John Stack had first conceived a high-speed research airplane in 1933, but his paper design had been merely an object for theoretical performance evaluation. He had wanted to explore how fast an imaginary airplane with all known favorable features could go when due allowance was made for the adverse effects of compressibility on drag and propeller performance.
With the coming of the compressibility crisis by 1940 and the growing recognition that there was some barrier preventing the acquisition of useful transonic data in existing wind tunnels, Stack began to campaign privately for NACA and military support for an actual airplane for high-speed research. By 1944, however, there were engineers, like Ezra Kotcher at Wright Field, and even some of Stack's colleagues at Langley, who had competing ideas for the requirements of a high-speed research aircraft. Though many of the particulars of Stack's research airplane concept would provide a solid foundation for the design of what became the Bell XS-1, the first plane to fly supersonically, some of its particulars would not be accepted and others would undergo major compromise.
Stack worked first on his contacts in the army. Citing the primary role the army had played in procuring the P-59 Airacomet, the first American turbojet plane, he pressed Col. Carl Greene and his assistant Jean Roche of the Air Materiel Command liaison office at Langley Field to persuade their superiors to develop a transonic research aircraft; within a few weeks a delegation from the Materiel Command, which included Ezra Kotcher, traveled from Wright Field to Langley to renew discussions with the NACA about the requirements of such an aircraft. At the first of two meetings in mid-May 1944, Kotcher reported the results of Wright Field's "Mach 0.999  study," the principal objective of which was to compare the theoretical performance of turbojet and rocket airplanes at high Mach numbers.1 He told Stack and his colleagues that the experience of the P-59 proved that turbojets could not propel aircraft to transonic speed, while the results of the Mach 0.999 study indicated that rockets could. Kotcher then showed the Langley engineers a rough drawing of a rocket plane Wright Field had in mind.
Stack responded to Kotcher's message by repeating a long-held NACA opinion: the application of rockets to airplanes was too unsafe. Stack knew that Melvin Gough, Langley's chief test pilot, had privately issued the edict, "No NACA pilot will ever be permitted to fly an airplane powered by a damned firecracker!" He let Kotcher know that the majority of Langley test pilots had opposed the idea of the transonic research airplane in the first place; they had felt that they were being asked to risk their lives because wind tunnel personnel were unable to do the necessary work on the ground. Now pilots were going to be asked not only to sit in the cockpit of a radically new airplane, atop a heavy load of explosive fuels, but also to rely on only a rocket to keep them aloft! No pilot in his right mind would want to fly this plane, Stack said. (It is not clear whether Stack or his associates knew anything yet about the experience of the Germans with the ME 163 rocket plane.) Furthermore, a rocket plane simply could not meet research needs as well as a turbojet. Because it could not stay in the air as long, it could not gather the kind and volume of systematic data that everyone required. Lastly, Stack argued that the performance of an experimental rocket aircraft surely would not be as applicable to the future development of aviation as that of the turbojet. At the end of this conference, however, it was agreed that the NACA would continue its separate study for the design of a transonic airplane and, upon completion, transmit a report about it to the army for comment.2 Stack held back submitting his design to Wright Field until 10 July, and then it still incorporated a turbojet rather than a rocket engine. The purpose of his airplane as he conceived it was to collect transonic data (space was provided for 400 pounds of research instrumentation), rather than to fly supersonically.
At another round of meetings at Langley on 13 and 14 December 1944, army representatives-many of whom Kotcher had persuaded personally to support his idea of a rocket-propelled transonic airplane-rejected the NACA's proposal for a turbojet as too conservative.3 The Stack team had designed an airplane to fly in the speed range from Mach 0.8 to 1.0, with a typical high-speed dash velocity of Mach 0.85 (650 MPH); the army wanted a plane that could fly supersonically to about Mach 1.2 (800 MPH). This apparently irreconcilable difference of intent was resolved easily: the army....
...."was putting up the money and they decided to do it their way." 4 One week after the meetings at Langley, the army started negotiations with the Bell Aircraft Company to procure a rocket plane. Bell immediately called together a design team headed by Robert Stanley, a California Institute of Technology aeronautical engineering graduate who had been the pilot of the first American turbojet, the XP-59. Under project designation MX-524, Stanley's team began development of the "Experimental Sonic-l" aircraft, or "XS-1" for short.5
Stack did not give up the idea of procuring the kind of transonic research airplane he wanted. In fact, as soon as Kotcher made it clear to him in the summer of 1944 that the army was going to insist on a rocket plane, he had contacted the navy. He wrote letters and telephoned various friends and acquaintances in the Bureau of Aeronautics, telling them that the rocket plane the army was procuring would probably not survive many flights. With the help of George Lewis and Capt. Walter Diehl (Lewis's good friend), Stack arranged to detail his engineer Milton Davidson to  Washington to work with personnel in BuAer's aviation design research branch on specifications for a transonic research aircraft.6
Since the navy had done very little in the way of research airplane studies up to this time, it was more ready than the army had been to accept the NACA's advice and general guidelines. In September 1944, a BuAer engineer (Abraham Hyatt, a Marine Corps officer and an aeronautical engineering graduate of the Georgia Institute of Technology) formally proposed that the navy procure a high-speed research airplane capable of meeting both military and NACA research requirements. Though the navy blueprint proposal called for some details different from those already set out by the Stack team (for example, side inlets instead of a nose intake, so as to free the nose for an armament installation), it basically matched the NACA's conservative design: the plane would take its power from a turbojet, not rocket, engine; the plane would take off from the ground and land under its own power; the plane would have good enough low-speed handling characteristics that data gathered from its flight test program could be applied directly to the design of future navy aircraft; and, finally, the plane would have a maximum velocity not exceeding the speed of sound. Together, these details would make an airplane far different from the XS-1 being planned by the army.
In late December 1944 Davidson informed Langley that the navy had taken the first step toward procurement of this airplane: BuAer had shown a representative of the Douglas Aircraft Corporation, one of the prime contractors for naval aircraft, a preliminary specification of the proposed experimental plane and asked him whether Douglas would be interested in working on it. Apparently the representative had immediately taken the offer back to his company's main office in California, Davidson reported. The report was accurate. By the first weeks of 1945, Douglas engineers were busy considering the design criteria for what would become "Douglas Model 558, High-Speed Test Airplane," the "D-558" for short. BuAer made it clear to its contractor from the start that the navy "was only interested in obtaining an airplane which met with the full approval of the NACA."7
Thus by early 1945 the development of two different transonic research airplanes was under way in the United States: the rocket-powered XS-1, being built by Bell under Army Air Forces sponsorship, and the turbojet-powered D-558 being built by Douglas under navy sponsorship. Though researchers at Langley would actively assist in the development and flight-testing of both airplanes, they would have reason to prefer helping with the D-558. It was most like the research airplane they wanted.
As soon as the Army Air Forces decided to procure an experimental rocket-powered aircraft, Langley researchers helped Bell engineers to determine the vehicle's basic design criteria. In December 1944, they estimated the instrument requirements for the XS-1: 370 pounds of instruments and 130 pounds of auxiliary equipment (wiring and tubing), all to fit within a space of nine cubic feet.8 This estimate would form the basis for the package of instruments eventually installed in both the XS-1 and D-558-1. In January 1945, they finished calculation of the load requirements of the airplane: a load factor of 18g, or 50 percent higher than the usual load factor of fighter aircraft. (With a load factor of 18g, the aircraft could accept the stresses and strains of aerodynamic forces equivalent to 18 times its own weight.) Stack suggested this figure because he wanted a wide margin of safety for the plane's first flights.9
One of Langley's most important recommendations for Bell's design of the XS-1 was its call for a thin wing section to minimize the buffeting,  loss of lift, and control problems that the experimental aircraft could probably experience at supercritical speeds. Langley thought long and hard before making this recommendation, but not because its research staff lacked knowledge about the effects of wing thickness ratio on transonic performance. By early 1945 the staff knew from Hugh Dryden's earlier work at the Bureau of Standards, from preliminary data from Gilruth's wing-flow tests (described in the previous chapter), and from a recent report of their own high-speed airfoil group that "airfoils of large thickness ratio should not be used at high Mach numbers because of radical adverse changes in their characteristics at supercritical speeds." The shock-stall effects were just too severe.10 Langley engineers disagreed sharply, however, over whether Bell should deliberately design wings to throw the XS-1 most quickly into the troubling region of deep shock stall, from Mach 0.75 to 0.90. There were two schools of thought on this question at the lab, one led by John Stack and composed mostly of wind tunnel people, and the other led by Robert Gilruth and made up primarily of his fellow flight researchers.
Stack and his followers advocated a wing section of average (12 percent) thickness. They did so for reasons that Stack made clear in late 1944 in a handwritten note to himself in preparation for a conference with army personnel about its transonic airplane designs:
In sum, Stack wanted Bell to choose a thick wing because it would force the research airplane to encounter exactly those drastic flow changes occurring at critical Mach numbers that aerodynamicists were most interested in studying and correlating with wind tunnel results. The research benefits would be greater.
Gilruth and his followers strongly opposed Stack's point of view. They opposed it, not because as proprietors of the NACA's wing-flow method they possessed some knowledge that Stack and his wind tunnel engineers did not have about thin wings retaining their lift at transonic speeds, but  because they had a different concept of the airplane's safety requirements. Gilruth believed that Bell should design the XS-1, the first aircraft to penetrate deeply into the supercritical zone, with every feature it knew could contribute to the airplane's safe operation. "If you put a thick wing on it," Gilruth warned, "it's bound to have problems." On the other hand, if you put a thin wing on the XS-1 (he suggested using wings as thin as five percent thickness-chord ratio), not only would you have a safer airplane, but you might be able to fly through the speed of sound with it.12 Ironically, Gilruth's conservative concept of the safety requirement was leading him to consider the possibility of the XS-1 flying supersonically, while Stack's adventurous attitude toward that requirement was keeping the airplane he had in mind to speeds well below Mach 1.
Before the NACA could recommend to Bell a thickness ratio for the wings of their airplane, Langley management had to resolve this disagreement between the Stack and Gilruth groups. Resolutions of this sort were essential to the success of the lab, for it was an organization of people from many diverse disciplines. The assistant chief of research, Floyd Thompson, with nearly 20 years of broad experience in NACA flight testing and understanding of many different fields, had the responsibility of assessing the contradictory recommendations given to him by his specialists. Thompson talked at length with both groups of engineers, studied all the relevant data collected by them, and made his decision: Gilruth was right; the XS-1 needed to have a thin wing.
Stack pushed for a compromise: perhaps the research airplane could have two sets of wings, one not quite as thin as Gilruth wanted and the other not quite as thick as Stack wanted. Thompson and the rest of Langley management concluded that splitting the difference was a good idea. It was doubtful that Bell could fabricate a wing as thin as five percent with the desired over-strength load factor of 18g anyway. Between March and July 1945, the NACA decided to advise Bell to build two sets of wings, one eight percent thick and the other ten percent thick.13 Bell followed this advice. The company built the XS-1 to fly first with the thin wing, but later, in order to provide the data the wind tunnel people wanted, to fly with the somewhat thicker wing.*
Wing thickness seems to have been the only design criterion for the XS-1 about which any members of the Langley research staff seriously disagreed. Both Stack and Gilruth wanted Bell to design the airplane's horizontal tail using a thinner airfoil section than it used for the wings, for they knew that if the wing and the tail had the same section thickness, both surfaces would reach the critical Mach number at the same time. The simultaneous experience of high drag rise of the wing and other compressibility effects from the tail could easily cause the pilot to lose control of the plane and crash. Stack and Gilruth also insisted that Bell make the horizontal tail  surface all-moving-that is, make the entire horizontal stabilizer adjustable by the pilot in flight. They realized that at subsonic speeds a pilot could ordinarily retain control of his aircraft, if a problem arose, by moving the elevator on a fixed horizontal stabilizer. At transonic speeds, however, they feared that this type of control probably would not be possible. At Langley's suggestion, the NACA also advised Bell to put the adjustable stabilizer high on the vertical fin of its airplane. This position, the laboratory staff had said, would keep the control surface safely above the wing wake.14
In early 1945, there was a virtual consensus at Langley-and at the Army Air Materiel Command-on one other basic design feature of the XS-1: no one wanted the NACA to advise Bell to design the transonic research airplane with anything but a conventional straight wing. This was true even though one of the lab's best aerodynamicists had explored a "new" theory suggesting that an aircraft could penetrate the sound barrier more easily if its wings were swept backwards.
Robert T. Jones was an extraordinary aerodynamicist who made important contributions to NACA research without having completed a college education. As a boy in his hometown of Macon, Missouri, Jones read all the aviation magazines available on the local stand. His favorite was the journal Aviation, which carried technical articles by eminent aeronautical engineers and notices of forthcoming NACA technical reports. Jones ordered copies of many of the NACA reports from the Government Printing Office for ten cents each, and even received some free simply by writing NACA headquarters in Washington. He perplexed many of his high school English teachers by writing essays for them on aeronautical subjects.
Jones attended the University of Missouri for only two semesters before taking a job rigging wings on airplanes for a flying circus that gave aerial shows at county fairs across the Midwest. In 1929 he took a manufacturing job with the Nicholas-Beazley Airplane Company in Marshall, Missouri, helping to build its new Barling NB-3, a low-wing cantilever monoplane of metal construction (except for fabric covering). Then came the Depression, the collapse of the Nicholas-Beazley company, and a succession of different jobs, in various towns, broken up by periods of unemployment. In 1933 he got a job operating an elevator in a government building in Washington, D.C. At night he took classes in aeronautics at Catholic University taught by former Langley chief of aerodynamics Max M. Munk.
In 1934 the Public Works Administration opened up a number of temporary scientific positions in the federal government. On the recommendation of his hometown congressman, Jones secured a nine-month appointment at Langley laboratory. The NACA made him an "assistant scientific aide" and assigned him to the 7 x 10-Foot Wind Tunnel section, where he soon proved to have exceptional talents, particularly for addressing theoretical problems pertaining to airfoils and to aircraft stability and control. For the next two years Langley managed to keep Jones by arranging for a series of temporary and emergency reappointments. It could not promote him to even the lowest professional or engineering grade, however, because to rate that grade, civil service regulations said that an individual had to have a college degree. In 1936 the lab finally found a way to keep him permanently, and to pay him what he was worth: it gave him the next grade above the lowest professional grade-for which the academic requirement, though presumed, was not specifically mentioned. A few years later Jones became head of the stability analysis section.15
While John Stack worked to win the military services over to his idea of the transonic research airplane, Robert T. Jones was busily engaged in studying the aerodynamic configuration of guided missiles. By the end of  1944, Jones had finished designing an experimental air-to-air missile for the Army Air Forces (the JB-3 or NACA "Tiamat") and was in the midst of studying the potential of a proposed glide bomb having a low-aspect ratio delta (triangular) planform.16 This unconventional planform had been brought to Jones's attention in August 1944 during a meeting at Langley with Roger W. Griswold, president of Ludington-Griswold of Saybrook, Connecticut, a manufacturer of flying weapons. In 1942 Griswold's company had built a wind tunnel model of a dart-shaped missile conceived by Michael Gluhareff, a Russian émigré who was chief of design for the Vought-Sikorsky Aircraft Division of the United Aircraft Corporation; now, in 1944, Griswold was using the results of Vought-Sikorsky tunnel tests with the model to convince the AAF and the NACA that the new missile should be developed. At their Langley meeting, Griswold showed Jones data plots predicted for the Gluhareff model on the basis of Ludwig Prandtl's lifting-line theory, a mathematical theory involving a series of physical assumptions that made the problems of lift and drag accessible to analysis.17
Jones knew that Prandtl's 25-year-old theory of lift was applicable to bodies with high aspect ratio but that it did not work for bodies-like Gluhareff's dart-shaped missile-with low aspect ratio. Jones was intrigued by the prospect of the new missile, however, and, as soon as Griswold left Langley, he began to study its unconventional shape on the basis of a new theory of his own making. This theory, developed by Jones especially for the lifting characteristics of slender delta wings, resulted in formulas and analytical solutions that were very simple, and in some key respects similar to those derived for flow around airships in 1924 by Max Munk, his mentor at Catholic University, and to those derived for supersonic flow around projectiles and other slender bodies in 1938 by Hsue-Shen Tsien of Caltech's Guggenheim Aeronautical Laboratory. For the moment, though, Jones chose not to pursue publication of his theory. He thought the theory "so crude" that "nobody would be interested in it," especially since it was based on incompressible flow at very low subsonic speeds. He placed it in a drawer of his desk and temporarily forgot about it.18
One day early in 1945, while playing with the highly sophisticated mathematics of potential flows at supersonic speed,** it dawned on Jones that he was obtaining the same simple formulas with compressible flow equations as he had derived from his crude lifting theory for incompressible flow. He now recalled that Professor Tsien had reported finding that certain  slender projectiles exhibited no influence of compressibility when revolving at high speed. Jones immediately got his old paper out of his desk drawer and incorporated the compressible flow equations into it. To his growing wonderment, he discovered that for very slender wings there seemed to be no compressibility effect, no effect of Mach number.
Jones sought a physical explanation for the total lack of compressibility effects on the theoretical performance of slender wings. After performing a series of complicated calculations, he recognized that the physical explanation was related to the effect of sweepback on the lift of large-span wings. This effect, Jones remembered, had been noted by Munk in 1924 in a paper published by the NACA dealing with the stability of wings. In this paper, Munk had stated that in level flight, only the component of velocity normal (that is, perpendicular) to the planform's leading edge was "effective for the creation of lift.19 This statement by Munk-namely, that the air force on a wing depends on the normal component of velocity-was the first statement of the basic effect of sweepback made by anyone, and it was surely more than a coincidence that it was Jones, Munk's prize student, who now recalled it. Though Munk had made this statement for the purpose of comparing the relative effect of dihedral and sweepback on airplane stability in incompressible (low subsonic) flow-and thus not in connection with high Mach number effects-Jones now had good reason to suspect that Munk's principle could be incorporated meaningfully into his slender-wing theory. The result was a new theory that covered the entire sweep range from zero to 90 degrees, and was not limited just to very slender wings.
Jones guessed that his sweep theory would show that the effective Mach number would be much less than that of the flight Mach number even for moderately swept and thick wings. He did not realize how much less the effective Mach number could be until he tried sweeping the leading edge of a slender wing back behind the Mach cone, the idealized cone-shaped zone of disturbance that theoretically emanates from a body moving through the air (or any other fluid medium) at supersonic speed. The effective Mach number of the highly swept wing then appeared to be in the astonishing range of three to five times less than that of straight-wing planforms. The sweep smoothed out the sharply bending streamlines of supersonic flow that otherwise would have affected the wing adversely. This enabled a purely subsonic type of flow to exist on the wing's surface, a phenomenon which worked to eliminate the wave drag and compressibility shock of high-speed flight almost entirely. Jones now had a physical explanation for the missing compressibility effect shown by the mathematics of his theory.
At the time Jones did not know that Adolf Busemann, a German aerodynamicist who would come to work at Langley after World War II,....
....had introduced the idea of sweeping wings to diminish the wave drag at supersonic speeds ten years earlier, in a paper he presented at the Volta Congress on High-Speed Aeronautics in 1935.20 (Busemann had kept the wing ahead of the Mach cone, however, so that the cross-flow was still supersonic.) Jones's colleague and close friend, Eastman Jacobs, had attended the meeting in Italy but did not remember the "arrow-wing" concept-one of many highly theoretical ideas in Busemann's paper-as anything important. Neither did Theodore von Karman or Hugh Dryden, the only other American representatives at the Volta meeting.21
Jones discussed his sweep concept first with Langley's other theoreticians, and with supersonics expert Arthur Kantrowitz in particular; then he brought it to the attention of his division chief, Hartley Soulé. In mid-February 1945 he outlined his concept for Jean Roché, civilian liaison officer at Langley for the Air Materiel Command, and described it for Ezra Kotcher. (At the time Jones was working with Kotcher to help the army copy the German V-1 missile.) In his conversations with both Roché and Kotcher, Jones tried to make clear his belief that sweep benefits were  progressive-that is, that the adverse effects of compressibility were reduced as the sweep angle of the wing increased-and that these benefits were not limited to the very slender wings of his original theory. He advised the army engineers that wings designed for flight at supersonic speeds should be swept back to an angle that would assure that the component of velocity normal to the wing's leading edge was less than the critical speed of the airfoil sections. On 5 March 1945, he sent a memo to Gus Crowley, Langley's chief of research, announcing that he had
Jones asked Crowley to approve tests of experimental wing shapes "designed to minimize compressibility effects."22
Jones's articulation of his theory was still in raw form, however; he would not finish a formal report on his theory until late April.23 Then the report ran into trouble in Langley's in-house editorial committee. Theodore Theodorsen, head of the Physical Research Division, chaired this committee. Theodorsen had serious reservations about the publication of Jones's paper; he felt that parts of the presentation were too intuitive and asked that Jones clarify the "hocus-pocus" with some "real mathematics." More importantly, Theodorsen was sure that supersonic flow was so completely different in nature than subsonic flow that it was most unlikely to be accompanied by the subsonic flow that Jones predicted on a wing traveling at supersonic speeds. He called Jones's insight into the potential of swept wings "a snare and a delusion."24 At the end of his committee's deliberations, Theodorsen insisted that Jones take the part about sweep theory out of his paper.25
NACA management supported the judgment of Theodorsen and his editorial committee and withheld publication of Jones's report until the sweep theory was confirmed experimentally. 26 This confirmation did not take very long. Even before Jones had finished the first draft of his controversial report, Robert Gilruth's flight research section had started a series of wing-flow and drop-body tests to verify the favorable effects of sweepback on wing drag predicted by Jones. By the end of May 1945, results from these free-flight tests validated the swept-wing concept in convincing fashion: they showed a reduction of wing drag by a factor of almost four.27 Shortly thereafter, Macon C. Ellis and Clinton Brown verified this dramatic reduction of drag by testing a section of wire at a large angle of sweep in Langley's model supersonic tunnel.28
In early June, Langley transmitted Jones's report to NACA headquarters for publication. In the transmittal letter, engineer-in-charge Reid stated that "Dr. Theodore Theodorsen [still] does not agree with the arguments presented and the conclusions reached and accordingly declined to participate in editing the paper."29 On 21 June, the NACA issued Jones's report, "Wing Plan Forms For High-Speed Flight," as a Confidential Memorandum Report (CMR L5F21), part of a series the Committee's executive officers prepared chiefly for the information of the army or navy. Before the paper was published, Jones's colleague at Langley, Robert Hess, found an overlooked copy of Busemann's earlier paper (a British translation dated April 1942) in the LMAL library, and Jones included a reference to it.30 Three weeks later, the NACA reissued Jones's paper as an Advance Confidential Report (ACR L5G07), a type of publication the NACA sent not only to both services and to its own subcommittee members but also by registered mail to those members of the aircraft industry who had signed secrecy agreements with the services and who had a "need to know."31
 Verification of sweep theory and publication of Jones's report came too late for the XS-1. On 10 March 1945, the Army Air Forces had. notified the NACA that it was awarding Bell a contract to develop the rocket powered research airplane with straight wings. By this time, all, three parties involved had known something about Jones's theory that sweeping a wing would probably alleviate compressibility shock and generally improve performance, but they would not have changed their minds about the design in any case. There was no proven reason for them to recommend changing from conventional straight-wing planforms to swept wings, an almost completely unknown quantity. Flight test research with full-scale aircraft had to proceed cautiously and conservatively. They were doing enough bold things with the XS-1 as it was. Five days later, the Materiel Command at Wright Field had held the first design review of the XS-1. No one seems to have made any mention of sweep theory.32
Because John Stack was in Europe at the time, Langley had sent Stack's top assistant John V. Becker, who was head of the 16-Foot High-Speed Tunnel section, to the XS-1 design review as its representative. At Wright Field Becker found that Bell had accepted all of the NACA's ideas for the design of the airplane except for Stack's longstanding recommendation for a more conservative power plant (turbojet) and speed range (Mach 0.8 to 1.0). Because Bell seemed to be planning for the XS-1 to take off from the ground rather than to be launched from the air, Becker reported that the proposed design was acceptable to the NACA. In climbing by itself up to the Mach 1.2 supersonic cruising speed that the army specified, through the Mach 0.8 to 1.0-speed region the NACA most wanted to know about, such an aircraft would provide realistic data on a full range of flight considerations.33
Two months after the design review at Wright Field, however, Bell opted to change the research airplane to air launch: a specially configured B-29 would carry the XS-1 to an altitude of 30,000 feet and then release it for flight. Though there was disagreement among Bell engineers over the wisdom of this decision, the company made the change because, after technical deliberations, it saw no way for the airplane to achieve the supersonic speed required by the army if it had to take off and climb from the ground.34 The rocket engine would simply consume too much of the precious fuel allotment. This was true even though two-thirds of the gross weight of the airplane was to be in fuel, which would have been an extraordinarily high proportion for any nonrocket military aircraft.35
This change from ground takeoff to air launch further dampened Langley's enthusiasm for the XS-1. According to the lab's experts, air launching was a cumbersome method and the second major violation of the NACA's basic notion that a research airplane should operate as conventionally as possible (the first violation having been the use of rocket propulsion). Moreover, air launching also meant that in all probability the little rocket plane would never be operated out of Langley, a busy flying field close to highly populated areas-if the XS-1 came loose accidentally, without a pilot, from the B-29 in flight, the resulting crash could kill many people. At another field, the NACA would not be able to manage the program of flight tests for the XS-1 as directly as it wanted.36
Langley objected to the evolution of Bell's XS-1 from another standpoint besides the launch mode. Because Bell believed that the unavailability of the complex new rocket fuel pumps (then being developed by Reaction Motors of Pompton Lakes, New Jersey) called for by the original design would probably hold up flight tests of the transonic airplane, it decided in April 1945 to redesign the first XS-1 with pressurized fuel tanks of some  simpler type already existing. Though the company's design team realized that use of pressurized tanks instead of the new pumping units would reduce the duration of the airplane's maximum thrust by approximately 3 minutes, from 5.4 to approximately 2.6 minutes, and thus force a reduction in the plane's cruising altitude, it judged that "it would be better to have an airplane which would enable preliminary flights to be made at a reduced altitude, rather than to have an airplane on the ground awaiting a pumping unit. 37
John Stack reacted strongly when he heard about Bell's revised plans. In a memorandum to Langley's chief of research, he warned that the transonic airplane under development "may prove quite unsuitable." Stack noted that everyone had agreed at the initiation of the project that five items were the basic requirements of the research airplane:
These requirements had since been sacrificed to the point where the project was now
Although he agreed completely with Bell's view that it was best to get an airplane flying as quickly as possible, Stack wanted the NACA to remind everyone that the "basic purpose of all of this work," as he had originally conceived it, was to obtain in actual flight compressibility data that could not be acquired in wind tunnels in certain speed regions. Bell could not let a little thing like the present unavailability of the correct rocket fuel pump destroy the basic purpose of the entire project. Stack recommended that "a much larger effort be devoted to the development of this pump, an effort that is as large as the project demands." He urged that the army be asked to call in engineering organizations other than Reaction Motors to help develop the pump, if necessary.38
 For the rest of 1945 Langley did whatever the army or its contractor asked it to do to help complete the design of the XS-1 and ready it for test flights. It did these jobs well, as was expected of the NACA, even though the XS-1 was far from the research airplane that it wanted. The laboratory oversaw the design and preparations for installing the XS-1's instrument panel. By the end of the year, wind tunnel tests provided data reliable enough for the lab to predict the rocket plane's flying characteristics up to about Mach 0.90 for the low-angle-of-attack conditions which were of most significance for the XS-1 and D-558 flights. Wing-flow data took these NACA predictions up to about Mach 0.93.39
Army criticism of the NACA for not releasing Jones's sweepback theory earlier made giving this assistance a somewhat unpleasant task. In May 1945 a special intelligence unit of the U.S. Navy had discovered among the countless abandoned documents of the aerodynamical laboratory at Gottingen solid evidence that the Germans had been aggressively studying for some time the advantages of sweepback in designs of their jet-propelled aircraft.40 The army heard news of this startling finding at least as soon as did the NACA.41 Some of its leaders thought that here was another example, like the turbojet revolution, of the NACA failing to keep the United States on a par with Europe in aeronautical development.
In October 1945 Brig. Gen. Alden R. Crawford, chief of the Production Division of the AAF, asked Jerome Hunsaker, the NACA chairman, why there had been no mention of Jones's theory during the first XS-1 design review at Wright Field the preceding March or during follow-up visits of Air Materiel Command personnel to Langley later that spring. Applying 20/20 hindsight, Crawford indicated that the NACA might have announced its vital new information in time to change the design of the XS-1 from straight to swept wings. Because such a change at this time "must delay the project and increase the cost to the Government," Crawford lamented, now the only thing the Air Forces could do was contract with Bell for the development of new XS airplanes with swept wings (which it did in December 1945).
The NACA knew that its organization could not justly be held responsible for the XS-1's conventional wing planform; after all, the Materiel Command had made the decision for straight wings, not the NACA. Moreover, R. T. Jones had described his theory for both Jean Roche and Ezra Kotcher by the time of the first design review in March. Floyd Thompson (the LMAL assistant research chief who had arbitrated the original Stack Gilruth difference over XS-1 wing thickness) prepared for Hunsaker a polite but taciturn answer to General Crawford's letter. To have recommended changing the XS-1 in March 1945 from straight to swept wings could have been a "blunder of the greatest magnitude," Thompson wrote. "Not only  was experimental evidence lacking [especially about the low-speed characteristics of the swept wing] but our best theoretical minds were divided as to the validity of the theory."42
Douglas proposed to build for the navy six D-558 transonic research airplanes initially. Each aircraft would be powered by a General Electric TG-180 turbojet engine and equipped with alternative wing and nose duct configurations; maximum speed would be about Mach 0.90. In phase two of the program, Douglas would change two of the planes to Westinghouse 24C turbojets plus supplementary rocket propulsion units. These modified aircraft would gather aerodynamic data from Mach 0.89 to about Mach 1. In phase three, Douglas would use results acquired during phases one and two to construct a combat version of the D-558. Douglas estimated the total cost of the three-phase program to be just less than $7 million.43
This proposal was not what Douglas originally had in mind. In February 1945 company representatives had submitted a proposal to the navy for just one airplane. This airplane was to be designed around an available turbojet unit capable of delivering, with the help of supplementary rocket propulsion units, a maximum thrust of 3000 pounds for 40 seconds. It would reach Mach 0.9 in level flight and Mach 1.0 after a 25-degree dive from 35,000 feet.44 With only a few minor modifications, this airplane was to be adaptable as a navy fighter. Its development could thus lead to volume production and considerable profits to the contractor. In most essentials, it was the same plane Douglas later proposed for development during phase one.
The NACA had objected to Douglas's original proposal for this research airplane in very strong terms. Its spokesmen argued that a true research airplane should not be compromised for military or volume production requirements. In meetings with navy officials, they called Douglas's idea for a research airplane "wholly inadequate," "a half-way measure" that would result in an airplane, which would "be obsolete by the time it was built." Milton Davidson, John Stack's colleague on special assignment to BuAer's airplane design research branch, reported to his superiors that he had fully outlined the NACA transonic research airplane specifications during meetings with Douglas representatives in early February. What the NACA desired, Davidson said he had explained, was an airplane that would "take off, climb to operational altitude [20,000-foot minimum, 35,000-foot maximum], operate for 10 minutes at a velocity near the critical speed at altitude, have a 2-minute burst at maximum thrust, and return to  the airport with power." The airplane Douglas suggested building would be deficient in duration and amount of power and fuel to meet these requirements. Davidson also indicated that he had made clear to the representatives that the plane had to have adequate space for a sizable package of NACA recording instruments; the airplane Douglas proposed did not have enough such space. On many occasions since the first round of meetings at BuAer, he had gone over their preliminary engineering sketches of high-speed research airplanes, enumerating the changes necessary to make the D-558 satisfactory. Apparently Douglas had chosen to ignore NACA advice, Davidson concluded. 45
The navy had supported the NACA's objections to the original Douglas proposal. Captain Diehl told Douglas representatives at a meeting at BuAer on 23 February that he thought "the NACA had spent a good deal of time studying the problem, and since the NACA was in the best position to know what was wanted in a research airplane," Douglas's airplane proposal should "measure up to NACA specifications. " Four days later, Comdr. Emerson W. Conlon, head of BuAer's structures department, opened another meeting with Douglas representatives by stating that the NACA would have to "heartily approve of any airplane" before its procurement by the navy. 46 This double-barreled NACA-navy criticism quickly led Douglas to the decision to commit itself to a new design proposal, the three-phase plan that the company ultimately submitted in April.
BuAer quickly approved Douglas's preliminary designs for the phase one and phase two aircraft and outlined a development program that guaranteed, among other things, the NACA's role in the management of the flight tests and immediate access to at least one of the airplanes: Douglas test pilots would fly the D-558-1 to acquire data applicable to the design of combat aircraft, and NACA test pilots would fly it to gather fundamental aerodynamic information about air loads, stability and control, flutter, and engine performance at high Mach numbers.47 Although John Stack, in particular, had some serious reservations about the adequacy of Douglas's phase one aircraft, especially in comparison with the proposed Bell XS-1 and the German ME 163, he soon became satisfied that the phase two program would result in the transonic research airplane he wanted. 48
Douglas held the first mock-up inspection of the D-558 at its main office in El Segundo, California, from 2 to 4 July 1945. As its representatives, the NACA sent Stack, Thompson, and Gough from Langley, as well as Milton Ames, a technical assistant assigned to NACA headquarters, and H. Julian "Harvey" Allen from Ames laboratory. The five NACA representatives made all of the various sources of their dissatisfaction with the Douglas design known during the first day of the inspection. Among other things,  they suggested that Douglas needed to increase the size of the space allotted for NACA recording instruments, generally enlarge the fuselage, change the design of the cockpit canopy and the side inlets, and improve the ducting of the nose inlet. Douglas concurred immediately. Following the ideas agreed upon by everyone during the technical meetings of the first day, the company prepared new drawings of the mock-up as modified. During the last day of the inspection, the NACA delegation got Douglas and navy spokesmen to agree to support the NACA's development of an afterburner unit at its engine research lab in Cleveland. The application of this afterburner, they argued, could probably provide the phase one research airplane with enough additional thrust to permit flight "at extremely great supercritical speeds." Langley and NACA headquarters representatives flew home to the east coast satisfied that they had finally gotten Douglas to commit itself to making the extensive changes that were necessary to make the D-558 into an adequate transonic research airplane.49 At a second D-558 mock-up inspection held 14 to 17 August 1945, NACA representatives found that Douglas had indeed made the canopy and inlet changes in accordance with the requirements they had outlined at the meeting five weeks earlier.50
Even before the first mock-up inspection in early July, John Stack had talked to Captain Diehl about Langley's experimental confirmation of R. T. Jones's sweep theory. The head of Langley's Compressibility Research Division thought it might be wise, considering recent developments, for the navy to ask Douglas to incorporate a 35-degree swept wing on one of its D-558s. Both Stack and Diehl realized that swept wings for the phase one airplane made no sense; it could not be powered by the proposed power plant to a high enough Mach number for the performance of swept wings to be fully evaluated. They also knew that the navy would want Douglas to proceed cautiously, with straight-wing configurations, until there was absolutely no doubt in the minds of the experts that sweep was the best way to go when designing an airplane wing for high-speed flight. They agreed, however, that wind tunnel evaluation of swept wings should be included immediately in the D-558 program for possible incorporation into the design of the phase two aircraft. Soon after the first inspection at El Segundo, the Langley High-Speed Panel, which Stack chaired, asked NACA headquarters to arrange for permission to incorporate swept wings on the model of the D-558 that was to be tested in the lab's 8-Foot High-Speed Tunnel. This request was supported at a joint army-navy-NACA research meeting at the NACA's Washington office on 13 July 1945.51
In early August couriers arrived at the Bureau of Aeronautics in Washington and at the Douglas company in El Segundo with microfilm of the captured German swept-wing reports. BuAer shared and analyzed....
 .....the new information with the NACA almost immediately. Their common evaluation of this microfilm led to a joint request at the second mock-up inspection for Douglas to initiate a study for the design of a D-558 with swept wings. The two men responsible for capturing and microfilming the papers at Göttingen for the navy (L. Eugene Root and A. M. 0. Smith) were Douglas employees. The company embraced the navy-NACA request for a design study of a swept-wing D-558 with both turbojet and rocket power for development during phase two and gave the NACA the job of specifying many of the design requirements, including complete responsibility for high-speed wind tunnel testing.52
While Langley tried to use its influence to get the navy and its contractor to accelerate the development of the swept-wing phase two aircraft, the laboratory staff continued to aid development of the straight-wing D-558-1 in every way it could. Stack encouraged the dozens of engineers, scientists, technicians, and mechanics involved in carrying out Langley's comprehensive high-speed research program to extend themselves in every way to meet the needs of the D-558 project quickly and successfully.53 This group included the staff of his own 8-Foot HST section, who were kept busy testing the aerodynamic characteristics of scale models of the D-558 configuration. It also included many personnel of the Flight Research Division, who were using the wing-flow method to test D-558 models mounted on the wings of a P-51 Mustang, and most of the personnel of the Spin Tunnel section, who had modified a scale model of the Bell XS-1 to simulate the spin behavior of the D-558.
Throughout the early development of the Bell XS-1 and the Douglas D-558, Langley engineers displayed their long-term preference for airbreathing propulsion units over rockets. This preference can also be seen in a project designed by a team of engineers in the lab's 9-Inch Supersonic Tunnel section to study the feasibility of powering a small airplane to Mach 1.4 with a ramjet engine.
By the spring of 1944 the Campini jet propulsion system had disappeared from Langley's list of research interests, even though the system's champion, Eastman Jacobs, seems never to have formally acknowledged that it was unworthy of additional consideration. During the summer, Jacobs moved to the Aircraft Engine Research Laboratory in Cleveland after the NACA dissolved the Air-Flow Research Division and made him, its former chief, a "consulting engineer." 54 (He would remain at AERL for only a short time before retiring from government service to do independent....
.....consulting work in California.) Before leaving Langley, however, Jacobs had encouraged the two-man staff of the 9-Inch Supersonic Tunnel section Clinton Brown and Macon C. Ellis, Jr.-to investigate the potential of a new ramjet propulsion unit. Another simple type of jet engine, this unit consisted. of a specially shaped tube or duct open at both ends. It required no mechanical compressor. The forward motion of the engine shoved all the air necessary for combustion into the duct and compressed it. In the engine, the compressed air passed through a specially designed chamber, or diffuser, and mixed with fuel; together the fuel and air burned rapidly. Exhaust gases then issued as a propulsive jet from a rear opening.55
In December 1945 Ellis and Brown finished a report, which showed to the NACA's satisfaction that a small ramjet research aircraft was feasible. Accelerating through the transonic region would require rocket boosters, but once the airplane flew to the speed of sound the ramjet could take over and power it for a short distance (about 60 miles) at a supersonic speed of Mach 1.4. Ellis and Brown envisioned either airplane tow to altitude or air launch as the ramjet's takeoff mode-a plan not without a certain irony, given Langley's opposition earlier in the same year to the plan to air launch the XS-1, and given Langley's overall commitment to developing research airplanes that could operate as conventionally as possible.56
 Langley managers, at John Stack's instigation, briefly considered advising one or both of the military services to add the ramjet aircraft to the fleet of transonic research airplanes under development, but there was no ramjet engine under development at the time: the engine Ellis and Brown had assessed was hypothetical. This meant that the ramjet proposal "had virtually no chance of support" outside the NACA, especially with the designs of the XS-1 and D-558 now well under way. Stack, still strongly committed to the idea of operating a research airplane as conventionally as possible and apparently satisfied with the direction of the D-558 program, let the supersonic ramjet aircraft proposal die at home in conference.57
Bell completed construction of the first XS-1, without the rocket motor, in December 1945-the month of the Ellis-Brown ramjet proposal. Under terms of its contract with the Army Air Forces, the company now had to test the XS-1 to the speed of Mach 0.8 before official acceptance. The AAF and the NACA had determined even before the delay in completing the rocket motor that Bell pilots should first fly the new airplane through a series of glide tests. These glide tests would identify quirks in the air launch method and address the feasibility of operating the rocket plane from conventional flying fields (like Langley) near population centers. In November 1945 the army had selected isolated Pinecastle Field in central Florida as the site of the glide tests. It was the NACA's understanding that these preliminary flight tests scheduled for Pinecastle with the unpowered airplane were to determine the feasibility and safety of operation from Langley Field.58
The NACA sent two Langley engineers, Walter C. Williams and Gerald M. Truszynski, to Pinecastle to join Bell test pilot Jack Woolams and the B-29 launch crew for glide tests of the XS-1. Williams, a 1939 graduate in aeronautical engineering from Louisiana State University, had worked with Stack as early as 1942 on research aircraft studies. Recently, as a member of the flight research section, he had been responsible for advising NACA pilots about how far to push the P-51 in dive tests for transonic wing-flow data. At Pinecastle Williams monitored flight test preparations and supervised on-the-spot analysis of the resulting glide path data. Truszynski, a 1944 graduate in electrical engineering from Rutgers, had been designing radar and telemetry equipment in Langley's Instrument Research Division. At Pinecastle he took charge of the radar tracking equipment.59
The XS-1 glide test program at Pinecastle lasted about three months, from January through March 1946, while Bell readied the second XS-1 for powered-flight trials. The aircraft showed itself to be aerodynamically  sound, with good low-speed handling qualities, and the air launch method proved practicable, but problems landing the XS-1 safely at Pinecastle demonstrated the inadequacies of a conventional airfield for operating the plane.60 These last two test results erased the NACA's vestige of hope that some method of ground launching would be found for the XS-1, making it possible for the aircraft to fly from Langley Field.
During the final days of glide testing at Pinecastle, the AAF chose its flight base at Muroc Dry Lake in southern California as the site where the powered tests of the XS-1 would be made. In the opinion of the army, Muroc Dry Lake was the best possible location for several reasons: a flight test base was already there, complete with facilities and a contingent of military personnel; America's first turbojet aircraft, the Bell XP-59A, had flown for the first time at Muroc; the weather was usually excellent; an enormous stretch of desert and dry lake provided more than adequate space for emergency landings; and the remoteness of the site removed the worry and danger of overflying and crashing into populated areas.61
The NACA endorsed the AAF choice for the XS-1 test site and, in late September 1946, detailed a group of 13 engineers, instrument specialists, and technical observers from Langley laboratory to Muroc on temporary assignment.62 Hartley Soulé, chief of the Stability Research Division and project manager for the research aircraft program at Langley, designated this group the "NACA Muroc Flight Test Unit" with Walter Williams, veteran project engineer for the XS-1 glide tests at Pinecastle, unit leader. Williams, who reported to Soulé, was authorized by Langley's engineer-in-charge "to make all necessary contacts and decisions for the NACA at Muroc."63 The assignment given to this special unit was to supervise the complete instrumentation of the second XS-1, gather and analyze all possible data during the period of its powered test flights, and more generally to try to make sure that NACA research interests were considered in planning and carrying out the in-flight program.
From the beginning, different people had different purposes in mind for the XS-1. Stack wanted the aircraft to collect as systematically as possible the detailed transonic data unobtainable in the wind tunnel, whereas Gilruth and Thompson, more in line with the thinking of the AAF, wanted to design a good high-speed aircraft and to get that aircraft to fly supersonically as quickly as possible so that it could serve as a prototype of an operational supersonic aircraft. Both the AAF and the NACA had recognized early in the XS-1 development period that these purposes, and the methods for achieving them, were contrasting and in certain ways even contradictory, but they had agreed to coordinate their plans so that the research aircraft could be built and flown for their mutual purposes.
When Langley detailed the special flight test unit to Muroc in September 1946, it was seriously concerned about Bell's intention to make the acceptance tests of the XS-1 airplane in as short a time as possible. Though the lab recognized that Bell's test program lived up to the legal requirements of the army-Bell contract for the XS-1, it worried that the flight tests required of the company would cover only demonstration of the "limiting conditions." "The mere flying of the airplane to a Mach number of 0.8 and making an 8g pull-out is not considered suitable preparation for the research flying," Langley emphasized. The program its staff had outlined for the acceptance tests of the XS-1 included "systematic exploration of the stability and control characteristics and structural loading at successively higher speeds up to a Mach number of 0.8." The lab had based its program on the understanding that "before asking anyone to proceed with the extremely hazardous flying above a Mach number of 0.8 everything would be done to make certain that the airplane was satisfactory in all aspects in the  speed range up to Mach 0.8." The acceptance-test program was thus the NACA's means of assuring itself that the airplane's subcritical characteristics were satisfactory.
Since the likely level of such assurance seemed too low, Langley informed NACA headquarters that it did not want its pilots to undertake research flying in the XS-1 following the limited acceptance tests at Muroc proposed by Bell. It recommended that the army shift part of the flight test program originally included in the acceptance phase of the contractor program to the NACA research program phase. That way Bell could receive its payment for the airplane as quickly as it wanted without lowering the safety and overall value of the research airplane program.64
Bell began flying the XS-1 number two at Muroc on 11 October 1946. Two months later, on 9 December, Bell test pilot Chalmers H. "Slick" Goodlin flew the airplane with its rocket power successfully engaged for the first time. (The company had selected Goodlin to fly the plane in September after Jack Woolams was killed in the crash of a P-39 Airacobra that had been modified to compete in air races.) On 8 January 1947, during a buffet-boundary investigation, Goodlin reached Mach 0.8 at 35,000 feet, the speed and corresponding altitude required by the contract before the AAF would accept delivery of the aircraft. Three months later Bell began flying the XS-1 number one, which had been out of action since its last glide flight at Pinecastle in March 1946 in order to have a rocket engine installed. In mid-May Bell successfully put number two through a required 8g pullouts and final airspeed calibration flight. After a total of 21 powered flights (14 by number two and 7 by number one), the contractor program was complete; both the AAF and the NACA were satisfied that the experimental airplanes were airworthy. Now the XS-1s belonged to the military. It was up to AAF flight engineers and test pilots to "break the sound barrier" and to do it in as few flights as possible.65
Concurrently with the beginning of the AAF's accelerated transonic flight program, the NACA got ready to conduct its own series of flight tests with the XS-1. The AAF had agreed informally early in the development program to lend the NACA a finished XS-1 for a separate series of flight tests. According to the terms of the agreement-which was completed at an NACA-AAF conference at Wright Field on 30 June 1947-the NACA would use XS-1 number two. It would furnish fuel, maintenance, and a flight crew for the experimental airplane, while the army would furnish the same for the launch B-29.66 In March 1947 the NACA Muroc Flight Test Unit had prepared a more complex instrument package for installation in number two, as was necessary for making a thorough examination of the airplane's flying characteristics and loads. In late May, after the last flight in Bell's .....
....contractor program, Walter Williams wrote Melvin Gough at Langley that "we want to fly it [what Gough two years earlier had called 'that damned firecracker that no NACA pilot will ever be permitted to fly'] at the earliest possible date because everyone is quite anxious to get going."67
Williams's special unit could not fly the XS-1, however, until the NACA received the number two plane and got it ready-and for a time in the long hot desert summer of 1947 it seemed that neither thing would happen very soon. In early June, the airplane the NACA was going to get was damaged seriously in a freak on-the-ground accident and had to be ferried back via B-29 to Bell's hangar in Buffalo, New York, for repairs. When number two returned to Muroc in July, progress on it was slow because of the intense level of activity on number one. The preparation of the army plane required so many of the mechanical crew that there were usually none left for the NACA plane.
Gough's prediction of 1945 was coming back to haunt NACA personnel. In August 1947, while World War II combat ace Capt. Charles E. "Chuck" Yeager took up number one for more glide and the first accelerated power flights, the two NACA pilots on the scene-Herbert Hoover from  Langley and Howard Lilly from the Aircraft Engine Research Laboratory in Cleveland-had to be satisfied with taking number two through a series of ground runs. Through the first three weeks of September Walter Williams tried "stalling the Army off as much as possible until [the NACA could] get the NACA tests underway."68 When the NACA airplane was finally ready for its acceptance test on 25 September, the NACA pilots were not. Since NACA management had thought it imprudent for its pilots to take up number one on pilot familiarization flights-and thus risk doing any damage to it-neither Hoover nor Lilly had gotten checked out in the XS-1. Thus the task of flying number two through its NACA acceptance test fell to Captain Yeager.
In a preflight planning session on the morning of 14 October 1947, the NACA advised Yeager to take the rocket plane on its ninth powered flight to a maximum speed of Mach 0.97. Walter Williams and De Elroy Beeler emphasized for Yeager's sake that it would be unwise to go any faster until a complete examination of the data obtained from the previous flights was completed. They warned him to exceed Mach 0.97 only if absolutely certain that it was safe to do so.69 Yeager ordinarily did not like NACA "eggheads" trying to "dictate" the planned speed of his flight-he recalls attending "highly technical NACA preflight planning sessions and post flight briefings" and not knowing "what in the hell" Walter Williams was talking about. After NACA briefings Yeager usually sat down with fellow army pilot Jack Ridley to "decide whether or not we wanted to stick with [the NACA] recommendation." Invariably they determined to fly faster than the NACA engineers wanted them to. (Yeager has written that the NACA was "so conservative that it would've taken [him] six months to get to the barrier" if he had followed the NACA's instructions exactly.) The way he felt that morning, though, hurting from a broken rib suffered two days earlier in a fall from horseback, a speed of Mach 0.97 was at that moment all he thought he would care to handle.70
At about 10:00 A.M. Yeager got into the launch B-29, the rocket plane shackled in its bomb bay, for the approximately 20-minute climb to altitude. At 5000 feet Yeager climbed down the transfer ladder into the tiny cockpit of the XS-1. At approximately 20,000 feet, NACA radar cleared the B-29 to let loose the XS-1. Sixty seconds later, at 10:26 A.M., Yeager's plane dropped free. What followed was the first manned supersonic flight in history.
Though Langley laboratory got word of Yeager's achievement immediately, it did not find out the details of the sensational flight until it received  a letter from Walter Williams more than a week later. Williams described the flight in measured, technical language as part of his regular bimonthly report to Gus Crowley, Langley's chief of research:
There was nothing in the tone of Williams's letter to suggest the fears and inhibitions that had been blocking the work of aeronautical researchers and aircraft designers since 1935 when Hilton inadvertently coined the term, and the concept, of the "sound barrier." Williams made not even an oblique reference to the concept of the "sound barrier" in his letter. In the public mind, however, news of Yeager's flight-once it was finally announced some weeks later-meant only that the awesome sonic barrier had finally and miraculously been pierced.***
The day before Langley received this report from Williams, one of its own test pilots, Herbert Hoover, had initiated the NACA program on the XS-1 with a familiarization glide flight in the number two airplane. On landing, however, Hoover misjudged the height of the airplane and made several contacts with the ground, the last of which caused the nose wheel to collapse, before skidding to a stop on the desert runway with damage to the landing strut.72 Repairs and bad weather kept the NACA airplane grounded for the next seven weeks.
 During this interim from October to December 1947, test pilot Howard Lilly made the first two NACA test flights of a D-558-1, the navy airplane built by Douglas and so much favored by John Stack over the Bell XS-1. D-558-1 number one had arrived at Muroc with a company test team for the contractor program in April; at the end of summer, D-558-1 number two, the aircraft planned for extensive NACA service, arrived at the California site. It was number two that Lilly flew in November 1947. The NACA's systematic flight tests of the XS-1 number two began on 16 December 1947 when Herbert Hoover became the first NACA pilot to fly a rocket plane. He reached Mach 0.71. By the end of January 1948, Hoover had made six more powered flights in the XS-1, working the speed up to Mach 0.925, and Howard Lilly had checked out in the plane. On 10 March, Hoover achieved the NACA's first supersonic flight. Three weeks later, Lilly repeated Hoover's supersonic performance.
Between the time of these first civilian supersonic flights in early March 1947 and the time of the NACA's replacement by the National Aeronautics and Space Administration in the summer of 1958, the NACA made in the neighborhood of 100 research flights either in the XS-1 number two or in one of its sister ships in the X-series. In the same period, the NACA made nearly 300 flights in the D-558-1 or D-558-2, the latter being the first research airplane with swept wings.73 But this chapter's examination of Langley's role in this genesis and development of the transonic research airplane program ends with December 1948. At that time, the National Aeronautic Association selected John Stack of Langley laboratory to share in its 1947 award of the Robert J. Collier Trophy, the association's annual prize for the greatest achievement in American aviation. In a ceremony at the White House, President Harry Truman presented the Langley engineer the award citation, which read:
In accepting his citation, Stack insisted that he should not have been singled out for a share of the Collier award. The NACA's contribution to the supersonic flight of the XS-1, he said, had been a team effort.75
It was ironic that Stack won a share of the Collier Trophy, commonly rated the highest honor in American aviation, for his part in the success of the Bell XS-1. Supersonic flight had not even been Stack's original interest;  his idea for the research airplane program had been only to get information in the transonic speed range. He had opposed the army's decision to build a bold new air-launched rocket plane designed especially for the purpose of pushing through the speed of sound. He had favored a rather conservative, turbojet-powered airplane designed to take off from the ground, as airplanes had always done, and explore the high-speed frontier from Mach 0.8 to 1.0. It was the Douglas D-558 that had actually followed Stack's concept. It was the development of the D-558, not the XS-1, that Stack had most encouraged NACA researchers to advance.76
After the successful supersonic flight of the XS-1 in October 1947 the NACA said nothing to indicate what Stack's real position on the development of the rocket plane had been. Rather, it emphasized the cooperative nature of the entire research airplane program. During a public presentation in June 1949 Stack said: "The research airplane program has been a cooperative venture from the start among the Air Force, Navy, the airplane manufacturers, and the NACA. The extent of this cooperation is best illustrated by the facts that the X-1, sponsored by the Air Force, is powered with a Navy-sponsored rocket engine, and the D-558-1, sponsored by the Navy, is powered with an Air Force-sponsored turbojet engine."77 Stack repeated these two sentences in speech after speech in the late 1940s and early 1950s.78 As NACA spokesmen reiterated Stack's message, people believed that the research staff at Langley laboratory had in fact planned from the beginning for the XS-1 and D-558-1 to be complementary research vehicles, with the idea that the army plane would push through Mach 1 to supersonic flight while the navy plane simultaneously studied the transonic region from Mach 0.8 to Mach 1.79
In reality, of course, Stack had argued strongly from 1944 to 1946 that the rocket plane the army was procuring from Bell was unsafe and in important ways unsuitable for studying the intractable transonic speed region. This attitude eventually produced two further ironies. First, it was the conservative, slower-speed D-558-1 turbojet preferred by Stack, partly for safety reasons that killed NACA pilot Howard Lilly in May 1948 due to engine failure during a ground takeoff. The faster air-launched XS-1s had a good safety record at Muroc. Second, it was the D-558-1, not the XS-1 that ended up the greater anachronism. Shortly after the NACA began testing its D-558, service airplanes like North American's F-86 Sabre flew in the Mach 0.8 to 1.0 speed range that the NACA most wanted to explore. The NACA could have instrumented one or more of these service planes as it did the D-558-2 at Muroc and could have conducted extensive transonic flight research using them. As Stack's associate John V. Becker wrote in The High-Speed Frontier, "If the D-558-1 could have been promoted in the...
...early forties, it would have been timely. But coming into the flight picture as it did in 1947, it was unnecessary." 80
There was another reason why the D-558-1 was unnecessary by 1947. Not only were certain service airplanes flying fast enough to be instrumented for transonic flight research, but NACA engineers had discovered a variety of ways (see next chapter) to circumvent the problem of wind tunnels choking just below and just above the speed of sound-the problem, then thought to be insoluble in the short term, that had led Stack and his associates to the idea of the research airplane program in the first place.
Although it was ironic that John Stack shared the 1948 Collier Trophy for the supersonic flight of the rocket-powered XS-1, Stack and the NACA certainly deserved recognition. Supersonic flight depended unquestionably upon their prior successes. Almost singlehandedly, the Langley engineer had initiated the research airplane program and had sold it to military services heavily preoccupied with fighting a world war. As has been shown, this was not an easy accomplishment: the army did little with Stack's initial proposal other than to put it in a desk drawer at Wright Field. After the Bell XS-1 was in procurement, NACA ideas (including some from Stack) and new research information (provided by LMAL research teams led by Stack)....
 ....contributed greatly to the airplane's rapid development. NACA personnel, overcoming vested interests and the "not-invented-here" syndrome (their own and others'), became enthusiastic about cooperating with the military and its contractor to improve the chances of the experimental rocket plane. The cooperation that resulted supplied the American aircraft industry with the data base it needed for the safe and efficient design of the transonic and supersonic aircraft the U.S. military now wanted.
In the end, the research airplane program seems to have furthered the cause of the NACA almost as much as the NACA furthered the cause of the research airplane program. The transonic problem stimulated the development of important new free-flight and ground-based test techniques: the wing-flow, drop-body, and rocket-model methods. Working on the XS-1 and D-558-1 provided Langley researchers with a focus and a goal that were needed after the end of World War II. Winning the Collier Trophy in 1948 for the supersonic flight of the XS-1 (by then designated the X-1) and again in 1952 for the invention of the slotted-wall transonic wind tunnel bolstered the reputation of the NACA and boosted the morale and self-confidence of all NACA employees, at Langley and elsewhere. This was timely therapy after the criticism they had suffered at the end of the war by news of the American "failure" to seize the practical usefulness of the turbojet as quickly as the rival British and German aeronautics communities had.81
For good or bad, involvement in the ambitious research airplane program required the NACA to become more complex organizationally, to do more intra-agency planning, and to formalize some of its methods of management. Planning and monitoring the flight-testing of the XS-1 and D-558 at Muroc was not a small or simple task, especially when it entailed supervision from a mother laboratory some 2500 miles away from the engineers and equipment doing the work. Concern for proper management led the NACA in 1948 to create a special research airplane projects panel and in 1949 to establish a larger NACA High-Speed Flight Research Station (HSFRS) at the California air force base. Langley continued to manage this station until 1954, when NACA headquarters decided to make it an autonomous field installation, the NACA High-Speed Flight Station (HSFS). In 1958, this installation became NASA's Flight (later Dryden) Research Center.
* "As it turned out, the most important region for comparison of flight and tunnels was from Mach 0.9 to 1.1, and thinner wings served as well as a thicker one would have. The region of deep shock stall, Mach 0.75 to 0.9, [the study of] which Stack advocated, proved relatively unimportant from the correlation standpoint." Becker, The High-Speed Frontier, pp. 97-98.
In 1965, at a history meeting of the American Institute of Aeronautics and Astronautics (AIAA), Stack acknowledged the correctness of Langley's thin-wing decision as if he had agreed with it at the time. "We knew it should have a thin wing," Stack told his audience. (Draft of Stack's statement at AIAA History Committee session, San Francisco, Calif., 28 July 1965, p. 6, in "John Stack, Special Collection," Langley Historical Archive.)
** In the theory of fluid mechanics, a potential flow is a type of fluid motion in which the rotation of the fluid element is zero (or irrotational). This type of flow is also called vortex-free flow. The term potential derives from the mathematical concept of the velocity potential. See Theodore von Karman, Aerodynamics, pp. 36-39.
*** It is illuminating to compare Williams's dry technical report with General Yeager's colorful and exciting account of the epic flight published years later in his autobiography for it sheds light on why Williams and Yeager had such a hard time communicating. But Yeager's outspoken reminiscences shed even more light on the differences and personal frictions between test pilots who are engineers (like most of those employed by the NACA), who try always to fly precisely, systematically, and after meaningful data, and pilots like Yeager who are accustomed to "living dangerously and flying the same way." At Muroc in the late 1940s a real grudge apparently grew up between these two types of pilots. According to Yeager's autobiography, the NACA "wasn't thrilled" with the army's selection of him as the XS-l's test pilot: "The NACA team [at Muroc I thought I was a wild man," a macho fighter jockey with no education and no real experience in flight research, whose cockiness might very well lead to tragic mistakes. Yeager, who remembers being treated with some condescension, calls the NACA pilots "the most arrogant bunch" at Muroc; "there was nothing worthwhile that a military pilot could tell them .... I rated them about as high as my shoelaces." See Yeager and Leo Janos, Yeager: An Autobiography (Toronto and New York: Bantam Books, 1985), pp. 129-131, 180-183.