This is the story of an NACA research authorization. It tells how and why the authorization was opened, executed, and closed. While research authorization 201 had some idiosyncrasies-it lasted longer than most and produced fewer practical results-still it is sufficiently representative to give some idea of how the NACA went about aeronautical research. It is particularly enlightening on the respective roles of headquarters and the laboratory in selecting and conducting research projects, on changes in those roles over the years, on publication policies of the Committee, and on the relations of NACA staff members with clients and colleagues.
Boundary-layer research had been going on in Europe for 20 years before the NACA took any official interest. Only when the Europeans began to achieve some success in boundary-layer control did the NACA launch a program of its own. The NACA was always more interested in application than in theory; it never wanted to understand the wind so much as to control it.
The boundary layer is a thin film that forms on the surface of a solid body moving through a viscous fluid, like the wing of an airplane moving through the air. Within the film, velocity increases parabolic ally, from zero at the solid surface up to the free-stream velocity at the outer edge of the boundary layer. The depth of the layer varies with the smoothness of the surface, the viscosity of the fluid, and the speed of the flow, but it is never very large. At 5 cm from the leading edge of a fiat plate moving through standard sea-level air at zero angle of incidence and 120 m/sec, the boundary layer will be only .04 cm deep.1
The boundary layer was first identified and labeled by Ludwig Prandtl in 1904 in a classic paper that revolutionized this branch of fluid mechanics. The Göttingen University  professor actually used the term "boundary layer" only once, while he used "transition layer" seven times. But "boundary layer" became the accepted term, and boundary-layer theory became the descriptor of choice for the entire field. Prandtl had based his paper on empirical investigations, but his concept remained only a theory until it was verified in the 1930s and 1940s by more sophisticated research instruments and techniques. Even today, some of the more complex behavior of the boundary layer is explained only by unconfirmed theory.2
Applications of boundary-layer research are as diverse as the circumstances of fluid flow itself. Prandtl was studying the use of a jet of air to blow away sweepings in a factory. Others looked into the flow of fluids in pipes. Many turned their attention to the infant technology of flight, seeking to improve the flow of air over wings.
The flying qualities of wings can be enhanced in two ways, and boundary-layer control can help in both. The first is to decrease drag; the second is to increase lift. The most desirable way to decrease drag is to maintain laminar flow within the boundary layer and prevent a transition to turbulent flow. Laminar flow occurs when successive layers of air within the boundary layer slide smoothly over one another, from the stationary film at the surface up to the free-stream velocity of the outside air. Turbulent flow within the boundary layer occurs when these "streamlines break up and a fluid element moves in a random, irregular, and tortuous fashion," as when the smoke rising from a cigarette in a still room ceases to travel smoothly up but tumbles instead in eddies and curls. Over a normal wing, the boundary layer remains laminar over only a small portion of the wing chord before breaking up into turbulent flow. The area of turbulent flow experiences significantly greater skin-friction drag than the laminar flow.3
The second way to improve the flying qualities of a wing through boundary-layer control is to increase the lift, especially the maximum lift, of the wing. Maximum lift can be increased by delaying the onset of separation of the boundary layer. As a wing's angle of incidence increases-as its leading edge is tipped up above the plane of flow of the....
 ....free-stream air-its lift also increases, up to a point. Finally, however, the boundary layer on the upper surface breaks free of the wing altogether, reducing lift drastically. This is known as stalling. If the boundary layer can be kept from separating, the maximum lift of the aircraft can be increased, an important consideration in increasing takeoff-weight capacity and reducing landing speed. Furthermore, the same energizing of the boundary layer that delays separation can also help to maintain the boundary layer in fast laminar flow, increasing total lift even at low angles of incidence.
In the early years of boundary-layer theory, two methods of boundary-layer control were proposed by the Europeans, who dominated the field. Prandtl and his proteges at Göttingen developed mechanisms to suck the boundary layer along the upper surface of wings, thus maintaining laminar flow and preventing separation. Others studied ways of blowing air into the boundary layer near the leading edge, to energize the boundary layer and prevent separation.
The latter technique, with its promise of practical application, first drew the NACA into boundary-layer research. In 1926, Elliot G. Reid, a junior aeronautical engineer and a member of the Langley Memorial Aeronautical Laboratory's research council, wrote to the engineer-in-charge of the laboratory about European work in boundary-layer control. He noted that the research of Handley-Page and Lachmann in England "constituted the first successful attempt to control flow in the boundary layer," thus improving the performance of wings. He cited NACA Technical Memorandum 374, published just that year, describing the work of the Göttingen group under Prandtl. And, most important, he referred to John J. Ide's recent visit to Vienna, where the NACA's European representative had talked with Richard Katzmayr, director of the Vienna Aero dynamical Laboratory. Katzmayr was trying to increase lift by blowing compressed air over the upper surface of airfoils, and had already published some promising results. He gave Ide an extremely optimistic account of his work to date. On the basis of Ide's report, Reid suggested that the NACA Aerodynamics Committee should authorize 'Experiments on Airfoils with Modified Boundary Layer Flow," looking into Katzmayr's blowing technique as well as the obverse method, the suction technique advocated by the Göttingen group.4
Reid's proposal immediately fell foul of the bureaucracy both at the laboratory and at headquarters in Washington. It was first forwarded for comment to Max Munk. Munk, the enfant terrible of the NACA, was a brilliant and temperamental aerodynamicist  who had done more than anyone to set up the program and facilities of the Langley laboratory and to distinguish the NACA by important contributions to aeronautics. Just now he was in charge of the Aerodynamics Division, and he resented an outsider from the Engine Research Division suggesting programs for his fiefdom. "I suggest that Mr. E. G. Reid be advised to draw his memorandum back," he replied icily, "and to ask it to be forwarded to the Aerodynamics Division, if he cares to."5 Apparently Reid did not care to, trying instead an end run around Munk directly to the NACA Aerodynamics Committee. That ploy brought him into collision with George W. Lewis, the NACA director of aeronautical research. Lewis advised the laboratory on 11 November-just one week after Munk's rebuff of Reid-that in the future all research recommendations would go through the Director of Aeronautical Research and would not be proposed directly to a technical committee or subcommittee.6
The engineer-in-charge duly forwarded Reid's recommendation to Lewis, with a copy to Munk. Munk's response, now more formal, displayed the temperament that would finally undo him at Langley:
Part of this argument was mere self-serving rationalization, an attempt by Munk to keep his own field inviolate and to have the last word on what was done within it. To this extent it is petty and at odds with the way the Langley staff operated at its best-encouraging a free flow of ideas and suggestions and cutting across administrative boundaries as the demands of aeronautical research dictated. But Munk's argument contained a kernel of truth, and the investigation of boundary-layer control by the NACA might have proved more successful had Munk's advice been taken. Like all complex research activities, aeronautical research requires an informed supervisor able to see the big picture, to distinguish the forest from the trees, to separate the random interesting idea from the cumulatively productive next step in a long-term investigation. Reid's suggestion, though full of interest and potential, still bore no guarantee that it would prove the best way to use the limited personnel and facilities available to the NACA.
The engineer-in-charge sent Munk's comments along to George Lewis. For two weeks nothing happened. Then, on 3 December, Lewis sent Langley new photographs and test results from Katzmayr and directed the lab to check them in the atmospheric wind tunnel, earliest and crudest of the tunnels then at Langley. The laboratory staff may have considered this a tentative approval of Reid's suggestion, but the real source of this authorization apparently was in Washington. On the previous day, Captain E. S. Land, assistant chief of the Navy's Bureau of Aeronautics, had formally requested that the NACA test in a wind tunnel and in flight the Katzmayr method of increasing lift.8 Land was then one of two Navy members of the NACA Main Committee, a member of the Executive Committee, and a frequent visitor to the NACA offices. George Lewis may well have asked him if the Navy were interested in checking out Katzmayr's claim. The lower takeoff and landing speeds resulting from improved lift could appreciably help the Navy in its early attempts at carrier aviation.
 Whether through collaboration or not, on 6 December Lewis forwarded Land's letter for comment to Langley with a revealing postscript: "A research authorization (No. 201) will be submitted to the Executive Committee for approval at its next meeting to carry out this request of the Bureau of Aeronautics." 9 In other words, Lewis might hesitate to submit laboratory recommendations to the Executive Committee-especially when opinion in the Langley staff was divided-but as soon as he had a Navy request in hand he had a research authorization in draft. Capt. Land's letter added nothing to the technical justification for the research, but it did add the political justification Lewis seems to have felt he needed.
While Langley was waiting for the Executive Committee to act on the proposal, the staff agreed in conference on the desirability of investigating Katzmayr's scheme as well as the suction method of boundary-layer control recently demonstrated by Theodore von K·rm·n, one of Prandtl's most accomplished proteges.10 Flight testing and research in the atmospheric wind tunnel were prescribed, should research be authorized. Meanwhile, Katzmayr had visited Ide and provided full blueprints of his invention, which were duly forwarded through Lewis to Langley. Thereafter Katzmayr persistently sought word of NACA results, especially of comparing his blowing method with the suction method proposed by his rivals at Göttingen. Ide was eager to give him the information in return for his cooperation, but the Langley personnel at first reported "indefinite" results because of the "contradictory nature of the data" submitted by Katzmayr. By the middle of the following year, they were viewing the suction method as more promising, though before they reached a firm conclusion they wanted to know how Katzmayr had achieved the high pressures he claimed. At that juncture, Katzmayr disappears from the records.11
While this modest investigation was proceeding, Lewis won the endorsement he had promised the laboratory on 6 December. Research Authorization 201, "Investigation of Various Methods of Improving Wing Characteristics by Control of the Boundary Layer," was approved by the Executive Committee and signed by its new chairman, Joseph Ames, on 21 January 1927. Broad as the title of the authorization was, its "Why" and "How" sections made it all too specific. The purpose of the investigation was to "Determine the possibilities of improving wing characteristics" by using the blowing and sucking methods suggested by Katzmayr and by the University of Göttingen  respectively. The Katzmayr method was to be tested in the atmospheric wind tunnel and in flight, the Göttingen method only in the wind tunnel. Under "Remarks" it was noted: "Investigation requested by Bureau of Aeronautics."
With the research authorization in hand, the Langley staff offered a spate of suggestions: Max Munk suggested rounding the trailing edge of a wing as a means of controlling the boundary layer, and requested "that priority of this invention be taken down for the writer for a future application for a patent." The engineer-in-charge forwarded this suggestion with a note that it could be readily incorporated in the program for the atmospheric wind tunnel; this procedure was quickly approved by the Executive Committee and charged to research authorization 201. E. G. Reid, who at the beginning of this story had suffered at Munk's hands, now turned the tables by asking Munk to describe in a memo just exactly what he was recommending. When Munk failed to reply, the engineer-in-charge put the same question to him in writing. Still no answer. At last-apparently after a personal interview-the engineer-in-charge recorded formally on his own memo: "Dr. Munk has no suggestion to make." These were Munk's last days at the NACA, and this behavior was typical of the animosity and friction between him and the staff that made his departure inevitable. 12
Suggestions by other members of the Langley staff fared better. A laboratory assistant recommended investigation of "electrical lubrication"-electrically charging the wing, with the expectation that the adjacent air would take on the same charge and be repelled, thus repelling the boundary layer and eliminating skin-friction drag altogether. Although entirely unrelated to the blowing and sucking methods specified in the research authorization, this idea won quick approval from the Executive Committee. Three other engineers recommended that research on the suction method proposed by Göttingen and reported in NACA TMs 374 and 395 be conducted in the variable-density wind tunnel. This newest tunnel, the brainchild of Max Munk, was the NACA's first radical research tool, a device for getting results more closely approximating those of an airplane in flight (see chapter 4). Though this too was a departure from the original specifications of research authorization 201-which had called for research in the atmospheric wind tunnel-the Executive Committee nonetheless gave it similarly quick endorsement.13
If a pattern was emerging here, it consisted of cautious NACA approval of a basic research authorization, preferably based on a specific request from one of the military services, followed by ready endorsement of laboratory suggestions on how to conduct the research. The Committee approved new ideas from all professional members of the staff (Munk's objections notwithstanding) and had no apparent compunction about straying from the precise language of the authorization. The authorization served rather as a foundation on which the laboratory, with the approval of Lewis and the Executive Committee, could build a program of its own choice.
First hard results of this particular investigation were disappointing. Thomas Carroll's report on "Preliminary Flight Tests of a Method of Boundary Layer Removal," submitted 2 September 1927, concluded that "the improvement in performance is negligible" for the first arrangement of sucking slots that the staff tried on an aircraft wing. One of the engineers quickly cautioned the engineer-in-charge not to publish these results, because they were for one method of installation only; circulation of the report, he warned, "might make persons less familiar with the subject skeptical of any possible improvement in wing characteristics by boundary layer control." The engineer-in-charge concurred and advised headquarters not to publish the report, as other wind-tunnel tests then underway might suggest better arrangements. 14
In spite of this tendency to play their cards close to the vest, the Langley staff was already amassing a useful store of knowledge. In October 1927, less than a year after work on this research authorization began, J. S. McDonnell, Jr. then a struggling young engineer in private employ, later to become one of the giants of the aircraft  manufacturing industry-wrote to ask if the NACA was doing research on the blowing and sucking methods of boundary-layer control such as that reported from Göttingen in the NACA TMs or from the Army's McCook Field in the Journal of the Society of Automotive Engineers. H. J. E. Reid was able to reply that the laboratory's research to date indicated that overall efficiency increased with use of suitable slots, that suction was more economical than blowing, and that a blunt nose on the airfoil appeared better than a sharp one. These results were worth publishing, and Reid in fact stated that the NACA was preparing a preliminary report that would include "a complete bibliography which may be considered as a guide to the work done on this subject by other research organizations."15 Though the NACA was not itself publishing preliminary results, it was apparently following very closely the results published by other laboratories.
Another year was to elapse before the NACA actually published its first report under this research authorization. In the meantime it embarked on several new departures. In January 1928 George Lewis wrote the laboratory that in a recent conversation with Orville Wright, then a member of the NACA Main Committee, the pioneer aviator told of experimenting with a wing having a split trailing edge. This produced a considerable negative pressure along the line of the split; at high angles of attack it was possible that openings from the split to the interior of the wing could suck air into the wing and perhaps control the boundary layer. Lewis directed-apparently on the basis of this conversation alone-that Wright's concept be included in the work done under research authorization 201.16
Henry J. E. Reid, in a letter drafted for him by Elton Miller, new head of the Aerodynamics Division, replied to Lewis that Langley would do the work, but the staff was not optimistic. The split flap, they thought, would probably increase drag, decrease lift, and produce the kind of turbulent wake that accompanies separation. Lewis responded with a Washington Navy Yard report which he believed contradicted the staff predictions; one of the engineers at the lab countered that the trailing edge described in that report was a downward flap only, not the split flap recommended by Wright. When the research was concluded and a report prepared the following year, it confirmed the staff's skepticism. Said Reid in forwarding the report to Lewis, "The results obtained in this investigation are mainly negative and it has been doubted whether the paper is worthy of publication." Once more (as in the preliminary sucking-slot tests) the staff at Langley was recommending suppression of negative results, but in this case Lewis seems to have overridden their objections. Five months after the negative recommendation by Langley, the same report, now edited and retitled, was forwarded for publication as a technical note. 17
The laboratory was more successful in suppressing the results of another investigation tacked onto research authorization, 201. In June 1928, a Langley engineer brought to the attention of George Lewis some Japanese research, which concluded that rows of transverse flaps across the upper surface of a wing would prevent the backflow of air along the upper surface at high angles of attack-a precondition of separation with little effect on drag. Tests seemed warranted. Lewis agreed, and authorized tests under research authorization 201. But again the results were disappointing. Early in 1929, Henry Reid forwarded to Lewis a report that he said was based on "somewhat crude" research equipment. He did not recommend the report for publication, and he did not have the personnel to continue the research. The chief test pilot, however, was more optimistic about the technique, as was Lewis, who told LMAL that he found the results interesting and wanted more research done. But there the record stops. Queried about the Japanese technique in 1935, the laboratory staff could find no memorandum report on its research, or even any record of tests beyond some notes in the chief test pilot's own files. In this one case, at least, the laboratory succeeded in smothering a project it did not want to pursue.18
 Not until the summer of 1929 did the Langley laboratory forward the first findings under research authorization 201 that the staff judged suitable for publication. On 23 August, H. J. E. Reid forwarded a document by Montgomery Knight and Millard J. Bamber, "Wind Tunnel Tests on Airfoil Boundary Layer Control Using Backward Opening Slot," recommending its publication as a technical note. Two months later it appeared as NACA TN-316. Less than two years after that came the culmination of the work under research authorization 201, Millard J. Bamber's "Wind Tunnel Tests on Airfoil Boundary Layer Control Using a Backward Opening Slot. " In forwarding this report to headquarters, H. J. E. Reid recommended its publication as a technical report, the top of the NACA line and the intended end product of all research authorizations. Reid specifically noted that "the work covered by this report was done under Research Authorization No. 201 and completes the work to be done under this authorization." The following year the report was published as NACA Report 385. In it, Bamber mentioned the personnel limitations on the investigation and suggested that this research was all the NACA was going to conduct on this topic. Reading the records only to this point might lead to the conclusion that research authorization 201 had run its course.19
In fact, however, research authorization 201 was just getting under way. Even as Bamber's report was being edited for publication, another report by another engineer went from Langley to headquarters, carrying a note by Reid that the research was conducted under R.A. 201 and did not complete the work to be done under that authorization. And, in the same year, another young engineer at Langley, Hugh B. Freeman, submitted a preliminary report on an investigation conducted under research authorization 201, this time on pressure distribution about an airship model, an entirely new departure in NACA boundary-layer research.20 Langley records do not explain why they suggest why R.A. 201 expanded into an umbrella for work not directly connected with the blowing and suction techniques suggested by Katzmayr and Göttingen, the initial targets of the research. There was other research authorizations active under which boundary layer investigations could be-and in fact were being-conducted. The most likely explanation is the promise offered in Bamber's final report of actually controlling the boundary layer by suction and blowing, and Lewis' reluctance to abandon the research especially when he held an authorization explicitly requested by one of the armed services. Better, perhaps, to keep the authorization open and use it for targets of opportunity: if a promising new departure in research appeared, it could be pursued within the mandate of this research authorization without going back to the Executive Committee and asking for approval of what might appear in embryo a far-fetched line of research.
Whatever the reasons, research authorization 201 remained open, and under its protective cover all manner of boundary-layer research went on. In 1932, for example, the newly opened NACA tow tank-a model basin intended primarily for experimentation with seaplane hulls-was drawn into a Navy investigation of soaring birds in still air. The scheme, not especially well received at Langley, was to harness seagulls, buzzards, and Seahawks to a movable carriage in the NACA tank and pull them along at varying speeds, to measure the lift their wings developed at different attitudes and degrees of extension. Constructing a balance that could measure the results of these tests became a major research project per se, and the Langley staff found itself not only yielding up precious tank time to the enterprise but also becoming immersed in procuring test specimens and designing and supervising construction of the balance.21
But these tank tests were merely a distraction and an aberration. The real center of activity on research authorization 201 in the next phase of its long career was to be Hugh B. Freeman, the young engineer who reported late in 1931 on airship research. In a memorandum to the chief of the Langley Aerodynamics Division in April 1932,  Freeman argued that boundary-layer control had enormous potential that was being overlooked. His work on airflow around airships had convinced him of this, and he was dismayed to learn that the NACA's only major work on the subject was Bamber's technical report. Freeman considered Carroll's earlier work on the blowing slot in a wing section a step in the right direction, and he outlined a program to continue that research. 22
On the same day that Freeman formally presented his proposal, H.J.E. Reid wrote to headquarters that key staff members agreed on its possibilities. Recommending approval of the proposed research, Reid noted that "it may be advisable to request an extension of Research Authorization No. 201 ... to permit work being carried out in the propeller research tunnel. It will be recalled that the above research authorization authorizes work in the atmospheric tunnel." Why this deviation from the original authorization needed new approval, when Lewis had freely approved R.A. 201 work in the variable-density tunnel and the NACA tank (not to mention departing from the type of work originally prescribed), Reid did not say. Perhaps this looked to him like a major new investigation, which should bear a reaffirmation of R.A. 201 from the outset. Lewis apparently did not share Reid's view, but told the laboratory to draw up a detailed program; when that reached his hands some three months later, he quickly approved it for inclusion in research authorization 201 with no apparent endorsement by the Executive Committee. 23
In essence Freeman proposed to investigate lowering the drag on airships by using boundary-layer control to delay transition. This was truly a new departure in the history of R.A. 201.24 Previous efforts had sought for ways to delay separation and increase the velocity gradient within the boundary layer. Freeman would concentrate on delaying the transition from laminar to turbulent flow. The idea was by no means original with him, but his work on airships and his reading of earlier NACA efforts convinced him that this was a promising line of research and one with which the NACA should be deeply involved. He was right. It would be in this area-though not under this R.A.-that the NACA would make its greatest contribution to boundary layer control, the laminar-flow airfoil.
While Freeman was occupied with this research, over the long stretch between proposal and publication of results, word of his study was abroad in aeronautical circles. One who heard about it was Clark B. Millikan, a young aeronautical engineer at the California Institute of Technology who would in time become one of America's leading aero dynamical theorists. In 1933 he was assistant to Theodore von Karman, Ludwig Prandtl's most famous American protege, and in July of that year he asked the NACA for Freeman's boundary-layer data to use in the work he and von Karman had been conducting for more than a year. George Lewis asked the Langley lab what Miller (chief of the Aerodynamics Division) thought of the request, and Reid replied for the laboratory -[that] the tests made thus far are of a preliminary nature intended mainly to establish the satisfactory working of the equipment and that the results are not of a nature suitable for release by the committee."25 The NACA's long-standing reluctance to share preliminary data with industry, lest they be misinterpreted, was being extended to the scientific community where colleagues customarily shared preliminary results-even negative ones-so long as they advanced the common store of knowledge. This sort of answer, unsatisfactory even to industry, was sure to be doubly unpalatable to scientists.
Some few, however, were privy to the NACA's closed work on boundary-layer control. George Lewis showed charts of Freeman's early results in the propeller-research tunnel to Walter Diehl, a Navy captain who was a prolific contributor to NACA technical publications and for years was the Navy's principal working-level contact with the NACA. Diehl was interested. He believed that wing flaps had largely solved the navy's landing problems but takeoff was still a major difficulty. Some of the  current planes needed as much as 1000 feet in which to take off, a distance that could only increase with increased speeds. Boundary-layer control offered a possible solution to this problem. Diehi reported that an engineer at one of the leading aircraft manufacturers had suggested cooling engines by a blower fan in the wing; this seemed a good source of pressurized air to be released through forward slots. Diehl recommended tests of the idea.26
Freeman replied for the laboratory to Diehl's letter. First he set the captain straight: flaps had not entirely solved landing problems. Lateral control was still a difficulty, especially if the flaps extended the full span of the wing and interfered with the ailerons. But even here, said Freeman, boundary-layer control offered a solution, for it promised a high lift coefficient, elimination of stalling, and a smooth flow conducive to good aileron control at all angles of attack. He reported that data were not yet available on the use of boundary-layer control for improving takeoff, but were expected soon. As to the suggestion by the manufacturing engineer, Freeman treated it with a trace of institutional defensiveness: "The scheme proposed by Mr. Leighton seems entirely practicable. Indeed the idea had been discussed in this office (before we heard of Mr. Leighton's suggestion) as probably the most promising method of boundary-layer control for very large air transports and bombers in which the motors can be placed inside the wing.27
Shortly after this exchange, Freeman submitted his first report on the work he had been doing for more than a year and a half. His memorandum, "Some preliminary results of force tests on a thick stub wing on which the boundary layer was removed by suction arid pressure," dated 25 January 1934, set forth lift results and promised that results on drag would soon follow. Major conclusions were that boundary-layer control to increase lift was "much more favorable than previous model tests have indicated," that separation could be entirely eliminated, that suction was more efficient than blowing, that the power for suction or blowing could be obtained from a throttled engine or a "windmill of practicable dimensions" (à la Leighton), and that the results should be checked on a full-span wing.28
The personnel at Langley were uniformly encouraged by Freeman's report, though they used the cautious, dry language that characterizes engineering correspondence: their comments ranged from "rather interesting" to "most promising." But beneath the restrained wording was clear evidence of excitement. One man suggested sending the report to Lewis, since it revealed why previous tests at the laboratory had been unproductive: the slots had been too small. Another engineer expected that drag would be no problem. Two others had schemes to run the blowers off the propellers; this suggestion led Freeman to alter his plans and run more tests in the propeller research tunnel before proceeding to full-scale in flight tests. Reid sent all this material to Lewis (save Freeman's last reservations) and-in what was becoming laboratory style for this research authorization-recommended getting more complete results before considering the report for publication.29
Before answering this correspondence, Lewis discussed it with the staff during one of his frequent visits to the laboratory. The conferees approved Freeman's proposal with one significant alteration: Freeman wanted to run the tests on a symmetrical airfoil; one shaped the same on the top and the bottom. This was not surprising, given his preference for research using theoretically satisfactory shapes, like the body-of-revolution offered by the airship model used in his earlier work. This preference was in tune with current theoretical literature and presumably would give results applicable to all airfoils. Lewis and the Langley staff, however, insisted that Freeman use a NACA 2415 airfoil, a slightly cambered shape from a family of NACA wing sections just then achieving promising results in lift/drag tests at Langley. For several years the NACA had been running exhaustive tests on families of wings whose design components-thickness, camber, taper, etc.-were minutely altered for each succeeding wing to  document the change in flight characteristics. This was turning out to be among the most popular and most useful research conducted at Langley, for it gave aircraft designers a whole range of wings from which to choose, as one might select home-furnishing or automobile accessories from a catalogue. Wings thus developed in the laboratory became known by the Committee's name with a number code identifying the features of the wing. The most famous was the "two-thirty" family of wings, introduced in l935.30
 Exactly, why Lewis arid the Langley staff forced Freeman to use the 2415 wing section, the written record does not say. If Lewis wanted merely to ensure mention of a NACA airfoil in Freeman's published results, he could have prescribed any one of a number of Committee-developed symmetrical sections. Lewis may have wanted to spotlight the 24 groups of NACA airfoil sections, just then being touted by the Commit-tee as superior to the Clark Y and the R.A.F. 6, two of the most popular airfoils of the time. Whatever the reason, the decision seems to have been purely political, an instance where Lewis allowed his own judgment about the best interests of the Committee to overrule the judgment of the researcher in the laboratory. That Lewis chose to reach this conclusion orally with the Langley staff, rather than to commit his reasons to writing, reinforces this impression.31
Like a good soldier, Freeman did as he was told, bringing to the new experiments the same enthusiasm and creativity that had marked his entrance into this field of research. Shortly after selection of the slightly cambered NACA 2415 wing for the tests, Freeman suggested a new slot design to improve the characteristics of such wings at low angles of attack. He proposed a connection between the front bottom of the wing and the rear top. Natural pressures of the airflow at low angles of attack would, according to Freeman, suck air into the wing at the top rear and in turn suck the same air out of the wing at the front bottom, thus moving the boundary layer across the top. Several months later Freeman changed the proposal to put the top intake near midchord instead of at the rear of the wing. In the meantime, he also suggested that the boundary layer might be controlled by adding to the trailing edge of a wing a retractable flap of adjacent tubes, using the Venturi effect to draw air into the tubes and pull it further aft. Freeman thought the latter idea so promising that the govern-ment might want to consider a patent.32
To Freeman's first suggestion, Lewis gave quick assent for inclusion in research authorization 201. But the notion of the Venturi flap drew a more cautious, more revealing response: First Lewis observed that, at the ninth annual NACA industry conference held recently at Langley, many considered the demonstration of boundary-layer control in the smoke-flow tunnel and the charts illustrating the results of this investigation to be the most interesting exhibition. Furthermore, the Navy Bureau of Aeronautics had expressed interest, so the idea was certainly worth pursuing. But Lewis was reluctant to continue personally evaluating every new departure in the program without more staff work at the laboratory:
Perhaps Lewis was just overworked. Perhaps the technicalities of boundary-layer control were simply becoming too much for him. Perhaps the Langley laboratory whose professional staff had more than doubled since it first proposed boundary-layer research-had simply grown too large to handle on a personal basis. Whatever his reason, Lewis was directing the laboratory to formalize its procedures for administering boundary-layer research, and in the process was giving up to the laboratory some of his autonomy. He approved Freeman's suggestion, just as he had approved all those before it; if the Langley staff were going to speak with one voice on future recommendations, he would be even less likely to override their collective judgment.
Lewis's delegation of authority at this time, his loosening of the reins on the Langley staff, should not be taken out of context. It was no more than an extension of the policy he had followed from the outset of his NACA career. He had always insisted upon the freest discussion and the most open flow of ideas within the Langley staff. He  distrusted organizational arrangements that hampered interdisciplinary and interdivision collaboration. When he visited the laboratory, as he did often in the early years, he convened informal staff meetings at which everyone was encouraged to present his views. Lewis fostered, and the laboratory ensured, informal discussions where rank and protocol mattered less than the worth of the ideas. The most junior engineer could corner his division chief in the cafeteria and argue a case over lunch without fear of overstepping bounds. In setting up a special committee on boundary-layer control, Lewis was trying to guarantee the continuation of this kind of interplay and cooperation even if he were unable to personally supervise and participate. By doing so, he was also laying the groundwork for the collaboration that would lead to the laminar-flow airfoil.34
Evidence of Lewis's increasing workload and his need to delegate responsibility came the following month when P. E. Hemke of the Case School of Applied Science wrote to Lewis asking for boundary-layer information. He was conducting some wind-tunnel tests of boundary-layer control and had learned of the NACA's work through the last industry-conference report. Could the NACA advise him on the best positioning of slots, the best method for achieving even airflow distribution along every slot, and the best wing thickness? Lewis, instead of forwarding this to Langley lab for a draft reply, as he would have done in previous years, let John Victory handle the correspondence.35
This subtle shift did not mean that Lewis had relinquished his final say. In fact, Lewis's intervention in Freeman's wind-tunnel program was about to bear fruit. Freeman reported disappointing results in the tests on the NACA 2415 wing, "as was expected with the use of such a low cambered wing section," he added somewhat acidly, in an "I-told-you-so" tone. He recommended that the tests continue on a more highly cambered and tapered wing, apparently believing that, short of the symmetrical wing section he preferred, he would get best results with a section of greater camber than the 2415 preferred by George Lewis. Others in the division agreed, though Eastman Jacobs-one of the most brilliant and influential men ever to work at Langley-thought that "possibly Freeman has been a little hasty in condemning the slightly cambered airfoil." Still, he agreed with Freeman on the need for a more tapered wing, and the two men selected a satisfactory shape. H.J.E. Reid, trying to proceed as he imagined Lewis would wish, reported that the program would proceed with tests on something like the NACA 8318 airfoil unless disapproved by Lewis, and he added that "the future program will be planned to show the value of boundary-layer control in take-off." Silence from Lewis was interpreted as assent.36
Keeping the military services happy was not Lewis's only concern; industry too had become interested in boundary-layer control and was increasingly difficult to put off. Eclipse Aviation Company, for example, learned of NACA research on boundary-layer control and wanted to know if it was too early to consider manufacturing a power supply for the blowers to be used in wings for suction or blowing. Freeman thought this might be an excellent chance for the laboratory to get a prototype manufactured free for testing, but more conservative voices at Langley prevailed; Eclipse Aviation was finally told that the requirements were not clear enough for manufacturing. 37
Of greater concern to Lewis was a request from the Northrup Company, which in early 1935 was having boundary-layer control tests conducted at the California Institute of Technology. Lewis discussed this and other boundary-layer research in the United States with Donald H. Wood when the latter visited headquarters from his post in the Aerodynamics Division at Langley. Would it be possible, Lewis wanted to know, to publish some results of the work already done at Langley and continue the testing on an actual airplane? As Wood made clear when he returned to Langley,38 airplane tests had been proposed by Freeman more than a year previously when Lewis intervened and insisted on tunnel tests on an NACA airfoil." It appears now," lamented Wood,  "that airplane tests would be very useful in establishing the priority of our investigations but it is a pity that this was not realized a year ago when the tests were suggested here."
This lost opportunity prompted Wood to examine how the boundary-layer research program at Langley had been conducted. e noted (as had been suggested elsewhere) "the work on the general project has not been pushed sufficiently," a failure he attributed to shortage of personnel and the press of "other projects deemed of equal or greater importance." Less forgivable was the "constant shuffling about of personnel in the drafting room and shops to work on projects of momentary and changing first importance." Compounding these shortcomings in the laboratory was the premature announcement of research programs at the annual industry conferences. "The fact that we announced results of incomplete tests at the last mfg conference," he concluded, "has stimulated interest and the fact that we have published nothing now puts us in an embarassing situation," one that "will continue ... so long as we continue to give out advance information each year." The only way out of the present dilemma, he believed, was to override Freeman's reticence and get something into print. "I know that Mr. Freeman is somewhat adverse to putting out information on the inconclusive tests so far made," argued Wood, "but I think that under the circumstances it might be well to get out a confidential note on the results obtained to date. This would place us on record and give Northrup a starting point for his tests which I don't think he would misuse."
The chief of the Aerodynamics Division agreed with Wood, though he doubted that design of an airplane wing could begin until tunnel tests were completed. So, while the work proceeded apace, Freeman prepared "Large-Scale Boundary-Layer Control Tests on Two Wings in the N.A.C.A 20-foot Wind Tunnel" as a Confidential Memorandum Report.39 In forwarding this report to headquarters, Reid advised that it had not been edited and was not intended for wide circulation in its present form. Its contents would be included with other material in a future Technical Report.40
Outsiders from industry, the services, and academia were not the only ones inquiring after the progress of boundary-layer research at Langley and thereby affecting the course of the research program. For example, Charles H. Helms, a headquarters aeronautical engineer specializing in advanced design studies, patents, and inventions, offered two ideas for boundary-layer control to Lewis, who sent them along to Langley for comment. Helms suggested an endless belt along the upper surface of a wing to keep the boundary layer moving at the speed of the airstream, and he suggested vibration to shake the boundary layer loose. Eastman Jacobs responded with a perfunctory "no comment." Freeman replied that the endless-belt idea, which had been patented in Germany in 1917, was impractical, while the vibration technique was "like attempting to lose one's shadow. No matter how quickly the surface is moved by vibrating it, the air is forced to follow."41
This exchange would not have affected research authorization 201 except that it involved Helms actively in the program. Two months after Helms made his suggestions, Lewis asked him to comment on a new idea of Freeman's. Prompted by a private individual's expression of support for the Venturi effect as a means of controlling boundary layer, Freeman dusted off his proposal of the previous year and sent Lewis an expanded version of it for approval as part of research authorization 201. Helms's comment was that Langley should be aiming at reduced profile drag* as the real product of boundary-layer control, not at increased lift as envisioned by Freeman. Said Helms, "If there is anything to boundary layer control, and I think there is, we should be the ones to lead the way, even to the point of actually applying it. I have been of the  opinion for a long time that in this particular phenomenon is the graveyard of all slots, slits, slats, auxiliary air foils, and flaps." 42
Whatever the virtues of this appraisal, it was sufficiently at odds with the Freeman proposal to place Lewis between conflicting technical recommendations. He sent Helms's comments to Langley for the staff's reaction before approving the Venturi research.43
In the meantime, however, Freeman had drafted an entirely new proposal to bring order out of the chaos engulfing boundary-layer research at Langley. Freeman took time to express agreement in principle with Helms while lecturing him on the technical inaccuracies of his analysis.44 but past deeds seemed unimportant now, for Freeman was caught up in a drive to have his rationalized research program in boundary-layer control approved at both the laboratory and headquarters.
Freeman's proposal went to the chief of the Aerodynamics Division in a memorandum dated 5 August 1935. In it, Freeman noted that most drag on aircraft is skin friction. Boundary-layer phenomena influence skin friction. Two types of boundary layers are known, laminar and turbulent, but little is known of the transition from one to the other. It had been proposed three years earlier to study thick wings, but the sections used were found unsuitable. Then, said Freeman, he had suggested a study of boundary-layer phenomena about an airship hull, which seems to have been a ploy to get around Lewis's insistence on a NACA airfoil and to work with a more theoretically satisfactory shape. This last proposal had even been approved by the Subcommittee on Airships, but was finally dropped because of the "stigma" attached to these craft, or so Freeman guessed. Now he thought it was time for a comprehensive approach along all the most promising lines. He recommended three major areas of boundary-layer research: conventional and very thick wings, effects of surface texture, and effects of surface lubrication-i.e., oils and soaps.
As Lewis had requested the previous year in dealing with proposals, Reid called a conference of the leading aerodynamicists at Langley to evaluate Freeman's new proposal and make a report to headquarters. Besides Wood, Miller, Jacobs, Reid, and Freeman-the men primarily involved in boundary-layer research at Langley-the con-ferees included Theodore Theodorsen, a theoretician comparable in position (though not in personality) to Max Munk, and Albert E. von Doenhoff, a young engineer on the verge of a major role in boundary-layer research at Langley. The conferees agreed in principle with Freeman's proposal and explicitly endorsed his suggested use of a symmetrical wing, revealing wide opposition at the laboratory to Lewis's insistence on a NACA 2415 airfoil. They did not feel, however, that research on surface texture and lubrication was of primary importance; rather, they recommended more promising avenues of study. Von Doenhoffs proposal for smoke-tunnel tests of boundary layer would be pursued unless it duplicated work at the Massachusetts Institute of Technology or elsewhere, or unless Hugh L. Dryden, head of the National Bureau of Standardís boundary-layer research program, thought it ill advised. Dryden's laboratory had been conducting sophisticated research on measurements of fluid flow about a solid body, some of it under contract to the NACA, and his opinion was highly valued, not just at Langley but also throughout the aeronautical community in the United States and abroad. The conferees further agreed that boundary-layer research should aim at high lift, and that several high-lift devices should be tested in the Variable Density Wind Tunnel. Whereas this plan seemed to contradict Helms's recommendation in his critique of Freeman, it was actually (as Jacobs pointed out) the opposite side of the same medal. As Reid reported Jacobs's thoughts on the subject-perhaps in a conscious effort to appease Helms-he introduced what would become a turning point in boundary-layer research at Langley:
The memorandum Jacobs turned in six days later may properly be called the result of Freeman's dissatisfaction with the pathless ness of work on research authorization 201, Helmís criticism of the pursuit of lift instead of drag reduction, the independent work Jacobs had been doing under another research authorization, and finally Jacobs's own genius for synthesis and conceptualization. He had found that increased wing loading of a "normal airfoil" produced "surprisingly large" increases in speed. He hypothesized seven reasons for this, some of which he felt had been neglected. The reasons ranged from the transparently logical to the seemingly incongruous. Smaller wings, for example, would clearly result in reduced wing-surface cover weight. But the argument that higher speed would result in fuel-weight savings sounds to the uninitiated like hurrying up to get there before the gas runs out.46
In addition to the favorable features of increased wing loading, Jacobs saw some unfavorable ones. For example, structural weight increases tended to result from shortening wings while maintaining the cross-sectional proportion." This may be avoided," suggested Jacobs, "by the use of thicker sections, but the analysis has shown that the change to thicker sections is usually not justifiable owing to their higher drag. In fact one of the most important results of the analysis to date is the tentative conclusion that the sections in common use on cantilever wings are too thick." It was  indeed one of the most important conclusions, for from it would flow in time the low-drag airfoil and a radical shift in emphasis in boundary-layer research at Langley.
Jacobs went on to suggest rethinking some of the conventional wisdom about aircraft designs in view of the potentials of increased wing loading. Aircraft of higher wing loading required faster landing speeds, but perhaps airfields should be designed for aircraft, not aircraft for airfields. Aircraft design should anticipate high-altitude flying, for the difficulties involved in climbing and descending seems outweighed by the advantages, and other difficulties appeared negligible. If high-altitude flight were the goal, work on turbo superchargers and more powerful engines would be required first. Jacobs was in fact calling for a major reexamination of the assumptions underlying contemporary aeronautical research. To pursue the promising leads already in hand, he insisted on more than the $300 currently allotted to his work.
Jacobs's memorandum was forwarded to Lewis together with the report of the conference that prompted it. Lewis seems to have been overwhelmed. The laboratory was speaking in several voices, and some were not as clear as he might have wished. He sent the whole corpus back to Langley for yet another conference, this time to reach a consensus on the next step in boundary-layer research. This time Miller, Jacobs, and Reid of the first conference were joined by three different engineers in managerial positions; absent were the junior engineers actively engaged in the program. This more senior group concluded that, while boundary-layer control would produce no great savings in drag at high speed, there just wasn't enough knowledge to justify a conclusion on friction drag. The proper course, therefore, was to proceed beyond the models and wing sections already tested and to experiment with a 15-foot-chord wing in the full-scale wind tunnel. This, they hoped, would give them data on friction drag around zero lift with high Reynolds number: that is, with close correlation to actual flight conditions. Presumably they would thus gain a better idea of the most promising path of research.47
Data already available had begun to produce publishable results. For example, young von Doenhoffs report "An Application of the von Karman-Millikan Laminar Boundary-Layer Theory and Comparison with Experiment" reached headquarters for publication as a technical note just before the second committee met at Langley to decide on the future course of boundary-layer research. But the heart of Research Authorization 201-the work being done by Freeman-was still withheld from publication. Even when Edward P. Warner, formerly of the Langley staff and a current member of the NACA itself, asked to use Freeman's confidential memorandum report of the previous year in revising his textbook Airplane Design-Aerodynamics, Freeman still balked. Lewis insisted that NACA publish the report in the open literature, as a Technical Note or a Technical Report, before it turned up as a citation in a secondary source, but Langley objected, saying that Freeman's results should not be published until they had been checked in tests for tunnel blocking: i.e., to see if the presence of the model in the wind tunnel created variations in the wind pattern that would undermine the validity of the findings. These tests, said Langley, could not be completed in time to meet Warner's deadline. Warner, one of the most knowledgeable men in the field of aeronautics at the time, was presumably competent to judge the proper use of Freeman's preliminary results; but those results were once more withheld from publication in an attempt to further refine and check them. It turned out in the blocking tests that, though "the presence of a lifting body in the airstream modified the distribution of velocity in the test section, and thereby changed the tunnel calibration obtained with the tunnel empty," the change was less than 3 percent and could be ignored.48
After almost ten years, research authorization 201 was becoming a classic example of normal research-the kind most often conducted but too seldom reported. It was in sum a rather pathless excursion through an important field. While everyone attested to  the potential of the investigation, no one seemed entirely clear as to where it should go or how it should get there. Instead, different avenues of attack were followed simultaneously. New results led to refinements of the program or new lines of research. Most often these were suggested by the staff at Langley (usually junior members), discussed at laboratory conferences, and referred directly to Lewis for approval by him alone. About halfway through the life of the authorization, the results had been disappointing and the future was cloudy. In June of 1936, Smith J. DeFrance reported on a conference with Freeman and Eastman Jacobs:
Such a conclusion is hard to argue with, and George Lewis did not: he quickly approved it.50 By the same token, it represents no advance in the state of the art after ten years of work. Surely the laboratory was now trying to look at the forest, but ten years amongst the trees had not done much for the researchers' perspective, and Lewis seems simply to have been rubber-stamping their recommendations.
Even though what Lewis approved was not really a program, new lines of attack did emerge from it. For example, von Doenhoff visited Dryden at the National Bureau of Standards to learn how to measure mean air speeds over a solid surface with a hotwire anemometer, a technique pioneered by Dryden and his staff. And Jacobs reported in July 1936 the conclusion of another Langley staff conference that "adequate systematic investigation [of the boundary layer] requires the construction of special wind-tunnel equipment like the proposed 2-dimensional flow tunnel."51 This endorsement added weight to the growing demand for a low-turbulence tunnel and brought closer the research that would finally break the NACA through the boundary-layer research impasse.
Some results began to appear, though in the same old pattern: Freeman's report on "Boundary-Layer-Control Tests of a Tapered Wing in the N.A.C.A. 20-Foot Wind Tunnel," originally planned as a Technical Report, was (according to Reid) "too incomplete and too inconclusive."52 It was not to be published or released to manufacturers, but distributed only to the armed services as a confidential memorandum report. Von Doenhoff was characteristically more open with results submitted the following year (1937) in "Notes on a Preliminary Investigation of Boundary-Layer Transition along a Flat Plate with Adverse Pressure Gradient." He asked that a copy be forwarded to Dryden for comment, with a view to publication. Dryden recommended its publication as a Technical Report, though he cautioned that part of the discussion should be presented less dogmatically, "to convey the idea of a stimulating speculation rather than that of an established theory" for computing scale effect on maximum lift. Von Doenhoff complied, and the report appeared as a Technical Note three months later.53
In spite of von Doenhoffs example, Langley still tended to suppress less-than final results obtained under research authorization 201. A glaring instance occurred in February 1938, when Clark B. Millikan tried again to obtain some preliminary results. Millikan wrote to Lewis that he had read in the Committee's 23d annual report about Langley's boundary-layer control work and felt that the results would be useful to him and his staff at Cal Tech. Alive to the fact that the NACA results might still be inconclusive, Millikan wanted to use them as a guide to keep from plowing the same ground. "It would be very valuable to us," he told Lewis, "if we could have the benefit  of some of your experience before starting our program, so that we need not repeat tests for which you have already found answers."54
Even the normally reticent Freeman found this request persuasive, and he recommended that his last two confidential memorandum reports should be released to Millikan. Another engineer objected, however, noting that CMRs were normally released only to manufacturers and he saw no reason "to forget this rule in the present case." And others at Langley agreed with him, even though Millikan had freely provided information about his program in his letter to Lewis and even though the theoretical work of Millikan and von K·rm·n had been the foundation of von Doenhoffs research on laminar flow. The laboratory's answer to Lewis seems to have completely missed the reasoning behind Millikan's request:
If the laboratory staffers were still trying to chastise Lewis for insisting on the NACA airfoil for Freeman's original tests instead of the more theoretically satisfactory symmetrical wing, they were cutting off their noses to spite their faces. Millikan was not after final results. To deny him the benefit of their experience was to alienate people who would prove important not only to aeronautics, but to the reputation of the NACA as well.
But these early vagaries of research authorization 201 would soon be overshadowed by a breakthrough that changed irrevocably the course of boundary-layer control research at Langley. In the very month that the laboratory's new low-turbulence wind tunnel went into operation-the tunnel for which Jacobs and von Doenhoff had been arguing for years-Jacobs reported to Reid that the most promising future research would be in an almost entirely unanticipated direction. " We can now conclude definitely," he wrote, "that the most likely form of boundary-layer control to reduce drag is through the use of the flow conditions and pressures ordinarily attainable over the section through changes of the section shape to provide the desired control to maintain laminar flow."56 It was not to be sucking or blowing, then, as presupposed in research authorization 201, that would most successfully control the boundary layer, but wing shape; furthermore, the greatest advantage would be derived from maintaining laminar flow over as much of the upper wing surface as possible and not (as suggested earlier) from energizing or moving the boundary layer itself. Verification of these revolutionary conclusions was made possible by the low-turbulence tunnel, where Jacobs had reduced drag 30 percent and maintained laminar flow to 75 percent of the chord of the wing behind the leading edge. Much work remained to be done, and he would need a new job order to proceed.
Lewis seems not to have immediately appreciated the importance of this breakthrough at first, for he approved Jacobs's request only on condition that it not interfere with icing research scheduled for July and August in the low-turbulence tunnel.57 Icing of wings and control surfaces on aircraft was just then one of the most important problems the NACA had to solve for commercial operators and the armed services, and Lewis wanted usable results as soon as possible. This icing research had been a major justification for building the low-turbulence tunnel in the first place, and good politics would require using it for that purpose, at least at the outset. In spite of these limitations, Jacobs found sufficient tunnel time to follow up on the laminar-flow research.
The impact of this discovery on research authorization 201 was swift and dramatic. It seems to have killed off interest in Freeman's work and presaged Freeman's  departure from the NACA within a few months. Shortly after Jacobs's discovery was announced, Langley had prepared a new proposal by Freeman to study boundary-layer control on bodies of revolution, but by the end of the year Langley withdrew the proposal on the grounds that "increased knowledge of boundary-layer conditions since this letter was written indicates that the proposed program would hardly be worthwhile."58
Publication policy also began to change. Langley was now willing to give wider circulation to Freeman's 1936 report, perhaps in the belief that further research in that area seemed unlikely. At the same time von Doenhoff began publishing in the newer field of laminar-flow research. His first report, "A Method of Rapidly Estimating the Position of the Laminar Separation Point," was sent to headquarters for publication as a Technical Note within three months of Jacobs's memo.59 Others followed less rapidly, but upon their appearance reversed the procedure used in Freeman's case early in the 1930s. Now, the first results published were on a symmetrical airfoil in a low-turbulence tunnel. Only after those theoretically satisfactory results were printed did the NACA begin issuing data on a family of cambered airfoils, the new laminar flow or low-drag wings. In this research, von Doenhoff was joined by names new to research authorization 201. By 1942, these experiments had graduated into flight tests, and the first practical application-a low-drag wing on an operational aircraft-was already being used on the P-51 Mustang.60 The performance of that aircraft in World War II was one of the gems in the NACA diadem, an example ever after of the contribution of NACA research not only to the advance of American aeronautics but also to the winning of World War II.
But then von Doenhoff began following the same course Freeman had taken almost a decade before, recommending changes in the research program. In fact von Doenhoff proposed to study the very problem for which research authorization 201 was opened in the first place: whether blowing or suction could be used to control boundary layers, on the surfaces of wings as well as internally in ducts and passages.61
Times, however, had changed. There was a war on and George Lewis was reluc-tant to approve new proposals as he had during the 1930s. He told Langley that von Doenhoffs suggestion would be referred to the next meeting of the Aerodynamics Committee, but four months went by without any action. H.J.E. Reid finally wrote to Lewis asking about the proposal, advising the director that "the Laboratory has already initiated work on this job pending approval of this project by your office."62 Apparently Lewis's rubber-stamping of Langley proposals during the 1930s had bred in the laboratory a habit of autonomy that considered approval by Washington a mere bureaucratic routine.
Lewis was working in a new atmosphere, however, and could not accept the old justifications. "It would be desirable," he advised the laboratory, "if the proposed investigation of turbulent boundary layers could be conducted in connection with some specific project having a direct bearing on applications to wings or duct designs in preference to a long-range study such as has been proposed by the laboratory." In essence Lewis was saying there could be no more fundamental research for the duration. All NACA effort must contribute to the war effort; if basic research was authorized, it would have to show promise of practical application to the war. Reid met Lewis's demand with a bromide so general as to be virtually meaningless. In proposing to conduct the work under research authorization 201, he assured Lewis that the research "is of a fundamental character and the results will be applicable to current problems relative to military aircraft both to wing development and to ducting problems as well." Couched in those terms, the research was quickly approved by Lewis, apparently without reference to the Aerodynamics Committee.63
Even the reports being generated under research authorization 201 had to be oriented to practical applications. When a report by von Doenhoff and another  engineer on "Determination of General Relations for the Behavior of Turbulent Boundary Layers" went to headquarters early in 1943, it was returned with the recommendation of a staff engineer that it be expanded to show practical application. Reid replied that more research was required before the practical applications could be determined, indicating that (at least in this instance) the laboratory continued to do fundamental research even as headquarters was insisting on practical results.64
This demand by headquarters for practical results may account for the diminished level of work done under research authorization 201 in the final years of the war. Although work did continue during 1943 and 1944, it was on a comparatively reduced scale, producing in the latter year only one advanced confidential report and one confidential bulletin. A representative of Lockheed Aircraft Corporation who visited Langley in January 1945 was told by the staff that "although we are very interested in boundary layer control we have no new data on the subject."65
As World War II drew to a close, Lewis reported from Washington a "revival of interest in boundary layer control as a means of reducing landing speeds." He asked the Langley staff about reissuing some of the old classified reports. A committee at the laboratory had already addressed that question and concluded that Freeman's 1935 report was worth reissuing, but the 1936 report was not. Nor was more enthusiasm shown when Hugh Dryden reported from Europe on the German work in boundary-layer control during the war. The staff at Langley could find nothing new in his report, even though Dryden (surely no novice in the field) found much of interest.66 With Technical Note 1007 (the reprint of Freeman's 1935 report) published in January 1946, the Langley laboratory closed the book on research authorization 201. The laboratory would do more work on boundary-layer control but not under this research authorization-which, after all, had been opened nearly twenty years earlier to compare the suction method of the Göttingen aerodynamicists with the blowing method of Richard Katzmayr.
Whatever the achievements and shortcomings of research authorization 201, its history provides a classic example of the research process, with implications far beyond the details of how the NACA conducted aeronautical research in the second quarter of the 20th century. Among the many lessons to be learned here, a few are particularly clear and poignant. George Lewis attempted-with diminishing success as time went on-to maintain personal control over the research program at Langley and to channel it in directions of political use to the NACA and immediate practical use to its clients. The Langley staff displayed a continuing reluctance to publish preliminary results or even to share them with knowledgeable colleagues like Clark Millikan and Edward P. Warner. The Langley laboratory maintained an openness to new ideas and suggestions, even from junior staff members, that seemed so extreme at times as to make of RA 201 a rudderless craft with too many hands at the sheets, precisely the problem envisioned by Max Munk when he chastened E. G. Reid at the outset. Eastman Jacobs's discovery of boundary-layer control through modification of airfoil shape illustrates the serendipitous nature of research, and the way in which one line of investigation will often lead to more fruitful conclusions than those anticipated. It also shows how a revolutionary piece of research equipment-in this case, the low-turbulence wind tunnel-can dramatically alters the course of research. Finally, it should be noted that boundary-layer control is still an intriguing and elusive technique that attracts and frustrates aerodynamic researchers even 35 years after the close of research authorization 201. If the research under that authorization left a trail that now appears aimless and confused, it was for that very reason typical of most research into the unknown.
* Profile drag is friction drag plus drag due to separation.
1. The example of boundary-layer depth is drawn from John D. Anderson, Jr., Introduction to Flight: Its Engineering and History (New York: McGraw-Hill, 1978). pp. 123-24. I have drawn heavily on this excellent volume in preparing the discussion of the boundary layer. I have also profited greatly from the following works: Joseph Flatt, "The History of Boundary Layer Control Research in the United States," in G.V. Lachmann, ed., Boundary Layer and Flow Control: Its Principles and Application (2 vols.; New York, Pergamon Press, 1961), I, pp. 122-43; Hugh L. Dryden, "Exploring the Fundamentals of Aerodynamics," Journal of the Washington Academy of Sciences 37 (15 May 1947), 145-56; Neal Tetervin, "A Review of Boundary Layer Literature," NACA Technical Note 1384 (1947); and H. Schlicting, "Some Developments in Boundary Layer Research in the Past Thirty Years," Journal of the Royal Aeronautical Society 64 (Feb. 1960), 64-79.
2. Hugh L. Dryden, "Fifty Years of Boundary-Layer Theory and Experiment," Science 121 (18 Mar. 1955), 375-80. One reason for choosing "boundary layer" over "transition layer" was that transition came to be the preferred term to describe the change from laminar to turbulent flow.
3. The quote is from Anderson, Introduction to Flight, p. 118. The discussion here refers to in-compressible flow, the kind experienced by an airplane traveling below the speed of sound. During most of the life of research authorization 201, even the air velocity over wings seldom exceeded the speed of sound.
4. Reid to engineer-in-charge, 2 Nov. 1926; Ide to George W. Lewis, 22 Sept. 1926; H. Lee Dickinson to Walter Bonney, "The Katzmayr Effect," 25 July 1956.
5. Engineer-in-charge to Munk, 3 Nov. 1926; Munk to engineer-in-charge, 4 Nov. 1926. On the importance of Munk, see Joseph Sweetman Ames, "A Resume of the Advances in Theoretical Aeronautics Made by Max M. Munk," NACA Report 213 (1925).
6. Lewis to Langley Memorial Aeronautical Laboratory (hereafter LMAL), 11 Nov. 1926. Evidence that this memo was prompted by Reid's attempt to present the idea directly to the Aerodynamics Committee is in Munk's memo, "Recommendation for new research," 16 Nov. 1926.
7. Munk, "Recommendation for new research."
8. Lewis to LMAL, 3 Dec. 1926; E.S. Land to NACA. 2 Dec. 1926.
9. Lewis to LMAL, 6 Dec. 1926.
10. J.W. Crowley to engineer-in-charge, 14 Dec. 1926; A. J. Fairbanks to engineer-in-charge, 10 Dec. 1926; George J. Higgins to H.J.E. Reid, 10 Dec. 1926; Thomas Carroll to H.J.E. Reid, 10 Dec. 1926.
11. Ide to NACA, 8 Dec. 1926; Crowley to H.J.E. Reid, 17 Jan. 1927; Katzmayr to Ide, 21 May 1927; Lewis to LMAL, 8 June 1927; H.J.E. Reid to NACA, 15 June 1927.
12. Munk to Lewis, "(Through official channnels)," 29 Jan. 1927; H.J.E. Reid to NACA, 31 Jan. 1927; Lewis to LMAL, 4 Feb. 1927; E.G. Reid to engineer-in-charge, 14 Feb. 1927; H.J.E. Reid to Munk, 3 March 1927, with Reid's subscript of 22 March 1927.
13. Max Sherberg to engineer-in-charge, 19 Jan. 1927; Lewis to LMAL, 19 Feb. 1927; George J. Higgins, Eastman Jacobs, and J.M. Shoemaker to engineer-in-charge, 15 Feb. 1927; Lewis to LMAL, 2 March 1927.
14. Thomas Carroll, "Preliminary Flight Tests of a Method of Boundary Layer Removal," 2 Sept. 1927; John W. Crowley, Jr. to H.J.E. Reid, undated; Reid to NACA, 10 Sept. 1927. Even though RA 201 stated that the suction technique was to be tested only in the wind tunnel, the first experiment run by the lab was a flight test.
15. J.S. McDonnell, Jr. to LMAL, 10 Oct. 1927; H.J.E. Reid to NACA, 14 Oct. 1927.
16. Lewis to LMAL, 23 Jan. 1928.
17. H.J.E. Reid to Lewis, 19 Jan. 1928; Lewis to LMAL, 23 Jan. 1928; Reid to NACA, 15 March 1929; Reid to NACA, 27 Aug. 1929.
18. Starr Truscott to Lewis, 25 June 1928; Lewis to LMAL, 2 July 1928, forwarding Karoku Wada, "Some Experiments on the Feathered Wing"; Reid to NACA, 9 March 1929; Thomas Carroll to Reid, 11 March 1929; Lewis to LMAL, 20 March 1929; Reid to NACA, 10 Sept. 1935. The chief test pilot referred to in this last letter was not Thomas Carroll but his successor, Melvin Gough.
19. Reid to NACA, 23 Aug. 1929; Reid to NACA, 1 Dec. 1930. An earlier report, Elliot G. Reid and M.J. Bamber, "Preliminary Investigation on Boundary Layer Control by Means of Suction and Pressure with the U.S.A. 27 Airfoil," NACA TN-286 (1928), was apparently the result of work done under a different research authorization.
20. Reid to NACA, 31 March 1931, forwarding 1.H. Abbott, "Experiments with an Airfoil Model on Which the Boundary Layer Is Controlled without the Use of Supplementary Equipment"; Hugh B. Freeman, "Preliminary Report of the Measurement of Pressure Distribution on the ZRS-4 Airship Model," dated 25 Nov. 1931. Freeman observed in "Pressure-Distribution Measurements of the Hull and Fins of a 1/40-Scale Model of the U.S. Airship 'Akron,'" TR-443 (1932), that "experimental pressure-distribution results are ... useful . . . indirectly, in computing the frictional forces on the surface of the hull." See also Hugh B. Freeman, "Measurements of Flow in the Boundary Layer of a 1/40-Scale Model of the U.S. Airship 'Akron,' " TR-430 (1932), which resulted from the same experiments.
21. Starr Truscott to engineer-in-charge, 5 April 1932; see also the correspondence between the NACA and the Bureau of Aeronautics between Dec. 1932 and March 1933, leading up to Lewis to LMAL, 5 May 1933.
22. Freeman to chief, Aerodynamics Division, 18 April 1932.
23. Reid to NACA, 18 April 1932; Lewis to LMAL, undated; Freeman to chief, Aerodynamics Division, 6 July 1932; Reid to NACA, 12 July 1932; Lewis to LMAL, 18 July 1932.
24. Eastman N. Jacobs to engineer-in-charge, "Notes on the history of the development of the laminar-flow airfoils and on the range of shapes included," 27 Dec. 1938.
25. Millikan to Lewis, 24 July 1933; Lewis to LMAL, 28 July 1933; Reid to NACA, 2 Aug. 1933.
26. Lewis to LMAL, 7 Nov. 1933.
27. Reid to NACA, 14 Nov. 1933.
28. Freeman to engineer-in-charge, 25 Jan. 1934.
29. Smith J. DeFrance to Elton W. Miller, undated; Donald H. Wood to Miller, 21 Dec. 1933; John W. Crowley, Jr., to Miller [ca. 28 Dec. 1933]; Fred E. Weick to Miller, 9 Jan. 1934; Freeman to Miller, 2 Jan. 1934; Reid to NACA, 25 Jan. 1934.
30. Lewis to LMAL, 5 April 1934; Reid to NACA, 13 April 1934. On the NACA families of airfoils, see George W. Gray, Frontiers of Flight: The Story of NACA Research (New York: Alfred A. Knopf, 1948), pp. 98-112. What the NACA was actually testing at the time were airfoil sections: i.e., cross-sections of wings and other airfoils cut from front to rear. In common parlance, however, many of the NACA engineers would refer to the section as simply an airfoil. On the subject of the NACA 2415, for example, the classic report (Ira H. Abbott, Albert E. von Doenhoff, and Louis S. Stivers, Jr., "Summary of Airfoil Data," TR-824 (1945)) says "the NACA 2415 airfoil has a 2-percent camber at 0.4 of the cord from the leading edge and is 15 percent [of the cord] thick."
The reasoning behind the NACA program to develop families of airfoil sections was revealed in Eastman N. Jacobs, Kenneth E. War, and Robert M. Pinkerton, "The Characteristics of 78 Related Airfoil Sections from Tests in the Variable-Density Wind Tunnel," TR-460 (1933):
The NACA intended to remedy this shortcoming by developing its own family of airfoils based on tests in the variable-density wind tunnel, where high Reynolds numbers could be achieved.
31. The Annual Report for 1933 cited an investigation then under way on airfoil shapes:
An earlier technical note had reported that slightly cambered airfoils like the 24 series were superior to comparable symmetrical wings (Eastman N. Jacobs and Kenneth E. Ward, "Tests of N.A.C.A. Airfoils in the Variable Density Wind Tunnel: Series 24," TN-404 [19321) and another report two years later found a 24-series airfoil superior to all others tested at high speeds (John Stack and Albert E. von Doenhoff, "Tests of 16 Related Airfoils at High Speeds," TR-492 ).
32. Freeman to chief, Aerodynamics Division, 21 April 1934; 11 June 1934; and 9 Oct. 1934.
33. Lewis to LMAL, 25 April 1934; and 15 June 1934.
34. Lewis's insistence on free discussion often made it difficult to determine where an idea originated. See HJ.E. Reid, "Notes for Dr. Hunsaker with reference to Dr. Lewis' part in establishing the Langley Laboratory," 4 Aug. 1948; and John V. Becker, "The High-Speed Frontier: Case Studies of Four NACA Programs, 1920-1950," NASA SP-445 (1980), p. 22.
35. P.E. Hemke to Lewis, 17 July 1934; Victory to LMAL, 24 July 1934. Reid replied for the laboratory that a tapered slot placed near midchord was the answer to questions one and two. There was no firm answer to three. A slight gain in lift over drag was experienced for coefficients of lift above .25, the gain increasing rapidly with coefficient of lift. This was due more to an increase in the coefficient of lift than to a reduction in drag. Reid to NACA, 27 July 1934.
36. Freeman to chief, Aerodynamics Division, 15 Nov. 1934; Frederick E. Weick to chief, Aerodynamics Division, 5 Dec. 1934; Jacobs to chief, Aerodynamics Division, 5 Dec. 1934; Reid to NACA, 4 Dec. 1934 [sic].
37. R.P. Lansing to Lewis, 19 March 1935; Freeman to Elton W. Miller, 2 April 1935; Donald H. Wood to Miller, undated; Reid to NACA, 3 April 1935; Lewis to Lansing, 5 April 1935.
38. Donald H. Wood to Elton W. Miller, 6 Feb. 1935.
39. Dated 20 March 1935. Remarks by the chief of the Aerodynamics Division were added to Wood's memorandum, over the date of 9 Feb.
40. Reid to NACA, 22 March 1935.
41. Helms to Lewis, 13 May 1935; Jacobs to Miller, undated; Freeman to chief, Aerodynamics Division, 15 May 1935.
42. Lewis to LMAL, 2 Aug. 1935; Helms to Lewis, 25 July 1935 and 29July 1935.
43. Lewis to LMAL, 1 Aug. 1935.
44. Helms had cited in defense of his interpretation Millard Bamberís Technical Report 385. Freeman countered that Bamberís report had shown only that measured drag could be reduced by suction and blowing methods of boundary-layer control; it did not include the drag corresponding to the power consumed by the required blower. It was therefore inconclusive on the overall effect on drag. Reid to NACA, 5 Aug. 1935.
45. Reid memorandum for files, 15 Aug. 1935. With respect to Doenhoffs proposal for smoke-tunnel research on the boundary layer, Jacobs and Doenhoff agreed that such results must be conducted in a near-zero-turbulence tunnel, which the Langley laboratory then lacked.
46. Jacobs to engineer-in-charge, 21 Aug. 1935. The full paragraph on these reasons follows:
Jacobs added to this memorandum the caveat that the results reported should "be considered strictly confidential and subject to revision."
47. Reid to NACA, 21 Aug. 1935; Smith J. DeFrance to engineer-in-charge, 31 Oct. 1935.
48. Reid to NACA, 17 Oct. 1935, forwarding von Doenhoffís report, which was published early the following year as Technical Note 544; Lewis to LMAL, 13 Sept. 1935; Reid to NACA, 18 Sept. 1935; Abe Silverstein and E. Floyd Valentine, memorandum report to engineer-in-charge, "Blocking Tests in the 20-Foot Tunnel," 13: Feb. 1936.
49. Smith J. DeFrance to chief, Aerodynamics Division, 10 June 1936.
50. Lewis to LMAL, l6June 1936.
51. Von Doenhoff to chief, Aerodynamics Division, 30 June 1936; Jacobs to chief, Aerodynamics Division, 20 July 1936. This recommendation echoed the one that Jacobs and von Doenhoff had made the previous year in the conference on Freeman's proposed program of research. See note 42.
52. Reid to NACA. 11 Nov. 1936.
53. Reid to NACA, 19 Nov. 1937; Dryden to NACA, 14 Dec. 1937; Reid to NACA, 21 Feb. 1938. Von Doenhoffís report was published as TN-639 in March 1938.
54. Millikan to Lewis, 8 Feb. 1938.
55. Freeman to Donald H. Wood, 18 Feb. 1938; Wood to Elton W. Miller, undated; Smith J. DeFrance to Miller, undated; Reid to NACA, 28 Feb. 1938.
56. Jacobs to engineer-in-charge, undated [ca. 27 June 19381. The low-turbulence tunnel had gone into operation at Langley the very month in which Jacobs sent his report to Reid, indicating that this was among the first projects to win tunnel time in the new facility. Low turbulence was obtained by screening the air in the tunnel and by increasing the contraction ratio: i.e., the ratio of the widest part of the tunnel to the lowest part, the test section. The old variable-density tunnel, with a contraction ratio of 4 t 1, had a 2-percent turbulence. By 1941 the NACA had a tunnel with a contraction ratio of 20 to 1 and turbulence of less than .015 percent. Two-dimensional flow was achieved by placing a wing section completely across the test section, to eliminate airflow anomalies at the wing tip and the wing-to-fuse-lage interface. See Gray, Frontiers of Flight, pp. 47-50.
57. Lewis to LMAL, 6 July 1938.
58. Reid to NACA, 13 Oct. 1938. Freeman left the NACA in 1939.
59. Reid to NACA, 6 Aug. 1938, in response to a letter from Vega Aircraft Company, requesting Freeman's results. Reid to NACA, 23 Sept. 1938; von Doenhoffs report was published as TN-671 the following month.
60. Albert E. von Doenhoff, "Investigation of the Boundary Layer About a Symmetrical Airfoil in a Wind Tunnel of Low Turbulence," Advance Confidential Report, Aug. 1940; J.W. Wetmore and J.A. Zalovcik, memorandum for files, "A Flight Investigation of the Boundary-Layer Characteristics and Profile Drag of the NACA 27-212 Laminar Flow Airfoil," 15 Aug. 1940; von Doenhoff and Neal Tetervin, "Investigation of the Variation of Lift Coefficient with Reynolds Number at a Moderate Angle of Attack on a Low-Drag Airfoil," Confidential Bulletin, Nov. 1942; Wetmore, Zalovcik, and Robert C. Platt, "A Flight Investigation of the Boundary-Layer Characteristics and Profile Drag of the NACA 35-215 Laminar Flow Airfoil at High Reynolds Numbers," Memorandum Report, May 1941; Zalovcik, Wetmore, and von Doenhoff, "Flight Investigation of Boundary-Layer Control by Suction Slots on an NACA 35-215 Low-Drag Airfoil at High Reynolds Numbers," Advance Confidential Report 41529, Feb. 1944, first submitted 8 April 1942.
61. Von Doenhoff to chief, Aerodynamics Division, 23 June 1942.
62. Lewis to LMAL, 20 July 1942; Reid to NACA, 26 Nov. 1942.
63. Lewis to LMAL, 19 Jan. 1943; Reid to NACA, 2 Feb. 1943; Lewis to LMAL, 8 Feb. 1943.
64. Reid to NACA, 29 April 1943; R.E. Littell to Lewis, 5 May 1943; Reid to NACA, 24 May 1943. This report (Albert E. von Doenhoff and Neal Tetervin, "Determination of General Relations for the Behavior of Turbulent Boundary Layers") was published in 1943 as Technical Report 772.
65. See Advance Confidential Report L4G14 of Feb. 1944 and Confidential Bulletin L4H1O of Aug. 1944; I.H. Abbott to chief of research, 19 Jan. 1945.
66. Lewis to LMAL, 8 Feb. 1945; S. Katzoff to chief of research, 20 Jan. 1945; Reid to NACA, 5 Dec. 1945, commenting on Dryden's letter of 8 Aug. 1945.