sign in sign up
<<
Home , Clinton Davisson, Lester Germer

: fifty years ago

A look back at the experiment that established the wave of the electron, at the events that led up to the discovery, and at the principal investigators, and .

Richard K. Gehrenbeck

An article that appeared in the December Before finishing his undergraduate discovery of ; it was 1927 issue of Physical Review, "Diffrac- degree at Chicago, he became a part-time also at this time that he was joined by a tion of by a Crystal of Nickel," instructor in at Princeton Uni- young colleague, Lester Halbert Germer, has been referred to in countless articles, versity, where he came under the influ- just discharged from active service. monographs and textbooks as having es- ence of the British Owen Rich- tablished the wave nature of the elec- ardson, who was directing electronic re- An adventurous New Yorker tron—in principle, of all matter.1 Now, search there. Davisson's PhD thesis at Germer was born on 10 October 1896, fifty years later, it is fitting to look back at Princeton, in 1911, extended Richardson's the first of two children of Hermann the events that led up to this historical research on the positive ions emitted from Gustav and Marcia Halbert Germer, in discovery and at the discoverers, Clinton salts of alkaline metals. Davisson later Chicago, where Dr Germer was practicing Davisson and Lester Germer. Figure 1 credited his own success to having caught medicine. In 1898 the family moved to shows them in their lab in 1927, together "the physicist's point of view—his habit Canastota in upper state, the with their assistant Chester Calbick. of mind—his way of looking at things" childhood home of Mrs Germer. Ger- from such men as Millikan and Richard- mer's father became a prominent citizen A shy midwesterner son. in the little town on the Erie canal, serving Clinton Joseph Davisson, the senior in- After completing his degree, Davisson as mayor, president of the board of edu- vestigator, was born in Bloomington, Il- married Richardson's sister, Charlotte, cation and elder in the Presbyterian linois, on 22 October 1881, the first of two who had come from England to visit her church. children. His father, Joseph, who had brother. After a honeymoon in Maine Germer attended school in Canastota settled in Bloomington after serving in the Davisson joined the Carnegie Institute of and won a four-year scholarship to Cor- Civil War, was a contract painter and Technology in Pittsburgh as an instructor nell University, graduating from there in paperhanger by trade. His mother, in physics. The 18-hour-per-week the spring of 1917, six weeks early because Mary, occasionally taught in the Bloom- teaching load left little time for research, of the outbreak of the war. The local ington school system. Their home was, and in six years there he published only newspaper, after applauding 18-year-old as Davisson's sister, Carrie, characterized three short theoretical notes. One nota- Lester for working as a laborer for the it, "a happy congenial one—plenty of love ble break during this period was the local paving contractors during his sum- but short on money." summer of 1913, when Davisson worked mer vacation, proceeded to ridicule his Davisson, slight of frame and frail with J. J. Thomson at the Cavendish lazier contemporaries for sitting "day throughout his life, graduated from high laboratory in England. after day in the lounging places of the school at age 20. For his proficiency in In 1917, after he was refused enlistment village," saying there is "nothin' doin' " mathematics and physics he received a in the military service because of his and that "a young feller has no chanst in one-year scholarship to the University of frailty, Davisson obtained a leave of ab- this durn town." (Lester, must have Chicago; his six-year career there was in- sence from Carnegie Tech to do war-re- taken a bit of ribbing from the "idle boys" terrupted several times for lack of funds. lated research at the after this appeared!) Germer's studies at He acquired his love and respect for Company, the manufacturing arm of the Cornell were partly self-directed; in their physics from Robert Millikan; Davisson American Telephone and Telegraph junior year he and two classmates, finding was "delighted to find that physics was Company, in New York City. His work themselves "unsatisfied with the course the concise, orderly science [he) had im- was to develop and test oxide-coated in electricity and magnetism given ... agined it to be, and that a physicist [Mil- nickel filaments to serve as substitutes for bought a more advanced text and met likan] could be so openly and earnestly the oxide-coated platinum filaments then regularly in the vacant class room .. . and concerned about such matters as colliding in use. At the end of World War I he really learned something." bodies." turned down an offered promotion at LTpon graduation from Cornell, Germer Carnegie Tech to accept a permanent obtained a research position at Western position at Western Electric. It was at Electric, which he held for about two Richard K. Gehrenbeck is an assistant professor this time that he began the sequence of of physics and astronomy at Rhode Island Col- months before volunteering for the Army lege, Providence, Rhode Island. investigations that ultimately led to the (aviation section of the signal corps). He

34 PHYSICS TODAY / JANUARY 1978 0031-9228/78/3101-34 $0.50 © 1978 American of Physics apparently made no contact with Davis- years 1913 to 1916, before Davisson and Physical Review in 1920, concluding that son then. Lieutenant Germer, among Germer appeared on the scene. Never- positive-ion bombardment has a negligi- those piloting the first group of airplanes theless, because AT&T was deeply con- ble effect on the electron emission from on the Western Front, was officially cerned about the efficiency and effec- oxide-coated cathodes.' credited with having brought down four tiveness of its triode amplifiers—key With this problem settled, a related German warplanes. Discharged on 5 components in its recently constructed question came up: What is the nature of February 1919, Germer was treated in transcontinental telephone lines—Arnold secondary electron emission from grids New York City for severe headache, ner- assigned Davisson and Germer the task of and plates subjected to electron bom- vousness, restlessness and loss of sleep, conducting tests on oxide-coated cath- bardment? Davisson was assigned this conditions attributed to his military odes. They published their results in the new task and given an assistant. Charles campaigns, but he refused to file for compensation because "others were worse BELL LABORATORIES off." After three weeks of rest, he was re-hired by Western Electric—and had as his first assignment the preparation of an annotated bibliography for a new project being directed by his new supervisor, Davisson. That fall Germer married his Cornell sweetheart, Ruth Woodard of Glens Falls, New York. Electron emission—in court The assignment that engaged Davisson and Germer in their first joint effort re- flects one of the chief interests of the parent company, AT&T, at this time: to conduct a fundamental investigation into the role of positive-ion bombardment in electron emission from oxide-coated cathodes. Although Germer later re- membered this project as having been directly related to the famous Arnold- Langmuir patent suit, that occupied Western Electric (Harold Arnold) and General Electric (Irving Langmuir) from 1916 until it was finally settled2 by the US Supreme Court (in favor of Western Electric) in 1931, a careful examination of the documents makes it clear that Dav- isson and Germer's project could have related to it only in a very indirect way. The patent case concerned improvements Davisson, Germer and Calbick in 1927, the year they demonstrated electron diffraction. In their New York City laboratory are Clinton Davisson, age 46; Lester Germer, age 31, and their assistant to the earliest deForest triode tubes with Chester Calbick, age 23. Germer, seated at the observer's desk, appears ready to read and record metallic (tungsten or tantalum) cathodes; electron current from the galvanometer (seen beside his head); the banks of dry cells behind Davisson it dealt with evidence obtained in the supplied the current for the experiments. Figure 1

PHYSICS TODAY , JANUARY 1978 35 link together their data, the model and the predictions were anything but defi- nite—quite out of keeping with the Rutherford-Geiger-Marsden tradition. Although Davisson (and Kunsman) must have been somewhat disappointed at the limited success of their initial ven- ture, they pressed on with additional ex- periments. In the next two years they built several new tubes, tried five other metals (in addition to nickel) as targets, developed rather sophisticated experi- mental techniques at high vacuum ("the pressure became less than could be mea- sured, i.e., less than 10~* mm Hg,") and made valiant theoretical attempts to ac- count for the observed scattering inten- 50 100 150 200 sities. The results were uniformly un- ENERGY OF SCATTERED ELECTRONS (eV) impressive; several of the studies were not Electron-scattering peak. The energy of the scattered electrons varies from almost zero to that even published. In fact, the generally of the incident beam (indicated by the arrow). This is a reconstruction of the type of observation disheartened atmosphere that seems to that led Davisson and Charles Kunsman to conclude that some electrons were being scattered have prevailed by the end of 1923 is indi- elastically. Davisson saw these as possible probes of the electronic structure of the atom, in analogy cated by the fact that Kunsman left the to Rutherford's use of alpha particles to explore the nucleus. Figure 2 company and Davisson abandoned the scattering project. H. Kunsman, a new PhD from the Uni- Davisson's own words, A year later, however, Davisson was versity of California. For this work they "The mechanism of scattering, as we ready to have another try at electron were able to convert the positive-ion ap- pictured it, was similar to that of scattering. Was this change of heart paratus to an electron-beam apparatus. alpha ray scattering. There was a prompted by Davisson's strong attraction Meanwhile Germer was shifted to a certain probability that an incident to the project? Was it his eagerness to project on the measurement of the electron would be caught in the field obtain additional information about the thermionic properties of tungsten, a topic of the atom, turned through a large extranuclear structure of the atom? In he pursued for about four years, both angle, and sent on its way without loss any case, in October 1924 Germer was put under Davisson's direction and as part of of energy. If this were the nature of back on the scattering project in place of a graduate program he undertook at it would be possi- the departed Kunsman. Germer, who nearby Columbia University part time. ble, we thought, to deduce from a sta- had already completed several therm- tistical study of the deflections some ionic-emission studies, was returning to A startling observation information in regard to the field of Western Electric after a 15-month illness. Soon after Davisson and Kunsman the deflecting atom . . . What we were Regarding his development as a physicist began their secondary electron emission attempting . . . were atomic explora- by this time. Germer later recollected: studies, they observed an unexpected tions similar to those of Sir Ernest "I learned relatively little at Columbia phenomenon that was to have crucial Rutherford ... in which the probe . . . but was nevertheless fortunate in importance for their future experimental should be an electron instead of an working . . . with Dr C. J. Davisson. I program: A small percentage (about 1%) alpha particle." learned a simply enormous amount of the incident electron beam was being In fact, Davisson was so enthusiastic from him. This included how to do scattered back toward the electron gun about a full-scale assault on the atom that experiments, how to think about with virtually no loss of energy—the he was able to convince his superiors to let them, how to write them up, how even electrons were being scattered elastically. him and Kunsman devote a large fraction to learn what other people had pre- Figure 2 reconstructs this phenomenon. of their time to it, and to give them the viously done in the field ... I am quite Previous observers had noticed this effect necessary shop backup. certain that I do really owe to Dr Dav- for low-energy electrons (about 10 eV), The basic piece of apparatus, built to isson much the best part of my educa- but none had reported it for electrons of order by a talented machinist and glass- tion, and I am not really convinced energies over 100 eV. blower, Geroge Reitter, was a vacuum that it is so inferior to that obtained in Although this discovery undoubtedly tube with an electron gun, a nickel target more conventional ways. It is cer- had no immediate impact on the stock- inclined at an angle of 45° to the incident tainly different." holders of AT&T, it affected Davisson electron beam and a Faraday-box collec- profoundly. To him these elastically tor, which could move through the entire A "lucky break" and a new model scattered electrons appeared as ideal 135° range of possible scattered electron So the scattering experiments were fi- probes with which to examine the ex- paths; it is diagrammed in figure 3. The nally resumed. One can easily imagine, tranuclear structure of the atom. Ernest Faraday box was set at a voltage to accept then, the feelings of disappointment and Rutherford announced his nuclear model electrons that were within 10% of the in- frustration that Davisson and Germer of the atom in 1911, the year Davisson cident electron energy. must have shared when, soon after the completed his PhD; Hans Geiger and After two months of experimentation, project had been restarted, they discov- Ernest Marsden completed their defini- Davisson and Kunsman submitted a ered a cracked trap and badly oxidized tive experimental tests of Rutherford's two-column paper to Science, in which target on the afternoon of 5 February theory and announced his they sketched the main features of their 1925, as the notebook entry in figure 4 planetary model of the atom in 1913, scattering program, presented a typical shows. What it meant in simple terms when Davisson worked with Thomson at curve of their data, proposed a shell model was that the experiments with the spe- Cambridge. So it is not surprising that of the atom for interpreting these results, cially polished nickel target, discontinued Davisson was enthusiastic about the and offered a formula for the quantitative for almost a year, were to be delayed prospect of using these electrons for basic prediction of the implications of the again. Apparently Germer's attempts to research on the structure of the atom. In model.1 I Unfortunately their attempts to revitalize the tube after its long period of

36 PHYSICS TODAY / JANUARY 1978 disuse by repumping and baking (out- gassing) were to be for nought; an addi- tional delay for repairs was necessary. Electron gun This was not the only time that a tube had broken during a scattering experi- Nickel target Incident beam ment, nor was it to be the last. Nor was the method of repair unique, for the method of reducing the oxide on the nickel target by prolonged heating in Filament vacuum and hydrogen had been used once before (unsuccessfully; that time it had led to the formation of a "black precipi- Scattering angle tate" and "no apparent cleaning up of the nickel"). This particular break and the subsequent method of repair, however, had a crucial role to play in the later dis- covery of electron diffraction. By 6 April 1925 the repairs had been completed and the tube put back into operation. During the following weeks, as the tube was run through the usual se- Faraday box (collector) ries of tests, results very similar to those obtained four years earlier were obtained. Then suddenly, in the middle of May, unprecedented results began to appear, as Arc on which Faraday box rotates shown in figure 5. These so puzzled Davisson and Germer that they halted the experiments a few days later, cut open the To galvanometer tube, and examined the target (with the assistance of the microscopist F. F. Lucas) to see if they could detect the cause of the Scheme of the first scattering tube, which served as a prototype for the group's later models. new observations. Davisson and Germer later included mechanisms for rotating the target azimuthally 360° about the beam axis and for changing the angle of the incident beam with respect to the normal to the What they found was this: The poly- target. In their 1926-27 work the incident beam was perpendicular to the target face, and the crystalline form of the nickel target had scattering angle was called the "colatitude angle." Figure 3 been changed by the extreme heating until it had formed about ten crystal fac- ets in the area from which the incident fleeted their expectation of finding certain ing arrangements for the children: "It electron beam was scattered. Davisson "transparent directions" in the crystal seems impossible that we will be in Oxford and Germer surmised that the new scat- along which the electrons would move a month from today—doesn't it? We tering pattern must have been caused by with least resistance. They expected should have a lovely time—Lottie dar- the new crystal arrangement of the target. these special directions to coincide with ling—It will be a second honeymoon— In other words, they concluded that it was the unoccupied lattice directions. and should be sweeter even than the the arrangement of the atoms in the first." Something was to happen on this crystals, not the structure of the atoms, More than a "second honeymoon" particular trip, however, to turn it into that was responsible for the new intensity Having suffered disappointment with more than the "second honeymoon" pattern of the scattered electrons. the results of the original scattering ex- Davisson envisioned. Thinking that the new scattering pat- periments performed with Kunsman, Theoretical physics was undergoing terns were too complicated to yield any Davisson must have been doubly fundamental changes at this time. In the useful information about crystal struc- disheartened by the meager returns he early months of 1926 Erwin Schrodinger's ture, Davisson and Germer decided that and Germer obtained with the new tube. remarkable series of papers on wave me- a large single crystal oriented in a known After an entire year spent in preparation, chanics appeared, following Louis de direction would make a more suitable and with a new tube and a new theory in Broglie's papers of 1923-24 and Albert target than a collection of some ten small hand, they obtained experimental results Einstein's quantum-gas paper of 1925. facets randomly arranged. Because nei- that were even less interesting than those These papers, along with the new matrix ther Davisson nor Germer knew much from the earliest experiments. The new of , Max about crystals, they, assisted by Richard colatitude curves showed essentially Born and Pascual Jordan, were the Bozorth, spent several months examining nothing, and even the new azimuth curves subject of lively discussions at the Oxford the damaged target and various other gave at best only a weak indication of the meeting of the British Association for the nickel surfaces until they were thoroughly expected three-fold symmetry of the Advancement of Science. Davisson, who familiar with the x-ray diffraction pat- nickel crystal about the incident beam. generally kept abreast of recent develop- terns (note!) obtained from nickel crystals Davisson must have been quite pleased ments in his field but appears to have in various states of preparation and ori- with the prospect of getting away for a few been largely unaware of these recent de- entation. months during the summer of 1926, when velopments in , at- By April 1926 they had obtained a he and his wife had planned a vacation tended this meeting. Imagine his sur- suitable single crystal from the company's trip to relax and visit relatives in England. prise, then, when he heard a lecture by metallurgist, Howard Reeve, and cut, Mrs Davisson recalled that this summer Born in which his own and Kunsman's etched and mounted it in a new tube that had been chosen for the trip because her (platinum-target) curves of 1923 were sister, May, and brother-in-law, Oswald cited as confirmatory evidence for de allowed for an additional degree of free- 5 dom of measurement; the collector could Veblen of , were Broglie's electron waves! now rotate in azimuth (the 360° angle available to stay with the Davisson chil- After the meeting Davisson met with circling the beam axis) as well as in cola- dren at that time. As Davisson wrote to some of the participants, including Born titude. The design of the new tube re- his wife, then at the Maine cottage mak- and possibly P.M.S. Blackett, James

PHYSICS TODAY / JANUARY 1978 37 December. These attempts revealed only "very feeble" peaks. The situation changed dramatically on 6 January 1927, however; the data for that day are ac- companied by the remark, in Calbick's neat handwriting: "Attempt to show 'quantum bump' at an intermediate [co- latitude] angle. Bump develops at 65 V, compared with calculated value for 'quantum bump' of V = 78 V." Then, stretched across the bottom of the page in Germer's unmistakable bold strokes, is the additional remark: "First Appear- ance of Electron Beam." A portion of the notebook page is reproduced in figure 6. The data for this curve are extremely interesting. Noting from the figure that the readings were taken in one-volt in- The notebook entry (or 5 February 1925 records, in Germer's handwriting, the discovery of the tervals on either side of 79 volts, whereas broken tube that interrupted the scattering experiments once again. It was this break, however, the steps are 2, 5 and then 10 volts else- which initiated a chain of events that eventually led to the preparation of a single crystal of nickel where, we see that a peak was expected at as the target, and to a shift of Davisson's interest from atomic structure to crystal structure. Re- produced by courtesy of Bell Laboratories. Figure 4 about 78 volts. But the experiment yielded a single large current at 65 volts. The experimenters took immediate notice Franck and Douglas Hartree, and showed fashion, however, this quest was preceded of this spike, making a second run in them some of the recent results that he by a period of careful preparation, in- one-volt steps around 65 volts, which on and Germer had obtained with the single cluding an important change in the ex- a graph shows a clear peak centered on 65 crystal. There was, according to Davis- perimental tube. As Davisson wrote to volts. It is easy to imagine the excitement son, "much discussion of them." All this Richardson in November, that must have accompanied this sudden attention might seem strange in light of "I am still working at Schrodinger and turn of events, moving Germer to sprawl the relatively feeble peaks Davisson and others and believe that I am beginning his glad tidings across the bottom of the Germer had obtained, but even these may to get some idea of what it is all about. page! have been exciting to already In particular I think that I know the With this single critical result in hand, convinced of the basic correctness of the sort of experiment we should make the experimental situation changed sud- new quantum theory. It may also reflect with our scattering apparatus to test denly. The next day, 7 January, they ran the fact that several European physicists, the theory." several additional voltage curves, one for Walter Elsasser (Gottingen), E.G. Dy- each of four different colatitude positions. mond (Cambridge, formerly Gottingen Found—a "quantum bump" A voltage peak appeared at a colatitude and Princeton), and Blackett, James It was three weeks before the "thorough angle of 45° that was even greater than Chadwick and Charles Ellis of Cam- search" was begun. The importance that that at 40°, where the collector had been bridge6 had attempted similar experi- Davisson (and ) had come to set the previous day. On the eighth, a ments and abandoned them because of attach to this project can be surmised new colatitude curve was run at a voltage the difficulties of producing the required from the addition to it of a new assistant, of 65 volts, and the first true and unmis- high vacuum and detecting the low-in- Chester Calbick, a recently graduated takable colatitude peak was observed— tensity electron beams. Apparently they electrical engineer. After about a month this was what Davisson had been looking were encouraged by these results, which of experimenting, during which time for since 1920! Skipping Sunday, they appeared so unimpressive to Davisson. Calbick took charge of operating the ex- next ran an azimuth curve at 65 volts and At any rate, Davisson spent "the whole of periment, they gave the newly prepared a colatitude of 45°. This time the three- the westward transatlantic voyage . . . tube a thorough set of consistency tests. fold azimuthal symmetry was immedi- trying to understand Schrodinger's pa- During one attempt by Germer to reacti- ately apparent. Figure 7 shows these pers, as he then had an inkling . . . that the vate the tube in late November the tube curves. explanation might reside in them"—no broke, but with little damage. (Strangely, The experiments that were carried out doubt to the detriment of the "second little damage can be considered "lucky" during the next two months show that honeymoon" in progress. in this case, whereas it would have been Davisson, Germer and Calbick, having Back at Bell Labs (as the engineering "unlucky" in the case of the 1925 finally found and positively identified one arm of Western Electric has been called break!) set of electron beams, could now find and since 1925), Davisson and Germer exam- The first experiments with the new identify others quickly. This block of ined several new curves that Germer had tube yielded no significant results; the experiments continued through 3 March, obtained during Davisson's absence. colatitude and azimuth curves looked when Calbick left for a month on family They found a discrepancy of several de- much as before, and the new experiments business. Comparing this with earlier grees between the observed electron in- added by Davisson "to test the theory" periods of Davisson's long contact with tensity peaks and the angles they ex- were uninformative as well. These tests electron scattering, we see that not since pected from the de Broglie-Schrodinger consisted of varying the accelerating the early days of the original Davisson- theory. To pursue this matter further voltage, and hence electron energy E, for Kunsman experiments had there been they cut the tube open and carefully ex- fixed colatitude and azimuth settings, and such intense and concentrated effort in a amined the target and its mounting. were designed to see if any effect could be single well defined direction. The pres- After finding that most of the discrepancy discerned for a changed electron wave- ence of a clear, unambiguous goal cer- could be accounted for by an accidental length A, according to the de Broglie re- tainly must have been a major factor in displacement of the collector-box open- lationship, X = h/(2mE)1/2. the two cases, an ingredient lacking at ing, they "laid out a program of thorough A concerted search for "quantum other times. search" to pursue the quest of diffracted peaks" (voltage-dependent scattered electron beams. In typical Davisson Another factor undoubtedly urging electron beams) was launched by late Davisson on to rapid (but careful) exper-

38 PHYSICS TODAY / JANUARY 1978 indentation and possible early publication True to form, however, Davisson and was his feeling that others might be Target Germer did not sit back and rest on a "job pursuing similar investigations at that Incident beam well done"; they recognized the consid- time. Recalling his conversations at erable work necessary to resolve a number Oxford and the comments that had been of questions still outstanding. Among made about the interest of others in this these were: matter, he sent off an article to Richard- • the problem of the "anomalous" beams son in March with the accompanying mentioned above, note: Davisson and Kunsman 1921 • the ad hoc "contraction factor" that "I hope you will he willing, if you they had found necessary to attribute to think it at all desirable, to get in touch the nickel crystal and with the editor of Nature with the • extension of their electron energies idea of securing early publication. over a greater range, and sharpening and We know of three other attempts that refining their diffraction peaks. have been made to do this same job, and naturally we are somewhat fearful Instant acclaim that someone may cut in ahead of us." Toward this end they initiated an ex- As it turned out these efforts had long tensive experimental and theoretical at- been abandoned, but he had no way of Davisson and Germer 29 April 1925 tack that lasted from 6 April (when Cal- knowing that. Nevertheless, another bick returned from his month's absence) investigator, unknown to Davisson at that until 4 August. At that time the tube was time, was indeed making progress at re- cut open for a final careful examination of vealing the phenomena of electron dif- the target and the other tube components. fraction with high-voltage electrons and As it turned out, this intention was foiled thin metal foils. This was J.J.'s son, G.P. when, in the process of being brought Thomson; his and Andrew Reid's first back to room temperature, the tube "blew note was published in Nature just one up and [was] partially ruined . .. the leads month after Davisson and Germer's.' being broken, filament also, and a large Davisson and Germer part of the nickel oxidized." A broken A conservative note and a bold one tube had served to initiate the decisive Davisson and Germer's Nature article Before and after the accident of 5 February experiments on 5 February 1925, and a 1925. Although the first scattering curves after broken tube ended them on 4 August was an extremely conservative expression the repair of the broken tube (middle curve) re- of the new experimental evidence for sembled the 1921 results of Davisson and 1927, two and a half years later. The 8 electron diffraction. Its title, "The Kunsman (top curve), striking peaks soon made cover of this issue of PHYSICS TODAY Scattering of Electrons by a Single Crystal a sudden appearance (bottom). This develop- shows the Davisson-Germer tube as it of Nickel," bears a closer connection to ment led Davisson and Germer to make a major appears today. the early work of Davisson and Kunsman change in their program. Figure 5 The most interesting of this last group than it does to the new wave mechanics. of experiments was a series designed to Although the paper included a table investigate "the anomalous peaks after linking the scattered electron peaks to the Davisson's thoughts (and certainly Ger- bombardment," which appeared for a corresponding de Broglie , it mer's as well) on the subject were not restricted period of time after the target was not until the last two paragraphs that nearly as reserved as the Nature article had been heated by bombardment. The a tentative suggestion was made about the suggests. experiments showed that the nature— important implications of the work: The One other public announcement of the even the existence—of certain beams was results were "highly suggestive ... of the recent discoveries was made at this time. not static but varied with temperature ideas underlying the theory of wave me- In a paper presented at the Washington and time (and hence conditions of the chanics." meeting of The American Physical Soci- target in terms of occluded gases). The This cautious attitude may have been ety on 22-23 April 1927 and abstracted in notebook entries include a great variety due to the problem that Davisson and the Physical Review in June,1" Davisson of different terms, diagrams and calcula- Germer had in making the proper corre- and Germer basically repeated what they tions designed to try to make sense out of lation between their data points and the had stated in their Nature article, and these data. Davisson and Germer found theory; they found it necessary to hy- then added an intriguing final paragraph. a "gas crystal" model, in which "gas atoms pothesize an ad hoc "contraction factor" Referring to the three anomalous beams fit into the crystal," to be the most effec- of about 0.7 for the nickel-crystal spacing that could not be fitted into the analysis tive. to get approximate correspondence be- in the Nature article, they suggested that The task of welding data and inter- tween the de Broglie wavelengths and these "offer strong evidence that there pretation into a comprehensive report for their data. Even at that, only eight of the exists in this crystal a structure which has publication was begun in mid June, well thirteen beams described were clearly not been hitherto observed for nickel." before the experiments were completed. amenable to this analysis. This statement implies Davisson and It appears that Davisson was responsible This cautious attitude appears to have Germer had already gone beyond the for most, of not all, of the writing; in a been abandoned in a concurrent article by point of using the "known" structure of letter to his family at the summer cottage Davisson alone for an in-house publica- the nickel crystal to find out about the he wrote: tion, the Bell Labs Record.9 The very possibility of the wave properties of the "I'm busy these days writing up our title, "Are Electrons Waves?" suggests electron; they were now using the experiment—It's an awful job for me. this difference. After reviewing the evi- "known" electron waves to learn new facts I didn't get much done yesterday as dence that led to think of about the nickel crystal. Between March, Prof. Epstein from Pasadena turned x rays as being wave-like, he cited his and when the Nature article was submitted, up and had to be entertained and Germer's recent work with electrons, and April, when the Phys. Rer. abstract shown things—and today I'm too slee- urging a similar conclusion in this case. was prepared, results that had been em- py [after having spent last evening at Although this article gave its readers no barrassing to the theory had become a the theater with Karl Darrow]. How- actual data on the experimental evidence potential new application of that very ever, I must keep at it." for electron waves, it clearly indicates that theory! More than three weeks later (23 July) he

PHYSICS TODAY / JANUARY 1978 39 BELL LABORATORIES inherent in the paper itseli. 74 This may be illustrated by some re- DATE marks made by prominent physicists .cr^ prior to the publication of t he Phys. Rev. ^ L-+9 article. In the reports and discussions of ^1 the fifth Solvay Conference held in r.ji Brussels in October 1927, Niels Bohr, de «i Broglie, Born, Heisenberg, Langmuirand r.f Schrodinger all hailed the experiments of m rv Davisson and Germer (as described in the rl.wcjfvrrj Nature article) as being, in the words of de Broglie, "very important results which [appear] to confirm the general provisions and even the formulas of wave mechan- ics."12 Bohr, speaking before the Inter- national Congress of Physics assembled in Como, Italy, on 16 September 1927, drew upon these experiments in estab- lishing his views on complementarity: "... the discovery of the selective re- flection of electrons from metal crys- tals . . . requires the use -of the wave theory superposition principle .., Just as in the case of light ... we are not dealing with contradictory but with complementary pictures of phe- nomena."1' Planck, addressing the Franklin Insti- tute on 18 May 1927, even before he had heard of the Davisson-Germer results, stated about the electron: "[Its] [in the atom] resembles . . . the vibrations of a standing wave . . . [Thanks] to the ideas introduced into science by L. de Broglie and E. Schrodinger, these prin- ciples have already established a solid foundation."14 Yet in the same address Planck stated that he was still (in 1927, four years after the decisive Compton experiments) reluctant to accept the corpuscular implications for electromag- netic radiation inherent in his own The sixth of January 1927 might well be regarded as the birthday of electron waves, for it was the quantum hypothesis! It appears that day that data directly supporting the de Broglie hypothesis of electron waves were first observed. physicists were willing to accept the ex- Note the peak deflection at 65 volts, and the detailed study of the region directly below. Calbick's perimental evidence for electron waves handwriting is neat and cautious; Germer's is bold and expansive. Davisson made no entries in almost before those experiments were any of the research notebooks kept in the Bell Labs files. Figure 6 performed! was at last able to exclaim, gested "contraction factor" or an "index The world that is physics "I finished the first draft of our paper of refraction" proposed by Carl Eckart, Davisson and Germer succeeded where this morning. It is going to take a lot A.L. Patterson, and Fritz Zwicky in in- others had failed. In fact, the others of going over and revising ... I will dependent responses to the Nature arti- mentioned above (Elsasser, Dymond, leave [the drawings) to Lester—and cle)." Summarizing the evidence, the Blackett, Chadwick and Ellis), who had also the thing is full of blanks in which paper concluded that of the 30 beams that the idea of electron diffraction consider- he will have to stick in the right num- had been observed, 29 were adequately ably ahead of Davisson and Germer, were bers." accounted for by attributing wave prop- not able to produce the desired experi- A week later he made his final changes erties to free electrons. It acknowledged, mental evidence for it. G.P. Thomson, before departing for Maine. Germer, too, however, that the wave assumption im- who did find that evidence by a very dif- needed a break, and after finishing his plied the existence of eight additional ferent method, testified to the magnitude tasks he left on 14 August for a canoe trip beams, which had not been observed. of the technical achievement as follows: with several friends. The final copy was The discrepancies between theory and "[Davisson and Germer's work] was sent to the Physical Review in August and experiment, apparently fairly minor, that indeed a triumph of experimental the article appeared in December. Davisson and Germer recorded, evidently skill. The relatively slow electrons The paper itself was a detailed, com- did not reduce their fundamental belief |they] used are most difficult to han- prehensive report on experiments per- that free electrons behave like waves. dle. If the results are to be of any formed, conclusions reached and ques- The physics community appears to have value the vacuum has to be quite out- tions left unanswered. One of the sig- concurred, for I have not found a single standingly good. Even now [1961]... nificant features of the paper was its voice raised in opposition. This may well it would be a very difficult experi- thoughtful examination of the possible have been due as much to the success of ment. In those days it was a veritable ways of interpreting the systematic dif- the earlier theory of wave mechanics and triumph. It is a tribute to Davisson's ferences between observed and calculated the acceptance of a wave-particle duality experimental skill that only two or electron wavelengths (either the sug- for light as to the force of the evidence three other workers have used slow

40 PHYSICS TODAY / JANUARY 1978 Incident beam

Nickel target

90 180 270 360 AZIMUTH (degrees)

New colatitude and azimuth curves. The black lines show the ap- The colored curves are from data taken after 6 January 1927, when the pearance of the colatitude (left) and azimuth (right) distributions of the first "quantum bump" was observed. The azimuth curves also confirm scattered electrons when Davisson took the curves to England in 1926. the threefold symmetry of the nickel crystal. Figure 7

electrons successfully for this pur- "Why Davisson and Germer, and not 1. C. J. Davisson, L. H. Germer, Phys. Rev. pose. "15 someone else?" one's thoughts leap to 30,705(1927). Davisson and Thomson shared in the such things as the "luck" of the broken 2. US Reports 283, 665 (1931). for physics in 1937 for their tube in 1925 and the trip to England in 3. C. J. Davisson, L. H. Germer, Phys. Rev. accomplishments. Germer and Reid, as 1926. Davisson and Germer themselves 15,330(1920). junior partners to Davisson and Thom- freely admitted the key importance of 4. C. J. Davisson, C. H. Kunsman, Science 54, son, did not share in the prize. Reid was these events. But to dwell on them ex- 523(1921). tragically killed in a motorcycle accident clusively would be a mistake. Neither of 5. M. Born, Nature 119, 354 (1927). shortly after his and Thomson's definitive these events would even have been re- 6. W. Elsasser, Naturwissenschaften 13,711 papers appeared in 1928. membered had they not been followed by (1925); E. G. Dymond, Nature 118, 336 Davisson and Germer actively pursued thorough, careful and creative experiment (1926). the topic of electron diffraction for about and reflection. Perhaps of equal impor- 7. G. P. Thomson, A. Reid, Nature 119, 890 three years after 1927, publishing, to- tance is the habit of attention to technical (1927). gether and separately, about twenty more detail established by Davisson in his stu- 8. C. J. Davisson, L. H. Germer, Nature 119, papers on the subject; reference 16 gives dent days and extended in the long series 558(1927). three of the most important. By the early of Davisson-Kunsman and earlier Dav- 9. C. J. Davisson, Bell Lab. Record 4, 257 1930's, both Davisson and Germer had isson-Germer experiments. Another (1927). turned to new fields: Davisson to elec- important factor is the time for pure re- 10. C. J. Davisson, L. H. Germer, Phys. Rev. tron optics (including early television); search provided by Western Electric-Bell 29,908(1927). Germer to high-energy electron diffrac- Labs, and the technical support in areas 11. C. Eckart, Proc. Nat. Acad. Sci. 13, 460 tion and later still to electrical contacts. such as high vacua and electrical detection (1927); A. L. Patterson, Nature 120, 46 Davisson retired from Bell Labs in 1946 techniques available at that industrial (1927); F. Zwicky, Proc. Nat. Acad. Sci. 13, and spent the remaining twelve years of laboratory. 518(1927). his life in Charlottesville, Virginia, sum- All in all, this case history on the dis- 12. L'Institut International de Physique Sol- mering as usual in Maine. Germer re- covery of electron diffraction appears to vay, Electrons et : Rapports et gained his interest in low-energy electron illustrate the complex nature of the world Discussions du Cinquieme Conseil de diffraction in 1959-60, at which time he that is physics, the difficulty of singling Physique, Gauthier-Villars, Paris (1928), and several co-workers at Bell Labs per- out any one factor as being responsible for pages 92, 127, 165, 173, 274, 288. fected a technique, eventually referred to a great discovery, and the importance of 13. N. Bohr, Atomic Theory and the De- as the "post-acceleration" technique,17 scription of Nature, Macmillan, New York establishing and nurturing the ties that (1934), page 56; italics supplied. which had been devised in 193418 and bind together the generations of physi- then abandoned, by Wilhelm Ehrenberg. 14. M. Planck, J. Franklin Inst. 204, 13 cists, as well as the physicists of each (1927). With this work Germer was able to follow generation. up with great success the study of sur- 15. G. P. Thomson, The Inspiration of faces, to which he had been attracted in Science, Oxford U.P., London (1961); re- References printed by Doubleday, Garden City, New his original work with Davisson; the field York (1968), page 163. of low-energy electron diffraction (LEED) References to correspondence, personal re- is now widespread and very active. Ger- marks and other archival material are docu- 16. C. J. Davisson, L. H. Germer, Proc. Nat. Acad. Sci. 14, 317 (1928); Proc. Nat. Acad. mer retired from Bell Labs in 1961 and mented in the author's PhD dissertation, "C.J. Davisson, L. H. Germer, and the Discovery of Sci. 14, 619 (1928); Phys. Rev. 33, 760 remained active in this "new" field and in (1929). his favorite recreation, mountain climb- Electron Diffraction," University of Minne- sota, 1973, available from Xerox University 17. A. U. MacRae. Science 139, 379 (1963). ing, until his death in 1971. Microfilms, 3000 North Zeeb Road, Ann Arbor, 18. W. Ehrenberg, Philosoph. Mag. 18, 878 In trying to answer the question of Michigan 48106, Order No. 74-10 505. (1934). D

PHYSICS TODAY / JANUARY 1978 41