran Conference was to get the Soviet views on strategy as a basis for coordinating United Nations' policy, it was Marshal Stalin who at that first meeting called the tune. He wasted no time making clear the Russian view. After listening to some preliminary oratory by the President and Prime Minister, he cut short his own introductory remarks with a blunt "Now let us get down to business." Within a few minutes he announced all the major elements of USSR strategy as it affected the Western Powers. He declared first that, as soon as Germany had been defeated, the Soviet Union would join with the United States and Great Britain in the offensive against Japan. Announced casually as though it were a point well understood, this was actually his first official assurance of Russian intentions in the Pacific, and it had, as will be seen, profound effect on Anglo-American strategy. Stalin then took up the Italian front. Allied victories there, he thought, had been important, but "they are of no further great importance as regards the defeat of Germany." The USSR, he continued, believed that the most suitable point of attack against Germany was northwest France. Thus in a few sentences he scouted General Deane's prognostications of a change in the Soviet attitude. The Americans were pleased, if somewhat surprised.
The Prime Minister wanted to know whether operations in the eastern Mediterranean to take some weight immediately off the USSR would not be acceptable even though they might mean a delay of a month or two in mounting OVERLORD. Stalin replied that he did not consider it worth while to scatter British and U.S. forces. On the other hand he was very much interested in the suggestion made by both the President and Mr. Churchill that an invasion of southern France was being considered as a diversion for OVERLORD. He did not look on it exactly as a diversion. Simultaneous attacks from the northwest and south appealed to him as a single pincer attack the pattern of which was familiar from many Red Army victories.
Stalin's interest in an assault on southern France caught the U.S. and British
TEHRAN CONFERENCE. Left to right: Marshal Stalin, President Roosevelt, and Prime Minister Churchill.
delegations unprepared.151 The President and Prime Minister told Stalin that no detailed examination of the project had yet been made, but that their staffs would study it. It was a curious approach, possibly calculated to avoid making any firm commitment. But it ignored a large amount of quite detailed planning work already done for the operation. It ignored General Eisenhower's recommendations made to the Combined Chiefs of Staff on 29 October, opposing the operation in view of inadequate resources.152 Most important of all it ignored the fact that the bulk of the planning staffs had been left in Cairo. The only material on hand at Tehran to "start" the study was a copy of the 9 August outline plan which was very much out of date.153 It was on the basis of this old plan that the skeleton planning staff at Tehran drew up a draft memorandum for the President's consideration. Following the August plan, the memorandum recommended a two-division assault with a build-up to ten divisions. There was, however, no reliable information as to the resources likely to be available. On the critical matter of landing craft supplies, the memorandum was necessarily vague; its general conclusion was that there probably would be enough craft "to move two assault divisions by short haul from Corsica and Sardinia to Southern France under favorable weather conditions...."154
On the basis of this study and under considerable urging by Marshal Stalin, the Americans and British before leaving Tehran committed themselves to mounting OVERLORD with a supporting operation against the south of France during May 1944. It was clear that Marshal Stalin considered the two operations as a single, inseparable military undertaking.
This was the Soviet view of what Anglo-American armies should do to hasten the end of the war in Europe. The view was hammered in during the conference, as in every meeting the Soviet representatives, abetted by the Americans, sought to pin down the most unequivocal possible agreement on OVERLORD.155 All other operations in the Mediterranean Stalin waved aside as diversion. "He had no interest in any . . . [Mediterranean] operations other than those into Southern France."156 He admitted the desirability of getting Turkey into the war, but doubted that it could be done. In any case he felt that Turkey's participation was a comparatively unimportant matter. The important point was that he did not wish the Western Allies to contemplate any diversion whatsoever from OVERLORD. OVERLORD was the main question, not Turkey or Rhodes or the Balkans. To make sure that there would be no wavering in the preparations for OVERLORD, Marshal Stalin
urged that a supreme commander be appointed at once. He felt that the operation could not progress until someone had clear and single responsibility for it. Both Stalin and Churchill emphasized that the choice was the President's to make; all agreed that the appointment should be settled within the next fortnight.
The air was cleared when the Americans and British gathered again at Cairo. The question of postponing OVERLORD in favor of Mediterranean diversions was scotched at last. The one big problem that remained alive and would grow from now on more robust and obstreperous was how to carry out the commitment to invade southern France (Operation ANVIL) at the same time as OVERLORD. In agreeing to ANVIL, the Prime Minister had stipulated that the strength of the assault be not less than the two divisions which the Tehran staff had indicated were feasible. But back in Cairo the Combined Planners looked over the landing craft figures and reported that there was actually enough lift for only one division in the assault and two-thirds of a division in the immediate follow-up. Estimating that a two-division assault would require a simultaneous lift for 4,500 men and 7,740 vehicles, planners calculated a probable shortage of lift for 6,500 men and 3,200 vehicles. That deficit might be made good, they reasoned, by diverting from the Pacific one month's allocation of landing craft-or a total of 26 LST's and 26 LCT's-and then taking 5 LCT's from OVERLORD. With reference to the latter recommendation, the planners noted that OVERLORD was already getting 24 more LCT's than had been counted on at the Quebec Conference and that therefore the diversion of these five craft was "not materially at the expense of that operation." The diversion of a month's supply of craft from the Pacific would probably force the postponement of an operation planned against Truk, and they suggested that, if that operation could not be delayed, then it would be necessary to make up the deficit for it by transferring craft from the South Pacific.157
Here was the sign of things to come. The "numbers racket" of shuffling allocations of landing craft around the globe, a half dozen here, a half dozen there, had begun and it would not end until late in 1944.
One possible source for more landing craft was the cancellation of a proposed amphibious assault (BUCCANEER) against the Andaman Islands, in connection with operations to open the Burma Road. The British, who had always, felt that offensive action against Japan could wait until after the defeat of Germany, now took the initiative in urging cancellation. "It appeared," said Air Marshal Portal, "that in order to carry out a successful operation in the South of France, other operations would have to suffer."158 Specifically, he thought, the operation that should suffer was BUCCANEER. The Joint Chiefs of Staff were opposed at first, but after consultation with the President they finally agreed to accept a smaller amphibious assault requiring not more than half the lift. The surplus craft could thus be diverted to Europe.159
In the course of the discussion Admiral King expressed alarm over the tendency to regard the Pacific area as a pool of resources for the European theater.160 But in view of the priority of OVERLORD and ANVIL as the paramount operations of 1944, and the chronic shortage of landing craft, it was in fact inevitable to regard any non-European theater of operations as potentially a stock pile to be raided in order to fatten the ETO. The point was perhaps best made by a cable which General Marshall sent in December to all theater commanders and defense commands. "The landing craft situation is critical," General Marshall wired, "and will continue to be so for some time to come. Any possible increase in production is far behind the increasing demand for landing craft. You are directed to make every landing ship and craft available for and apply them to the maximum battle effort."161 The word went around the world: to Cairo, Algiers, Tehran, Chungking, Southwest Pacific Area, Fort Shafter (Hawaii), Noumea (New Caledonia), Quarry Heights (Canal Zone), Anchorage, and Adak Island (Alaska). Every craft saved was precious; wherever it was on the globe its fate was tied up with the fate of OVERLORD.
The net effect of the reallocation of craft at the Cairo Conference was to give OVERLORD an additional 26 LST's, 24 LCI(L)'s, and 64 LCT's above the allocations set at Quebec; and to give ANVIL an additional 41 LST's, 31 LCI(L)'s, 3 XAP's, and 6 LSI(L)'s.162 These were significant additions but their impressiveness would pale before the increasing demands of the big invasion, as continued planning brought out the military realities of the task to be done.
Table of Contents ** Previous Chapter (2) * Next Chapter (4)
 CCS 95th Mtg, 24 May 43.
 JCS 422/1, QUADRANT, 25 Jul 43.
 See Naval Int Div (Br), History of U-Boat Policy, 1939-1945, 26 Feb 43. Navy Dept files. This consists of translations of extracts from documents in the files of the U-Boat Division of the German Admiralty.
 Ibid. Actually the Allies in February lost 83 ships totaling 420,238 tons. See Commander-in-Chief Atlantic Fleet (U.S. Naval Administration in World War II), MS, p. 463. Navy Dept files.
 Commander-in-Chief Atlantic Fleet, pp. 481ff.
 Berghof Conference, 31 May 43, Office of Naval Intelligence, Fuehrer Conferences, 1943. Fuehrer Conferences is a selection of translated documents from German naval archives.
 Independent sailings were authorized for all ships except those carrying aviation fuel and those slower than eleven knots. See Commander-in-Chief Atlantic Fleet, p. 581.
 Commander-in-Chief Atlantic Fleet, p. 462.
 Figures are from Statistical Summary of Eighth Air Force Operations, European Theater, 14 Aug 1942-8 May 1945, Hq Eighth Air Force, 10 Jun 45.
 Bad weather in April reduced operations so drastically that a comparison of April sorties and bomb tonnage with May figures would be misleading. For a detailed account of operations see Craven and Cate, The Army Air Forces, II, Ch. 10.
 CCS 109th Mtg, 16 Aug 43.
 General der Kavallerie Siegfried Westphal, Der Feldzug in Italien, MS, Ch. IV. Hist Div files. The story of Italian collapse and surrender is treated in full by E. M. Smyth in Sicily and the Surrender of Italy, a volume now under preparation in this series.
 Meetings at Algiers between Churchill, Marshall, Eisenhower, and others, 29 and 31 May 43. Min of TRIDENT Conf.
 CCS 102d Mtg, 16 Jul 43. The debate during July on what to do in the Mediterranean may be followed in the CCS 268 series and in CCS 103d Mtg, 23 Jul 43, and CCS Spec Mtg, 26 Jul 43.
 Rpt, Combined Planning Staff, 15 Jul 43, CCS 268/2; Msg to Eisenhower, 21 Jul 43, CCS 268/5.
 CCS 268/2.
 CCS 103d Mtg, 23 Jul 43.
 CCS Spec Mtg. 26 Jul 43.
 Both the Joint Strategic Survey Committee and the Joint War Plans Committee made thorough examinations of British techniques of argument and ways of meeting them. See JPS 189, Preparations for the Next U.S.-British Staff Conference, 25 May 43; JCS 422/1, QUADRANT, 25 Jul 43; Memos and staff studies in OPD file ABC 337 (25 May 43) passim.
 The British brought a staff of 93 to the Washington Conference in May, and in addition made considerable use of the Joint Staff Mission and diplomatic personnel already in this country. The Americans by contrast took only 56 officers to Quebec. See JPS 189 and various commentaries thereon in OPD file ABC 337 (25 May 43) .
 Cf. Winston S. Churchill, The Grand Alliance (New York, 1950), p. 28; Eisenhower, Crusade in Europe, p. 61.
 Ismay held this position under Churchill all during the war. Before Churchill became Prime Minister, Ismay was secretary of the Chiefs of Staff Committee. Interv, F. C. Pogue with Gen Ismay, 17 Dec 46. Hist Div files.
 See Note by the Minister of Defence, 7 Oct 41, cited in Churchill, The Grand Alliance, pp. 498-500.
 This was the interpretation of both General Barker and General Morgan on the basis of their experience on the COSSAC staff. See Intervs, F. C. Pogue with Barker, 4 Oct 46, and Pogue with Morgan, 8 Feb 47. Hist Div files. It is worth noting in this connection Churchill's impatience with the stodgy military professional which runs through his account of World War I as a kind of leitmotif. See esp. Winston S. Churchill, The World Crisis: 1915 (New York, 1923), pp. 540-46.
 Interv with Gen Ismay, cited n. 22.
 On the question of possible British political interests in the Mediterranean see below, p. 96, and cf. Eisenhower, Crusade in Europe, pp. 194, 284; Stimson and Bundy, On Active Service, p. 447.
 The President was always conscious of this popular feeling. See, for example, Mtg, Roosevelt, Hopkins, and JCS on board ship, 19 Nov 43. Dep C/S file 110.00 A 48-41 dr 15. The President said "that we should not get roped into accepting any European sphere of influence. We do not want to be compelled, for instance, to maintain United States troops in Yugoslavia." See also Memo, Col Gailey (OPD) for Gen Handy, Information from the White House, 30 Jun, quoting portions of Cbl, Roosevelt to Churchill, 29 Jun 44, in reference to the British proposal to divert ANVIL resources to the Adriatic. Roosevelt wrote: "I would never survive even a minor set-back in Normandy if it were known that substantial troops were diverted to the Balkans." OPD files, exec 10, item 71 (Information from the White House) .
 The term "United Nations" to designate the Allies was apparently first used by Roosevelt at the end of 1941. See Sherwood, Roosevelt and Hopkins, p. 458; cf. Ibid., pp. 548-49.
 An exception, of course, was the President's initiative in pushing the North African operation. See above, Ch. I.
 Cf. on this point Stimson and Bundy, On Active Service, pp. 428, 439, and Ch. XVII passim; Sherwood, Roosevelt and Hopkins, p. 446.
 For instance, see JCS 83d and 84th Mtgs, 17, 18 May 43; see also above, Ch. 1.
 JPS 231, Adequacy of the TRIDENT Strategy, 26 Jul 43. Colonel Bessell, Army member of the JWPC, concurred in the memorandum but his concurrence did not represent War Department opinion. See JCS 444/ 1 .
 Copy of JPS 231 (with attached buck slip) in C/S file 381 II.
 Morgan's notes on telephone conversation with Barker, 5 Aug 43. SHAEF SGS file 381 Ia.
 The original paper (JPS 231), which set forth the naval point of view, was never presented to the Joint Chiefs of Staff, but it formed the basis for JCS 444, which did reach them. The latter paper is a slightly toned down version of JPS 231 based on the same strategic principles. The formula at the 7 August meeting was to call for a "reconciling" of the reports of the Navy planners, JCS 444, and the Army planners' rejoinder, JCS 444/1. In fact what happened was that the Army views were accepted in toto and formed the basis for the U. S. Chiefs of Staff memorandum submitted to the British at Quebec as CCS 303. See OPD file ABC 381 Europe ( Aug 43) and ABC 337 (2 May 43).
 On the other hand, General Marshall was scrupulous in letting Admiral King speak first on Pacific matters.
A graduate of the Virginia Military Institute in 1901, Marshall was commissioned in the infantry. During World War I he was assigned to general staff duty and went overseas with the 1st Division, with which he served at Luneville, St. Mihiel, Picardy, and Cantigny. He worked on the plans for the St. Mihiel offensive and then became Chief of Operations of First Army in the midst of the Meuse-Argonne offensive. In May 1919 Marshall was appointed aide to General John J. Pershing and held that assignment for the next five years. In 1936 and again in 1937 he commanded Red Forces in army maneuvers. The following year he came to the War Department first as Assistant Chief of Staff, War Plans Division, and then as Deputy Chief of Staff. After a two months' tour as chief of a military mission to Brazil in 1939, he was appointed Chief of Staff of the Army.
 JCS 104th Mtg, 15 Aug 43.
 SCAEF 12th Mtg, 27 Mar 44. SHAEF SGS file 381 I. Brooke was at the time refusing to make commitments on the invasion of southern France as a diversion for OVERLORD (see Ch. V). He said that "it was not militarily sound to run a war on a lawyer's agreement by making commitments ahead of time, thereby losing the necessary and desired flexibility to meet the ever changing conditions and circumstances."
 See n. 26. It is impossible to assess what force political consideration may have had at any juncture in the debate, for neither the Prime Minister nor the British Chiefs ever put a political argument on the contemporary record. Even at this time (1950) Mr. Churchill apparently still does not defend the Mediterranean strategy on political grounds. See Churchill, The Grand Alliance, pp. 660-61.
 Joint Board Estimate of United States Over-all Production Requirements, 11 Sep 41, JB 355, ser 707. OPD files.
 Both of these estimates were primarily for purposes of planning war production and not mobilization, but they are nonetheless symptomatic of the Army's expansive thinking. Cf. other estimates in early 1942 cited in K. R. Greenfield, R. R. Palmer, and B. I. Wylie, The Organization of Ground Combat Troops (Washington, 1947), p. 198.
 JCS 293, Limited Operations in the Mediterranean in 1943 44, 7 May 43.
 CCS 83d Mtg, 13 May 43.
 The fear of a German move through Spain was frequently expressed. See, for instance, JCS 52d and 57th Mtgs, Jan 43; CCS 151/2,17 Feb 43 and JCS Spec Mtg, 26 Jul 43. Appreciation that such a move was no longer possible was contained in JCS 438/1, JIC Rpt, Estimate of the Enemy Situation, 1943-1944, European-Mediterranean Area, 7 May 43.
 See discussions of the Quebec Conference agenda, CCS 288 series.
 CCS 108th Mtg, 15 Aug 4.
 CCS 303/1, Strategic Concept for the Defeat of the Axis in Europe, 16 Aug 43.
 CCS 303/3, 17 Aug 43. TRIDENT was the code name for the Washington Conference of May 1943.
 JCS 106th Mtg, 16 Aug 43.
 Interv with Gen Ismay cited n. 22.
 Capt Mansergh (COSSAC Naval Staff), Rpt on QUADRANT at COSSAC (43) 23d Mtg, 30 Aug 43.
 1st Plenary (Citadel) Mtg, 19 Aug 43. OPD files, Min of QUADRANT Conf.
 CCS 328/1 Directive to General Eisenhower, 27 Aug 43, Annex V. The origins and development of the southern France operation will be discussed in detail by Maj. J. D. T. Hamilton in Riviera to the Rhine, a volume under preparation in this series.
 Memo, Capt Mansergh for COSSAC, Appreciation of the Effect of the Provision of Extra Landing Craft and Shipping for Operation OVERLORD, 27 Jul 45. SHAEF SGS file (1944 Operations).
 COS (43) 186th Mtg (0), 12 Aug 43.
 Ltr, Morgan (COSSAC) to Maj Gen W. D. Morgan (CofS 21 Army Group), 18 Aug 43. SHAEF SGS file 560 I.
 COSSAC (43) 50, Landing Craft for OVERLORD, 10 Sep 43; Memo, Devers for COSSAC, Support Craft Required per U.S. Assault Division, 15 Oct 43. SHAEF SGS file 560 I.
 COS (43) 175th Mtg (0), 29 Jul 43.
 Memo, Adm Creasy for COSSAC, Provision of Support Craft-Progress Report, 2 Sep 43. SHAEF SGS file 560 I.
 Ltr, King to ComNavEu, 10 Aug 43. SHAEF SGS file 560 I.
 Information supplied by the British Admiralty.
 COS (43) 566 (0), Availability of Landing Craft (Tanks) for OVERLORD, 22 Sep 43.
 Cbl cited at COS (43) 209th Mtg (0), 7 Sep 48.
 COS (43) 96 (0), 30 Sep 43.
 COS (43) 236th Mtg, 4 Oct 43.
 Cbl, Nelson to Wilson, 27 Sep 43, cited in George E. Mowry, Landing Craft and the WPB (Historical Reports on War Administration: WPB Special Study No. 11), rev. ed. (Washington, 1946), p. 29. Wilson was chairman of the Production Executive Committee, War Production Board.
 Mowry, Landing Craft and the WPB, pp. 33-34, 51-52.
 CCS 111th Mtg, 18 Aug 43.
 CCS 314/3, Allocation of Landing Craft (Operation OVERLORD-Vehicle Lift), 20 Aug 43.
 JCS 462/3, Landing Ships and Craft-Means to Increase U.S. Production, 1 Oct 43, with incl., Ltr, Vice Chief of Naval Operations to Chiefs of Bureaus, 13 Sep 43.
 Ltr, Morgan to Barker, 28 Oct 43. Barker Papers. See Bibliographical Note.
 Ltr, Morgan to Hollis (Secy of War Cabinet), 2 Sep 43. SHAEF SGS file 322.011/2.
 Annex I to COS (43) 206th Mtg (0), 3 Sep 43.
 Ltr, Morgan to Devers, 2 Sep 43. SHAEF SGS file 322.011/2.
 CCS 75/3, System of Command for Combined United States British Operations, 21 Oct 42.
 The Chiefs of Staff Conference, ABC-4, JSSC 2, 25 Dec 41. ARCADIA Conf Bk. (Marshall's words at this meeting are quoted in Sherwood, Roosevelt and Hopkins, pp. 455-57.) The principle of unity of command was embodied first in the Directive to the Supreme Commander in the ABDA Area, ABC-4/5, 10 Jan 42. ARCADIAID Conf Bk. The directive assigned command over "all armed forces, afloat, ashore, and in the air," that belonged to the four participating powers (Australia, the Netherlands, United Kingdom, and United States).
 In July General Barker had written: "It can be accepted as an absolute certainty that the PM would not, for one moment, allow the assault to be made wholly by American troops. The same is true with relation to the U.S. Government. We must be practical about this and face facts." Memo, Barker for Morgan, 4 Jul 43. Barker Papers.
 See, however, below (pp. 110-11) for citation of the Salerno experience in a different connection.
 COS (43) 416 (0), Operation OVERLORD, Report and Appreciation, 30 Jul 43, Part I, par. 40. SHAEF SGS file 381 la.
 At the first meeting of the COSSAC staff, 17 April 1943, Morgan announced that Casablanca had decided the invasion was to be commanded by a British supreme commander with an American deputy.
 COS (43) 206th Mtg (0), 3 Sep 43.
 COS (43) 525 (0), Operation "OVERLORD"-Command and Control, 11 Sep 43.
 COS (43) 217th Mtg (0), 16 Sep 43.
 COS (43) 138th Mtg (0), 26 Jun 43. See above, Ch. II.
 CCS 113th Mtg, 20 Aug 43.
 CCS 304/2, Directive to the Supreme Allied Commander for Operation "Overlord," 12 Oct 43.
 CCS 304/3 and 304/4, 19 Oct 43. All the paper in the 304 series bear the title as given in note 87 except 304/1 and 304/4. The subject of these latter two is "Command and Control for Operation 'Overlord'." 8 See below, pp. 112-14
 See below, pp. 112-14.
 Cbl, Eisenhower to Marshall, W8967, 31 Dec 43. WD Cable log.
 Minute, 24 Oct 43, Annex II to Min, COS (43) 209th Mtg (0), 25 Oct 43.
 Ltr, Devers to Morgan (unsgd), 4 Sep 43; Memo, Comments on General Devers' Letter on Plan for Command and Control of Operation "OVERLORD." SHAEF SGS file 322.011/2.
 OPD file ABC 381 (22-1-43) sec. 1.
 The Supreme Allied Commander.
 Draft Memo, Col Roberts (JWPC) for General Handy for CofS, 18 Oct 43. OPD file ABC 384 Europe (5 Aug 43) sec. la.
 Ltr, organ to Barker, 28 Oct 43. Barker Papers.
 Ltr, Barker to Morgan, 3 Nov 43. Barker Papers.
 Ltr, Morgan to Barker, 8 Nov 43. Barker Papers.
 CCS 304/4, 19 Oct 43.
 CCS 124th Mtg, 22 Oct 43.
 CCS 304/7, 4 Nov 43. This is a "split" paper. The first five paragraphs of the draft directive inclosed were agreed on by U.S. and British planners. The final paragraphs, dealing with the control of the strategic air forces, were reported in two version representing opposing U.S. and British views.
 Cbl, Br COS to JSM, 10 Nov 43. SHAEF SGS file 322.011/2.
 CCS 126th Mtg, 5 Nov 43.
 The directive was issued through COSSAC, saving the letter of SAC's right to appoint his subordinates.
 COS (43) 717 (0), Formation of an Allied Expeditionary Air Force, 17 Nov 43. Air Defence of Great Britain replaced Royal Air Force Fighter Command. "Despatch by Air Chief Marshal Sir Trafford Leigh-Mallory" (submitted to the Supreme Allied Commander in November 1944), Fourth Supplement to The London Gazette No. 37838, 31 December 1946, pp. 37-38.
 After the paper referred to in note 101 (dtd 4 Nov 43), the next CCS memorandum on command (CCS 304/8) was dated 6 January 1944.
 For full text of this directive see below, App. B.
 Interv with Gen Ismay, cited n. 22; cf. Stimson and Bundy, On Active Service, p. 439.
 Sherwood, Roosevelt and Hopkins, p. 615.
 Stimson and Bundy, On Active Service, p. 439.
 Ltr, Col Edwards (OPD) to Gen Edwards (CofS ETOUSA), 5 Sep 43. Pre-Inv files, Devers Correspondence. Colonel Edwards wrote: "The newspapers have just announced and rather firmly, that SAC is to be Gen. Marshall and suggest that Gen. Eisenhower may be C/S [Chief of Staff of the Army]. This should clear the atmosphere."
 Interv with Gen Ismay, cited n. 22.
 See the account in Katherine T. Marshall, Together: Annals of an Army Wife (Atlanta, 1946), pp. 69ff.
 CCS 408, Command of British and U.S. Forces Operating Against Germany, 25 Nov 43; CCS 126th Mtg, 5 Nov 43. During a meeting with the Joint Chiefs of Staff on shipboard on the way to Cairo, President Roosevelt said that it was his idea "that Marshall should be commander-in-chief against Germany commanding all British, U.S., French and Italian troops." JCS Mtg, 15 Nov 43. Dep C/S file 110.00 A 48.41 dr 155.
 CCS 408/1, 26 Nov 43.
 Ltrs, Devers to Marshall, 18 and 19 May 43. Pre-Inv files, Devers Correspondence.
 Ltr, Devers to Marshall, 6 Jul 43. Pre-Inv files, Devers Correspondence.
 Ltr, Barker to Devers, 30 Aug 43. Pre-Inv files, Devers Correspondence.
 For Eisenhower's appraisal of Bradley's qualifications, see Eisenhower, Crusade in Europe, p. 215.
 See Cbls, Marshall to Eisenhower, No. 5968, 25 Aug 43; No. 6595, 1 Sep 43; No. 6904,.4 Sep 43. WD Cable log.
 Cbl, Devers to Marshall, W4421, 13 Sep 43. WD Cable log.
 Marshall's views on command in Europe are stated in Cbl, Marshall to Devers, R3267, 18 Sep 43; Dir to Devers, 24 Sep 43. OPD Misc Exec office file.
 Cbl, AGWAR to ETOUSA, 18 Sep 43. Pre-Inv files, Devers Correspondence.
 Ltr, Marshall to Devers 24 Sep 43. OPD file ABC 381 (22-1-43) sec. 1.
 Cbl, Marshall to Devers, R5583, 11 Nov 43. WD Cable log.
 Ltr, COSSAC to CinC 21 A Gp, Operation 'OVERLORD,' 29 Nov 43, COSSAC (43) 76.
 Ltr, Col Edwards (OPD) to Gen Edwards, 7 Sep 43. Pre-Inv files, Devers Correspondence.
 Montgomery, however, never received the formal designation of ground commander, and there actually remained some confusion, at least on his part, as to the duration of his command over ground forces on the Continent. The ground commander problem will be fully discussed by F. C. Pogue in The Supreme Command, a volume under preparation in this series.
 CCS 125th Mtg, 29 Oct 43.
 Cbl, Eisenhower to CCS, NAF 486, 24 Oct 43. WD Cable log.
 CCS 379, Operations in the Mediterranean, 26 Oct 43.
 CCS 379/1, 29 Oct 43.
 CCS 379/5, 4 Nov 43. At the Cairo Conference, the date for transfer of the craft was extended to 15 January 1944. CCS 132d Mtg, 30 Nov 43; Cbl, CCS to Eisenhower, FAN 281, 1 Dec 43. WD Cable log.
 Cbl, CCS to Eisenhower, FAN 271, 5 Nov 43. WD Cable log.
 CCS 126th Mtg, 5 Nov 43.
 2d Plenary Mtg at Cairo, 24 Nov 43. OPD files, Min of SEXTANT Conf.
 See CCS series 365, Future Operations in the Eastern Mediterranean.
 See above, Ch. I.
 Cbl, Deane to JCS, No. 51, 9 Nov 43. WD Cable log.
 Ltr, Barker to Morgan, 20 Oct 43. Barker Papers; JCS 533/7, Recommended Line of Action at Next U.S.-British Staff Conference, 18 Nov 43.
 Min of Mtg, Roosevelt, Hopkins, and JCS, l9 Nov 43, cited n. 27. Cf. below, n. 144.
 The principal U.S. representatives were Secretary of State Cordell Hull, Ambassador W. Averell Harriman, and Maj. Gen. John R. Deane. Deane, following the conference, was to establish a U.S. military mission to the Soviet Union. The British were represented by Foreign Secretary Anthony Eden and Gen. Sir Hastings Ismay
 General Deane reported in detail to the Joint Chiefs of Staff in a series of cables in October and November 1948. See esp., Cbls No. 2, 19 Oct; No. 4, 21 Oct; No. 28, 29 Oct; No. 84, 31 Oct; No. 47, 6 Nov; No. 51, 9 Nov. WD Cable log. The story of the conference is briefly but accurately told in Deane's book: John R. Deane, The Strange Alliance (New York, 1947), pp. 16ff., 85.
 See Mtg, Roosevelt, Hopkins, and JCS, 19 Nov 43, cited n. 27. At that meeting Marshall expressed the opinion that the British might like to "ditch" OVERLORD now in order to go into the Balkans. Roosevelt thought the Russians might press for an attack up the Adriatic to the Danube to tie up with Russian forces entering Romania.
 JCS 127th Mtg, 22 Nov 43.
 2d Plenary Mtg, 24 Nov 43.
 CCS 409, "OVERLORD" and the Mediterranean, 25 Nov 43.
 CCS 131st Mtg, 26 Nov 43.
 This and the following narrative of the Tehran (EUREKA) Conference are from the official U.S. minutes of the meeting, bound with the minutes of the Cairo (SEXTANT) Conference. OPD files. The records kept by the U.S. and British secretaries of the plenary political-military meetings at Tehran were not afterward coordinated. For purposes of the discussion here, however, the discrepancies are unimportant and the U.S. minutes have been used throughout.
 This can readily be deduced from the records and was specifically stated by Colonel (now Brig. Gen.) Bessel in conversation with the author, 5 March 1947. It is further confirmed by statements of General Ismay to Dr. F. C. Pogue, in the interview cited above, note 22.
 Cbl, Eisenhower to CCS, NAF 492, 29 Oct 43. WD Cable log.
 JPS 249, Plan for Invasion of Southern France, 9 Aug 43.
 Unnumbered Memo, Operation Against Southern France, 29 Nov 43. OPD file ABC 384 Europe (5 Aug 43) sec. 9a.
 It is clear that the Soviet views had force in the conference because they were also American views. For a plausible analysis of President Roosevelt's role at Tehran see Sherwood, Roosevelt and Hopkins, pp. 789ff.
 Plenary Mtg, 29 Nov 43. Statement not reported in British minutes.
 CCS 424, Amphibious Operations Against the South of France, 5 Dec 43.
 CCS 133d Mtg, 3 Dec 43.
 This decision and its impact on strategy in Burma as well as on U.S.-China relations will be discussed in full by C. F. Romanus and R. Sunderland in Stillwell's Command Problems, a volume under preparation in this series.
 CCS 135th Mtg, 5 Dec 43; JCS 132d Mtg, 28 Nov 43.
 Cbl, R6820, 11 Dec 43. WD Cable log.
 CCS 428 (rev), Relation of Available Resources to Agreed Operations, 15 Dec 43.