Langley first built its reputation as an outstanding aeronautical research institution on the strength of the variable-density wind tunnel. Max M. Munk, the NACA's German aerodynamicist, proposed this unique and, in some respects, revolutionary piece of experimental equipment in 1921. Two years later Munk's so-called VDT went into operation at the lab. The test results it yielded were s superior to those obtained with any previous tunnel design, especially retarding wing performance, that they made the NACA a world leader in aerodynamic research for at least the next ten years. Aircraft companies, engineering schools, and even foreign research establishments, such as the National Physical Laboratory of Great Britain, sent crews to Langley to study the VDT and return home with ideas for building improved versions of it.
Considering this achievement, it s curious that the history of the VDT involves as much controversy as it does. There is the controversy over credit for inventing the tunnel: Was Munk the true father of the VDT concept, or was it the Russian Vladimir Margoulis, who in 1920 was working as an aerodynamical expert and translator for the NACA's Paris office? Even if Munk does deserve credit as the originator of the design concept, does credit for actually designing a feasible VDT rightfully go to Munk or to the engineering staff at Langley? There is also the controversial "revolt" against Munk at Langley, which, though secondary to the VDT achievement, is important for what it reveals of the Langley personality and for what it suggests about the intercultural transfer of technology. Also somewhat controversial in retrospect are the quality of the tunnel design and the quality of its test results. Was the VDT the total aerodynamic triumph trumpeted in the NACA brochures, or was it in fact riddled by shortcomings? Finally, at the end o VDT history, there is the matter of laminar-flow airfoils (which allowed drag to be reduced and speed to be  increased; to be discussed in chapter 4). Was their practical achievement by Langley researchers a reality or a myth?
Many of the major developments in early aeronautics depended largely on findings achieved through intelligent use of research equipment. The laboratories that pioneered aeronautical research and development possessed a surprising panoply of tools, many primitive but a few highly sophisticated. They included whirling arms, dynamometer cars, water channels (also called towing tanks), engine test beds, as well as flying machines. In a given lab, various technical departments supported the work of this experimental equipment. One department might devise and calibrate pressure gauges, balances, recording dynamometers, chronographs, and the like, while another built and repaired test models. Mechanics tuned engines and maintained drive units. Photographers developed cameras to visualize airflow and techniques to measure aircraft movements in real and simulated flight. Successful research and development required careful planning and management of this intricate, expensive, and bedeviling equipment, with personnel organized into teams.
Within the diversity of facilities, the wind tunnel predominated. Francis Wenham built the first known tunnel at Greenwich, England, for the Aeronautical Society of Great Britain in 1871. The tunnel consisted of a steam-driven fan that blew air through a wooden box 12 feet long and 18 inches square, and open at both ends. All succeeding tunnels shared certain features of the Wenham design: a drive system turned a fan that produced a controlled air stream, the effects of which on a scale model mounted in a test section of the tunnel were precisely observed. Balances and other instruments measured the aerodynamic forces acting on the model and the model's reaction to them.1 The progressive integration of improved versions of these wind tunnel components rendered all other experimental aerodynamic research tools, with the exception of full-scale experimental aircraft in free flight, secondary or obsolete by the end of World War I. Subsequent advances in aerodynamics have generally been closely linked to the course of tunnel development.
The physical law behind the wind tunnel was not fully understood until the late nineteenth century though it had been deduced by da Vinci and refined quantitatively by Newton: a fluid flowing past a stationary object produces the same interactions as those that occur when the object moves through the fluid at rest.2 For aeronautical researchers, this meant  that flight conditions could be simulated by holding an aerodynamic surface stationary within an air stream moving at flight velocity. And a tunnel was the ideal place to conduct and observe such simulations. A tunnel was relatively versatile, safe, and economical. A full-size experimental airplane cost a great deal more money than a wind tunnel test model. Reconfiguring as testing progressed cost more at fill size, too. And flying experimental planes could cost lives. Moreover, some test conditions could be measured and controlled more accurately on the ground than in flight, and some instruments could be mounted and read more easily, and lasted longer, in a tunnel.
Though the invention and early (pre-Kitty Hawk) evolution of tunnel technology provided vital knowledge of the forces affecting wing surfaces (specifically about the surface area required to support a given weight, as well as the surface's optimum shape, the wind tunnel's full potential was not entirely obvious in aviation's earliest days. Several developments led to recognition of its importance. First, the Wright brothers relied heavily on wind tunnel data to design, build, and fly the first powered manned airplane in 1903. Second, the electric power industry developed a cleaner and more compact motor to replace older steam-driven monstrosities powering wind tunnel fans. Third, between 1908 and 1915, German aerodynamicists at the University of Göttingen leapfrogged earlier designs when they built the first closed-circuit tunnel.
The real significance of the wind tunnel became gradually more apparent beginning with aviation's dramatic event of 1903. Legend depicts the Wright brothers as simple bicycle mechanics whose hard work led to success, but in truth engineering knowledge underpinned their flying achievements. After their 1901 glider tests revealed major inaccuracies in published aerodynamic data, the Wrights turned t the wind tunnel for reliable design information. (Only one tunnel seems to have operated in America before 1900-that at MIT, built to check drag measurements Samuel Langley had made with a whirling arm in his Washington lab.) In their Dayton shop, the Wrights first built a makeshift tunnel from an old starch box and later a more sophisticated wooden one (with a 16-square-inch test section). By testing the lift of each of nearly 200 airfoil models, they obtained much of the critical information needed to build the highly successful 1902 glider and its derivative, the landmark airplane of 1903. The wind tunnel had proved indispensable to the first successful powered flight.3
Electric power also contributed to the growth of the wind tunnel's importance. No wind tunnel before [910 had more than 100 horsepower. Steam engines powered most of the early tunnel drive systems, at relatively low speeds. After the turn of the century, however, electric motors powered  more and more of the tunnels at faster and faster speeds. The first tunnel fan driven by electricity in the United States was most likely Albert Zahm's at Catholic University in 1901. Zahm's later 8 x 8-foot tunnel at the Washington Navy Yard attained airspeeds in 1913 of up to 160 miles per hour, equivalent to the diving speed of World War I military aircraft.4 Cheap and increasingly available, electricity permitted precise adjustment of tunnel speed and reliable performance at higher horsepower in a quieter and cleaner environment. (The availability of electric power was to become a very important factor in the planning and operation of wind tunnels at Langley, especially during the facilities boom of the World War II era.)
Nearly all of the pre-World War I wind tunnels, starting with Wenham's and including the Wrights', had open circuits; that is, they drew air into the test passage directly from the atmosphere and released it back into the environment. The classic examples of the non-return, open-circuit tunnel are those Gustave Eiffel (1832-1923) built in and around Paris in the early 1900s. His 1.5-meter-diameter tunnels at Champs de Mars completed in 1909, sucked air through a test section at 20 meters per second (roughly 45 miles per hour). Eiffel's later tunnel at Auteuil, built in 1911 and 1912, improved the design. Producing airspeed of 32 meters per second (roughly 72 miles per hour), it was the last great open-circuit design of the era.5 (Open-circuit tunnels are still used today, for special purposes.)
The aerodynamics research staff of the great German physicist-engineer Ludwig Prandtl (1875-1953) changed the direction of tunnel development in 1908, when it finished the first continuous-circuit, return-flow machine at the University of Göttingen. This new tunnel had three inherent advantages over open circuits: first, it reduced power requirements (through partial recovery of the kinetic energy of the air leaving the diffuser); second, by incorporating improved screens and honeycombs, it produced and maintained airflow that was much more uniform than that in open circuits; and third, it permitted pressurization and humidity control. The primary problem peculiar to the closed circuit-turning the airflow 360 degrees-was solved by introducing efficient turning vanes. A settling chamber upstream of the test area in Göttingen's second-generation closed-circuit tunnel, completed in 1916, further dampened air stream turbulence, and a contraction cone at the test section entrance further increased its velocity.6 Thus, the new closed-circuit tunnels produced faster, smoother, drier, and more reliable airflow than any tunnel had produced before.
By the time Langley laboratory came to life in 1920, the closed-circuit tunnel had proved its superiority over the open-circuit type. But the NACA, cautious because its original staff had so little wind tunnel experience, still chose to design its first tunnel with an open circuit. It patterned the design after that of a successful tunnel, which had been in operation for some time at the British National Physical Laboratory. The leaders of the Committee apparently felt that it was better to improve the NFL tunnel design and to get some immediate firsthand operating experience with the proven machine than it was for novices in the field to proceed boldly with the creation of newer experimental technology. In fat, NACA engineering personnel were so inexperienced that they were told to construct and operate a one-fifth-scale model of the English tunnel before going ahead with the design of the actual facility.
In the fall of 1920, very soon after completing the tunnel that resulted from experience with this model, Langley researchers discovered that results from tests in Tunnel No. 1 could not really be applied to the performance of full-size airplanes. Because the circular test section of the new facility was only five feet in diameter, it was impracticable to use models wider than three and a half feet, or about one-twentieth scale. NACA engineers and other informed aerodynamicists knew how to convert or "scale up" data determined from airflow over such a small object, but systematic testing now made it clear to them that the empirically derived factor customarily used to approximate full scale was largely unreliable.
The problem concerned Reynolds number. In the 1880s, Osborne Reynolds (1842-1912) of the University of Manchester had identified this crucial scaling parameter. In a classic set of experiments dealing with the flow characteristics of water through pipes, Reynolds had demonstrated that the responses of an object to that flow depended on the object's size, the speed with which it (or the water) was moving, the density of the water, and the viscosity of the water. He concluded from a mathematical study of the relationship between the flow patterns over a scale model and those patterns over the same shape at actual size that if in both cases a certain flow parameter (the ratio pVd/µ, where p[Greek letter rho] = density, V velocity, d = diameter, and z = fluid viscosity) was the same, the flow pattern in both cases would also be the same. Understanding and using this ratio, known thereafter as Reynolds number, soon became vital to wind tunnel work....
 ....because it provided a rational Oasis for extrapolating experimental data from scale-model testing. The closer a tunnel's airflow came to producing the value of the full-scale Reynolds number, the closer its test measurements came to indicating the aerodynamic forces of actual flight.7
In the older form of atmospheric wind tunnel the Reynolds number usually amounted to only about one-tenth that of actual flight. This limitation was critical in the aerodynamic region known as maximum lift, which determines landing speed, and equally critical in the region near zero lift, or minimum drag, which determines maximum speed. (According to some aeronautical engineers, minimum drag is "mostly fictional" and thus strongly dependent on Reynolds number.)8 Since NACA Tunnel No. 1 was a low-speed facility, which necessarily involved one-twentieth-scale models, the Reynolds numbers of its tests were recognized as being too low by a factor of 20 for comparison with flight performance of the actual aircraft. Though the researchers at Langley knew that it was possible theoretically to increase Reynolds number in their tests by increasing model size, increasing the speed of the airflow, or by increasing the density or decreasing the viscosity of the air, none of these alternatives seemed feasible given the nature of the existing facility.
In 1921 Max Munk, working as a technical assistant in the NACA's Washington office, suggested to the Committee that experimental results comparable to full-scale flying conditions might be realized in a sealed airtight chamber, the air in which would be compressed "to the same extent as the model being tested." His basic idea was simply to achieve higher Reynolds numbers approximating the flight values of contemporary aircraft by using denser air. Specifically, he proposed immediate construction at the LMAL of a variable-density tunnel. This facility, Munk argued, would compensate for the small size of the one-twentieth-scale models by increasing the density of the air in the tunnel up to 20 atmospheres. Though the chief physicist at Langley argued that through flight research his staff could obtain airfoil data at high Reynolds numbers without this expensive new facility, the NACA Executive Committee authorized construction of Munk's compressed-air tunnel in March 1921.9
By the time he arrived in 120 at the port of Boston from his native Germany, en route to Washington, D.C., to confirm his appointment as technical assistant to the NACA, Max M. Munk, just 30 years old, was already a prominent aeronautical engineer. His aptitude in mathematics and the sciences was such that Munk as a young teenager had convinced....
 ....his parents-lower-middle-class dews from the worldly old Hansa city of Hamburg-that he should leave rabbinical school for German academe. In 1914 he earned an engineering diploma at the Hanover Polytechnical School (where to sound more Germanic he started to use his middle name Max in place of his first name Michael) and in 1917 two doctorates at the University of Göttingen, one in engineering and one in physics.
At the university he had been one of Ludwig Prandtl's most gifted students, assisting Prandtl in his effort to achieve higher Reynolds numbers by using oversized models in the new closed-circuit tunnel. During World War I, a significant number of Mink's analyses of wind tunnel experiments appeared as secret military reports. "Nevertheless," according to Munk, "they were translated in England a week after appearance and distributed there and in the U.S." In his doctoral thesis, "Isoperimetrische Probleme aus der Theorie des Fluges," Munk used shrewd intuitive mathematics to solve the problem of how to make the induced drag of a wing (a concept originated by Munk) as small as possible. (He showed that the minimum induced drag of an airfoil was obtained mathematically if the distribution of the lift over the span corresponded to an ellipse.) At the end of the war, he worked a short time for the German navy and then became an employee of the airship manufacturing company Luftschiffbau Zeppelin, where he designed a small atmospheric wind tunnel and proposed the design of a much larger (1000 horsepower) one for the testing of large airship models. This incredible facility was never built, but according to Munk's plan, would have produced a Reynolds number equivalent to the flight conditions of a full-size airship by having a 152-kilometer-per-hour (nearly 100 miles per hour) closed-circuit airflow pressurized to 100 atmospheres.10 An airflow of this speed under such high pressure would have produced a Reynolds number much higher than that produced by any other wind tunnel at that time.
Leaders of the NACA were greatly impressed with what they thought to be the scientific orientation of European aeronautical researchers like Munk and of their parent organizations. Joseph Ames, professor of physics at Johns Hopkins University and chairman of the NACA Executive Committee from 1920 to 1937, wrote in January 1922 that
 Future NACA member Jerome C. Hunsaker had spent a few weeks in 1913 at Prandtl's Göttingen laboratory as a representative of the U.S. Navy while touring several major European aerodynamic labs. On his return he reported his special admiration for this particular German research organization, where a steady stream f promising young doctoral candidates under an accomplished academic mentor provided the lifeblood of the research effort.12 Thus after the war and despite its residual ill will, the NACA generally and Hunsaker specifically would be predisposed to listen closely to any request by one of these young aeronautical scientists for employment. According to Munk's own version of his 1920 migration from Germany to the United States,' Prandtl had contacted Hunsaker soon after the end of the war about a job for Munk. (Munk was interested in going to America partly because a distant uncle had made a fortune in mining here.)13 Hunsaker informed Ames of Munk's interest and availability, and Ames persuaded the rest of the Committee, which was then hard pressed for talented aerodynamicists (Edward P. Warner having just resigned as Langley's chief physicist), to offer Munk a position as technical assistant. Munk's employment required two orders from President Woodrow Wilson: one to get a former enemy into the country, the other to get him a job in government.14 (At the end of the next world war, another special arrangement would bring the German rocket specialists led by Wernher von Braun to work for the American government as part of "Operation Paperclip.")15
For six years Munk was stationed in Washington, where he worked mostly on theoretical problems. He contributed theories of flow around airships, and of moments and positions of center of pressure on other aerodynamic shapes. He introduced a significant advance in airfoil theory, in the form of a linearization that permitted the calculation of certain airfoil characteristics in terms of easily identified parameters of the profile. During the six-year period the NACA published over 40 of Munk's papers. His contributions were considered so outstanding by the Committee that in 1925 it published a paper (TR 413) by Joseph Ames entitled "A Résumé of the Advances in Theoretical Aerodynamics Made by Max M. Munk."
Munk had arrived at his idea for pressurizing air to increase the Reynolds number in wind tunnel experiments at just about the same time that Russian-born Wladimir Margoulis (former collaborator of aerodynamicist Nikolai E. Joukowski) considered the feasibility of a closed-circuit wind tunnel using carbon dioxide as the test medium. Though the ideas of Munk....
....and Margoulis were elaborated in different ways, their basic concept was the same-that dynamical similarity between scale models and full-size prototypes could be achieved by using a. fluid that had a lower density/viscosity ratio (the (p/µ term in the Reynolds number). The virtual simultaneity of the two men's thinking has, since the 1920s, prompted some people to question the priority: Who had the idea first, Munk or Margoulis?
Margoulis first proposed using carbon dioxide for wind tunnel testing in his paper "Nouvelle méthode d'essai de modeles en souffleries aérodynamiques," which appeared in the Comptes rendus de l'Académie des Sciences, Paris in November 1920. Five months later, the NACA published Margoulis's own English translation of his paper as Technical Note (TN) 52, under the title "A New Method of Testing Models in Wind Tunnels." Munk proposed his idea for a compressed-air tunnel in NACA Technical Note 60, "On a New Type of Wind Tunnel," which appeared in June 1921. Thus, it appears that the first published proposal to increase Reynolds number in wind tunnel experiments by using a fluid of low kinematic viscosity came from Margoulis. On the other hand, Munk had proposed for Zeppelin the design of a pressurized tunnel even before 1920.
The Munk-Margoulis priority issue was not energetically debated in aeronautical circles until the British began to design their own variable density tunnel at the National Physical Laboratory in the late 1920s;....
 .....then the British-American competition for the lion's share of credit for developing the tunnel concept began. In the twentieth Wilbur Wright Memorial Lecture, delivered in England in May 1932, H. E. Wimperis, vice president of the Royal Aeronautical Society, claimed that British engineers had extrapolated the variable-density tunnel idea from Margoulis's paper and had put forward a considered design for a compressed-air tunnel before hearing a word about the NACA design suggested by Munk. Spokesmen for the American aeronautical research establishment disputed this British claim. Walter S. Diehl of the navy's Bureau of Aeronautics, for example, wrote: "While it is quite natural for Mr. Wimperis to argue in favor of the British equipment, I get the impression from his lecture that there is a lot of sour-grape' background and that he is being unfair to the National Advisory Committee for Aeronautics in his statements and comparisons." In Diehl's mind, there was "no doubt whatever ... that Munk originated the idea" and that the British were trying to steal the credit for Margoulis and themselves.16
Though all manner of aerodynamic studies were attempted in the VDT, the facility's primary purpose was to test airfoils. Wing design was one of the most important aeronautical research problems facing NACA Langley in its early years From the time that Sir George Cayley (1773-1857) had identified the inclined plane as "the true principle of aerial navigation by mechanical means" in the 1830s, aerodynamicists had tried in earnest to know better the complex flow phenomena through which the airfoil generates the lift necessary for flight. In the eight decades of sporadic aeronautical development between Cayley's major work and the establishment of the NACA, they had tried everything from crude cut and-dry to rather sophisticated experiments. All of the successful methods of wing design had been empirical. Cayley had feared that the whole subject of aeronautics was "of so dark a nature" that it could be more usefully investigated by experiment than by theoretical reasoning; thus he had tested various airfoil shapes on the end of a whirling arm. In 1879 the Aeronautical Society of Great Britain had reinforced this commitment to empiricism, opining that mathematics had been "quite useless to us in regard to flying." One of the Society's. most prominent members, Horatio Phillips (1845-1924), had conducted primitive wind tunnel tests "of every conceivable [wing] form and combination of forms."17 The Wright brothers had later used a rough version of experimental parameter variation to determine how much lift and drag could be expected from various wing  sections. (Parameter variation has been described as "the procedure of repeatedly determining the performance of some material, process, or device while systematically varying the parameters that define the object or its conditions of operation"; see chapter 5.)18 During World War I, European research teams at the NPL in England, the Eiffel Institute in France, and Prandtl's laboratory in Germany had refined this method. Their five or six best shapes, plus close derivatives, provided nearly every wing section in use at the end of the conflict.
Ironically, the empirical method had been providing designers with some basic misinformation about wings. Since the tests were made at the low Reynolds numbers then available in the small atmospheric wind tunnels, thin, highly cambered (arched) wing sections seemed to have the most favorable properties. At low Reynolds numbers, airflow over thick sections "separated" early and resulted in unsatisfactory performance.* Furthermore, the Wrights had achieved their successful flight in 1903 with a long, slender airfoil. Convinced that the longest span with the thinnest sections generated the greatest lift, some German designers of propellers even went so far as to make their blades from mere fabric stretched by centrifugal force. Nearly all World War I aircraft, with the important exceptions of some advanced aircraft designed by Junkers and Fokker, employed extremely thin wings requiring for external strength and rigidity a messy conglomeration of wires, struts, and cables. 19
In its first Annual Report to Congress in 1915, the NACA called for "the evaluation of more efficient wing sections of practical form, embodying suitable dimensions for an economical structure, with moderate travel of the center of pressure and still affording a large angle of attack combined with efficient action." The Committee could not carry out this work itself, of course, because Langley laboratory was at that time no more than a dream. The best the NACA could do toward improving wing design was to support wind tunnel tests at MIT, which were under the auspices of the airplane-engineering department of the Bureau of Aircraft Production. This experimental program resulted by 1918 in the introduction of the U.S.A. series, the largest single group of related airfoils developed in America up to that time.20
The NACA supplemented its support of the MIT wind tunnel program with a laborious effort by its small technical staff in Washington to bring  together the results of airfoil investigations at the European laboratories. In June 1919 the Committee opened an intelligence office in Paris to collect, exchange, translate, and abstract reports, and miscellaneous technical and scientific information relating to aeronautics. Then, through its Committee on Publication and Intelligence, the NACA planned to distribute this information within the United States.21
One of the early fruits of this labor was NACA Technical Report (TR) 93, "Aerodynamic Characteristics of Airfoils," a comprehensive and handy digest of standardized test information about all the different airfoils employed by the Allied powers. The report, published in the NACA Annual Report of 1920, offered graphic illustrations of the detailed shapes and performance characteristics of over 200 airfoils, as well as four index charts that classified the wings according to aerodynamic and structural properties. The intention was to make it easier for an American designer to pick out a wing section suited to the particular flying machine on which he was working. In retrospect it is plain that many of the plots were totally unreasonable-no doubt because the NACA personnel who interpreted the collected data, like those who made the original tests, did not really understand how and why certain shapes influenced section characteristics as they did. Despite the flaws, however, the effort that went into the preparation of this report and others like it mobilized the NACA staff to manage a solid program of airfoil experiments once research facilities were ready at Langley.22
When the LMAL began routine operation in June 1920, the empirical approach was by far the most sensible way to better wings. Wing section theory, as developed before World War I by Europeans Martin W. Kutta (1867-1914) and Nikolai E. Joukowski (or Zhukovski, 1847-1921, director of the Eiffel laboratory during World War I and consultant to the NACA's Paris office after the war), permitted the rough determination of lift-curve slopes and pitching moments, but little else. It was possible to transform from the pressure distribution around a circle, which was known theoretically, to the flow distribution usually measured around an airfoil, and thus create an approximate airfoil shape, but the mathematics required for the transformation was too abstruse for the average engineer. Further, there was no way to measure the practical value of the mathematical formulations other than via systematic wind tunnel testing. Prandtl had refined the Kutta-Joukowskj method, but his refinement still allowed only for the rough calculation of wing section characteristics.23
Some of the most popular airfoils of the 1920s were produced by highly intuitive methods-cut-and-try procedures based neither on theory nor on systematic experimentation. For the wing section of his successful seaplane,  Grover Loening took the top curvature of the Royal Air Force's number 15 wing section and for the underside crew a streamlined curve with a reverse in the center, which enclosed the spars. The net result of this cut-and-try method was so good that Loening, who did not want other people to copy his product, decided not to submit the wing for tests anywhere. Col. Virginius Clark, USA, designed one of the 1920s' most popular airfoils for wings, the Clark Y, simply by deploying the thickness distribution of a Göttingen airfoil above a fiat undersurface; he chose the fiat feature only because it was highly desirable as a reference surface for applying the protractor in the manufacture and maintenance of propellers.24
The cut-and-try method, though successful in the hands of a few talented practitioners, had too spotty a success record. Aeronautical engineers understood that a wide range of effective airfoils would be created only by using some more systematic analytical method involving tests in a significant and reliable wind tunnel.
From the standpoint of significant and reliable research results, Langley's original atmospheric wind tunnel had been largely unproductive; however, the earliest tests in the new Variable-Density Tunnel, which began operation in October 1922, demonstrated that the NACA's experimental equipment had come of age. Tests in the compressed air of the VDT raised the dynamic scale significantly, validating Munk's design principle and making it possible to estimate full-scale performance more correctly by observing small model wings.
Langley began its first experimental investigation of a series of wing sections in the VDT in 1923. Though the research approach was to be essentially empirical, the idea behind the design of the series derived from a highly intuitive theoretical statement. In the "General Theory of Thin Wing Sections," published by the NACA in 1922, Max Munk had reversed the classic Kutta-Joukowski method. Convinced that contemporary aerodynamicists would fail to produce significantly improved airfoils if they continued to let the wing section be dictated by this mathematical method, Munk decided to "start with a wing section, any technically valuable wing section, and fit the mathematics to the wing section." Even though the method required some simplifying assumptions and did not permit the calculation of maximum-lift coefficients, Munks idea was still a major breakthrough, if not a watershed in the history of airfoil design.25 By replacing the airfoil section with an infinitely thin curved line, it permitted the calculation of certain airfoil characteristics (e.g., lift-curve slope, pitching moments, and....
....chord-wise distribution) directly in terms of easily identified parameters of the shape.26
Munk's analysis suggested to the NACA that a design having a slight upward camber near the trailing edge would result in a stable center of pressure travel. So, starting with a mean line pulled out analytically from one of the better contemporary airfoils, the VDT research team wrapped a thickness form about the upper and lower surfaces of an airfoil. Then, by pulling the mean line or camber out, going to a symmetrical section, and changing all of the ordinates to correspond to the correct proportion of thickness, it prescribed a family of 27 related airfoils. The NACA named the members of this experimental series "M sections" after Munk.27
The range of parametric variation having been determined and the shapes prescribed, the wind tunnel program then followed a rather typical course. First, technicians prepared the precious scale models from a heat-treated aluminum alloy strong enough to take the stresses to be encountered in 20 atmospheres of pressure.** Second, an instruments expert fine-tuned  the tunnel balance and other pieces of recording equipment. Third, an engineer mounted the carefully prepared model in the test section, using a sensitive inclinometer to set its chord parallel to the airflow and calibrating the given angle of attack. In successive tunnel runs, he would change that angle in increments of approximately 1.5 degrees. Finally, someone hit the switch to compress the air in the tank to the desired pressure.
As the test began, two researchers peered through small glass portals in the side of the tank, operating a signal system that triggered different lights as the airspeed became constant or when a problem arose. These men called out their readings of the balance scales to a recorder who simultaneously read aloud, from his panel of instruments, tank pressure and temperature of the manometer liquid. For scale-effect comparisons, the VDT staff made the tests at a constant airspeed (approximately 50 miles per hour) and at five different tank pressures (usually 1, 2.5, 5, 10, and 20 atmospheres) and then tested airfoils of closely related characteristics at 20 atmospheres only. Modifying a particular feature of a model while keeping all its other characteristics constant enabled the staff to compare the aerodynamic effects on each new shape with those on the original. When all the necessary readings had been taken, someone shut the drive motor off and opened a blow-off valve, which released the pressurized air. The calculation, plotting, and final processing of data took weeks. "Computers" existed in those days-but being human; they had to eat lunch, and wanted coffee breaks!
The NACA reported the results of its "Model Tests with a Systematic Series of 27 Wing Sections at Full Reynolds Number" in 1925, declaring that they showed "remarkable agreement" with Munk's theory and had resulted in the design of several sections (especially the M-6 and M-12) with excellent characteristics. 28 Langley's VDT had established itself as the primary source for aerodynamic data at high Reynolds numbers in the United States, if not in the world.
In 1926, following the initial success of the airfoil research program in the VDT, the NACA transferred Munk to Langley full time as chief of the Aerodynamics Division. Munk was to supervise the work of all the wind tunnel, flight research, and analytic sections.29 The only man above him in the laboratory organization was the engineer-in-charge.
Within a year the engineers who worked for Munk were in full revolt against him, and he chose to resign.
Serious tension between Munk and Langley engineers dated back to the design and construction of the Variable-Density Tunnel in 1921 and 1922.
 Munk visited Langley occasionally during this period to monitor the work. Leigh Griffith, Langley's first engineer-in-charge, was apparently grateful for this assistance. At one point he informed George Lewis that "the results of Dr. Munk's [recent] visit to the laboratory have clearly demonstrated that such visits are very desirable as a means of securing a correct understanding of the conditions in the wind tunnel and of harmonizing the opinions of the men immediately concerned with his work." Griffith recommended that Munk visit Langley again in two or three weeks.30
But it was easier for Griffith, an engine man, to feel this way about Munk than it was for Langley chief physicist Frederick Norton, in charge of the aerodynamics sections, who detected in. Munk a stubborn unwillingness to take personal responsibility for transforming the idea of the compressed air tunnel into reality. In 1921 Norton complained to Washington about the chaos brought on by Munk's vague yet overbearing direction of the construction of the Variable-Density Tunnel. He reported that the work of designing the interior and balance system of the VDT was being carried out very inefficiently, due chiefly, be believed, to the lack of sympathy between Munk and the Langley draftsmen and engineers. "Dr. Munk does not seem to have any clear idea as to what he wishes in the engineering design," Norton reported, "excepting that he is sure that he does not want anything that [I or my men] suggest." According to Norton, many portions of Munk's design were quite unsatisfactory:
Later, Norton lamented that be was "getting so disgusted" that he wanted the NACA to keep as much of the tunnel design work as possible in Washington. By this, he meant also for the Committee to keep Munk away from Langley as much as possible, no matter how dissatisfied the German was to stay in Washington working entirely on theory.31
The NACA did not listen to Norton's appeal. When the VDT was ready for operation in late 1922, George Lewis began sending Munk to Langley for periods of four to eight weeks to take charge of the device. During these stays at the Tidewater facility Munk was "responsible for the preparation of the research program, the control of the operation of the apparatus and....
 ....the preparation of reports."32 In 1923 Norton resigned as Langley's chief physicist to work in industry (and later, academics). His successor, David Bacon, an engineer, opposed Munk's presence at Langley even more actively than had Norton. According to an order from the NACA in Washington, Bacon was to turn over the direction of the VDT to Munk during 1924 for a period of four weeks. The order read:
Bacon would not do it. Munk wired George Lewis in Washington: BACON REFUSES TO SURRENDER THE TUNNEL AND FILES. PLEASE SEND INSTRUCTIONS. 33 Bacon relented, but one month later he too resigned from the NACA.34
To understand why the Langley engineers ultimately found it impossible to work with a man like Munk, hardly anything else is needed than an account of their troubles with him over the design of a balance for the lab's Propeller Research Tunnel. When Mink began his full-time duty at Langley as chief of aerodynamics in January 1926, engineer Fred E. Weick was busy designing a support and balance system for the aircraft components and models to be placed in the new tunnel. Weick, a former employee of the navy's Bureau of Aeronautics, had known Munk in Washington. In fact, Weick and Munk were in different ways originally responsible for the NACA's decision to build the propeller research equipment.*** Weick had  great respect for Munk's abilities. On the other hand, he did not want his balance design turned down at the last minute, so he had taken the pain to take each detail of design- which was mostly on cross-section paper-up to Munk to get his approval, and got his initials on every single one of them. This, Weick thought, would certainly assure Munk's final approval.
Weick proceeded to build the tunnel balance atop a structural steel framework. A couple of days before Weick planned to try out the balance using a little Sperry Messenger airplane with its engine operating, Munk made an unannounced visit to the PRT building. Just as he walked into the bare-walled 50-foot cubicle that housed the test section, a loud horn squawked, calling someone to the telephone. According to Weick,
Visualizing the entire structure vibrating to the point of failure and the whole airplane and balance crashing to the ground, the perturbed Munk ordered Weick to tear down the balance entirely and design a new foundation and framework for it. The chief of aerodynamics then went back to his office a couple of blocks away.
Naturally Weick, too, was perturbed. Munk, after all, had approved every detail of that balance. After giving the German aerodynamicist some time to cool down, Weick went to the chief's office and, as calmly as he could manage, mentioned that he thought the natural frequencies of the long diagonal members would be so low that vibrations would not be incited by the more rapid impulses from the engine and propeller. The engineer suggested that inasmuch as the balance was ready to be tried out, they should make a careful trial starting at low speed, gradually increasing it, before dismantling the apparatus. Munk agreed, but demanded to be present when the test was made.
Weick did not like that idea one iota. To start the engine, the Messenger's propeller had to be cranked by a man standing on a ladder. This sweaty business often took some time. It was not the kind of operation he wanted the excitable Munk to watch. Weick executed an end run around Munk, his division chief, and discussed his problem with the engineer-in-charge, now Henry Reid. Together, Weick and Reid decided to check out the tunnel balance system in Munk's absence. This was easily done, as Munk worked on theoretical problems in his room at a Hampton boarding house every afternoon. Weick set up the test run and ran through the speed....
 ....range without any difficulty from the balance. He then made some minor adjustments and satisfied himself that all rough spots had been smoothed out.
The problem of convincing Munk remained. Weick could not simply tell him about the successful test, so he and the engineer-in-charge agreed to arrange another "first test" for Munk to witness. Reid escorted Munk to the tunnel the next morning Weick casually said, "Good morning," walked up the ladder, and pulled through the Messenger's prop. Luckily, the engine started on the first try. Weick moved the ladder away, ran the engine through its entire range, and then shut it down. There was no noticeable vibration in any part of the balance. Weick, who had wondered what Munk's reaction would be, later recounted: "He walked toward me with his hand outstretched and congratulated me on the success of the operation." 35 The balance system operated satisfactorily with engines of up to 400 horsepower into the late 1930s, when it was replaced by a new and better one.
The matter of the PRT balance design resolved, Weick later had to deal with Munk over the technical issue of the best propeller blade-section coefficients-the numbers representing the lift, drag, and pitching moment characteristics. Munk thought that the coefficients should be put on the same logical foundation as that on which wing coefficients were based. While Munk's were more precise and elegant, Weick urged the use of coefficients that would be easier for designers to apply. (As an employee of the Bureau of Aeronautics, Weick had authored NACA TN 212, "Simplified Propeller Design for Low-Powered Airplanes," to help people make their own props for home-built aircraft.) One day in Munk's office Weick argued for his viewpoint. Not flinching, Munk-thumbs in the arm holes of his vestended the conversation with his version of compromise:
At that moment Weick did agree, but, back at work, he continued to use his own coefficients.36
Munk's subordinates did what they could during 1926 to work with him and then around him, but they finally rebelled. In early 1927 all of the section heads of the Aerodynamics Division resigned in protest against Munk's supervision: Elton Miller, head of the PRT section; George Higgins, head of the VDT section; Montgomery Knight, head of the Atmospheric Wind Tunnel (AWT) section; and John Crowley, head of the flight test  section. Engineer-in-charge Reid, in office for barely a year and already stuck between the devil and the deep blue sea, tried to resolve the crisis by reassigning Munk as his adviser. Lewis tried halfheartedly to pacify Munk by asking him to return to Washington, even though Lewis probably did not want that to happen.37 But Munk, his pride hurt, refused all options left open for him with the NACA, emphatically refusing to be holed up again in a small office away from research facilities, and resigned. Peace, and the section heads, returned to Langley, but only at the cost of losing one of the best theorists ever to work there.
Because a profile of an outsider can help to outline the character of an inside group, just as a clear statement of antithesis clarifies a thesis, further exploration of the question "Why was it impossible for Munk to survive as chief of aerodynamics at Langley?" seems worthwhile. Such exploration might reveal important aspects of the historical personality of Langley laboratory, and might suggest much about the intercultural sharing of technology.
Clearly Munk was unusual in the Langley setting. The first thing that any group of Americans would have noticed about him, once hearing him speak, was that he was a foreigner. No doubt his thick accent and unfamiliar inflections made him seem more eccentric than he really was. What was worse in the early 1920s-a time of rampant nativism-Munk was a German, a "hated Hun," only recently the enemy of the United States and its allies.**** In 1921 Frederick Norton, Langley's chief physicist, informed the NACA in Washington that if Munk were to stay at the lab on a regular basis-as he believed (correctly) Munk desired to do-it would be "extremely difficult to fit him into the organization." Army officers at Langley Field, he reported, would not take kindly to the presence of the German.38
Like the great majority of Langley researchers, Munk was an engineer but the professional norms he had learned to value were those of German engineering and Göttingen applied science. Though some Langley engineers understood the importance of theory and were mathematically competent according to the American (not the Gottingen) standard, they saw Munk's  theoretical orientation as a factor separating him from themselves. George Lewis remarked that Munk's works were "of such a highly scientific character that they are not appreciated by the average aeronautical engineer, and can be appreciated only by those who have a very extensive training in mathematics and physics."39 Besides believing this personally, Lewis had heard others complain about the highly mathematical character of Munk's works. For instance, in response to Munk's criticism of one prospective NACA report, Langley engineer-in-charge Leigh Griffith had advised the Washington office that "criticism of research reports dealing with actual laboratory results should not be undertaken by theoreticians since the viewpoint of the theoretician is usually so radically different from that of the laboratory research man." The engineer-in-charge tied his unwillingness to accept the judgment of this particular theoretician to Munk's use of foreign criteria. It "is rather unfortunate," he told Lewis, that Munk was "not more familiar with current standard American nomenclature" and was therefore "inclined to criticize terminology not in agreement with his own peculiar ideas."40
Not only American engineers but American scientists as well thought that Munk's report writing, both its style and substance, was excessively vague and obscure. After reading the draft of Munk's July 1925 report "On Measuring the Air Pressures Occurring in Flight," Joseph Ames, a physicist, wrote Lewis that Munk's discussion of general problems in the paper was "excellent" but his style "impossible." It was "neither fish, flesh nor fowl." Ames nevertheless recommended that the report be published after extensive editing.41 However, Walter S. Diehl of the Bureau of Aeronautics (a profoundly influential individual in NACA history; see especially chapter 6) reviewed Munk's prospective report and asserted that "there can be no real argument about the style desirable for a scientific report." It had to be "clear, concise, and without grammatical errors or rhetorical flourishes." "I feel that Munk has carried the matter entirely too far and that he is substituting rhetoric for scientific facts," Diehl charged.***** Though he  considered the piece of interest and value, Diehl questioned the advisability of publishing it. He suggested holding out for a "conventional report," one in which the observed data-all of it-held center stage.42
In addition, Munk showed personality quirks that went far beyond those tolerable even in NACA Nuts.****** Having internalized the social relations of German academic life, Munk considered himself the absolute master of the division he directed. He intended to set the research goals and, like a German university professor, himself receive whatever credit was forthcoming. A proud genius, Munk was frequently autocratic and arrogant in his dealings with people, treating his men at Langley as graduate students and obliging some of them to attend a seminar on theoretical aerodynamics he conducted in a way that at least two talented young men, Elliott Reid and Paul Hemke, a Ph.D. in physics from Johns Hopkins University, found rude and condescending. (In 1927 Reid and Hemke both resigned from the LMAL, "strongly influenced in their decisions to leave the Committee because of their unpleasant relations with Dr. Munk." )43 While still learning English by reading Macaulay and Oscar Wilde, Munk had the audacity to offer Langley employees and their wives a night class on English literature. The class met once; Munk's sweeping criticisms of some of the class members' favorite authors and books alienated his audience completely.44
In sum, one may interpret the revolt of the Langley engineers against Munk as a clear instance of non-adaptation between different national cultures of science and engineering, or as a case in point showing how "culture shock" may affect technology transfer. American history is full of outstanding examples of skilled European technologists, such as Samuel Slater, Benjamin Latrobe, and John Roebling during the early nineteenth century, emigrating to the United States and successfully transplanting the....
....new or "hot" technologies of the Old World in the fertile soil of their New World. But these transplantations were not automatic. There were certain forces in American society that resisted technology transfer. We know for example that American clients criticized Latrobe for his stubborn insistence on doing everything to English standards, such as building with expensive stone instead of wood. 45 We know also that the canal builders of western Pennsylvania at first rejected (also mainly for financial reasons) Roebling's advice to substitute wire for hemp in their winch-driven canal cables.46
Transfers of technology, like transplantations of living organisms, are difficult, and do not always succeed. You cannot raise cotton in Michigan, after all, or sugar cane in Maine. Alligators will not live in the Bering  Sea. Taken away from native nutrients, confronted with the challenges of a radically different environment, species have to prove adaptable genetically or they die out. The story of Max Munk at Langley should encourage students of American science and technology to be more sensitive to the kinds of attitudes and arrangements on both sides of a transfer that in some circumstances facilitate and in others impede the flow of knowledge between different peoples.
* "At small angles of attack the flow has little difficulty in following the surface. As the angle is increased, however, the air finds it increasingly difficult to maintain contact, especially on the upper surface, where it his to work its way against increasing pressure, and it separates from the surface before reaching the trailing edge." Theodore von Karman, Aerodynamics (Ithaca, N.Y., 1954), pp. 46-47.
** In the early days of the VDT, Langley ordered most of its models from the W. H. Nichols Machine Company, Waltham, Mass., which possessed a piece of equipment perfect for cutting metal airfoils. Later the NACA bought this machine from Nichols and made many of its own models. Because precision in construction of a model was essential for an accurate experiment, with special attention necessarily being paid to exactness of contour, fineness of the polished surface, and the chance of error resulting from excessive machine tool wear, LMAL managers felt that this essential manufacturing job should not be left in the hands of contractors.
*** In response to a request from George Lewis in 1923 about how the NACA might better help naval aviation, Weick (B.S., University of Illinois, 1922) mentioned the need for full-scale propeller tests at high tip speed, where compressibility losses became evident. (Compressibility is the physical property by which the volume of matter-air, in this case-decreases to some extent as pressure is brought to bear on it.) According to Weick, the British had made high-tip-speed tests on two-foot models in a wind tunnel, but the Reynolds number was so low that the results were questionable when applied to full scale. At that time, however, Weick could not envision any practical way of making full-scale tests other than in flight.
About a year later, Lewis asked Weick how he would like to see a wind tunnel capable of making the tests. Weick laughed, remarking that in order to make practical tests on a ten-foot propeller at speeds up to 100 miles per hour, the diameter of the tunnel's throat would have to be at least 20 feet, or four times the largest wind tunnel at Langley up to that time, with a drive unit capable of providing at least a couple thousand horsepower. "Yes," Lewis agreed, "but I've been talking it over with Dr. Munk, and we think that such an arrangement might be practical." Asked if h would like to come down to Langley Field and run the propeller tunnel for the NACA, if built, Weick agreed without hesitation. Weick, "Historical Reminiscences," tape 3 (side 1); transcript in the Langley Historical Archive (LHA), p. 6.
**** I have found no evidence that Munk suffered from any anti-Semitism at Langley; in fact, Munk, during my most recent (20 August 1985) interview with him, refused to admit that this sort of ill will, to whatever extent it existed, had anything to do with his problems at the NACA laboratory.
***** Typically a paper by Munk included very few references to relevant published literature or rigorous mathematical demonstrations which would show readers exactly how he came to his conclusions. His manner of thinking was so highly intuitive that he proceeded in research as if he were the only person working in the field. His collection of technical books (which Munk recently donated to the Langley Historical Archive) is remarkably meager-perhaps indicating the great extent to which he relied on no human being but himself for revelation of knowledge. In "My Early Aerodynamic Research: Thoughts and Memories" (in the Annual Review of Fluid Mechanics, 13 , pp. 4 and 6), Munk declared: "Mathematics comes from within .... Undertaking research for the advance of mathematics is more difficult than using established mathematics. It requires more curiosity, diligence, and aimful thinking. The researcher's character in general, I believe, has also much to do with it. The pure in heart shall see." When asked (during an August 1985 interview) how he arrived at the thin airfoil theory of 1922, for example, Munk answered, "How do such things happen? They are miracles!"
****** Langley old-timers Harold R. Turner, Sr., James G. McHugh, and Hartley A. Soulé love to tell a tall tale about Munk learning to drive a car; over the years, the story has become a sort of local legend, an extravagantly exaggerated one. When he arrived, the story goes, Munk had never driven an automobile. One of the wind tunnel technicians tried to teach him how to do it right, but Munk thought there was a better way. So, he drew a map of the road between Hampton and Langley Field, figured the exact distances between the curves of the road, calculated the curvatures of the mandatory turns, hung a string down from the top of the steering wheel, and applied numbered pieces of tape to indicate the manipulation of the wheel required for the car to follow each turn. Then, by driving at a predetermined speed, he could, with the help of his map and a stopwatch, make it safely from home to work. According to Soulé, "It was that type of story that caused the local people to assume that everyone from the NACA was a screwball" (Soulé interview with Walter Bonney, 28 March 1973, p. 2 of transcript in LHA).