On arriving in Cambridge, Bardeen proceeded immediately to Harvard’s Jefferson Laboratory, a massive three-story brick structure completed in 1883. It was the first university physics building in America designed for research as well as instruction. Bardeen climbed its granite staircase and found the department office, where he spent some time on administrative chores.

He next checked into Lowell House, the student dormitory complex where he would live for the next three years. He found Building G in a group of adjoining brick buildings surrounding a courtyard. He climbed to the third floor and, entering his new home, suite 31, noticed that his living room was equipped with a wood-burning fireplace. It was far too hot to even think of using it now. He opened several windows to let a breeze into the stuffy apartment.

He decided to unwind with a short stroll around Cambridge. Walking a block or so, he came to the Charles River. He stopped to watch a group of rowboats gliding past. Then he entered Eliot House, another Harvard residential complex resembling Lowell House. He tried to find the room where the junior fellows would gather once a week for a formal dinner. After this brief tour, John went back home to Lowell House to unpack.

If Princeton was a graduate student’s paradise, then Harvard was a postdoc’s heaven, especially for a junior member of the Society of Fellows. When the society was established in 1933, “The idea was to give promising young students free time to do whatever they wanted in their research.” Bardeen soon realized: “There were no obligations.” They could work in any area they found compelling, with access to all of Harvard’s libraries, laboratories, and other facilities. Bardeen would flourish there.

The deeper aim of the society was to groom exceptionally promising men (and later women) between twenty and thirty years of age for intellectual nobility. Selected by the senior fellows of the society “for their promise of notable contribution to knowledge and thought,” the younger members were expected to advance into positions of leadership in politics, science, literature, or the arts. Interdisciplinary interaction was encouraged, for the society’s elders believed in the benefits of study at the borders of disciplines “where growth is still going forward.” Seven or eight junior fellows were appointed each year. At the time Bardeen came into the society, about twenty junior fellows were in residence.

Bardeen might have reverted to his usual pattern of dividing his time between physics and games had not the society imposed certain social rituals on its fellows. The junior fellows were expected to eat lunch with each other on most days and to dine formally with the senior fellows every Monday evening at Eliot House. Bardeen came to consider these dinners “a great privilege and inspiration, even though I was mostly a listener.” After the traditional drinking of sherry—a time “when the ordinary small talk of academic life is passed”—the group sat down to dinner at a U-shaped table. The high point of most dinners was a discussion with some eminent guest, such as a Supreme Court justice, or a nationally known poet, invited by one of the senior fellows. These guests contributed to the “general broadening influence of the society on its members,” Bardeen later reflected.

In time, most of the junior fellows assumed positions of influence. Of the three studying physics in Bardeen’s class, one was Ivan Getting, then conducting experiments in cosmic-ray and nuclear physics. He later became vice-president of engineering and research at Raytheon, chairman of the Naval Warfare Panel, and president of the Aerospace Corporation in California. The other was the theorist James Fisk, who would climb the Bell Labs ladder to

become its president. Fisk would also serve on many high-ranking scientific committees, including Dwight D. Eisenhower’s President’s Science Advisory Committee (PSAC), where he was influential in his support of a nuclear test ban. Fisk’s term on PSAC would overlap with Bardeen’s. In later years Fisk became a Harvard senior fellow.

Bardeen also befriended many junior fellows in fields other than physics. One was David Griggs, a geophysicist who would later became noted for demonstrating the flow of rocks under high-pressure conditions. Bardeen and Griggs both took a strong interest in the famous high-pressure experiments of Harvard physicist Percy W. Bridgman. Other friends were the philosopher Willard Van Orman Quine, who was at the time immersed in mathematical logic, and the behavioral psychologist Burrhus Frederick Skinner. Bardeen’s career would benefit many times from his acquaintances with former Harvard junior fellows. For instance, in 1945 Fisk would be instrumental in bringing Bardeen to Bell Labs.

Bardeen’s interaction with the other fellows was not confined to intellectual matters. The mathematician Garrett Birkhoff sometimes borrowed Bardeen’s car “for the nominal sum of $3 per month in 1937–38 when, unbeknownst to each other, we were both courting our future wives in Wellesley.” Birkhoff “used John’s car one evening to drive to Jake Wirth’s restaurant in Boston for a German meal with three distinguished friends: Heinrich Brüning, the former chancellor of Germany, John von Neumann, and Stan Ulam. The front seat would only hold three of us, so Stan had to sit in the rumble seat.”

The junior fellows were not required to teach, but were encouraged to do so if they thought the experience might enhance their training. Bardeen decided to offer a course on general relativity, basing his lectures on notes he took in Robertson’s course at Princeton, supplementing them with material from Arthur Eddington’s authoritative text, The Mathematical Theory of Relativity. Eddington’s philosophy of science matched Bardeen’s evolving sense of what physics should be. Eddington wrote on the first page of his book, “The physicist is not generally content to believe that the quantity he arrives at is something whose nature is inseparable from the kind of operations which led to it.” He explained:

Bardeen found that he felt at home in a department of physics. He attended seminars and immersed himself in the physics literature. He continued his habit of stopping regularly at the library to read important new papers. At that time, as Bardeen later reflected wistfully, “It was still possible to read all of them.” He reread Bethe and Sommerfeld’s major 1933 article on the quantum theory of solids, a review of all the major solved and unsolved problems in the theory of solids.

By then Bethe had resettled in the United States. He was among the 100 or so German-Jewish physicists who fled Germany and came to England or America after Hitler came to power. In 1935, after spending two years in England, Bethe emigrated to the United States and accepted a permanent post at Cornell. He became the central figure in Cornell’s growing program of nuclear physics— one of the three most rapidly growing subfields of physics, along with solid-state and quantum electrodynamics. Having created the basis for graduate training in solid-state theory with the Bethe-Sommerfeld article (mostly written by him), Bethe now played the same role in nuclear physics, publishing a series of detailed nuclear physics articles in The Reviews of Modern Physics. They became known as the “Bethe Bible.”

Bardeen was attracted to Bethe’s no-frills, practical style of physics. The tables of experimental numbers that Bethe often included in his publications reassured Bardeen that the theory he described was in step with observed data. Bardeen also joined in Bethe’s appreciation for the American tendency to take part in team research. Bardeen readily accepted Bethe’s invitation to visit Cornell, where the two theorists conferred about applying solid-state theory to nuclear physics in calculating the energy level density of heavy nuclei.

Bardeen also appreciated the “style and approach” of two excellent British texts that appeared in 1936 on the quantum theory of solids: Nevill Mott’s and Harry Jones’s, Theory of the Properties of Metals and Alloys and Alan Wilson’s The Theory of Metals. The point of view of these texts was compatible with his own. Written to address the needs of experimenters, they stressed the “close

correlation between experiment and theory.” Not yet in print was Fred Seitz’s The Modern Theory of Solids, which would appear in 1940. The first comprehensive textbook in modern solid-state physics, Seitz’s text became an instant classic. The book would help to establish solid-state physics as a distinct area of research. It was used for decades in physics departments throughout the world.

While in Cambridge, Bardeen wrote a paper with John Van Vleck on a method for approximating the interactions between electrons in solids, the “tight-binding method.” “I had more interaction with Van Vleck than anyone else at Harvard. He was one of the first there who was interested in solid state, and so I’d talk about work with him, but did the work pretty much on my own.”

Bardeen often visited MIT where he interacted so much with John Slater and his students that “years later, in his memoirs, Slater remembered me [incorrectly] as a postdoc at MIT,” Bardeen said. Slater had been among the first to recognize the great opportunities offered by the Wigner-Seitz method. As the head of MIT’s Department of Physics since 1930, Slater encouraged students to compute the band structures of materials such as lithium, copper, and sodium chloride. Although interested in many of the same problems of solids that had excited Wigner, Slater approached them less creatively. He referred to himself as “of the prosaic, matter-of-fact type.”

Slater had taken his Ph.D. at Harvard in 1923, before any theorists in physics were on the faculty. Working under Percy Bridgman, he attempted to explain some of his advisor’s pioneering measurements on solids under high-pressure conditions. By increasing the pressure from a few thousand to 100,000 kg/cm², Bridgman had managed to alter the spatial ordering of the atoms in the crystal lattice. That, in turn, affected the material’s electrical, magnetic, thermal, and optical properties, which Bridgman had carefully measured.

Bridgman encouraged Slater to apply quantum mechanics to his data; in particular, to explain the compressibility of alkali halides (materials in the same class as ordinary salt). Having studied physics in the pre-quantum-mechanical era, Bridgman had no working knowledge of the modern theories. Slater did his best to develop an explanation in terms of a balance between the electrical attractions of alkali and halide ions, and postulated repulsive forces of an unknown nature. But he could not account for the source of

the repulsive forces. “I was convinced by these facts that the quantum theory of 1923 was not adequate to describe the nature of molecules and solids.” After studying for a time in European centers of physics research, he accepted MIT’s offer to become the chair of its physics department. Karl Compton, who had successfully built up Princeton’s programs in math and physics, was now the president of MIT and was encouraging more growth in these areas.

Conyers Herring spent two years (1937–1939) working in postdoctoral study at MIT under Slater, overlapping with Bardeen in 1937–1938. Herring had met Slater as a student at Princeton and had been impressed by the speed with which Slater produced physics papers. Slater would say, “Well here’s something that needs to be calculated,” and go on to quickly do so, writing down the equations, and “zip zip zip, he’s got a paper.” One of the papers that Slater wrote in his short time at Princeton was his classic work on the approximation method known as the Augmented Plane Wave method. Unfortunately, Herring soon recognized that Slater’s outlook “was somewhat different from mine.” He did not appreciate Slater’s emphasis on “grinding out” results.

The mismatch in style between Herring and Slater showed up when Herring encountered difficulty in calculating the energy of a distorted lattice. After some weeks Slater suggested that Herring drop the problem and do “something that you could get finished with.” Herring preferred “to just keep butting away at it.” His approach, like Bardeen’s, demanded perseverance. His continuing attempts to solve the problem of the distorted lattice paid off in his development of a far more powerful approximation method, the widely used orthogonalized plane-wave method. “And that was real pay dirt.”

Herring’s “strongest intellectual contacts” in Cambridge were with Bardeen, not Slater. Herring recalled “that the things he [Bardeen] was working on were very close to the things I was working on.” For example, Bardeen was calculating the wave function for various alkali metals, “and I wanted to do it for beryllium,” said Herring. “I used to go up to Harvard every so often and talk with him about some of these band techniques and I learned quite a few things from him.”

At MIT, Bardeen also interacted often with Slater’s cocky graduate student from California, William Shockley, who was completing his doctoral thesis under Slater on sodium chloride. Bardeen

had already met Shockley while he was studying under Wigner at Princeton.

Seitz and Shockley were in those years close friends. The two had taken a memorable drive together across the country in September 1932. Both were traveling from Stanford, California, to their East Coast graduate schools—Seitz to Princeton and Shockley to MIT. Seitz had found Shockley “strongly influenced by the Hollywood culture of the day, fancying himself as a cross between Douglas Fairbanks, Sr. and Bulldog Drummond, with perhaps a dash of Ronald Coleman.” Seitz felt uneasy when he realized that Shockley had stowed a loaded pistol in the glove compartment of his 1929 DeSoto convertible. “I was handy with a rifle at that time, but still looked askance at traveling thousands of miles in the company of a loaded pistol.”

One night during the trip Shockley startled Seitz by firing his pistol at a group of howling coyotes. Having encountered torrential rains, they had “parked the car off the highway at a rise” and were “wandering through the desert.” The next morning a gas station attendant warned them “two desperadoes were loose in the area.” He assured them “the local police had been alerted.”

Bardeen and Shockley often interacted during their year of overlap in Cambridge, 1935–1936. “We went to joint seminars. I talked about his work with him.” Bardeen enjoyed Shockley’s quick intelligence, imagination, and self-confidence. The two had many common interests and spoke at length about calculating wave functions for alkali halides. They also discussed Shockley’s current work on surface states, which Bardeen would draw on a decade later in his research that led to the transistor.

Shockley’s surface states work followed up on studies conducted several years earlier by the Soviet physicist Igor Tamm. Cutting off periodic structures at the surface of a sample, Tamm showed, gives rise to “dangling bonds” and causes the formation of surface states. Shockley showed that surface states also arise when energy bands cross.

Shockley’s thesis calculation required many hours of number crunching on office calculators. Because it was easy to introduce errors during this process, Shockley devised checks along the way, including a novel “empty lattice test,” to see if the approximations provided agreement for the case of free electrons. It was a “case in which we knew what the exact answers would be,” Shockley explained, an example of bridging unknown with known physics.

In March 1936, during Bardeen’s second semester at Harvard, Shockley received an enviable job offer. The director of physical research at Bell Telephone Laboratories, Mervin J. Kelly, had received his physics doctorate under the noted University of Chicago physicist Robert Millikan. Kelly was one of the first leaders of industrial research to recognize the promise that solid-state physics held for industry. He learned about Shockley on a recruiting visit to MIT. When he then heard that Shockley was preparing to accept a physics instructorship at Yale, Kelly quickly countered with an offer of a research position at Bell Labs having better terms. Kelly assured Shockley that Bell Labs would offer him the freedom to conduct basic research in solid-state physics. Shockley was working at Bell Labs by July 1936.

Herring recalled that Shockley left behind a “residue of stories” at MIT. In one Shockley had cleverly rewired an elevator “so that when you pushed for floor one you’d go to floor four and when you pushed floor three you’d go to the basement. And then he took secret observations to see how long it took people to get used to it.”

Bridgman was perhaps the greatest influence on Bardeen at Harvard. The common intellectual “ancestor” of Van Vleck, Shockley, and Slater (the physics “grandfather” of Shockley through Slater, and of Van Vleck through Edwin Kemble), Bridgman impressed on his physics progeny his strong conviction that observation, not theory, was of primary importance in physics.

Bridgman was widely known as a solitary researcher who spent far more time in his lab gathering data than he spent in the classroom. On weekends and in summers, Percy (Peter to his friends) would immerse himself in solitary gardening at his New Hampshire home or in solitary hikes in the mountains. Although usually courteous and helpful, he was known to be “fierce in his inner disdain of sloppy or wishful thinking.” Some of Bridgman’s colleagues noted that he “indulged in no elaboration that would not improve the measurements.” The economy with which Bridgman worked in his laboratory probably impressed the frugal Bardeen.

On the foundations of Bridgman’s work, wrote Bridgman’s biographer, Maila Walter, “the new field of high pressure physics was established.” Bridgman had invented an important tool when he was a graduate student that allowed him to apply high pressures to solids. In subsequent decades he improved the apparatus and made countless precise measurements of matter under extremely high pressures.

In the 1920s Bridgman had hoped that his student John Slater could explain some of his experimental results using the new quantum physics, but the quantum theory was not yet sufficiently developed. A decade later, when Bardeen appeared on the scene, Bridgman showed him some of his data on alkali halides. Bardeen was excited to have the opportunity to work with the unsorted data of this master experimenter who, Bardeen recognized, was “at the height of his productivity.”

Bardeen’s teamwork with Bridgman established a model for Bardeen of highly productive collaborative work between a theorist and an experimentalist. He would in the years ahead create similar collaborations whenever possible. The two often discussed details of measurement techniques and experimental strategies. Bridgman suspected that through pressure alone he could bring about a phase change in cesium at about 22 kilobars. Before attempting this feat, however, he wanted to be armed with a better theory. He discussed with Bardeen why he expected the high pressure to pack the ions more closely and cause the cesium to transform from its “body-centered” structure to the more densely packed “face-centered” structure. Bridgman suspected that the change was related, somehow, to the atomic forces involved, but he did not know enough quantum theory to explore this hunch. Here was Bardeen’s chance to stand up in the physics community he had recently adopted and address a problem posed by one of its most prominent senior members.

Bardeen constructed a semiempirical theory that allowed him to predict the exact pressure at which cesium would change from a body-centered to a face-centered structure. When Bridgman tried the experiment, he found the transition at just the pressure Bardeen had predicted. “This was one of the first cases of being able to calculate a structure change before it was found experimentally,” Bardeen boasted. “Bridgman was very impressed with John,” claimed the Harvard physicist Edward Purcell. “For the first time one could calculate things that Bridgman was gathering data for.”

In another calculation, Bardeen used atomic spectral data to extend his earlier theory of the work function and compute the binding energy of lithium and sodium as a function of both volume and pressure. “The alkali metals were of particular interest to me,” Bardeen explained, “because there are large changes in compressibility over the 45 kbar range which should be calculable from a

first-principles Wigner-Seitz calculation.” He readily engaged in the long calculations. “Many days were spent working with a desk calculator on a calculation that could be done in minutes with modern computers.” For the case of sodium, the results “checked very well with experiments over the entire pressure range.” Unfortunately, there were small but significant departures (about 15 percent in compressibility) for the case of lithium. The problem, Herring discovered somewhat later, came from using the preliminary, rather than the final, values for the electric field, which Seitz had inadvertently published. “John never forgave me,” joked Seitz.

None of Bardeen’s colleagues recalled Bardeen ever mentioning the philosophy of science for which Bridgman is remembered today. “Operationalism” was Bridgman’s attempt to explain how physicists acquire scientific knowledge and achieve clear thinking. Maila Walter interprets Bridgman’s search for scientific meaning as a way of coping with the new quantum physics, for which he “was intellectually and emotionally unprepared.” Bridgman wrote copiously on operationalism during the 1920s and 1930s; Bardeen could not have remained untouched by the grounded convictions of the man who developed this philosophy.

“The concept is synonymous with the corresponding set of operations,” Bridgman wrote in 1927 in his celebrated monograph, The Logic of Modern Physics. He considered the operations by which measurements are made the true basis of scientific knowledge; physics concepts were merely “penumbra” surrounding the operations. Bridgman believed that studying Einstein’s route to relativity would help students understand how important new concepts could and should be formulated. He explained that Einstein’s writings reveal how he formulated the theory of special relativity starting with the operations basic to the concepts of time, length, and simultaneity.

In his later years, Bridgman also emphasized the importance of what a theory does—how it helps physicists make sense of the world and contribute to social good. The emphasis on usefulness resonated with the progressive values of physicists nurtured in the early years of the twentieth century.

Bridgman had another, less tangible but possibly as important, influence on Bardeen. Like four of Bardeen’s other teachers—Peter Debye, Paul Dirac, Eugene Wigner, and John Van Vleck—Bridgman was a future Nobel laureate. In 1946 Bridgman became the first

physicist on the Harvard faculty to win a Nobel Prize, received for the invention of an apparatus for producing high pressures and for his pioneering experiments establishing the field of high-pressure physics. (Dirac’s Nobel Prize in 1933, shared with Erwin Schrödinger, had been for fundamental work on the theory of quantum mechanics; Debye’s prize in 1936 was for chemistry. Seven years after Bardeen first became a Nobel laureate in 1956, Wigner would share his Nobel Prize with Maria Goeppert Mayer and J. Hans D. Jensen for the application of group theory to the atomic nucleus. And in 1977 Van Vleck would share the Nobel Prize with Philip Anderson and Nevill Mott for their studies of the electronic structure of magnetic and disordered systems.)

The Nobel Prize connections between certain scientist mentors and their students have been examined by the sociologist Harriet Zuckerman in her classic study of Nobel Prize–winning scientists. She suggests that the scientists who are destined to become Nobel laureates and their future Nobel laureate mentors recognize each other intuitively. There is a reciprocal relationship between them. The mentors “selectively recruit” students of Nobel caliber, and the students “self-select” to work with these mentors, often years before either receive their awards. According to this interpretation, in studying with physicists who were “Nobel-bound,” Bardeen was “being socialized” for a “position in the aristocracy of science.”

At Harvard, Bardeen continued to concentrate on the type of problems he had begun to address in his thesis calculation—“many-body” problems, in which the interactions between electrons, or between electrons and the lattice, play a significant role. He was often frustrated because physics was not yet equipped with adequate tools for treating these interactions. Indeed, he could not solve most of the problems he struggled with at Harvard. He worked on them anyway. John’s brother Bill had once described this attitude of his younger brother in the context of their playing football as youngsters. “John just hangs on and won’t let go.”

One of the problems Bardeen worked on at Harvard concerned the sharpness of the “Fermi surface,” a two-dimensional surface drawn in momentum space. This construction bounds the physical states occupied by the electrons in a metal. By the 1950s, Fermi surface studies would become an active topic of research, but in the 1930s few physicists recognized the importance of this geo-

metrical surface. An exception was Bethe, who discussed the Fermi surface in his 1933 Handbuch article with Sommerfeld. “It was clear to me that [these Fermi surfaces] would be important,” Bethe later said, “and that it made a great difference whether they were nearly a sphere or were some other interesting surface.”

One of the mysteries of the Fermi surface was the fact that states inside are fully occupied, whereas those outside are completely empty in the ground state. This puzzling property known as the “sharpness” of the Fermi surface had been observed at MIT in 1934, in Henry O’Bryan and Herbert Skinner’s measurements of the low-energy X-ray emission spectrum of light metals. Many-body effects, such as electron exchange or correlation interactions and particle motion due to finite temperatures, were expected to destroy the sharpness, but they did not. Bardeen tried to explain this result, but he was unsuccessful. His contribution only reformulated the problem as a question of explaining why the scattering lifetimes of electrons at the Fermi surface grow longer as the temperature falls.

The sharpness question was eventually answered in 1951 by Viktor Weisskopf, who showed that the increase in lifetimes of particles close to the Fermi surface was a consequence of the Pauli exclusion principle. Not until 1957 did Joaquin Luttinger, Walter Kohn, and Arkady Migdal more fully explain the sharpness of the Fermi surface within the context of Lev Landau’s Fermi liquid theory.

Another many-body problem on which Bardeen worked unsuccessfully concerned nuclear energy levels. Collaborating with Eugene Feenberg, who had taken his doctorate under Kemble in 1933 and was now a postdoctoral research associate at Princeton, Bardeen attempted to move beyond his earlier discussions with Bethe on employing many-body methods in computing the density of the energy levels in heavy nuclei such as uranium. He also tried to extend earlier work by Wigner on isotopic spin to a study of the effects of symmetry on nuclear energy levels, but this too “was too elaborate a calculation for its time.”

Similarly, Bardeen tried—but failed—to explain the “exchange forces” that arise as a consequence of the symmetry of the quantum-mechanical wave function when two particles exchange position. These forces seemed to increase the calculated velocity of an electron at the Fermi surface all the way to infinity, an obviously anomalous effect of the mathematics employed. He suspected that

the difficulty arose from an incorrect formulation of the long-range part of the electron interaction. But in 1936 Bardeen could not yet understand the deeper meaning, namely that electron–electron interactions are electrically “screened”; the electrons in such a system attract positive charge to them and travel with a cloud of positive charge around them. The result is that the charge (of electrons plus their positive cloud) appears neutral. This calculation also did not result in a full publication (only an abstract), but the work prepared Bardeen for his later research.

Bardeen also struggled at Harvard with how to treat the electrical forces between displaced ions inside crystals in the presence of electrons that screen the ions. His simplifying assumption—that the unscreened potential of an ion moves along with ions in the crystal—was equivalent to the widely used approximation method known as the “random phase approximation” (RPA). Again he made no breakthrough on the problem itself.

Why then did Bardeen often describe his time at Harvard as “the most influential years of my life”? He was probably recognizing the critical training effect of his work in those years, a training that derived from his persistent efforts to overcome challenging problems of physics. In struggling with problems too difficult to solve, he was honing his skills as a theoretical physicist.

Explaining superconductivity was another of the problems that Bardeen grappled with unsuccessfully at Harvard. In 1911 the great Dutch experimentalist Heike Kammerlingh Onnes had observed that at very low temperatures, certain metals and alloys lose all their electrical resistance. Over the next four and a half decades nearly every theorist working in physics sought to explain this puzzling effect. The work resembled a search for the Holy Grail.

In the late 1920s and early 1930s, following the development of the quantum theory of solids, theoretical physicists engaged in a flurry of work that produced many theories of superconductivity. They all failed. In his 1933 Handbuch article with Sommerfeld, Bethe lamented that superconductivity still resisted quantum-mechanical treatment: “Only a number of hypotheses exist, which until now have in no way been worked out and whose validity cannot therefore be verified.” Nevertheless, Bethe remained confident that “despite lack of success up until now, we may assert that superconductivity will be solved on the basis of our present-day quantum mechanical knowledge.”

All the theories of superconductivity put forth at that time were based on novel assumptions about the electronic structure of superconductors. For instance, a theory of Felix Bloch, Lev Landau, and Yakov Frenkel assumed that the ground state of a superconductor bears a “spontaneous current,” suggesting that this is why the energy is at a minimum in this state. Bloch’s calculations, however, could not find any current in the minimum-energy state, a result that came to be known as Bloch’s first theorem on superconductivity. He soon added a second theorem: “Every theory of superconductivity can be disproved!” This widely quoted second theorem reflected the frustration of the many who failed to explain superconductivity in the 1930s.

Another conception was put forth by Niels Bohr and Ralph Kronig, who imagined the electrons in a superconductor joined in a chain or lattice. Walter Elsasser based his theory on the assumption of relativistic electrons, while Richard Schachenmeier assumed exchange forces between conduction electrons and bound electrons. Léon Brillouin associated superconductivity with electrons trapped in metastable states. All these theories were hampered by incomplete data and inadequate theoretical tools. As a postdoc just sensing his prowess, Bardeen could not resist pitting himself against the many who had failed to crack the riddle of superconductivity.

Developing a “feel” for what was known experimentally was to be another of Bardeen’s characteristic first steps in approaching a new physics problem. In his efforts to learn what experimenters already had observed about superconductivity in the late 1930s, Bardeen read a short monograph recently published by the Cambridge physicist David Shoenberg. Its message echoed Bethe’s: “Although wave-mechanics has found a satisfactory qualitative explanation for most metallic phenomena, superconductivity has remained as anomalous as ever from the theoretical point of view.” Recent experiments, Shoenberg wrote, “have not materially changed this position,” but “they have at least made possible a coherent statement of what it is that the theory has to explain.” Fifteen years later Bardeen would recommend Shoenberg’s book, by then in its second edition, to his postdoc Leon Cooper.

In 1933 a breakthrough in superconductivity occurred. The London brothers, Fritz and Heinz, published a speculative theory that had remarkable predictive power, although it did not make

use of the full machinery of quantum mechanics. At the time Bardeen was among the few who recognized the importance of this theory. The Londons had fled Hitler’s Germany. Settling temporarily in Oxford, they collaborated in formulating a theory of superconductivity designed to explain the surprising result, reported in 1933 by Walther Meissner and Robert Ochsenfeld, that superconductors expel magnetic fields. This phenomenon implied that the transition between the normal and superconducting state of a metal is reversible and can thus be described by thermodynamics.

The Londons devised two equations relating the superconductor to the electric and magnetic fields. The vanishing of the resistance observed by Meissner and Ochsenfeld followed from the second equation. What the Londons realized, and expressed in their second equation, is that what is proportional to the electric field in a superconductor is not the current, as in the normal case, but the change in the current with time. If there is no electric field, the change in current will be zero, and an existing current will flow forever.

Solving the “London equations” together with Maxwell’s equations for the electric and magnetic fields yielded the observed experimental results, for instance, that the magnetic field decays exponentially as it enters the superconductor, with a penetration depth between 10^–6 and 10^–5 cm. Moreover, the current flow in the penetration layer near the surface shields the interior from the external field, the so-called “Meissner effect.” The Londons’ assumption that the normal current (which exists along with the supercurrent and satisfies the usual Ohm’s law of resistance) is short-circuited by the superconducting current under steady-state conditions explained how superconductors respond to electromagnetic waves.

Attempting to place this theory into a quantum-mechanical framework, the Londons suggested that there is “rigidity” of these ground-state wave functions in the presence of a magnetic field. In other words, if one applies weak fields, there is no effect on the wave function. But if one applies a sufficiently large magnetic field, the system ceases to be superconducting. An analogy would be pounding on a table. If one pounds gently, nothing happens because the table is rigid. But if one applies great force, the table will break.

Why should the wave function be rigid? The Londons said it was so because there exists an energy gap in the electronic structure

of the solid, a finite interval between the energy of the quantum-mechanical ground state and the low-lying excited states. Because of this gap it takes a great deal of energy to reach the lowest excited state. Bardeen sensed that this notion of the gap was the key to explaining superconductivity.

Physics was not the only challenge Bardeen faced while at Harvard. Although he thought that his romance with Jane was flourishing, she disagreed. Nearing thirty, she wanted to know whether she and John were going to be a permanent couple. Unless they could see more of one another, how could they decide? “Every time I saw him I liked him better,” but “I got small doses.” It frustrated her that John avoided all talk of marriage. “In those days, it was thought that you should support your wife if you got married,” Bardeen later explained. He could not yet do so.

Jane made a move. She again arranged to work at Woods Hole in Massachusetts. While studying the reproductive patterns of wasps there for her master’s thesis, she had many opportunities to see John during the summers of 1936 and 1937. From Boston she wrote home, “Dinner at the Wayside Inn Thursday night, the Pops concert at Symphony Hall Friday night, dinner at the Blue Ship on T Wharf and a drive last night… . John and I are having a grand time.” After that weekend, they drove to Pennsylvania to see Jane’s family.

As John continued to avoid commitment, she took another step. “Having decided that absence had not produced the result I wanted, I decided to try presence.” In August 1937 she wrote home that she was “resigning from Tech” and taking a position as a teacher of biology at the Dana Hall School in Wellesley, Massachusetts. Jane’s family knew that John was nearby in Cambridge. “But his proximity has not been the influence you may think it is in bringing me to move East,” she protested. She claimed she had “left that factor out, since I still am not sure he would make me a good husband.” She rationalized that she had done all she could with regard to her career at Carnegie Tech and that in Massachusetts she would be able to take courses toward her doctorate at Radcliffe or Wellesley. “I can’t pass up all chance for getting ahead in my own profession while waiting for some man to discover what a good wife and mother I could be.” Her family was not fooled.

During her job interview at Dana Hall, the schoolmistress, Miss Helen Cooke, asked Jane whether she had any “plans.” Jane prevaricated that she had none. She saw no need to volunteer her hopes of a future with John, for it was frowned on for a married woman to work outside the home. By November Jane’s Dana Hall colleagues were showing “great interest in ‘that nice young man from Cambridge,’ who comes to see me occasionally and frequently telephones when I am out.”

Jane still expressed reservations about John. “He really is fond of me and always wants to be with me when he seeks entertainment, but he works more than he plays and Physics is his first love.” Almost in the same breath, she resolved to be “more philosophical about the situation,” because “I do not see how there can ever be any other man for me.”