Title Banner: The Army Air Forces in World War II

Chapter 3

THE MATTERHORN plan reflected the predominant interest in strategic bombardment that existed in AAF Headquarters. Essentially it was an effort to introduce into the Japanese war the objectives and techniques of the Combined Bomber Offensive in Europe: so to batter the industrial fabric of an enemy nation by long-range bombardment that armed resistance would be enfeebled. The circumstances under which the new campaign would be conducted, however, contrasted sharply with those in Europe. In the ETO the heavy bombers of the Eighth and Fifteenth Air Forces were opposed by a determined and relentless enemy, but they operated under material conditions that were, for wartime, quite favorable. In the United Kingdom the Eighth's bases had been built by the British--of materials, by methods, and to standards comparable to those of the AAF. Supply and maintenance depots were large and lavishly equipped. Supply routes, both within the British Isles and from the United States, were as highly developed as any in the world. True, submarines menaced the sea lanes, port facilities still bore the marks of the Luftwaffe's blitz, railways were choked with munitions, materials and civilian labor were short; but the communications network was a going concern, and the CBO enjoyed a high priority in most logistical matters. There was no serious shortage of fuel, few long-term shortages on any items in production, and bases were by field standards luxurious. In Italy base development did not begin until autumn 1943, but the use of Italian Air Force facilities sped the task. And in Italy, as in England, Army air forces enjoyed the inestimable advantages of working in an industrialized community. In the CBI most of these advantages were lacking. Bases had to be


made to order for the B-29's. Ideally, they should have been built from locally available materials and by native manpower, but in India both U.S. Army engineers and U.S. materials had to be used to supplement local resources. Supply and maintenance installations were understocked and overworked; there was little industrial organization upon which the AAF could draw. Supply lines to the United States and United Kingdom were excessively long, the CBI's shipping priorities low, and supply lines within the theater were unequal to current demands and incapable of rapid expansion. All in all, it did not seem the ideal theater, logistically, in which to shake down an untried, complex, and gluttonous bomber. Those difficulties were realized by the Washington planners (though not as keenly as by officers in the theater) and had deliberately been accepted for want of a better base area within reach of the Japanese homeland and because of Roosevelt's desire to bolster Chinese morale. The President's concern placed an emphasis on speed, but as delays, many unavoidable or unpredictable, pyramided, the time schedule formulated in the autumn of 1943 was not even approximated. By June 1944 the bases and essential installations were in operation, the supply system was functioning in its own complicated way, and XX Bomber Command was ready for its first mission. But this belated readiness had been accomplished only by scrapping some of the essential features of the MATTERHORN logistics plan, and it was already evident that further compromises would be necessary to support a sustained bombardment campaign.

The Bases

Theater officers had begun in August 1943 an on-the-spot investigation of potential base sites in India and China. Their tentative choices and their estimates of CBI capabilities in airfield construction served as practical guides for the Washington planners. Basic assumptions were: 1) that airfields could be built in China without recourse to American aid other than financial support and technical advice; 2) that in India it would be profitable to bring up to B-29 specifications airdromes already in existence or being built; and 3) that the India bases could be built on schedule only by importing certain materials and using U.S. Army construction units with their organizational equipment, as well as local labor. Under these conditions, it would require one U.S. aviation engineer battalion four months to


complete each India base; the Chinese could build two fields in two months, four in four months, five in six months.1 Because of Roosevelt's desire for an early D-day, preliminary arrangements were initiated and completed before the CCS had given their belated sanction to the VLR project in the final report at the Cairo conference. On 10 November, one day after he had informally approved MATTERHORN, the President informed Churchill and Chiang Kai-shek of the plan and asked for aid in securing the necessary airfields. Both leaders responded promptly and favorably.2

Responsibility for construction fell to General Stilwell, as U.S. commander in the CBI, and under him, to his ranking Services of Supply officer, Maj. Gen. W. E. R. Covell. To supervise the task in both theater sectors, the Air Engineer, Brig. Gen. S. C. Godfrey, was sent out from the States. Actual construction work was directed in India by Col. L. E. Seeman, in China by Lt. Col. W. I. Kennerson.3 It was characteristic of CBI operations in general that in spite of the unified command provided by Covell's office, the two base areas were developed separately and by methods which differed sharply.Southern Bengal had been chosen as the rear base area for reasons acceptable to all: its position vis-à-vis China, relative security from attack, the port facilities of Calcutta, and rail and road communications that were good by Indian standards. In the territory surrounding Midnapore, some seventy miles west of Calcutta and on the edge of the rolling alluvial plain of the Ganges, Eastern Air Command had laid out twenty-seven airdromes and twenty-three satellite strips, each designed to accommodate two squadrons of B-24's; by extending and strengthening the 6,000-foot runways of some of these fields, CBI engineers hoped to make them serviceable for B-29's.4 A TWILIGHT Committee headed by Brig. Gen. Robert C. Oliver of ASC made a preliminary survey of the airdromes and on 17 November tentatively designated for early development as B-29 fields the following: Bishnupur, Piardoba, Kharagpur, Kalaikunda, and Chakulia. This choice was approved by an advance party of XX Bomber Command staff, except that Dudhkundi was substituted for Bishnupur.5 General Wolfe inspected the sites in mid-December and picked Kharagpur as his headquarters. Some sixty-five miles out of Calcutta on the main-line Bengal-Nagpur railway, Kharagpur was an important junction point, with a branch line that served most of the other proposed air base sites. The deciding factor was the existence at the adjacent village


Map: Kharagpur Area Airfields
Kharagpur Area Airfields


of Hijili of a large new building (the Collectorate, designed as a political prison) which would house the bomber command's headquarters.6

Original plans, with an assumed deployment of two combat wings, had called for eight fields housing one B-29 group each and a ninth for transport planes. In January 1944 it was decided instead to build four fields with two-group capacity (fifty-six hardstands) at least as a temporary measure.7 The decision in April to divert the second B-29 wing to the Marianas obviated the necessity of completing the additional fields. Meanwhile, delays in the building program had made it necessary to utilize temporarily one other B-24 airfield, Charra, where the existing runway was extended by two 900-foot steel mats. The permanent fields developed were Kharagpur, Chakulia, Piardoba, Dudhkundi, and Kalaikunda--the last as a transport base.8

General Godfrey in early November had set the requirements for U.S. construction units which had been written into the MATTERHORN plan: one aviation engineer regiment (less three battalions) for administration, four regular and one airborne aviation engineer battalions, four dump-truck companies, and two petroleum distribution companies.9 To meet the 1 April target date which AAF planners had set in answer to the President's urgency, those units should have been in place by the beginning of December. Since they had to go out from the United States, that was obviously impossible; even to have the fields operational by 1 May, the date finally accepted at SEXTANT, would require rapid action and some good fortune. On 13 November General Arnold recommended that the War Department divert certain designated construction units from previous assignments and ship them out on the 15 December convoy.10 The JCS, en route to Cairo, approved Arnold's requisition for the units on 17 November, but added limitations which would scale down by about half the troops to be dispatched on 15 December.11 On request, Stilwell reluctantly granted the necessary shipping priorities; the first increment of troops sailed on schedule, transshipped in North Africa in early January, and arrived in India in mid-February.12 This was two months later than the original ideal estimate and a month later than had been hoped in November, and it was but half the required force.Late in November responsible officers had begun preliminary work on the Bengal fields, with AAF casuals driving some 500 trucks borrowed


from the China Defense Service and the Ledo Road.13 General Stratemeyer proposed that two engineer battalions be borrowed from the Ledo Road until the expected units arrived from the States; about Christmas, however, he learned that these units would not arrive until February.14 At SEXTANT, the provision that the B-29 project be conducted "without materially affecting other approved operations" had been interpreted to allow the temporary diversion of certain "resources" from the Ledo Road. These included the trucks but not engineer units. Stilwell, committed to the road-building both by his interpretation of his directive and by conviction, refused Straterneyer's request but was willing to refer it to Washington.15 General Marshall backed Stilwell's view, but when apprised of the CBI's pessimistic estimate of the construction schedule, was willing to indorse the theater's suggestion (acceptable to Stilwell and Mountbatten) that engineers assigned to July amphibious operations in SEAC be loaned to MATTERHORN. Marshall accordingly assigned to the latter on 13 January the 1888th Engineer Aviation Battalion, on orders to sail from the west coast early in February and due in India in April.16

This move offered no early relief; the JCS on 15 January had to inform their British counterparts of the lag in the schedule, and the AAF considered postponing the target date for the operation to 30 June.17 In this emergency Stilwell reversed his earlier stand. On 16 January he consented to lend from the Ledo Road the 382d Engineer Construction Battalion (Separate), and the unit was moved by air to Kharagpur. Further, when the 853d Engineer Aviation Battalion arrived in India on 1 February, it also was reassigned to the B-29 project and sent to Chakulia. With this reinforcement, the project officers could hope to have two fields barely operational by 15 March; by using two auxiliary fields temporarily, they could accommodate the B-29's as they arrived.18 When the units from the 15 December convoy came in during mid-February, they were assigned to the several fields: the skeletonized 930th Engineer Regiment to Kalaikunda; 1875th and 1877th Engineer Aviation Battalions to Dudhkundi and Chakulia, respectively; 879th Engineer Battalion (Airborne) to Piardoba.19 That last unit, with its light equipment, was unsuited for heavy concrete work and was later reassigned, as were the two units on loan from Stilwell.20

In all, construction forces numbered some 6,000 U.S. troops and 27,000 Indian civilians,21 the latter working under India's Central Public


Works Department by contract. A complicated system of requisitioning and the traditional slowness of native methods required much "expediting" by the Americans, and until the U.S. engineers arrived, Colonel Seeman was essentially a liaison officer with the Anglo-Indian organization. There was some overlapping of tasks, but in general the natives did those jobs which could be accomplished by hand labor, the U.S. troops those requiring skilled labor and heavy machinery.

The first large task completed was installation of the pipeline system. This called for a six-inch line from Budge-Budge on the Hooghly River to Dudhkundi, with four-inch pipes to the four other fields and internal lines and steel tank storage for each. Light-weight "invasion" type pipe was used, but it was buried to avoid injury from accident or native curiosity. Four petroleum distribution companies did the work-the 700th, 707th, 708th, and 709th. Beginning the job on 15 January, those companies by 15 March had fuel flowing to the three fields then approximately ready to receive B-29's, and later they completed the whole circuit.22

Runway construction was a more considerable task. Grading for the strips accounted for more than half of the total of 1,700,000 cubic yards of earth moved on the project. In spite of urgent requests from the CBI, most units arrived without the heavy machinery needed for earth moving; some machinery was borrowed from the British and kept in service even after the unit equipment came.23 Specifications called for extending the B-29 strips to 7,500 feet instead of the 8,500 feet designated by Washington.24 New concrete pavement was ten inches thick, and old pavement had an additional seven inches poured on. Both chevron- and horseshoe-type hardstands were used, and eventually rectangular parking areas were paved. The British system of dispersal was abandoned in favor of a more concentrated layout.25

Ideally, the fields should have been built of local materials. Sand was available in streams near each field and coarse aggregate (gravel and crushed basalt) was found in the immediate neighborhood. Indian cement, however, was both scarce and inferior, and much imported U.S. cement had to be used. Concrete was produced locally by means varying in efficiency according to equipment on hand. On all the fields save Kalaikunda, which was paved in July after all heavy machinery arrived, concrete was spread by hand by native workers.26 Buildings on the several bases showed no little variety. The Collectorate, prize structure of the rear area, required extensive modification.


Troops were under tents at first, but eventually were housed in hutments of native "basha" construction--hard earth or concrete floors, bamboo and plaster walls, thatch roofs. Administrative and technical buildings included basha, U.S. plywood prefabs, Nissens borrowed from the British, and some ex-Italian prefabs imported from Eritrea, bullet-marked and somewhat shopworn. MacComber shops with overhead traveling cranes and Butler hangars with steel frames and canvas covers proved useful but difficult to erect because of damage and loss of steel parts. Most of the utilities--water and electric systems--were installed by U.S. engineers.27 Fortnightly reports to Washington after February were apt to read "work progressing on schedule," a schedule, of course, far in arrears of early plans. Actually, the fields were not completed until September.28 But, by using the B-24 field at Charra (until July), General Wolfe found it possible to receive and house his four combat groups as they flew in with their Superforts in April and May. The cost of the five bases is difficult to determine because of the several agencies involved; Colonel Seeman considered $20,000,000 an approximate estimate.29

That figure was modest in comparison with the cost of the fields in China, where indeed, finances proved the chief headache for the Americans. The advanced B-29 bases were situated in the neighborhood of Chengtu. Chennault had preferred Kweilin which was closer to industrial Japan, but Stilwell had estimated that fifty Chinese divisions would be needed for ground defense and Washington had named Chengtu because of its greater security.* Chengtu, capital of the province of Szechwan, was located about 200 miles northwest of Chungking and 400 miles from the Hump terminal at Kunming. An ancient city, a seat of commerce and of culture, Chengtu lay in the valley of the Min River. About 2,200 years ago a semi-mythological engineer, one Li Ping, had harnessed the river as it burst from the mountains northwest of the city and had diverted it into several large canals and a myriad of smaller ones. His ingenious irrigation system, still operated with due respects to beneficent deities, had made at the Min valley a sort of artificial delta of extraordinary fertility. The delta or plain, no more than 70 miles long and some 1,700 square miles in extent, supported a population of about 2,200 persons to the square mile. In many respects Chengtu was admirably suited for a base area.

* See above, p. 21.


There were rugged mountains to the north and west, but the immediate vicinity was level enough, and weather was reasonably good for flying. But the fertility of the valley and its teeming population meant that airfields could be built only at the expense of some economic and social dislocation, and there were serious political implications as well.30

After Chiang Kai-shek had agreed to Roosevelt's proposal to build B-29 bases in China, engineers of General Oliver's TWILIGHT Committee surveyed the region and by 28 November had tentatively selected sites for five VHB fields.31 These the Generalissimo approved provisionally on 16 December; he also approved, in principle, other sites which would lessen Chennault's objections to Chengtu: Niuchang, near Kunming, as a ferrying base and Kweilin and Suichwan in the east as staging fields. Within a fortnight the list for Chengtu had been modified somewhat to include Hsinching, Kiunglai, Kwanghan, Pengshan, Chungchingchow. Except for Kwanghan these sites had strips already. Availability of materials and labor and the relative amount of interference with the irrigation system were deciding factors.32 In January XX Bomber Command staff officers, then Wolfe himself, approved the sites; later Chungchingchow was stricken from the list.33

Chennault, responsible for air defense, located the fighter fields at Fenghuangshan, Shwangliu, Pengchiachiang, and Kwanghan (at the bomber base) in the immediate neighborhood, and an outer arc of strips at Mienyang, Kienyang, and Suining.34 Chennault pressed for the staging fields in the east in a letter written directly to Arnold who referred him back to Stilwell.35

At Stilwell's request, Chiang finally consented to improve a number of B-24 fields for Superfort use: Chengkung and Luliang near Kunming; Kweilin, Li-Chia-Tsun, and Liuchow in Kwangsi province; Hsincheng and Suichwan in Kiangsi.36 These plans were later interrupted by changes in the tactical situation; by November, only Luliang (usable) and Hsincheng (under construction) were still on the active list.37 VHB operations were conducted, as had been planned in MATTERHORN, from Chengtu.

Chennault did not have enough engineers to furnish the supervision called for in the agreement with Chiang Kai-shek, but the AAF furnished the needed personnel on requisition-a few specialists who flew out with General Godfrey early in December 2nd a larger party which arrived at the end of the month. Over-all supervision fell to


Chennault's chief engineer, Col. H. A. Byroade. One of Godfrey's party, Lt. Col. W. I. Kennerson, was in charge of U.S. engineers at Chengtu.38 The Americans did the planning and supervision while Chinese engineers directed actual construction. Airfield construction in China was a responsibility of the Minister of Communications, American-educated Dr. Tseng Yang-fu, who delegated most of his task to the Deputy Commissioner of Engineering and to the Chief Engineer.

Map: Chengtu Area Airfields
Chengtu Area Airfields

Chengtu office of the Chinese Engineering Office handled administrative and financial matters. Dr. Tseng Yang-fu selected the executive engineers, one for the whole project and one for each field. They came up from Kunming early in January, each bringing his own staff, some 300 in all. Few of them were experienced in airfield construction, but after briefing by Colonel Kennerson they were ready to take over.39

The labor problem was handled forthrightly and with little concern for the laborer. China's greatest source of strength lay in her inexhaustible reservoir of manpower--unskilled by western standards and wholly devoid of modern machinery but patient and sturdy and bound by a social organization that could be transferred directly to


the new task. This great reservoir the Chungking government proposed to tap by the custom-hallowed process of conscripting farmers from the Min valley for the heavy construction work; housing was to be erected by more skilled contract labor. The project was to challenge credulity by the magnitude of the force involved. Western witnesses sought analogies in the building of the Great Wall of China or in Herodotus' account of the building of the great pyramid of Cheops. Tools and methods employed at Chengtu were not dissimilar from those used in the ancient works, but the time schedule was characteristically American.40

In early January the Chinese directors and Colonel Kennerson estimated the labor force required, setting the figure at 240,000.41 The Governor of Szechwan drafted the men for 11 January; two weeks later something like 200,000 had appeared and work had begun on most fields.42 In mid-February the governor agreed to draft 60,000 more men in an effort to catch up with the schedule, and in March 30,000 more for the fighter fields. To these 330,000 conscripts must be added 75,000 contract workers. Reports from various U.S. sources vary as to the total number of men who actually appeared, and it is doubtful that Chinese statistics were meticulously accurate, but with the inevitable turNover there may have been well more than a third of a million men on the job.43 The story of Chengtu, wrote a correspondent, was "a saga of the nameless little people of China," for the fields were built by the "hand, muscle and goodwill . . . of 300,000 to 500,000 farmers."44 They came from villages within a radius of 150 miles from Chengtu on the basis of 50 workers from each 100 house-holds. On the job the coolies were organized into units of 200, still preserving something of the village structure; local officials kept the payrolls.45

An enterprise of such magnitude could hardly fail to. effect a sharp economic and social reaction. Chengtu's geographical remoteness from the war was favorable, but there was as well psychological and political remoteness. Szechwan has been compared, whether accurately or not, with our own pre-Pearl Harbor midwest. Seemingly immune to Japanese attack, the province was isolationist, apathetic toward the war, and potentially anti-foreign. Powerful local warlords looked on the MATTERHORN project as a scheme whereby the Chungking government could encroach upon their quasi-autonomy. Men of property feared, needlessly, that their lands would be confiscated


without recompense and, with more justification, that the project would aggravate the current inflation. Men of whatever class feared that the establishment of the airfields would bring Japanese bombers to Chengtu and that U.S. soldiers would be disorderly. The conscripts also knew that they were being torn away from home in the New Year holiday season and feared that they would not get back in time for the rice planting. Only the Chungking government, the politicians, and the contractors could hope to profit from the project.46

To most American officials, the attitude of the Chungking government did not seem too generous. The President had assured the Generalissimo on 10 November of American financial aid through lend-lease, but in the early negotiations no specific terms were suggested. In mid-December Chiang Kai-shek set the total cost of the fields at "over $2,000,000,000"Chinese National (CN) currency and asked Roosevelt for a guarantee of that amount.47 This guarantee the President was willing to make, but his administration was interested in the rate of exchange. The current open (black market) rate was about $100 CN for $1 U.S. The Chinese government, as an anti-inflationary device, had set the rate arbitrarily at twenty to one. At the open rate, the cost of the fields would have been high, but "not unreasonable"; at the official rate, the cost would have been exorbitant.48 Negotiation dragged on for several months. The U.S. Treasury and State Departments, interested in the broader problem of U.S.-Chinese financial relations, wished to adhere to the open rate. The War Department, though anxious to secure the fields at a reasonable cost, felt the pressure of time more keenly and was willing to compromise by accepting the twenty-to-one ratio but requiring the Chungking government to deposit $80 CN for each $20 CN advanced by the United States.49 Chiang Kai-shek refused to compromise; holding fast to his demands, he began to point out that failure to agree to terms would delay construction.50 To keep the project going, Stilwell had to guarantee payment of the sum demanded at a rate which was to be decided by subsequent negotiations.51

Negotiations were complicated by a number of factors. Funds in China were frozen, making it difficult to meet obligations at Chengtu. The Chinese Minister of Finance, Dr. H. H. Kung, insisted that there was an actual shortage of CN notes, and although U.S. officials thought there was a reserve of $10,000,000,000, it became necessary for ATC to fly in from India a supply of notes for immediate needs


in Chengtu.52 Two hundred million dollars in small bills bulks up; hauling Chinese money became, as one observer remarked, "definitely a factor in the tonnage operation over the hump."53 There was too the matter of financing the extra fighter defense fields chosen by Chennault and the proposed B-29 fields in the east and near Kunming. The War Department was willing to pay for the former out of MATTERHORN funds, not for the latter.54

Finally, there was the matter of Chiang's demand at Cairo for a loan of one billion dollars CN. Stilwell ascribed the request to a desire for prestige and the Generalissimo's postwar plans rather than current needs, but refusal to grant it complicated the MATTERHORN deal.55 Negotiations continued through the winter months and into the spring. The Chengtu project was kept going by occasional advances of currency without agreement as to ratio, but at times construction was handicapped by lack of ready funds.56 By early March, estimates for the Chengtu fields (bomber and fighter) had risen to $4,450,000,000 CN, and the final figures were not far from this sum.57 Final settlement was not reached until after conferences in July between Dr. Kung and Secretary of the Treasury Henry J. Morgenthau, Jr. A lump sum of $210,000,000 U.S. was finally paid to China, but this included other items as well as the Chengtu fields and an accurate breakdown is impossible to achieve.58

In spite of the tremendous cost, many Chinese suffered. Landowners did receive compensation for their fields, but not promptly and not at a favorable price. Inflation was increased by the project and with the depreciation of currency those who had to sell land at government prices lost.59 The Governor of Szechwan set ceiling prices on materials used by the contract builders (some $400,000,000 CN were involved), but with only partial success.60 The conscript workers suffered most-from the squeeze and from low pay. Paid on a piecework basis, they averaged perhaps about $25 CN per day, which with rising prices (by September the black market rate had risen to $270 CN for $1 U.S.) hardly sufficed for food, so that many workers had to be partly supported by their families.61 Even with these difficulties the disorders feared by some did not occur. There was much grumbling and a few small riots, occasioned when overeager U.S. engineers moved in before the land had actually been purchased. But there was no general resistance on the part of the Szechwan citizens, and eventually they came to take some proprietary pride in the B-29 project.62

The four bomber strips were built to a length of about 8,500 feet


and a thickness of about 19 inches, with 52 hardstands each. The fighter strips were approximately 4,000 by 150 feet, with a thickness varying from 8 to 12 inches and having 4 to 8 hardstands. The base course consisted of rounded rocks from streams set with gravel and sand, wet and rolled. The wearing course was a sort of native concrete called "slurry," a mixture of crushed rock, sand, clay, and water. Rolled and finished, this gave a texture and tensile strength not unlike adobe. The fields were literally "handmade." Materials were carried from nearby streams in buckets or baskets slung from yokes, in squeaky wooden-wheeled wheelbarrows, or infrequently in carts. Excavation was by hoes. Crushed rock was patiently beat out with little hammers and stones were laid individually by hand. Rollers were drawn by man (and woman) power, the slurry puddled in pits by barefoot men and boys.63

Work began on 24 January, when the first rice paddies were drained. At that time it was thought that two fields would be operationally complete by 31 March, two others by 30 April,64 but the financial disputes and other difficulties retarded that schedule. On 24 April, three months to the day after the first paddy wall was breached, General Laverne G. Saunders set down the first B-29 at Kwanghan.65 By 1 May all four VHB fields were open to B-29 traffic,66 and by 10 May all runways were finished and some fields were operationally complete.67 The fighter fields had been finished somewhat more nearly on schedule.68 In spite of the delays, the whole job excited the wonder and admiration of most Americans who saw it in process. And, in a fashion not always true in war, it was the man at the bottom who got most of the headlines, the man with the hoe and the complaining wheelbarrow. The historian of XX Bomber Command wrote: "The Chinese coolies--the John Q. Public of the Chengtu Plain--demonstrated effectively the best features of their nation.69

There was a third base area from which the B-29's were to operate, but there XX Bomber Command had no permanent installations. As a compromise with those strategists who had wanted to base the Superforts in Australia and bomb petroleum targets in the Netherlands East Indies, MATTERHORN planners had suggested that VLR missions could be staged out of Bengal against Palembang, Sumatra's chief oil center, by refueling in Ceylon. This suggestion was accepted at SEXTANT and the target date set at 20 July 1944.70 At the conference the British reported on the airfield situation in Ceylon. Conventional


fields existed at Sigiriya, China Bay (Trincomalee), and Ratmalana; two others (at undesignated locations) under construction might be extended to B-29 specifications.71 Only part of the large island of Ceylon

Map: Ceylon Airfields
Ceylon Airfields

lay within B-29 range of Palembang, and none of the fields named had been situated with that target in mind. Poor internal communications would make it difficult to build fields in the southeast, the area


Image: Building the Chengtu Airfields Image: Building the Chengtu Airfields
Building the Chengtu Airfields
Image: Matterhorn Operations: Before Yawata Mission
Before Yawata Mission
Image: Matterhorn Operations: Attack on Mingaladon Cantonment Area, Rangoon
Attack on Mingaladon Cantonment Area, Rangoon
Matterhorn Operations:
Image: C-46's on the Assam-China Route: The First Ridge on the Hump Run
The First Ridge on the Hump Run
Image: C-46's on the Assam-China Route: Loading Gasoline Drums
Loading Gasoline Drums
C-46's on the Assam-China Route
Image: The Ledo Road Image: The Ledo Road
The Ledo Road


best oriented toward Sumatra, and, indeed, the leisurely pace of construction normal in Ceylon discouraged the selection of any virgin site.Construction of the fields would be a responsibility of Mountbatten as SACSEA. He had known of the MATTERHORN project at Cairo but had left before its final approval. He hoped to use the B-29's in a drive southeastward toward Singapore but had received no definite order to provide the VHB fields in Ceylon. On 5 March General Kuter, then in New Delhi on a mission for AAF Headquarters, received Lord Mountbatten's promise to build the B-29 fields when requested.72 The request came soon through Stilwell, whose directive of 6 March specifically called for staging fields in Ceylon.73 En route to Australia, Kuter stopped off at Colombo and conferred there with SEAC officers. He learned that the British were currently working on two bomber fields with long strips, apparently the ones referred to at Cairo. They were located at Kankesanturai, near Jaffna at the north end of the island, and Katunayake, in the west near Negombo; neither was favorably oriented, and completion dates were scheduled for late 1944 or 1945. As alternates the British suggested Minneriya and China Bay. Kuter preferred Matara in the extreme south, but that was vetoed because of its inaccessibility.74 Finally the four sites named by the British were accepted; China Bay and Minneriya, with high priorities, were scheduled for completion in July.75 In April it appeared that the date could not be met and, with JCS permission, work at Minneriya was temporarily suspended. Engineers from XX Bomber Command and AAF IBS worked with the British at China Bay in an effort to meet minimum requirements. Accommodations there were increased to handle two B-29 groups (fifty-six planes), and by concentrating on the one field SEAC was able to approximate the schedule. Some equipment, including a fifty-six-point fueling system, was sent in by XX Bomber Command. By mid-July, a 7,200-foot runway, the hardstands, and the fuel distribution system were complete, and when the first mission was run, belatedly, on 10 August, the field was fully operational.76

Movement Overseas

In January 1944 the Joint Intelligence Committee, considering the various base areas under consideration for use by the B- 29's, rated Chengtu as the locality offering the greatest logistical difficulties.77


Few persons in the MATTERHORN planning staff would have challenged that judgment. Referring to air operations in China, General Arnold had recently said for public information that "to supply our growing air strength in that country has been perhaps the greatest single challenge to the efficiency of the Air Force.''78 The B-29 project promised to aggravate an already complex situation. The most obvious difficulty lay in the lack of an adequate system of communications within the CBI, and the problems arising therefrom will be described in the next section of this chapter. Even to get the necessary men and supplies to India, however, taxed the ingenuity of officers in Washington and the CBI. Three factors, especially, handicapped their efforts: the inordinate distance from the United States to India; the low priority accorded the CBI in the allotment of shipping; and the insistence on an early commitment of the B-29, which left little time for readjusting existing transportation schedules.

The B-29's could be flown out by their own combat crews, a mere matter of 11,500 miles by the route chosen. Highest-priority passengers and freight could go out by ATC's planes via Natal, Khartoum, and Karachi, a trip which might be made in six days with luck but which for some XX Bomber Command personnel took more than a month. Eventually a special "blend" service was installed-by surface ship from Newark to Casablanca and thence to Calcutta by ATC transport--which required four to five weeks for passage. But the great bulk of troops and supplies had to be moved by water. Some units proceeded eastward via North Africa, where they transshipped in British vessels and went on through the Mediterranean and Suez. Other units and most supplies went westward from the States, around Australia, whence supply ships went up through the Bay of Bengal to Calcutta, and troop ships sailed to Bombay where the soldiers entrained for an uncomfortable week of travel to Kharagpur. One lucky contingent made it from Los Angeles to Bombay in thirty-four days, but most units were eight to ten weeks in passage from American ports of embarkation to their Bengal stations. A Liberty cargo ship could make a trip out in sixty days and accomplish two turn-arounds in a year. Ports in India were few, overtaxed, and inefficiently operated; even Calcutta was rated by a XX Bomber Command officer as "a good port with bad habits."

MATTERHORN was not, by standards of the ETO, a tremendous undertaking. The logistics tables used at SEXTANT called for bottoms


to accommodate 20,000 troop spaces and 200,000 tons of dry cargo between 1 January and 30 June, and something more than 20,000 tons of POL per month after 1 April.79 Bottoms were hard to find (a SEXTANT cable declared succinctly, "shipping is bottleneck") but sinkings by submarines in the last quarter of 1943 were fewer than had been anticipated and tonnage and troop spaces might be had by ingenious juggling of schedules and by accepting some delays. Troop transports were harder to find than cargo ships.80 To secure either type it was necessary to interpret liberally the proviso with which MATTERHORN was accepted--that it be mounted "without materially affecting other approved operations." The first units dispatched, the engineers who went out on the 15 December convoy, were provided for out of trooplift regularly assigned to the CBI.81 Stilwell had agreed to this but with the understanding that extra shipping would be allocated for other MATTERHORN needs. At SEXTANT additional lift for 3,000 troops was allotted to the CBI and was earmarked for two service groups, an air depot group, and various smaller units.82 By Christmas shipping had been found for all troops and supplies scheduled for XX Bomber Command through July.83 Allocation did not insure prompt delivery. It was important that initial organizational equipment go out with the units. In this category, Air Service Command items were dispatched with some promptness, but not so Army Service Forces items. It was necessary to set up special priorities for the latter in February, and by the 19th some 52,000 tons had been shipped, leaving a backlog in U.S. ports of only 4,000 tons.84 The late start was reflected in the need, already described, of borrowing heavy construction equipment in India.

Before the end of February most of the troop transports were at sea.85 One large contingent of men, including seven bomb maintenance squadrons, sailed from Newport News on 12 February in a convoy of Liberty ships bound for Oran, transshipped in the British-operated Champollion, and reached Bombay on 1 April. Other units, sailing on 22 February via Casablanca, went on from there in the Vollendam, arriving at Bombay on 25 April. More fortunate were those units, including eight bomb maintenance squadrons, which sailed from Los Angeles in the Mt. Vernon on 27 February and arrived at Bombay on 31 March.86 Other troops arrived in Bombay during April and went on to Bengal by rail. A station list of 10 May


showed 21,930 men on hand. This included some CBI and a few British troops attached to the command and those MATTERHORN personnel who had come out by air. But in all, something like 20,000 men, most of whom had come by sea, had arrived in India in March and April, and had been processed and put to work.87

Because of the pressure of time, air transport was of great importance in moving out personnel and high-priority freight. Small advance parties went out by regular ATC service. The first important movement was that of the twenty C-87's assigned to the command. Led by General Wolfe, those planes left Morrison Field on 5 January, carrying key personnel and some equipment, and arrived at New Delhi on the 13th.88 The original plan of ferrying out all combat crews, regular and extra, and some other passengers in the B-29's was scrapped. With the R-3350 engine still untried, it was considered necessary to have more than the usual number of spares, and it was decided to haul one engine in each B-29 in lieu of passengers. Even so, ATC would have to help. AAF Headquarters estimated requirements from that service as: February, 90 tons; March, 130 tons; April, 240 tons; May, 230 tons.89 passenger total, 1,252 Stilwell agreed to underwrite these amounts from his allotments.90 The movement of personnel from the various headquarters (command, wing, groups, and squadrons) began on 20 February. Priorities were low and there was the usual amount of "bumping" in favor of VIPs; some men were as long as thirty-five days en route, a little longer than those on board the Mt. Vernon. They arrived in India with some recently acquired geographical lore, souvenirs picked up in three continents, and loud gripes about ATC.91

Meanwhile, it had become obvious that the AAF's estimate of needs was too low and that some additional airlift must be provided temporarily, especially for the R-3350 spares.92 A special surface-air transport service was established, with passengers and freight going to Casablanca by ship and thence to Calcutta by ATC. For this, twenty-five C-54's were assigned to ATC's North African Wing.93 The shuttle service, known as "Mission 10," lasted from 8 April to 1 June, hauling about 250 engines and 1,225 passengers. Time in passage from the States wasthree to four weeks.94 This was only a stopgap for the crucial months of April and May. In mid-March Arnold had informed Wolfe of the intention of providing him with three "bomber support" squadrons with initial unit


equipment of eighteen C-46's each. Arnold's idea was that the first squadron might be used to augment the command's Hump tonnage and the second and third to operate on the Casablanca-Calcutta shuttle until October.95 The first squadron arrived in April, a month later than had been promised, and was put on the Hump run. The other units, now designated Ist and 2d Air Transport Squadrons (Mobile), were assigned to ATC's North African Wing and began the so-called "Crescent Blend" service on 6 June. This guaranteed to XX Bomber Command 333 tons per month (including about 225 engines) in June and July, slightly more thereafter.96 The service was something of a chore to ATC. The C-46 lacked the range of the C-54's normally used on the Casablanca-Calcutta run, and a new operational procedure had to be set up. The mobile squadrons had no service personnel attached; they had to "live off the land" and the land in this case was ATC.97 But the Blend was a valuable service for XX Bomber Command at a time when engine spares were essential to operations. In addition, a fifty-ton allotment of all-air delivery from the United States to India was assigned to Wolfe's command out of ATC's "Fireball" service.98

The overseas movements of the B-29's justified the expectation that R-3350 spares would be needed in substantial quantities. That movement had been postponed repeatedly, in anticipation, by delays in production and modifications of the B-29's and in the construction of the Bengal fields. By the end of January it appeared that most of the initial complement of 150 B-29's would be ready early in March and that by using various temporary expedients provision could be made for receiving them in India. Thus early March became the target date for dispatch.99

According to a plan worked out in Salina and Washington and coordinated with Eighth Air Force Headquarters, the first B-29 went to England via Natal and Marrakech.100 In part, this initial flight was to test the new bomber in long over-water flights, as well as to serve as a cover plan. The B-29, a hard plane to hide under a bushel, had been publicly announced by Arnold as ready for combat in 1944.101 The Japanese were aware of the existence of the abnormally long runways near Calcutta and Chengtu, and when the Superforts arrived in the CBI, it would take no mastermind to deduce their probable target. The cover plan called for controlled leaks to create the impression that the B-29 would be used in the ETO for combat but that because its range had not lived up to expectations the bomber would be used


in the CBI only as an armed transport.102 Stilwell gave "news" releases to that effect in his theater.103 Col. Frank Cook flew the B-29 to England early in March and exhibited his plane as directed.104 Flight data transmitted to Salina indicated no serious variations from previous experience.105 Cook went on to Kharagpur, arriving on 6 Apr1;106 his B-29 was the second to reach the goal. On 1 March General Arnold had informed the CBI of the flight schedule for the 58th Wing.107 The planes would go out in daily increments of nine or ten planes, beginning on 10 March; with a five-day trip planned, this would put all the B-29's at their stations by the 31st.

The designated route was:
Salina to Gander Lake 2,580 miles
Gander Lake to Marrakech 2,700 miles
Marrakech to Cairo 2,350 miles
Cairo to Karachi 2,400 miles
Karachi to Calcutta 1,500 miles
Total 11,530 miles

By 10 March it was necessary to retard the initial flight and each subsequent increment by a fortnight; according to the new schedule, all the B-29's would arrive between 1 and 15 April.108 The lead plane almost made it in on time. Very much impressed with the "historic" significance of this first arrival, public relations officers staged an elaborate welcome, with a fighter escort aloft and a plentiful supply of brass, sound film trucks, and reporters on the ground. After several false alerts and eleventh-hour changes in the ETA, the audience had lost something of the sense of drama when Col. L. F. Harman eased his Superfort onto the runway at Chakulia on 2 April.109 By 15 April only thirty-two planes were at their stations. Save for one forced landing at Presque Isle, the B-29's had made the ocean passage without trouble, but then misfortune set in. First came a total wreck at Marrakech on 13 April, then a partial one at Cairo on the 10th, and then, in rapid succession, five serious accidents including two planes completely lost at Karachi. All planes along the route were grounded from 21 to 29 April. Investigation proved that most accidents had occurred from engine failures, some of which could be blamed on inexperienced crews.110 When flight was resumed the ferrying went more rapidly. On 8 May, 148 of the 150 planes had reached Marrakech and 130 had arrived at their home fields.111 The movement was under control of ATC and both that organization and XX Bomber Command profited by experience. This is shown by the


safety record. Of the original 150 planes, 5 were lost and 4 seriously damaged; yet by March 1945 when the movement of B-29's to India ceased, 405 planes had been ferried out with only 8 total losses--that is, 3 out of the last 255.112

The elaborate cover plan seems to have fooled no one-at least not Japanese intelligence. There is a report to the effect that Colonel Harman's arrival at Chakulia was greeted by an enemy radio broadcast which identified the B-29, and Japanese announcers continued to comment on the VHB fields near Calcutta and Chengtu.113 XX Bomber Command and ATC made mutual accusations of security breaches along the ferry route, and in both India and China the Japanese had many agents.114 Whatever the source of the leak, when the enemy had a brief test of the B-29's armament in an interception of an over-the-Hump run on 26 April, he seems to have had no illusion that he had tangled with the long-range armed "supertransport" of the news releases.115

The several units settled into their Bengal bases: XX Bomber Command Headquarters and 468th Group, Kharagpur; 58th Wing Headquarters and 40th Group, Chakulia; 462d Group, Piardoba; and 444th Group, Charra (temporarily).116 Neither the India nor the China bases were operationally complete, but the successive delays in arrival of the B-29's made that of less importance than it had appeared, On 26 April Arnold wrote to Wolfe: "The airplanes and crews got off to a bad start due to late production schedules, difficult modifications, inclement weather, and the sheer pressure of time necessary to meet the early commitment date.117 Perhaps the last was the most important element, for from November 1943 on it had made impossible any close articulation of the various stages in the deployment plan. Thus in early May, with his combat elements on hand or momentarily expected, Wolfe was still faced with the task of building up a stockpile before he could launch his first mission, already overdue by the Cairo schedule of operations. Both stockpiling and the B-29's themselves were endangered, now that the MATTERHORN designs could be sensed by the Japanese, by the late arrival of the fighter defense forces for Chengtu.

Air defense of the B-29's in China was Chennault's responsibility. In September 1943 he had stated his requirements as "at least 1 Gp of fighters (150 P-51's recommended);118 the MATTERHORN plan had called for two fighter groups. At Cairo the CCS decided to transfer


two P-40 groups from Italy, re-equipping them with P-47's.119 Stratemeyer asked that the P-47's be sent from the United States to Karachi in January and February, and that the pilots arrive in time to complete transitional training there.120 The units could not be released, however, until after the initial phase of the Anzio operation (D-day, 22 January), and by ordinary surface shipment the P-47's could not get to Karachi before 1 May. At the AAF's request, the Navy agreed to ferry out 100 P-47's on the CVEs Mission Bay and Wake Island; the other 50 would go by regular transport.121 The units selected were the 33d and 81st Fighter Groups, veterans of the North African, Sicilian, and Italian campaigns. The ground echelons sailed from Taranto and arrived at Bombay on 20 March; the flight echelons proceeded by air in mid-February.122 The two CVE's brought the P-47's into Karachi on 30 March and two weeks later transitional training was begun.123

To provide for proper control of the fighters, the Fourteenth Air Force on 1 March activated the 312 th Fighter Wing, of which Brig. Gen. A. H. Gilkeson, just arrived from the States, assumed command on 25 March.124 When the first B-29 landed at Chakulia on 2 April, the wing was still only a skeleton organization, inadequately staffed, with its personnel scattered from Karachi to Chengtu, and with only a few P-40's available for immediate combat. The situation caused some justifiable alarm. There was little fear for the Bengal fields, for though Calcutta had been bombed during Christmas week, the B-29 bases lay farther west, at extreme bomber range for the Japanese, and RAF and Tenth Air Force fighters were considered adequate protection. The dangers in China were much more apparent, and Chennault grew progressively more pessimistic in his analyses of enemy capabilities.125 He attempted to get additional fighters to guard the Assam-China air route and to hasten the delivery of two squadrons of P-61 night fighters promised for July. He wished also to increase the number of fighters allotted to Chengtu, and to re-equip his new units with P-51's, much more economical of fuel than the P-47's, though he had accepted the latter planes under the assumption that they would be supplied by XX Bomber Command transports.126

Stilwell shared Chennault's anxiety and early in March had suggested that the target date for B-29 operations be postponed a month to allow the defense forces to be readied.127 When this request was refused, it was decided to send one squadron of the new wing on to Chengtu with their old P-40's, and allow the other five squadrons to


follow as they were re-equipped with P-47's.128 The 59th Squadron flew into Szechwan province with its P-40's and was the only local fighter defense when the B-29's began their transport activities late in April. The other two squadrons of the 33d Group (58th and 60th) followed with their P-47's in May. On 15 May, the 92d Squadron of the 81st Group arrived at Kwanghan, but it was two months later before the other two squadrons (the 91st and 93d) came.129 Japanese attacks on the Chengtu fields were to prove less intensive than had been feared, and the late arrival of the fighters should have eased somewhat the task of stockpiling fuel for B-29 missions. That task became the chief concern of XX Bomber Command and the needs of the 312th Wing continued to be an important part thereof.

Transport Within the CBI

The MATTERHORN logistics plan was a long document, but its essence was compressed into a single sentence by an early emissary of XX Bomber Command in the CBI. "Remember too," he wrote to a friend at Salina, "that every single goddam thing that we send into China has to be flown in." There was little opportunity to forget that fact. MATTERHORN transportation difficulties began at factories and depots, at air bases and seaports in the United States, and dogged each ton and passenger along the slow trip to India. Yet it had been possible to move out XX Bomber Command and its equipment without disrupting too seriously existing shipping schedules; resupply would be comparatively simple. The rear area bases were well located, with rail and motor road connections with Calcutta and the facilities grouped around the city--the port, the ATC terminus and the Bengal (28th) Air Depot at Barrackpore, and ASC's installations at the Alipore airport. Surface transportation routes in the region, good by India's standards, proved unequal to the new demands and the command had to rely in part on an inter-base air-shuttle service in Bengal. But this was a minor evil; the crucial stage in the MATTERHORN supply route was the Calcutta-Assam-Chengtu haul, with the fabulous Hump as its midriff.The distance, while great, was not prohibitive: a B-29 with cargo could easily make the 1,200 miles or so from Kharagpur to Hsinching in five to five and one-half hours. The movement of goods along existing theater channels was much slower: ordinarily a shipment would


Map: Hump Routes of XX Bomber Command
Hump Routes of XX Bomber Command


proceed from Calcutta to Assam by river barge and rail, and thence via Kunming to Hsinching by ATC plane, taking several weeks in transit. That mode of transport did not figure originally in the MATTERHORN plan. The India-China Wing (ICW) of ATC had materially improved its Hump tonnage during the autumn months of 1943. In December 12,594 tons were delivered in China, more in January and February, and though the totals fell off in each of the spring months of 1944, there would be a marked increase from June on. But that tonnage was jealously regarded by the several using agencies, of which the Fourteenth Air Force was chief. The various CBI commands had accepted the MATTERHORN plan without enthusiasm and with a clear understanding that the VLR project would not impinge upon current allocations for transport. Approval at Cairo had carried the same proviso. The key feature of MATTERHORN was that XX Bomber Command would supply its own staging bases, using its B-29's and its twenty C-87's.

Unable from the beginning to sustain itself, the command had to turn to ATC for aid. This antagonized other theater agencies and, when the aid proved insufficient, led to mutual recriminations between them and the VHB command. The latter tended to blame ATC, while ATC and the Fourteenth looked on the bomber command as an interloper with specious claims of independence and a habit of sponging on the strained services of ICW. This lack of understanding is reflected in the several accounts, which differ sharply according to provenance, of some of the important agreements. Even more disconcerting is the wide variation among the statistical records, which make it impossible to establish exactly the tonnages allotted, on loaded, or actually delivered to MATTERHORN users. Some inaccuracies were unavoidable under the circumstances--the ICWs historian wrote of the Chengtu area: "Records-of-tonnage allocations and deliveries were not kept primarily because no personnel were on hand to keep such records for ATC. The personnel at Hsinching were, for the most part, mechanics." But figures emanating from better-staffed headquarters have to be used with caution, and it is rare that perfect agreement can be found among several sources.*

Fundamentally the MATTERHORN supply plan was uneconomical, as must be any based on long hauls by air with fuel available at

* The figures which XX Bomber Command gives on its own transport activities can be checked against the daily record sheets; they are quite accurate. But in regard to ATC's contribution, XX Bomber Command estimates vary widely from those of ICW.


only one terminus. This Washington had always granted. Probably the transport resources added to MATTERHORN in successive increments might have yielded more tonnage had they been assigned to ATC, but the Twentieth Air Force feared to lose control of transport aircraft without a firmer guarantee than could be had. Given time,

Wolfe might have been able to approximate his original design. But he worked always with an impracticable target date, and delays originating in the United States became cumulative in the CBI--delays in the arrival of troops, equipment, and aircraft, in the preparation of fields and installations. Tactical emergencies in the CBI interfered too with stockpiling for the first missions, so that the initial strike against Japan was repeatedly postponed, and when finally launched, its weight was well below earlier plans.

Wolfe had to establish his forward area base, move up the requisite equipment and personnel, nourish the latter (the 312 th Fighter Wing and the 315th Service Group), and build a stockpile for his initial missions. For these transport tasks he had counted on the tactical B-29's and the twenty C-87's assigned his command. Wolfe brought the C-87's out to India in mid-January (losing one en route) with ATC crews on ninety-day temporary duty but with no organizational or maintenance personnel. AAF Headquarters had intended that the 308th Bombardment Group (H) should operate the planes for Wolfe's benefit.130 General Stratemeyer objected to this additional burden for the 308th and won Wolfe's approval of another arrangement.131 The nineteen C-87's would be turned over to the ICW in return for a guaranteed monthly tonnage; on 15 April the ATC crews would return to the States, and XX Bomber Command would resume operation of the transports.

This arrangement constituted a slight but real departure from the

regard to ATC?s contribution, XX Bomber Command estimates vary widely from those of ICW.


1944 Feb. March Apr. May June July Aug. Sept.
XX BC C-46's -- -- 14 117 280 1,162 798 707
Tactical B-29's -- -- 27 518 404 1,083 -- 504
Tanker B-29's -- -- -- 22 396 753 1,106 814
C-109's -- -- -- -- -- -- -- 415
Total XX BC -- -- 41 657 1,080 2,998 1,904 2,440
ATC 427 2,608 1,399 1,293 308 976 1,904 2,141
GRAND TOTAL 427 2,608 1,440 1,950 1,388 3,974 3,382 4,581


doctrine of self-sufficiency. For February, ICW promised Wolfe 1,650 tons from the first 10,250 flown over the Hump, and half of all surplus up to 11,500--a possible total of 2,275. This seemed more than the C-87's would haul, and the theater proposed to make good the deficit out of the allowance for the Burma-China pipeline.132 ATC made 12,920 tons that month, but XX Bomber Command profited little: Wolfe released to Chennault 1,534 tons (of the basic 1,650) to be repaid later and Chengtu got only about 400tons.133

March was a bad month for ICW; with a gas shortage in Assam and a serious diversion of C-46's in favor of Burma ground operations, Hump tonnage fell to 9,587.134 Yet 1,997 tons were allocated to MATTERHORN, and ICW reported that it carried for the project 3,603 tons, the guarantee plus 1,606 tons to repay the February loan to Chennault.135 Wolfe's version of the transportation was different. Of the 3,603 tons onloaded in Assam for MATTERHORN, 682 had been diverted to "other activities" and only 2,921 delivered at Chengtu. Of this amount, Chennault, who was badly squeezed by the light haul in March, claimed 800 for April delivery-the 312th Fighter Wing had to be set up at Chengtu.136 By either reckoning, stockpiling was badly in retard. Stilwell's directive of 5 March called for the B-29's to stage one shakedown mission from Calcutta and one regular mission from Chengtu in April, three in May. With the late arrival of the B-29's and the slow build-up of supplies in China, that directive had to be scrapped.

In this crisis, Washington resorted to an expedient suggested earlier by the CBI--assignment of additional transports to MATTERHORN.137 These were the C-46 bomber support squadrons mentioned in a previous passage, of which the first contingent reached Bengal on 10 April.138 Some of the C-46's were put on the inter-base shuttle in Bengal; others, based in the Kharagpur area, began the Hump run, but during April hauled only fourteen tons into China.139 The self-service B-29's did little better: by 1 May, once looked on as D-day, they had laid down in China twenty-seven tons, just enough to support one combat sortie.140 The main burden in April was still on ATC. Wolfe claimed an allotment of 2,000 tons but received only 1,399, the other 600 going to Stilwell's Yoke Force on what Wolfe thought was a loan.141 In all, 1,440 tons went forward in May.

Using a planning factor of 23 tons per B-29 combat sortie from Chengtu, Wolfe had hoped to have by 1 May a 6,000-ton stockpile to


support two 100-plane strikes. According to his figures, he had received less than 4,800 of which 800 were claimed by Chennault; much of the balance went for uses other than the stockpile. High-octane gasoline was particularly short, with only 380,000 gallons in storage instead of an anticipated 660,000. With the transport capabilities of the B-29 appearing less impressive in practice than in anticipation and with a fixed charge for support of the 312th Fighter Wing now facing him, Wolfe felt that he might have to scale down the weight of attacks against Japan.142 Additional transport equipment would see him through the present emergency, and Wolfe hoped to secure that help in the form of the C-46 squadrons assigned to the Crescent Shuttle for his support. Control of those squadrons (and of the C-87's) became then a matter of grave importance, much discussion, and several short-lived agreements between XX Bomber Command and ATC during April and May. None of these agreements was wholly satisfactory, nor was the arrangement worked out in a Washington conference on 12 May between representatives of AAF Headquarters and ATC.143 A week later Stratemeyer had engineered another compromise between Wolfe and Brig. Gen. T. O. Hardin of ICW. The remaining C-87's and thirty-six C-46's would be assigned to Hardin, and the 1st Air Transport Squadron (Mobile) to Wolfe. ICW would lay down 1,500 tons monthly at Chengtu, of which 1,000 tons would be carried from Calcutta to Jorhat by Wolfe's planes, 500 by ICW.144 All this shuffling of units--some of which had not even arrived--effected no great improvement in May deliveries. Wolfe hoped to get from ATC his 1,500-ton guarantee, plus the 600-ton "loan" to the Yoke Force. The latter, however, had been written off by Chungking, and only 1,293 tons were offloaded in the Chengtu area.145 The C-46's operated by the command, still new on the Hump run, carried 107 tons;146 the tactical B-29's delivered 540 tons in 141 sorties, far less than the early estimates and about one-third of Wolfe's revised figures.147 That record would be bettered as the full complement of planes swung into the job and as crews and ground organizations improved. But Wolfe had come to feel that the "use of B-29 as a cargo carrier has definite limitations and any large scale operations should be dependent upon regular cargo-type aircraft for supplies." He also pointed to the obvious fact that regular use of the B-29 as a transport would shorten its combat life.148

This attitude was a negation of the very essence of the MATTERHORN


plan. Wolfe and the Washington planners must have realized from the first that it would have been more economical to supply the B-29's by regular cargo planes than by the Superforts' own efforts; but lack of cargo planes in sufficient numbers, pressure of time, and perhaps fondness for the AAF concept of the bomber unit as a self-contained entity had led to the adoption of a logistical system which had already been modified and which was facing collapse. The one hopeful statistic was too small to be appreciated yet--the twenty-two tons hauled in May by B-29 "tankers." Wolfe had stripped some planes of all combat equipment except tail guns and a minimum of radar, and thus was able to haul seven tons of aviation gasoline (avgas) per trip as against three in a tactical plane. This stripping was questioned in Washington, but planes could be combat-readied in a week, and until the stockpile grew, there could be no combat missions.149alculated that his first two missions (100 sortes each) would require 4,600 tons plus what the tactical B-29's carried.150 This he could not transport rapidly with resources presently available; by 26 May he estimated that, by reaching a total of 4,840 tons in June, he could stage his first strike about the 20th and a second in July.151 This schedule the enemy spoiled. At the end of May the Japanese pushed off in their long-anticipated drive for the Canton-Changsha railroad. On 4 June Stilwell diverted to the Fourteenth tonnage previously guaranteed to MATTERHORN. The JCS sanctioned this, but they refused Chiang Kai-shek's request, forwarded by Stilwell without indorsement that the MATTERHORN stockpile be turned over to ChChennault in the emergency.152 After questioning Wolfe as to his immediate capabilities, the JCS on 8 June ordered him to put at least seventy B-29's over Japan on the 15th--this to relieve pressure in east China and to coordinate with the landing on Saipan.153 Even a strike of this reduced weight could be achieved only by increasingly drastic economies in the forward area.

Since the war, General Chennault has stated publicly that such economies were not effected:

The Twentieth Air Force refused to face the realities of the China supply situation. Even when gas was so low at Chengtu that their defending fighters could not fly local interceptions, the Twentieth refused to live off the land and operate on skeleton tables of organization. They continued to fly in thousands of tons of American food and excess personnel into [sic] China at the expense of gas and bombs. . . .They always retained indelible recollections of the Pentagon standard of living.154


A different version of the story came from XX Bomber Command. In 1944, while the supply problem was still a very live issue, the command's historian wrote:

Faced with the necessity of executing a combat mission on the directed date, despite its reduced transport capacity, the command had only one alternative: to reduce the delivery of equipment, supplies and personnel to all units in the forward area to the bare essentials required to sustain life and permit the airplanes to take off for the target. These instructions were so stringent that all surface transportation to [sic] the forward area ceased with the exception of one vehicle per base. No supplementary rations were supplied to the garrisons in the area. All supplies of PX rations were eliminated. There was no shipment of clothing, less than 25 percent of the mail. No hospital rations and no additional personal or organizational equipment were supplied. Indeed, insofar as supply was concerned, personnel in the forward area were isolated and limited as if they had been on a desert island. Full colonels walked two miles to their airplanes.155

This passage is, for the period concerned, an almost point-by-point denial--five years in advance--of Chennault's blanket charges. The detail of the walking colonels may tax the credulity of some readers, but during the emergency of late May and June there does not seem to have been much "Pentagon standard of living" at Chengtu.

Unfortunately for intra-theater amity, the economies begun in May had been applied to the Chengtu organizations belonging to the Fourteenth Air Force. By agreement between Wolfe and Chennault, the 312th Fighter Wing and the service forces were to get 1,500 tons per month--half of the figure originally demanded. In May this allotment was reduced to 1,000 tons, apparently without consulting Chennault and certainly without full coordination with the 312th.156 For want of a priority list from the fighter units, XX Bomber Command determined what goods should go forward as well as total tonnage. In the June emergency the 312th fared even worse; Chennault claimed the wing got "practically nothing," while XX Bomber Command statistics said 600 tons.157 When the command flew its first combat mission from Chengtu, Gilkeson had enough gasoline to fly only four two-hour sorties with 60 per cent of his fighters; not unnaturally he was alarmed.158

The pinch was felt by others as well. In spite of economies and strenuous efforts to increase net tonnage, XX Bomber Command on 15 June could hardly support at Chengtu the minimum demand for a seventy-sortie mission. This effort so bled the forward fuel stocks that


some planes could not return immediately to Calcutta.159 ATC's deliveries, cut off on 5 June, amounted in the month only to something over 300 tons; XX Bomber Command had done somewhat better for itself than in May, with 280 tons by C-46's and 800 tons divided about equally between B-29 tactical planes and tankers. The two combat missions, the shakedown on 5 June and the trip to Japan on the 15th, had interfered sharply, and the month's total of 1,388 tons was the lowest since February.160

To make up the deficit caused by the diversion of ATC tonnage, Arnold reassigned to XX Bomber Command the 2d and 3d Air Transport Squadrons (Mobile), then working on the Crescent Blend. The 2d was on the Hump run before the end of June, the 3d by 8 July.161 By the latter date Wolfe had some 60 C-46's and his B-29's to meet requirements for his July target directive: a 15-sortie mission early in the month, a 100-sortie effort during the last 10 days.162 To insure a build-up, he again cut back the 312th Wing, this time to 850 tons. By his staff's calculation, this should give the fighter groups ten hours' flying time per pilot and a small reserve.163 Admittedly it was a slim margin, and though more than June deliveries, 850 tons fell far short of the original agreement and of the 312th's idea of a safe minimum. (Ironically, XX Bomber Command returned Chennault's charges of exaggerated standards of living, saying that the 33d and 81st Fighter Groups, accustomed to the luxurious life of the MTO, could not adjust to the scarcity economy of China.)164

Chennault on 25 June informed Arnold of the "deplorable conditions" and stated flatly that "under existing conditions I cannot be held responsible for defense of Chengtu."165 Settlement of the immediate problem fell to Stratemeyer who had logistical responsibilities for both the Fourteenth Air Force and XX Bomber Command.166 The correspondence from the generals concerned which passed over Stratemeyer's desk during the next week was acrimonious in the extreme. Chennault accused Wolfe of cutting back fuel deliveries to the 312th beyond the safety line and of determining cargo priorities arbitrarily ("Gilkeson has no idea as to what he will receive and is entirely at the mercy of Wolfe who controls the purse strings. . . .")167 Wolfe denied the accuracy of Chennault's figures on fuel deliveries and affirmed that the amounts scheduled had been agreed to by Gilkeson.168 On 3 July Wolfe had to accede to Chennault's demand that the previous guarantee of 1,500 tons monthly for the 312th be restored,


with the further concession that Chennault determine the breakdown of the tonnage.169 Four days later the arrangement was modified somewhat by mutual agreement. Wolfe turned over to Chennault, effective 20 July, XX Bomber Command's claim to 1,500 tons monthly from ATC and was relieved thereby of all logistical responsibility toward the 312th and its service organizations. This was an excellent deal for the command, ending a long and bitter dispute, and cutting down on staff work.170

Furthermore, the total lift for July set a record. The ATC allotment, restored during the first 20 days of the month, amounted to 976 tons, just enough to meet the 312th's quota. The bomber command's enlarged fleet of C-46's hauled 1,162 tons, the B-29 tankers 753, the tactical B-29's 1,083. The latter record was accomplished in spite of a halt for the two designated missions which were run off as scheduled with a combined total of 115 sorties.171 The improvement came as the command learned more about the B-29 and C-46, and more about the air transport business. Lt. Col. Robert S. McNamara's Statistical Section practically ran the show, watching carefully the variable factors: aircraft in tactical use, aircraft out of commission, turn-around time, gross load, and net offload.* The first factor was of course out of their control, but in the others careful study brought marked improvement. Thus, between May and the end of July the average gasoline consumption on a round trip was reduced from 6,312 to 5,651 gallons; the net offload rose in the same period from 495 to 1,326 gallons, and at the end of July tankers were laying down 2,496 gallons net.172 At either period it was expensive, but at worst it meant burning twelve gallons of gasoline to put down one in Chengtu, at best two for one; the margin was the measure of the command's adjustment to the transport task.



1944 April May June July Aug. Sept.
B-29 transport trips 7 238 164 237 116 206
C-46 transport trips   58 150 419 368 265
B-29's in commission (for transport or operations)   38% 37.5% 41.3% 41.1% 50%
B-29 abortive rates 18.2% 14.7% 18.1% 11.5% 7.5% 9%
B-29 turn-around time in China--in days   2 1.8 1.4 1.5 1.5
B-29 average net offload per trip--in tons   2.25 4.87 7.66 9.53 6.40


The flight by a B-29 was a "through" trip via the Assam valley and over the Hump without making Kunming. In the early months fear of enemy interception sent the planes along a northern or southern route, according to which had weather dangerously clear or overcast enough to render interception difficult; later each group had its own route. But Japanese fighters caused little trouble: there had been six or seven contacts by the end of July but no determined attacks.173 The route was a dangerous one, nonetheless, with its jagged ranges and uncertain weather and communications, so that combat time was allowed for all transport trips. In the same period an even dozen B-29's were lost, mainly from engine failures, as against six C-46's by September.174 Most of the crews were saved. Some bailed out over friendly territory and received hospitable treatment from the Chinese. Others fared less well, coming down in the dread Lolo country. Their walkout reports and the report of Capt. Frank Mullen of Air Ground Aid Service, who penetrated the Lolo land on a rescue mission, portray a wild country and a people as untouched by western civilization as in the days of Marco Polo.175

In July for the first time XX Bomber Command approached its ideal of self-sufficiency; the 3,000 tons hauled forward by its own planes just about supported the 115 sorties. But this was the peak of performance by the B-29's and the weight of attack against Japan was but half what had been anticipated earlier. If the resources already poured into MATTERHORN were to be more fully realized, the supply system must be revamped. The changes were to come in late summer when a change in strategic plans in the Pacific called for a more intensive air effort in China.


Table of Contents
Previous Chapter (2) * * Next Chapter (4)

Notes to Chapter 3

1. CM-IN-9027 (11 Sept. 431, AQUILA #2106 TA, 11 Sept. 1943; CM-IN-IN-19022 (31 Oct. 43), New Delhi to AGWAR, #W2542, 31 OCt. 1943.

2. CM-OUT-417 (10 Nov. 43), President to PM; CM-OUT, number un-known (10 Nov. 43), President to CKS; CCS 400/1, 23 Nov. 1943 (quoting Churchill's reply) ; CM-IN-8594 (14 Nov. 43), Chungking to AGWAR, #876, 14 Nov.1943.

3. AAFRH-12, pp. 144-45, 165.

4. CM-IN-9027 (11 Sept. 43), AQUILA #2106 TA, 11 Sept. 1943.

5. CM-OUT-10738 (17 Nov. 43), Stilwell to AGWAR, AG 2882, 17 Nov. 1943; memo for Gen. Perrin from Gen. H. A. Craig, 25 Dec. 1943.

6. Hist. XX Bomber Command, Third Phase, p. 47 ff. [hereinafter cited XX BC, Third Phase].

7. CM-IN-8578 (14 Jan. 441, Stratemeyer to Arnold, W111, 14 Jan. 1944.

8. AAFRH-12, pp. 142-44.

9. B-29 Project, Reference Data (n.d., but ca. 7 Nov. 1943); JPS 320, App. A, 9 Nov. 1943.

10. JPS 113th Mtg., 9 Nov. 1943; memo for C/S from CG AAF, 13 Nov. 1943.

11. JCS 600/2; JCS 124th Mtg., 17 Nov. 1943; CM-IN-15448 (25 Nov. 43), SEXTANT 10037, 25 Nov. 1943; CM-OUT-10879 (27 Nov. 43), SEXTANT 1113, 25 Nov. 1943.

12. CM-OUT-10177 (26 Nov. 43), AMMDEL #4506, 25 Nov. 1943; CM-OUT-10880 (27 Nov. 43), AMMDEL #4538, 27 Nov. 1943; CM-IN-110 (1 Dec. 43), AMMDEL #3085, 30 Nov. 1943.

13. CM-IN-17764 (28 Nov. 43), AQUILA #W2755, 28 Nov. 1943; memo for CG ASF from Gen. Lutes, 25 Dec. 1943.

14. CM-IN-133569 (22 Nov. 43), AQUILA #W2696, 22 Nov. 1943; CMIN-15435 (24 Dec. 43), AMMDEL #3517 24 Dec. 1943.

15. Memo for CG ASF from Gen. Lutes, 25 Dec. 1943; CM-IN-16659 (27 Dec. 43), AQUILA #3020, 27 Dec. 1943; CM-OUT-9949 (27 Dec. 43), AM-MISCA #4140, 28 Dec. 1943; CM-IN-987 (2 Jan. 44), Stilwell to Marshall.

16. AAFRH-12, p.149.

17. JCS 600/3, 15 Jan. 1944; memo for C/S from Arnold, 15 Jan. 1944.

18. CM-IN-10519 (16 Jan. 44), Sultan to Arnold, W146, 16 Jan. 1944; AAFRH-12, p. 150.

19. CM-IN-155155 (22 Feb. 44), New Delhi to WD, #509, 22 Feb. 1944.

20. AAFRH-12, pp. 151-52.

21. Construction Service, SOS USAF CBI, Final Rpt. B-29 Bases in India, November 1944, p. 8. This excellent report by Col. Seaman [hereinafter cited Final Rpt.] has proved the most useful source for construction details.

22. CM-IN-5843 (9 Jan. 44), Stratemeyer to Arnold, W80, 9 Jan. 1944; Final Rpt., pp. 12-13, 34-35.

23. CM-IN-12900 (21 Nov. 43), TIGER GW 1160, 20 Nov. 1943; CM-IN-13730 (22 Nov. 43), New Delhi to ASC, W895, 22 Nov. 1943; CM-IN-16925 (24 Feb. 44), SACSEA to C/S, SEACOS 103, 23 Feb. 1944; Final Rpt., p. 26.

24. CM-OUT-1234 (3 Dec. 431, AQUILA #3890, 3 Dec. 1943.

25. Final Rpt., p. 40.

26. Ibid., pp. 26-32.

27. XX BC, Third Phase, pp. 47-64; Final Rpt., pp. 13-15.

28. AAFRH-12, p. 158 and accompanying progress chart.

29. Final Rpt., p. 4.

30. For a general description of the valley, see W. C. Lowdermilk, "China Fights Erosion with U.S. Aid," National Geographic, LXXXVII (June 1945), 641-81; also the well-written appreciation by the historian of XX Bomber Command, in Hist. XX Bomber Command, Fourth Phase: The Forward Area, pp. 3-9 [hereinafter cited Forward Area].

31. CM-IN-13509 (22 Nov. 43), New Delhi to WAR, W2696, 22 Nov. 1943; CM-IN-17764 (28 Nov. 43), New Delhi to AGWAR, 28 Nov. 1943.

32. Ltr., Stratemeyer to Arnold, 5 Jan. 1943; CM-IN-550 (1 Jan. 44), New Delhi to AGWAR, W2, 1 Jan. 1944.

33. XX BC, Third Phase, pp. 23, 35.

34. CM-OUT-6080 (15 Jan. 44), Arnold to Stratemeyer, #4311, 15 Jan. 1944.

35. AAFRH-12, p. 163; ltr., Chennault to Arnold, 26 Jan. 1944.

36. Engr. Sec., 14th AF, Airfields in Unoccupied China, 20 July 1944.

37. Ibid., 30 Nov. 1944.

38. AAFRH-12, pp. 164-65.

39. Ibid., pp. 165-66; Forward Area, pp. 14-17.

40. AAFRH-12, p. 166.

41. Forward Area, p. 15.

42. Ltr., Godfrey to Col. George Mayo, 8 Jan. 1944, Forward Area, p. 16.

43. Forward Area, pp. 18-20.

44. Brooks Atkinson in N.Y. Times, 17 June 1944.

45. Forward Area, p. 20.

46. XX BC, Third Phase, pp. 161-65; memo for Gen. Somervell from Col. F. K. Newcomer, 10 Apr. 1944.

47. CM-IN-11893 (19 Dec. 431, Chungking to AGWAR, #946, 19 Dec. 1943.

48. CM-IN-15685 (25 Dec. 43).

49. CM-OUT-8195 (19 Dec. 43), AM-MISCA #4100, 19 Dec. 1943; memo for record, 20 Dec. 1943, D. 2 MATTERHORN.

50. CM-IN-351 (1 Jan. 44), Chungking to AGWAR, #1, 1 Jan. 1944.

51. CM-OUT-1223 (4 Jan. 44), Marshall to Stilwell, #4199, 4 Jan. 1944; CMIN-4645 (8 Jan. 44), Stilwell to Marshall, AG 18,7 Jan. 1944.

52. AAFRH-12, p. 170.

53. Ltr., Voynow to Farthing, 4 Jan. 1944.

54. AAFRH-12, p. 171.

55. CM-IN-13539 (21 Jan. 44), Chungking to AGWAR, #46, 19 Jan. 1944; Stilwell Papers, pp. 251-52; CM-IN-3024 (5 Feb. 44), Chungking to AGWAR, #86, 4 Feb. 1944.

56. AAFRH-12, p. 171.

57. CM-IN-4096 (6 Mar. 44), AQUILA #W649, 6 Mar. 1944.

58. For a fully documented account of the financial negotiations, see Rpt. of Proceedings of Officers to Investigate the Chengtu Airfields, 24 Sept. 1945 ; Chennault, Way of a Fighter, p. 239 puts the cost at $350,000,000.

59. XX BC, Third Phase, p. 162; Forward Area, p. 18.

60. Forward Area, p. 18.

61. N.Y. Times, 17 June 1944.

62. Forward Area, pp. 59-63.

63. Airfields in Unoccupied China, 20 Feb. 1944; Forward Area, pp. 10-22.

64. CM-IN-6334 (10 Jan. 44), Wedemeyer to Marshall, #97, 10 Jan. 1944

65. AAFRH-12, p. 174.

66. Forward Area, pp. 20-21.

67. Div. #I, SOS USAF CBI Wkly. Progress Rpt., 18 May 1944.

68. Ibid.

69. Forward Area, p. 22.

70. CCS 426/1, 6 Dec. 1943.

71. AVM Moore, Ceylon Airfields, Heavy Bomber Type, 7 Dec. 1943 (SEXTANT).

72. CM-IN-3162 (5 Mar. 44), AQUILA #W634 RACS, 5 Mar. 1944.

73. CM-OUT-2115 (6 Mar. 44), AMMISCA #4685, 5 Mar. 1944; CM-IN-11904 (17 Mar. 44), Kuter to Arnold, W161 RG, 17 Mar. 1944.

74. CM-IN-20648 (29 Mar. 44), Kuter (SWPA Hq.) to Arnold, AX30385, 28 Mar. 1944.

75. CM-IN-21136 (29 Mar. 44), AQUILA #W903, 20 Mar. 1944.

76. AAFRH-12, p. 178; Hist. XX BC, Fourth Phase: Double-Strike, Palembang-Nagasaki, pp. 7-14.

77. JIC 152/1, 6 Jan. 1944.

78. Report of the Commanding General of the Army Air Forces to the Secretary of War (4 Jan. 1944), p 39.

79. JPS 320, App. B, Shipping Requirements by Months at Calcutta, 9 Nov. 1943.

80. AAFRH-12, p. 189.

81. Ibid., p. 147.

82. CM-OUT-2209 (6 Dec. 43), AM-MISCA #3990, 6 Dec. 1943; CM-OUT-2529 (7 Dec. 43), AMMDEL #4739, 7 Dec. 1943; CM-OUT-8157 (21 Dec. 43), WAR to Stilwell, #4056, 21 Dec. 1943.

83. CM-OUT-8157 (21 Dec. 43), WAR to Stilwell, #4056, 21 Dec. 1943.

84. Memo for DC/AS from Brig. Gen. H. A. Craig, 25 Dec. 1943; ltr., Maj. F. A. Thomas, Jr. to Gen. Wolfe, 15 Jan. 1944; AFAEP Daily Activity Rpt., 8 Feb. 1944; CM-OUT-8383 (19 Feb. 44), AQUILA #4694, 19 Feb. 1944.

85. CM-OUT-8383 (19 Feb. 44), AQUILA #4694, 19 Feb. 1944.

86. AAFRH-12, p. 190 and n 26.

87. Hist. XX BC, Apr. 1944, p. 2: XX BC Station List of 10 May 1944.

88. CM-OUT-1560 (5 Jan. 44), AMMISCA #4205, 5 Jan. 1944; CM-IN-8970 (14 Jan. 44), New Delhi, W113, 13 Jan. 1944.

89. Memo for CG ATC from Arnold, 7 Feb. 1944.

90.CM-IN-9483 (14 Feb. 44), AMMDEL #940, 13 Feb. 1944.

91. XX BC, Third Phase, pp. 73-89.

92. AAFRH-12, p. 192.

93. CM-OUT-390 (1 Mar. 44), AQUILA #4824, 1 Mar. 1944.

94. Hist. North African Wg. ATC, May 1944, pp. 21-22.

95. CM-OUT-6275 (15 Mar. 44), AQUILA #5021, 13 Mar. 1944.

96. Hist. NAW ATC, June 1944, p. 34.

97. AAFRH-12, p. 194.

98. Memo for CG North African Div. from Gen. George, 14 July 1944.

99. AAFRH-12, pp. 194-5.

100. Ibid., p. 196.

101. Report, as in n 78 above, p. 31.

102. See the long list of cables in AAFRH-12, p. 197, n 55.

103. Reported from Chungking in N.Y. Times, 14 Feb. 1944, by Brooks Atkinson.

104. CM-OUT-3033 (8Mar. 44), Arnold to Spaatz, FGZO, 8 Mar. 1944.

105. CM-IN-8596 (12 Mar. 44), Cook to Arnold, 5101, 12 Mar. 1944.

106. XX BC, Third Phase, p. 96.

107. CM-OUT-129 (1 Mar. 44), AQUILA #4812, 1 Mar. 1944.

108. CM-OUT-4343 (10 Mar. 44), AQUILA #4964, 10 Mar. 1944.

109. See the circumstantial account of the ceremonies in XX BC, Third Phase, pp. 89-94.

110. Ibid., pp. 95-101.

111. Daily B-29 Arrival and Accident Rpt., 9 May 1944.

112. OC&R Daily Rpt., 8 Mar. 1945.

113. Digests of Domei News Service broadcasts of 23 and 24 April and 15 May in XX BC Air Intelligence Digest, Vol. I (20 May 1944) and Vol. II (27 May 1944).

114. Hist. NAW ATC, Apr., May, June 1944; CM-IN-9888 (14 Mar. 44), Wolfe to Arnold for ATC, #162D, 14 Mar. 1944.

115. AAFRH-12, p. 200.

116. XX BC, Third Phase, p. 101.

117. Ltr., Arnold to Wolfe, 26 Apr. 1944.

118. A document, marked simply "Plan," printed out by hand and dated at 14th AF Hq., Chungking, 3 Sept. 1943. It is initialed by Chennault, Beebe, and Harmon, and indorsed by Stratemeyer on 10 Sept. 1943.

119. CCS 428 (Revised); memo for Marshall from Giles, 31 Dec. 1943.

120. Plans, Daily Digest, 1 Jan. 1944; CM-IN-2139 (4 Jan. 44), Stratemeyer to Arnold, W25, 4 Jan. 1944.

121. Memo for Hansell from Loutzenheiser, 17 Jan. 1944; memo for Capt. A. K. Doyle, USN, from Hansell, 18 Jan. 1944.

122. Med. Hist. 33d Ftr. Gp. (an unusually well-written unit history) ; 81st Ftr. Gp. War Diary, Mar. 1944.

123. Same two sources.

124. 14th AF GO 17, 13 Mar. 1944; CM-OUT-3952 (10 Dec. 43), Arnold to AQUILA, #3952, 10 Dec. 1943; CM-IN7131 (11 Dec. 43), Stratemeyer to Arnold, W2878, 11 Dec. 1943.

125. AAFRH-12, p. 204, especially ltr., Maj. G. A. Stinson to Col. George Carey, quoted in n 91.

126. CM-IN-1650 (24 Feb. 44), Chennault to Arnold, COGUK W22FB, 24 Feb. 1944; CM-IN-17774 (25 Feb. 44), Stratemeyer to Arnold, AQUILA #W559, 25 Feb. 1944.

127. CM-IN-3289 (5 Mar. 44), Stratemeyer to Arnold, W641 AOX, 5 Mar. 1944.

128. CM-IN-5347 (8 Mar. 44), Stratemeyer to Arnold, W678, 8 Mar. 1944.

129. Med. Hist. 33d Ftr. Gp.; 81st Ftr. Gp. War Diary, May, June, July, 1944.

130. Col. W. P. Fisher, Asst. A-3 58th 766 Bomb. Wg., had been sent out to command the group; XX BC, Third Phase, p. 47.

131. CM-IN-12153 (19 Dec. 43), AQUILA #W2944, 19 Dec. 1943; ltr., Stratemeyer to Arnold, 3 Feb. 1944; CM-IN-3499 (5 Feb. 44), Stratemeyer to Marshall.

132. CM-IN-4866 (7 Feb. 44), Stratemeyer to Arnold, W362, 7 Feb. 1944; CM-IN-954 (2 Feb. 44), AMMDEL AG 666, 1 Feb. 1944; memo for Kuter from Loutzenheiser, 2 Feb. 1944.

133. CM-IN-11476 (16 Apr. 44), CAB 16169, 15 Apr. 1944. Note the discrepancies in figures on the Chengtu deliveries: ATC, 383 tons; XX BC Stat. Sec., 427 (Wolfe, XX BC Prog. Rpt. #4, 3 Apr. 1944); ASC CBI, 116 tons (CM-IN-11476 [16 Apr. 44], Stilwell to Marshall, CAB 16169, 15 Apr. 1944).

134. ICW ATC, Story of Spring Diversions, Feb. to June 1944; Hist. India-China Div. ATC, 1944,Vol. I, chap. 4.

135. CM-IN-11476 (16 Apr. 44), Stilwell to Marshall.

136. CM-IN-16339 (21 Apr. 44), Wolfe to Arnold, YB1733, 21 Apr. 1944; XX BC Prog. Rpt. #4, 30 Apr. 1944. See Chennault's version of this whole deal in Way of a Fighter, pp. 278-84.

137. Ltr., Maj. G. A. Stinson to Lt. Col. J. D. Garcia, 22 Dec. 1944; CM-IN-15468 (24 Dec. 43), AQUILA #W2988, 24 Dec. 1944.

138. CM-IN-11079 (15 Apr. 41, Wolfe to Arnold, T29A, 15 Apr. 1944.

139. CM-IN-16401 (23 Mar. 44), Stratemeyer to WAR, W836 RACX, 23 Mar. 1944; XX BC Memo 55-9, 15 Apr. 1944; CM-IN-20510 (28 Apr. 44), Wolfe to Arnold, YB2005, 27 Apr. 1944.

140. AAFRH-12, p. 213.

141. CM-IN-16339 (21 Apr. 44), Wolfe to Arnold, YB1733, 21 Apr. 1944; Prog. Rpt. #4, 30 Apr. 1944.

142. Prog. Rpt. #4.

143. AAFRH-12, pp. 215-17; memo for CG AAF from C. R. Smith, DC/S ATC, 15 May 1944; CM-OUT-36950 (15 May 44), Arnold to Stratemeyer, Wolfe, Hardin.

144. Ltr., CG AAF IBS to CG XX BC and CG ICW, 26 May 1944.

145. Hist. XX BC, Fourth Phase: The Transport Prospect, pp. 32-33 [herein-after cited Transport Project].

146. Ibid., p. 31.

147. Ibid., pp. 33-34; memo for AC/AS Plans from Lt. Col. J. W. Wilson, 26 Apr. 1944; Prog. Rpt. #4, p. 11.

148. CM-IN-90 (1 May 44), Wolfe to Echols, A-22135, 3 Apr. 1944; WD-TT-518 (3 June 44), Washington-Kharagpur.

149. AAFRH-12, p. 219.

150. CM-IN-11782 (16 May 44), XX BC to WD, YB2918, 16 May 1944.

151. Memo for Staff Secs. from CG XX BC, 26 May 1944.

152. CM-IN-2827 (4 June 44), Stilwell to JCS, 1148, 4 June 1944; CM-OUT-46820 (6 June 44), JCS to Stilwell, WARX 46820, 6 June 1944; CM-IN-5027 (7 June 44), Stilwell (to WD, CFBY 18238, 6 June 1944; CM-OUT-47296 (7 June 44), Stilwell to Marshall, CHC 1173, 9 June 1944.

153. CM-OUT-46999 (6 June 44), WD to Wolfe; CM-IN-5597 (7 June 44), Wolfe to WD, 4269A; CM-OUT-47759 (8 June 44), WD to Wolfe.

154. Chennault, Way of a Fighter, p. 278.

155. Transport Project, p. 39.

156. CM-1N-1319 (2 July 44), Chennault to WD, CADX 4378, 2 July 1944.

157. Transport Project, p. 38.

158. Ibid., p. 41.

159. CM-OUT-51560 (16 June 44), Hansell to Marshall and Arnold at London.

160. Transport Project, p. 40.

161. CM-OUT-46999 (6 June 44), WD to Wolfe; T-CON-OUT (no number), Washington-Kharagpur, 23 June 1944; T-CON-OUT (no number), Washington-Kharagpur, 10 June and 21 June 1944; memo for CG ATC from Arnold, 10July 1944; T-CON-IN-YB231, Kharagpur-Washington, 9 Aug. 1944.

162. CM-OUT-56673 (27 June 44), WD to CG XX BC.

163. Transport Project, p. 41.

164. Ibid., p. 43.

165. CM-IN-21219 (26 June 44), Chennault to WD, CADX 4019, 25 June 1944.

166. Direc., CG AAF IBS to CG XX BC, 13 June 1944.

167. CM-IN-1319 (2 July 44), Chennault to WD, CADX 4378, 2 July 1944.

168. Transport Project, p. 43.

169. CM-IN-2546 (4 July 44), Stratemeyer to WD, CABX 2917, 3 July 1944; CM-IN-24632 (30 June 44), Stratemeyer to Chennault, CABX 2796, 30 June 1944; CM-IN-1914 (3 July 44), Chennault to WD, CADX, 2 July 1944.

170. Transport Project, p. 47.

171. AAFRH-12, p: 227.

172. Transport Project, pp. 48-57, and stat. charts attached thereto.

173. Ibid., pp. 80-83.

174. Ibid., pp. 68-70.

175. Walkout Rpts., XX BC; Lolo Rpt. by Capt. Frank Mullen.


Table of Contents
Previous Chapter (2) * * Next Chapter (4)


Transcribed and formatted for HTML by Terry Welshans for the HyperWar Foundation