The nuclear device the Soviet Union exploded in August 1949 also exploded comforting illusions: that our nuclear monopoly made us impregnable, that we could safely and sharply cut our spending on defense, that we were technically superior, and, more subtly, that long-range missiles, because they could not be well aimed, were not worth the investment. This shattering of transient postwar comfort by that single event reshaped the United States in ways that are still with us—and they reshaped my own life. I had embarked on an enormously satisfying and very busy life of family, teaching, research, and government service. I had proven to myself that I could be a university professor, do research, and enjoy it.

In September 1947, de facto become de jure when the Army Air Corps became a separate service—the United States Air Force. The Air Force had come out of the war a fervent believer in the power of engaging first-class science and technology with military purposes. That fervent belief got its “bible” in the report Toward New Horizons mentioned in the previous chapter. Late in 1944, the Army Air Forces’ commanding general, General Henry “Hap” Arnold, had asked Theodore von Kármán to lay out the postwar role of technology for the Air Force. Von Kármán and his colleagues did that, in 13 volumes delivered to an ill Arnold on December 15, 1945. One outcome was the creation of the Air Force Scientific Advisory Board (SAB),¹ a lineal descendant of the wartime Scientific

Advisory Group, which Major General Curtis E. Lemay said “rendered such signal service to the army air forces during the war that it has made obvious the necessity for continuation of such a service as an essential part of Headquarters staff planning.”²

On the strength of my guided-missile experiences, I became a member of the SAB and began an ever-closer relationship with its chairman, Theodore von Kármán. At Cal Tech I learned about the unique abilities and character of this genius from some of my friends in the aeronautical department. With his heavy Hungarian accent, they had a lot of trouble understanding him. But his blackboard notes were priceless, and for his research students his close contact night or day created a loving family. When Bunny met him in the receiving line at a reception for my chief scientist job, he grasped her hand, looked her in the eye, and said, “Ze wife of ze Guy is a Doll.” He was a charmer to women and friendly to his colleagues. I became an admiring follower.

I remained an SAB member through 1968, except the year plus when I served as chief scientist of the Air Force. It was through the SAB and the implications of the Soviet bomb that I got involved in work on defending the country against bomber and missile attacks. General Muir S. Fairchild, vice chief of staff of the Air Force, called an emergency session in November 1949 requesting assistance in strengthening the nation’s air defenses. Until that time the Air Force had expressed concern but not alarm that it could not put up a minimal air defense of our nation’s peoples and industries. In 1945, an Air Force officer pointed out:

This assessment carried little weight at the time in part because of the absence of reliable ways for a weapon to find its target, especially at intercontinental distances. Suddenly, with the explosion of the Soviet bomb, that mattered much less. A miss of even a few tens of miles might be good enough. More to the point, if the Soviets could develop a bomb, why not missiles? They also had German rocket scientists. At the same time, our own missile program, especially ballistic⁴ missiles, was in disarray. The Air Force ballistic missile program was essentially terminated for budget reasons in 1947. In July 1947, the Air Staff concluded:

Long-range missiles were a fourth priority, with vertical bombs of the World War II type fifth. The Air Force knew that long-range missiles would become strategically important, but it didn’t have the money and figured that development time would be at least 10 years (not a bad estimate, since Atlas, the first intercontinental ballistic missile was declared operational by the Air Force in September 1959).

It was the third priority—air defense—that SAB was now asked to deal with. Our defenses were in bad shape. The Soviets had the bomb, and the United States essentially had no effective defenses against bombers carrying nuclear weapons. “The Battle of Britain had been won through the long-term attrition of Nazi bombers. However, in the atomic age, a near perfect defense against the penetration of bombers was required.”⁶ Yet in 1949 there was no integrated air defense system, no effective radar system for low-flying aircraft. Airborne interceptor radars failed when looking down at low-flying bombers, and while ground radars worked well at long ranges against high-flying aircraft, they could not detect low flyers.⁷ Bob Robertson, a colleague of mine from my wartime days in

Britain and France called the nation’s air defense “scandalous and disgraceful.”⁸ In response, the SAB formed the Air Defense Systems Engineering Committee (ADSEC) but made it an unusual committee in that it gave a lot more than advice. It had an operational role, to set out systems requirements for an effective air defense, identify technical needs to meet the requirements, do demonstration experiments, and provide procurement data and criteria to the Air Force. ADSEC was chaired by my Massachusetts Institute of Technology (MIT) colleague George Valley, who had led the Radiation Laboratory in its development of the H₂X radar bombsight for all-weather bombing of Germany, and before World War II had done research on cosmic rays, going, as I did for my thesis, to high altitudes for his data, in his case to 12,000 feet on Mount Evans in Colorado.⁹

George and I spent time together in Europe during the war when he was on extended visits from the Rad Lab.¹⁰ I played chess with him on a little portable set while we were traveling through the English countryside to visit research laboratories or in London sitting out an air raid. I came to know him well.

ADSEC fell into a pattern of meeting every Friday night at the Air Force Cambridge Research Laboratories, which become the venue for placing contracts and providing funds in response to ADSEC recommendations delivered quite informally. We realized at the outset that there were several things we would have to know to do our job: (1) the state of the Soviets’ development of long-range bombers; (2) the capabilities of our long-range radar systems, how they would improve over what time period, and whether and how they would deal with countermeasures used against them; (3) the current state of interceptor aircraft and missiles, both air-to-air and ground-to-air; and, not least, (4) the state of the communications, command, and control (C³ in today’s parlance) to link data on incoming aircraft with our defending missiles and interceptor aircraft.

Al Donovan, of the Cornell Aeronautical Laboratory, did the calculations showing “that a bomber flying in over the north polar region at high altitude could always detect the ground radar before the radar detected it;

it could thereupon descend under the radar beam and continue undetected at low altitude.”¹¹ For judgments on C³ systems we would have to gauge the ability of current computer systems to handle the immense amount of data that would come in and be processed at very high speed. And we’d have to be able to project that into the future. We became well acquainted with Project Whirlwind, the very impressive efforts by Jay Forrester (23 years old) and Robert Everett (26) to develop a digital computer reliant on vacuum tubes.¹² Indeed, it was in 1949, when ADSEC began its work, that Forrester and Everett changed their tactics from development of a high-speed analog computer for an aircraft trainer simulator to a high-speed digital general-purpose computer. Two years later came the key inventions necessary to place into operation the first real-time, synchronous, parallel digital computer, including the magnetic core memory that revolutionized computing.¹³ ADSEC in its work suggested the creation of a distributed network of small radars to meet the threat of low-flying bombers, with vital presumption of the availability of large digital computers to handle and analyze the data harvest.

I’m jumping ahead. George Valley asked Al Donovan and me to look at the state of guided missiles and of fighter aircraft, including interceptors. We moved quickly by scheduling in early January 1950 a week of visits to the centers of radar and missile development in the Los Angeles area—Cal Tech’s Jet Propulsion Laboratory, Northrop, RAND, Hughes Electronics, North American Aviation, and Douglas Aircraft. I was amazed by the progress since I first visited in 1942. Not least, transcontinental travel had changed dramatically. In 1942, with Office of Scientific Research and Development (OSRD) priority, I took my first commercial flight from Dayton to Los Angeles in a TWA DC3, which stopped about every 250 miles, flying low when the weather was bad—and weather or equipment delays were common. By the end of the war, there were longer DC4s and Lockheed commercial flights. We didn’t get commercial passenger jets until a decade later.

Now in 1950 we were talking about subsonic interceptor aircraft that with turbojets would go to higher subsonic Mach numbers. And in the works were turbojets with afterburners¹⁴ that would take the aircraft into the supersonic region. Also in 1950 there were experimental prototypes, some in early production, of radar-guided missiles to replace or

complement bombers using celestial navigation. Inertial navigation had not yet arrived, but its components such as gyroscopes and accelerometers that Charles Stark Draper and his colleagues were developing at MIT were entering aircraft control and fire control systems. It wasn’t until 1953–1954 that the structure of an inertial navigation system became clear, and its operational status was confirmed in 1955, when the Air Force made it the primary guidance system for Thor, an intermediate-range ballistic missile.¹⁵

ADSEC helped push that technology and made plain the urgency for large-scale digital computing to the nation’s air defenses. And ADSEC established a style of working that was to continue well into the 1950s:

The very success of ADSEC—the very large technical territory it traversed and mapped—made it obvious that there was much more than could be done by part-time volunteers: a full-time laboratory was needed. That judgment led to the creation of the MIT Lincoln Laboratory in 1951 and to the mutually agreed upon dissolution of ADSEC in January 1952.¹⁷ The Lincoln Laboratory made real the recommendations of ADSEC; for example, early research at the laboratory focused on the design and prototyping of a network of ground-based radar and aircraft control centers for continental defense.

I’m doing a gross injustice to a much more complicated story, wonderfully told by George Valley.¹⁸ The struggle to create the Lincoln Lab was fierce. After several meetings, punctuated by the usual push and haul of academic politics, MIT accepted the ADSEC recommendation and established a laboratory. MIT stipulated a triservice laboratory, especially having Navy support, which it got rather grudgingly because the Navy

was putting all the money it could afford for MIT into the Research Laboratory for Electronics, a lineal descendant of the Rad Lab, and the Army was just as heavily invested in Bell Laboratories. The final score on support by the Air Force, Navy, and Army was proportionately 10:1:1.¹⁹ MIT also demanded that it choose the director. Agreed. Another condition was that the new laboratory, to be called the Lincoln Laboratory, would not be on campus. Agreed. It was put next to Bedford (later Hanscom) Field at the intersection of the towns of Bedford, Lexington, and Lincoln. Finally, MIT stipulated that the ADSEC recommendations for a laboratory focused on air defense were to be vetted and detailed by a larger group of people through a study called Project Charles. Done but not without heat. “Had Margaret Mead attended Project Charles,” George Valley wrote, “she might have written a sequel to her well-known book: Growing Up Among the Physicists.”²⁰ Though I agree to some extent with George’s sentiments about Project Charles, it did succeed in greatly broadening the knowledge of ADSEC and did entice a number to join the effort at Lincoln Laboratory. Britain’s Royal Air Force assigned Air Commodore Sir Geoffrey Tuttle, a much decorated “Battle of Britain” fighter pilot to Project Charles. Geoffrey’s wide knowledge and good spirits contributed greatly to the study, which ended in January 1952.

I was pleased with what ADSEC did. The work pushed new technologies, such as real-time processing of radar data, which in turn girded improved detection and tracking of aircraft and ground vehicles. The most dramatic was SAGE—Semi-Automated Ground Environment— conceived by George Valley, in which hundreds of continuous-wave radars were connected via telephone lines and computers to provide detection of low-altitude attacks unseen by conventional radar systems. It became the largest computer project of the 1950s and the training ground for many of the people who led the computer revolution in the following decades and created many of the companies of a new industry. SAGE was the first large control system to use a digital computer and, in fact, because of its central command structure, it was the first computer network.²¹

On June 25, 1950, 90,000 North Korean troops led by 150 T-34 Soviet tanks crossed the 38th Parallel to attack South Korea.²² The South Korean capital, Seoul, was taken within a few days, and the onslaught

pushed south, threatening to drive the South Korean forces into the sea. Within two days President Truman committed U.S. ground troops. Weaknesses in our military capabilities, especially in the Army, quickly became apparent. Some new weapons were introduced but overall very few, which was not surprising given the tight purse strings after World War II. We were again in a shooting war, with pretty much the same weapons, the same tactics. Charles Stark Draper at MIT was delighted with the fire control systems that he helped the Air Force install in some of our aircraft, particularly the F-86 Sabre jet, the country’s first swept-wing fighter, which during the Korean War attained a 7:1 kill ratio.²³

Military budgets went up sharply. The Air Force was still putting together the new Air Research and Development Command, recommended by a panel of the SAB, chaired by Louis Ridenour. There was a strong need, the panel believed, for a command for research and development distinct from logistical functions, such as supply, procurement, and mobilization.²⁴ But having been established in January 1950, it hadn’t advanced far enough to be fully trusted by the top brass to do the work. It just didn’t have the capability. So in July, a month after the North Korean attack, the Air Force asked the SAB to review its guided-missile program, in effect mirroring the work of ADSEC looking at air defenses. Specifically, the panel was asked to judge which guided missiles then in development were most promising and when they might become operational. The Air Force also wanted to know if and how these programs could be accelerated if more money became available, as seemed likely. Louis Ridenour,²⁵ who had chaired the earlier SAB panel that led to the Air Research and Development Command and also served as the first chief scientist of the Air Force, took this one on.²⁶ It turned out that our ADSEC work fit like a glove with this task, and so we were able within a few weeks to get our briefings and frame recommendations. In fact, the Air Force jumped on the recommendations we discussed at our meeting in July even before we finished our report and accepted a lot of them. The key recommendation we made was to quit thinking of guided missiles as a special case—as an add-on to “conventional” Air Force programs—but rather to integrate them fully as a normal and essential part of its planning. That echoed the systems approach flavoring our ADSEC work. We told the Air Force that if it was to efficiently move guided

missiles into deployment to regard them “as being the natural next steps in the cultivation of various aspects of air warfare.”²⁷ This work, paralleling that by other SAB committees, proved over time an excellent kickoff for the new Air Research and Development Command under Major General David M. Schlatter, who commented very favorably on the help given by the SAB and who got the order from the Air Force chief of staff, General Hoyt S. Vandenberg, that the ARDC become by May 1951 “an independent, self-sufficient major Air Force command.”²⁸

The SAB worked very effectively. There were usually about a dozen panels in existence with titles like “Electronics” or “Armaments” to which each member was assigned, with occasional dual assignments. At each of two annual meetings of the board, following general briefings by leaders, the panels would break out on their own to discuss any topic of interest in their field together with the ad hoc and special committee reports. The panel chairs would then report to the entire board. Thus each board member got complete coverage. That feature broadened my views, stretching me toward the generalist’s point of view that helped me in later work.

The arrival of the Air Research and Development Command to focus and crystallize research and development for the Air Force was one piece of the postwar structure for federal science and technology developing rapidly around 1950. That structure in its fundamental form endures to this day; it drove the exuberance and excellence of American science and technology after World War II. The philosophical underpinning was, and to a considerable degree remains, Science, the Endless Frontier, by Vannevar Bush.²⁹ The organization that Bush created and led during World War II, and for which I worked during the war, had been an incredible success, but Bush knew that it was not sustainable in peacetime and fought against bitter resistance to shut it down. At the same time, he argued fiercely that the central lesson of the OSRD, the essentiality of linking the best science to national goals, had to endure, and that new organizations fitted to peacetime were needed. He argued:

Bush clearly articulated those “other fields” for peacetime federal research:

That Bush report essentially molded and made durable the bond between government and civilian science. The underlying faith that civilian science was essential to national security, meaning in those days military security, set the table for the postwar structure of federal research. The contracting instruments created by the OSRD became standardized in peacetime research supported by the federal government. The decentralization of research entered mainstream belief, as did the faith that, while the government could set general goals, results came by supporting the best. Not least, the enormous success of wartime research through the OSRD model made the military and the American public passionate believers in high-quality science.

Bush was the principal in making that happen. Still he ran into trouble when he tried to impose his ideas for turning the faith in science and scientists into federal structures and budgets. Bush wanted a “National Research Foundation,” with sweeping powers vested in civilian scientists. It would be federally funded, but civilian scientists would control its expenditures and choose its director. The foundation would support work in the physical sciences and medicine³² and would set priorities for long-term military research.

Typical for Bush, this was an audacious, even confrontational plan. Audacious because it would be even more powerful than the OSRD and

confrontational because it would weaken if not eliminate military control over its science and technology programs. And it was doomed. The biggest reason was, ironically, the very success of OSRD in convincing the military to love research. The three services fervently believed that if the OSRD went, something else had to rise in its place and that “something else” was going to be controlled by the military, not by a board of civilian scientists setting military research priorities over which they had no control. And Bush’s argument for a director appointed by a civilian board didn’t sit well with President Truman, his budget director, nor with many in Congress, notably Senator Harley Kilgore (D-W.Va.), who proposed a “National Science Foundation,” in a form that he largely shaped. Created in 1950, the NSF was a pale version of what Bush wanted: it had no military division and no medical division, the director was appointed by the president, the board shared its power with the director, and its budget was capped by Congress at $15 million. In fact, it got $250,000 the first year. Bush was understandably not happy and would have little to do with the new agency; for example, he wouldn’t let researchers from the Carnegie Institution, which he headed, apply for NSF grants.³³

What emerged by 1950 was to an outsider a confusing mixture of agencies supporting civilian research, each with different structures and missions. These included not only the NSF but also the National Institutes of Health, the Office of Naval Research, the Atomic Energy Commission, the Air Research and Development Command, and in the fall of 1951, the Air Force Office of Scientific Research. Many people, including me, thought this wasn’t bad—that it was a good thing to have many sources of support for scientific research and to have several agencies doing applied research and development. This melange gave the Bureau of the Budget, now the Office of Management and Budget, a strong role in putting together these different pieces.

Clearly, my work for the SAB—I had been a member since 1947— brought me in touch with some of these struggles, although I did not get involved in the hassling about turf, how to organize the various agencies, or budgets. At this time in my life, I began to realize that I didn’t relish these kinds of organizational struggles and often took extremes to avoid them. I was much more interested in research itself, in development, and in education. Thus, I wasn’t all that upset that I hadn’t stayed on in

Washington to work for Vannevar Bush, but rather pushed to return to MIT, first with Project Meteor and then on the faculty. If I’d stayed on, I would have been just another fellow fighting for this or that idea in the complex and confusing organization that is the U.S. government.

I remember a visit I had with Wilbur Goss of the Bumblebee missile program at the Johns Hopkins’ Applied Physics Laboratory, someone I respected greatly from the time I first met him in 1944. Wilbur and I philosophically reminisced about how our lives had changed from what we wanted to do. For example, absent the war, I would probably have gone to Stanford in 1941 to start a career in physics rather than to the Rad Lab. Yet we had no regrets in having plunged into wartime work, he into ordnance and proximity fuzes and I into radar and guided missiles. It was that way with many Americans. We couldn’t face once again the horrors of war after we just fought the war “to end all wars.” Nor did we want to face a world split into two hostile camps, this time with the risks multiplied many times by fission and soon fusion weapons.

I decided that my push into an academic career had been right and that I was going to keep pushing that no matter what. I also realized that with all the investments I had made in certain fields of military importance I should continue to help out, in the main through part-time advising, as I was doing in 1950 with the Air Defense Systems Engineering Committee and with the Ridenour Committee looking at the state of the Air Force’s guided-missile program. It worked both ways: my chosen field of aeronautics meant that some of my advising contacts were valuable in deciding my research program.

In the spring of 1951, I was promoted to associate professor and given tenure. It was 10 years after I got my doctorate in physics at Cal Tech and made the decision to go to MIT and the Rad Lab instead of Stanford and physics. Tenure in a university was what I wanted and now I had it. It was in a department in the engineering rather than science school. I certainly still had a strong interest in basic science, particularly physics and more particularly in my own fields of cosmic rays and Geiger counters, and I tried to keep up by talking with friends in those fields and reading the journals. My path at MIT was clearly going to put the finishing touches on making me a professional in aeronautical engineering, or, what it became later, aeronautics and astronautics. Aeronautics

of course became one of the great drivers of the postwar American economy, through the development of long-range, wide-bodied aircraft carrying large numbers of people and goods around the world. And certainly not least I now had a family—a wife, two children, a boy and girl, a dog, a lovely home, and a foot in a beautiful place in Randolph, New Hampshire, where Bunny’s parents had a home.

Getting a promotion and tenure in the spring of 1951 started me thinking even harder about my academic career. I started thinking about that Banquo’s ghost of academic life, “publish or perish.” I discovered that the topics one could publish on in engineering were different than in science. In academic science, basic research is the ideal of the research that goes with teaching in the universities. In engineering there is wider latitude. Some professors believed that basic research in the engineering sciences was proper for engineering professors, while many others believed that work on major projects not quite basic in their approach would do fine for their own intellectual work and that of their students.

In my three years of graduate work at Cal Tech, I got three nice papers out of two areas of research, on cosmic rays and Geiger counters.³⁴ When I entered the classified fields of wartime radar and then guided missiles, I found unsurprisingly that I wasn’t getting any good publications in the open literature, even though I did an awful lot of writing on reports that were classified. This pattern of hard writing but for a classified audience continued after the war. Still, that MIT gave me tenure signaled that my work, however limited the audience, met the institution’s intellectual standards. But I wasn’t satisfied. I wanted to change the pattern. I turned to open publications of the work I had started on three projects: (1) the hypersonic wind tunnel with the offshoot technical field of condensation in high-speed flows; (2) the transonic aircraft control project covering high-speed aerodynamics from subsonic through transonic to supersonic; and (3) the shock tube program, which put me into a special area of compressible flow. These all served as a source of publish-able work, for me and for my students.

In the fall of 1950, Kenneth Rathbun, a mechanical engineering

graduate thesis student of mine and I handed in our report, titled Theoretical and Experimental Investigation of the Condensation of Air in Hypersonic Wind Tunnels, to the National Advisory Committee for Aeronautics for publication as an NACA document. That in turn led to me submitting a paper entitled “Condensation of Air in Hypersonic Wind Tunnels” to a meeting in England on heat transfer of the Society of Mechanical Engineers. This occurred at a very busy time for me. I was learning to teach new courses and doing a lot of advising—for ADSEC, Project Charles, and the Ridenour Committee advising the Air Force on its guided-missile program. That meant I had to plan the trip to England very carefully. And I did, not only giving my talk but catching up with what Britain was doing on missile defense, visiting the famous air show of the Society of British Aircraft Constructors at Farnborough, and not least visiting with Geoffrey Tuttle and seeing wartime friends. I returned from three weeks in Britain on a Sunday, for MIT registration day on Monday and to meet a new class on guided missiles on Tuesday. And I had to quickly reconnect to the transonic aircraft control and to the shock tube project, never mind getting our home ready for autumn and winter.

I continued to publish via journal articles and book chapters—on the growth of the boundary layer behind a shock wave, on condensation in high-speed flows, and, in general, on high-speed aerodynamics and jet propulsion. I was engaged in the intellectual network of aerodynamics. In 1954 I published with Ray Bisplinghoff an article in the Proceedings of the National Academy of Sciences entitled “The Shock Tube in Aerodynamic and Structural Research.” It was used to support Ray’s election to the academy and some years later as partial support for my election.

It was an extraordinary time for me, but just as the onset of World War II had sharply altered my academic career, so too were events of the early 1950s to change my life again. A signal event was the acceptance of intercontinental missiles mated with nuclear weapons as the country’s prime weapon in maintaining its national security—and with that even more

focus on how to defend the continental United States against Soviet intercontinental ballistic missiles (ICBMs).

The performance of liquid rocket motors had advanced since World War II. There also were widespread calculations on what could be done with them. In my rocket course I analyzed the performance needed to accomplish intercontinental trajectories, to establish a satellite trajectory, to escape the earth, even to escape the sun. Scientific visionaries like Arthur Clarke foresaw manned space flight; scientists planning the International Geophysical Year, July 1957–December 1958, began to think of flying an instrumented satellite; and military men began planning ballistic rockets of ever-increasing range. The Air Force was assigned the ICBM program. In 1950 Louis Ridenour chaired a committee on which I served on the Air Force’s guided-missile program.

In 1952 Theodore von Kármán enlisted another former Hungarian, John von Neumann,³⁵ to work on the problem of linking the development of ICBMs with that of nuclear weapons, and in 1953 the von Neumann panel was started, and soon pointed out that progress on nuclear weapons was so fast that they would not be the stumbling block to the development of an effective ICBM. Von Neumann and most of his committee were soon co-opted into another venture, the “Tea Pot” Committee, to map out the development and eventual deployment of ICBMs.

The “Mike” thermonuclear tests were set off in October 1952 at Eniwetok in the Marshall Islands. Less than a year later, in August, the Soviets set off their own hydrogen device. Two months later, in October, the “Tea Pot” Committee met, chaired by von Neumann, a fervent believer in lighter and very destructive thermonuclear weapons delivered on rockets. The committee was organized and its members were selected by Trevor Gardner, who in 1953 was Special Assistant for Research and Development to the Secretary of the Air Force and who passionately believed that the nation needed an operational ICBM if it was to avert nuclear disaster. Only a few months later, in February 1954, the committee issued its report, strongly arguing for a crash program on ICBMs, and in 1954 the United States went from a desultory pace to running hard to build its first ICBM. This was Atlas, liquid fueled, first radio and then inertially guided, with a range between 6,400 and 9,000 miles and declared operational late in 1959.

This is cartoon history, and the reasons for these radical changes are complex and still argued over. Until the early 1950s the Air Force had concentrated on building bombers to deliver nuclear weapons and on fighter aircraft for the Korean War. Resistance to an ICBM program also rested in part on lack of accurate guidance; if V-2s traveling a few hundred miles had an operational error of about 4 miles, ICBMs traveling 3,000 miles would miss their target by 60 miles.³⁶ But the power displayed in the first U.S. thermonuclear test, Mike, weakened the accuracy argument. As the chairman of the Atomic Energy Commission reported to President Truman: “The shot island Elugelab is missing, and where it was there is now an under-water crater of some 1500 yards in diameter.”³⁷ With “Mike” in principle promising a vastly improved yield-to-weight ratio, the lift needed in an ICBM eased downward. But since Mike weighed 60 tons, that was truly “in principle,” and it wasn’t until the Bravo test of a much lighter bomb in March 1954 that the yield-to-weight argument became more realistic. Trevor Gardner had taken advantage of a review of missile programs to eliminate unnecessary duplication ordered by President Eisenhower’s defense secretary, Charles Wilson, when he created the Tea Pot Committee, formally the Strategic Missile Evaluation Committee, chaired by von Neumann. Gardner deliberately put the committee beyond Air Force control but appointed as its military representative Colonel Bernard Schriever, then a fairly lonely advocate for missiles in the Air Force. This Tea Pot Committee³⁸ reported in February 1954, urging a strong push toward ICBMs—that is, the Atlas missiles—judging that an ICBM could be operational within six years (i.e., within six years of the Tea Pot report).³⁹ As to what the Soviets were up to, the committee offered artful language: “While the evidence does not justify a conclusion that the Russians are ahead of us . . . this possibility can certainly not be ruled out.”⁴⁰ Not least, the committee strongly urged that a new organization be created to develop the Atlas. The Air Force got the point, and moved to establish a facility for accelerating Atlas: the Western Development Division, located in Inglewood, California, under the Air Research and Development Command, and led by Bernard Schriever, now a general officer.⁴¹ General Schriever, faced with the daunting task of rapidly developing a very com-

plex piece of technology, adopted what was then a new management approach called systems engineering. It had many parts, including “change control” to track the effects of changes in a single component on the entire system and “integrated system testing” to look at the whole system under realistic operating conditions. Systems engineering was a management revolution then but is now commonplace.⁴²

So in the space of a few years—from about 1950 to 1955—I saw remarkable events, participated in some, and was affected by all. Civilian scientists continued to have a major and to a degree independent role in military issues, not least through the SAB. The need for a separate and strong research and development organization in creating major new weapons systems had been recognized in creating the Air Research and Development Command. A major commitment to command and control systems that were reliant on computers had been made in building SAGE to protect the United States against low-altitude bombers not detectable by other early-warning systems. The United States had committed itself to missiles as the primary delivery system for nuclear and thermonuclear weapons. And, finally, I had just started an SAB study on an anti-intercontinental ballistic missile defense, AICBM.

In late September 1954 the SAB held its semiannual meeting at Offutt Air Force Base, in Omaha, Nebraska, the home of the Strategic Air Command (SAC). I gave a preliminary discussion of our AICBM study. Our host was General Curtis Lemay, the powerful voice in the Air Force for strategic bombardment and if not hostile at least quite skeptical of the role of missiles to carry a war to the enemy several thousand miles away. SAC was heavily dependent on two subsonic bombers: B-47s and the new, faster, and more powerful B-52s, which first flew that year, 1954, and entered service a year later. For intercontinental-range bombardment, either aerial refueling or a worldwide set of bases was needed, and the Air Force worked on both. General LeMay was especially interested in the SAB’s advice on longer-range aircraft that could carry heavier loads (i.e., more and heavier bombs).

At the meeting, several Air Force officers, led by Lieutenant General

Donald Putt, who had recently become deputy chief for development at Air Force headquarters,⁴³ took me aside to ask if I would become the next chief scientist of the Air Force. It was and is a rotating position. Chief scientists normally came from academia and served for about a year. The term for the current incumbent, Chalmers Sherwin, was ending at the beginning of 1955, and they were beginning to work on a replacement. General Putt made a strong sales pitch, and as I rode home from that meeting I began to think that professors do better if now and then they take a sabbatical. The usual time for a sabbatical is every seven years, and I’d been a professor for eight. And I also felt that it would give me a chance to look again at the work I was doing in the field and maybe take on some new things. Finally, I felt that the results from my research programs were peaking—the transonic control project and the shock tube and atomic blast programs. It would do the projects and me good to have others take those over. I wanted to do it.

I talked this over with Bunny, and she was quite willing; indeed, she had told me she would live happily with me in Boston, San Francisco, or Washington. And it was for supposedly only one year, the nominal tenure for a chief scientist, although it turned out to be closer to 18 months. We had just had our fourth child, Roy, and we would be going to Washington with four children, with Roy six months old and Guy, Jr., seven, bracketing the ages of the two girls, Sarah and Margo—plus one handsome Irish Setter. My mentor Julius “Jay” Stratton, now chancellor of MIT, also thought it was good for me to accept.

Also supportive was Jimmy Doolittle, who was becoming a close friend and who was vice chairman of the SAB, along with Mervyn (“Iron Mike”) Kelly of Bell Telephone Laboratories. Kelly became chair of the SAB in January 1955, when the chairman, Theodore von Kármán, decided he could no longer continue.⁴⁴ So Kelly and I took office at the same time, although Kelly left before I finished my year, with Jimmy succeeding as SAB chair.

Back to Washington to meet with General Nathan Twining, Air Force chief of staff, and General Putt, and to accept the appointment. Moving from MIT to Washington for a year and half was tougher on Bunny in making all the family arrangements than it was for me. I had secured an attractive rental in Belle Haven, between Alexandria and Mount Vernon,

Virginia, an easy commute to my office at the Pentagon. Early February 1955, one week after I started my job, I returned to Belmont, and Bunny and I drove our two cars to Virginia. Our passengers included four children, one wonderful but aged Irish Setter, and a fine helper, Nanny Blyth, who helped Bunny in many ways, not least the babysitting for a couple of weeks until Bunny got the major chores done.

Washington had the reputation of being a sleepy southern city. From the work standpoint, it was not. We did not get into much of the social and political life of greater Washington, being away both summers and having many friends and colleagues from elsewhere visit us. The local elementary school was a disaster for Guy, Jr., who had dyslexia. Sarah attended a private school, Burgundy Farms Country Day School, run by friends, Kay and Eddie Mayer, and had a long bout of mononucleosis. Bunny had her hands full.

Jimmy Doolittle had an office next to General Putt’s and mine. Both met me as soon as I came aboard and told me that, while I knew “a lot about guided missiles and rockets and fighter aircraft and all of that, we now want you to get more information about real airplanes.” “Real airplanes” meant big ones: bombers.

So off we went, on a plane piloted by Don Putt, flew overnight, and then spent a week flying around to West Coast aircraft manufacturers. It turned out that what these companies wanted to talk with Putt and Doolittle about was the competition to design and/or build the KC-135, which was to be the new jet-propelled refueling tanker for our B-47 and B-52 bombers. I was a bit puzzled and asked Don and Jimmy why these companies were putting so much emphasis on that particular airplane. They laughed, and told me “the company that won the final production contract [would] have a substantial leg-up on the development of the first commercial jet transport.”⁴⁵

On the last day of our trip, a Saturday morning, we flew from Los Angeles to Edwards Air Force Base in the desert. When we touched down in our big C-54 transport, we saw extensive black skid marks that rather randomly serpentined down the runway for some distance, finally going off the side of the runway and leaving deep gouges in the turf. This was

the mark of the first B-52 Stratofortress, about to be delivered from Boeing to the Air Force. Clearly, the landing had been close to a disaster and would have been were it not for rapid action by Captain Magruder, the Air Force officer in charge of the acceptance tests. The jar of the B-52 hitting the runway had released the landing gear, which began to fold up. Magruder, in the copilot’s seat, recognized what happened, grabbed the landing gear knob, pushed it back in place, and held it! Nobody knew how he could react that fast and push that hard, but he did. The landing gear unfolded again, but one side locked. Therefore, the wild skidding on and off the runway. But the plane was saved, no one was hurt, and the locking mechanism for the landing gear was promptly redesigned.

I spent some of my time during this tour of the West Coast looking at the status of Convair’s B-58 program, since my first real assignment as chief scientist was to a board looking at this plane. The B-58 Hustler was to be the country’s first supersonic bomber, going into service in March 1960, about the same time the country’s first ICBM went operational. The supersonic B-58 was considered the next step after the subsonic B-52. Smaller turbojet fighter aircraft and even some fighter bombers were reaching the supersonic range, and many in the Air Force now wanted it for bombers. But there were problems, not least that the Strategic Air Command, and especially its leader, General Lemay, didn’t much like the plane. Getting to supersonic speed meant slimming the fuselage and wings to reduce the ferocious drag at transonic and supersonic speeds. While more powerful turbojets were then in development to lessen size pressures, the smaller size translated into fewer bombs and not enough space for the electronics and jamming equipment to sufficiently protect a bomber aircraft over hostile ground. The B-52, even though it flew at subsonic speeds, carried the electronics to defend itself, carried more bombs, and with the onset of the KC-135 and aerial refueling, had tremendous range and very long flying times.

My role on the B-58 board was well defined—to judge the aerodynamics of the aircraft. It turned out that aerodynamics were the least of

the plane’s problems because Convair had worked hard on that, so much so that other elements suffered.⁴⁶ For example, there were great difficulties simply cooling the electronics during flight. This was when vacuum tubes hadn’t yet been replaced by integrated circuits. Nevertheless, even with its handicaps, we recommended that the program continue. It did, and in December 1955 a contract for 13 B-58s and 30 external weapon pods was let to Convair. The first plane was delivered in August 1960 and by 1964 90 B-58s were deployed.⁴⁷ The last plane was retired in 1970. Its problems included relatively high accident rates, that the first ICBMs came on line the same year the plane did, and that the Soviets in the late 1950s built surface-to-air missile capability that forced low-level penetration of enemy defenses, where the plane performed poorly.

It wasn’t until I met with Jim Killian that I began to get some glimmerings of what my term as chief scientist would be like. Jim, then president of MIT, chaired the Technological Capabilities Panel, better known as the Killian Committee, established in July 1954 at the request of President Eisenhower to look at three major national security issues for the country: striking power, continental defense, and intelligence. Many things converged to create the study, but Jim Killian summed it up crisply when he later wrote that “there was a growing realization that thermonuclear weapons in the hands of the Soviets posed a threat of terrible dimensions that required urgent efforts to construct new defenses, to give greater emphasis to the deterrence of war, and to seek arms limitations.”⁴⁸ The committee was due to report the same month, February 1955, that I was to start my tour as chief scientist. And it did so.

Not only did the Killian Committee report in February, Trevor Gardner was finally confirmed in February 1955 to his appointment as Assistant Secretary of the Air Force for Research and Development, then a new position.⁴⁹ Apt timing because Gardner was partly responsible for the creation of the Killian Committee.

The new Eisenhower administration and the Air Force had only moderate interest in missile development, when Gardner became Special Assistant for Research and Development to the Secretary of the Air Force

in February 1953. Two and half years later, on September 13, 1955, President Eisenhower designated the ICBM program as “a research program of the highest priority.”⁵⁰ Gardner got there in good part because of the Killian report, but his methods weren’t always appreciated. Jim Killian, in his memoirs about his service in the Eisenhower administration, grudgingly acknowledged Gardner’s take-no-prisoners style, describing him as “technologically evangelical.”^5a He learned early of the kind of gadfly pressure that Trevor Gardner could put on his superiors and that Trevor could be abrasive. Jim in a private conversation I had with him used stronger language than in his book in describing Gardner. Others were equally blunt, describing him as “sharp, abrupt, irascible, cold, unpleasant, and a bastard.”⁵³ But he was also “a doer, always direct, skilled at cutting red tape, yet with an amazing breadth of knowledge and interests.”

The Killian study⁵³ was clearly critical. It involved the future role of nuclear and thermonuclear weapons, large rockets, long-range radars to detect missiles and to guide them, and inertial navigation, a technology we were beginning to get familiar with, although as an idea it had been around a long time. It got the strong endorsement of President Eisenhower, who “stressed the high priority he gave to reducing the probability of military surprise.”⁵⁴ Strategically, the Killian Committee looked at the entire spectrum of military problems facing the United States and the world in the mid-1950s. Its recommendations matched the scope of its charter. It confirmed what Gardner and others already believed, that because “our defense system is inadequate . . . SAC is vulnerable and the U.S. is open to surprise attack.” Further, “evidence is accumulating that the Soviets are developing their long-range delivery capability.” And most alarmingly, “because of our vulnerability, [the] Soviets might be tempted to try a surprise attack. They might be so tempted in order to attack before we achieve a large multimegaton capability.”⁵⁵

There was much more to their findings and recommendations, which they divided into missile programs, continental defense, and intelligence, parsed by progressive time periods, traversing from our present very strong advantage to a future standoff when “an attack by either side would result in mutual destruction.”⁵⁶ The report of the Killian Committee was presented a couple of weeks after I arrived in Washington to take up the chief scientist job, and the president and his National Security Coun-

cil moved quickly to give top priority to the development of an ICBM by the Air Force. It also started actions to improve our intelligence, not least starting the development of the U-2 reconnaissance plane. It gave the Army the go-ahead for intermediate-range missiles and encouraged the Navy to develop ballistic missile capability for its warships, particularly its future nuclear-powered submarines. In effect, the report set the table for Atlas, the first ICBM; the Army’s intermediate-range Jupiter missile; the Poseidon missile to go into submarines; and the U-2 spy plane.

And it certainly set the table for me. The report “hailed the establishment by the Air Force Scientific Advisory Board of an Antiballistic Missile panel, and urged that it give early consideration to the formation of a full-time technical group to carry out a rapid but thorough study of defense against ICBMs.”⁵⁷ The SAB did establish the panel—the Anti-ICBM, or AICBM, panel—and I chaired it. It started before I became chief scientist and was the major item on my plate during my tour. While we were flattered that the White House “hailed” the work of the antiballistic missile panel of the SAB, it also added to our challenges, by asking us to examine the ABM problems facing the Canadians and the British. The Canadians of course also had the same problems with respect to ICBMs that we did, although they were likely to see an incoming one before we did. And the British and NATO faced the problem of IRBMs—intermediate-range ballistic missiles—that we knew the Soviets were developing.

We had a strong membership for the task, including Canadian-British representation.⁵⁸ And we had at least one unusual member: Charles A. Lindbergh. Trevor Gardner nominated him. The appointment was unusual in several respects, not least that FDR had banned him from wartime service to the government, angered by Lindbergh’s vicious attacks on him and his “America First” campaign before the war. Somehow Gardner got Lindbergh reinstated and on the panel, and he did indeed serve a useful role as a skeptic. Lindbergh later commented about his service on the AICBM committee in a book, saying that antiballistic defense was like trying to stop one rifle bullet with another, and that he

felt when he started on the committee some of the technical people were too optimistic that it could be done.⁵⁹

In fact, the problem was worse than hitting a bullet with a bullet, because incoming ballistic missiles are faster than a rifle bullet. That was solvable because recent studies suggested that the incoming missiles didn’t have to be hit on the nose. They could be stopped if not destroyed by nearby aerial bursts, by hitting shrapnel, or by atomic explosions generating intense X rays that could possibly burn through the lightweight covering of the warhead. And we looked at other esoteric ways to destroy incoming missiles, studies that continue to this day.

We had growing intelligence information about the Soviet Union’s work on long-range missiles. This came in good part from a large radar built on the initiative and leadership of my predecessor as chief scientist, Chalmers Sherwin, on the Turkish coast of the Black Sea at Samsum, close enough to the Soviet border to detect and monitor test firings of intermediate-range rockets flying some 1,500 miles into the central Asian desert. But it was not until the highly secret flights of the U-2⁶⁰ over Soviet territory beginning in June 1956 that we could “see” Soviet ICBM facilities deeper in the heartland, in Kazakhstan.⁶¹ Further, the threat of a ballistic missile attack had become much more real now that a thermonuclear warhead could offset bad aim.

The interesting part, of course, was defending against weapons that didn’t yet exist and whose specific capabilities and vulnerabilities could only be guessed at. What seemed clear was that a ballistic missile was most vulnerable in its initial phase when the rockets were just beginning to accelerate off the ground. It was tempting to think of antiballistic missile systems attacking at liftoff. That meant the defensive system had to act right over enemy territory. And that in turn suggested very advanced systems in which satellites circled all the time, some to detect and track a launch, others to carry warheads to intercept and kill it. That kind of thinking in the mid-1950s was premature—after all, satellite reconnaissance didn’t begin until the 1960s—but people were working on the possibilities of such systems.

The middle range of the flight of a ballistic missile would be very high, far away, and very difficult to intercept. So from the start of the launch we turned to its final stages just before it struck. In that final

trajectory, all of the major parts of the launching rockets would have split off, and the incoming warhead would be something like an ice cream cone. It would be well stabilized to descend smoothly through the heavier parts of the atmosphere, the cone pointing forward and protected against the tremendous heat of reentry by shielding. Work was then being done on ablative materials, which would absorb and carry off the intense heat. The first time I heard of this idea was when I was asked to evaluate some Soviet intelligence referring to a mysterious hard wood coating. We were quite puzzled until we realized that hard oak would be a very good ablating coating for the nose cone of a missile: the wood would char, and the char would actually be cooled as carbon particles evaporated into the atmosphere, going instantly from solid to liquid to gas and in the process carrying a lot of heat away. Also, as the charring continued and deepened it would become a very good heat insulator. In fact, that’s the kind of coating that eventually came about, not wood but embodying the same principles of char simultaneously carrying away heat and insulating.

So those of us in the ABM business at that time considered these incoming cones with a great deal of interest, moving at great speed, not maneuverable since they were in the later stages ballistic, and very hard to hit. That the nose cone couldn’t duck gave some potential advantage to a defense that could fire at the expected trajectory of the missile. The first need was for radar to detect a very small nose cone with very low radar reflectivity. Second was a computer system that would integrate and analyze the radar information in time to launch a defending missile.⁶² Another part of the puzzle was what kind of destruction was needed, whether one could use just shrapnel and let the incoming nose cone destroy itself when it hit shrapnel or whether one could depend on a blast some distance out from the nose cone to destroy it. Of course, talk about damage at a distance immediately invoked nuclear weapons.

One of the largest technical problems then—and still today—was an ABM system differentiating the real thing from decoys. Of all the various schemes we thought about, the best one rested on weight differences. The decoys were likely to be lighter than the actual weapon and hence would decelerate faster from atmospheric drag on entry into the atmosphere and then be detectable. We spent a lot of time calculating how heavy decoys could be and how many were needed and still have an

operational missile. Of course, if one could send decoys, you could with a little more push in the rockets also send multiple warheads. That would make defense even more difficult.

We also looked at reshaping the nose cone so that it slowed down just as the decoys did. That stratagem was unlikely to work because, if the nose cone was shaped differently than the decoys, it would have a different and likely detectable radar signal. So the game of measure and countermeasure was played but all on paper, since at the time there were neither ICBMs nor the kind of technologies available that would be embodied later in, say, the Nike series.

We had our final meeting in late March 1955, concurrently with a spring meeting of the SAB and also with the SAB nuclear weapons committee established in 1953 and chaired by John von Neumann. Johnny, who died in 1957, led the panel in drawing for the Air Force the future role of nuclear and thermonuclear weapons. We had a joint session with Johnny’s panel to discuss how nuclear weapons fit into the whole ballistic missile defense program, specifically the size and type of nuclear weapons likely to be available by the end of the 1950s. In it we emphasized the urgency of ICBM development.

Even before our report was finished in May, I had to begin briefing it, first to the full SAB. While we felt that it might be possible to intercept a missile as it was coming down, intercepting it in space was then beyond the state of the art. Systems had been looked at to do it, but the technology just wasn’t there. But ground interception fit into the aircraft defense role given the Army, which was pursuing it very strongly through its Nike program. I was a little disappointed that we weren’t able to come up with stronger and more brilliant recommendations on ballistic missile defense. The ones we did offer were good as far as our technology went, and several would come into play as new technology developed.

At this time in my life, I established a great friendship with Jimmy Doolittle, a friendship that lasted until his death in 1993. Jimmy had retired to the business world after his exploits in World War II but came in from time to time as a special consultant to the Air Force chief of staff.

Jimmy was a superb mentor—he was so good with people, thinking of their interests, receptive to their ideas, and throwing himself fully into anything he did with them. He had a strong role in brokering relationships between the Air Force and the White House at a time of tectonic changes in thinking about national security, most sharply expressed in the charge to the Killian Committee, of which he was a member. Indeed, he was a close associate of Jim Killian, a member of the MIT Corporation, and a distinguished graduate, earning a doctorate in aeronautics in the mid-1920s—indeed, he was one of the first in the United States to earn this degree.

My personal relationship with Jimmy Doolittle did not end when he finally finished his tour with the SAB in 1958. He helped me get established with a small club of avid landlocked salmon fishermen in northern Maine. The club opened only when the lake was ice-free, the smelt were numerous, and the salmon would rise to feed on them. Fly fishing with smelt-like lures was done from a canoe motoring just fast enough to make our imitation smelt attractive to a salmon. It proved at times very effective. One fisherman sat in the middle, another in the bow, and the guide in the stern, handling a paddle or a small outboard. Jimmy and I also spent summers climbing in the low coastal mountains of California, doing so until he was close to death. When he was in town during my time in Washington as chief scientist, he would come to our home for dinner. Our children remember him for his wonderful Doctor Doolittle stories. Jimmy taught me a lot, not the least of which was to keep my speeches short! I did.

Washington was justly famed for its summer heat and humidity. In those days, the mid-1950s, it was also known for the fact that along with most everybody else it had no air conditioning. The children could take it better than their parents could, but we also had gotten used to living in Boston, where in the summertime a cooling east wind often prevailed. What made the 1955 summer terrifying was a serious polio outbreak the summer before. Despite the arrival of the Salk vaccine, people were frantic that in the cities polio would again become a serious threat. Bunny’s mother, Lillian Newell Risley, herself not too well, thought this situation too frightening and invited Bunny and the children to come to their place in Randolph, New Hampshire. It turned an ominous summer into

a glorious one, but it ended sadly. The day we returned to Washington for the start of school we got a telegram that Nana Risley was dead, her heart gimpy for all the dozen years of our marriage had plain given out. We thanked her in our prayers.

That sad change in our lives was accompanied by many others. My term as chief scientist would nominally be done by the end of the year, and I was to face new challenges on my return to MIT.⁶³ More broadly, I was now enmeshed as one cog in the deterrence and defense posture the country had adopted. For deterrence the Eisenhower administration had adopted the policy of massive retaliation, the threat of using nuclear weapons delivered on an aggressor by bombers, cruise missiles, and soon ICBMs. For defense the country was to depend on early-warning radars, fighter aircraft, and long-range antiaircraft missiles, such as the Army’s Nike series. And just to keep things interesting, another card was soon to be added: entry into space, to detect what the enemy was doing and to attack him.⁶⁴