Page 255

Chapter 10

OVERLORD Versus the Mediterranean at the Cairo-Tehran Conferences

by Richard M. Leighton

(See Chapter 8 for information on the author.)

The long debate between U.S. and British leaders over the strategy of the European war reached a climax and a turning point at the great mid-war conferences at Cairo and Tehran late in 1943. Since the decision to invade North Africa, a year and a half earlier, the debate had focused on the war in the Mediterranean, the British generally advocating a bold, opportunistic strategy, the Americans a more cautious one. On the surface, they had disagreed on specifics rather than fundamentals. Few on the American side advocated complete withdrawal from the Mediterranean, and U.S. leaders were as quick as the British to respond to the opportunity offered by the disintegration of Italian resistance in early summer of 1943. They opposed the British primarily on the choice of objectives, especially east of Italy. For their part, the British never questioned the principle that the main attack against Germany in the West, and the decisive one, must eventually be made from the northwest (OVERLORD) not the south. In the meantime, they argued, aggressive operations in the Mediterranean were not merely profitable but even essential in order to waste the enemy's strength and to contain and divert enemy forces that might otherwise concentrate on other fronts. But the debate was embittered by American suspicions that the British intended somehow to sidetrack, weaken, or indefinitely postpone the invasion from the northwest, subordinating it to peripheral and indecisive ventures in the Mediterranean that would serve their own long-range political ends. Since the British consistently disclaimed such intentions, the issue of OVERLORD versus the Mediterranean could not be debated as such-and, indeed, cannot now be proved even to have existed outside the minds of the Americans. For them, nevertheless, it was the real issue, and the question actually debated at Cairo and Tehran,

Page 256

whether OVERLORD should be postponed a few weeks in order to permit certain small-scale operations in the eastern Mediterranean, was only the shadow. From the American point of view, the great achievement of the conferences was not the compromise reached on the latter question-essentially a technical matter, worked out on the staff level-but the decision of the Big Three-Roosevelt, Stalin, and Churchill-to make OVERLORD and its southern France complement, ANVIL, the supreme effort of the Western Allies against Germany in 1944. After Cairo-Tehran, Mediterranean strategy continued to be a source of friction, but American leaders seemed to consider the cross-Channel invasion as assured and the issue of OVERLORD versus the Mediterranean as closed. [1]

Tools of Amphibious War

In November 1943 Allied military fortunes were high. On the Eastern Front Soviet armies had crushed the Germans' summer of-

[1] (Footnote 1 is essentially a list of suggested readings and has been moved to the end of the file to enhance readibility.)

Page 257

[Map 7]

Page 258

fensive against Kursk before it had got well under way, and had launched a series of powerful counteroffensives which by late November had driven the enemy across the Dnieper, isolated the Crimea, and, farther north, pushed almost to the Polish border. British and Americans in the Mediterranean had swept through Sicily in July and August, forced the capitulation of Italy, and invaded the peninsu1a in September, bogging down finally in the mountains south of Rome. (See Map 7.) The strategic bombing offensive against the German homeland continued with mounting intensity, despite heavy losses, and in early autumn the hitherto lagging build-up of American invasion forces in the British Isles swelled to massive proportions. In the Pacific war, the New Guinea offensive had reached the Huon Peninsula with the capture of the important enemy base at Finschhafen, the South Pacific campaign had advanced to Bougainville in the northern Solomons, and the push across the central Pacific had begun with fiercely contested landings on Tarawa and Makin in the Gilberts. In the Atlantic, the U-boats had been decisively defeated, and shipping losses reduced to negligible proportions. On all fronts except Italy and Burma the Allies were advancing. Germany's defeat was now predicted for October 1944, and the American planning committees had been ordered to produce a scheme for ending the war against Japan within the year following. Optimism ran high. (See Map I, inside back cover.)

Behind this optimism lay the realization that men and materials would be available on the scale needed to sustain and quicken the momentum already gained. The immense weight of Soviet manpower and industry, after more than two years of mobilization, was now making itself felt; Soviet armies, backed by masses of reserves and munitions, now had a capacity for sustained offensive warfare that the enemy no longer possessed. In the West, Britain's war effort had passed its peak, with armed forces fully deployed and manpower and industrial capacity fully engaged. American armed forces, on the other hand, though approaching the limits of their planned expansion, were still mostly uncommitted; the bulk of the U.S. Army was at home, waiting to be deployed overseas, and only a handful of divisions had actually seen action. American war industry by now had achieved a productivity that in many categories of munitions actually surpassed foreseeable needs. Military supply programs were already being cut back, and pressures were building up to expand production of civilian goods. The most spectacular achievement of American war production was in shipbuilding. American shipyards in this year poured out 19.2 million deadweight tons of merchant shipping, which was more than two and one-third times as much as had been built in 1942. Early

Page 259

in August, the Joint Chiefs of Staff informed the chairman of the War Production Board that they no longer expected merchant shipping to be the bottleneck of the overseas war effort. [2]

Only one category of supply-landing craft-threatened seriously to limit Allied strategy in 1944. At Cairo and Tehran, indeed, the apparent necessity of choosing between a postponement of OVERLORD and abandonment of planned or proposed amphibious operations elsewhere was dictated by the shortage of landing craft-more particularly, of one type of landing vessel, the Landing Ship, Tank (LST). Less than 300 LST's were in existence in November 1943, almost all built in the United States. Of these, 139 were in the Mediterranean-67 of them allocated to the British under lend-lease-and, except for a small contingent, were all earmarked for transfer to the United Kingdom for OVERLORD as soon as the amphibious phases of the Italian campaign were completed. For OVERLORD, in addition, the United States had agreed the preceding spring to provide 62 more new LST's during the coming winter. The remaining new production of LST's was allocated to the war in the Pacific. [3]

Production of LST's and other landing ships and craft in the United States had been late in getting under way, reaching large volume only in the winter of 1942-43 and then rapidly falling off. This first wave of production, aimed originally at the now discarded plan for a cross-Channel invasion in spring of 1943 (ROUNDUP), had proved generally adequate, together with the smaller output of British factories and yards, to meet the rather modest needs of Allied amphibious operations before mid-1943.4 Even the invasion of Sicily, in some respects the most massive amphibious operation of the entire war (eight divisions were landed simultaneously), was adequately mounted without drawing upon a substantial reserve of U.S. assault shipping in

[2] (1) Frederick C. Lane, Ships for Victory: A History of Shipbuilding 

Under the U.S. Maritime Commission in World War II (Baltimore, Md.: 

Johns Hopkins Press, 1951), pp. 601-05. (2) Gerald J. Fischer, A 

Statistical Summary of Shipbuilding Under the U.S. Maritime Commission 

During World War II (Washington, 1949), Table A-4. (3) Leighton and 

Coakley, Global Logistics and Strategy, 1943-1945, MS, Ch. X.

[3] A few LST's of special design were constructed in Britain early in 

the war. In origin, the LST was a British-designed vessel, like 

virtually all the landing ships and craft used in World War II. Under 

wartime agreements, the United States constructed most of the merchant 

and amphibious shipping used by both countries, thus enabling Britain, 

with limited building capacity, to concentrate on expanding its Navy.

For figures on distribution of LST's in November 1943 see CCS Memo for 

Info 175, 23 Nov 43, Landing Craft Reports, 1 Nov 43.

[4] SLEDGEHAMMER, the tentatively planned emergency cross-Channel attack 

in fall of 1942, had been canceled, partly because of a shortage of 

assault lift, but for other reasons as well.

Page 260

the Atlantic or interfering with planned deployments of craft to the Pacific. [5]

From the Navy's point of view the whole landing craft production program had been undertaken at the worst possible time-when the Navy was straining to rebuild sea power destroyed or immobilized at Pearl Harbor and in later engagements, in order to gain supremacy in the Pacific, while at the same time trying to break the strangle-hold of enemy submarines upon the sea lanes in the Atlantic. The program competed with many other lines of war production for materials, above all for the steel, engines, and facilities needed to build other types of combatant vessels. A Navy official commented bitterly in April 1943 that the high rate of landing craft construction achieved late in 1942 had been obtained "only by cutting across every single combatant shipbuilding program and giving the amphibious program overriding priority in every navy yard and every major shipbuilding company. The derangement . . . will not be corrected for about six months." [6] As landing craft schedules were terminated or cut back that winter and spring, the Navy pushed the building of escort vessels to meet the revived menace of the German U-boats in the Atlantic, which in March reaped a harvest of more than a million deadweight tons of Allied shipping. [7] Navy officials candidly wanted no more emergency landing craft programs.

By August 1943, however, pressures were building up to increase the output of landing craft, at a time when, as a result of the abatement of the submarine menace, the Navy was cutting back its escort and antisubmarine vessel programs. While the landings on Sicily had been successful, with losses lower than anticipated, the greater part of the entire amphibious fleet in the Mediterranean was tied up for weeks after the initial landings moving supplies over the beaches and performing other administrative tasks. Other amphibious undertakings were in prospect in the Mediterranean, in southeast Asia, and on the two main avenues of advance toward Japan across the Pacific. The biggest prospective deficit of amphibious shipping, however, loomed in the planned cross-Channel assault, then scheduled, as a result of decisions at the Washington Conference of May 1943, for

[5] (1) Richard M. Leighton and Robert W. Coakley, Global Logistics and 

Strategy, 1940-1943, UNITED STATES IN WORLD WAR II (Washington, 1955), 

pp. 376-82, 602-03, 682-83. (2) George Mowry, Landing Craft and The War 

Production Board (WPB Study No. 11, July 1944), Ch. II. (3) Jeter A. 

Isley, and Philip A. Crowl, The U.S. Marines and Amphibious War 

(Princeton: Princeton University Press, 1951), pp. 1-4, 47-48. (4) 

Harrison, Cross-Channel Attack, pp. 60-61. For a description of the 

various types of landing ships and craft, see ONI 226, Allied Landing 

Craft and Ships (Office, Chief of Naval Operations, April 1944).

[6] JPS 152/1, 3 Apr 43, title: Production of Landing Craft.

[7] Leighton and Coakley, Global Logistics and Strategy, 1943-1945, 

MS, Chs. I and III.

Page 261

the spring of 1944. For this undertaking, against what was expected to be the most heavily defended coast line in the world, the assault as then tentatively planned was to be on a scale of only about three and a half divisions with two more afloat, a limit imposed arbitrarily by the predicted availability of assault shipping. At the Quebec Conference of August 1943 Prime Minister Churchill bluntly called for an immediate increase in American landing craft production, pledging at the same time a maximum effort in his own country, in order to strengthen the OVERLORD assault. The same demand was voiced on the American side in various quarters.

In September and again late in October, the Navy with JCS approval ordered large increases in landing craft production. Navy officials framed the new program, however, with an eye to the war in the Pacific, not the war in Europe. Large segments of the program were devoted to new types of vessels mainly adapted to warfare in the Pacific-notably armored amtracks (LVT's) and the new LCT(7), actually a small landing ship-inevitably at the expense of the older types desperately needed in Europe. Nor would the increases become effective in time to help OVERLORD if the operation were carried out early in May 1944. Allocations of new production to OVERLORD were limited to about three months' output, at current low rates, at the end of 1943. To this Admiral Ernest J. King, on 5 November, promised to add something less than a month's production of LST's, LCI(L)'s, and LCT's, and not all of these seemed likely to arrive in the United Kingdom in time to be used in the invasion. [8]

A month before the Allied leaders assembled at Cairo, U.S. representatives at a conference of foreign ministers in Moscow pointed to the new landing craft program as indisputable proof that the long-postponed second front would be opened the following spring. In Washington, by contrast, the British were being warned privately that no more new landing craft would be forthcoming for OVERLORD. [9] It was the latter assumption that shaped the options laid before the conferees at Cairo and Tehran.

Alarums and Excursions

American military leaders and their staffs, on the eve of the Cairo-Tehran Conferences, were in a mood to force a showdown on the strategy of the European war. [10] As they viewed it, Allied strategy

[8] Ibid., MS, Chs. X, XI.

[9] Ibid.

[10] This section is condensed from Leighton and Coakley, Global 

Logistics and Strategy, 1943-1945, MS, Chs. VI through X. For a good 

summary, see Harrison, Cross-Channel Attack, Ch. II.

Page 262

since the decision to invade North Africa had been drifting steadily away from the northwestern Europe orientation, agreed on in April 1942, and into a peripheral line of action that could only end in stalemate. Preparations for a cross-Channel invasion in spring of 1943 had been suspended, the British Isles almost denuded of American troops, and American resources had been diverted into the development of a new line of communications and a new invasion base in North Africa. The decision at Casablanca in January 1943 to attack Sicily had ensured that the Mediterranean would continue to be the main theater in Europe during 1943 and that no cross-Channel invasion could be attempted until 1944 at the earliest. Since then British persuasion and the ineluctable logic of momentum had drawn the Allies deeper into the Mediterranean-into Tunisia, Sicily, Italy, Sardinia, Corsica-and a long, uphill struggle still loomed ahead in Italy. Most alarming to the Americans was the persistent effort of the British to broaden the Mediterranean front eastward-by pressure on Turkey to enter the war, by proposals to seize ports on the Dalmatian coast and to step up aid to the Balkan guerrillas, and, most recently, by an ill-advised incursion into the Dodecanese Islands which had cost the British several thousand troops killed and captured and untold loss of prestige. Persistent dabbling by the British in this region raised, in American minds, the dread specter of military operations and political involvement in the Balkan peninsula, a land of inhospitable terrain, primitive communications, and turbulent peoples.

In the light of developments in the Mediterranean, American military leaders discounted the repeated pledges of loyalty by the British to the cross-Channel invasion strategy. They tended to gloss over or ignore the immense investment Britain already had in the cross- Channel operation, the heavy contributions of British shipping to the build-up of American invasion forces and material in the United Kingdom (almost half the entire tonnage used), and the persistent pleas of British leaders for a strengthening of the OVERLORD assault. It was widely believed in American official circles that British leaders feared to come to grips with the German Army on equal terms, that they were haunted, as Secretary of War Stimson put it, by the "shadows of Passchendaele and Dunkerque." [11] Army staff officers, wrestling with the paradox, could only conclude that the great Anglo-American invasion army amassing in the United Kingdom was intended by the


[11] (1) Henry L. Stimson and McGeorge Bundy, On Active Service in Peace 

and War (New York: Harper & Brothers, 1948), pp. 435-38. (2) Bryant, 

Turn of the Tide, pp. 573-76.

Page 263

British to be "a gigantic deception plan and an occupying force" after the expected German collapse. [12]

In recent weeks American suspicions of British intentions had quickened. At the Quebec Conference in August, the OVERLORD plan prepared by the Anglo-American planning staff in London had been accepted by both sides with little discussion. But the British had rejected an American demand that OVERLORD be given an "overriding" priority over operations in the Mediterranean. Prime Minister Churchill, then and subsequently, had stressed, to a degree that aroused American misgivings, the stipulations written into the OVERLORD plan to the effect that additional forces would be needed if German strength in France exceeded certain levels. As yet, the Americans had not made an issue of this point, since the JCS had approved the stipulations along with the plan. Then, late in October, a crisis had suddenly developed in Italy when it appeared that the Germans were winning the build-up race south of Rome and might soon be in a position to launch a crushing counterattack. Churchill and the British Chiefs had reacted with what the Americans considered unjustified alarm. The British had proposed the temporary retention in the Mediterranean of all assault shipping earmarked for OVERLORD and, more disturbing, had intimated that if the situation in Italy continued to deteriorate it might be necessary to postpone OVERLORD beyond its present target date of 1 May 1944. After some discussion, it had been agreed that sixty-eight OVERLORD LST's should remain in the Mediterranean until mid-December, as the theater commander, General Dwight D. Eisenhower, had requested, to help mount an amphibious turning movement around the enemy's right flank south of Rome. But the JCS and their staffs were still worried, as the time for the Cairo meetings approached, over the implied British threat to OVERLORD. For the British Chiefs had bluntly warned that they intended, at the forthcoming conference, to bring up for reconsideration "the whole position of the campaign in the Mediterranean and its relation to OVERLORD." [13]

Finally, early in November, the Americans received a disturbing hint of the role the Soviet leaders might play at Tehran. At the foreign ministers' conference in Moscow late in October, Marshal Joseph V. Stalin had displayed a lively interest in the operations of his allies in the Mediterranean, and, to the astonishment of the Western repre-

[12] OPD paper [about 12 Nov 43], U.S. Courses of Action in Case Sextant 

Decisions Do Not Guarantee OVERLORD, Exec 5, Item 12a. See also various 

staff studies in ABC 381 Strategy Section Papers (7 January 1943), 

Numbers 131-95.

[13] Memo, Representatives COS, in CCS 379, 26 Oct 43, Opns in 

Mediterranean.

Page 264

sentatives, had reacted with bland unconcern when British Foreign Secretary Anthony Eden had hinted that, owing to the worsening situation in Italy, it might be necessary to postpone OVERLORD. After the conference, Maj. Gen. John R. Deane, head of the U.S. military mission in Moscow, had been bombarded by complaints from the Soviet staff about Allied inaction in the Mediterranean. Deane had concluded from all this that the Soviets, as he informed Washington, "want to end the war quickly and feel they can do it," and therefore were less interested in OVERLORD, six months or more distant, than in immediate action to draw German strength from the Eastern Front. Deane warned his superiors to expect a Soviet demand at the forthcoming conference for a greater effort in the Mediterranean, including "some venture in the Balkans," even if that meant delaying OVERLORD. [14] Deane's warning caused a furor in Washington, where it was apparently taken at face value. Combined with the disturbing hints from London of an impending attack on OVERLORD, it conjured up nightmarish visions of a concerted Russo-British demand at Tehran for a major shift of effort to the Mediterranean-or worse, to the eastern Mediterranean-at the expense of OVERLORD. [15]

First Cairo: The Options

At Cairo, the Americans found that the British, too, were ready for a showdown. "It is certainly an odd way of helping the Russians," declared Churchill after a scathing review of recent setbacks in the Mediterranean, "to slow down the fight in the only theatre where anything can be done for some months." [16] The British Chiefs of Staff seized the initiative with a blunt criticism of American insistence on the "sanctity of OVERLORD."

We must not ... regard OVERLORD on a fixed date as the pivot of our whole strategy on which all else turns. In actual fact, the German strength in France next Spring may, at one end of the scale, be something which makes OVERLORD completely impossible and, at the other end, some-


[14] (1) Msg 51, Deane to JCS, 9 Nov 43, with related papers in Exec 5, 

Item 15, Env. 3. (2) John R. Deane, The Strange Alliance (New York: The 

Viking Press, 1947) p. 35. (3) Churchill, Closing the Ring, pp. 286-89. 

(4) Cordell Hull, Memoirs of Cordell Hull, (New York: The Macmillan 

Company, 1948), Vol. II, p. 1301. (5) Ehrman, Grand Strategy, V, 100-

101, 156-157.

[15] Leighton and Coakley, Global Logistics and Strategy, 1943-1945, MS, 

Ch. XI.

[16] Quoted in Churchill, Closing the Ring, pp. 332-33. The Cairo-Tehran 

meetings lasted from 22 November through 7 December 1943. At Cairo, 

Roosevelt, Churchill, and their military advisers met formally for the 

first time with Chiang Kai-shek. The Tehran meetings (28 November-1 

December) brought the two Western leaders together with Premier Stalin 

for the first time. A second series of meetings, attended by British and 

Americans only, was held at Cairo on 2-7 December. See Matloff, 

Strategic Planning for Coalition Warfare, 1943-1944, Ch. XVI.

Page 265

thing which makes RANKIN not only practicable, but essential. ... This policy, if literally interpreted, will inevitably paralyze action in other theatres without any guarantee of action across the Channel. ... It is, of course, valuable to have a target date to which all may work, but we are firmly opposed to allowing this date to become our master.

...

They were prepared, they asserted, to carry out the cross-Channel invasion "as soon as the German strength in France and the general war situation gives us a good prospect of success," but they insisted that unless the Allies pursued an aggressive course of action in the Mediterranean during the coming winter and spring, such conditions were unlikely to develop. [17]

The ominous implications of this manifesto were hardly borne out, however, by the British Chiefs' concrete program for the Mediterranean. They wanted to advance beyond Rome only as far as the Pisa-Rimini line (the same limit the JCS had in mind); to extend more aid to the Balkan partisan forces in the form of weapons, supplies, technical assistance, and commando raids; to try to bring Turkey into the war before the end of the year; and, with Turkish consent, to open the Dardanelles (shortest route to the USSR) to Allied shipping. The opening move, provided Turkey's support were assured, would be an attack about February 1944 on the largest of the Dodecanese Islands, Rhodes, which commanded the approaches to the Aegean and the Straits. Finally, control of the whole Mediterranean area would be unified under a British commander. (This last the Americans were already prepared to concede, in return for the appointment of a U.S. commander for OVERLORD.) In short, the British hoped by means of a major effort in Italy and what Ambassador Winant called "bush-league tactics" east of Italy to force the Germans back along the entire Mediterranean front. By the Prime Minister's reckoning, the eastern Mediterranean operations would involve not more than a tenth of the combined British and American resources in the whole theater. But, while all the troops and other means needed were available in the area, the landing ships and craft were scheduled for early transfer to the United Kingdom. To retain them for the required time might mean postponing OVERLORD as much as six weeks or two months-that is, until about 1 July 1944. [18]

As far as Italy and the Balkans were concerned, the U.S. Chiefs had no quarrel with these proposals. They even saw certain advantages in gaining Turkey as an active ally, provided the price paid

[17] (1) CCS 409, Note by COS, 25 Nov 43, OVERLORD and the 

Mediterranean. (2) Ehrman, Grand Strategy, V, 109-12. RANKIN was one of 

several alternative plans for crossing the Channel in the event of a 

German collapse before the OVERLORD target date.

[18] (1) CCS 409, cited n. 17(1). (2) Ehrman, Grand Strategy, V, 104-21, 

165-67. (3) Min, 2d Plenary Mtg, SEXTANT, 24 Nov 43.

Page 266

for intervention was strictly limited. But they doubted the ability of the Turks to hold their own if attacked by the Germans, and felt no enthusiasm for another try at Rhodes, so soon after the recent disaster. Moreover, the American staffs challenged the timetables and requirements of the British plan at many points. They doubted whether the Rhodes operation could be fitted into the LST movement schedules, even if OVERLORD were postponed to 1 July. [19]

The British, however, had an alternative proposal: the necessary assault shipping for the Rhodes operation might be taken from south-east Asia. In August a new Allied command had been set up in southeast Asia under Vice Adm. Lord Louis Mountbatten, embracing Burma, Ceylon, Siam, the Malay Peninsula, and Sumatra (but neither China nor India). Since then the basic divergence of British, American, and Chinese purposes in the area, not to mention the differences within each camp, had been sharpened. British aspirations looked primarily south and southeast, toward a restoration of Britain's prewar possessions and influence in Malaya and the East Indies. The Americans were more interested in increasing China's effectiveness as an ally and in gaining bases in China for bombing and, ultimately, invading -Japan. For the British, therefore, Burma was a stage on the road to Singapore and beyond-one that might, perhaps, be bypassed-while for the Americans it lay on the route over the Himalayas into China. Although construction had begun early in 1943 on a road from Ledo, in India's Assam Province just over the Burmese border, to connect with the old Burma Road where it crossed into China, contact between China and her allies depended for the present on the airlift. Throughout 1943 supplies delivered over the Hump from India to China each month had not exceeded, on the average, what could be carried by a single medium-sized freighter. Competition for this trickle of cargo was fierce. Lt. Gen. Joseph W. Stilwell, Mountbatten's American deputy and commander of the U.S. China-Burma-India theater (lying partly within Mountbatten's command), wanted to use the supplies mainly to equip Chinese forces in China in order to help in the reconquest of northern Burma, scheduled to begin early in 1944 before the onset of the spring monsoon. Maj. Gen. Claire L. Chennault, commanding the U.S. air forces in China, believed the airlift should be greatly expanded and devoted entirely to support of an air offensive against Japanese communications in China and with the home islands. Chennault's program, which promised quicker results at lower cost than Stilwell's long-range plan of regenerating Chinese armies and restoring land communications with China, appealed both to the

[19] The discussion of these points is described in detail in Leighton 

and Coakley, Global Logistics and Strategy, 1943-1945, MS, Ch. XI

Page 267

President and to Chiang Kai-shek, though the latter naturally demanded an airlift large enough to support both programs. Since the preceding spring the bulk of supplies brought over the Hump had in fact gone to Chennault's air forces. However, Roosevelt's broader aims for China coincided with Stilwell's. His purpose in inviting Chiang to Cairo (over British objections) was, in part, to discuss further economic and military aid-which was imperative, Chiang said, if China were to continue fighting-and, in part, to enlist his co-operation in the forthcoming Burma offensive.

The general plan of this offensive was to launch converging drives into northern and central Burma-by British-Indian forces from the west, by Stilwell's American-trained Chinese from the northwest, and by Chiang's own armies from southern China. Subsidiary features of the plan included a British naval demonstration in the Bay of Bengal by fleet units released from the Mediterranean after the Italian surrender, and an amphibious operation-although where this would be carried out was still somewhat uncertain. A year earlier Stilwell's plan for the reconquest of Burma had included a major amphibious attack on Rangoon; at the Washington Conference of May 1943 this had been scrapped, for lack of assault shipping, in favor of smaller landings near Akyab and on Ramree Island, along the Burma coast just below the Indian border. To mount these operations, a contingent of half a dozen attack transports, eighteen LST's, and a number of smaller craft, had been sent to India, arriving there finally, after a protracted hold-over in the Mediterranean, early in the fall. Churchill, meanwhile, had come out strongly for an "Asiatic-style TORCH" in the form of a surprise descent on the northern tip of Sumatra (CULVERIN), which the Americans opposed as eccentric to the main effort and his own advisers thought would require more resources than were available. The theater commanders, finally, had proposed a more modest substitute in the form of landings on the Andaman Islands, southwest of Rangoon, in March or April 1944. This operation (BUCCANEER) had been tentatively endorsed by both the British and the U.S. Chiefs of Staff, though it, along with the remainder of the whole plan, still awaited formal approval. BUCCANEER was, then, the amphibious part of the general plan (CHAMPION) submitted to Chiang at Cairo. [20]

Immediately the plan ran into heavy weather. Hardly anyone, in fact, had much enthusiasm for BUCCANEER, except perhaps Chiang,


[20] (1) Ehrman, Grand Strategy, V, 148-53. (2) Romanus and Sunderland, 

Stilwell's Command Problems, Ch. II. (3) [Mountbatten] Report to CCS by 

Supreme Allied Commander, South-east Asia, 1943-1945 (London: H.M. 

Stationery Office, 1951), p. 27. (4) Matloff, Strategic Planning for 

Coalition Warfare, 1943-1944, Ch. XIV and Ch. XVI, pp. 2-3.

Page 268

who had not been informed of its objective, but who independently suggested the Andamans as a suitable objective. Its most serious defect was that it seemed to have little connection with the mainland operations it was intended to support, and hardly represented a threat serious enough to provoke a strong reaction. The U.S. Chiefs of Staff preferred it to CULVERIN, but were not committed to any particular operation; Admiral King himself favored a landing on the mainland near Moulmein with a view to cutting across the isthmus to Bangkok, but such an undertaking was not thought feasible with the assault shipping available. Evidently the most that could be said for BUCCANEER was that it would provide a base for future amphibious landings on the mainland and for bombing the new Bangkok-Moulmein railroad, which gave the Japanese in Burma direct overland connections with the Gulf of Thailand. [21] Churchill made no secret of his distaste for BUCCANEER and had earlier declared that if he could not have CULVERIN he would send the British assault shipping back to the Mediterranean. At Cairo he expanded on the idea: if the Americans would not accept CULVERIN, and if they refused to postpone OVERLORD the few weeks necessary to carry out the attack on Rhodes and move assault shipping back to the Mediterranean, then why not take the shipping needed for Rhodes from southeast Asia? BUCCANEER might be postponed rather than canceled. As Churchill remarked, "There really cannot be much hurry. The capture of the Andamans is a trivial prize compared with Rhodes, and also it can be undertaken at any time later in the year." [22]

That Churchill was willing to entertain the idea of doing BUCCANEER at all, despite his candidly expressed scorn for the operation, was the result of the position taken at Cairo by the Chinese Generalissimo. Chiang immediately branded the whole Burma plan (CHAMPION) as inadequate. As a price for his participation in a more ambitious one, moreover, he demanded an immediate increase in the airlift far beyond the capacity of available transport aircraft and explicit guarantees from the British that the land operations would be supported simultaneously by major co-ordinated naval and amphibious attacks. The unreasonableness of the airlift demand, and the arrogance shown by Chiang's subordinates in discussing it with the Western military leaders, caused the latter to close ranks, and drove even Gen-

[21] (1) Mountbatten Report, p. 27. (2) Romanus and Sunderland, 

Stilwell's Command Problems, p. 51. (3) Ernest J. King and Walter Muir 

Whitehill, Fleet Admiral King, A Naval Record (New York: W. W. Norton 

& Company, Inc., 1952), pp. 509-10. (4) Ehrman, Grand Strategy, V, 

162. (5) Min, CCS 129th Mtg, 24 Nov 43. 

[22] Msg, PM for CsofS Com, 21 Nov 43, quoted in Churchill, Closing the 

Ring, p. 686. (2) Ehrman, Grand Strategy, V, 114, 159. (3) Min, 2d 

Plenary Mtg, SEXTANT, 24 Nov 43. (4) Romanus and Sunderland, Stilwell's 

Command Problems, p. 66.

Page 269

eral Marshall to exasperation. [23] A moderate increase in the airlift was ordered, but the Chinese were told unequivocally that they must choose between an offensive in Burma and expanded ferry operations, since both competed for transport aircraft. As for BUCCANEER, the U.S. Chiefs of Staff did not at first take a strong stand, agreeing to postpone action pending decisions yet to be taken on the broader strategy of the war against Japan and the British role in it. In the CCS, therefore, Chiang's demand for an amphibious operation was carefully and noncommittally "noted," with a promise merely of future "consideration." Churchill, however, sharply challenged Chiang's view of the interdependence of the naval and amphibious phases of CHAMPION and the land operations. He pointed out that, in the absence of accessible bases and because of the time needed to refit and redeploy British naval forces from the Mediterranean, direct naval support could not be provided in the forthcoming spring campaign, even though he could promise that by March strong naval forces would be operating in the Bay of Bengal. Finally, he told Chiang emphatically that no definite undertaking to carry out an amphibious operation in conjunction with the land campaign could be given. [24]

Chiang thus faced defeat on all his demands. Early in the afternoon of the 25th he agreed to the CHAMPION plan as drawn, with the sole stipulations that the British should gain naval superiority in the Bay of Bengal-which Churchill had already promised-and that the plan should include an amphibious operation-to which Churchill was willing to agree if the Americans met his own conditions in the Mediterranean. At the same time, however, Chiang was demanding that President Roosevelt give him something to show for having attended the conference. [25] The President obliged. On the same afternoon he told Stilwell and Marshall he had decided, as a further concession to Chiang, greatly to enlarge the program of equipping Chinese divisions, and, some time on the same day, he seems to have given Chiang a pledge that BUCCANEER would be carried out on the scale and at the time planned. [26]

[23] (1) See Marshall's outburst quoted in Joseph W. Stilwell, Theodore 

H. White, ed., The Stilwell Papers (New York: William Sloane Associates, 

Inc., 1948), p. 255. (2) Min, 129th Mtg CCS, 24 Nov 43. (3) Min, 130th 

Mtg JCS, 25 Nov 43. 

[24] (1) Churchill, Closing the Ring, p. 328. (2) Ehrman, Grand 

Strategy, V, 162, 164-65, 571. (3) Min, CCS 128th Mtg, 23 Nov 43. (4) 

Romanus and Sunderland, Stilwell's Command Problems, p. 65. (5) Min, 1st 

Plenary Mtg, SEXTANT, 23 Nov. 43. 

[25] See Marshall's remark at the JCS meeting earlier in the day. Min, 

JCS 130th Mtg, 25 Nov 43. 

[26] The evidence on this last point is strong but not conclusive. 

Churchill (Closing the Ring, page 328) and Leahy (I Was There, page 201) 

both assert unequivocally that the pledge was given and Ehrman accepts 

this as fact (Grand Strategy, Volume V, page 165). Matloff (Strategic 

Planning for Coalition Warfare, 1943-1944, Chapter XVI) regards it as at 

least highly probable. Romanus and Sunderland, who give little attention 

to the amphibious phases of the war in Burma, do not mention the pledge, 

though they do mention the promise to equip more Chinese divisions. It 

may be significant that the President, in the interview with Stilwell 

and Marshall mentioned above, seemed from his remarks to have the 

Andamans operation on his mind. But the most convincing evidence is to 

be found, as shown below, in the abrupt change in the attitude and 

position of the JCS on the morning of 26 November.

Page 270

The President's pledge left his military chiefs very little room for maneuver. If the Soviet leaders at Tehran should insist, as the Joint Chiefs fully expected them to insist, on an immediate major effort by the Western allies in the Mediterranean, with or without OVERLORD, approval of the British program seemed assured. The assault shipping allotted for BUCCANEER now sacrosanct, could not be made available for Rhodes. If OVERLORD shipping currently in the Mediterranean were used instead, how could it be replaced in time to meet the OVERLORD target date? New American production after January was allotted to the Pacific, and Admiral King bristled at the suggestion of further inroads on this source. The only remaining possibility seemed to be to postpone OVERLORD a few weeks as the British had proposed, thus giving more time to redeploy assault shipping and, incidentally, making available another one or two months' production of landing craft. As Admiral Leahy remarked, the problem was brutally simple: the JCS had to decide whether or not they could accept a delay in OVERLORD; if they could not, "the problem appeared insoluble." [27]

The President, at least, had been thinking of delay. Back in Washington Justice James F. Byrnes, Director of the Office of War Mobilization, had received on the 23d a "very urgent" message from him inquiring whether the output of landing craft could be increased, by means of an overriding priority, during the first five months of 1944-an inquiry that made sense only under the assumption that OVERLORD might be postponed beyond 1 May 1944. Byrnes' reply, dispatched on the 25th, indicated that substantial increases might be possible in April and later, but virtually none before then. Roosevelt probably knew, therefore, when he promised Chiang an amphibious operation, that if OVERLORD were postponed to July, it could be bolstered by the addition of some twenty-two new LST's, not to mention ten more now allocated but unlikely to reach the United Kingdom in time for a May assault-and this without encroaching on existing Pacific allocations of February and later output. [28]

Final decision had to wait, then, until the Russians showed their hand. At the last Cairo meeting with the British (on the 26th) before

[27] Min, JCS 131st Mtg, 26 Nov 43.

[28] (1) Msg, FDR to Byrnes, Dir OWM, 23 Nov 43, Exec 5, Item 14. (2) 

Msg, Byrnes to President, 25 Nov 43, in JCS Memo for Info 171, 27 Nov 

43, ABC 561 (30 Aug 43). (3) Mowry, Landing Craft and the WPB, pp. 30-

32.

Page 271

going to Tehran, the U.S. Chiefs of Staff stressed the sanctity of BUCCANEER, but said little about OVERLORD or the Mediterranean. Sir Alan Brooke asked them whether they understood that "if the capture of Rhodes and Rome and Operation BUCCANEER were carried out, the date of OVERLORD must go back." Marshall assured him they did. Would it not be better, urged Brooke perplexedly, to postpone BUCCANEER rather than OVERLORD? What if the Russians should demand both a strong Mediterranean offensive and an early OVERLORD? The situation had become embarrassing. Finally, Admiral Leahy blurted out a broad hint: the U.S. Chiefs of Staff "were not in a position to agree to the abandonment of Operation BUCCANEER. This could only be decided by the President and the Prime Minister." There was little more to say. The Americans accepted the British program as a basis for discussion at Tehran, but on the contradictory assumption that it "would in no way interfere with the carrying-out of Operation BUCCANEER." The British left with the distinct impression, as Lt. Gen. Sir Hastings L. Ismay reported to the Prime Minister, that the Americans, now rigid against any tampering with BUCCANEER, contemplated a postponement of OVERLORD with equanimity." [29]

Tehran: Enter ANVIL, Compromise on OVERLORD

At the opening general meeting at Tehran, on 28 November, the three principals, at Stalin's brusque suggestion, promptly got down to business. Roosevelt noted in his opening remarks the possibility that OVERLORD might have to be postponed "for one month or two or three," and spoke of the various operations in the Mediterranean that were being considered to relieve enemy pressure on the Eastern Front-in the Aegean, at the head of the Adriatic, and in Italy. OVERLORD, he pointed out, would draw away more German divisions than any of these, and he urged that, if possible, it not be delayed "beyond May or June." Churchill presented the British case, elaborating on the promising opportunities that could be exploited in the eastern Mediterranean without detriment either to the campaign in Italy or to OVERLORD. How would the Soviet Union, he asked, regard this prospect "even if it meant as much as about two months' delay in OVERLORD? [30]

Up to this point the atmosphere had been cordial. To the pleased surprise of the Westerners, Stalin opened his remarks with an almost casual promise that the Soviet Union. would intervene in the war

[29] (1) Min, CCS 131st Mtg, 26 Nov 43. (2) Ehrman, Grand Strategy, V, 

166-67.

[30] Min, 1st Plenary Mtg, EUREKA, 28 Nov 43.

Page 272

against Japan as soon as Germany was defeated. This confirmed and strengthened the more tentative offers the Soviet marshal had made on earlier occasions. But his next words brought the discussion abruptly to a tense climax. He declared bluntly that the whole Mediterranean program appeared to him to involve an excessive dispersion of forces. OVERLORD should be made the "basic" operation for 1944, and all other operations, however attractive, regarded as diversions. He saw only one useful possibility in the Mediterranean, an attack on southern France (which Churchill had mentioned in his opening remarks) followed by a drive northward toward an eventual junction with the main OVERLORD forces-the classic pincers formula, which the Russians had applied so often in their own theater. Why not, he suggested blandly, suspend the Italian campaign immediately in order to release forces for this operation, and then launch OVERLORD two or three months later? [31]

General Marshall must have been reflecting sardonically, while Stalin was dropping his bombshell, on an innocent remark he himself had made in a meeting of the Joint Chiefs that morning, to the effect that the Soviet demands, whatever they might be, "would probably simplify the problem." [32] Whatever the reasons for the sudden evaporation of Stalin's recently displayed interest in Mediterranean operations, and for his return to the old familiar insistence on a "second front"-perhaps because, with Soviet armies now at a standstill in the Ukraine, a grand convergence on southeastern Europe no longer promised quick victory-his proposals immeasurably complicated what had been an essentially simple, if baffling, dilemma. The Mediterranean was a "going" theater of war in which the Western allies had a heavy investment. To stop short on the present line in Italy would be almost as repugnant to the Americans as to the British, and Churchill promptly and emphatically asserted that from the British point of view the capture of Rome was both strategically and politically imperative. Stalin seemed, moreover, not to have grasped the limiting role of shipping and landing craft, or the central problem of timing and sequence that grew out of it. He had to be reminded that the troops in the Mediterranean, except for seven divisions already in transit to the United Kingdom, were irrevocably bound there for lack of shipping to deploy them elsewhere. He missed the point that the southern France operation and the landings in the Adriatic had been suggested as mutually exclusive alternatives and that the Rhodes operation was very modest in scope. When Churchill reminded him of this last fact, he conceded that on those terms the capture of

[31] (1) Ibid. (2) Ehrman, Grand Strategy, V, 174-76. 

[32] Min, 132d Mtg JCS, 28 Nov 43.

Page 273

Rhodes might be worthwhile. But if both the Rome and Rhodes operations were to be carried out, or even only the latter, how could a landing in southern France, two or three months before OVERLORD (for which Stalin had stipulated no date) be worked into the schedule-unless OVERLORD was postponed? [33]

At this juncture the President, who had been silent during the above exchange, suddenly interposed. Stalin's proposals, he said, had raised a serious problem of timing. A choice must be made: either undertake Churchill's Aegean operations, which would delay OVERLORD a month or two, or, as the Soviet premier had suggested, "attack [southern] France one or two months before the first of May and then conduct OVERLORD on the original date." His own preference, Roosevelt added, was for the latter alternative. [34]

Churchill was caught off balance. Nothing in the President's earlier remarks had suggested any intention to insist on adherence to the 1 May target date for OVERLORD. He had, in fact, seemed to accept the idea of postponement, urging only that it be brief. His military advisers, by the end of the Cairo meetings, had seemed resigned to the inevitability of some delay. But by implying now that Stalin himself had demanded a 1 May date (which he had not, in fact, done), the President evidently hoped to enlist his support. If so, it was an adroit maneuver, for Stalin failed to challenge the President's implication. Its significance was not lost on Churchill, who immediately protested against the idea of condemning twenty or more divisions in the Mediterranean to inactivity "solely for the purpose of keeping the May date for OVERLORD," and chided the President for the "rigid timing" of the program he had proposed. [35]

The Russians had shown their hand. For the Americans the nightmare of an Anglo-Soviet demand for a shift to the Mediterranean had been dissipated in the comforting assurance that the Soviet leaders once more stood solidly for the primacy of OVERLORD and shared the American aversion for operations in the eastern Mediterranean. It quickly became clear, moreover, that the Russians also shared American suspicions as to British motives, for in the course of the next day (the 29th) both Churchill and Brooke, under Soviet grilling, were repeatedly obliged to go through the ritual of affirming their loyalty to OVERLORD. [36] At a meeting of the military representatives on this

[33] (1) Ehrman, Grand Strategy, V, 175. (2) Churchill, Closing the 

Ring, p. 355.

[34] (1) Min, 1st Plenary Mtg, EUREKA, 28 Nov 43. (2) Ehrman, Grand 

Strategy, V, 176. [Italics supplied.]

[35] (1) Min, 1st Plenary Mtg, EUREKA, 28 Nov 43. (2) Ehrman, Grand 

Strategy, V, 176.

[36] (1) Min, Military Mtg, EUREKA, 29 Nov 43. (2) Min, 2d Plenary Mtg, 

EUREKA 29 Nov 43. (3) See also Churchill's account of Stalin's attack on 

General Brooke at the banquet on the evening of the 30th, Closing the 

Ring, pages 386-88.

Page 274

same day the Soviet leaders indicated no very specific notions as to what should be done in the Mediterranean or when. When Sir Alan Brooke pointed out the risk that a landing in southern France so long in advance of OVERLORD might be crushed before OVERLORD could get under way, the Soviet representative, Marshal Klementy Voroshiloff, merely reiterated rather woodenly his master's statement that the operation would be a valuable complement to OVERLORD. Anyway, he added, Stalin did not insist on a southern France operation. All other undertakings in the Mediterranean, "such as Rome, Rhodes, and what not," were diversions that, if carried out at all, should be "planned to assist OVERLORD and certainly not to hinder it." Evidently the Soviet leader intended to let his allies squabble unhindered over Mediterranean strategy. According to Voroshiloff, however, Stalin did insist on OVERLORD-and "on the date already planned." [37]

Thus the issue was finally joined on the timing of OVERLORD. On this same 29 November Roosevelt, now committed to fight for a May OVERLORD and evidently confident that with Soviet support he could win, sent a message to Washington tardily instructing Justice Byrnes to call off the proposed speed-up in landing craft production, since "the increase in critical types . . . does not become effective enough to justify change in present construction programs." [38] At the plenary meeting that afternoon, Stalin set forth his position in the language of an ultimatum. He also pressed for an early appointment of a commander for the operation. Soviet forces, he promised, would match the invasion from the west by a simultaneous offensive from the east. [39] Churchill held the floor for most of the session with a spirited defense of the British Mediterranean program. He vainly tried to draw out Stalin on his proposal for a southern France operation, for which, as he pointed out, no plan had yet been drafted, and he warned, as Brooke had already done, that if the attack were too weak or launched too early, it would invite disaster. If, on the other hand, a two-division amphibious lift could be left in the Mediterranean, bright possibilities opened up-turning movements along the Italian coasts, then a swift

[37] Min, Military Mtg, EUREKA, 29 Nov 43.

[38] (1) Quoted in Mowry, Landing Craft and the WPB, p. 31. See also pp. 

32-33. Actually, the program had already been accelerated, in response 

to Roosevelt's message of the 23d, since certain measures had to be set 

in train immediately, without waiting for the President's order. (2) See 

Leighton and Coakley, Global Logistics and Strategy 1943-1945, MS, Ch. 

X.

[39] Min, 2d Plenary Mtg, EUREKA, 29 Nov 43. This was hardly a 

compelling argument for a May OVERLORD since the Russians, as the 

British pointed out, had never launched a summer offensive in that 

month. According to Churchill, Stalin told him at lunch on the 30th that 

he wanted OVERLORD in May or in June in order for it o synchronize with 

the Soviet offensive. In the event, the latter jumped off on 23 June, 

two and a half weeks after OVERLORD. See Churchill, Closing the Ring, 

pp. 380, 383.

Page 275

capture of Rhodes, and finally an invasion of southern France in conjunction with OVERLORD. This might mean setting back OVERLORD by six to eight weeks, or-and Churchill here introduced the alternative for the first time at Tehran-the needed assault shipping could be brought back from India. At all events, Churchill concluded, if the handful of vessels needed for Rhodes could not somehow be found, it was unreasonable to suppose that the larger number required for an invasion of southern France or any other diversionary operation in support of OVERLORD could be provided. He reminded the Soviet premier that OVERLORD could not be undertaken at all unless there was a reasonable expectation of success based on certain specified conditions of enemy strength. This brought from Stalin his celebrated query: Would OVERLORD be ruled out if there were thirteen instead of twelve mobile German divisions in France and the Low Countries on D Day? Churchill assured him it would not. [40]

Stalin made no effort to answer Churchill's arguments. He ignored the allusion to BUCCANEER, restated his demand for a May OVERLORD, and indicated his preference for a southern France invasion two or three months before OVERLORD; if this proved impossible, the operation might be launched simultaneously with OVERLORD or even a little later. All other operations in the Mediterranean he regarded as diversions. Roosevelt finally interposed to suggest a date for OVERLORD "certainly not later than 15 or 20 May, if possible." Stalin chimed agreement. Churchill promptly and emphatically dissented, and the atmosphere again became tense. Finally, the problem was referred to the military representatives to work out before the next afternoon, when final decisions would be reached. [4l]

Despite the appearance of a deadlock, the germ of a compromise had already emerged. Both Stalin and Roosevelt had refrained from demanding a 1 May date. Before lunch the next day (30 November) Churchill decided to agree to a date sometime in May, and the British Chiefs of Staff came to the meeting with their American opposites that morning with specific proposals worked out on this basis. [42] General Eisenhower would be allowed to keep the sixty-eight OVERLORD LST's in the Mediterranean until 15 January in order to ensure the early capture of Rome. This meant, by British calculations, that OVERLORD could not be earlier than June-but the British Chiefs were will-

[40] (1) Min, 2d Plenary Mtg, EUREKA, 29 Nov 43. (2) Churchill, Closing 

the Ring, p. 371. (3) Ehrman, Grand Strategy, V, 179.

[41] (1) Min, 2d Plenary Mtg, EUREKA, 29 Nov 43. (2) Churchill, Closing 

the Ring, p. 370. (3) Ehrman, Grand Strategy, V, 180. (4) Sherwood, 

Roosevelt and Hopkins, p. 788. The British accounts indicate that both 

Roosevelt and Stalin gave the OVERLORD date as "in May," or words to

that effect.

[42] (1) Churchill, Closing the Ring, p. 376. (2) Ehrman, Grand 

Strategy, V, 181.

Page 276

ing, in order to satisfy Stalin, to define this as "in May." They were also prepared to support an operation against southern France and, most important, to agree that no assault shipping earmarked for OVERLORD should be retained in the Mediterranean specifically for the Rhodes operation. The key to this last concession lay in their final proposition: as a result of Stalin's momentous pledge on the 28th to enter the war against Japan after Germany's defeat, they argued, the role of China in the coalition had been automatically reduced, and the whole case for an offensive in Burma in spring 1944, including BUCCANEER, had been weakened. The British now hoped, in short, to persuade the Americans to cancel BUCCANEER and send its assault shipping back to the Mediterranean, where it could be used to help mount the southern France operation-and, as a likely by-product, the attack on Rhodes as well. If the Americans refused to cancel BUCCANEER, the burden would be upon them to find the assault shipping for southern France elsewhere, leaving the same probability that it could also be used for Rhodes. [43]

Meanwhile, the U.S. Chiefs of Staff, feeling confident in the assurance of Soviet support, had worked out their own position. The assault shipping already in the Mediterranean could be safely kept there until mid-January, as General Eisenhower had asked, to support the Italian campaign, without endangering an early May OVERLORD. With what remained after the withdrawals, the staff estimated, it would be possible to mount a two-division assault against southern France (now labeled ANVIL). This operation, for tactical and strategic reasons, should be launched no earlier than three or four weeks before OVERLORD rather than on the date suggested by Stalin. But the later date would not leave time, the Americans emphasized, to shift any of the landing craft over to the eastern Mediterranean for an attack on Rhodes and get them back to Corsica in time to refit for the ANVIL landings. Ergo-no Rhodes operation. The problem, as Admiral Leahy triumphantly summed up, "seemed to be a straightforward one of the date of OVERLORD."

The argument that a southern France operation would be feasible but a Rhodes operation would not hinged on logistical calculations of an extremely speculative character. While these calculations, involving forward projections of landing craft availability, could not be positively disproved at the time-although the British challenged them

[43] Ehrman, Grand Strategy, V, 181.

[44] (1) Min, 132d Mtg, CCS, 30 Nov 43. (2) Harrison, Cross-Channel 

Attack, p. 125. (3) James D. T. Hamilton, Threat to Southern France, in 

draft MS on the southern France operation, OCMH. The Americans had 

little reliable data on the southern France operation with them at 

Tehran. See Matloff, "The ANVIL Decision: Crossroads of Strategy," 

below.

Page 277

at every point-the case for the ANVIL landings seemed particularly flimsy. Sir Alan Brooke could cite against it the verdict of General Eisenhower, a month earlier, that the assault shipping remaining in the Mediterranean would suffice for no more than a one-division lift, that the build-up following the assault would be very slow, and that no attack on such a scale would be likely to succeed. [45] The British did not believe the OVERLORD shipping could be moved back to the United Kingdom, after mid-January, in time for a May D Day, and they feared, incidentally, that the landing craft allotted for OVERLORD were inadequate.

Caught between contradictory logistical estimates, the discussion deadlocked. Nevertheless, the afternoon deadline was at hand, and the Russians had to be given an answer. The military leaders agreed, therefore (falling back on the subterfuge suggested by the British), that the Russians could be told "we will launch OVERLORD during May, in conjunction with a supporting operation against the south of France on the largest scale that is permitted by [available] landing craft," with a target date, for planning purposes, the same as that for OVERLORD. The advance in Italy would continue as far as the Pisa-Rimini line, and the sixty-eight LST's requested by Eisenhower would be left in the Mediterranean until 15 January. The fate of BUCCANEER and the Aegean operations was reserved for discussion at Cairo. [46] Thus the difficult questions of timing and provision of means raised by Stalin's bombshell on the 28th were left unanswered, and no breath of discord ruffled the meeting of the principals on the afternoon of the 30th, when the military decisions were ratified. It was inconceivable, Churchill declared, "that the two nations, with their great volume of production, could not make the necessary landing craft available." [47]

Whatever else they did, the Tehran decisions did not spell defeat for the British program in the Mediterranean. The heart of that program-capture of Rome and advance to a defensible line beyond-now seemed assured, even though it had been the first target of Stalin's attack. The modest proposals for the Balkans had been accepted. American opposition had centered on the Aegean operations, for which Admiral King had warned he would not under any circumstances turn over American landing craft. [48] Nevertheless, the prospects of

[45] (1) Min, 132d Mtg CCS, 30 Nov 43. (2) Msg NAF 492, Eisenhower to 

CCS, 29 Oct 43, quoted in Ehrman, Grand Strategy, V, MS, pp. 188-89.

[46] (1) Min, 132d Mtg CCS, 30 Nov 43. (2) Ehrman, Grand Strategy, V, 

182. (3) Msg FAN 281, CCS to Eisenhower, 1 Dec 43, Exec 3, Item 13.

[47] Min, 3d Plenary Mtg EUREKA, 30 Nov 43.

[48] (1) Min, JCS 131st Mtg, 26 Nov 43. (2) See Hopkins' strong 

statement in Sherwood, Roosevelt and Hopkins, pages 793-96.

Page 278

mounting the attack on Rhodes had been immeasurably improved by the introduction of an assault lift requirement for ANVIL, and Stalin had supported the British view, which was written into the formal conclusions of the Tehran Conference, that Turkey should, if possible, be brought into the war before the end of the year. The attitude of the Turks themselves, on which the whole enterprise would depend, was soon to be tested anew in negotiations at Cairo. At all events, the British left no doubt in American minds that they intended to press forward with their Aegean plans, and that, in view of Stalin's firm pledge of participation in the war against Japan, they now regarded BUCCANEER as fair game. [49]

Second Cairo: Scratch BUCCANEER

Back at Cairo, the CCS faced the task of finding enough assault lift to carry out (1) a late May or early June OVERLORD, (2) a simultaneous southern France operation, (3) the British attack on Rhodes, as soon as possible following the impending landings south of Rome, and (4) BUCCANEER, still scheduled for March. The British promptly renewed their attack on BUCCANEER. The operation was now even more vulnerable than before, since the receipt of Mountbatten's most recent plan which provided for a considerably stronger assault, with increased requirements for assault shipping and carrier-borne aviation. The ends in view seemed hardly commensurate with the cost, for more than 50,000 troops were to be concentrated against a garrison estimated at only about 5,000. The British insisted, moreover, on debating the larger issue of the whole campaign in Burma, which, in view of American plans for the Pacific and Stalin's firm promise to enter the lists against Japan, seemed to them to make little sense. The JCS were mainly worried lest, in the absence of an Allied offensive in Burma, the Japanese might seize the initiative and overwhelm the precariously defended American air bases in China. But they found it difficult to defend BUCCANEER on its merits. General Marshall candidly admitted that if the operation could be dropped without wrecking the mainland campaign, "he personally would not be seriously disturbed." [50]

Whatever the defects of BUCCANEER, the JCS were, of course, no more inclined than before to release its assault shipping if the craft

[49] CCS Memo for Info 165, 2 Dec 43, Military Conclusions of the EUREKA 

Conference.  It was also noted that Stalin had undertaken to attack 

Bulgaria if the latter attacked Turkey.

[50] (1) Min. CCS 135th Mtg, 5 Dec 43. (2) Ehrman, Grand Strategy, V, 

185-86. (3) Romanus and Sunderland, Stilwell's Command Problems, pp. 65-

67. (4) Morton and Morgan, The Pacific War: Strategy and Command, The 

Road to Victory, MS, Ch. I.

Page 279

were to be used to mount an attack on Rhodes. But the British adroitly shifted ground. They now soft-pedaled their Aegean plans (which depended mainly on the outcome of negotiations with the Turks, anyway), and concentrated on the problem of mounting an adequate attack against southern France, to which the Americans were firmly committed. ANVIL, they argued, must not be tailored to the leavings of other undertakings (as implied in the Tehran formula), but should be made strong enough to form a genuine complement to OVERLORD. This meant an assault by at least two divisions, perhaps three. But when the staffs checked their hasty Tehran estimates against the more ample data available at Cairo, they found that the residual assault lift in the Mediterranean, after OVERLORD withdrawals, would not exceed one and two-thirds divisions and might be even less. After a halfhearted attempt to hew to the Tehran line, the JCS conceded the need for at least a two-division assault, and on 4 December Admiral King, in a surprise move, offered to meet the ANVIL assault shipping deficit from new production previously allotted to the Pacific. [51]

King's offer opened no breach in the opposition of the JCS to the Rhodes operation, since, as he made clear, the new ships and craft could not reach the Mediterranean in time to be used for it. On the other hand, although they almost covered the calculated deficit against a two-division ANVIL assault, they did not guarantee this operation. [52] They left no margin for unforeseen contingencies, and many on the American as well as on the British side considered even a two-division assault too weak. There was, moreover, growing uneasiness over OVERLORD'S own weakness, even after the allocations of 5 November. Time was growing short. OVERLORD and ANVIL were now designated the supreme operations for 1944; the responsible command-

[51] (1) Min, CCS 133d Mtg, 3 Dec 43. (2) Min, 3d Plenary Mtg, SEXTANT, 

4 Dec 43. (3) CPS 131/1, 3 Dec 43, Amph Opns Against South of France. 

(4) Msg 10131, Adm Badger to VCNO, 5 Dec 43, Exec 5, Item 13. (5) CCS 

424, Rpt by CPS and CAdC, 5 Dec 43, Amph Opns Against South of France. 

(6) Ehrman, Grand Strategy, V. 184, 187, 195.



King's offer was accompanied by a warning that it might result in 

setting back the operation against Truk, the main Japanese base in the 

Caroline Islands. He did not mention that the JCS had already decided, 

some three weeks earlier, to suspend the attack on Truk pending the 

results of carrier raids to test whether it might be feasible to bypass 

the fortress. See Min, JCS 123d Mtg, 15 Nov, and 124th Mtg, 17 Nov 43, 

and Robert Ross Smith, The Approach to the Philippines, UNITED STATES 

ARMY IN WORLD WAR II (Washington, 1953), p. 6.

[52] The total extra lift required was figured at 3 XAP's (modified 

assault transports) 12 MT ships (freighters fitted for vehicle 

carriage), 26 LST's, and 31 LCT's. King promised to provide the XAP's, 

the LST's, and 26 of the LCT's. The MT ships were, or would be, 

available in the area; the five LCT's could be taken from craft 

earmarked for OVERLORD, to be replaced by others in the contingent 

promised for OVERLORD on 5 November. See CCS 424, 5 Dec 43, cited. n. 

51.

Page 280

ers were about to be named, [53] and few doubted that when they reviewed the existing plans they would demand a more ample provision of means. At the plenary meeting on 5 December Harry Hopkins elicited from the military leaders, after some sharp cross-questioning, the remarkable admission that although they had given the stamp of approval to a two-division ANVIL and a three and a half division OVERLORD, they believed nevertheless that both operations should be strengthened. [54]

After two days of discussion at Cairo, the problem had thus taken on new dimensions. It was no longer a question of mounting the ANVIL assault at a fixed scale. Now it seemed necessary to provide a pool of assault shipping large enough to mount both ANVIL and OVERLORD on a scale as yet undetermined but adequate to give each a reasonable margin of safety. Precisely how much shipping would be needed could not be known until the plans themselves were revised and developed in detail. The very uncertainty on this score lent force to the British argument that it would be folly to commit precious assault shipping irrevocably to a venture in southeast Asia that even the U.S. Chiefs of Staff conceded to be of secondary importance.

At the plenary meeting on the 5th, Churchill bluntly pointed out that only the President's unilateral pledge to Chiang stood in the way of agreement. He suggested that Chiang might be offered some lesser substitute for BUCCANEER, which itself would then be postponed until after the monsoon. The remainder of the campaign could be carried out as planned. Hopkins supported this idea. The President, obviously unhappy, finally agreed to the suggestion that Mountbatten's representatives, then in Cairo, and Mountbatten himself should be queried as to what small-scale amphibious operations might be undertaken if he had to give up the bulk of his assault lift. At the same time the CCS were ordered to re-examine forthwith the two main European operations "with a view to increasing the assaults in each case." Roosevelt's full capitulation swiftly followed. That same afternoon, after consulting with his advisers (only King, in the JCS, held out against postponing BUCCANEER) he sent Churchill a brief message: BUCCANEER is off." [55]

The Joint Chiefs were not informed of the decision until the next day, but they must have realized, after their meeting with the Presi-


[53] The President announced General Eisenhower's appointment as 

OVERLORD commander on 6 December.

[54] Min, 4th Plenary Mtg, SEXTANT, 5 Dec 43.

[55] (1) Churchill, Closing the Ring, p. 411. (2) Min, 4th Plenary Mtg, 

SEXTANT, 5 Dec 43. (3) Ehrman, Grand Strategy, V, 190-92. (4) King and 

Whitehill, Fleet Admiral King, p. 425. (5) William D. Leahy, I Was There 

(New York: Whittlesey House, 1950), p. 213. (6) Sherwood, Roosevelt and 

Hopkins, p. 801.

Page 281

dent, that it could not long be delayed. What now had to be decided were the precise alternatives to be offered Chiang. On the night of the 5th the British and U.S. planners made a list of various amphibious operations that might be undertaken in Burma during the spring, assuming arbitrarily that the shipping to be withdrawn from the theater would comprise most of the LST's, combat loaders, and small aircraft carriers. It was not an impressive list. The Joint Chiefs, studying it the following morning, were inclined to conclude that it might be better to give up serious amphibious ventures in the Southeast Asia Command altogether during this season, and transfer all the BUCCANEER assault shipping back to European waters. The British agreed. The CCS accordingly recommended that major amphibious operations in the Bay of Bengal be delayed until after the monsoon, and that Chiang Kai-shek be offered two alternatives: the mainland offensive as planned, with British naval control of the Bay of Bengal assured, but without BUCCANEER, for which would be substituted carrier strikes, commando raids, and bombardment of Bangkok and the railroad; or postponement of the mainland offensive, compensated for by increased airlift to China and more rapid development of the long-range bombardment program from bases in China. Later that day Mountbatten's reply came in, stating flatly that seaborne operations smaller than BUCCANEER would not be worth the effort. He proposed that, in anticipation of Chiang's probable reaction, only limited land operations in northern and central Burma and along the Arakan coast be undertaken, and that the aim of opening the land route to China during this season be abandoned. [56]

BY evening of the 6th all knew that the President, without informing the JCS, had decided to abandon BUCCANEER and, moreover, had already cabled Chiang the bad news, presenting the same alternatives arrived at by the Chiefs of Staff that morning. [57] Chiang's reply had not yet been received, but the President was due to leave Cairo the following morning and the conference decisions could not wait. Accordingly, the two alternatives presented to Chiang were both included

[56] (1) Min, 136th Mtg JCS, 6 Dec 43. (2) Min, 136th Mtg CCS, 5 Dec 43. 

(3) Min 137th Mtg CCS, 6 Dec 43. (4) CCS 427, Rpt by CPS, 5 Dec 43, 

title: Amph Opns in Southeast Asia Alternative to BUCCANEER. (5) Romanus 

and Sunderland, Stilwell's Command Problems, p. 70. (6) Ehrman, Grand 

Strategy, V, 192-93.

[57] (1) Msg, President to Chiang, 5 Dec 43, Exec 10, Item 70. (2) Min, 

5th Plenary Mtg, SEXTANT, 6 Dec 43. Since the CCS recommendations were 

those approved by the JCS and had been drafted by General Marshall, it 

is possible that Marshall had earlier shown this draft to the President 

and that the latter used it, but without informing Marshall. At all 

events it seems unlikely that the Chiefs of Staff could have known on 

the morning of the 6th that the President had already cabled Chiang, 

since they were discussing their own draft with a view to submitting it 

to the President. See Matloff, Strategic Planning for Coalition Warfare, 

1943-1944, MS, Ch. XVI, p. 60.

Page 282

in the final SEXTANT paper approved by the President and Prime Minister at the plenary meeting on the night of the 6th.

In the light of the Generalissimo's known attitude, there could be little doubt that he would reject the first; there was considerable doubt that he would accept even the second. Actually, by ruling out any worthwhile substitute for BUCCANEER, and so informing the Chinese leader forthwith, the President had thrown away an option that might have been acceptable to Chiang, inasmuch as the latter had never been told precisely what sort of operation was contemplated, but only that it would be a major one. At the time the conference decisions were approved, however, the leaders had Mountbatten's word for it that nothing less than BUCCANEER would serve. Later in the month Mountbatten changed his mind, but by then the President's message had left Chiang in no mood for compromise. In any case, Mountbatten's small residue of assault shipping was soon to be swallowed up in the maw of swelling European requirements. On 7 December the world-wide redeployment of assault shipping dictated by the SEXTANT decisions began as the CCS ordered Mountbatten to send fifteen LST's and six assault transports-the bulk of his amphibious fleet-back to European waters. [58]

It has become almost a commonplace in American interpretations of World War II to say that at Tehran the British were forced to abandon their reservations concerning OVERLORD. Thus, it is asserted, the primacy of OVERLORD vis-a-vis the Mediterranean, and, indeed, its execution were finally assured. [59] Like the classic query, "When did you stop beating your wife?" this interpretation accepts as fact what is actually the nub of the issue, namely, the American allegation that the British, and Churchill in particular, had never intended to go through with OVERLORD and only resigned themselves to do so under Soviet pressure at Tehran. In reality, both Churchill and Brooke, forced repeatedly by the Russians to state their intentions concerning OVERLORD, held firmly to their position. At the end of the conference it was what it had been before: OVERLORD would be the main

[58] (1) Ehrman, Grand Strategy, V, 193, 211-12. (2) Mountbatten Report, 

p. 29. (3) Romanus and Sunderland, Stilwell's Command Problems, pp. 

75ff.

[59] For example, Sherwood (Roosevelt and Hopkins, page 788) states that 

Churchill at the plenary meeting on the 29th "bowed to the inevitable"-

i.e., accepted OVERLORD-by promising Stalin that "Britain would hurl 

every ounce of her strength across the Channel at the Germans." Admiral 

Leahy in his memoirs (I Was There, page 209) speaks of the decision on a 

May OVERLORD (which he represents as a capitulation by the British, not 

a compromise) in the same sense-e.g., the British "fell into line." See 

also Harrison, Cross-Channel Attack, pp. 125-26, Cline, Washington 

Command Post: The Operations Division, UNITED STATES ARMY IN WORLD WAR 

II (Washington, 1951), p. 229; Matloff, Strategic Planning for Coalition 

Warfare, 1943-1944, Ch. XIII, pp. 71-72; "The Decision to Invade 

Southern France," MS, pp. 8-9; Greenfield, The Historian and the Army, 

p. 54; Higgins, Winston Churchill and the Second Front, pp. 212-13, 244.

Page 283

effort of the Western Allies in Europe, and, as far as the British were concerned, it would be carried out, as Churchill told Stalin on 30 November, "provided the enemy did not bring into France larger forces than the Americans and British could gather there." [60] In essence, this was the reservation already spelled out in the OVERLORD outline plan and accepted by the U.S. Chiefs of Staff themselves. Whether British leaders secretly harbored reservations of a more far-reaching nature is not known now (except by themselves) and probably will never be known. Certainly, the Americans had no basis at the time, other than hearsay, for suspecting that they did. The historian's position is likely to depend largely on where he decides to place the burden of proof-on the Americans to demonstrate that their suspicions were based on fact, or on the British to show that their professions were sincere.

As for Stalin's stand on OVERLORD, it was no more than a restatement of the familiar "second front" theme dinned into Western ears from the time of the German invasion of Russia down to the Moscow Conference of October 1943. It may be doubted whether Stalin was taken in by the transparently vague formula finally decided on to define the target date for the operation, but there is no indication that he attached any importance to it. His whole attitude at Tehran toward the timing of OVERLORD and supporting operations in the Mediterranean was one of lofty indifference. At all events, his pronouncements on OVERLORD added nothing to earlier Anglo-American agreements on the relation between the cross-Channel invasion and the Mediterranean. The most significant effect of Stalin's position was, not the essentially empty characterization of OVERLORD and ANVIL as "supreme" operations in 1944, but the CCS decision on 5 December to explore the possibility of strengthening the two assaults. This decision which virtually invited the responsible commanders to demand the means they considered necessary, formally recognized-what the JCS since spring of 1943 had refused to concede-that the limit placed on the size of the OVERLORD assault at the TRIDENT Conference was arbitrary and unrealistic. In principle, it represented a real vindication of the stubborn efforts by the British since early 1942 to obtain more American landing craft for OVERLORD. How many would actually be forthcoming remained to be seen. For the present, over and above the allotments made at the TRIDENT and QUADRANT Conferences, the planners could count on the vessels released from southeast Asia, most of about two months of American production of LST's, LCI(L)'s, and LCT's, pledged by Admiral King on 5 November and 4 December, a handful of U.S. and British assault transports, and an indeterminate amount of new British LCT's. These additions, it was

[60] Churchill, Closing the Ring, p. 380.

Page 284

COMMAND DECISIONS

expected, "should provide a satisfactory lift both for OVERLORD and ANVIL." The expectation proved to be overoptimistic. [61]

With relation to the war in the Pacific, Stalin raised an issue that was welcome to the British and may have been embarrassing to the U.S. Chiefs of Staff. No debate on the question is recorded, but the CCS, in making OVERLORD and ANVIL the "supreme" operations for 1944, agreed that "nothing must be undertaken in any other part o the world" to jeopardize their success. Never before had the cross-Channel operation been underwritten in such sweeping terms; the statement wiped out provisos, insisted upon by the JCS at the TRIDENT Conference, that in the face of reverses in the Pacific the United States would be obliged to expand her operations there, even at the expense of the effort in Europe. In principle, at least, the war in the Pacific was now subordinate to the war in Europe. [62]

Coming at the time they did; the decisions at Cairo and Tehran relating to the war in Europe have inevitably taken on a retroactive luster from the dramatic events of the following summer-the invasion of Normandy and southern France, the advance up the Italian peninsula, the sweep across France to the Rhine. The decisions foreshadowed the events; it is less certain that they shaped them as well. To contemporaries, indeed, it seemed as though the whole conference program was going awry almost before the ink was dry. As the Cairo meetings ended, the outlook for the attack on Rhodes was good. The British could reasonably count on using for it the ex-BUCCANEER assault shipping now on its way back to the Mediterranean, since the conference agreements stipulated only that any operations undertaken in the Aegean must be "without detriment" to OVERLORD and ANVIL. [63] Negotiations with the Turks were going well. Less than a week later, the Turks suddenly raised the ante on military aid demanded as the price of intervention, and by the 25th Churchill had abandoned his Aegean plans. Similarly, by the middle of the month the planned

[61] The additional lift was listed as follows: 
  For OVERLORD For ANVIL
LSTs 23 U.S. 3 British 36 U.S. 5 British
LCI(L)s 24 U.S.  
LCTs 19 U.S. 45 British 31 U.S.
Assault transports   3 U.S. 6 British

See CCS 428 (Rev.), 15 Dec 43, Annex V. [62] CCS 426/1, 6 Dec 43, Rpt to President and Prime Minister. [63] (1) CCS 428 (Rev.), 15 Dec 43, Annex V. (2) CCS 426/1, 6 Dec 43, Rpt to President and Prime Minister. (3) A summary of the Cairo-Tehran decisions prepared for the Army Service Forces on 15 December stated that Turkish intervention and surrender of Bulgaria were considered "probable," and that in this event it would be necessary "to mount such operations as may be practicable in the Eastern Mediterranean...." See Memo, Gen Wood for various addressees, 15 Dec 43, sub: SEXTANT Decisions, ASF Planning Div Folder SEXTANT Decisions.

Page 285

landings south of Rome, taken almost for granted at Cairo and Tehran, had been canceled owing to the failure of the American Fifth Army to reach positions within supporting distance of the target area. Later in the month, they were revived on a larger scale and, after a frantic search for the necessary assault lift, finally carried out late in January at Anzio. ANVIL, too, came under fire almost immediately, as the OVERLORD commanders laid claim to its allotted assault lift, and after protracted debate the plan was canceled. When the Allies finally invaded southern France in mid-August, the operation was no longer strategically related to OVERLORD and could not have been justified by the arguments used at Tehran.

As for OVERLORD itself, Stalin's unequivocal insistence upon the operation undoubtedly enhanced the likelihood that it would be carried out, even in the face of an unforeseen increase in German power. On the other hand, the massive preparations for the invasion had already generated a momentum difficult if not impossible to arrest. Any radical change of direction or of emphasis at this time-let alone later-would have caused an upheaval in plans and preparations more costly than many military defeats. As a practical matter, the war in Europe had progressed beyond the point of no return. Even the date was hardly any longer in the realm of strategic decision. After Tehran strategic planning was pointed toward a late May or early June OVERLORD (though the administrative staffs continued for some time to work toward an early May deadline), but in the end the actual date of the launching was shaped, as Churchill has remarked, mainly "by the moon and the weather." [64]

[64] Churchill, Closing the Ring, p. 376. 

[1] The present essay is condensed from several chapters of Richard M. 

Leighton and Robert W. Coakley, Global Logistics and Strategy, 1943-

1945, a forthcoming volume in the series UNITED STATES ARMY IN WORLD WAR 

II, based on original research in the records of the Joint and Combined 

Chiefs of Staff, including the official minutes and papers of the Cairo-

Tehran Conferences, and Army records in the custody of the Deputy Chief 

of Staff for Operations, G-3, filed in the Federal Records Center of the 

National Archives. A contrasting interpretation of the Cairo-Tehran 

Conferences will be found in another volume in this series, Maurice 

Matloff, Strategic Planning for Coalition Warfare, 1943-1944 

(Washington, 1959). The conferences are also described from various 

points of view in a number of other works in the series: Gordon A. 

Harrison, Cross-Channel Attack (Washington, 1951); Charles F. Romanus 

and Riley Sunderland, Stilwell's Command Problems (Washington, 1956); 

Forrest C. Pogue, The Supreme Command (Washington, 1954); and Louis 

Morton and Henry Morgan, The Pacific War: Strategy and Command, The Road 

to Victory (in preparation). Three other American studies deal with the 

Cairo-Tehran Conferences at some length: Herbert Feis, Churchill, 

Roosevelt, and Stalin (Princeton: Princeton University Press, 1957); 

William Hardy McNeill, America, Britain, and Russia, Their Co-Operation 

and Conflict, 1941-1946 (London: Oxford University Press, 1953) and 

Robert Sherwood, Roosevelt and Hopkins: An Intimate History (New York: 

Harper & Brothers, 1948). The outstanding British interpretation is John 

Ehrman, Grand Strategy, Vol. V, August 1943-September 1944 (London: H.M. 

Stationery Office, 1956), in the official British History of the Second 

World War. (Permission to quote from this work has been received from 

the Controller of H.M. Stationery Office.) Of the large memoir 

literature, Winston S. Churchill's Closing the Ring (Boston: Houghton 

Mifflin Company, 1951) contains the most detailed and valuable account. 

The published memoirs of Admiral William D. Leahy, Admiral Ernest J. 

King, General Dwight D. Eisenhower, and Maj. Gen. John R. Deane are 

also useful, though sketchy on the conferences themselves. Arthur 

Bryant's two-volume biography of Lord Alanbrooke (General Sir Alan 

Brooke), Turn of the Tide, and Triumph in the West (New York: Doubleday 

and Company, 1957, 1959), is also useful for the British position. Three 

American studies, recently published, contain brief, provocative 

analyses, from the American point of view, of the Anglo-American debate 

on European strategy in World War II: Kent Roberts Greenfield, The 

Historian and the Army (New Brunswick, N. J.: Rutgers University Press, 

1954); Samuel Eliot Morison, Strategy and Compromise (Boston: Little, 

Brown and Company, 1958), and Trumbull Higgins, Winston Churchill and 

the Second Front (New York: Oxford University Press, 1957).