Part 1: Strategy and Operations
Blank page
Chapter 1: The Debate Over Southern France
Although ultimately proving to be one of the most important Allied operations of World War II, the invasion of southern France has also remained one of the most controversial.1 From start to finish and even long afterwards, Allied leaders hotly debated its merits and its results. Most judged the enterprise solely on the basis of its effect on the two major Allied campaigns in western Europe, the invasion of northern France and the invasion of Italy. Supporters, mainly American, pointed out its vital assistance to the former, and detractors, mostly British, emphasized its pernicious influence on the latter. Even many years after these events, surprisingly few have ever examined the campaign in southern France itself or added anything to the original arguments that surrounded the project from its initial inception to its execution some fourteen months later. Yet the debate over the invasion of southern France was central to the evolving Allied military strategy during that time and became almost a permanent fixture at Allied planning conferences in 1943 and 1944.
The Protagonists
President Franklin D. Roosevelt and Prime Minister Winston S. Churchill together headed the Anglo-American coalition and made or approved all political and strategic decisions. They were assisted by their principal military advisers, the American Joint Chiefs of Staff (JCS) and the British Chiefs of Staff (BCS).2 Each
national group met separately and formulated plans and programs, but then came together in a single committee called the Combined Chiefs of Staff (CCS) to discuss matters further and arrive at joint decisions. The CCS spoke for the president and the prime minister, allocated resources among the theaters of operation, and directed Allied theater commanders. From time to time, when the two Allied political leaders came together to resolve key issues, the CCS accompanied them and sought to reconcile the often divergent views and interests of the United States and Great Britain. Occasionally, Roosevelt and Churchill conferred with other Allied leaders, such as Joseph Stalin, and on those occasions the military chiefs were also consulted.
At first the British tended to dominate the Anglo-American strategic deliberations. They had been in the conflict from the beginning, had amassed more experience, and had more military forces engaged than the poorly prepared Americans, who entered the struggle more than two years later. As the Americans committed increasing manpower and matériel to the war, they gradually became the more important partner and had correspondingly greater influence on the courses of action adopted by the alliance.
In 1942 Churchill had proposed the North African invasion, and Roosevelt, over the objections of the JCS, had acquiesced. At the Casablanca Conference in January 1943, the British had recommended and the Americans had reluctantly accepted the seizure of Sicily upon the conclusion of the Tunisian campaign. What differentiated the outlooks of the two parties was where and when to make the Allied main effort in Europe.
The Americans wished to launch an immediate cross-Channel attack from England to the Continent, followed by a massive and direct thrust into the heart of Germany. General George C. Marshall, Chief of Staff of the U.S. Army, a member of the JCS, and the dominant figure among American planners for the war in Europe, was the primary spokesman for this operational strategy, and Marshall tended to judge other ventures by evaluating their possible effect on what he thought should be the main Allied effort. In contrast, the British preferred a peripheral, or “blue water,” strategy, undertaking lesser operations around the rim of Europe to wear down Germany and Italy before launching the climactic cross-Channel strike. What the British wanted in particular was to continue the offensive momentum in the Mediterranean area as opportunities unfolded.
British preference for operations in the Mediterranean, especially along the eastern shores of the sea, was motivated in part by postwar political considerations that were not shared by their American Allies. As such they were rigorously opposed by Marshall and his cohorts. But the United States, in turn, had similar political commitments in Southeast Asia and
China that had little to do with British interests overseas. Differences in military strategy and postwar political concerns thus colored all discussions of future Allied operations and in the end forced the CCS to adopt a series of compromises not entirely satisfactory to either side. Such, inevitably, is the nature of coalition warfare, and the southern France invasion represented one of the major compromises of the Anglo-American partnership during World War II.
TRIDENT, May 1943
The proposal for an invasion of southern France formally arose in May 1943 at the TRIDENT Conference, a series of meetings between the American and British staffs held in Washington, D.C. At the time, the Allies had cleared North Africa and were preparing to invade Sicily, but had not yet decided on subsequent operational objectives. In their preliminary gatherings, the JCS had their eyes firmly fixed on an assault across the English Channel, eventually code-named OVERLORD; following a successful Sicilian campaign, they wanted to begin transferring all Allied military resources out of the Mediterranean theater to support an OVERLORD invasion sometime in the spring of 1944. But this, they realized, was hardly feasible. The campaign in Sicily promised to be over by the end of summer, and the prospect of suspending all ground operations against the Axis until the following year, a gap of possibly eight or more months, was unacceptable. Some interim operations beyond Sicily were required, and the considerable Allied establishment in the Mediterranean argued for further activity in the area, especially if it could divert German resources from northern France without greatly impeding the Allied buildup in Great Britain.
The JCS considered a number of potential target areas, including southern France, southern Italy, Sardinia and Corsica, the Genoa area of northwestern Italy, Crete and the Dodecanese Islands in the eastern Mediterranean, the Balkans, and the Iberian Peninsula (See Map 1). The majority of American planners regarded an early invasion of southern France as extremely risky: an exploitation northward would require more strength than the Allies were likely to leave in the Mediterranean, and the operation would demand the prior occupation of Sardinia and Corsica, causing another diversion of Allied resources. As for the Iberian Peninsula, earlier fears that the Germans might move against Gibraltar had disappeared, and no one saw Spain as a potential invasion route to anywhere. In the eastern Mediterranean, air support requirements for operations appeared to depend on Turkish entry into the war on the Allied side, an unlikely event. Many also believed that operations in either the eastern Mediterranean or the Balkans would lead to a major Allied commitment in southeastern Europe, where logistical and geographical problems could preclude the application of decisive strength. Although the Italian peninsula appeared to be an immediately feasible objective, the JCS feared that the invasion would evolve into a major campaign that would divert resources from OVERLORD. Instead,
Map 1: Western and Central Europe, 1 September 1939
American planners concluded that the seizure of Sardinia, and probably Corsica as well, would prove the most desirable action in the Mediterranean. The operations would keep some pressure on the Axis, while from bases on Sardinia and Corsica the Allies could pose strong threats against both Italy and southern France, thereby pinning Axis forces in place.
British planners attending TRIDENT were better prepared. They quickly agreed with their American opposites that an early invasion of southern France would be difficult and, furthermore, doubted the value of seizing Sardinia and Corsica. Instead, the BCS proposed that action in the Mediterranean be aimed at eliminating Italy from the war in 1943. The collapse of Italian resistance would not only provide the Allies with a great psychological victory but would also, the BCS argued, compel Germany to redeploy strong forces to Italy to hold the German southern flank, thereby promoting the success of OVERLORD as well as relieving German pressure on the Russian front. As an alternative, the BCS were prepared to propose a move into the Balkans, estimating that Germany would divert strong forces from both the east and west to hold southeastern Europe. In the end, however, they persuaded the Americans that an early invasion of Italy was the best solution.
Thus the idea of an invasion of southern France attracted some attention at TRIDENT, but was dropped from primary consideration. Instead, with some reluctance on the part of the JCS, the CCS approved the concept of knocking Italy out of the war in 1943. Almost immediately they directed General Dwight D. Eisenhower, commanding Allied forces in the Mediterranean,3 to draw up plans for invading Italy and tying down the maximum number of German divisions in the Mediterranean area. In late July, with Allied success on Sicily assured and with the sudden collapse of the Mussolini government in Italy, the CCS, at Eisenhower’s behest, agreed that an invasion of southern Italy would best achieve the ends set forth at TRIDENT; the Allied high command scheduled the invasion for early September 1943.
Another Look at Southern France
Throughout the summer of 1943 American planners continued to regard the Mediterranean theater with mixed feelings. The Joint War Plans Committee of the JCS emphasized the advantages of fighting the major western European battles in Italy if the Germans so elected, pointing out that once the Allies had cleared the Italian peninsula, the newly renovated French Army could invade southern France with relative ease. The Joint Strategic Survey Committee, thinking along similar lines, suggested that after Italy had been eliminated from the war, the Allies might well launch a major invasion of southern France in conjunction with a much smaller
OVERLORD effort undertaken with forces left over from the Mediterranean. Navy planners favored a major effort in Italy, believing that a southern approach would allow more amphibious resources to be switched to the Pacific.
Although the JCS easily resisted these internal arguments against the basic OVERLORD concept, they were still unable to decide how best to exploit the decline of Axis power in the Mediterranean with the resources left in the theater after OVERLORD requirements had been met. The situation was further complicated by the lack of sufficient shipping to move the bulk of Mediterranean resources to England for OVERLORD. In many cases it was easier to ship OVERLORD forces from America to England than from the Mediterranean; for example, the Allies would probably never have enough shipping to move the French Army from North Africa to England for participation in OVERLORD.
American planners thus believed that after all OVERLORD requirements had been met they would still have enough strength left in the Mediterranean to maintain strong pressure against German forces in Italy; to seize Sardinia and Corsica; to establish air bases on the Dodecanese Islands; and, in conjunction with OVERLORD, to launch some kind of assault against southern France. Even if the Allies halted all offensive action in the Mediterranean, they would still have to leave twelve to fourteen divisions in the theater to maintain security and to pose threats. Better to have these forces engage in at least limited offensives than to have them waste away from inaction. A small, multi-division landing in southern France would obviously complement OVERLORD, representing a secondary, southern prong of the Allied attack on German-occupied France. Current OVERLORD plans in July 1943 even called for such a diversionary effort against southern France at the time of the cross-Channel assault. But American planners now began to propose that the southern landings be more than a diversion and be upgraded to a larger effort—one that would provide continued assistance to OVERLORD, would make immediate use of the French Army, and, incidentally, would preempt any British proposals to employ excess Allied strength in the eastern Mediterranean.
For these reasons the JCS decided in August 1943 formally to support an invasion of southern France, code-named ANVIL, which would be launched either before, during, or after OVERLORD as the situation permitted; they ultimately concluded that the operation would have to follow OVERLORD.4 American planners reasoned that a successful ANVIL would probably depend on OVERLORD to deplete German strength in southern France, while the seizure of the southern ports and a subsequent drive to the north would force the Germans to defend the approaches to their own country from two directions. Still giving overriding priority to OVERLORD, the JCS thus settled on a three-phase plan for the Mediterranean: (1) eliminating Italy from the war and clearing the Italian peninsula as far north as Rome; (2) capturing
Sardinia and Corsica to increase the width and depth of the Allied air penetration into Europe; and (3) creating a situation in the Mediterranean favorable to the launching of ANVIL about the time of OVERLORD. Specifically, the JCS plan for southern France called for the seizure of a beachhead in the Toulon–Marseille area, the development of Toulon and Marseille into major supply ports, and an exploitation northward up the Rhone valley to support OVERLORD. This was the basic ANVIL concept on which all planning for the invasion of southern France turned for nearly another year.
The QUADRANT Conference
Soon after the end of the Sicilian campaign in August 1943, Roosevelt, Churchill, and the CCS met in Quebec at the QUADRANT Conference. There the British accepted the JCS program for the Mediterranean in principle. However, the BCS also pointed out that the Allies would have to make OVERLORD a much stronger assault than current plans envisaged and that the forces already allocated to OVERLORD would require more amphibious lift than had been planned. If the United States was unwilling to make the additional lift available from Pacific allocations, it would logically have to come from the Mediterranean, inevitably threatening ANVIL. The BCS also believed that an effective ANVIL would require a three-division assault, but British projections indicated that by late spring of 1944 the Allies would have only a mixed collection of ships and landing craft left in the Mediterranean, capable at best of putting a single reinforced division ashore. Eisenhower agreed with the BCS that anything less than a three-division ANVIL would not be feasible unless Allied forces in Italy had first reached the Franco-Italian border.
The JCS admitted the necessity for increasing the OVERLORD assault echelon, but they convinced the BCS that the Allies should continue planning for at least some kind of ANVIL operation on the basis of the limited means expected to be available in the Mediterranean at the time of OVERLORD. Accordingly, a QUADRANT decision by the Combined Chiefs of Staff directed Eisenhower to prepare an ANVIL plan by November 1943. The directive was somewhat vague and did not differentiate between the British view that ANVIL should be reduced to a threat at the time of OVERLORD and the American desire to make ANVIL a major operation directly connected to OVERLORD. But obviously the pressures against ANVIL were growing. A JCS insistence on a three-division ANVIL would create a natural competition between OVERLORD and ANVIL for amphibious resources. Meanwhile, the demands of operations in Italy would generate their own momentum at the expense of both ANVIL and OVERLORD. Finally, the British, at QUADRANT, expressed continuing interest in limited operations in the eastern Mediterranean, operations that would also divert resources from ANVIL.
Eisenhower submitted his reduced ANVIL plan in late October 1943. By then the heady optimism of the summer had faded. The Allies had successfully invaded southern Italy in September, Italian resistance had collapsed,
Members of U.S. and British staffs conferring, Quebec, 23 August 1943.
Seated around the table from left foreground: Vice Adm. Lord Louis Mountbatten, Sir Dudley Pound, Sir Alan Brooke, Sir Charles Portal, Sir John Dill, Lt. Gen. Sir Hastings L. Ismay, Brigadier Harold Redman, Comdr. R. D. Coleridge, Brig. Gen. John R. Deane, General Arnold, General Marshall, Admiral William D. Leahy, Admiral King, and Capt. F. B. Royal.
and the Germans had evacuated both Sardinia and Corsica. But they had also quickly moved reinforcements into Italy, while the Allied buildup was slow. By late October the Germans had twenty-five divisions in Italy as opposed to eighteen for the Allies, and the Allied commanders faced a stalemate if not a serious reverse. Hopes that Allied forces might reach Rome before the end of 1943 had disappeared.
Eisenhower’s ANVIL plan made it clear that the Allies would be able to mount little more than a threat to southern France at the time of OVERLORD, then scheduled for about 1 May 1944. There was simply not enough amphibious shipping in the European theater for two major assaults. Eisenhower himself felt that there was little chance for the Allies to be far enough north in Italy by the spring of 1944 to launch an overland invasion of southern France from that quarter; the best he could promise was the seizure of a small beachhead in southern France in the unlikely event that the Germans withdrew the bulk of their forces from the south. He concluded that the Allies might do better at spreading out German defenses by
continuing the Italian offensive with all resources available in the Mediterranean. As for the eastern Mediterranean, Eisenhower, with the concurrence of the BCS, judged that no operations could be undertaken in that area until the Allied forces on the Italian mainland were at least as far north as Rome. In the end, Eisenhower recommended that ANVIL remain indefinite, as one of several alternatives the Allies should consider for the future in the Mediterranean.
The British requested that the JCS accept Eisenhower’s concept as a basis for future planning, a step the JCS reluctantly took early in November. As a result, plans for an ANVIL operation in conjunction with OVERLORD were dropped from consideration before the next CCS meeting, and ANVIL was absent from the SEXTANT Conference agenda when the CCS convened at Cairo late in November 1943.
The Cairo and Tehran Conferences (November–December 1943)
The meeting of the western Allied leaders with Generalissimo Chiang Kai-shek at Cairo, code-named SEXTANT, on 22–26 November 1943, was followed by a second conference with Joseph Stalin in Tehran, 28 November–1 December, and then a final session at Cairo, 3–7 December.5 At the initial SEXTANT meetings, the British again pressed for increased Allied efforts in the central and eastern Mediterranean. They proposed a schedule that called for an advance in Italy as far as Rome by January 1944; the capture of Rhodes during February; a drive in Italy as far as the line Pisa–Rimini (about halfway from Rome to the Po River in northern Italy); and increased support to Yugoslav guerrillas (including the establishment of minor beachheads on the east coast of the Adriatic). To provide the amphibious lift needed to support all these operations and to increase the amphibious allocations for OVERLORD, the BCS recommended canceling amphibious undertakings in Southeast Asia and postponing OVERLORD until July 1944. Eisenhower appeared to support the British outlook. He acknowledged the value of harassment operations across the Adriatic and suggested that after the Po valley had been reached, the French Army could move westward into southern France and the other Allied forces could advance northeast.
The reaction of the JCS was predictable. The British and Eisenhower presentations threatened the place of OVERLORD as the centerpiece of the war in Europe. Although the May target date was hardly sacrosanct, a delay until July was intolerable, and waiting until the Allies had reached the Po River valley would probably set the invasion back to August. Moreover, commitments to Chiang Kai-shek at Cairo made it virtually impossible for the JCS to cancel Southeast Asia operations. To General Marshall all these proposals posed serious threats to OVERLORD and represented a return to the British peripheral strategy or, worse, one that would have had American troops fighting in
the Balkans and the Italian Alps.
Settling nothing at Cairo, the CCS moved on to Tehran for consultations with Stalin and Russian military leaders. During preliminary conferences in Moscow among American, British, and Russian officials in late October and early November, the Americans had gathered that the Russians favored increased efforts in the Mediterranean and perhaps some operations in the Balkans to divert German strength from the eastern front. U.S. representatives had also received the impression that the Russians were no more than lukewarm toward OVERLORD.6 However, at Tehran Russian representatives vehemently objected to further Anglo-American operations in the Mediterranean-Balkan area that might detract from OVERLORD, which, the Soviets insisted, had to be launched in May 1944. To the surprise of both the Americans and British, the Soviets also proposed an invasion of southern France in support of OVERLORD. While not insistent about ANVIL, the Russians firmly opposed other major offensives in the Mediterranean and took a stand against any advance in Italy beyond the lines the Allies had already attained. Stalin maintained that any major operation in the Mediterranean other than ANVIL would prove strategically indecisive and could lead only to the dispersal of OVERLORD resources. He appeared intrigued with the pincers aspect of a combined OVERLORD–ANVIL campaign, but urged that not even ANVIL should be permitted to interfere with OVERLORD.
The CCS promised Stalin that OVERLORD would be launched toward the end of May 1944, a compromise between the American date of 1 May and the British proposal of 1 July. The CCS also assured Stalin that they would execute ANVIL concurrently with OVERLORD on the largest scale possible with the amphibious lift left in the Mediterranean in May, and they agreed to carry the offensive in Italy no farther than the Pisa–Rimini line, about 150 miles north of Rome, but 100 miles short of the Po valley.
Returning to Cairo, the CCS took another look at the amphibious lift available for ANVIL. Even by scraping the bottoms of all potential barrels, they estimated that sufficient lift for a one-division assault with a quick follow-up of two-thirds of a division was all that could be assembled by May 1944. As a remedy, the BCS again proposed canceling projected amphibious operations in Southeast Asia or reducing Pacific allocations. With great reluctance, Admiral Ernest J. King, Chief of Naval Operations and a member of the JCS, agreed to divert enough Pacific lift resources to execute a two-division ANVIL assault.7 However, King’s
offer helped little, permitting only a minimum, risky ANVIL assault; making no allowance for unforeseen contingencies in the Mediterranean; and still leaving the amphibious resources for OVERLORD at a level that many planners considered inadequate.
The British were still dissatisfied and requested that even more amphibious shipping be allocated from the Pacific theater. Reminding the JCS that Russia would ultimately enter the war against Japan, the BCS argued that major amphibious operations in Southeast Asia were thus unnecessary, making it possible to transfer more such resources to the European theater. But the JCS at first refused to accept the British rationale, and the matter reached a temporary impasse. However, on 5 December President Roosevelt, changing commitments to China, agreed to cancel some of the planned operations in Southeast Asia, and the CCS thereafter began dividing the excess lift between OVERLORD and ANVIL.
Thus, at the beginning of December 1943, ANVIL was again on the agenda and, instead of a diversionary threat, was to be an integral adjunct to OVERLORD. Indeed, the CCS now went so far as to agree that OVERLORD and ANVIL would be the “supreme” operations in Europe during 1944 and that no other campaigns in Europe should be allowed to prevent the success of those two. Prospects that other operations in the Mediterranean, at least, would not interfere with ANVIL were also brightened by a British agreement to halt in Italy at the Pisa–Rimini line and by the fact that a British condition for the capture of Rhodes—Turkish entry into the war—could not be met. The only other Mediterranean threat to ANVIL was the possibility that outflanking amphibious maneuvers in Italy, such as the one planned for Anzio in early 1944, might reduce ANVIL allocations. But most American planners foresaw that the real danger to ANVIL, if any, would come from pressures to strengthen the OVERLORD assault.
ANVIL Canceled
The turn of the year saw a general reshuffling of command structures and boundaries in the European and Mediterranean theaters. The CCS appointed Eisenhower as supreme Allied commander for OVERLORD, and he left the Mediterranean in December. General Sir Bernard L. Montgomery, who was to be Allied ground commander for OVERLORD, and Lt. Gen. Walter Bedell Smith, who was to be chief of staff at Eisenhower’s new command—Supreme Headquarters, Allied Expeditionary Force (SHAEF)—followed. The Mediterranean, previously divided between Eisenhower and the commander of the British Middle East Theater, General Sir Henry Maitland “Jumbo” Wilson, became unified under Wilson.
After reaching London Montgomery and Smith began reviewing the draft OVERLORD plans and pressing for major increases in the size of the OVERLORD assault, a step the preliminary planners had been urging on the CCS for months. Knowing that no other ready source but ANVIL existed
from which to draw the amphibious lift needed to enlarge OVERLORD, they recommended that the lift be taken from ANVIL and that ANVIL be reduced to a one-division threat. The BCS supported these recommendations, reiterating their position that ANVIL should not be permitted to interfere in any way with OVERLORD.
The renewed pressure against ANVIL put Eisenhower in an ambiguous situation when he reached London in mid-January.8 One of his last tasks as Allied commander in the Mediterranean had been to prepare a new ANVIL plan in accordance with a CCS post-SEXTANT directive. Eisenhower’s plan had again been built around a three-division ANVIL assault followed by exploitation northward. Only the three-division ANVIL, Eisenhower believed, would provide strong Support to OVERLORD—support that a mere threat could not provide. He reminded the BCS and his principal subordinates that the CCS had promised ANVIL to the Russians, and repeated his arguments about the most effective use of the French Army. If ANVIL were canceled, many French and even many American divisions might well be locked in the Mediterranean, wasted for lack of shipping to take them to northern France, for lack of port capacity in the OVERLORD area to support them, and for lack of room in Italy to deploy them. Eisenhower made clear his reluctance to reduce ANVIL to a threat and proposed that every other possible means of strengthening OVERLORD be sought. The JCS generally agreed.
The next step in the debate was a highly technical argument between British and American logistical planners over the capacity, serviceability, and availability of assault shipping and landing craft already allocated to OVERLORD. Employing American figures, the JCS concluded that the Allies could significantly increase the size of the OVERLORD assault force and still provide the lift necessary for at least a two-division ANVIL. The JCS carried this argument almost to the point of insisting on a two-division ANVIL, with OVERLORD being undertaken with the means left over after the ANVIL demands were met. Perhaps happily for their peace of mind, other problems arose before the JCS were forced to push their argument to its logical conclusion—giving a two-division ANVIL priority over OVERLORD.
As had been the case earlier, the war in Italy now began to influence the fate of ANVIL. On 22 January 1944, in an attempt to outflank German defenses and speed the capture of Rome, the U.S. VI Corps surged ashore at Anzio, on Italy’s west coast some thirty miles south of Rome. But the Germans reacted vigorously, and the Allied landing forces soon found themselves confined to the beachhead, unable to move toward Rome or even to establish contact with Wilson’s main armies to the south. The Italian campaign had again bogged down, and the Anzio venture began to consume resources that planners had already earmarked for ANVIL.
The situation in Italy prompted Churchill to recommend immediate
reinforcement of the theater and the abandonment of ANVIL. Italy was where the Allies had the best opportunity to tie down German divisions and thus contribute to the success of OVERLORD. It was unjustifiable, he held, to deny resources to the Italian campaign for the sake of ANVIL. Far better to transfer the bulk of the scarce amphibious lift earmarked for ANVIL to OVERLORD, retaining perhaps enough shipping in the Mediterranean for a one-division threat and ultimately moving the French Army to northern France when more shipping became available. ANVIL, he contended, was too far from Normandy to give direct support to OVERLORD.
The BCS agreed. The Allies were attempting to execute three major campaigns, OVERLORD, ANVIL, and Italy, and had given none of them sufficient resources for success. The British arguments boiled down to two simple propositions: if the campaign in Italy went poorly, then it would be necessary to commit the ANVIL resources there; if the campaign in Italy went well, then ANVIL was unnecessary.
By mid-February 1944 ANVIL had lost most of its prominent supporters among Allied planners in both England and the Mediterranean. Even Eisenhower had begun to waver. He still wanted ANVIL, but thought that the Allies would be unable to disengage sufficient strength from Italy to execute a meaningful landing in the south at the time of OVERLORD. Moreover, he was still anxious to obtain additional resources for OVERLORD. Within the JCS, General Marshall felt that the Allies would probably have to cancel ANVIL as an operation more or less concurrent with OVERLORD unless Wilson’s forces in Italy reached Rome before April 1944. Marshall was willing to forego ANVIL if Eisenhower insisted, but suggested that the JCS could accept a stabilized front in Italy south of Rome if such a step would enhance the chances of executing ANVIL about the time of OVERLORD. He still believed that canceling ANVIL out of hand was unwise and still hoped that improvements in the Allied situation at some future date might make the operation again feasible and perhaps even necessary.
The CCS finally reached another compromise. The JCS agreed to allocate all Mediterranean resources to Italy temporarily for the purpose of seizing Rome by May 1944, and the BCS instructed Wilson, the theater commander, to continue planning for ANVIL to be launched as circumstances in Italy permitted. The CCS deferred a final ANVIL decision until late March, but this delay left the projected assault at the mercy of pressures from both Italy and OVERLORD.
In March Eisenhower, still attempting to increase the size of the OVERLORD assault force, recommended first postponing a decision on ANVIL and then canceling the operation entirely. Concerned over the ability of Wilson to transfer additional amphibious resources from the Mediterranean in time for OVERLORD, he wished to assume no unnecessary risks for the sake of ANVIL. At the time, launching both OVERLORD and ANVIL concurrently seemed impossible. The earliest prospect of gaining Rome was mid-June, which meant the earliest possible date for ANVIL was mid-July. Thus he proposed immediately reducing
Mediterranean resources for ANVIL to a one-division lift, transferring the excess amphibious assets to OVERLORD, and reducing ANVIL to a strategic threat that might possibly be executed around the time of OVERLORD if the circumstances permitted. With the British agreeing to Eisenhower’s proposals, the JCS, on 24 March, reluctantly concurred.
Late in March, however, the JCS appeared to be trying to reverse the decision and insisted on scheduling at least a two-division ANVIL for 10 July. The American staff may have been concerned at this point about a renewed British interest in the Balkans. Recently, for example, Wilson had proposed using the one-division lift left in the Mediterranean for a variety of operations, including the establishment of a beachhead at the head of the Adriatic. As might be expected, the JCS opposed this project as well as another Wilson suggestion that the offensive in Italy be pushed to the Po River. Wilson had already estimated that he would have difficulties deploying more than eight divisions north of Rome until he had seized major ports in northern Italy; the Americans thus believed that he would not be able to employ usefully all of the Allied forces left in the theater. The JCS also estimated that the Germans could hold a defensive line north of Rome for six months or more, while at the same time retaining the ability to redeploy significant strength from Italy to the OVERLORD area. To the JCS, a reasonably early and strong ANVIL still appeared to provide the best means of supporting OVERLORD and of effectively employing Allied resources in the Mediterranean.
To accomplish this, the JCS promised the British that sufficient amphibious lift would be made available from American resources to execute a two-division ANVIL in July.9 However, the Americans also specified that the additional lift could be used only for the purpose of executing ANVIL on or about 10 July, and accompanied the offer with a proposal that Wilson halt his offensive in Italy south of Rome so that the ANVIL target date could be met.
Allied strategic discussions over the matter now reached an impasse. Reiterating old arguments against ANVIL and for Italy, the BCS submitted counterproposals assigning priority to Italy and allowing Wilson to use the additional amphibious lift as he saw fit. The JCS remained adamant, dissatisfied that the British were unwilling to accept an offer of additional resources without making any concessions in return. The BCS, in turn, believed that the JCS were attempting to force Wilson to adopt an American “strategy” in a theater for which the British had had primary responsibility since January.
Churchill, taking a hand in the discussions, proposed that the Allies again defer a final decision about the relative priority of Italy and ANVIL. He declared that unless the United
States made good its offer of additional amphibious lift, there could be no choice in the Mediterranean—priority would go to Italy by default.
Marshall, replying for the JCS, pointed out that unless the Allies began immediate preparations for ANVIL, there would also be no options in the Mediterranean. Moreover, the United States could not make any more resources available for a campaign—Italy—in which the Americans had no faith. If ANVIL was to assist OVERLORD, Marshall argued, it would have to take place before the end of July. To meet such a target date, Wilson would have to release ANVIL units from Italy by mid-May. Marshall estimated that Wilson could continue the offensive in Italy without the units needed for ANVIL despite BCS concern over a projected infantry shortage there.
On 8 April Wilson, completing plans for a spring offensive in Italy, informed the CCS that he could no longer wait for an ANVIL decision. The renewed offensive in Italy would require his entire strength, including those divisions earmarked for ANVIL. The earliest he could execute ANVIL was probably late July, and late August appeared more realistic.
Without consulting the JCS, the BCS directed Wilson to carry out his planned deployments.10 At the same time, they prepared a directive for a general offensive in Italy, employing all resources available in the Mediterranean, and submitted it for JCS concurrence. Almost as a footnote, the proposed directive also instructed Wilson to prepare the most effective threat possible against southern France at the time of OVERLORD.
Unable to obtain any commitment for ANVIL, the JCS approved Wilson’s new directive, but as their price they withdrew the offer of additional amphibious lift. Although Wilson was thus left with scarcely enough shipping for a one-division assault, the JCS agreed that he could use it as he saw fit. To all intents and purposes, the latitude of Wilson’s directive and the withdrawal of the American offer of amphibious lift meant the end of ANVIL.
For Wilson the ANVIL deferral was a welcome relief. Aside from the question of amphibious lift, the decision ostensibly removed the competition between ANVIL and Italy for cargo shipping, combat aircraft, U.S. and French Army divisions, and logistical resources. The decision also settled months of uncertainty regarding the means Wilson would have for the Italian campaign and enabled him to make final preparations for his spring offensive, scheduled to begin about 10 May, almost a month before OVERLORD.11
ANVIL Restored
For an operation that had so few consistent supporters, ANVIL proved to have remarkable staying power.
Wilson’s new directive was but a few days old when both the JCS and BCS, perhaps motivated by a desire to heal wounds left from the sometimes acrimonious exchanges of early April, took still another look at the possibilities in the Mediterranean.
After reexamining his resources late in April, Wilson informed the CCS that when his main forces had linked up with the Anzio beachhead, he could begin releasing sufficient strength from Italy for a major amphibious operation, but not with the one-division assault lift left to him. Among other alternatives, Wilson proposed an invasion of southern France or, with an ever-ominous sound to the JCS, a landing at the head of the Adriatic.
The BCS, studying Wilson’s proposals, hinted at the possibility of a small-scale landing in southern France by the end of June, and Churchill suggested a descent on France’s Atlantic coastline along the shores of the Bay of Biscay. Encouraged, the JCS renewed their offer of amphibious lift from Pacific allocations, but made it clear that they still favored a southern France invasion. In response the BCS suggested a number of alternative landing sites, including southern France, the Bay of Biscay, the Gulf of Genoa in northwestern Italy, or the west coast of Italy between Rome and Genoa. With obvious regard for American sensibilities, the BCS omitted any mention of a landing at the head of the Adriatic, although such an operation had recently loomed large in BCS planning discussions. Preparations, they recommended, should start immediately for participation by American and French units in whatever operation the CCS selected. On 9 May the JCS, although still primarily interested in ANVIL, accepted the British proposals as the basis for further planning. Although the CCS had still been unable to reach a decision on the future course of operations in the Mediterranean, at least they had reopened the door to the possibility of ANVIL sometime in July.
Early in June, after the Allies had seized Rome earlier than expected and had come ashore on the Normandy beaches, the Anglo-American debate on Mediterranean strategy reopened. Five alternative courses of action seemed feasible: (1) an ANVIL landing in the Marseille–Toulon area followed by an exploitation north up the Rhone valley; (2) an ANVIL landing in the Sete area, west of Marseille, and an exploitation northwest to Bordeaux; (3) an assault in the Bay of Biscay area, but only after OVERLORD forces had advanced as far south as the Loire River; (4) an advance in Italy north to the Po, followed by a drive west into France or northeast into Hungary through the Ljubljana Gap; and (5) a landing at the head of the Adriatic with a subsequent exploitation northeast through the Ljubljana Gap. Both American and British planners agreed that the first three alternatives required halting the advance in Italy at the Pisa–Rimini line.
Wilson regarded a drive to the Po and then northeast through the Ljubljana Gap as strategically more decisive. His operational planners estimated that an advance to the Po would force the Germans to deploy ten more divisions to Italy, thereby relieving pressure against Eisenhower
in Normandy. Yet reaching the Po would also probably necessitate retaining most of the divisions marked for ANVIL. Churchill and the BCS furthermore approved continuing the offensive in Italy to the Po, but deferred a decision on whether to swing west from there into France or northeast toward Hungary.
The JCS, while still firmly opposing an entry into southeastern Europe, had now become vitally interested in securing another major port to support Eisenhower’s OVERLORD forces. Although strongly favoring a landing in southern France, the American chiefs appeared willing to settle for the Sete–Bordeaux or Bay of Biscay alternatives. Whatever amphibious operation the Allies selected, the JCS insisted that it should comprise a three-division assault and should take place on or about 25 July. In addition, they assured the British that the United States would make available most of the amphibious lift required for a three-division landing.
The need to secure more ports to support OVERLORD lent a new note of urgency to the debate over southern France. The JCS had been concerned about the port situation since the inception of OVERLORD planning, believing that the southern French ports would prove vital for funneling more Allied divisions into France, especially the French forces in North Africa. While discussing the problem with Wilson earlier in June, Marshall had pointed out that the Channel ports clearly lacked the capacity to support all the forces that the Allies planned to pour into France. His remarks were underlined shortly thereafter by a great storm in the Channel that severely upset OVERLORD unloading schedules and that, coupled with the tenacious German defense and eventual destruction of Cherbourg, made the need for another major port to support the invading Allied armies even more obvious.
By this time Eisenhower saw the original three-division ANVIL concept as the best and most rapid method of securing a supplementary port. While hoping for an early ANVIL operation, he believed that a landing in southern France would still be of considerable help to OVERLORD if undertaken before the end of August. If this target date proved impossible to meet, Eisenhower continued, then all French divisions along with one or two veteran American divisions from Italy should ultimately be shipped to northern France through the Channel ports.
American planners quickly came to the conclusion that the second half of August was the only practicable time to initiate ANVIL. The Channel storm made it impossible for Eisenhower to release various types of landing craft from OVERLORD supply and reinforcement runs before that time; furthermore, because of meteorological conditions, mid-August was about the latest possible date for an ANVIL assault. Naval planners in the Mediterranean estimated that Allied forces in southern France might have to be supported over the beaches for at least thirty days, that is, until the seizure and rehabilitation of the port of Toulon, if not Marseille as well. But they could not guarantee over-the-beach support after the mistral, the strong, northerly winds along the coast of southern France that would
begin about 1 October. Therefore, 1 September was the latest safe date in 1944 for executing ANVIL, and any earlier date would ease potential problems.12
Wilson, meanwhile, had concluded that if the seizure of a major port was the primary consideration, then ANVIL was the best choice. The BCS, while reserving judgment about the need for another port, seemed to lean in mid-June toward the idea that ANVIL might prove desirable and necessary. By the third week of June, the JCS had decided that any further delay in reaching a decision could only result in another cancellation of ANVIL. Accordingly, on the 24th the JCS recommended to the BCS that the Allies halt in Italy at the Gothic Line (a German defensive network just north of the Pisa–Rimini trace) and that Wilson launch ANVIL as close to 1 August as possible.
Churchill, who still had his heart set on continuing the Italian campaign with a thrust northeast from the Po valley, now appealed directly to Roosevelt. Admitting that his proposals contained political overtones, he maintained that political objectives must be taken into consideration. Although introducing little that was new into the debate, he forcefully repeated all his old arguments against ANVIL and for Italy, pleading with Roosevelt not to wreck one campaign for the sake of starting another.
Advising the president, the JCS pointed out that Allied forces in Italy would suffer a net loss of only three divisions if Wilson executed ANVIL and argued that Wilson still had ample strength to drive to the Po. The JCS were convinced that Churchill’s real aim was to commit major Allied strength to the Balkans—although the British expressly denied such intentions—and did not even comment on the British contention that the capacity of the Channel ports could be easily expanded without recourse to the seizure of ports on the Atlantic and Mediterranean coasts.
Replying to Churchill, Roosevelt brought up political considerations of his own. He reminded the prime minister that the United States would hold national elections in November and noted rather obliquely that even a minor setback in the OVERLORD campaign would assure the president’s defeat. Diverting significant American strength into the Balkans or Hungary was dangerous. Roosevelt also reminded Churchill that the Allies had promised ANVIL to Stalin; as for Churchill’s reiterated geographical objections to ANVIL, Roosevelt pointed out that the terrain in the Ljubljana Gap region was even worse than that along the Rhone valley.
On 30 June the BCS backed off. They informed Churchill that, although they considered the Po valley-Hungary plan sounder, they were prepared for the sake of Anglo-American unity to approve ANVIL. Churchill, fearing another impasse in the Mediterranean, gave way, but could not resist the temptation to prophesy that Stalin would be pleased, for the execution of ANVIL would leave southeastern Europe open to Russian domination.
ANVIL was at last back in the Allied operational program. On 2 July the CCS directed Wilson to launch a three-division ANVIL on 15 August, reinforcing the amphibious assault with airborne units and following up with French Army divisions. The missions of the ANVIL forces were to seize the ports of Toulon and Marseille and exploit northward to Lyon to support future Allied operations in Western Europe. Wilson was to build up the ANVIL force to a total of at least ten divisions (most of them French) as soon as the tactical and logistical situations in southern France permitted. The Allies were to throw into Italy all other resources left in the Mediterranean, and Wilson was to push on up the Italian peninsula as best he could.
Churchill’s Last Stand
Churchill continued to view the decision with foreboding and some pique, feeling that the Americans had forced ANVIL down his unwilling throat. Even after temporarily dropping the issue, he continued to believe that if the Allies were to employ major strength west of Italy, they should seize a port on the Atlantic coast of France. Until early August the tactical situation in the OVERLORD lodgment area was such that Churchill could make no good case for an Atlantic coast venture. However, after the Normandy breakout he resumed his struggle against the ANVIL decision, arguing that the landing should be switched to Brittany where the American ANVIL divisions could play a more direct role in OVERLORD and where American reinforcements from across the Atlantic could be more easily introduced to the northern European battlefields.
The JCS quickly scuttled these last-minute proposals, noting that the Allies had little information about the conditions or defenses of the Breton ports; that the Atlantic beaches were beyond effective air support range of Mediterranean bases; that ANVIL would have ample air support; that the Allies had no plans for Atlantic coast operations; and that the Breton ports lay so much farther from Mediterranean staging and supply bases than southern France that insoluble shipping problems would be created for both assault and follow-up echelons. The JCS could see no merit in abandoning a carefully planned and prepared operation for the sake of securing what they considered only a hypothetically better line of supply and reinforcement for OVERLORD.
Stubbornly Churchill turned to Eisenhower, hoping to persuade him to recommend cancellation of ANVIL in favor of the Italian campaign. Judging that Churchill was still primarily interested in pursuing his Hungarian and Balkan projects, Eisenhower evaded his pleas, stating that he could speak only from the basis of military considerations and suggesting that if the prime minister wanted to ground his arguments on political premises then Roosevelt was the person to approach. Even Wilson proved of no real help to Churchill. Although supporting the Atlantic coast switch in principle, the Mediterranean commander-in-chief pointed out that the change would require at least a two-week delay in launching any assault by the forces already loading for
ANVIL. Churchill unhappily gave up his fight, but it was not until 11 August, only four days before the scheduled date for ANVIL, that the BCS issued Wilson a final directive to execute the operation.
To the end, Churchill remained unreconciled to the endeavor, termed it a “major strategic and political error,” and predicted it would prove a “costly stalemate” and ultimately a “cul-de-sac,” or dead-end.13 Perhaps the British prime minister had good cause for concern. The American forces leading the assault had been fighting in Italy for over a year and had left the battle area only recently; preparations and training for the amphibious landing, one of the most complex types of military operations, had been hurried and incomplete; and there was barely enough shipping to send troops over the beach and support them. The Germans in southern France clearly outnumbered the initial attackers by a figure of three or four to one; had strongly resisted every other Allied attempt to land on the Continent; and could easily send reinforcements from Italy or other fronts if they wished. In contrast, many of the French units scheduled to follow were untested and short of trained personnel and equipment; no further Allied reinforcements from Italy or Great Britain could be expected; and air support would have to be staged out of Corsica, nearly one hundred miles away. The Americans, Churchill feared, were taking on much more than they could hope to handle. On 15 August, as the ANVIL landings began, he arrived in a British destroyer for a ringside seat at what many believed was one of the gravest Allied strategic mistakes of the war.