Chapter XIV
Counting the Costs of TORCH
August-November 1942

The plan for Operation TORCH, as it finally took shape after the compromise of early September 1942, left London some room to hope for a quick victory in North Africa, while providing Washington with some assurance against the fear of a demoralizing defeat.1 The most likely result of the compromise was a long, expensive operation. The plan adopted was unfavorable to the prospect not only of a short, cheap campaign in North Africa but also of a campaign of any kind in Europe in 1943. A long campaign in North Africa would use the men and munitions, the ships and naval escort, needed for a great sustained operation of the kind the War Department has proposed to launch in Europe in 1943. And the steadfast unwillingness in Washington to risk everything on speed and surprise in North Africa did not favor the Prime Minister's hope of carrying out bold attacks by small mobile forces against other positions on the periphery of German-controlled territory.

The effect of TORCH on British and American strategy gradually became apparent in the late summer and the fall of 1942. First, the military staffs had to recalculate the initial requirements--in particular naval escort and air support--for the three simultaneous landings. These increases did not, of course, measure the increase in the total cost of the operation, which the staffs could not even estimate until after the landings, when they could at last decide what to expect, for the purposes of planning, from French authorities in North Africa, the German High Command, and the Spanish Government. If there should be serious initial opposition on the part of the French forces in North Africa, if there should be a strong German reaction in Tunisia followed by the movement of large reinforcements to the front, or if the Spanish Government should allow the movement of German forces into Spain and Spanish Morocco, the entire operation might be endangered and would certainly be prolonged.2 But even while so much remained uncertain, the two governments


and their military staffs had to begin reckoning the costs. If these were higher than the British staff had estimated as necessary to obtain the objective and higher than the American staff had believed the objective to be worth, it was also true that the costs could in part be charged off to the delays and compromise accepted for the sake of reaching agreement. If the two governments set a high value on agreement, they had to stand ready to pay the price for it.

The actual and prospective costs of TORCH, as they were calculated and recalculated from August through November 1942, had effects not only on planning for later British and American operations in Europe but also on making and fulfilling commitments to Allied forces in the other theaters of war. In the Middle East the threat of a renewed attack by the Afrika Korps, though eased by the arrival of British and American reinforcements in the late summer and early fall, remained real and immediate until the great British victory at EL Alamein, just preceding the TORCH landings. elsewhere the Allied situation remained precarious throughout the period.

On the Russian front German forces had overrun the Don and were penetrating the valley between the Don and the Volga. The Battle of Stalingrad, begun in August, lasted throughout the period. The Battle of the Atlantic was still going badly. The Chinese war effort was almost completely demoralized, and the prospect of a counteroffensive in Burma, based on India, was still very uncertain. in the Pacific the battle for, control of the Solomon Islands had become a desperate test of the troops engaged and of the intentions of the Japanese and American high commands. The initial and subsequent requirements of TORCH limited and unsettled American plans for helping all Allied Powers and conducting all American operations and thus gradually blurred the outlines of American strategic planning.

The Order of Priorities for Shipping

The principal projects for shipping American troops and materiel abroad that were bound to be affected by TORCH were five: (1) all BOLERO movements of ground and air force units to the United Kingdom; (2) the movement of U.S. Army Air Forces units and missions personnel to the Middle East and India; (3) the convoys to the USSR; (4) the relief of British troops in Iceland by part of a U.S. division; and (5) the movement (under CCS 94) of a U.S. division and fifteen air groups (to be diverted from BOLERO) to the South Pacific. On 4 August the British Chiefs of Staff recommended a revised order of priority for shipments, as follows: (1) TORCH, (2) convoys to the Middle East, (3) movement of U.S. Army Air Forces units to the United Kingdom, (4) the relief of Iceland, and (5) BOLERO.3 This proposal, which seemed to the U.S. Army planners reasonable, was brought before a meeting of the CCS two days later by Sir John Dill.4 The combined planners recommended that a high priority also be assigned to the Pacific theater.5 As


amended and approved by the CCS on 13 August, the new order of priority read:
  1. TORCH--(To take precedence over other shipping in the Atlantic while being mounted).
    Middle East
    Pacific Ocean
    Russian supplies shipped by way of the southern route.
  2. U.S. Army Air Forces to the U.K. and to China.
  3. Relief of Iceland.
  5. India and China

    NOTE.--If supplies are to be sent to Russia via the northern route, priority 6 is recommended.6

The fulfillment of the requirements of TORCH had a direct bearing on the execution of the rest of the program of shipping U.S. Air Forces units and missions personnel to the Middle East and India. Taking into account the primary needs of TORCH, the CCS on 13 August approved the recommendation of a committee of British-American transportation experts that the rest of the shipments scheduled for the Middle East and India be carried out, but that they should not be accelerated even though they had fallen behind schedule. These shipments could be accelerated only by using the Queen Mary and the Queen Elizabeth, which were the only ships left that were fast enough to make the North Atlantic run for BOLERO service unescorted. Since TORCH would for some time take up all available escort, the two Queens were the only troop-carrying ships that could be used on the run during the North African operation. Further interference with that run the CCS were not then prepared to accept.7 As it later turned out, the schedule as then approved for Middle East shipments left too little leeway for TORCH requirements.8 However, as King pointed out at the time of the decision, the CCS must then reserve ships for sending units to the Middle East in order to retain the option of sending them.9

The withdrawal of shipping and naval escort from the sea lanes in time to mount TORCH was certain to call into question other important commitments of the United States and United Kingdom. A striking example was the interruption of the convoys that went by the northern route to the Soviet Union. How long to continue sending these convoys depended on what date would be set for TORCH. On 12 September, when the mid-September convoy had sailed and the next was half loaded, the question as formulated in London was how likely it was that TORCH might be postponed beyond 8 November 1942. If that were likely, it might be desirable to run at least one convoy, accepting the postponement of TORCH until 15-19 November or perhaps later, if losses during the voyage were unusually great.10 While the mid-September convoy was still in dangerous waters, reports came in that


twelve ships had been lost.11 When the mid-September convoy was run, thirteen out of forty ships had been sunk, even though there had been an escorting group of seventy-seven ships of various types protecting the convoy. The Prime Minister attached so much importance to the continuation of the northern route convoys that even then he considered proposing that TORCH be put off long enough to allow for one more convoy.12 The Prime Minister ended by proposing instead to inform Stalin that, though large-scale convoys like that of mid September would be impossible for the rest of the year, he and the President were looking for some way to keep on sending supplies by the northern route on a smaller scale. At the same time he brought up again the possibility of operations in northern Norway. The chief strategic purpose would be to secure the northern route to Archangel and Murmansk. And to open staff conversations with the Soviet military staff on those operations, he believed, might in the meantime help offset the effect on the Soviet Government of interrupting the convoys.13

Both proposals received a cool reception in Washington.14 Nothing more was said, for the time being, about operations in northern Norway. Shipments were reduced to the movement of unescorted merchantmen, one at a time. from Reykjavik, to the Russian White Sea ports. In mid-December, convoying began again on a smaller scale.15

Pacific requirements were not so readily reduced. To the continued heavy demands of the Pacific bases were added, during the TORCH period, the requirements for sustaining the Solomons operation in the South Pacific.16 The. Solomons operation was in direct competition with TORCH for combat loaders.17 Arid the needs for naval support of TORCH, as finally planned, were so great that it was out of the question to transfer from the Atlantic to the Pacific any U.S. Navy units to help meet the critical situation in the Solomons. The situation was so tight that it was not until early September, when the President and


the Prime Minister were about to agree on a compromise version of TORCH, that the Navy finally furnished a definite list of U.S. naval vessels available for TORCH.18 According to that list, the most that the Navy could spare for the North African venture was one modern battleship, two old battleships, one aircraft carrier, two converted aircraft carriers, two 8-inch cruisers, three large 6-inch cruisers, forty destroyers, and six fast minesweepers.

The most dangerous weakness in both oceans, as Admiral Turner had feared, was the want of aircraft carriers. By October 1942 four of the seven carriers with which the United States had entered the war had been Sunk in the Pacific--the Lexington, the Yorktown, the Wasp, and the Hornet.19 The latter two carriers were lost during the contest for Guadalcanal.20 In addition, the Saratoga and Enterprise had been damaged by the Japanese during the naval battles for Guadalcanal. In November the Pacific Fleet was down to its last active aircraft carrier, the Enterprise, and even that survivor was damaged and out of action for most of the month. The only large aircraft carrier remaining was the USS Ranger of the Atlantic Fleet, and since the Ranger was the only carrier at all likely to be available to protect General Patton's forces during the landings on the Atlantic coast of French Morocco, it could not be withdrawn from the Atlantic to reinforce the U.S. Pacific Fleet.

The new urgent demands for shipping and escort affected other claims on shipping and escort, lower on the list of strategic priorities, until the success of operations in North Africa and the Solomons was assured. It was necessary once more to put off the long-planned relief of the British troops that remained in Iceland.21 The movement of service troops to Iran had also to wait on developments in North Africa, in spite of the desire of the President and the Prime Minister to accelerate the movement of Soviet lend-lease traffic through the congested Persian Gulf ports to northern Iran.22

The want of ships and naval escort furnished the War Department strong grounds for pleading once again that the United States could not give substantial military Support to China, much less satisfy Chiang Kai-shek's "three demands" of 28 June 1942. These three demands represented Chiang's summary of requirements in terms of ground and air forces, and lend-lease tonnage for the maintenance of the China theater--three American divisions, 500 planes, and 5,000 tons monthly airlift into China.23 The War Department recommended to the President on 9 October 1942 that

. . . the extremely serious shortage of ocean shipping for troop transport, including Naval escorts for such convoys through dangerous waters, not to mention the long turn around to India, make it utterly impracticable this fall to send and maintain United States Divisions in the China India theater. . . . The United States is waging this war on far flung


fronts and demands for men and particularly materials and ship tonnage are now beyond our present capacity.24

Similar restrictions also had a direct bearing on postponing operations for ejecting the Japanese from the Aleutians. At a time when all available means were being used either to mount TORCH or to bolster the precarious position in the southwestern Pacific, the United States could not afford to begin operations in what was, by common consent, an indecisive theater. During October and November 1942, General Marshall repeatedly refused General DeWitt permission to assemble forces for an operation in the Aleutians. The Army and Navy agreed that neither the shipping nor the troops could be made available.25

The search for escorts for TORCH focused the attention of U.S. Army planners and the military chiefs on Allied programs of shipbuilding and ship allocation, which needed to be reviewed in the light of the new plans and the heavy toll of Allied shipping still being taken by German submarines in the Atlantic.26 The program for producing landing craft under the BOLERO plan had delayed the completion of aircraft carriers and superseded the construction of escort vessels. It seemed clear that U.S. naval construction should be shifted back from landing craft to escort vessels.27 In early October the CCS approved allocations of American production of landing craft to cover the revised operational needs for the rest of 1942.28 Before the close of 1942 the JCS took measures to secure a review of the whole Allied shipbuilding program, and an increase in the production of escort vessels and merchant shipping.29

These actions at the end of 1942 constituted an acknowledgment that the effects of TORCH on the Allied shipping situation would be prolonged far into 1943. Allied operations in North Africa, at first severely limited by existing port and overland transport capacity, and still limited by the size and frequency of the convoys that the British and American naval commands would run with the available escorts, could not


as yet be sustained on a big enough scale to overcome the large forces the Germans were moving into Tunisia. In North Africa, as in the Solomons, the issue became a test of the willingness and ability of both sides to meet the demands of air operations for which neither side was well prepared--to maintain the flow of their own supplies and reinforcements and to interdict the flow of enemy supplies and reinforcements to the front. The effects of haste and waste, the rate of attrition, and the scale and duration of the effort in North Africa depended largely on the willingness of the German High Command to invest in the continued defense of a position that must sooner or later be abandoned. It was, therefore, impossible to calculate with any certainty just how serious the limiting effect of TORCH on Allied shipping schedules might be. But it was evident that the demands of TORCH and the losses incurred would bear heavily on Allied shipping schedules. The War Department planners concluded that in any event, unless current commitments were altered or canceled, no new operations could be launched by the United States for several months to come.30

Allotment and Preparation of Ground Troops

The problem of making ground strength available for TORCH was complicated for planners on both sides of the Atlantic by their uncertainty how many divisions would be used in the operation, and what would be the precise composition of assault and follow-up forces. The original decision that only American troops should be used in the assaults soon had to be changed. Only ten regimental combat teams, two armored combat commands, and a Ranger battalion were available.31 Few of these troops, moreover, had received the necessary amphibious training. In setting aside ground forces in the United States for TORCH and in allocating the necessary priorities, Army planners in the United States calculated in the summer of 1942 on a basis of seven divisions from the United States.32 In one combination or another, these almost always included the 3d, 9th, 36th, and 45th Infantry Divisions, 2d and 3d Armored Divisions, and the 4th Motorized Division, in addition to the 1st and 34th Infantry and 1st U.S. Armored Divisions in the United Kingdom. Accepted political strategy and logistical considerations required that the United States furnish as large a part as possible of the total expeditionary force. A more definitive determination of the total number of troops to be employed--both British and American--was introduced with the promulgation of the 20 September outline plan. According to that plan the United States was ultimately to furnish about seven divisions and two regimental combat teams; the British would furnish four to six divisions.33

The problems of furnishing fully trained and equipped troops for the assault forces


from the United States and United Kingdom continued to plague the planners almost to the eve of the actual launching of the operation. Combat-loading troop transports were to be available in time for the operation--at the immediate expense of troop shipments to the United Kingdom--but there was all too little time to train and rehearse crews to handle the debarkation of men and equipment and the assault troops themselves.34 The need for such training affected not only the date of launching the operation but also the choice of troops, for it required the use in TORCH of all available Army troops that had had any training in landing operations.35

Informal agreement had been reached on 18 July between War and Navy Department representatives on "amphibious" training and organization.36 This arrangement provided for training three Army engineer amphibian brigades and an amphibious corps of two or more Army divisions.37 The original reason for the Army's undertaking to train amphibian brigades was the anticipated need for the projected cross Channel operations (SLEDGEHAMMER-ROUNDUP) and the inability of the Navy to provide sufficient boat crews within the prospective time available.38 After the shift to TORCH, the need for training amphibian brigades continued to exist--and with time pressing more heavily on Army authorities than ever. Though the Army Navy understanding of 18 July was never formally approved by the JCS, it continued to serve as if it had been, so far as preparations for TORCH were concerned.39

Even before the terms of the agreement were presented formally to the JCS in early August, three amphibian brigades had been activated and were in training. One of these brigades, with a strength of about 7,000, was set tip to load, man, and unload assault craft for an entire division. As a result of the TORCH decision, however, the Army postponed the organization of two additional brigades that it had scheduled for activation in August.40


The training of Army divisions for assault landings--which was also subject to dispute with the Navy--was thrown into even greater confusion by TORCH, confusion aggravated by the uncertainty that existed during August over the composition of TORCH forces, and especially over the composition of the assault forces that were to sail from the United States. As Handy observed on 7 August, the assault force I from the United States must consist either of two infantry divisions or of one infantry division and one armored division.41 These possibilities affecting the disposition of the 3d and 9th Infantry Divisions raised a number of corollary questions for the Army planners. If only one of these divisions were used in the assault landings, which one would be chosen: Should the other be used in the follow-uip for TORCH or be dispatched to meet commitments to the Pacific? 42 Faced with the necessity of speeding amphibious training for the assault forces for TORCH, the Army planners in early August disregarded, for the moment at least, possible far-reaching consequences of setting aside both divisions for possible use in TORCH landings.43 Making allowance for the uncertainty of the composition of the assault force from the United States, military authorities moved quickly to set tip the Atlantic Amphibious Corps (Maj. Gen. Jonathan NV. Anderson, commanding) with the 3d and 9th Divisions and the 2d Armored Division.44

There was no unity of command in TORCH until the expedition set sail from the United States. For training, the Atlantic Amphibious Corps, designed as part of Patton's task force for TORCH, came under the general supervision of Admiral Hewitt, Commander Amphibious Force Atlantic Fleet. Army and Navy authorities tried in the summer of 1942 to straighten out the lines of command for that corps--a test case in joint Army-Navy planning and training.45 The temporary arrangements adopted for amphibious training and organization in preparation for TORCH by no means settled, but rather drew attention to, the jurisdictional problems that would have to be resolved.if training for assault landings was to keep pace with plans for amphibious warfare in the Atlantic and the Pacific.

Problems of training, equipping, and utilizing amphibious forces for the TORCH landings arose also across the Atlantic, in the British Isles. In the summer of 1942 Eisenhower's headquarters had to decide whether the 1st Infantry Division, already in England, could be used in the amphibious assault force sailing from the United


Kingdom. It was better trained than the 34th Division, stationed in Northern Ireland, and was, therefore, the choice for leading the assault force. In mid-August, however, a ship carrying nearly all its medium and about a third of its light artillery, weapons together with other equipment went aground off Halifax on the voyage from the United States to the United Kingdom. Eisenhower was at that time planning on an early or mid-October date for launching TORCH. This mishap required him to train assault troops from the 34th Division instead of from the 1st.46 In Washington the Army planners speedily set in motion War Department machinery to send to the New York Port of Embarkation weapons to replace those carried in the ship which had run aground.47 Even so, the schedule left so little leeway that Eisenhower was unwilling to commit himself to using the 1st Division though he ordered it held ready to be trained in the event the invasion was put off until November.48 Late in August, when it became clear that the operation would not be launched until November, plans were made to use the1st Division along with elements of the 34th Division in the assault force sailing from the United Kingdom. These plans were confirmed upon the agreement of the President and the Prime Minister on 5 September.49

Equipping and training armored forces introduced further problems of urgency and difficulty for the Army planners. In, early August planning for TORCH generally began to assume that the assault force for Casablanca would probably be one armored division and one infantry division instead of two infantry divisions, even though this change would require additional combat loading vessels.50 Besides the probable use of the armored division for the assault, all plans called for another armored division from the United States. In order to provide another trained armored division besides the 2d then receiving amphibious training, the 3d Armored Division was transferred from Camp Polk, Louisiana, to the Desert Training Center, California, for training and maneuvers. On 2 September it was designated for General Patton's Task Force "A." After completing maneuvers in mid-October 1942, it was transferred on 24 October to Camp Pickett, Virginia, for the Western Task Force. Shortage of shipping, however, finally precluded its being used in TORCH.51

The shuttling of the 3d Armored Division back and forth across the country illustrated the difficulties of planning during the summer and fall of 1942 as a result of uncertainty over the probable deployment even of major combat elements. The movement of that division was one of three large rail


movements to which the Chief of Staff in the fall of 1942 called the attention of his staff. Though he conceded that sudden demands, state of training, and deficiencies in equipment had forced such moves in the past, he believed that there were more of them than necessary.52 The Army planners explained the shuttling of the 3d Armored Division on the grounds that no similar unit near the east coast had had desert training.53 The two other large-scale transcontinental movements noted by the Chief of Staff were also related by the Army planners to the uncertainty over TORCH. Both the 43d and the 29th Divisions, involved in these shifts, had been moved to new stations in the uncertain period before the final determination of requirements for TORCH. The tentative allocation of seven divisions to TORCH left very few divisions available in the United States for other uses.54

The build-up for TORCH drew heavily on U.S. ground and supporting units in the United States and in the United Kingdom. As Marshall pointed out at the close of October, eight or nine divisions in the United States had been stripped of so many trained men to fill units for TORCH that six to eight months would be required to restore them to their former level of efficiency.55 Efforts to meet Eisenhower's needs for service troops, he added, had resulted "almost in the emasculation" of remaining American units.56 The reserves of the Army were drained for TORCH. To the demands of TORCH on units in the United States were added the heavy demands on American strength in the British Isles--the 1st and 34th Infantry Divisions, the 1st Armored Division, and the 1st Ranger Battalion, with supporting troops transferred to North Africa in the fall of 1942 for service with II Corps.57

Of course, with the heavy demands for troops went correspondingly heavy demands for equipment. According to the calculations by Army planners on 2 August, two infantry and two armored divisions in the United States would be equipped on or about 10 October, and three additional divisions (one motorized) could be equipped later in the fall.58 In effect, the only divisions in the United States that would be fully equipped before the close of 1942 were divisions that had to be ready for TORCH. The actual demands of TORCH on divisional equipment. in the summer and fall of 1942 confirmed--in large measure--these calculations. Divisions in training in the United


States and available for shipment to other theaters were stripped of equipment.59 The extent of that depletion led Marshall to observe in the fall of 1942 that in mounting TORCH the War Department had "scalped" units in the United States for equipment.60 The demands of TORCH also cut deeply into the American supplies and equipment that had been accumulated in the British Isles, arid were due to limit accumulation during the next few months.61

Provision of Air Units

It was evident from the beginning that most of the American air units for operations in North Africa. like most of the ground and service troops, would have to come from resources previously allotted for the projected major cross-Channel operation. CCS 94 expressly provided that all American heap N arid medium air units in the United Kingdom would be available for TORCH. It had soon thereafter been accepted that TORCH could not be carried out on any other basis.62 The rest of the TORCH air force would come from the United States, from units scheduled to go to the United Kingdom arid to the Middle East..

In early August Army planners estimated that withdrawals for TORCH would leave very little air strength in the United States for other uses in 1942. 63 Activation of new units in the United States would have to be deferred to provide replacements for losses in TORCH.64 All that could be shipped to England during the rest of 1942, over and above TORCH requirements, would be five bomber groups in September and six troop carrier groups from August through October. Of the fifteen groups to be diverted from BOLERO to the Pacific (under CCS 94), the first would not become available till December.

The TORCH air force, as projected in mid-August, was to consist of two heavy bomber groups, three medium bomber groups, one light bomber group, four fighter (two P-38 arid two Spitfire) groups, and one troop carrier group.65 The Eighth Air Force, that in the early stages of testing the American doctrine of high altitude daylight


bombing, held the main A AF' resources as Well as the most highly trained men available for service in Africa.66 The Eighth Air force was charged with the organization, planning, and training of the new air force for North Africa. The Eighth was also to contribute, its heavy bombers, and on an order from Eisenhower on 8 September it had to discontinue operations from the British isles, notwithstanding the protests of the Army Air Forces.67 Except for heavy bomber units, most of the commands of the TORCH air force the Twelfth) were activated in the United States from units previously designated for the Middle East.68 These had to be hurriedly prepared and sent to England in time to be indoctrinated and assimilated, a task hard in itself and made harder by bad weather in the North Atlantic, which played havoc with the ferrying of medium and light bombers.69

The most pressing and serious problem in allocation of air units for TORCH was a shortage of fighters and observation planes, particularly long-range models. General McNarney stated the problem on 5 September in response to a proposal from the Navy that P-38 reinforcements be sent to the South Pacific:

The reinforcements which propose can only be effected by diversion front TORCH. All the P-38's now in the U.K. or being organized in tire U.S. for movement to U.K. are allotted to TORCH and the number is believed to be insufficient. other tighter planes can make the long initial flights required across, the Atlantic or from U.K. to Casa Blanca [sic] and Oran but the P-38 type. If we withdraw these planes we, in effect. impose a drastic change, if not the abandonment of TORCH.70

The shortage of fighter planes was so serious that it could not be met by using all American units in the United Kingdom together with those in the United States available for BOLERO. American planning for a TORCH air force--pushed by Patton and Doolittle--proposed, therefore, using P-39's in England in transit to the Soviet Union and the 33d Pursuit Group (P-40's) which w as in the United States and awaiting shipment to the Middle East. 71 The release to TORCH Of the P-39's en route to the Soviet Union was arranged by Eisenhower with the Prime Minister. The United States undertook to replace them via Alaska as soon as practicable.72

The release to TORCH of the 33d Pursuit Group was less readily arranged. On 8 September the formal proposal was submitted in a War Department letter to the JCS.73 The letter stated that the reallocation of the 33d Group was required for the U.S. air force planned for TORCH. Reaction in Washington to this proposal--as in London to a similar proposal of General Doolittle--


was mixed, because of a rather general belief that Allied air superiority in the Middle East would help assure the success of TORCH.74 But the JCS agreed to recommend the War Department proposal to the CCS and at the same time authorized General Arnold to seek the informal concurrence of the British Chiefs of Staff. 75 Arnold thereupon wrote to Air Marshal Douglas C.S. Evill of the British Joint Staff  Mission for his concurrence. Evill did not concur, in view of the need for fighter planes for the Middle East.76 In order to resolve the problem the CCS agreed on 18 September to refer it to Eisenhower for his views.77 Following a discussion with Doolittle, commander of the Twelfth Air Force for North Africa, Eisenhower agreed that the 33d Pursuit Group should be diverted to TORCH as proposed, but he also recognized the need for sending fighter planes as reinforcements to the Middle East and the bearing on TORCH of air superiority in the Middle East. The British Chiefs of Staff, concurring, called attention to Eisenhower's reservations.78 The 33d Group was assigned to the Twelfth Air Force and its P40's were launched from an auxiliary aircraft carrier accompanying the assault convoy to Casablanca. Though the Middle East had been given a priority in shipping second only to TORCH itself, the limited Allied resources available in the summer and fall of 1942 left little leeway beyond the fulfillment of requirements of the number one priority, TORCH.

In meeting the claims of TORCH the Army also left unsatisfied the Navy's continued demands for substantial air reinforcements for the Pacific. In August 1942 the problems of immediate and eventual air reinforcements for the Pacific were merged with the question of TORCH requirements. Since August General Marshall had conceded that one group of heavy bombers should go to Hawaii and had relaxed restrictions on the use in the South Pacific of bombers assigned both to Hawaii and to Australia.79 But there remained as a source of disagreement between the services the broader question of priorities to govern the assignment of the remainder of the fifteen groups scheduled for withdrawal from BOLERO as they became available in succeeding months. Army planners--in accord with AAF views--continued to argue in September that there be no further diversions to the Pacific--beyond the heavy bombardment group currently authorized for Hawaii until the requirements of TORCH, the Middle East, and the United Kingdom had been met.80 In supporting the AAF position in joint planning discussions, Army planners observed that there was some doubt that facilities available in the South Pacific could support more aircraft than were en route or present. Navy planners, agreeing that


TORCH and the Middle East should hold top priorities, countered that diversions to the Pacific should precede further deployment to the United Kingdom (BOLERO).

The decision to invade North Africa was not at all hard to reconcile with the great aim of the Army Air Forces--strategic bombing against Germany. Through the use of alternate air bases in the Mediterranean to complement long-range strikes from the United Kingdom, the Air staff hoped to minimize the effects of the change in plans. A difference of opinion arose with the Navy over the relations of the strategic air offensive to ground operations in Europe. The Navy held that the projected bomber offensive from the United Kingdom could not be considered apart from a European invasion and that TORCH had postponed the one as well as the other, thereby permitting the release of aircraft for use in the Pacific and elsewhere. The Air staff argued strongly that strategic bombardment, as originally conceived and as it must still be conceived, was a separate offensive operation, related to but distinct from a European invasion. Delaying the invasion had left a theater that, in the immediate future, would become purely an Air theater, requiring more than ever the concentration of air power against Germany.81

These divergent views were further elaborated on the JCS level. Arnold maintained that air forces operating in the United Kingdom and the Middle East were directly complementary to TORCH and must be kept in the same priority.82 He cited the view of Eisenhower, Patton, Clark, and Spaatz to support his argument. King continued to maintain, as in August, that the CCS had released the fifteen groups for deployment to the Pacific, and that the situation there demanded they be sent.83 Arnold replied that the decision to launch TORCH had not altered the Allied strategy of concentrating against Germany, and that TORCH--in conjunction with the development of strategic bomber offensive--promised the most decisive results of any pending Allied operation. He held that the withdrawal of any of the fifteen groups would preclude the success of the operation.84 Marshall and Leahy held to a middle-of-the-road policy: TORCH and the Middle East were to take precedence, and the allocation of new units would be decided as they became available.85 Marshall added (as he had earlier told Eisenhower) that he regarded the main purpose for the American proposal to withdraw the fifteen groups from BOLERO as the transfer of jurisdiction over their final assignment back to the JCS.86 Further discussions were postponed until Arnold, accompanied by Brig. Gen. St. Clair Streett, Chief, Theater Group, OPD, could make an inspection of the facilities available in the Pacific.

The upshot of the discussions in the joint staff and of the Arnold-Streett survey was


an agreement reached by the end of October 1942 that the uncommitted balance of the fifteen groups withdrawn from BOLERO was to form a part of a general United States strategic air reserve precisely as Marshall had intended.87 Claims on. air units for operations against Japan would, as before, be weighed against claims for operations across the Atlantic. In effect, General Marshall had regained some of the freedom of action he had lost in the spring by proposing to give absolute priority to the concentration of American forces in the British Isles.

Effects on Plans for a Cross-Channel Operation

The War Department Thesis

The great initial withdrawals of BOLERO units for TORCH, the related withdrawal of BOLERO air units for future disposal, the improbability that the American version of TORCH would allow of a quick victory and the corollary probability that many deferred claims against Allied resources would accumulate for several months, all tended to confirm the contention of American military leaders, expressed in the London conference of July, that TORCH would almost certainly entail the postponement of the major cross Channel effort scheduled for the spring of 1943.88 In early August, Marshall and his staff restated this view. They believed it probable that TORCH would not merely delay ROUNDUP but would be, in effect, a Substitution for that undertaking in 1943 They were quite certain that in any event the movement of troops to the British Isles would be considerably reduced for at least four months after the assembly of shipping and escorts for the assault,landings for TORCH began. And, in Marshall's opinion, the invasion of French North Africa, undertaken with due allowance for the uncertainties involved and with a determination to sec it through to a successful conclusion, would preclude the "offensive" operations "directly" against Germany contemplated in the original document on "American British Grand Strategy," dating from the ARCADIA Conference.89

Slowdown of BOLERO

By the late summer of 1942 the War Department had a fairly well-defined idea what revisions must be made in the BOLERO troop basis down to the spring of 1943 and how the mission of Army forces during that time should be redefined to fit the new conditions produced by the deviation from the strategy of SLEDGEHAMMER-BOLERO-ROUNDUP. According to the revised Army planning for its forces in the United Kingdom to the spring of 1943, the U.S. air force was to be built up in the United Kingdom to increase offensive operations against the Continent; a balanced ground force was to be maintained in the United Kingdom as a reserve for TORCH, for the defense of the United Kingdom, and in preparation for emergency


action on the Continent. Toward the close of the summer the Chief of Staff accepted the Army planners' proposal for a balanced ground force of 150,000 U.S. troops in the United Kingdom.90 They had pointed out to him, on 27 August, that the change in strategic policy from BOLERO to TORCH had by that time resulted in stopping the movement of major ground force elements to the United Kingdom.91 A great number of supporting combat and service troops had been prepared for movement to the United Kingdom on the basis of the BOLERO requirement of an over-all force of about one million men by April 1943. The continuation of shipments of these troops would not only result, his staff planners observed, in stripping the United States of such troops but would also lead to an unbalanced ground force in the European theater. They therefore called for a balanced ground force, similar to the one envisaged early in the war-for the purpose of relief or defense under the MAGNET (Northern Ireland) plan.

According to the revised War Department estimates of the late summer of 1942, the air forces in the United Kingdom would total approximately 95,000 by 1 April 1943.92 That number represented the original air force figure set up for BOLERO, less 100,000 to be diverted for TORCH. Services of Supply troops (about 60,000 to support this air force, as well as the projected balanced ground force) would give the United States a total force of about 305,000 in the United Kingdom by 1 April 1943. By 30 September 1942 the Army would have 160,000 troops in the United Kingdom or en route, over and above the forces required for TORCH. In order to bring the force in the United Kingdom up to the total strength of 305,000 by 1 April 1943, it would be necessary to ship 145,000 troops there. The use of the fast-sailing and unescorted Queens on the North Atlantic run appeared to be the most practicable means of expediting these shipments without interfering with TORCH.

In early November 1942 the War Department tentatively approved, for planning purposes, a new reduced strength for American forces in the United Kingdom set at approximately 427,000.93 This figure represented an increase of over 100,000 above the original estimates of the late summer. Shortly thereafter--on 12 November--in submitting his revised estimates for the European theater to General Marshall, Maj. Gen. Russell P. Hartle, Deputy Commander, European Theater of Operations, stated that, as of about 30 November 1942, there would be slightly more than 25,600 U.S. Services of Supply troops left in the United Kingdom.94 About 84,800 more men would be required to meet the estimated figure of 110,463 SOS troops. He indicated that after the withdrawals for the North African operation, United States ground forces in the United Kingdom would total, as of about 30 November, only 23,260 troops--including the 29th Infantry Division. Over 136,000 more ground


force troops would be needed to reach an estimated total of approximately 159,000. In an accompanying note General Spaatz, the commanding general of the Eighth Air Force, stated that combat units of the Eighth Air Force that would remain in the United Kingdom after the departure of the Twelfth Air Force would be seven heavy bomber groups, one single-engine fighter group, and one observation group. Additional combat units scheduled for the United Kingdom in November and December included one medium bomber group, one twin-engine fighter group, and one troop carrier group. General Spaatz pointed out that the Twelfth Air Force had priority in the European theater. The only ready source from which replacements for the Twelfth could be drawn was the Eighth Air Force, which was also actively engaged. The process of withdrawing aircraft and combat crews from the operating organization of the Eighth Air Force, he observed, had already begun. Unless steps were taken to counteract this trend, the Eighth was likely to be bled of its operating strength. He recommended that a sustained air offensive against Germany be made the principal mission. of American forces in the British Isles, and that their growth be controlled accordingly.

Thus the trend in Army planning during the fall of 1942 was to increase the proportion of air and supporting service troops in the British Isles, although the staff still planned to have a "balanced" ground force of about 150,000 there by the spring of 1943. The tentative plans for increasing American forces in the British Isles in part reflected the close dependence of they Twelfth Air Force on the Eighth. In part, they also reflected the agreement of Marshall, Arnold, Eisenhower, Spaatz, and their advisers that air operations against Germany should be resumed and intensified during the North African campaign.

Even on this reduced scale, the schedules for the BOLERO movements could not be met with the trained and equipped ground combat units and cargo shipping then available. In the latter part of October Army planners estimated that the troop lift of the four remaining convoys to the United Kingdom for the balance of 1942 would be only 4,000, 3,300, 8,000, and 8,000, these figures representing the maximum which cargo shipping could support.95 In early December the Chief of Staff called the attention of the President to the fact that the monthly flow of United States troops to the United Kingdom was then only 8,500.96 Troops were moving even more slowly than the Army had wished or expected.

The Army planners had not given up the idea that the United States and Great Britain must save their strength to engage and defeat the German Army in northwestern Europe.97 But this idea, the polestar by which the planners had steered, had been obscured; they had been thrown off their course; and they were no longer even sure of their position. The day of landing in France seemed as far away as it had six months before, or further. To gather huge ground forces in England to await a hypothetical


break in German military power appeared neither possible nor desirable, particularly in the light of other and more immediate demands. If the British remained unwilling to agree to a cross-Channel offensive until German military power was broken, there remained the "Pacific Alternative," and the Army planners once again argued for its adoption in that event. Clarification of the subsequent lines of strategic action in the European theater for 1943 for the ultimate defeat of Germany would have to await the outcome of current operations and basic decisions of top Allied political leaders. Meanwhile, the War Department staff strove to keep alive the idea that it would finally prove necessary to undertake a very large cross-Channel operation against a still formidable German Army, while the Air staff further explored the idea that in any event a great air offensive over the European Continent from bases both in the British Isles and in the Mediterranean should have the first claim on American air forces.


In a conference with General Eisenhower and his staff during the latter part of September, the Prime Minister took notice of the effect of the North African operation on the War Department's plans for 1943.98 As Eisenhower wrote to Marshall immediately after the conference, it appeared that "for the first time the Former Naval Person [Churchill] and certain of his close advisers" had "become acutely conscious of the inescapable costs of TORCH." Eisenhower went on to observe:

The arguments and considerations that you advanced time and again between last January and July 24th apparently made little impression upon the Former Naval Person at that time, since he expresses himself now as very much astonished to find out that TORCH practically eliminates any opportunity for a 1943 ROUNDUP.

The Prime Minister could no longer simply assume, Eisenhower pointed out, that TORCH could be reconciled with Soviet expectations of a second front and of material aid:

Although the memorandum prepared by the Combined Chiefs of Staff. when you were here, and later approved by both governments, definitely states that the mounting of TORCH would in all probability have to be a substitute for 1943 ROUNDUP, while the several memoranda you presented called attention to the effects of TORCH upon the possibilities of convoying materials to Russia and elsewhere, these matters have now to be met face to face, and with an obviously disturbing effect upon the Former Naval Person.

The Prime Minister was still quite unwilling to acknowledge that TORCH would strain United States and British resources to the utmost, for that would be, in effect, to acknowledge that the United States and Great Britain would remain in 1943--as they had been in 1942--unable to meet the expectations of the Soviet Government with reference either to the shipment of supplies or to the establishment of a "second front." He declared that the United States and United Kingdom could not confess to an inability to execute more than a thirteen division attack in the Atlantic theater during the next twelve months.99 They must not acknowledge that TORCH left nothing to spare.

The Prime Minister wrote to the President that the conference with Eisenhower


and other American officers had left him much troubled on that score, saying "I gained the impression at the conference that ROUNDUP was not only delayed or impinged upon by TORCH but was to be regarded as definitely off for 1943. This will be another tremendous blow for Stalin. Already Maisky [Soviet ambassador to Great Britain] is asking questions about the spring offensive." The Prime Minister ended his message by saying, "To sum up, my persisting anxiety is Russia, and I do not see how we can reconcile it with our consciences or with our interests to have no more P Q's [northern route convoys to Russia] till 1943, no offer to make joint plans for JUPITER, and no signs of a spring, summer, or even autumn offensive in Europe.100

The Prime Minister's discomfort over the probable elimination of ROUNDUP as a possibility--not necessarily to be realized for 1943 was all the greater when he learned, in the fall of 1942, of the War Department's definite plans for scaling down the BOLERO preparations in the United Kingdom. In the latter part of November there came to his attention a letter from General Hartle stating that under existing directives from the War Department any construction in excess of requirements for a force of 427,000 would have to be done by British labor and materials.101 Lend-lease materials, the War Department had stated, could not be furnished for these purposes. The Prime Minister took the occasion to sound out the President on the meaning of this great reduction from the original estimates tinder the BOLERO plan to have 1,100,000 American troops in the British Isles by 1 April 1943. He took the reduction to indicate that the United States had given up planning for an invasion in 1943. To abandon ROUNDUP, he declared, would be "a most grievous decision." He pointed out that TORCH was no substitute for ROUNDUP and only employed thirteen divisions against the, forty-eight projected for ROUNDUP.102 He reported that although his previous talks with Stalin had been based on a postponed ROUNDUP he had never suggested that a second front should not be attempted in 1943 or 1944. One of the arguments he himself had used against SLEDGEHAMMER, the Prime Minister added, was that it would eat tip in 1942 the "seed corn" needed for a much larger operation in 1943. Only by building up a ROUNDUP force in the United Kingdom as rapidly as other urgent demands on shipping permitted could the troops arid means be gathered to come to grips with the main strength of the European enemy nations. The Prime Minister conceded that, despite all efforts, the combined British-American strength might riot reach the necessary level in 1943. In that case, he believed that it became all the more important to launch the operation in 1944. He asked that another British American conference be held, either in London, with Hopkins representing the President (as in July), or in Washington as in June.

General Smith, Eisenhower's chief of staff, reassured the Prime Minister that the War Department directive on authorized construction in the United Kingdom referred only to the necessity of keeping BOLERO preparations in the United Kingdom


in line with the revised estimates in the anticipated troop build-up.103 He pointed out that, as had been agreed during the July conference in London, TORCH commitments made ROUNDUP improbable in 1943 and necessitated revision of BOLERO Estimates based on the temporarily reduced troop lift. Other operations that the Prime Minister Was urging could only be mounted at the expense of TORCH and would have the same effect. lie reassured the Prime Minister that none of these considerations, however, implied any change in the American conception of the BOLERO-ROUNDUP plan.

This was not the kind Of assurance the Prime Minister needed. The Prime Minister wanted to continue operations in the Mediterranean after gaining control of the coat of North Africa, with an operation against Sardinia (BRIMSTONE).104 American officers had therefore some reason to go on discounting the Prime Minister's assertions about ROUNDUP. They knew that lie was anxious lest American forces be committed to larger offensive operations in the Pacific, and lest it be alleged he had dealt in bad faith with the Soviet Union. The kind of operation actually being undertaken in French North Africa, over the protests of London, was hard to reconcile with the idea of undertaking an operation of am kind on the Continent in 1943. The Prime minister could hardly expect, therefore, unqualified reassurance that the President still thought that TORCH did not rule out ROUNDUP. But he could expect and wanted a declaration leaving open the possibility of some such operation.

Such a reassurance he soon received from the President.105 The President reminded him that the mounting of TORCH postponed necessarily the assembling of forces in the British Isles. The North African operations must continue to take precedence, against the possibility of adverse situations developing in Spanish Morocco or in Tunisia. The United States, the President added, was much more heavily engaged in the Southwest Pacific than he had anticipated a few months previously; nevertheless, a striking force should be built tip in the United Kingdom as rapidly as possible for immediate action in the event of German collapse. A larger force for, later use should be built tip in the event that Germany remained intact and assumed the defensive. Determination of the strength to be applied to BOLERO in 1944 was a problem, the President observed, requiring "our joint strategic considerations." The Prime Minister accepted the American explanations and wired the President that he was completely reassured.106 The idea of a cross-Channel operation in 1943 thus remained alive for purposes of negotiation arid of the staff planning associated therewith. It was evidently out of the question to plan on undertaking in 1941 the kind of cross-Channel operation the war Department had proposed, and necessary to defer to 1944 the great decisive campaign on the plains of northwestern Europe that the American planners, unlike the British planners, had always believed unavoidable.


Table of Contents
Previous Chapter (13) * Next Chapter (15)


1. See above, pp. 292-93.

2. For estimates, see: (1) memo, OPD for WDCMC, 3 Aug 42, sub: TORCH OPD, Tab 78, ABC 381 (7-25-42), 4-B; (2) msg, Br CsofS to Jt Stf CofS, 4 Aug 42, sub: Comd of TORCH Opns, WDCSA TORCH (SS) (3) memo, G-2 for CofS, 4 Aug 42, sub: Comd of TORCH Opns, WDCSA TORCH (SS); (4) msg, Eisenhower to Marshall, 15 Aug 42, CM-IN 5608; (5) memo, OPD for SW, 10 Nov 42, sub: German Capabilities in Iberian Peninsula, and (6) JCS memo for info . 35, 28 Nov 42, sub: Axis Capabilities in Mediterranean Area, both in ABC 381 Mediterranean Area (11-28-42); and (7) Howe, Operations in Northwest Africa, pp. 68-88 MS.

3. Msg [Br] CsofS to Jt Stf Miss (Washington), 4 Aug 42 (COS (W) 236), WDCSA TORCH, 1.

4. (1) Min, 35th mtg CCS, 6 Aug 42. (2) The comments of Army planners are in memo, OPD for CofS, 6 Aug 42, sub: TORCH, Tab 21, Item 1, Exec 5. This memorandum contains a systematic review of COS (W) 236.

5. Min, 28th mtg CPS, 7 Aug 42. The recommendations of the CPS were circulated as CCS 100, 8 Aug 42, title: Gen Order for Priority of Shipping Mvmts.

6. (1) CCS 100/1, 14 Aug 42, title cited n. 5. (2) Min, 36th mtg CCS, 13 Aug 42.

7. (1) CCS 87/4, 9 Aug 42, title: Shipping Implications of Proposed Air Force Deployment, rpt by CMTC, concurred in by CPS to the CCS. (2) See earlier papers in CCS 87 series, for background. (3) See min cited n. 6 (2), for subsequent approval of the provision of the report.

8. For the diversion of the 33d Pursuit Group from the Middle East to TORCH, see pp. 319-20, below.

9. Min, 28th mtg JCS, 11 Aug 42. It was necessary to decide at once what to do with the Aquitania and the Mauretania, which were en route to the United States. If they were not utilized on the Middle East run, and were rerouted, ships suitable for the unescorted voyage might not readily be found. (CCS 87/3 and CCS 87/4.)

10. Msg, Eisenhower to Marshall, 12 Sep 42, CM-IN 4988. This message contains a report on an important conference with Churchill at Chequers, the official country residence of the British Prime Ministers.

11. The Prime Minister then took up the question with the President. Msg, Prime Minister to President, 14 Sep 42, Tab 50, Item 1, Exec 5.

12. Pers ltr, Eisenhower to Marshall, 21 Sep 42, Paper 59, ABC 381 (7-25-42), 4-B. This letter reports a conference with the Prime Minister.

13. (1) Msgs, Prime Minister to President, 22 Sep 42, Nos. 151 and 154, Paper 57, ABC 381 (7-25-42), 4-B. (2) Pers ltr cited n. 12.

14. (1) Ltr, President to Prime Minister, 25 Sep 42, Item 42, Exec 10. (2) Memo, Deans for Marshall and King, 26 Sep 42, no sub, Item 63a, Exec 10. This memorandum transmitted information from Leahy, on the response of the President, Leahy, and Hopkins to the messages from the Prime Minister. Mention is made of a draft message prepared by the Chiefs of Staff and given to Admiral Leahy. This is probably identical with a text preceding the Prime Minister's messages and Leahy's memorandum. (Papers 56 and 58, ABC 381 (7-25-42), 4-B.)

15. For a description of the so-called Trickle movement and the renewal of convoys on a reduced scale in mid-December 1942, see Morison, Battle of the Atlantic, pp. 365-70.

16. According to the Army strategic planners, an examination of the required Atlantic and Pacific railings of cargo vessels for the latter part of 1942 and early 1943 revealed that Pacific shipping requirements were greater, and that the Navy's demands alone in the Pacific were almost equal to all requirements in the Atlantic. (See The Weekly Strategic Sum of Policy Cam, OPD, 28 Nov 42, Tab Policy Com, 17th mtg, ABC 334.3 Policy Com (1 Aug 42), 3.)

17. It appeared to the Army planners that the South Pacific undertaking complicated further an already muddled situation, since Allied shipping was handled by "four uncorrelated agencies" (the Army, the Navy, the War Shipping Administration, and the British Ministry of War Transport), and that the Navy had been "grabbing shipping" wherever it could be found. (See The Weekly Strategic Resume of Policy Com, OPD, 21 Nov 42, Tab Policy Com, 16th mtg, ABC 334.3 Policy Com (1 Aug 42), 3.)

18. (1) Msg, Marshall to Eisenhower, 4 Sep 42, CM-OUT 1673. (2) Draft msg [President to Prime Minister], 4 Sep 42, Item 9, Exec 1.

19. Samuel Eliot Morison, Operations in North African Waters: October 1942-June 1943 (Boston, Little, Brown and Company, 1947), p. 31.

20. Morrison, Struggle for Guadalcanal, pp. 133, 222. The Wasp was sunk on 15 September 1942 and the Hornet on 27 October 1942.

21. (1) Msg, Eisenhower to OPD, 13 Aug 42, CM-IN 4811 (8/14/42). (2) Msg (originator OPD) Marshall to USFOR, London, 5 Oct 42, CM-OUT 1787.

22. See pp. 336 ff below, for the establishment of the Persian Gulf Service Command (PGSC).

23. OPD draft memo [CofS for President, 2 Oct 42], sub: Support of China, Item 54, Exec 10.

24. Ibid. The proposed reply for Chiang Kai-shek was sent to the President on 9 October 1942] The United States persuaded Chiang to modify the terms of his three demands. (See (1) Romanus and Sunderland, Stilwell's Mission to China, Ch. VII; and (2) Ch. XVII, below.)

25. (1) Particularly useful War Department files on the story of strategic planning in connection with the Aleutians in the fall of 1942 are OPD 381 ADC and WDCSA Alaska (SS). (2) See also Conn, Defense of the Western Hemisphere.

26. In the month of November 1942 the total Allied losses of merchant vessels by submarine action in the Atlantic "for the first and only time surpassed 600,000 tons." (See Morison, Battle of the Atlantic, p. 324, and msg, Prime Minister to President, 2 Dec 42, No. 216, Tab 12/6, Item 4, Exec 5.)

Even then, the long debate on whether the Army or the Navy should control air operations against submarines had not been settled. For a partial explanation of the jurisdictional disputes, see: (1) Craven and Cate, AAF I, pp. 514-53; (2) Wesley F. Craven and James L. Cate, Europe; TORCH to POINTBLANK--August 1942 to December 1943, II, THE ARMY AIR FORCES IN WORLD WAR II (Chicago, The University of Chicago Press, 1949), pp. 377-411 (hereafter cited as Craven and Cate, AAF II); and (3) Morison, Battle of the Atlantic, pp. 237-47.

27. (1) Unused OPD draft ltr [President to Prime Minister], 25 Sep 42, Item 42, Exec 10. This letter was drafted by OPD and revised by Marshall. (2) See min, 40th mtg CCS, 18 Sep 42 for Marshall's views, in latter part of summer of 1942, on the revision of the shipping program.

28. Min, 42d mtg CCS, 2 Oct 42.

29. (1) Min, 48th mtg JCS, 29 Dec 42. (2) William Chaikin and Charles H. Coleman, Shipbuilding Policies of the War Production Board January 1942 to November 1945, Historical Reports on War Administration; WPB Special Study 26, pp. 14-18. (3) Civilian Production Administration, Industrial Mobilization for War, I, Program and Administration (Washington, Government Printing Office, 1947), 535-38, 602-09. (4) Leighton and Coakley, Logistics of Global Warfare, Ch. XVI, p. 29, MS.

30. The Weekly Strategic Resume cited n. 17. For a full treatment of supply factors (including the long holdovers and uneconomical routing of ships and inefficient port operations) limiting the TORCH operation, see: (1) Leighton and Coakley, Logistics of Global Warfare, and (2) Howe, Operations in Northwest Africa.

31. See CinC's Dispatch, N Af Campaign, 1942-43, p. 5, copy in OPD Hist Unit File.

32. This number was carried over from the War Department planners' studies of June and July. See memo, OPD for CGs AGF and SOS, 2 Sep 42, sub: Preparation of Units for Overseas Serv, OPD 370.5 Task Force, 3a.

33. CCS 103/3, 26 Sep 42, title: Outline Plan, Opn TORCH. The plan itself bears the date 20 September 1942.

34. The immediate result of the decision taken on 30 July to withdraw ten small ships for partial conversion to use as combat loaders was to cut by about 20,000 the scheduled August troop shipments to the United Kingdom. (See msg, Marshall to Eisenhower, 31 Jul 42, CM-OUT 9255.)

35. For the effect on the date of TORCH, see above, Ch. XII.

36. For the background of Army-Navy consideration of this subject, see ABC 320.2 Amph Forces (3-13-42), 1.

37. (1) Memo, for rcd, Lt Col Edward B. Gallant, 18 Jul 42, sub: Conf, Amph Trig, Jul 18, 1942, ABC. 320.2 Amph Forces (3-13-42), 1. (2) For the engineer amphibian brigades, see typescript by Herbert Rosenthal, entitled, The Engineer Amphibian Command, in Engr Hist Files, Baltimore.

38. Memo, King for Marshall, 5 Feb 43, sub: Army Engr Amph Boat Crews, with JCS 81/1 in ABC 320.2 Amph Forces (3-13-42), 1.

39. The agreement of 18 July 1942 had been signed by General Handy and Admiral Cooke, with the concurrence of Marshall and King. The Navy's refusal to ratify it formally was based on its unwillingness to be committed on a permanent basis to a program which included provision for training Army engineer amphibian brigades. The 18 July arrangement was circulated on 10 August as JCS 81. (Memo, JPS for JCS, 10 Aug 42, sub: Distribution and Composition of U.S. Amph Forces, ABC 320.2 Amph Forces (3-13-42), 1.) JCS 81 was superseded by JCS 81/1, approved by the JCS and circulated on 5 September 1942.

40. By early August the scale of U.S. assault forces was well enough established to make it relatively certain that the fourth and fifth brigades would not be needed for TORCH. Furthermore, suitable personnel were not, according to Lt. Gen. Lesley J. McNair, available. After consulting with various officers in SOS, G-3, and OPD, Handy of OPD authorized their suspension. (See tel conv transcripts, Handy with McNair, and with Brig Gen Floyd L. Parks, 7 Aug 42, Book 6, Exec 8.) Neither the 2d nor the 3d Engineer Amphibian Brigade was used in North Africa. Together with the 4th Engineer Amphibian Brigade-which was later activated-they were sent to the Southwest Pacific and kept under Army command after the disbanding of the other brigades and the discontinuation of such training for Army troops in early 1943. (See memo of agreement, CofS and COMINCH and CNO, 8 Mar 43, with JCS 81/1 in ABC; 320.2 Amph Forces (3-13 42), 1.)

41. Tel conv, Handy with Parks, cited n. 40.

42. In Operations in North African Waters (pp. 23-31), Morison summarizes the difficulties encountered in joint planning and training for this amphibious operation. He points out that service traditions and organization at that time made the Army and Navy more receptive to a "temporary partnership" than to "organic unity" in the conduct of amphibious operations. (For an account of amphibious training for TORCH, see also Howe, Operations in Northwest Africa.)

43. Ultimately parts of both the 3d and the 9th Infantry Divisions participated in the initial landings.

44. This arrangement was in accord with the provision of the 18 July agreement for an amphibious corps of two or more divisions.

45. (1) OPD draft ltr, DCofS to COMINCH and CNO, 6 Aug 42, sub: Employment of Amph Force, U. S. Atlantic Fleet, atchd to memo, OPD for DCofS, 4 Aug 42, sub: Amph Forces, OPD 353 Amph Forces, 29. This was sent out to the Navy on 6 August 1942. (2) See also tel conv, Handy with Parks, cited n. 40.

46. (1) Msg, Eisenhower to Marshall, 17 Aug 42, CM-IN 6236. (2) Memo, CofS for President, 20 Aug 42, sub: TORCH Opn, WDCSA TORCH, 1.

At the same time the shortage of combat-loading vessels available in the United Kingdom for amphibious training reduced from two to one the number of requested U.S. combat teams that could be trained there at the time. Amphibious training in the United Kingdom, therefore, began with one regimental combat team of the 34th Division in stead of two from the 1st Division, as originally contemplated.

47. Msg, Marshall to Eisenhower, 15 Aug 42, CM-OUT 4636.

48. Msg, Eisenhower to Marshall, 19 Aug 42, CM-IN 6998.

49. AFHQ, G-3 Sec, Outline Plan C (Provisional) Opn TORCH, 5 Sep 42, ABC 381 (7-25 42), 4-A.

50. (1) Ltr, Patton to OPD, 3 Aug 42, sub: Notes on Western Task Force--TORCH Opn, Book 6, Exec 8. (2) Msg, Eisenhower to Marshall, 8 Aug 42, CM-IN 2766.

51. The 3d Armored Division was held in the United States until August 1943 and then sent to England.

52. Memo, CofS for OPD, 25 Oct 42, no sub, OPD 370.5 Task Force, 101.

53. Memo, OPD for CofS, 29 Oct 42, sub: Shifting of Divs Prior to Overseas Mvmt, OPD 370.5 Task Force, 101.

54. Memo, McNair, AGF, for ACofS OPD (through G-3, WD), 13 Nov 42, sub: Excessive Tr Mvmts, OPD 370.5 Task Force, 101. Army authorities did their best to cut down such transcontinental movements and the consequent taxing of transportation facilities in the zone of interior, but, in the late fall, one division was needed for the British Isles and one for the South Pacific. The 29th and 43d were selected as best available and were again shifted within the zone of interior.

55. (1) Msg, Marshall to Eisenhower, 30 Oct 42, CM-OUT 10217. (2) See also Robert R. Palmer, Bell I. Wiley, and William R. Keast, The Procurement and Training of Ground Combat Troops, UNITED STATES ARMY IN WORLD WAR II (Washington, Government Printing Office, 1948), pp. 178-79.

56. A good example of the difficulties and confusion faced by SOS in the zone of interior in readying its units for overseas service for TORCH, as a result of lack of sufficient time, was the dispatch of the 829th Signal Service Battalion. (See Cline, Washington Command Post, pp. 435-42 [sic].)

57. Memo, Hq ETO for TAG and incl, 14 Dec 42, sub: Tr List, U.S. Trs Transferred to N Af from UK, Tab 58, Item 2, Exec 5.

58. Memo, OPD for CofS, 2 Aug 42, sub: TORCH, Paper 45, ABC 381 (7-25-42), 4-B.

59. The extent to which TORCH was absorbing divisional equipment was indicated in an estimate in the late summer of 1942 of divisions that might be available for shipment to the South Pacific. The Army planners estimated that three divisions (2d, 29th, and 38th) besides those allocated to TORCH could be equipped by 1 October 1942--except for ammunition--by stripping most of the reserve and training divisions of their equipment. Preparations for TORCH had meanwhile made themselves felt in another way; all three divisions were understrength. (See memo, Logistics Gp, OPD for S&P Gp, OED, 13 Sep 42, sub: Availability of Units for a Sp Opn, with JCS 97/1 in ABC 381 (7-25-42), 1.)

60. Msg cited n. 55 (1).

61. For a full account of the complicated transactions over the utilization in TORCH of supplies accumulated in the United Kingdom and over the future level of supply far units remaining in the United Kingdom or sent to the United Kingdom, see Leighton and Coakley, Logistics of Global Warfare.

62. See, for an early example, memo, OPD for DCofS, 2 Aug 42, sub: Air Force Deployment, Item 7, Exec 1.

63. (1) Ibid. (2) Memo, Theater Gp, OPD [Streett] for Handy, 15 Aug 42, sub: Unit Designations for BOLERO and Air Force Deployment, Item 7, Exec 1. (3) Statements of McNarney, DCofS, in min, 28th mtg JCS, 11 Aug 42.

64. As Army planners were to point out, there were barely enough combat planes assigned to defense and training. See, for example, OPD draft memo [CofS for President], 26 Oct 42, sub: Sit in S Pacific, OPD 381 PTO, 107.

65. (1) Msg, Eisenhower to OED, 13 Aug 42, CM-IN 4811 (8/14/42), transmitting the plans for TORCH air force as drawn up in London by Patton, Spaatz, and Doolittle. (2) See also memo cited n. 63(2). Various proposals of Army planners drafted in the summer of 1942 for air deployment for TORCH are in Item 7, Exec 1.

66. See Craven and Cate, AAF II, p. 51.

67. See n. 83, below.

68. See Craven and Cate, AAF II, pp. 24-25, 52.

69. Ibid., pp. 51-52, 59-60. This contains a brief description of the difficulties encountered in equipping, training, and readying far combat the medium and light bombers for TORCH. The original August plan provided for bomber groups-three medium and one light-to fly to England and there be indoctrinated, processed, and initiated into combat. Eventually, as a result of the experience with bad weather in the North Atlantic, the route was closed to twin-engine aircraft.

70. Memo, DCofS (for CofS) for COMINCH, 5 Sep 42, sub: Air Reinforcements for Guadalcanal Tulagi Area, Book 6, Exec 8.

71. (1) Ltr, Patton to CG AAF (through OPD), 26 Aug 42, sub: Air Support TORCH, Item 7, Exec 1. (2) Msg, Marshall to Eisenhower, 31 Aug 42, CM-OUT 0233. (3) Ltr, Patton to CG AAF (through OPD), 9 Sep 42, sub: Air Support TORCH, Incl B to JCS 97/1.

72. (1) Msg, Eisenhower to Marshall, 12 Sep 42, CM-IN 4988. (2) Msg, Marshall to Eisenhower, 12 Sep 42, CM-OUT 4316.

73. Ltr, Streett to JCS, 8 Sep 42, sub: Immediate Allocation of 33d Pursuit Gp, circulated as JCS 101.

74. See Craven and Cate, AAF II, pp. 25, 63.

75. Min, 32d mtg JCS, 8 Sep 42.

76. (1) Ltr, CG AAF to Evill, 9 Sep 42. (2) Ltr, Air Commodore S. C. Strafford (for Evill) to Arnold, 10 Sep 42. (3) Ltr, Arnold to Evill, 12 Sep 42. These ltrs are incls to CCS 112, 17 Sep 42, in ABC 381 (7-25-42), 1.

77. Min, 40th mtg CCS, 18 Sep 42.

78. (1) Msg, Eisenhower to Marshall, 19 Sep 42 (No. 2396), paraphrase atchd to CCS 112/1, 21 Sep 42, title: Immediate Allocation of 33d Pursuit Gp. (2) Msg, same to same, 13 Sep 42, CM-IN 5411. The War Department proceeded to comb U.S. air resources for replacements for the 33d Group for the Middle East. (See Craven and Cate, AAF II, p. 25, and min, 34th mtg CCS, 22 Sep 42.)

79. See above, Ch. XIII.

80. (1) OPD brief, Notes on . . . JPS 32d mtg, 2 Sep 42, with JPS 48 in ABC 381 (9-25-41), 3. (2) JPS 48, 28 Aug 42, title: Detailed Deployment of U.S. Air Forces in Pacific Theater. This contained views of Army and Navy planners on the JUSSC.

81. See Craven and Cate, AAF II, pp. 274-77.

82. JCS 97/1, 11 Sep 42, title: Relationship Between TORCH and Air Operations from the Middle East and the United Kingdom. This paper is a memo of CG AAF to JCS, with four inclosures. Arnold indicated that, at his suggestion, Eisenhower had been urged not to discontinue U.S. air operations from the United Kingdom on 8 September, but to maintain them until the last possible moment before TORCH, and thereby give no respite to the German Air Force.

83. JCS 97/2, 15 Sep 42, title: Detailed Deployment of U.S. Air Forces in the Pacific Theater (Need for Army Aircraft in Current Solomons Opn). For King's views in August, see above, Ch. XIII.

84. JCS 97/2, 15 Sep 42.

85. (1) Min cited n. 75. (2) Min, 33d mtg JCS, 15 Sep 42. (3) Min, 36th mtg JCS, 6 Oct 42.

86. For Marshall's earlier explanation to Eisenhower at the end of July, see above, Ch. XIII.

87. (1) Min 38th nag JCS, 20 act 42. (2) Min, 39th mtg JCS, 27 Oct 42. By 26 October, of the fifteen groups three had been comitted; one heavy bomber group to Hawaii; one half of another heavy bomber group to the South Pacific; and one and a half troop carrier groups divided among the Middle East, Alaska, and the South Pacific. Twelve groups were still uncommitted. (See JCS brief, 26 Oct 42, title: JCS 97/5 Deployment of U.S. Air forces in Pacific Theater, with JCS 97/5 in ABC 381 (9-25-41), 3.)

88. Msg (originator OPD), Marshall to Eisenhower, 6 Aug 42, CM-OUT 2023.

89. "American-British Grand Strategy," 31 Dec 41, ABC-4/CS 1, in ARCADIA Proceedings . . . See above, p. 99, n. 6, and p. 295.

90. Memo, Lt Col E.H. Qualls, OPD, for Gen Hull, 7 Sep 42, sub: BOLERO Com Mtg, File 3, ABC 381 BOLERO (3-16-42), 4. Colonel Qualls was a member of the committee for BOLERO movements in OPD's European Theater Section, then headed by General Hull.

91. Memo, OPD for CofS, 27 Aug 42, sub: 5440 Shipts of Tr Units, Paper 39, ABC 381 (7-25-42), 4-B.

92. Memo cited n. 90.

93. Msg (originator OPD), Marshall to USFOR, London, 8 Nov 42, CM-OUT 2704. In view of the tight shipping situation no precise date was set for the attainment of the 427,000 figure.

94. Ltr, Hartle to CofS, 12 Nov 42, OPD 381 ETO, 48.

95. OPD draft memo [CofS for President], 17 Oct 42, OPD 381 ETO, 35.

96. Min, mtg at White House, 1430, 10 Dec 42, Tab 42, Item 2, Exec 5. Present at this meeting with the President were Hopkins, Leahy, Marshall, Arnold, Deane, and Vice Adm. Richard S. Edwards.

97. (1) Summaries of Army planners' strategic views in the fall of 1942 appear in unused WD draft ltr, President to Prime Minister, 25 Sep 42, Item 42, Exec 10. This letter was drafted by OPD and revised by Marshall. (2) Memo, OPD for CofS, 8 Nov 42, sub: Amer-Br Strategy, with JCS 152 in ABC 381 (9-25-41), 3.

98. Pers ltr, Eisenhower to Marshall, 21 Sep 42, Paper 59, ABC 381 (7-25-42), 4-B. Eisenhower, Clark, and Smith attended this staff conference.

99. Ibid.

100. Msg, Prime Minister to President, 22 Sep 42, No. 151, Item 63, Exec 10.

101. Msg, Smith to Marshall, 26 Nov 42, CM-IN 11164. A copy of this message was sent to the President.

102. Msg, Prime Minister to President [25 Nov 42], No. 211, WDCSA 381, 1 (SS). This message was shown to Marshall and Handy in the War Department.

103. Msg cited n. 101.

104. Churchill, in Hinge of Fate (pp. 648-59) gives an exposition of his views in late 1942 on a 1943 ROUNDUP.

105. Sherwood, Roosevelt and Hopkins, pp. 657-58. Preliminary- drafts of the President's reply to the Prince Minister, drawn tip by Marshall and his staff' and bearing the date 25 November 1942, are contained in War Department files. For reply of President to Prime Minister's msg No. 211, see WD drafts of msg (25 Nov 42), Item 63a, Exec 10. Other copies are filed in WDCSA 381, 1 (SS).

106. Msg cited n. 101.

Transcribed and formatted for HTML by Patrick Clancey, HyperWar Foundation